EXPERIMENTAL ENGINEERING 
 
EXPERIMENTAL ENGINEERING 
 
 BY 
 
 U. T. HOLMES 
 
 Commander, U. S. Navy 
 
 
 ANNAPOLIS, MD. 
 THE UNITED STATES NAVAL INSTITUTE 
 
 1911 
 
COPYRIGHT, 1911, BY 
 
 W. B. WELLS 
 Sec'y and Treas. U. S. Naval Institute 
 
 J2or& (gafttmorc 
 
 BALTIMORE, MD., U. 8. A. 
 
EKEATA SHEET. 
 " EXPERIMENTAL ENGINEERING." HOLMES. 
 
 p. 12, llth line from foot of page, for " indeterminute " read 
 " indeterminate." 
 
 p. 17, 9th line, for " ob " read % db" 
 
 p. 22. Example in middle of page, first quantity in divisor under 
 solution should read 
 
 p. 31, 3d line, under " Special Slide Kules," strike out second 
 "for." 
 
 p. 34, 15th line, for "P" read " F." 
 
 p. 35, 5th line from bottom, for "grove" read "groove." 
 
 p. 57, Last line should read, " Suction in chamber D will be prac- 
 tically twice that in chamber C" 
 
 p. 58, 3d line, " the specific volume will be lower on passing 
 through B." 
 
 T T 
 
 p. 105, Equation 7 should read, <f> 8 = a log e + -= . 
 
 p. 157, Last line, for "adopted" read "adapted." 
 
 p. 161, Fig. 91. Drop a perpendicular from the center of the 
 shaft and indicate the horizontal distance between this 
 perpendicular and one at G as a. 
 
 p. 171, 4th line, formula should read: 
 
 MXL _2_ jr0_ 
 ~~ EXI ^ d == 180* 
 
 10th line, for "moment of inertia" read "modulus." 
 5th line from bottom, for " evaluting " read " evaluating." 
 
 p. 262, 3d line from bottom, for " gage " read " gauze." 
 
 p. 263, 21st line, for " oil " read " water." 
 
 p. 269, 3d paragraph should read: The analysis of the coal, 
 referred to combustible, i. e., coal less ash and water, 
 would be, in per cent: C, 89.6 (from 83.5 X 100 -=-93.2) ; 
 H, 5.15 ; 0, 3.43 ; N, 1.29 ; S, .536. 
 
p. 270, line 9, should read: 19.1 x. 896 = 17.1. 
 
 line 19, should read : ^| = .018. 
 .oo5 
 
 For the sum of H 2 0, " 56," read ".538." 
 lines 20 and 21, 19.1 + .54= 19.64. 
 line 21, 17.1 + .54= 17.64. 
 line 22, 17.64 x. 835 = 16.40. 
 line 24, 17.64-1 = 16.64. 
 p. 271, line 1, for " 17.41 read "17.64." 
 line 3, 17.64 x. 246x515 = 2233.5. 
 line 9, for ".56" read ".538." 
 line 11, .538x965.8 = 519.5. 
 
 . line fromfw ,t, . 
 
 p. 272, line 7, for "2206" read "2233.5." 
 line 9, for " 541 " read " 519.5." 
 line 10, for " 341 " read " 347.5." 
 line 12, for "1349" read "1346.5." 
 
 p. 273, 5th line from bottom, insert words " Close K " before 
 sentence beginning " Return leveling bottle," etc. 
 
PREFACE 
 
 In attempting to revise the volume of Notes on Experimental 
 Engineering compiled by the author in 1907, so much new matter 
 was at hand and so many changes were found necessary that it was 
 deemed advisable to rewrite the whole book. In doing this con- 
 siderable matter belonging properly to the subject of experimental 
 engineering has been brought in from other text-books. 
 
 The following authorities have been consulted : Carpenter's 
 Experimental Engineering; Experimental Engineering, by Pullen 
 and Popplewell; Physical Laboratory Notes, by Silas W. Holman; 
 Barton's Internal Combustion Engines ; Bieg's Text-Book on Naval 
 Boilers; Departmental Notes; Smart's Laboratory Practice; Prac- 
 tical Cement Testing, by Taylor; Journal of the American Society 
 of Mechanical Engineers ; Journal of the American Society of Naval 
 Engineers; Power and the Engineer; various manufacturers' 
 pamphlets. The notes on the temperature-entropy diagram have, 
 for the most part, been taken from the article on that subject by 
 Lieutenant-Commander L. M. Nulton, U. S. N., published in the 
 Journal American Society of Naval Engineers, Vol. VIII. The 
 matter relating to time-firing devices has been supplied by Com- 
 mander S. S. Eobison, U. S. N., of the Bureau of Steam En- 
 gineering. 
 
 My thanks are due to Mr. F. H. Rittenour, draftsman in the De- 
 partment of Marine Engineering and Naval Construction, for much 
 valuable assistance in the preparation of the illustrations and to Mr. 
 J. M. Armstrong for assistance in photographing apparatus. Also 
 to the several manufacturers of engineering apparatus for their 
 courtesy in supplying cuts of their laboratory appliances. 
 
 U. T. HOLMES, 
 
 Commander, U. S. N. 
 BUREAU OF STEAM ENGINEERING, NAVY DEPARTMENT, May, 1911. 
 
 242289 
 
CONTENTS 
 
 PAGE 
 
 CHAPTER I. 
 
 Introduction: Engineering calculations Limits of accuracy 
 Measurements Limits of error Significant figures Reports. . 9 
 
 CHAPTER II. 
 
 Instruments for Computing Experimental Data: The logarithmic 
 scale Different forms of the logarithmic scale The slide rule 
 Planimeters 17 
 
 CHAPTER III. 
 
 Instruments for Recording Experimental Data: The measuring 
 machine Pressure and vacuum gages Recording gages Gage 
 testing apparatus Thermometers and pyrometers Revolution 
 counters Tachometers Speed regulators 40 
 
 CHAPTER IV. 
 
 Measurement of the Quality of Steam: Definitions in thermody- 
 namics Steam calorimeters and their use The temperature- 
 entropy diagram and its uses 85 
 
 CHAPTER V. 
 
 Measurement of the Rate of Flow of Water: Water meters of 
 various forms Weirs 110 
 
 CHAPTER VI. 
 
 Measurement of the Rate of Flow of Air and Steam: The anemom- 
 eter Pitot's tube for low air pressures Steam flow meters 
 Air flow meters 129 
 
 CHAPTER VII. 
 
 Measurement of Power: Definitions The steam engine indicator, 
 errors to which it is subject and methods of calibration Other 
 means of measuring power Dynamometers The torsionmeter 
 Calculation of shaft horse power Various forms of the tor- 
 sionmeter 144 
 
 CHAPTER VIII. 
 
 Testing Materials of Construction: Definitions Testing ma- 
 chines and their operation Cements Classification of and 
 methods of testing 187 
 
 CHAPTER IX. 
 
 Engine Lubrication: Systems of lubrication Engine lubricants- 
 Testing lubricants, determinations required and apparatus used 
 Oil testing machine Tests on board ship 212 
 

 * 
 
 " CONTENTS 
 
 CHAPTER X. 
 
 The Selection and Testing of Fuel: Testing coal Fuel calorimeters 
 and their operation Testing fuel oil Practical tests with the 
 Navy Standard outfit Specifications for fuel oils 229 
 
 CHAPTER XL 
 
 Flue Gas Analysis: The Orsatt-Muencke and Hays apparatus 
 Automatic CO 2 recorders. Time Firing Devices: Various makes 
 used on naval vessels . 266 
 
EXPERIMENTAL ENGINEERING 
 
 CHAPTER I. 
 INTRODUCTION. 
 
 Experimental Engineering is that branch of the engineering pro- 
 fession which treats of the instruments and methods employed in the 
 collection of engineering data and includes the testing of all 
 materials and appliances employed in engineering work. Mechan- 
 isms are tested to determine the amount of work done on them, or 
 by them, and their efficiency. Materials are tested for the deter- 
 mination of their strength, or suitability for the purpose required. 
 
 The study of experimental engineering gives the student famil- 
 iarity with the various instruments employed in experimental work, 
 together with the methods of calibrating and of using them. 
 
 Materials under test are seldom so homogeneous that samples, 
 taken even from the same piece, show exactly the same strength. In 
 efficiency tests the personal error of the observer is usually such as 
 to cause slight variations in the results from different observations. 
 The results obtained in such work can therefore have no fixed and 
 certain values, following known laws, but will approximate more 
 or less closely to the average value which it is desired to obtain. 
 In order to approximate as closely as possible to the result, it is 
 desirable, where possible, to obtain the mean result from a series 
 of experiments. Where it is possible to secure only single observa- 
 tions, it is necessary to take more than ordinary precautions in 
 eliminating all possible sources of error. 
 
 ENGINEERING CALCULATIONS. 
 
 Accuracy with some persons is instinctive they practice it in 
 every thought and action ; with others it must be cultivated through 
 severe mental discipline, and they must be continually on their 
 guard against carelessness. No one will deny that accuracy is most 
 
10 EXPERIMENTAL ENGINEERING 
 
 desirable, yet there is always the chance of the exceedingly accurate 
 person becoming absorbed in details without giving due attention 
 to the main points. This is especially true with engineering cal- 
 culations, where it is almost as bad to use unnecessary refinements 
 as it is to carelessly use a number of crude approximations. From 
 this it should not be inferred that precision may be neglected in 
 purely mathematical operations. Forethought and discrimination 
 are required in determining the necessary degree of accuracy. This 
 involves an understanding of the relation which the calculation 
 bears to the problem under consideration, in order that the desired 
 result may be obtained without unnecessary labor. 
 
 Many problems involve assumptions or factors that cannot be ac- 
 curately measured and it would be absurd to carry out the cal- 
 culations to the third or fourth decimal place. For instance, in 
 computing the load that a certain member would carry, the assump- 
 tion is made that the point of rupture of that particular material 
 is 50,000 pounds per square inch. Now, this material may fail at 
 45,000 pounds per square inch or it may withstand 55,000 pounds 
 per square inch; hence, the futility of carrying the computations 
 to a high degree of refinement. 
 
 Again, in calculating the indicated horse power of a steam engine 
 from an indicator diagram, if a slide rule be used the result will be 
 about as near the true horse power as would be the case if the prob- 
 lem were multiplied out in detail. The error in finding the area 
 of the indicator diagram and the fact that the steam pressure in 
 the indicator may not exactly represent that in the cylinder, make 
 the result, at the best, only approximate. 
 
 On the other hand, there are many engineering calculations which 
 require a high degree of accuracy. The detail parts of a compli- 
 cated piece of mechanism, for example, must be worked out with 
 great precision in order that the whole will assemble correctly when 
 finished. 
 
 The advantage of being able to size up a problem at a glance with 
 the exercise of proper judgment as to the degree of accuracy re- 
 quired, increases with experience. In the computing departments 
 of large establishments great saving in time and labor may be 
 effected by the intelligent use of approximations. 
 
ENGINEERING CALCULATIONS 11 
 
 Measurements. Direct measurements are those that can be read 
 directly such as a measure of length, of time or of absolute quantity. 
 Indirect measurements are those which must be calculated from 
 data, the different parts of which may in turn be measured directly 
 or indirectly. Thus the horse power of an engine is measured in- 
 directly. The different factors entering into it are area of cylinder 
 and length of stroke, measured directly, mean effective pressure, 
 measured indirectly, and revolutions per minute which may be 
 measured directly by counting or indirectly from a recording device, 
 which gives the revolutions during an interval of time that must 
 be read elsewhere. 
 
 Limits of Error in Observations. The purpose for which a given 
 quantity is measured must determine the degree of accuracy which 
 we must endeavor to reach. This can usually be stated definitely 
 as a numerical quantity, or limits can be assigned; first an upper 
 limit below which the error must be kept and a second a lower 
 limit below which the error will do no injury. The extreme care 
 necessary to reduce the error below the lower limit will be useless 
 and the labor necessary for great refinements of accuracy is usually 
 out of proportion to the results obtained, so that the lower limit is 
 often of much importance. 
 
 If the measurement is direct it becomes only necessary to select 
 suitable apparatus and so operate it as to produce a result within 
 the assigned limits of accuracy.- If it is indirect, we should deter- 
 mine the degree of accuracy which is necessary in each of the various 
 component measurements from which it is to be computed. 
 
 Deviations from the True Measurement. Let a^ a 2 , . . . ., a n be 
 the various single measurements of a quantity in a series, each being 
 made with the same care and under the same conditions. These 
 values will not usually be the same, but will differ according to 
 the quality of the instrument used, the skill of the observer and 
 the conditions under which he is operating. In the case of a series 
 of rough measurements, say of a short distance to the nearest foot, 
 the various measurements may be all alike, but even then they might 
 differ by a foot if the correct measurement were approximately half 
 a foot over or under. If the attempt be made to measure the dis- 
 tance to within, say one per cent, the separate results will diverge. 
 
12 EXPERIMENTAL ENGINEERING 
 
 Suppose the different measurements to have been made with care, 
 and that a^ a 2 , etc., show sensible divergence and R is the true 
 value. Then a^ R e^ is the first error, a 2 R = e 2 is the second, 
 etc. But R is not known, else the measurements would not be re- 
 quired, hence the errors are also not known. It is therefore neces- 
 sary to establish a rule for selecting a result that will best represent 
 the correct measurement, and for a series such as that described, 
 experience has demonstrated that the arithmetical mean best ful- 
 fills this requirement. The most probable value is therefore in 
 such a case 
 
 A = i 2 n ? where n is the number of single ob- 
 
 servations. 
 
 If the single observations are not all equally reliable, so far as 
 known, then before they are combined, some numerical measure of 
 their relative reliability or weight must be taken. Let p ly p 2 , . . . . 
 p n be such numbers, then the most probable value will be the 
 weighted mean 
 
 4 - PI&I+ j? 2 ga+ ---- +Pnttn 
 Pl+P 2 + ---- +Pn 
 
 The variation of the single observations will be shown by their 
 deviation from the mean, viz.: a l A = d l ,a 2 A = d 2 , . . . . On A 
 = d n . A is, however, the mean, not the true result, hence these 
 deviations d ly etc., are not equal to e, etc. That is the deviations 
 are not the errors and must not be confounded with them. 
 
 Sources of Error are of two classes, determinate and indeter- 
 minute. Determinate errors are those inherent in the apparatus, 
 such as incorrect scales, etc., and can be corrected either absolutely 
 or approximately. It is for the elimination of errors of this char- 
 acter that apparatus must be carefully calibrated. After all such 
 corrections that are obtainable have been applied, there may still 
 remain a small error in the apparatus. This is designated as the 
 residual error. 
 
 In addition to the residual error there are many sources of error 
 affecting all observations, which cannot be discovered and allowed 
 for. These are all classed together as indeterminate, and are the 
 cause of the deviation of the mean from the true result. In a large 
 
ENGINEERING CALCULATIONS 13 
 
 * 
 
 series of observations made under various conditions the effects of 
 the indeterminate sources of error are likely to eliminate each other 
 to a large extent since the errors of equal magnitude are about 
 equally likely to be -f- or . But in any series of observations by 
 a single method there is a decided liability to the preponderance 
 of certain indeterminate sources of error, leading to a mean result 
 that may differ considerably from that obtained by a different 
 method, while in either series the single results may differ very 
 slightly from each other. There is thus, in any series of observa- 
 tions, a liability to a constant error. This can only be eliminated 
 by having the observations made by as many different methods, in- 
 struments and observers as is possible and average the different re- 
 sults, giving each its proper weight as nearly as can be determined. 
 
 Mistakes. Rejection of Observations. 
 
 Mistakes must not be classed with errors of observation. They 
 arise from mental confusion which leads to setting down a wrong 
 number, reading a scale division incorrectly, selecting a wrong 
 reference line, etc. Where one observation in a series differs widely 
 from the others it should be checked over very carefully to eliminate 
 mistakes. Great care and judgment must be exercised, after elimi- 
 nating all mistakes, in scrutinizing such observations, in order to 
 determine whether they should be used with the others or discarded. 
 Some authors offer mathematical rules for guidance in such cases, 
 but the best guide is the application of scrupulously honest un- 
 biassed judgment, in other words, common sense. Where possible 
 this should be given by some competent person other than the 
 observer. 
 
 Significant Figures. The term significant figures, in any ex- 
 pression, refers to the number of digits represented between and 
 including the first and the last, without reference to the position of 
 the decimal point. Thus the numbers 1234000., 12.34, 0.0001234, 
 5067, 5706, etc., each have four places of significant figures. In 
 computations involving the use of observed data it is important to 
 follow the rules for the retention or rejection of places of signifi- 
 cant figures, in order, on the one hand, to give the result all the 
 accuracy to which it is entitled from the data, and on the other 
 hand to avoid encumbering the work with a useless mass of figures. 
 
14 EXPERIMENTAL ENGINEERING 
 
 Let M be an original observed quantity and 8 its average devia- 
 tion. Let that place in M which corresponds to the second place 
 of significant figures in 8 be called the r th place. M is uncertain 
 by an amount 8. Suppose 8 = 0.034, then M is uncertain by 34 
 units in the r th place, by 3 units in the (r 1) place and is correct 
 in the (r 2) place. In general the first significant figure in the 8 
 may be anything from 1 to 9; hence M is uncertain by 1 to 9 
 units in the (rl) place and by 10 to 99 units in the r place. The 
 digit in the (rl) place in M is always uncertain and that in the 
 r place is so very uncertain as to make its retention useless. Two 
 places of significant figures in the 8 are all that are of service since 
 its only use is as an indication of the correctness of M. These 
 considerations govern the retention of significant places in the data. 
 Places may be rejected in the result whose value depends on the 
 doubtful places in the data. 
 
 Rules for Significant Figures. 
 
 (1) In averaging a series of observations, results should be carried 
 to two places only beyond the point where they become doubtful. 
 Where the digit in the first place is large the second digit is of 
 little or no value. 
 
 (2) In a single observation, the last place retained should be the 
 second beyond which the reading can be made positively, or if the 
 digit in the first place is large the second is unnecessary. 
 
 (3) In addition or subtraction, carry the sum or difference only 
 to the second significant figure of the least precise quantity in the 
 data. 
 
 (4) In multiplication or division, carry out the product or quo- 
 tient at any stage of the work; only to the number of places corre- 
 sponding to those retained in the least precise factor. 
 
 (5) In the use of logarithms, the mantissae should be retained to 
 as many places as there should be significant figures retained ac- 
 cording to rule 4. The characteristic is not to be considered as it 
 merely serves to locate the decimal point. 
 
 Take for example the formula for horse power in a simple engine, 
 viz. : 
 
 I. H. P. = const, x mean effective pressure x revs, per minute. 
 
ENGINEERING CALCULATIONS 15 
 
 The mean effective pressure will -usually be measured by an 
 averaging instrument, giving the mean height of card correct to 
 1/10 inch, with a fairly close approximation to the hundredths 
 place. Multiplying by the scale of the spring, the mean effective 
 pressure is obtained in which the first place is correct, 'the second 
 is probably correct, the third is doubtful and the fourth a very 
 doubtful approximation. The revolutions per minute is obtained 
 by counting while observing the revolution of the second hand of 
 a clock or watch. This gives two places accurate after which they 
 become doubtful. 
 
 Following rules 1 and 2, we use three or four places only in ex- 
 pressing the number of revolutions and the pressure. The constant 
 factor need only be carried out to the fourth place corresponding 
 with the other factors. From rule 4 it will be seen that it is use- 
 less to carry out the I. H. P. further than to four places. 
 
 Another example is that of a ship making standardization trials 
 over the measured mile. Several observers agree on the nearest 
 second, but differ on the tenths. Averaging the observations the 
 rule provides for carrying out the average to the hundredths place. 
 The number of revolutions on the mile is given by several counters 
 to the nearest whole number. The mean, applying the rule, is 
 carried to tenths. From these data the speed per hour and the 
 corresponding revolutions per minute while on the mile are cal- 
 culated, the speed going to hundredth knots and the revolutions 
 per minute to tenths. Having thus determined the number of rev- 
 olutions per minute for different speeds, we may construct a curve 
 which will give us the relation between revolutions per minute and 
 speeds per hour at all practicable speeds. To obtain the mean 
 speed during any particular interval of time, it then becomes only 
 necessary to observe the total number of revolutions, deduce the 
 revolutions per minute and apply to the curve. 
 
 It is obvious that to obtain the revolutions per minute with a 
 greater degree of accuracy than that obtained on the measured mile 
 will be an unnecessary refinement, and will entail useless pains on 
 the part of the observers. 
 
 Computations. The labor involved in working out the results is 
 often greater than that of taking the observations. In order to 
 
16 EXPERIMENTAL ENGINEERING 
 
 reduce this labor as much as possible the detail of the method of 
 computation should be gone over and simplified as much as possible, 
 rejecting useless places of figures and providing for frequent and 
 systematic checking. In the actual work of computation the slide 
 rule, as hereinafter described, will be found most useful for ap- 
 proximate results, while Fuller's and Thacher's instruments will be 
 found accurate for computations involving four places of signifi- 
 cant figures. The same degree of accuracy is obtained with four 
 place logarithms and these will be found sufficiently accurate for 
 most work. The usual five place tables may be used for practically 
 any work. 
 
 It is essential that all instruments used in engineering observa- 
 tions should be carefully calibrated. 
 
 Reports. In making reports on a test, the following should be 
 included : 
 
 ( 1 ) A summary of the orders under which the experiments or test 
 are made. The orders themselves or copies should be appended to 
 the report. 
 
 (2) A description of the plant and apparatus. 
 
 (3) The methods employed in carrying on the test. 
 
 (4) The observed and calculated data, tabulated as far as possible. 
 If blank forms are available these should be employed in making 
 out data sheets. 
 
 (5) Conclusions and recommendations. 
 
 Any illustrations or drawings to accompany the report should be 
 appended. Reports should be as concisely worded as possible, con- 
 sistent with clearness. 
 
CHAPTER II. 
 INSTRUMENTS FOR COMPUTING EXPERIMENTAL DATA. 
 
 The Logarithmic Scale. Take the logarithms of numbers from 
 1 to 10 and consider them as representing dimensions on a linear 
 scale as shown in Fig. 1. Let a and b be two pointers. Suppose a 
 
 i i i i i i i i i i 
 
 O-/oy/ toy 2 /oo3 *y4 /eyS /ogff /<y7 /oyd /oy9 /cyM 
 
 FIG. 1. 
 
 to be fixed in position and I capable of movement to the right or 
 left. Suppose also the scale to be capable of movement along its 
 length. 
 
 If we bring on the scale opposite a, and place b opposite log 3, 
 the distance ab will be equal to log 3. If then, without changing 
 the position of I, we move the scale to the left the distance 0& will 
 still be log 3. Suppose the scale to be moved so as to bring log 2 
 opposite a. Then Ob = Oa-\-ab = log 2+log 3 = log 6. It is obvious 
 that the mark log 6 must fall opposite b. If the scale be moved 
 further to the left, bringing log 3 opposite a, Ob will equal log 3 
 + log 3 = log 9. The mark log 9 will fall opposite b. If the scale 
 be moved far enough to the left b will be off the scale. This is 
 provided for by using a second fixed pointer a! 9 at a distance from 
 a=aa' = log 10. Suppose log 5 placed opposite a, then log 10 is to 
 the left of b. Shift the scale, bringing log 5 opposite a' (Fig. 2). 
 
 a b a' 
 
 \ \ \ 
 
 I . I . , . I 
 
 f I I I I I I \ I I 
 
 O-Ay/ /og3 /cy3 /oy4 /<yf fofff /cy7 /eyff/eyS/by/G 
 
 FIG. 2. 
 
 We have aa' = log 10 and &a' = log 10 log 3. But log 5 is opposite 
 a' and &a' = log 5 0&. 
 
 .'. log 10 Iog3 = log 5 Ob and 
 Ob = log 5 + log 3 log 10 = log 1.5. 
 
18 EXPERIMENTAL ENGINEERING 
 
 From this it appears that for problems in simple multiplication, 
 dropping the subscript log, if we set the scale to zero, i. e., with a 
 falling at 0, and a' at 10, bring the sliding pointer & opposite one 
 of the factors, then set the scale to bring the second factor opposite 
 either a or a', the result of multiplying together the two factors is 
 read at &. 
 
 For division, suppose that the dividend be placed opposite a' and 
 the sliding pointer moved opposite the divisor on the scale. Then 
 Oa'=log (dividend) and 0&=log (divisor). ?>a'=log (dividend) 
 log (divisor) = log (quotient). Bringing on scale, opposite & 
 the quotient is read off opposite a. 
 
 From the foregoing it should be clear that a logarithmic scale is 
 one on which the numbers set down are the linear measurements of 
 the logarithms of the numbers so marked. It affords a means of 
 mechanically adding or subtracting the logarithms of numbers and 
 permits the number corresponding to the resulting logarithm to be 
 read off directly. 
 
 By making the scale long enough and subdividing the divisions, 
 it may be made to represent the logarithms of numbers from to 
 1000, instead of to 10. 
 
 Sexton's Omnimetre. Construction. Inspection of the instru- 
 ment shows that it contains a number of circles, each circle, or set 
 of circles, containing a logarithmic scale of some function. The 
 circles are named and differently colored so as to avoid danger of 
 confusing them. With their several uses, they may be briefly de- 
 scribed as follows : 
 
 Outer Circle. Logs. This circle gives a direct reading of the 
 logarithms of numbers shown on the A circle. For any number 
 given on A, we read the decimal part of its logarithm opposite on 
 Logs, as though read from a table of logarithms. With the instru- 
 ment set to zero, i. e., with LB opposite 1A, or with the A and B 
 scales coincident, this circle also gives the logarithm of any function 
 shown on the various other circles. By the term " opposite," as 
 used herein, we mean on the same radial line, utilizing for accuracy 
 the straight line, or edge of the runner, as a guide. 
 
 A Circle. The subdivisions are to be read as though they were 
 figures. We may call the starting point 1, 10, 100, 1000, etc. 
 
INSTRUMENTS FOR COMPUTING EXPERIMENTAL DATA 19 
 
 The next subdivision to the right will be 1.005, 10.05, 100.5, 1005, 
 etc.; and the second subdivision to the right, 1.010, 10.10 101.0, 
 1010, etc., respectively, the value of each subdivision varying by 5 
 until we come to 2 (or, 20, 200, etc.) ; then we begin to read by 1 
 in the third place for each subdivision, 2.01, 20.1, 201, etc., until 
 later, when the value of the subdivision varies by 2 in the third 
 place, so that the first subdivision to the left of the starting point 
 is read as 9.98, 99.8, 998, etc. 
 
 FIG. 3. Sexton's Omnimetre. 
 
 B Circle. The A and B circles are exactly alike in subdivisions 
 and figures. With the . instrument set to zero, remarks under A 
 apply to B, as like numbers on the two circles coincide. We may 
 regard scale B of the instrument as the single logarithmic scale 
 described in the foregoing articles. The point 1A, or zero point, 
 becomes both a and a' of the description, while the swinging trans- 
 parent pointer with hair line becomes the movable pointer, &. Scales 
 A and B constitute the scale and slide respectively of an ordinary 
 double-scale slide rule, and may be used for ordinary calculations 
 according to the explanation given on page 28. In the following 
 detailed explanation of the method of using the omnimetre, in 
 order to avoid confusing the student, all operations are made to 
 
20 EXPERIMENTAL ENGINEERING 
 
 begin and end with the instrument set to zero. All the work is 
 done with the B scale only, the method of operation being that 
 applicable to a single-scale slide rule. Short-cuts will after a time 
 suggest themselves to the student, but these are not recommended 
 until after complete familiarity with the instrument is established. 
 
 Squares or N 2 . The value of the square of any number found on 
 Squares is opposite on B. If the number is on the inner Squares 
 circle, its square has 1, 3, 5, etc., figures; if on the outer circle of 
 Squares, 2, 4, 6, etc.,, figures. If the square root of any number is 
 desired, first find the number on B then look opposite on Squares 
 for the result, knowing from the foregoing which Squares circle 
 will contain it, and the number of figures in the result. 
 
 Cubes or N" 3 . The value of the cube of any number found on 
 Cubes is opposite on B. If the number is on the inner Cubes circle, 
 its cube has 1, 4, 7, etc., figures ; if on the second Cubes circle, 2, 5, 
 8, etc., figures; and if on the third or outer Cubes circle, 3, 6, 9, etc., 
 figures. If the cube root is desired, reverse the operation. Take the 
 number from B and, having in mind the foregoing explanation, the 
 proper Cubes circle in which to find the result and the number of 
 figures it should contain, will be determined. 
 
 Fifth Powers or N 5 . The value of the fifth power of any num- 
 ber found on an N 5 circle is opposite on B. For numbers on the 
 inner circle of N 5 , the fifth power has 1, 6, 11, etc., figures; on the 
 second circle, 2, 7, 12, etc., figures; on the third circle, 3, 8, 13, etc., 
 figures ; on the fourth circle, 4, 9, 14, etc., figures ; and on the fifth 
 or outer circle, 5, 10, 15, etc., figures. If the fifth root is desired, 
 the reverse operation is followed as explained before. 
 
 Trigonometric Functions. The numerical (natural) value of the 
 secant, sine, tangent, or versed sine of any angle is found on B 
 opposite the given angle on the circle of appropriate name. Setting 
 the instrument to zero, the corresponding logarithmic function is 
 read on the outer circle. The following explanation gives the 
 method of finding the decimal point in various cases. 
 
 Secants. There is no difficulty in locating the decimal point, as 
 the secant of any angle is greater than unity, and within the limits 
 of the instrument lies between 1 and 10. 
 
 Sines. If the angle is between 35' and 5 45', use the inner 
 
INSTRUMENTS FOR COMPUTING EXPERIMENTAL DATA 21 
 
 Sin. circle and read 0.01 to 0.10 on B. Between 5 45' and 86, 
 use the outer circle and read 0.10 to 1.00. 
 
 Tangents. If the angle is between 35' and 5 43', use the 
 inner Tang, circle and read 0.01 to 0.10 on B. If between 5 43' 
 and 45, use the middle circle and read 0.10 to 1.00 on B. If 
 between 45 and 84 15', use the outer circle and read from 1.00 
 to 10.00 on B. 
 
 Versed Sines. If the angle is between 2 35' and 8 05', use 
 the inner circle and read 0.001 to 0.01 on B. If between 8 .05' and 
 25 50', use the middle circle and read 0.01 to 0.1. If between 
 25 50' and 90, use the outer circle and read 0.1 to 1.0. 
 
 Multiplication and Division. In the use of the omnimetre as a 
 calculating instrument for formulae involving multiplication and 
 division of any of the functions given on the various slide circles, 
 a few experiments w r ill emphasize the following useful points : 
 
 (a) The formula should always be expressed in the form- of a 
 fraction in which : 
 
 /TX No. of factors in numerator 11-11 T .1 
 
 (b) T? - TT~~ j -- " should always be in the ratio 
 No. oi factors in denominator 
 
 of *1 . 
 
 X 
 
 Example 1 : Eequired the value of ax ^ xc . This should be 
 
 a x b x c 
 written -= - , unity being in all cases substituted for any of 
 
 the missing factors. 
 Example 2: 
 
 of expression is 
 
 Example 2: Eequired the value of - ^ - . The proper form 
 
 u x a x c 
 
 bxdxc 
 Example 3 : Eequired the value of x from the equation : 
 
 x _ a*xsmbxcxd s xe 
 
 t&nfxg 2 
 a* X sin b 
 
 Write the equation as *= 
 
 tan/x# 2 Xlxl 
 
 a, b, c, etc., may, of course, be given any value that can be found 
 on any circle of the slide.) 
 
22 EXPERIMENTAL ENGINEERING 
 
 In problems involving the use of the cosine, cosecant, or cotan- 
 gent, these functions should be written in the form : 
 
 , respectively. 
 
 secant sine tangent ' 
 
 (c) Having the problem expressed in proper form, as above, the 
 first move is to set the instrument to zero, i. e., IA on 15, so that 
 like numbers on scale and slide coincide. 
 
 (d) For all factors in the numerator move the runner over the 
 slide, keeping the scale and slide stationary. 
 
 (e) For all factors in the denominator move the slide under the 
 runner until the required number on the slide comes under the 
 radial line of the runner, keeping the scale and runner stationary. 
 
 (f ) Alternate the movements of runner and slide. 
 
 (g) The last move is to set the slide to zero (IA on 15), keeping 
 runner and scale stationary, and along the runner in the proper 
 circle will be found the required answer. 
 
 Example : Required the value of x from the equation : 
 
 0.296 X 73 X 0.00115 x 97.6 X 33.2 
 .012x671.5 
 
 Solution: 
 
 m _ 0.296 X 73 X 0.00115 X 97.6 x 33.2 
 0.12X671.5X1X1 
 
 (2) Set the instrument to zero. 
 
 (3) Move runner to 296 on slide B. 
 
 (4) Bring 12 on slide under runner. 
 
 (5) Move runner to 73 on slide B. 
 
 (6) Bring 6715 on slide under runner. 
 
 (7) Move runner to 115 on slide B. 
 
 (8) Bring unity on slide under runner. 
 
 (9) Move runner to 976 on slide B. 
 
 (10) Bring unity on slide under runner. 
 
 (11) Move runner to 332 on slide B. 
 
 (12) Set the instrument to zero. 
 
 (13) Eead numerical value of x on B (or A), as indicated by 
 runner. 
 
INSTRUMENTS FOR COMPUTING EXPERIMENTAL DATA 23 
 
 (14) Find position of decimal point by rough calculation. 
 
 Var. 1: If, instead of x, we have x 2 , x 5 , or x 5 , then, after (12), 
 read value of x from Squares, Cubes or N 5 , as the case may be. 
 
 Var. 2: If, instead of x, we have Vz, 8 Va?, 5 Vx, then, after 
 (12), select from B the number indicated by the runner, and with 
 this number enter Squares, Cubes, or N 5 , and the value of x will he 
 found opposite on B. 
 
 Var. 3 : If, instead of x, we have sec 0, sin 6, tan 0, or vers 
 sin B, then, after (12), the value of as an angle is read from the 
 circle Sec., Sin., Tang., or V. S., as the case may be. 
 
 As an example involving functions of the omnimetre, suppose 
 we assume 
 
 _ sin 14 x tan 4o x sec (39 20') x vers sin (27 45') x (4.7)3 
 
 (36.757 2 ~X (1.965)3 
 
 sin 14 x tan 43 X sec (39 20') x vers sin (27 45') x (4.7) 3 
 (36.75)2 X (l.65px IX 1 
 
 (2) Set the instrument to zero. 
 
 (3) Move runner to 14 on slide Sin. 
 
 (4) Bring 3675 on slide N 2 to runner. 
 
 (5) Move runner to 43 on slide Tang. 
 
 (6) Bring 1965 on slide N s to runner. 
 
 (7) Move runner to 39 20' on slide Sec. 
 
 (8) Bring unity on slide B to runner. 
 
 (9) Move runner to 27 45' on slide V. S. 
 
 (10) Bring unity on slide B to runner. 
 
 (11) Move runner to 47 on slide N 8 . 
 
 (12) Set instrument to zero. 
 
 (13) Eead value of x on B as indicated by runner. 
 
 (14) Make rough calculation for decimal point. 
 
 Special Cases. The method of using the instrument in the solu- 
 tion of a few well-known formulae may be illustrated as follows : 
 
 Case 1 : Should a formula contain a simple factor multiplied by 
 a radical, it is much easier of solution to introduce the simple factor 
 under the radical, and extract the root as a last operation. 
 
 Example: Find the diameter of a shaft to transmit 4850 I. H. P. 
 at 165 revolutions. 
 
24 EXPERIMENTAL ENGINEERING 
 
 The formula is : 
 
 where K = constant = 4. 3. 
 Solution: 
 
 (1) d=4.3^^= 7(4^X4850 
 V 1 V 
 
 165 165 
 
 (2) Set the instrument to zero. 
 
 (3) Move runner to 43 on slide N 3 . 
 
 (4) Bring 165 on slide B to runner. 
 
 (5) Move runner to 485 on slide B. 
 
 (6) Set instrument to zero. 
 
 (7) Eead value of d on N s as indicated by runner. 
 
 (8) Make rough calculation for decimal point. 
 
 Example 2: Find the thickness of a flat cylinder cover for a 
 pressure of 165 pounds. Tensile strength allowed 10000. Diame- 
 ter of cylinder 11.5 inches. 
 
 The formula is : 
 
 Solution: 
 
 -V 
 
 (5.75) 2 XX165 
 
 3X10000 3X10000 
 
 (2) Set the instrument to zero. 
 
 (3) Move runner to 575 on slide N 2 . 
 
 (4) Bring 3 on slide B to runner. 
 
 (5) Move runner to 2 on slide B. 
 
 (6) Bring 1 on slide B to runner. 
 
 (7) Move runner to 165 on slide B. 
 
 (8) Set instrument to zero. 
 
 (9) Read value of t from slide N 2 . 
 
 (10) Make rough calculation for decimal point. 
 
 Case 2: To find the square root of the sum of two squares 
 
INSTRUMENTS FOR COMPUTING EXPERIMENTAL DATA 25 
 
 Find the angle (6) whose tangent is --and divide a by the sine 
 of this angle. 
 
 The reason for this may be seen by referring to Fig. 4, in which 
 
 and 
 
 t an 0=4 
 
 o 
 
 sin 6 = p= == ; 
 
 vV+6 2 
 
 sin B ' p IG . 4. 
 
 Example : Find the equivalent twisting moment of a solid shaft 
 from the formula T e = M+ ^/M 2 + T 2 . 
 Solution: 
 ( 1 ) Consider the part under the radical and write 
 
 and 
 
 _ 
 
 ~ 
 
 sin B sin B ' 
 
 (2) Set the instrument to zero. 
 
 (3) Move runner to a on slide B. 
 
 (4) Bring b on slide B to runner. 
 
 (5) Move runner to unity on slide B. 
 
 (6) Set instrument to zero. 
 
 (7) Note value of B on slide Tang. 
 
 (8) Move runner to a on slide B. 
 
 ( 9 ) Bring 6 on slide Sin. to runner. 
 
 (10) Move runner to unity on slide B. 
 
 (11) Set instrument to zero and read value of x on slide B, plac- 
 ing decimal point. 
 
 (12) The value of T e = x + M. 
 
 Case 3: To find the square root of the difference of two squares 
 
26 
 
 EXPERIMENTAL ENGINEERING 
 
 Find the angle (0), whose sine is , and divide a by the secant 
 
 of this angle. The reason for this may be seen by referring to Fig. 
 5, in which 
 
 sin 0, - 
 
 b and 
 
 sec 9 
 
 sec0' 
 
 Example: Find the load that a screw thread will safely sustain, 
 having given the formula : 
 
 W=Kxnx~-X(d i 2 -d 2 2 ). 
 Solution: (I) Consider the part in parenthesis and write 
 
 sin 0= -~ = 
 a 
 
 , and x = 
 
 sec 
 
 oxl 
 
 sec0* 
 
 (2) Set the instrument to zero. 
 
 (3) Move runner to b on slide B. 
 
 (4) Bring a on slide B to runner. 
 
 (5) Move runner to unity on slide B. 
 
 (6) Set instrument to zero. 
 
 (7) Note value of 6 on slide Sin. 
 
 (8) Move runner to a on slide B. 
 
 (9) Bring 6 on slide Sec. to runner. 
 
 (10) Move runner to unity on slide B. 
 
 (11) Set instrument to zero and read value of x 2 from N* 9 plac- 
 ing decimal point. 
 
 (12) We now have W=KxnX ~ Xx 2 = 
 
 4 
 
 is in proper form for solution in the usual manner. 
 
INSTRUMENTS FOR COMPUTING EXPERIMENTAL DATA 
 
 Fuller's Calculating Instrument. This in- 
 strument, shown in Fig. 6, is a single logarithmic 
 scale about 83 feet in length, wound spirally on 
 a cylinder B. This cylinder is capable of a 
 movement of rotation, as well as a vertical move- 
 ment on the cylinder A. The fixed pointers a 
 and a' are carried on cylinder A, while the mov- 
 able pointer b is attached to the handle H and 
 is capable of a vertical, as well as a rotary move- 
 ment. The instrument is thus seen to be the 
 same in principle as that described above. The 
 length of scale permits calculations correspond- 
 ing in accuracy to those obtained with the use of 
 four place logarithms. 
 
 Sperry's Pocket Calculator. This is another 
 example of the single logarithmic scale and is 
 made up in the form of a watch as shown in Fig. 
 7. There are two dials on the two opposite faces, 
 as shown. The L dial contains a single logarith- 
 mic scale about 12J inches long, arranged in 
 three spirals, beginning and ending on the same 
 radial line. A mark on the glass face constitutes 
 the fixed pointer, which corresponds to both a 
 and a' of page 17, since and 10 are both found 
 on the same radial line. The dial carrying the 
 scale is movable, as well as a hand corresponding 
 to the movable pointer &. 
 
 FIG. 6. 
 
 L DIAL. 
 
 S DIAL. 
 
 FIG. 7. 
 
EXPERIMENTAL ENGINEERING 
 
 The S dial bears a scale of equal parts, a circular 
 logarithmic scale, and a scale of square roots. 
 
 This is a very convenient instrument for rough 
 calculations, but the scale is not long enough for 
 calculations requiring accuracy. 
 
 The Slide Rule. This instrument, the most 
 common adaptation of the logarithmic scale is 
 illustrated in Fig. 8. There are two scales, simi- 
 larly constructed, arranged to slide one on the 
 other. There is also a sliding pointer, or " runner," 
 used in reading off the results. For convenience 
 in the following description the two scales will be 
 designated as the "scale" and the "slide," re- 
 spectively. 
 
 Operation. For any position of the slide rela- 
 tive to the scale it is obvious that every number on 
 the slide will bear the same ratio to the number 
 opposite it on the scale. For example, bring 7 on 
 the slide opposite 10 on the scale. The ratio 7: 
 10 = I:- 1 /, will be found along the whole length 
 of the rule. The number on scale opposite any 
 other number on slide will be the product of that 
 number multiplied by 10, divided by 7. For ex- 
 ample, opposite 2.1 on slide we read 3. We have 
 
 then divided 10 by 7 and multiplied by 2.1. WX2 ' 1 
 
 = 3. Obviously unity may be substituted for any 
 one of these factors. Therefore to multiply any 
 two numbers together, bring unity on slide opposite 
 one of the numbers on scale* then opposite the 
 other number on slide read the product on scale. 
 In multiplying and dividing a number of factors, 
 by using the runner, the necessity for reading the 
 scale each time the slide is moved may be avoided. 
 For example: Multiply 6x7x3 and divide by 
 
 8x2. 
 
 6x7x3 
 
 X2 
 
 PIG. 8. 
 
 (28) 
 
INSTRUMENTS FOR COMPUTING EXPERIMENTAL DATA 29 
 
 The factors in the numerator show the successive positions which 
 the runner must take and those in the denominator the positions of 
 the slide. (1) Start with runner opposite 6 on scale. (2) Bring 
 8 on slide to runner. (3) Move runner to 7 on slide. (4) Bring 2 
 on slide to runner. (5) Move runner to 3 on slide. The result is 
 then read directly on the scale opposite runner. 
 
 The method of operation is thus seen to be one of alternately 
 dividing and multiplying. Until complete familiarity with the 
 operation of the instrument is established, the factors should be so 
 arranged, substituting unity for any of the factors that may be 
 missing. 
 
 FIG. 9. 
 
 The numbers on the slide rule are to be considered significant 
 figures, and to be used without regard to the decimal point. Thus 
 the number on the rule for 8 is to be used as 8 or 80 or 800, etc., 
 as may be desired, even in the same problem. The position of the 
 decimal point in the result is readily determined by a rough compu- 
 tation. In case the slide projects so much beyond the scale, that the 
 runner cannot be set at the required figure on the slide, bring the 
 runner to 1 on the slide, then move the slide its full length, until the 
 other 1 comes under the runner. Then proceed as before, moving 
 runner to number on slide and reading results on scale. 
 
 The ordinary straight slide rule as shown in Fig. 8 is in two 
 parts, one above and one below, either of which can be used. The 
 upper, being to half the scale of the lower, is not so accurate, but is 
 more convenient for rough computations. On the back of the slide 
 are two other scales, usually of logarithmic trigonometric functions. 
 The slide is reversed when it is desired to use them. 
 3 
 
30 
 
 EXPERIMENTAL ENGINEERING 
 
 Thacher's Calculating Instrument. This instrument, shown in 
 Fig. 9, is adapted to calculations requiring a considerable degree of 
 accuracy. Logarithmic scales, about 50 feet long are cut up into 
 sections of about 18 inches and the various sections for the slide 
 are placed on a cylinder about 4 inches in diameter. For the scale, 
 the sections are placed on a surrounding cage. The marking is 
 finely subdivided and a glass is provided for reading it. The method 
 of operation is identical with that of the slide rule. 
 
 The Duplex Slide Rule. This instrument, shown in Fig. 10, is 
 graduated on both faces, thus practically giving two rules in one. 
 
 FRONT VIEW OF DUPLEX SLIDE RULE. 
 
 BACK VIEW OF THE RULE, SHOWING THE INVERTED SCALES. 
 FIG. 10. 
 
 The front face, shown in the upper view, is an ordinary slide rule 
 in which both the scale and the slide are graduated from left to 
 right. On the rear face, shown in the lower view, the graduations on 
 the scale are the same as on the front face but the slide is graduated 
 from right to left. The scales on both sides have their indexes in 
 alignment so that the runner, which encircles the whole instrument, 
 permits coinciding points on all scales of either face to be read off 
 at once. 
 
 The advantage of the duplex slide rule lies in the reduction of the 
 number of settings required in performing calculations involving 
 several factors, thus saving time and giving increased accuracy. 
 This follows because when setting to quantities on the inverted 
 slide, the reciprocals of these quantities are given opposite on the 
 regular slide, and vice versa, so that the operations of multiplica- 
 
INSTRUMENTS FOR COMPUTING EXPERIMENTAL DATA 31 
 
 tion and division are reversed. What is a dividing operation with 
 the regular slide becomes a multiplication process with the inverted 
 one. Thus in compound multiplication the solution may be effected 
 by a combination of multiplication and division operations, by using 
 the regular inverted slides, with a consequent reduction in the 
 number of settings. A similar method may be followed with a 
 division operation when there are a number of factors in the de- 
 nominator, as the multiplication by a reciprocal performs a division 
 process. 
 
 Suppose it be required to solve a problem in the form a X b X c. 
 With this form of instrument, using the designations for the scales 
 as shown on Fig. 10, set & on C, to a on D and under c on C read 
 the result. 
 
 If it be required to solve , .. , set & on C to a on D and under c 
 
 u x c 
 
 on C, read the result. 
 
 These operations, which are performed at a single setting of the 
 duplex slide rule, require two settings with the ordinary slide rule 
 and three with the single logarithmic scale. 
 
 Complete familiarity with this instrument can only be acquired 
 from practice, when it will be found a great time saver. 
 
 Special Slide Rules. Many calculating instruments for special 
 purposes, have been made, involving the principles of the logarithmic 
 scale and slide rule. Two of these for for calculating horse power 
 will be found in another chapter. 
 
 PLANIMETERS. 
 
 The Amsler Polar Planimeter. This is the original planimeter 
 or mechanical integrator of Dr. Amsler and is shown in Fig. 11. 
 A radius rod OP contains a needle point P which acts as the center 
 around which it turns. This point is loaded with a weight to main- 
 tain it in position while in use. It is preferable to use the instru- 
 ment on a drawing board which has been previously covered with a 
 smooth, hard paper. Eough paper should not be used. The other 
 end of the radius rod is pivoted to a saddle which slides along the 
 square rod shown in a horizontal position, and which carries the 
 tracing point F. This saddle also carries the recording wheel D 
 
32 EXPERIMENTAL ENGINEERING 
 
 with its spindle and worm, together with the recording disc G and 
 a worm wheel for actuating it. The saddle, which is secured to the 
 tracing rod by means of a clamping screw S, contains the fine ad- 
 justment screw, shown at the left, by which it can be accurately 
 placed in position on the tracing bar. The several marks on the 
 tracing bar indicate the units in which the area to be traced may be 
 recorded by the instrument, with the multiplier that is to be used 
 in each case. 
 
 To use the instrument the saddle is placed on the tracing bar 
 and carefully brought to the mark corresponding to the unit in 
 which it is desired to find the area. The fixed center P is then 
 
 FIG. 11. 
 
 placed somewhere outside the area that is to be traced. The tracing 
 point F is then brought to a position on the outline of the figure 
 and a point pricked there. The graduated disc D f wheel G, and 
 vernier E are read and recorded. Trace the outline of the area, fol- 
 lowing in the direction of the hands of a watch, returning to the 
 point previously marked. Again take the reading and substract 
 from it the previous reading. Multiply this remainder by the con- 
 stant indicated on the tracing bar and the result will give the area 
 in terms of the required unit. 
 
 We may consider a circle of such diameter that with the fixed 
 point P as its center and the tracing point following its circumfer- 
 ence the tracing wheel will constantly move in a direction along its 
 own axis, producing no rotation of the wheel. This will occur when 
 
INSTRUMENTS FOR COMPUTING EXPERIMENTAL DATA 33 
 
 the angle PDF is a right angle. Obviously no change in reading 
 of the wheel can result and this circle is called the zero circle. In 
 case an area to be measured is so large that it cannot be traced with 
 the fixed point P placed outside, it may be placed inside the figure, 
 but in this case a correction must be added equal to the area of the 
 zero circle. This correction is marked on the tracing bar for each 
 of the several positions of the saddle. m 
 
 FIG. 12. 
 
 Theory of the Instrument. In the diagrammatic Fig. 12, OT 
 is the zero circle with radius F'P such that a perpendicular from 
 P on F'D' falls at D f , passing through the plane of the wheel. AFB 
 is a line lying outside of the zero circle with an area lying between 
 it and the zero circle that is to be measured. Suppose the pointer 
 F is made to trace an elementary portion of this area included in 
 the angle dB. The movement begins at F, extends radially to the 
 
34 EXPERIMENTAL ENGINEERING 
 
 zero circle, then along the zero circle through the angle d6, then 
 radially through the distance F'F to AFB, then along AFB through 
 angle d6, back again to F. It is evident that the first and third 
 movements are equal and opposite in direction, each causing a rota- 
 tion of the wheel, but with the two movements serving to nullify 
 each other. The second movement along the zero circle causes no 
 rotation of the wheel. There remains the fourth movement along 
 AFB which will be registered on the wheel when the pointer re- 
 turns to F. 
 
 Consider the pointer in the position F, with the wheel at D and 
 the pivot at C. While F is moving through the angle dO its in- 
 stantaneous center is P. C also swings about P which is therefore 
 the instantaneous center for the whole rod FD. The movements 
 of F and D are then in the proportion PF : 'PD. The linear move- 
 ment of P is PFdO and of D is PDdB. Drop a perpendicular from 
 P on FD extended. The linear movement of D, which is propor- 
 tional to PD, can be resolved into components, one parallel to FD 
 which is proportional to PM and causes no movement of the wheel, 
 the other perpendicular to FD, which is proportional to DM and 
 causes a linear movement of the circumference of the wheel that 
 we will call dR. 
 
 Let m = length of arm PC. 
 1 = length of arm OF. 
 n distance from pivot to wheel. 
 
 dR = DMdO=(mco$p-n)d0. (1) 
 
 The area of the element traced, 
 
 dA=%(r>-r' 2 )d0. (2) 
 
 In the triangle PFC we have 
 
 r 2 = m 2 + l 2 + 2mlcosfi. (3) 
 
 In the right triangle PF f D' we have 
 
 r'*=<PD' 2 +(l + n) 2 = m 2 -n 2 +(l + n) 2 = m 2 + l 2 + 2nl (4) 
 From (3) and (4), 
 
 and 
 
 dA = l(mcosj3-n)de. (5) 
 
INSTRUMENTS FOR COMPUTING EXPERIMENTAL DATA 35 
 From (1) and (5) 
 
 (6) 
 Integrating between the limits and R 
 
 This shows that the area is equal to the length of arm from pivot 
 to tracing point, multiplied by the space registered on the circum- 
 ference of the record wheel, and is independent of the other dimen- 
 sions of the instrument. 
 
 This is also true for areas not adjacent to the zero circle, or for 
 areas partly inside and out, as can be proved by subtracting the areas 
 between the zero circle and the given area. The demonstration is 
 general. 
 
 The instrument is usually constructed so that the arm / is adjust- 
 able in length, permitting its use for any scale or for various units. 
 In the No. 3 Amsler planimeter, shown in Fig. 11, two points are 
 shown on the back, one on the tracing bar and the other on the 
 saddle. The length between them equals I and by adjusting the 
 saddle so that this length is that of an indicator card, the reading 
 of the wheel will give the mean height of the card in fortieths of an 
 inch. 
 
 Coffin's Averaging Instrument. This instrument shown in Fig. 
 13 may be considered as a special form of Amsler's planimeter in 
 which the radius rod m is of infinite length, the point C, Fig. 12, 
 being constrained to move in straight lines. It is commonly em- 
 ployed for measuring mean effective pressures from indicator cards. 
 It consists essentially of a rod having on one end a tracing point 0, 
 which is moved over the outline of the diagram. The other end of 
 the rod contains a pin, which slides in the groove in the plate I, 
 on the left of the figure, the pin being maintained in contact with 
 the grove by the weight Q. The rod also contains the bearings for 
 the spindle of the graduated wheel, near the lower edge of the figure. 
 The rod is thus supported on three points, namely, the tracing point 
 Q, the pin Q, and the flange of the graduated wheel. Attached to 
 the board are a pair of clips C and K, of which the latter is capable 
 
36 
 
 EXPERIMENTAL ENGINEERING 
 
 of being moved parallel to itself by means of the slide at its lower 
 extremity. One of the horns of the rod which supports the gradu- 
 ated wheel is divided so as to form a vernier for the more accurate 
 reading of the graduations. 
 
 To obtain the mean height of an area (in the figure an indicator 
 diagram) slide the paper containing the area under the two clips 
 
 FIG. 13. 
 
 C and K t and arrange it so that a horizontal line of the diagram is 
 parallel to the horizontal edge of the clip C, and the left hand end 
 of the diagram touches the vertical limb of the same clip. In an 
 indicator diagram the atmospheric line is horizontal. Now push the 
 clip K towards the left until its inner edge touches the right hand 
 end of the diagram (in the figure at 0). Place the planimeter in 
 the position shown in the figure with the tracing point at where 
 the clip K touches the diagram. Press the head D of the tracing 
 
INSTRUMENTS FOR COMPUTING EXPERIMENTAL DATA 37 
 
 point so as to make a mark at , and then raise one of the horns 
 while the graduated wheel is turned to zero. The outline of the 
 diagram is now carefully traced over with the tracing point in a 
 clockwise direction until the starting goint is reached. Now 
 slide the tracing point along the clip K (keeping the eye on the 
 rolling wheel) until the wheel indicates zero. Then OA is the mean 
 height of the diagram and by using a scale graduated to represent 
 the scale of the indicator spring the mean effective pressure can be 
 read off directly. 
 
 Care must be taken to have the card properly adjusted as the 
 instrument mechanically divides the area traced by the horizontal 
 distance between the edges of the clips. It is necessary that this 
 distance accurately represent the length of the card. 
 
 Slight errors in the measurement of the line OA are likely to 
 occur. When greater accuracy is desired it is usual to read off 
 the area of the card on the wheel when the pointer has returned to 
 0. The wheel with vernier is graduated to read in square inches to 
 1/50 square inch. In this case it is better not to attempt to set the 
 wheel to zero before beginning to trace the card, but note the read- 
 ing of the vernier. The difference in reading at beginning and end 
 of the tracing will be the required area. Dividing by the measured 
 length of the card we obtain the length of the mean ordinate, which 
 on multiplying by the scale of the indicator spring gives the mean 
 effective pressure. 
 
 The Coffin instrument may be used to measure the area of any 
 figure that can be placed in position on the board. The size and 
 shape of the board, however, practically limits its use to the meas- 
 urement of indicator cards, for which it is primarily designed. 
 
 Theory of the Coffin Planimeter. Fig. 14 is a diagram repre- 
 senting the instrument, which, on inspection, is found to be the 
 Amsler instrument in modified form. Q is the pivot moving in a 
 vertical line CQ. The arm m becomes infinite. Continuing the com- 
 parison CQ may be considered as the circumference of the zero 
 circle whose radius is also infinite. The registering wheel D is 
 placed on an axis parallel to arm I to which it is fixed. It is evi- 
 dent that the exact position of D is immaterial, it being placed so 
 as to give the most convenient arrangement. Suppose we make the 
 
38 
 
 EXPERIMENTAL ENGINEERING 
 
 pointer describe an elementary area of height dx, bounded on one 
 side by CQ. The movement of the wheel in and out parallel to 
 OE will be equal in amount and in opposite direction, one move- 
 ment nullifying the other. No rotation can result from a move- 
 ment of the pointer along QC. There remains then only the move- 
 ment dx along the line OX during which the wheel moves also an 
 
 FIG. 14. 
 
 amount dx. The component of this movement in a direction tend- 
 ing to rotate the wheel is dx cos 0= = =dR. 
 
 The elementary area is dA=adx=ldR. 
 Integrating between and R 
 
 A = IR. 
 This expression is general, since all areas may be referred to CQ. 
 
 The Coffin Planimeter as an Averaging Instrument. After de- 
 scribing an area lying between EC and OX, bringing the pointer 
 back to the original starting point, suppose we move the pointer 
 along OX until the reading of the wheel is again the same as before 
 describing the area. Marking this point A, it is apparent from the 
 preceding that OA X OE will represent the described area. If then 
 
INSTRUMENTS FOR COMPUTING EXPERIMENTAL DATA 39 
 
 care has been taken to adjust the figure on the instrument so that 
 OE correctly represents its length, OA will represent its mean 
 height. This property of the instrument is made use of in obtain- 
 ing the mean height of indicator cards. 
 
 Several other planimeters have been devised, but in the U. S. 
 Navy, the Coffin and Amsler instruments are generally employed, 
 the former for obtaining mean effective pressures from indicator 
 cards and the latter for measuring the areas of miscellaneous 
 figures. 
 
CHAPTER III. 
 
 INSTRUMENTS FOE RECORDING EXPERIMENTAL DATA. 
 Measuring Machines. 
 
 As a basis for all manufacturing on the interchangeable system, 
 means of measuring with precision are imperative. The foun- 
 dation for such a system is a Standard Measuring Machine, by 
 which existing gages can be duplicated and new ones originated as 
 occasion requires. With the machine illustrated in Fig. 15, skilled 
 mechanics, accustomed to micrometric work, will measure positively 
 to 0.00005 inch, which is the practical limit of exact duplication, 
 and variations of 0.00001 inch may be readily detected. 
 
 The Pratt and Whitney Measuring Machine. Description. 
 This machine, shown in Fig. 15, has a heavy cast-iron bed supported 
 by two lugs at one end and one at the other. The bed is accurately 
 surfaced in straight lines and carries two head stocks, the one at 
 the left, A, being rigidly secured, and the one at the right, E, being 
 capable of movement along the bed. Each head stock has a spindle 
 through it. The one at the left, C, is capable of movement in an 
 axial direction and has a spring that tends to keep it to the right, 
 bringing the stops at D in contact. The spindle in the moving head 
 stock, E, is moved in and out by rotation of the graduated wheel F, 
 which operates a screw having 50 threads to the inch. This spindle 
 carries an index G which moves along a scale reading to 1/50 inch. 
 
 The wheel is graduated around its circumference, the main 
 divisions being numbered from to 200, and each main division 
 being further divided into two parts. Movements of the spindle 
 can therefore be read direct to 
 
 The end of each spindle is squared off and accurately ground and 
 scraped to a true surface. The moving head stock has a clamp H 
 for securing it firmly in position. It has also an adjusting screw 
 J for fine adjustments. This has another clamp, not seen in the 
 
INSTRUMENTS FOR EECORDING EXPERIMENTAL DATA 41 
 
 figure, which must be set up, H being slacked off, when fine adjust- 
 ments are to be made. 
 
 Attached to the moving headstock, is a microscope, K, with axis 
 vertical, which focuses on an index bar, not seen in the figure, 
 attached to the rear side of bed plate. This bar carries a series of 
 plugs with accurately surfaced faces, on each of which is drawn at 
 right angles two intersecting hair lines, the intersections being 
 
 FIG. 15. 
 
 spaced exactly one inch apart. "Under the eye piece of the micro- 
 scope is an adjustable glass, on which is a hair line drawn in a direc- 
 tion perpendicular to the plane of the machine. The wheel L en- 
 ables this glass to be moved so as to bring this line into coincidence 
 with the crossed lines on the plugs. N-N are rests for holding in 
 position the work that is to be measured. 
 
 Operation. To use the machine it must first be adjusted to 
 zero. Eun the index G to zero of the linear scale at top of head, 
 and bring the index pointer nearly to zero on the graduated wheel 
 F. Slide the head B nearly to contact between the measuring faces 
 
42 EXPERIMENTAL ENGINEERING 
 
 and adjust the latter by means of the adjusting screw J until the 
 drop, or indicating plug, at D shows a tendency to move. Then 
 clamp the head firmly and adjust F until the drop plug falls into a 
 vertical position, but not entirely out of contact with the faces with 
 which it is held. Then adjust the index for F to zero by means of 
 the screw M , and bring the line on glass under eye piece to the zero 
 mark on the index bar. 
 
 The machine is now adjusted for zero for its entire length, but 
 care must be taken, not to disturb the eye-piece line for any subse- 
 quent measurement of over one inch. If there is any doubt as to its 
 having been moved, return to zero and make sure of it. 
 
 To measure a length greater than one inch, this range being ob- 
 tained by the micrometer screw, bring the head B into range of the 
 plug on index bar from which the measurement is to be taken and 
 adjust the line under eye piece to coincide with mark on this plug, 
 using screw J to make this adjustment. Then clamp the head and 
 proceed. The micrometer screw should be run out to one inch if 
 the graduation on the bar is less than the amount to be measured, 
 and run back to zero if the graduated line is greater than the length 
 required to be measured. This will be obvious after a few trials. 
 
 The pressure of contact is uniform at zero and at any distance in 
 measurement of end gages, but precautions must be taken to avoid 
 variation in temperature of the end gages, especially those of con- 
 siderable length. 
 
 Flexure of the end gages must also be avoided, and if two sup- 
 ports are used they should be placed each at about one quarter of the 
 distance from each end. 
 
 Care of the Machine. Benzine, used with a soft woolen cloth, 
 will clean the polished surfaces of the graduated plugs on the index 
 bar, and a fine camelVbair brush will afterwards serve to remove 
 dust and not scratch them. The use of a fine grade of kerosene 
 will be useful to clean the surfaces, as it will not rust them, and a 
 little may be left on them without doing any injury or preventing 
 the clear definition of the lines at any time. 
 
 The microscope must be clamped in place with each adjustment. 
 It must be removed in order to place the covers over the index bar. 
 On replacing it, it must be readjusted for zero. 
 
INSTRUMENTS FOR RECORDING EXPERIMENTAL DATA 43 
 
 Measurement of Pressure. 
 
 Fluid pressures are commonly measured by some form of gage 
 in which the laws of deformation of elastic material are made 
 use of. Such gages are of two kinds, pressure gages, which record 
 pressures above the atmosphere, and vacuum gages, which record 
 pressures below the atmosphere. In America and England, pressure 
 gages are graduated to record pressures in pounds per square inch, 
 
 FIG. 16. 
 
 except for very heavy pressures which are sometimes recorded in 
 atmospheres. Vacuum gages register the difference between the 
 atmospheric pressure and the pressure in the vessel to which at- 
 tached. They are graduated to show this difference in inches of 
 mercury. This system has been adopted to facilitate comparison 
 with the barometer, which is read in inches of mercury. The dif- 
 ference between barometer and vacuum gage readings gives the 
 absolute pressure in inches of mercury in the vessel to which the 
 vacuum gage is attached. One cubic inch of mercury weighs 0.49 
 pounds. Hence to convert inches of mercury into pounds per square 
 inch, multiply by 0.49. 
 
44 
 
 EXPERIMENTAL ENGINEERING 
 
 Pressure gages and vacuum gages are similar in construction. 
 They are of two general types, (1) the diaphragm gage, one of 
 which is shown in Fig. 16, and (2) the bent tube or Bourdon gage, 
 shown in Fig. 17. In the first of these types the diaphragm is in 
 equilibrium under atmospheric pressure. If pressure be applied to 
 either side it is deflected an amount proportional to the pressure 
 applied. This deflection causes a movement of the attached strut, 
 which is communicated through the connections shown, to the cen- 
 tral spindle, carrying an index that moves around a graduated dial. 
 
 FIG. 17. 
 
 In the second type, there is a flattened tube communicating 
 with the fluid under pressure. If the pressure be increased, it tends 
 to round out the flattened section, and thus tends to straighten the 
 tube. Fig. 17 shows a single tube gage, in which the end of tube is 
 connected through levers with a sector that gears with a pinion on 
 the spindle-carrying index as before. If this tube be opened to a 
 vacuum, it will tend to still further flatten, and the ends will tend 
 to come closer together, with the reverse movement of index over 
 dial. The same construction is therefore used for both pressure and 
 vacuum gages. 
 
 The double tube gage, shown in Fig. 18, is of similar con- 
 struction, except that both ends of the tube are free to move, and 
 the actuating mechanism for the index is suspended between them. 
 
INSTRUMENTS FOR KECORDING EXPERIMENTAL DATA 45 
 
 The connection for introducing the pressure is in the middle of the 
 tube, which is secured in place. This form of gage has the advan- 
 tage that the tube may be drained completely when the pressure is 
 .off, which gives it somewhat greater durability. 
 
 A Compound Gage is often used for receivers, jackets, or other 
 places where the steam pressure is sometimes above and sometimes 
 below the atmosphere. The indicates atmospheric pressure. The 
 dial is graduated to the right of the for pressures above the at- 
 mosphere in pounds per square inch, and to the left of the for 
 
 FIG. 18. 
 
 pressures below the atmosphere in inches of mercury. It will be 
 noted that on a simple vacuum gage, the connections are such that 
 the index moves to the right, but on a compound gage, when regis- 
 tering vacuum, the index moves to the left. 
 
 Recording Pressure Gage. There are several forms of this 
 apparatus, in all of which the principle of operation is the same. 
 The mechanism for registering the pressure is similar to that in an 
 ordinary gage, but the moving index is longer, and the spring is 
 stiffer, so that the limit of pressure is recorded in a comparatively 
 small arc of the circle swept by the index. The end of index carries 
 a pen or pencil, and the pressure is recorded on a sheet of paper that 
 receives motion from a clock mechanism. The paper is laid off and 
 marked for intervals of time in one direction and for intervals of 
 
46 
 
 EXPERIMENTAL ENGINEERING 
 
 pressure in the other. The apparatus then gives a continuous record 
 of the pressure on the gage. In some forms of the apparatus, where 
 the paper is unwound from a spool, this record continues for several 
 days on the same sheet of paper. In other forms, as in that 
 described below, the paper must be changed every 24 hours. 
 
 FIG. 19. 
 
 Bristol's Recording Pressure Gage. As used in recording pres- 
 sures in pounds per square inch, this is to be seen in the labora- 
 tory, as shown in Fig. 19. The recording mechanism contains a 
 flattened tube, similar to that of the single tube Bourdon gage, ex- 
 
INSTRUMENTS FOR EECORDING EXPERIMENTAL DATA 47 
 
 cept that it is wound in a spiral, and the pressure when applied 
 causes it to unwind to an extent proportional to such pressure. The 
 index is carried from the end of the spiral and has a pen attached to 
 its free end. This traces a line on the revolving paper dial which 
 
 FIG. 20. 
 
 gives a continuous record of the pressure. The paper, when placed 
 on the apparatus, is adjusted so as to bring the time, as marked on 
 the paper for the instant of making such adjustment, under the pen. 
 This dial will then be good for 24 hours, when it must be removed, 
 
48 
 
 EXPERIMENTAL ENGINEERING 
 
 and is then marked with the date and any other data that it may 
 be desirable to record. 
 
 Fig. 20 shows another form of the same apparatus, used for 
 recording air pressures. The tube is bellows-shaped and its tend- 
 
 FIG. 21. 
 
 ency to elongate is resisted by a metal strip on one side. The pen 
 arm is attached directly to this diaphragm tube and pressures are 
 recorded on a dial in the same manner as before. 
 
 Fig. 21 shows the exterior of the gage, with the form of dial used 
 to give the continuous record. 
 
INSTRUMENTS FOR RECORDING EXPERIMENTAL DATA 49 
 
 Uehling Differential Pressure Eecorder. The working principle 
 of the Uehling Differential Pressure Eecorder includes a gravity 
 U tube, AEB, Fig. 22. B is suspended by a rod M from a bell float 
 I which is buoyed by mercury contained in a cylindrical vessel J. 
 M passes freely through a central tube H and in front of the record- 
 ing paper which is fed by clock work and kept taut by, and auto- 
 matically wound on, a receiving roller T. A small horse-shoe mag- 
 net N pivoted on a bracket fastened to rod M serves as a pen holder. 
 The pen is kept to the paper by the magnetism of the horse shoe 
 acting on a narrow iron strip 'P. B is connected to A and D by 
 means of flexible tubes E and F. F communicates with the top of 
 floating chamber B through C, and E with the bottom of said 
 chamber as shown. Z is an equalizing tube controlled by cock X. 
 W connects with the lower and Y with the higher pressure. 
 
 Connections having been made, for example, with a Venturi or 
 Pitot meter; open X then open W and Y. The pressure will be 
 equalized through Z and the mercury in A and B will stand at the 
 same level; B carrying the major part of the mercury will pull 
 down the bell float until the weight of mercury displaced by the 
 latter equals the weight of B with its mercury content and connec- 
 tions. When equilibrium is thus established, the pen will point to 
 the zero line. Closing X the higher pressure will be confined to D, 
 the lower pressure to A, and in consequence mercury will be driven 
 from B to A until the pressure difference is balanced by the mercury 
 head in A. As mercury flows from B to A the weight suspended by 
 M diminishes and the bell float rises, carrying with it the pen and 
 also chamber B. As the pressure difference diminishes, the mercury 
 returns to B, its weight increases and the reverse movement takes 
 place. 
 
 The cross sectional areas of the stationary leg A, the floating leg 
 B and the bell float I are so related that the pen will move over the 
 entire scale covering the available width of the chart for any pre- 
 determined pressure difference to be recorded. 
 
 This instrument can be calibrated to record pressure differences 
 from to 1 inch, to from to 30 inches of mercury head or more, 
 under a total pressure of 200 Ibs. and over. In all cases the calibra- 
 tion on chart covers 3 inches in width. 
 
50 
 
 EXPERIMENTAL ENGINEERING 
 
 FIG. 22. 
 
INSTRUMENTS FOR EECORDING EXPERIMENTAL DATA 51 
 
 The recorder can be conveniently fastened to any wall or column. 
 It makes a continuous rectilinear record of pressure differences or 
 rate of flow of either liquids or gases. Used in connection with a 
 Venturi or Pitot meter described in Chapter VI, it may be cali- 
 brated to record cubic feet per hour or other units as may be desired. 
 
 Testing Pressure Gages. Fig. 23 shows a late form of appara- 
 tus for testing gages, as made by Messrs. Schaeffer & Budenberg and 
 installed in the laboratory. It is in two parts, for testing pressure 
 gages and vacuum gages. On the left is the arrangement for test- 
 ing pressure gages. Pressure is applied by means of a screw 
 plunger, that is worked in and out in a cylindrical chamber by means 
 of the hand wheel shown at the extreme left. A reservoir, contain- 
 ing oil or glycerine, connects with this chamber through the valve A. 
 Valve B connects the chamber with a vertical cylinder, in which is 
 an accurately fitting piston of 14 square inch area of cross section. 
 This piston carries a tray on which the disc-shaped weights, shown 
 at the back, are placed for directly measuring the applied pressure. 
 The tray and piston, together, weigh li/4 pounds, so that when bal- 
 anced, without the addition of weights, the pressure per square inch 
 in chamber is 5 pounds. A pipe connects the chamber with the 
 mountings shown at the front, on which gages may be placed. The 
 gage under test is mounted at the left, the valve C serving to shut 
 off or turn on the pressure. The mounting at the right, to which 
 pressure is admitted through the valve D, is for a standard gage, 
 but may be used for ordinary testing work, in which case two gages 
 may be tested at once. 
 
 Operation. To test a gage, the connection to reservoir is opened 
 and all other connections are shut off. The plunger is forced in its 
 full length and the reservoir is filled with oil. Then the plunger 
 is withdrawn as far as it will go in the chamber, and the connection 
 to reservoir is shut off. The gage for test is mounted, its connecting 
 valve is opened, and the piston for carrying weights is put in place. 
 Opening the valve under the piston, weights are added to the tray 
 in increments of 5 or 10 pounds, and pressure applied by forcing 
 in the plunger until the piston lifts, each time a weight is added. 
 To prevent piston or index sticking, the piston is spun around each 
 time the gage is read, thus ensuring a correct reading. 
 
52 EXPERIMENTAL ENGINEERING 
 
 In case the gage is found to read incorrectly, the index should 
 be removed and properly set. In cheaply constructed gages a dif- 
 
 FIG. 23. 
 
 ferent error will sometimes be found at low and at high pressures. 
 In such cases the correction should be made at the average pressure 
 for which the gage is to be used. 
 
INSTRUMENTS FOR BECORDING EXPERIMENTAL DATA 53 
 
 Where many gages are to be tested at the same time, it will be 
 found most convenient to first test the " standard " gage, then shut 
 off the vertical cylinder and compare the readings of the other gages 
 with the " standard." A standard gage should, however, never be 
 relied upon, unless tested immediately before or after using. 
 
 Testing Vacuum Gages. The apparatus at the right in Fig. 
 23 is for testing vacuum gages. A vacuum pump at the extreme 
 rights connects through cock F with a mercury column and through 
 valve E with the mounting on which the gage to be tested is placed. 
 
 Testing Gages on Board Ship. The above apparatus is some- 
 times supplied to large vessels. The usual form of apparatus sup- 
 plied is similar to that described for testing pressure gages, but in 
 a light portable form with the vacuum pump omitted. Its opera- 
 tion is the same as that of the larger apparatus. 
 
 Vacuum gages on board ship are usually tested by comparison 
 with one another. If a more accurate test is desired and no means 
 are supplied for conducting such test, they may be taken ashore for 
 this purpose at a navy yard at frequent intervals. 
 
 Manometers. This name is applied to a gage used to register a 
 difference between two pressures, usually where the difference is 
 small. Such gages are used to show the air pressure in a closed fire 
 room, or the draft pressure in a flue passage, where the difference 
 between the variable pressure and the atmospheric pressure is ob- 
 served. The common form of manometer is a U-shaped glass tube, 
 partially filled with water, one leg of which is connected to the body 
 of air or gas, the pressure of which is to be measured, and the other 
 leg is connected to the atmosphere. A scale shows the difference in 
 height of the water in the two legs of the tube, and thus indicates 
 the pressure in inches of water. For indicating slightly greater 
 pressures, a mercury-filled manometer is sometimes used, in which 
 the indication is of course in inches of mercury. 
 
 Measurement of Temperature. 
 
 Ordinary temperatures of water and steam are measured by means 
 of mercurial thermometers. Before proceeding with any experi- 
 ment requiring accuracy, such thermometers should be carefully 
 calibrated, unless they are standard thermometers of known ac- 
 
54 EXPERIMENTAL ENGINEERING 
 
 curacy. The bulb of the thermometer and so much of the stem as is 
 ordinarily immersed in the liquid whose temperature is to be taken, 
 is packed in melting ice and after the mercury becomes stationary 
 the reading is taken. It is again placed in a vessel of boiling water 
 that is open to the atmosphere and the reading again taken. Eead- 
 ings are thus obtained at the known temperature of 32 and 212 
 on the Fahrenheit scale or and 100 on the Centigrade scale. 
 From these the proper corrections are derived to be applied to any 
 thermometer reading. It is usual to assume that the scale of the 
 thermometer is constant above the boiling point to the limit of the 
 tube. For this to be true it is necessary that the area of cross 
 section of the tube should be constant. 
 
 High Temperatures. Pyrometry. Mercury boils at 675 F., 
 under atmospheric pressure. It becomes necessary to resort to ap- 
 paratus other than the ordinary mercurial thermometer for the 
 measurement of temperatures above about 500 F. Such instru- 
 ments are known as high temperature thermometers or pyrometers. 
 Several different types have been constructed as follows : 
 
 Gas thermometers. 
 
 Pneumatic pyrometers. 
 
 Mercurial pyrometers. 
 
 Expansion pyrometers. 
 
 Calorimetric pyrometers. 
 
 Thermo-electric pyrometers. 
 
 Eesistance thermometers. 
 
 Eeflecting pyrometers. 
 
 The Gas Thermometer. This is used as the standard by which to 
 calibrate all other high temperature thermometers, and for research 
 work where great accuracy is required. It is not suited for ordinary 
 experimental work on account of its lack of portability and its 
 high cost. 
 
 Instruments of this type are usually specially constructed for the 
 service that is required, but the principle of operation in all such 
 instruments is the same. The relation between pressure volume 
 and temperature of a gas follows the law PV = CT, where P is the 
 absolute pressure, V is the specific volume, C is a constant and T is 
 the absolute temperature. A large bulb is constructed to contain 
 
INSTRUMENTS FOR KECORDING EXPERIMENTAL DATA 55 
 
56 
 
 EXPERIMENTAL ENGINEERING 
 
 a fixed volume of gas. Various materials have been used for the 
 bulb, but the best results are obtained with bulbs of platinum,, or 
 where very high temperatures are to be measured, of an alloy of 
 
 FIG. 25. 
 
 platinum and iridium. Air and hydrogen have been used in the 
 bulb, but the best results have been obtained with nitrogen. 
 
 The bulb being subjected to heat, the rise in temperature is 
 measured by the rise in pressure of the gas. This is registered on a 
 gage, graduated to read the degrees rise in temperature. 
 
INSTRUMENTS FOR EECORDING EXPERIMENTAL DATA 57 
 
 The Industrial Thermograph is a practical instrument based on 
 the principle of the gas thermometer. The general appearance of 
 the instrument is shown in Fig. 24, while the moving element is 
 shown in Fig. 25. There is a metallic bulb, connected by very fine 
 copper tubing with the coiled tube shown in Fig. 25. Instruments 
 on this principle are constructed for all ranges of temperature up 
 to about 1000 F. In instruments made for temperatures below 50 
 the bulb is filled with alcohol and the long capillary tube is omitted, 
 the bulb being placed close to the pressure tube. For instruments 
 having a range between 50 and about 500 sulphur dioxide (S0 2 ) 
 is the medium employed, while for instruments intended for a 
 range above 400 the bulb is filled with nitrogen, making the in- 
 strument in this case a true gas thermometer. The recording gage 
 shown in the figure may be replaced with an indicating gage. In 
 any case the gage is graduated for degrees rise in temperature. 
 
 TTehling's Pneumatic Pyrometer. The underlying principle of 
 
 B 
 
 FIG. 26. 
 
 this instrument is based on the same law that governs the opera- 
 tion of the gas thermometer. It is illustrated diagrammatically in 
 Fig. 26. A small steam aspirator, working at a uniform rate, 
 draws air through the apertures A and B, causing a partial vacuum 
 in chambers C and D. The same amount of air must necessarily 
 pass through each aperture. If A and B are the same size, and if 
 the air remains at a fixed temperature during a given length of 
 time, the suction in chamber C will be practically twice that in 
 
58 EXPERIMENTAL ENGINEERING 
 
 chamber D. If, however, this air is heated when it passes through 
 A, but again cooled to a lower fixed temperature before it passes 
 through B, the specific volume will be higher on passing through B 
 and therefore the pressure in C will be lower. In the same way any 
 change in temperature of the air flowing through A will have its 
 influence on the amount of vacuum in chamber D. Thus the ma- 
 nometer tube P may be calibrated to indicate the temperature of 
 the air passing through A. 
 
 Fig. 27 shows a diagrammatic disposition of all the parts com- 
 bined in the complete instrument. 
 
 The interior of the pipe e, f, g, li, i from aperture to aperture, to- 
 gether with the branches q and s, constitute the chamber C of Fig. 
 26. Its inlet from the atmosphere is through the opening a at the 
 bottom of filter I, and its connection with chamber C' is through the 
 pipe?. 
 
 The aspirator D exhausts into the chamber G, keeping it at a 
 constant temperature of 212. The steam and condensed water, 
 together with the air drawn through the aperture, escape through 
 the pipe t at atmospheric pressure. Opening the valve 6 steam 
 enters the aspirator D, and sucks the air through the tube m, out of 
 the chamber C' and produces a suction, which is kept constant by 
 the regulator H, as shown by the manometer p. With a constant 
 suction in C' and cocks 2 and 4 open, air will enter at a, pass 
 through the filter I, where it is purified, then through the connec- 
 tion b into the fire tube. It flows forward in the annular space 
 between the two tubes c and f; as soon as it reaches the platinum 
 tube d, which protrudes from the cooler, it becomes heated and 
 enters through the aperture A into the chamber C, at the tempera- 
 ture surrounding the exposed end of the fire tube, which is the tem- 
 perature to be measured. After passing A, the air flows through the 
 pipe e, f, g, h into the coil i, where it assumes the temperature of 
 212, at which it passes through aperture B, thence by the connec- 
 tion I into the chamber C", from which it is drawn by the aspirator 
 D through m, and discharged with exhaust steam and condensed 
 water. 
 
 The branch pipes s and q' connect respectively with the recording 
 gage L, and the manometer q, which is placed in proximity to the 
 temperature scale, as shown. 
 
INSTRUMENTS FOR EECORDING EXPERIMENTAL DATA 59 
 
 H 
 
 PIG. 27. 
 
60 EXPERIMENTAL ENGINEERING 
 
 This combination, therefore, fulfils all the conditions viz., air 
 is drawn through the instrument by a constant suction. It passes 
 through aperture B at a constant temperature. Aperture A is so 
 located that the air must enter at the temperature to be measured, 
 hence the indication of the manometer q will vary with the tem- 
 perature at A , as we have demonstrated, and can be read off directly 
 on the temperature scale placed beside the same. 
 
 Fig. 29 shows the instrument complete, with recording gage and 
 with portable fire tube. The fire tube is shown in section in Fig. 
 28. The aperture A of the diagram is located near the closed end 
 of a small platinum tube e, placed within a larger tube d, which is 
 also of platinum with closed end. Both d and e are brazed into 
 drawn copper tubes c and /, and these tubes are surrounded by a 
 water jacket F. Before entering aperture A the air is drawn 
 through the cotton filter and through the annular space between 
 tubes d and e, where it reaches the temperature to be measured. 
 
 Connections to the fire tube may be made of flexible tubing, thus 
 enabling the tube to be inserted successively into different furnaces 
 within a range of 150 feet. 
 
 This instrument is accurate and durable and has a temperature 
 range up to about 3000 F. While the fire tube is portable, the 
 instrument itself must be set up permanently and for this reason 
 it is not well adapted to ordinary marine work. 
 
 Mercurial Pyrometer. This is similar to the ordinary mercurial 
 thermometer in that there is a bulb and tube filled with mercury, 
 but the space in tube above the mercury is filled with nitrogen. 
 
 As the tube is heated the mercury expands and compresses the 
 nitrogen, the resultant increase in pressure on the mercury raising 
 its boiling point. These pyrometers are made in both portable and 
 stationary form. The end which contains the nitrogen may take 
 the ordinary form of a mercurial thermometer, or may terminate 
 in a pressure gage calibrated in degrees of temperature. The latter 
 form of instrument is more easily read and a recording gage may 
 be fitted from which a chart can be taken showing the temperature 
 changes throughout each twenty-four hours. In Fig. 30 is shown 
 a typical mercurial pyrometer with recording gage and flexible tube 
 connection. 
 
INSTRUMENTS FOR RECORDING EXPERIMENTAL DATA 61 
 
EXPERIMENTAL ENGINEERING 
 
 FIG. 29. 
 
INSTRUMENTS FOR KECORDING EXPERIMENTAL DATA 
 
 63 
 
 Mercurial pyrometers are fairly accurate, and are particularly 
 adapted to measuring uptake and chimney temperatures. They are 
 constructed to read as high as 1000 F. 
 
 Expansion Pyrometers. The difference in the rate of lineal ex- 
 pansion of two metals is the principle upon which expansion pyrom- 
 eters are based. For the same rise in temperature, copper expands 
 
 FIG. 30. Mercurial Pyrometer with Recording Gage. 
 
 60 per cent more than iron. In the expansion pyrometer a tube of 
 copper is inclosed in a tube of iron. At the end to be heated the 
 tubes are securely fastened to each other. At the other end they axe 
 attached to a set of multiplying gears which actuate a needle pointer 
 over the face of a properly calibrated dial. It is necessary to expose 
 the entire length of the expansion tubes to the full effect of the heat 
 to be measured; if this is not accomplished, error will result as the 
 proper amount of elongation has not been obtained. When the 
 
64 EXPERIMENTAL ENGINEERING 
 
 pyrometer is first inserted, the pointer will act rapidly in one direc- 
 tion or the other and give an untrue reading temporarily. This is 
 caused by the outer tube heating and expanding more rapidly than 
 the inner one. As soon as the inner tube heats up and expands 
 proportionately, the pointer will correctly indicate the temperature. 
 
 FIG. 31. Expansion Pyrometer. 
 
 When an expansion pyrometer has been used repeatedly for tem- 
 peratures near its limit, the indicator will no longer return to the 
 position indicating the temperature of the atmosphere. A perma- 
 nent change has taken place in the length of one of the tubes. By 
 loosening a set screw the dial may be adjusted to correct the varia- 
 tion. 
 
INSTRUMENTS FOR RECORDING EXPERIMENTAL DATA 65 
 
 A standard type of expansion pyrometer is shown in Fig. 31. 
 In spite of the fact that these pyrometers get out of calibration 
 rather easily they are capable of giving close results if understood 
 and carefully handled. They will indicate temperatures as high as 
 1500 F. 
 
 Calorimetric Pyrometers. The method of indicating temperature 
 by the calorimetric pyrometer is the reverse of that used in de- 
 termining specific heats by the water calorimeter. 
 
 A given weight of some metal, the specific heat of which is known, 
 is heated to the temperature to be measured and then instantly 
 plunged into a known weight of water. The rise in temperature of 
 the water is noted. The formula for finding the temperature is : 
 
 x=T+ wt 
 
 ws 
 in which, 
 
 X = Temperature to be measured in degrees Fahrenheit; 
 T= Final temperature of the water; 
 W= Weight of the water in pounds; 
 =Rise in temperature of the water; 
 w Weight of the metal, in pounds; 
 s Specific heat of the metal. 
 
 A calorimetric pyrometer is an apparatus by means of which the 
 temperature may be found without using each time the formula 
 previously given. It usually consists of a copper cup which is in- 
 sulated to prevent loss from radiation and which has gage marks 
 upon it for a definite amount of water. Fastened to the cup and so 
 arranged as to be properly immersed, is a small thermometer. A 
 sliding scale is attached to the thermometer. A copper or platinum 
 ball completes the outfit. The scale on the thermometer is calibrated 
 with the quantity of water which the instrument holds and with 
 the metal ball so that the rise in temperature of the water causes 
 the thermometer to indicate correctly the temperature being 
 measured. 
 
 This pyrometer is very liable to inaccuracy. The results which it 
 gives vary with the skill and care used in its manipulation. Its 
 cheapness and portability, however, recommend it where approxi- 
 mate results are satisfactory. It may be used for temperatures up 
 to 3000 F. 
 
66 EXPERIMENTAL ENGINEERING 
 
 Thermo-Electric Pyrometers. When rods or wires of two dis- 
 similar metals are joined at one end, they compose a thermo-electric 
 " couple " or " element." When the junction is heated, a difference 
 of electrical pressure is established between the cool ends. If a 
 circuit is made by joining the cool ends together, either directly or 
 b} r means of a conductor, a current will flow. The strength of the 
 current depends upon the nature of the " element " and the dif- 
 ference in temperature between the hot junction and the cool ends 
 or cold junction. The deflection of a galvanometer placed in the 
 circuit may be calibrated, then, to indicate the temperature at the 
 hot junction. 
 
 FIG. 32. 
 
 A thermo-electric pyrometer consists of the thermo-electric couple, 
 the galvanometer and a device for compensating for the changes in 
 temperature at the cold junction. The hot junction is protected 
 by a fire-proof insulation of asbestos and carborundum, and the 
 whole is encased in a common iron pipe for further protection. In 
 most makes for commercial use, the change in temperature of the 
 cold junction is taken care of by shifting the scale on the galva- 
 nometer a few degrees one way or the other when the temperature 
 at the cold junction is observed to change appreciably. Fig. 32 
 shows a thermoelectric pyrometer outfit of the usual type. A re- 
 cording device is sometimes added. 
 
 Thermo-electric pyrometers are adaptable to many uses. With 
 proper handling they will give accurate results. Their great ad- 
 vantages are convenience and simplicity. Repeated reheating will 
 cause the potential of the thermo-electric couple to change, necessi- 
 tating frequent recalibration and an occasional renewal of the 
 couple. Where high temperatures are measured and continuous 
 
INSTRUMENTS FOR EECORDING EXPERIMENTAL DATA 67 
 
 service required, the cost of upkeep may be considerable. Depend- 
 ing on the composition of the thermo-electric element, these pyrom- 
 eters will measure temperatures up to 3000 F. 
 
 For temperatures up to 1100 F., the element is made of nickel 
 and constantan, a composition consisting of 60% copper and 40%' 
 nickel. For temperatures ranging between 1100 and 2300 F., 
 nickel and a special carbon are used, and for higher temperatures 
 up to 3000 F., platinum and an alloy consisting of 90% plati- 
 num and 10% rhodium or iridium make up the element. 
 
 Resistance Thermometers. The electrical resistance of pure 
 platinum varies directly with its temperature. This property is 
 made use of in the measurement of high temperatures where the 
 problem becomes one of measuring the electrical resistance of a 
 piece of platinum wire, heated to the temperature under investi- 
 gation. 
 
 A coil of platinum wire is wound in crossed layers of mica and 
 then, with the lead wires, is encased in a porcelain tube which pro- 
 tects the wires and holds them in position. The tube, with its con- 
 tents, is known as the ~bul~b and is shown in section in Fig. 33. 
 
 FIG. 33. 
 
 Fig. 34 shows a diagram of the electrical connections in the 
 apparatus. A, B, and C correspond to the three contact points on 
 the head of the bulb. R is a fixed resistance which at normal tem- 
 perature corresponds to the resistance of the platinum coil, rr 
 are fixed and equal resistances, yy are the resistances of the two 
 lead wires. These are equal and may be taken as negligible in the 
 following explanation of the principles of the instrument. DE is 
 a wire, of uniform section and having a resistance L, joining the 
 two resistances rr. It is connected by a sliding contact F through 
 the battery G to C. 
 
 Let x=ihe resistance of the platinum coil, varying under change 
 of temperature. 
 
 Let z = the resistance of DF. 
 
68 
 
 EXPERIMENTAL ENGINEERING 
 
 It will be noted that the diagram is that of a Wheatstone bridge, 
 in which the balance is produced by varying the position of the 
 sliding contact F on the wire DE. From this we see that 
 
 R 
 
 r + z 
 
 (r+z) 
 
 FIG. 34. 
 For the minimum reading of the thermometer, z L and 
 
 For the maximum reading, 2 = and 
 
 The known relation between the temperature and the resistance 
 of pure platinum, as determined by the experiments of Le Chatelier, 
 Callendar, and others, gives a means for graduating the scale for 
 the sliding contact to indicate directly degrees rise in temperature. 
 
INSTRUMENTS FOR BECORDING EXPERIMENTAL DATA 69 
 
 Fig. 35 represents a diagram of the electrical connections of the 
 instrument, as installed in the laboratory. There are wire con- 
 nections for six different bulbs, enabling temperatures to be read in 
 rapid succession for different parts of the furnace and connections 
 of a boiler while making a test. The switch A in the upper left- 
 hand corner is, as will be noted, arranged for taking readings from 
 
 FIG. 35. 
 
 any bulb. The sliding resistance wire, above described, is arranged 
 in circular form, corresponding to a range of 2200 F. for the 
 instrument. A galvanometer is seen at C. B is a key for making 
 or breaking the connection from a battery as may be desired. Fig. 
 36 shows the outside appearance of the instrument. 
 
 To take a reading, all connections having been made, the bulb 
 is inserted in the place the temperature of which is to be measured 
 and the index is moved around until the galvanometer balances and 
 remains stationary, when the temperature is read off directly. To 
 
70 EXPERIMENTAL ENGINEERING 
 
 test the instrument for adjustment, the bulb should be immersed in 
 a well-stirred liquid of known temperature for a period of at least 
 15 minutes. With the index moved to the point on dial correspond- 
 ing to this temperature, the galvanometer should balance if the in- 
 strument is in proper adjustment. If not, the proper correction, as 
 thus obtained, should be applied. 
 
 FIG. 36. 
 
 This instrument is capable of registering continuously tem- 
 peratures up to 1800 F. By withdrawing the bulb each time after 
 taking a reading, temperatures up to the full limit of the instrument 
 may be taken without injury to it. It is very useful for taking 
 practically simultaneous temperatures of the flue gases at different 
 points in a boiler. It has been calibrated by comparison with an air 
 thermometer and found to compare with it very favorably in accu- 
 racy. It is portable and can be set up where needed, while this can- 
 not be done with the air thermometer. 
 
INSTRUMENTS FOR RECORDING EXPERIMENTAL DATA 71 
 
 Reflecting Pyrometer. This instrument is shown diagrammatic- 
 ally in Fig. 37 where the heat rays from AB enter the open end EF 
 of a tube and impinge on a concave mirror C , having one focus at 
 EF and another at D, where there is placed the hot junction of a 
 small thermo-couple. 
 
 The mirror C concentrates upon D the heat image of the aperture 
 EF filled with radiant heat from the hot bodies and this heat image 
 raises the temperature of the thermo-couple, giving rise to an elec- 
 tro-motive force. Connection is made to an indicating millivolt- 
 meter by means of a flexible cable and the indications of this milli- 
 voltmeter are calibrated to read in degrees of temperature direct. 
 
 The relative position of the mirror C, the sensitive device D and 
 the aperture EF are all fixed so that the focusing is done once for 
 all and is never changed in use. 
 
 -^ cl 
 
 FIG. 37. Diagram of the Receiving Tube. 
 
 In order that the aperture EF may be entirely filled with radiant 
 heat from the hot body A B it is necessary that the tube be brought 
 within a certain maximum working distance. This working dis- 
 tance varies with the size of the hot body and is measured from the 
 apex G of the cone GAB. 
 
 The rule for the working distance is that the center ring must be 
 within ten times the diameter or smallest dimension of the hot body. 
 Any distance less than this maximum working distance is satis- 
 factory and will not materially affect the reading. Looking at the 
 diagram in Fig. 37, it will be seen that if the hot body AB were 
 brought nearer to G the only result would be that the outer edges of 
 it would not be in the measurement as they could only radiate heat 
 to the walls of the tube and these are made non-reflexing. 
 
 The instrument is calibrated to give direct temperature reading 
 without corrections. It is best adapted for indicating furnace tem- 
 peratures for which purpose it is fairly accurate. 
 
72 EXPERIMENTAL ENGINEERING 
 
 Revolution Counters. Each main engine shaft is provided with 
 a continuous recording rotary counter for registering the number 
 of revolutions made. The general appearance of the instrument is 
 shown in Fig. 38 and some of the details of the mechanism, from 
 which its operation will be understood, in Fig. 39. 
 
 FIG. 38. 
 
 FIG. 39. 
 
 Maneuvering Indicator. On twin screw ships there is fitted in 
 each engine room an indicator consisting of a dial numbered from 
 to 100, with a red and a green hand for indicating the relative 
 speed of the engines. The green hand is actuated by the starboard 
 engine and the red hand by the port engine. Each of the hands 
 makes one complete revolution of the dial for 100 revolutions of 
 the actuating engine. 
 
 The indicators serve as guides in maneuvering. One engine, 
 usually the starboard, is operated at the required speed and, if the 
 same speed of revolutions is desired, the other engine is regulated 
 so that the two hands move together. In turbine-driven ships, 
 fitted with triple screws, the maneuvering indicators are worked 
 
INSTRUMENTS FOR EECORDING EXPERIMENTAL DATA 73 
 
 from the outboard shafts in the same manner as in twin-screw 
 ships. The maneuvering and backing turbines are fitted on the 
 outboard shafts. The center turbine does not come into action 
 until the vessel is fairly under way, and in working the ship it is 
 allowed to run idle. 
 
 Multiple Screw Installations. Differential Counter Gear. In 
 multiple screw installations it has been found that there may be 
 considerable variation in the revolutions made by the various shafts 
 and the speed will correspond to the mean revolutions of all the 
 shafts. It therefore becomes important to have a device for regis- 
 tering the mean number of revolutions. This is accomplished by 
 
 FIG. 40. 
 
 fitting differential gear to operate a counter. In Fig. 40 A and B 
 are operated each from different shafts. A differential wheel C is 
 operated from A and B through the miter wheels D , E and F. 
 The counter is actuated from C, the revolutions of which correspond 
 to the mean revolutions of the two shafts. A third differential gear 
 may be installed which will give the mean of the readings of the 
 other two, thus enabling us to obtain directly the average revolu- 
 tions of all the shafts in a four-shaft installation. 
 
 A differential counter has also been devised for giving the mean 
 revolutions of three shafts, for use on triple-screw ships. 
 
 Tachometers. These instruments are employed to indicate 
 directly the number of revolutions a shaft is making. Fig. 41 illus- 
 trates the general principle on which nearly all such instruments 
 are constructed. The shaft a gives rotation to a set of flying balls, 
 c i? c i, c s> c- With increased velocity the balls swing out around the 
 center b, against the restraining force of a spring, and in so doing 
 
74 EXPERIMENTAL ENGINEERING 
 
 pull down the links d and move the sector h, operating an index, 
 which is mounted on the axis of the wheel i. The index moves over 
 a dial graduated in revolutions per minute. The wheel i is a spur 
 wheel meshing with a small pinion, carrying a balance wheel at gf. 
 
 FIG. 41. 
 
 This prevents fluctuation of the index. A stop screw s, permits 
 adjustment of the index. 
 
 Tachometers of this type have been fitted on fast running shaft- 
 ing of turbine-driven vessels, where they are very useful in indi 
 eating the approximate number of revolutions. They cannot, as a 
 rule, be depended upon to indicate with exactness. 
 
INSTRUMENTS FOR EECORDING EXPERIMENTAL DATA 75 
 
 Portable Tachometers. An instrument of this type is shown in 
 Fig. 42. There is a set of change wheels operated by the milled 
 head at the right of the figure,, which serve to change the range of 
 the instrument. The setting is indicated automatically by a movable 
 plate as shown. 
 
 FIG. 42. 
 
 The Hutchison Marine Tachometer. 
 
 The Transmitter. Clamped around the propeller shaft P, Fig. 
 43, is split sprocket A. Rotation is imparted to driven sprocket B 
 by the silent chain C. Sprocket B is loosely mounted on shaft D. 
 Two oppositely coiled flat spiral springs E-E' transmit the rotation 
 of B to flywheel F and shaft D, one end of each spring being at- 
 tached to the sprocket B, and the other end to flywheel F. Any 
 irregularity of rotation of B, caused by variations in the angular 
 velocity of shaft 'P, is smoothed out by E-E', imparting to F and D 
 a constant resultant speed. The springs are protected against 
 breakage from sudden reversal of P by the radial arm G engaging a 
 pin H attached to flywheel. 
 
76 
 
 EXPERIMENTAL ENGINEERING 
 
 FIG. 43. 
 
INSTRUMENTS FOR BECORDING EXPERIMENTAL DATA 77 
 
 On the inside face of flywheel, at the end opposite from that 
 occupied by E-E r are cut gear teeth. These engage pinions which 
 actuate the shafts of magnetos L and M, supplying alternating 
 current. 
 
 One pinion is keyed to inductor shaft of its magneto M. 
 
 The other pinion is not keyed, but is so mounted that when the 
 direction of rotation of main shaft P is AHEAD the inductor of 
 magneto L is in the exact rotative relation to its armature and pole 
 shoes as that of magneto M . The current from L is therefore in 
 absolute phase with that from M. But when P is reversed in rota- 
 tion, the pinion rotates idly on shaft of magneto L until it has 
 traveled one hundred and eighty degrees before rotating same. 
 This causes the inductor of L to assume an exactly opposite relation 
 to its armature and pole pieces as obtains at the same instant in M, 
 and hence the current from L is one hundred and eighty degrees 
 electrically out of phase with M. 
 
 We have, therefore, two wires from L and two from M, running 
 to the indicators. At AHEAD these circuits are in phase. When P 
 rotates ASTERN, one is one hundred and eighty degrees electrically 
 out of phase with the other. 
 
 The Receiver. The indicating instrument, shown in Fig. 44, has 
 two coils a moving coil to which the pointer is attached, and a 
 fixed or field coil. The moving coil is connected electrically to one 
 of the magnetos, the fixed coil to the other. When the two magnetos 
 are in phase, the pointer is deflected to the RIGHT, indicating 
 R. P. M. AHEAD. When they are out of phase, the pointer is de- 
 flected to the LEFT, indicating R. P. M. ASTERN. The faster the 
 shaft P, Fig. 43, turns in either direction, the higher the voltage 
 generated, and the greater the deflection of the pointer calibrated to 
 conform thereto. 
 
 The Hopkins Electric Tachometer consists of a small direct-cur- 
 rent magneto generator and an indicating electrical voltmeter of 
 high grade. The two parts of the system are connected by a two- 
 wire insulated cable. 
 
 The principle of operation of this apparatus depends on the fact 
 that when a system of coils is rotated within a permanent magnetic 
 field, a potential is generated in direct proportion to the speed of 
 6 
 
78 EXPERIMENTAL ENGINEERING 
 
 rotation of the moving coils. It is therefore possible to calibrate 
 the voltmeter in terms of the speed, which in this case is repre- 
 sented by revolutions per minute. 
 
 Speed recorders working on the principle of the Hutchison and 
 Hopkins instruments have been proposed many times, but have not 
 been very successful, due principally to the imperfect mechanical 
 details. The indications at the voltmeter cannot be true if there 
 
 FIG. 44. 
 
 is any variation in the resistance of the system. The Hutchison in- 
 strument overcomes these difficulties in large measure by using 
 alternating current and thus avoiding the use of a commutator. In 
 the Hopkins instrument it is claimed that such difficulties have been 
 eliminated, by making the commutator bars on the magneto of pure 
 platinum and the brushes of 20 karat gold. With this construction 
 there are no oxidation changes or insulating salts formed. 
 
 McNab Marine Register and Indicator. A small air pump, with 
 piston 2 inches in diameter and 3 inches stroke, designated by the 
 inventor as the Agitator, is connected to the valve gear, or other 
 
INSTRUMENTS FOR EECORDING EXPERIMENTAL DATA 79 
 
 moving part of the engine capable of giving a small reciprocating 
 motion. This is connected by half -inch piping with one or more 
 Indicators and Registers in various locations about the ship as may 
 be desired. The indicator is a rod connected with a small piston 
 which is pulsated with each stroke of the engine. The register is a 
 counter which is worked through a ratchet by the indicator. There 
 is a separate register and indicator for ahead and astern, both on 
 the same board. A separate pipe is provided for ahead and astern, 
 with cocks at the agitator, so arranged that they are opened or 
 closed automatically when the engine is reversed. 
 
 The Davison Speed Regulator. This instrument is one of several 
 embodying the same general principles and has for its object the 
 keeping of a certain definite speed of the engine when this is needed 
 while steaming in squadron, standardizing over the measured mile, 
 etc. 
 
 The device consists essentially of the following parts : Eef er to 
 Fig. 45 (II) ; the vertical shaft S", carrying a worm wheel at its 
 lower end and a miter wheel at its upper end ; this small worm wheel 
 T is connected by means of a worm 8' shaft and bevel gear to some 
 part of the main engine in such a manner that shaft 8" makes one 
 revolution for each revolution of main engine. On this vessel the 
 worm and worm shaft 8' are connected to the vertical shaft of the 
 main engine counter. The miter gear on upper part of vertical shaft 
 8" meshes with another miter gear on horizontal shaft 8'"', this 
 shaft in turn carries a hard bronze friction wheel, or pinion F, and 
 is a little larger in diameter than the miter gear on shaft S f " '. The 
 position of friction pinion F on shaft S f " is determined by setting 
 of the micrometer screw E and micrometer wheel 0. F is made to 
 revolve with shaft 8"' by means of a feather and key. In the in- 
 strument, shown in Fig. 45, the micrometer arm E is graduated by 
 fives from forty-five revolutions to one hundred and twenty revolu- 
 tions, and the micrometer wheel is graduated by tenths from zero 
 revolution to five revolutions, so that the instrument is capable of 
 being set for revolutions from forty-five to one hundred and twenty, 
 by tenths. In other instruments the graduations on E and are 
 arranged to suit the range of speed of the engine for which the in- 
 strument is intended. 
 
80 
 
 EXPERIMENTAL ENGINEERING 
 
 5 PEED 'REGULATOR, 
 
 FIG. 45. 
 
INSTRUMENTS FOR EECORDING EXPERIMENTAL DATA 81 
 
 The locking nut D holds the setting for revolutions in permanent 
 place. There is an arm B which carries on its inner end a small 
 roller whose object is to hold the friction disk K f in constant fric- 
 tional contact with the friction pinion F. The friction disk K is 
 supported top and bottom by ball bearings, and carries with it a 
 velvet-lined pad TL, on which rests a timepiece. In this case the time- 
 piece is a large, ordinary commercial stop watch. The gear con- 
 nections to the engine, or moving part driven from the engine, are 
 such as to cause the timepiece to be carried in a counter-clockwise 
 direction. The whole mechanism is enclosed in a brass casing and 
 looks very much like a small compass binnacle. It is mounted on 
 the floor plate at the working platform in the engine room so that 
 the man on watch at the throttle looks directly down on the time- 
 piece and graduated circle in Fig. 45 (I). 
 
 Eeferring to Fig. 45 the circle on the top of the instrument, which 
 is fixed, as far as the mechanism is concerned, is graduated and 
 marked from left and right of the central point in yards, the scale 
 of marking depending on the number of yards made by the vessel 
 for each revolution of the engine. This distance and consequently 
 the whole number of revolutions corresponding to a total gain or loss 
 of a particular number of yards will vary, depending on the draft 
 of vessel, condition of bottom, state of sea, etc., but curves are fur- 
 nished with the instrument which permit corrections to be easily 
 applied. 
 
 The operation of the revolution indicator is as follows : The time- 
 piece being wound and fitted in its place on the velvet-lined pad 
 the hand will make one revolution every minute. By means of the 
 micrometer wheel and screw the instrument is set for the revolutions 
 ordered, then, as long as the throttle is manipulated so as to main- 
 tain this speed, the hand of the watch virtually stands still and 
 points constantly to the radial arrow shown in Fig. 45 (I). If at 
 any instant the engine room receives word from the bridge that the 
 ship is, say, forty yards behind position, the cap A is shifted with 
 reference to the hand of the watch, in the direction indicated by the 
 circular arrow, marked "gain," the radial arrow would then be 
 forty divisions to the left of the hand of the watch; and since the 
 hand of the watch moves at constant and uniform speed it would be 
 
82 EXPERIMENTAL ENGINEERING 
 
 necessary to impart a faster counter-clockwise motion to the time- 
 piece than it previously had, in order to bring the hand of the watch 
 into coincidence with the radial arrow. Vice versa, if the ship had 
 been ahead of position by a given amount the top rim would have 
 been shifted in the direction indicated by the curved arrow, marked 
 " lose, 7 ' and, in this case, the counter-clockwise motion of the instru- 
 ment would necessarily have to be reduced by throttling down until 
 the hand of the timepiece caught up again to the radial arrow. It 
 is the work of only an instant to set the instrument for any given 
 speed, and to lock it in place by means of locknut D, and the top 
 rim A can also be shifted instantly. 
 
 Special Engine Counters. For taking the number of revolutions 
 made by an engine during a short time, as for example, on the 
 measured mile, it becomes necessary to adopt a device for taking 
 the exact counter reading in fractional parts of a revolution at the 
 instant of making an observation. 
 
 The Taylor Counter. This instrument substitutes for the usual 
 numbered wheels in the ordinary type of counter, a set of printing 
 wheels. By pressing an electric contact at the instant of making the 
 observation, a printed record is made. This instrument gives ex- 
 cellent results in slow moving engines, but with high speed turbines 
 the records are frequently blurred and it becomes necessary to check 
 with other instruments. 
 
 The Bailey Counter. This consists of a train of wheels the shafts 
 of which carry indexes moving over dials. The wheels in each pair 
 are in the ratio 10 to 1 and the instrument when connected with the 
 shaft registers continuously up to 10,000,000 revolutions. The long 
 pointer travels around the full face of the instrument- once in 10 
 revolutions and its dial is graduated in tenths of a revolution. This 
 pointer and the two pointers in the hundreds and thousands places 
 have loose pointers which travel with them. On making an electric 
 contact at the instant of taking the observation these loose pointers 
 are held and the exact reading can be taken with certainty. The 
 loose pointers are then reset ready for the next observation. 
 
 This counter gives excellent results, but there is no permanent 
 record and for careful work its readings should be checked by in- 
 dependent observers. 
 
INSTRUMENTS FOR EECORDING EXPERIMENTAL DATA 83 
 
 As installed for trial trips each of these special counters is con- 
 nected to the shaft independently. 
 
 The Precision Tacograph. This instrument, shown in Fig. 46, 
 is a recording tachometer and is employed to record accurately slight 
 
 FIG. 46. 
 
 fluctuations in the revolutions of a shaft. A pair of weights, not 
 shown in the figure, are revolved about the shaft of the instrument 
 and tend to fly out against the force of restraining springs. Levers 
 connect the weights with a pen under which the tape shown in the 
 figure is caused to move by clockwork. The paper is perforated at 
 
84 EXPERIMENTAL ENGINEERING 
 
 one edge to suit the spacing of the teeth on a wheel so that its 
 movement is a correct function of the time. Lines on the paper 
 are drawn to represent variations in speed measured in per cent 
 of the standard. 
 
 This instrument makes it possible to investigate irregularities in 
 speed of motors and machines, and in the case of reciprocating 
 engines permits the measurement of irregularities in speed during a 
 single revolution, thus enabling us to investigate directly the effect 
 of the reciprocating, parts. 
 
 The same instrument may be used for a wide variation in speed 
 by employing a set of changeable cone pulleys. 
 
CHAPTER IV. 
 MEASUREMENT OF THE QUALITY OF STEAM. 
 
 The following terms in thermodynamics are explained in other 
 text-books, but as their meaning must be clearly understood in con- 
 nection with the study of the calorimetry of steam they will be 
 briefly reviewed here : 
 
 Heat as a term employed in thermodynamics must be understood 
 as that form of energy which when applied produces the sensation 
 commonly known as heat. 
 
 Quantity of Heat. Heat and work being mutually convertible, 
 the quantity of heat that passes to or from a body is capable of defi- 
 nite measurement and is directly proportionate to the amount of 
 work done in causing such heat to pass. 
 
 Temperature may be defined as heat pressure or heat intensity. 
 It is a measure of that quality which when two bodies are in contact, 
 causes heat to flow from the one of higher to the one of lower tem- 
 perature. In steam engineering research work in the United States, 
 temperatures are usually measured on the Fahrenheit scale, in which 
 at the atmospheric pressure the temperature of melting ice is 
 reckoned at 32 and the temperature of boiling water at 212. 
 
 Absolute Temperature is the temperature measured from the 
 absolute zero. This has been deduced from the known law govern- 
 ing the relation between pressures, temperatures, and volumes, 
 which is PV = c(t + k), where P stands for pressure, V for vol- 
 ume, t for temperature as measured by the thermometer, and c and 
 k are constants. For the Fahrenheit scale the value of k for all 
 gases has been found to be 461 or 493 below the freezing point of 
 water. T, the absolute temperature, then equals + 461, and 
 PV=cT. 
 
 Entropy of a fluid at any temperature is that quality by which 
 the relation between temperature and quantity of heat is measured. 
 It is best understood by study of the temperature-entropy diagram. 
 (See page 102.) 
 
86 EXPERIMENTAL ENGINEERING 
 
 Boiling Point is the temperature at which evaporation takes place 
 and depends chiefly on the pressure. It is raised slightly by the 
 addition of foreign matter to the water, in solution. 
 
 Thermal Unit. A British Thermal Unit (abbreviated B. T. U.) 
 is the quantity of heat required to raise one pound of water from 
 62 to 63 Fah. This is the unit adopted by Professor Peabody in 
 constructing his tables, on account of the ease with which its value 
 may be verified. The standard is usually defined as the quantity 
 required to raise water through 1 from its freezing point. It has 
 not been practicable to demonstrate this value experimentally. 
 
 Specific Heat of a substance is the number of B. T. U. required 
 to raise a pound of the substance 1 Fah. The specific heat of water 
 is not quite constant at varying temperatures. Its value at 32 is 
 1.0072, decreasing to a minimum value of 0.9948 at about 77, and 
 again increasing gradually to 1.046 at 395. At 60 and at 104 
 its value is unity. The specific heat of saturated steam is 0.478. 
 This value is also commonly used for the specific heat of super- 
 heated steam, but the true value of the specific heat of superheated 
 steam varies slightly with the temperature and pressure. 
 
 Sensible Heat at any temperature or pressure is the number of 
 B. T. U. required to raise one pound of water from the freezing 
 point to the boiling point. It is nearly but not quite equal to 
 t 32, where t is the Fahrenheit temperature at which evaporation 
 takes place. The sensible heat will be denoted by S. 
 
 Latent Heat is the number of B. T. U. required to convert one 
 pound of water at any temperature and pressure into steam at the 
 same pressure and temperature. It will be denoted by L. 
 
 Total Heat is the sum of the sensible heat and the latent heat and 
 will be denoted by H. 
 
 Joule's Equivalent. One B. T. U. has been found by experiment 
 to be equivalent to 778 foot pounds of mechanical work, which is 
 known as Joule's Equivalent. 
 
 Saturated Steam. When water at any pressure has been heated 
 to the boiling point corresponding to that pressure and converted 
 wholly into steam without further addition of heat it is said to be 
 saturated. 
 
 Wet Steam is steam containing a certain amount of moisture sus- 
 pended in it. 
 
MEASUREMENT OF THE QUALITY OF STEAM 87 
 
 Superheated Steam is steam to which heat has been added after 
 reaching the point of saturation. 
 
 Quality of Steam. In practice steam is seldom produced in 
 the perfectly dry saturated condition, but contains a certain amount 
 of moisture in the form of mist, which is suspended in it and carried 
 along with it. It is necessary to determine the amount of this moist- 
 ure during any steam experiments, though its complement, the 
 quality of the steam is usually spoken of. By quality we mean that 
 fraction of the whole that is pure, dry, saturated steam. It is usu- 
 ally denoted by x, and the fraction of the whole that is moisture or 
 hot water merely, is (1 x). The percentage of moisture is 
 
 100(1-2;). 
 
 Steam Calorimeters. 
 
 The Steam Calorimeter is an instrument for determining the 
 amount of moisture in steam and from that the quality. There are 
 three general classes in common use, as follows : 
 
 (1) Superheating Calorimeters. 
 
 (2) Separating Calorimeters. 
 
 (3) Condensing Calorimeters. 
 
 Fig. 47 shows Prof. Carpenter's Throttling Calorimeter, which 
 belongs to the first of these types. It consists of a small vessel A f 
 to which steam is supplied through a stop or throttle valve and a 
 tapering or converging orifice B, and contains in its center a very 
 deep cup, into which a thermometer can be inserted for obtaining 
 the temperature of the steam in the calorimeter. A cock C connects 
 with a mercury-filled manometer for measuring the pressure of 
 steam in the calorimeter. The exhaust steam is discharged from the 
 lower part of the calorimeter and may be permitted to escape freely. 
 
 The principle of operation follows from the superheating of steam 
 when it is allowed to expand freely without doing work. The whole 
 amount of heat contained in the steam must remain constant, but 
 the total heat of vaporization being greater at a higher than at a 
 lower pressure, the difference goes to superheat the steam of lower 
 pressure. 
 
88 EXPERIMENTAL ENGINEERING 
 
 Let p-L boiler pressure, absolute. 
 
 p 2 pressure in calorimeter, absolute. 
 t c temperature in calorimeter. 
 
 LI and $! = latent heat and sensible heat corresponding to p lm 
 H 2 and t 2 = total heat and temperature corresponding to p 2 . 
 s = specific heat of steam. 
 x = quality of steam required. 
 
 FIG. 47. 
 Then the heat in a pound of steam flowing to the orifice will be 
 
 and the heat in a pound of steam in the calorimeter after passing 
 through the orifice will be 
 
 Assuming that no heat is lost or converted into work these two 
 expressions must be equal, from which 
 
MEASUREMENT OF THE QUALITY OF STEAM 89 
 
 Graphical Solution for Throttling Calorimeter Determinations. 
 
 In the practical use of this instrument it is customary to exhaust at 
 atmospheric pressure, so that the normal temperature in the calori- 
 meter is the boiling point at atmospheric pressure. H 2 then be- 
 comes 1147, from steam tables, and t 2 becomes 212. Between the 
 temperatures at which the instrument will be ordinarily used it will 
 be found that the values of 8 l and L { can be represented by 
 
 ^ = 1.017^-32, and 
 
 ^ = 1115-0.705^. 
 
 The expression for x as found above, then becomes by substitution 
 
 _ .48* c -1.017< 1 /ov 
 
 1115-0.705^ 
 
 This equation may be written 
 
 , _ - 1 (1.017-0.705aO /v 
 
 ' 0.48 
 
 If x be given a constant value, this equation represents a straight 
 line of which t and t c are the coordinates. A diagram is constructed 
 by giving x a series of values approaching unity and drawing the 
 series of lines that will be represented. This is shown in Fig. 48, in 
 which t is measured on the vertical and t c on the horizontal axis. 
 x may be thus found graphically, knowing the values of ^ and t c . 
 
 In order to find ^ it is necessary to consult a steam table and 
 pick out the value corresponding to p^. To avoid this necessity a 
 diagram is constructed, as shown in Fig. 49, where the ordinates 
 represent initial absolute pressures instead of temperatures. Simi- 
 lar lines are drawn to represent x and the diagram is used in a man- 
 ner similar to the preceding, but it will be noted that, since the 
 pressure does not vary directly with the temperature, the values of 
 x are now represented by a series of curves, instead of straight lines. 
 
 Calibration Method. The throttling calorimeter is frequently 
 used to determine the quality of steam at a constant pressure, as in 
 boiler tests. In such cases if the discharge valve be closed so that 
 the only outlet is through the calorimeter, dry saturated steam will 
 flow into it. Let r=the corresponding temperature in the calorim- 
 eter. Equation (1) then becomes 
 
 L, 
 

 IN CALORIMET 
 Fio. 48. 
 
* 
 
 
 
 
 I::::::::;:::::::** 
 
 if H-RQT T t!liQ:LC;A'- J O 1 R!l'l^ FE! 
 
 SHE 
 
 FIG. 49. 
 
92 
 
 EXPERIMENTAL ENGINEERING 
 
 During the test, if t be the observed temperature in the calorimeter, 
 the boiler pressure being the same as before, we will have 
 
 and the percentage of moisture = 
 
 l_z- s ( y -0 _ OAS(T-t) 
 
 Log of Test. 
 
 j 191. 
 
 TEST WITH THROTTLING CALORIMETER. 
 Steam from 
 
 Duration 
 of 
 
 Barom- 
 eter. . 
 
 Boiler Pressure 
 
 Calorimeter Pressure 
 
 Calorimeter 
 Temp. 
 
 Quality 
 Steam 
 
 Gage 
 
 Abs. 
 
 Gag-e 
 
 Abs. 
 
 
 
 
 
 
 
 
 
 Limitations of the Throttling Calorimeter. If the percentage of 
 moisture is so great that the steam, expanding into the calorimeter, 
 does not become completely dried, the instrument is of no value. 
 The theoretical limit is found for any initial temperature ^ by put- 
 ting t c equal to t 2 in equations (1) and (2), pages 88 and 89. The 
 practical limit is somewhat lower and varies from 2.3% of moisture 
 at 50 pounds boiler pressure to about 1% at 300 pounds. 
 
 For a small percentage of moisture, the throttling calorimeter is 
 one of the handiest and most accurate forms of apparatus. 
 
 Professor Thomas' Superheating Calorimeter. This instrument 
 is the invention of Prof. Carl C. Thomas, of the University of Wis- 
 consin. The general arrangement is shown in Fig. 50, with a sec- 
 tion through the calorimeter proper in Fig. 51. 
 
 Steam is admitted and after passing through the instrument, 
 passes out to a condenser or to the atmosphere. The admission 
 valve, if there is one, is opened wide to permit free access of steam. 
 A thermometer measures the temperature of the steam in the in- 
 strument. 
 
MEASUREMENT OF THE QUALITY OP STEAM 
 
 93 
 
 As steam is passed through the instrument it is dried and then 
 superheated by an electric heater placed within its walls. The con- 
 dition of dryness is indicated by an immediate rise of temperature, 
 as shown by the thermometer, if more than the requisite amount of 
 electrical energy is supplied. 
 
 For convenience let E t represent the number of watts necessary 
 to exactly dry the quantity of steam that passes through the instru- 
 
 FIG. 50. 
 
 ment. After noting E l the steam is superheated by additional watts 
 E 2 up to a temperature i, above saturation, of 20, 30, 100, or some 
 other convenient number of degrees of superheat. This operation 
 is for the purpose of determining the rate of flow or the weight of 
 dry steam W l passing through the instrument when at the saturation 
 point. The weight W 2 passing through after superheating will be 
 less than W by a percentage which may, for a given pressure, be 
 represented by a constant (7; the specific gravity of superheated 
 steam being smaller than that of saturated steam at same pressure. 
 7 
 
EXPERIMENTAL ENGINEERING 
 
 FIG. 51. 
 
MEASUREMENT OF THE QUALITY OF STEAM 95 
 
 This constant has been determined by practical tests with the in- 
 strument, as has also the number of watts 8 necessary to superheat 
 a pound of steam to t at different pressures. The value 8 may be 
 used instead of the specific heat of steam, and has the advantage of 
 allowing for radiation losses. The quality of steam may be ob- 
 tained, either by the use of curves supplied with the instrument, or 
 by calculation, using the specific heat of steam at varying pressures. 
 By using the curves, possible errors due to uncertainty as to the 
 value of the specific heat are eliminated. 
 
 Let W 1 weight of dry steam passing per hour. 
 
 Then, since one watt = .0009477 B. T. U. per sec., or 3,412 
 B. T. U. per hour, we have E t watts ^E^x 3.412 B. T. IPs per 
 hour = the energy required to evaporate the water in W t pounds of 
 wet steam. 
 
 O A1 O TJI 
 
 Let H x = -^-TTT =the energy thus required, per pound of 
 
 steam, in B. T. IPs. 
 
 Let W 2 = CW l = weight of superheated steam passing per hour, 
 then E 2 watts = CW 1 8, where $ = the number of watts required to 
 
 ET 
 
 superheat one pound of steam through t, and W l =j T ^-. Substi- 
 tuting this value of W^ in the above expression for H x , we have 
 
 rr _ 3.412^ X 08 
 rL x - ^ - . 
 
 ^2 
 
 Since and S are constants for any given pressure and degree of 
 superheat, 3A12CS may be written as a constant, K, and values of 
 this constant are given for varying pressures and degrees of super- 
 heat, by a set of curves plotted from experimentally obtained data. 
 See Fig. 52. 
 
 The expression for H x then becomes 
 
 If H v represents the heat of vaporization of dry steam, obtained 
 from the steam tables for the pressure indicated by the original 
 temperature in the calorimeter, the quality of steam passing through 
 the calorimeter is 
 
 H-H 
 
MEASUREMENT OF THE QUALITY OF STEAM 97 
 
 It is to be noted that the constant K is independent of the weight 
 of steam flowing through the calorimeter during superheating, al- 
 though for clearness consideration of this weight has been included 
 in the above explanation. 
 
 Operation of the Apparatus. The calorimeter is attached in a 
 vertical position to the source of steam supply. This may be a 
 sampling tube of ordinary form, extending into a steam pipe. A 
 special form of sampling tube is supplied with the instrument, de- 
 signed for use in collecting samples of steam in the various passages 
 of a Parsons steam turbine. The same tube may be used for col- 
 lecting samples of steam from different portions of any steam pas- 
 sage without disconnecting the instrument. 
 
 The box, shown in Fig. 51, is used as a water rheostat for regu- 
 lating the amount of electrical energy supplied to the heating coils. 
 It is filled about half full of fresh water, to which one or two cups- 
 full of common salt are added. The inverted cone is lowered or 
 raised, to bring more or less of its surface in contact with the water, 
 thus varying the resistance and permitting a perfect adjustment of 
 the amount of electrical energy supplied. A voltmeter and ammeter 
 of ordinary form are fitted between resistance box and instrument. 
 Current is supplied from the lighting circuit at about 110 volts. 
 About ten amperes will be found sufficient for experiments. 
 
 Sufficient electrical energy is at first supplied not only to dry the 
 steam, but to superheat it to some convenient temperature. Let the 
 watts introduced be denoted by Et. Raise the cone of rheostat 
 slowly until the thermometer indicates that superheating no longer 
 is taking place. The steam will then be at the point of saturation. 
 Let EI denote the energy then being supplied, which will be the 
 amount required to just dry the steam, and E 2 , the watts required 
 to superheat through t=Et E 1 . 
 
 From the curves, Fig. 52, select the value of the coefficient K, 
 corresponding to the degree to which the steam was superheated, and 
 to the original pressure in the calorimeter, then calculate the value 
 
 W 
 
 of H x /rom the equation H x = Kx -- . 
 
 Find the quality of the steam from the second set of curves, Fig. 
 
 TT TT 
 
 53, representing the equation x *= - . The quality may of 
 
 H v 
 
 course be calculated from the equation if desired. 
 
MEASUREMENT OF THE QUALITY OF STEAM 99 
 
 Carpenter's Improved Separating Calorimeter. This instrument, 
 shown in Fig. 54, contains two vessels, one inside the other. The 
 outer vessel surrounds the inner, leaving a space which serves as a 
 steam jacket. The inner vessel is provided with a glass water gage 
 10 and scale 12. 
 
 The steam under test is admitted through the pipe 6. Striking 
 the bottom of a perforated cup 14, it is deflected nearly 180 degrees. 
 The water is thrown off and passes through the perforations into the 
 inner vessel 3, where the amount is indicated by the graduated scale 
 12 on gage glass. The steam passes across the top of perforated 
 cup and into outside chamber, from which it is discharged through 
 a small orifice 8, of known area, in the bottom part. 
 
 The orifice 8 is so small in comparison with any section of the 
 steam pipe or throttle valve that there is no sensible reduction in 
 pressure by passing through the calorimeter. The pressure in outer 
 chamber being the same as in the interior, it has the same tempera- 
 ture and consequently there is no loss by radiation from the interior 
 surface except what takes place from the exposed surface of the 
 gage glass. 
 
 It has been demonstrated that the flow of steam through a small 
 orifice is proportional to the absolute steam pressure, until the pres- 
 sure against which the flow takes place equals or exceeds 0.6 of that 
 of the vessel under pressure. A special form of steam gage is placed 
 on the outside chamber, the inner circle of which shows the gage 
 pressure and the outer circle shows the number of pounds of steam 
 that will escape through the orifice in 10 minutes of time. 
 
 The graduations on the scale 12 show the weight of water in 
 pounds and hundredths that is contained in the inner vessel. The 
 instrument is operated at a constant pressure for 10 minutes and w, 
 the weight of water collected, is read off; also W, the weight of dry 
 steam that escapes through the orifice, is read off on the steam gage. 
 Then 
 
 and t - ^ w 
 
 W+w' 
 
 An earlier form of this calorimeter is similar to the foregoing, 
 except that the escaping dry steam is condensed and weighed. 
 
100 
 
 EXPERIMENTAL ENGINEERING 
 
 12 - 
 
 Fio. 54. 
 
MEASUREMENT OF THE QUALITY OF STEAM 101 
 
 The separating calorimeter is accurate and applicable in all cases 
 where the steam contains moisture. It is not applicable in cases 
 where the steam is superheated. 
 
 The Barrel Calorimeter. This is of the condensing kind and is 
 improvised for steam tests in cases where other more accurate ap- 
 paratus cannot be obtained. 
 
 A heavy barrel, such as an oil barrel, of about 50 gallons' capacity, 
 is employed. This is placed on a platform scale for weighing. A 
 hose is provided for filling it with cold fresh water and a cock in the 
 bottom serves for draining this off. A pipe connection is led directly 
 over the barrel to bring the steam that is to be tested. This ends in 
 a short piece of rubber hose leading straight down into and nearly 
 to the bottom of the barrel. A stop valve is placed just over the 
 barrel, with which to turn on or shut off the steam. 
 
 The barrel is weighed empty, then filled about three-fourths full 
 of cold water, again weighed, and the temperature of the water 
 taken. The hose is removed and steam is blown through to heat it. 
 Steam is then shut off, and the hose inserted in the water, after 
 which steam is blown in and the water carefully stirred until the 
 temperature rises to about 110 Fah. The weight is again taken to 
 determine the weight of steam blown in. The pressure of steam is 
 noted from the gage and the corresponding values of 8 and L taken 
 from the steam tables. Then if W be the weight in pounds of cold 
 water in barrel at temperature t 19 and w pounds of wet steam of 
 quality x be blown in, bringing the weight of water up to W + w 
 at temperature 2 , and if no heat is lost during the operation, we 
 will have, reckoning from the freezing point, 
 
 Heat contained in cold water = W(t l 32). 
 
 Heat added in w pounds of wet steam = w(xL + S), and 
 
 Heat contained in water after blowing in steam 
 
 = (W+w)(t 2 -32). 
 From which 
 
 This method will not give the exact value of x, since it is im- 
 possible to prevent losses by radiation, and for accuracy a correc- 
 tion must be applied for the heat equivalent of the barrel. In prac- 
 
102 EXPERIMENTAL ENGINEERING 
 
 tice this is difficult of determination and the usual method is to neg- 
 lect it, but to allow for radiation approximately by first filling the 
 barrel and heating the water to about 20 Fah., above the final tem- 
 perature t 2 of the experiment. This water is drained off and the 
 barrel immediately refilled for the purpose of the experiment. 
 
 Care in Selecting Sample of Steam. With any calorimeter, care 
 should be taken in so placing it that a fair sample of the steam 
 is taken. Experiments have proven that steam taken from the cen- 
 ter of a large pipe contains less moisture than if taken near the 
 walls. For throttling calorimeters, the American Society of Me- 
 chanical Engineers recommend that the calorimeter pipe be ^ inch 
 in size, that it extend into the steam pipe to within J inch of the 
 opposite wall, that the inner end be plugged, and that it be provided 
 with not less than twenty J-inch holes, distributed along and around 
 its length, no hole being closer than ^ inch to the inner end. To 
 obtain satisfactory results, the same care should be exercised in 
 sampling the steam for test in any calorimeter. 
 
 The Temperature-Entropy Diagram. 
 
 The Temperature-Entropy Diagram is used by engineers for in- 
 vestigating the work done and the losses that occur in heat engines, 
 for which purpose it is of great practical value. In the indicator 
 diagram we have work or energy shown by an area, ordinates rep- 
 resenting force, and abscissas representing distance. In the tem- 
 perature-entropy diagram, an area represents heat, an ordinate rep- 
 resents absolute temperature, and abscissae represent entropy. 
 
 If a body takes in, or rejects, a quantity of heat dQ, at the abso- 
 lute temperature T, the entropy of the body is increased, or de- 
 creased, such increase or decrease of entropy being ^- . Let 
 
 entropy be represented by the symbol <, then the change of entropy 
 consequent upon the addition, or subtraction, of the quantity of heat 
 dQ, at the absolute temperature T, is expressed by 
 
 or with proper limits, 
 
MEASUREMENT OF THE QUALITY OF STEAM 
 
 103 
 
 If we now lay off the values of cf> as abscissae and the absolute tem- 
 peratures as ordinates in a diagram, we can plot the condition of a 
 body as defined by its entropy and absolute temperature. In Fig. 
 55, suppose p a point so defined, being the increase of entropy due 
 
 t 
 
 T 
 
 i 
 
 
 
 P 
 
 A 
 
 D 
 
 B 
 C 
 
 FIG. 55. 
 
 to the addition of a quantity of heat Q, and T the absolute tempera- 
 ture at which it was added. The area ABCD ^> x T, but this, from 
 equation (2) is equal to Q, the heat added, and brings us at once to 
 the principle of the temperature-entropy diagram. 
 
 In dealing with entropy, as in dealing with total heat, an arbitrary 
 point is chosen, entropy being reckoned from that point as a zero, 
 and the entropy of the substance for every other state will have a 
 value which is perfectly definite and may be calculated. In reckon- 
 ing entropy, the condition of water at 32 Fah. is taken as the zero. 
 
 FIG. 56. 
 
 The Diagram for Water. In Fig. 56, let OT and 0$ be the 
 axes for the measurement of absolute temperatures and entropy. 
 Let To be the absolute temperature corresponding to 32 Fah., then 
 < = at this point, as we have assumed the condition of water at 
 32 Fah. to be the zero of entropy. 
 
104 EXPERIMENTAL ENGINEERING 
 
 Now take one pound of water at 32 and add a small quantity 
 of heat to it. Then, with reference to the diagram, there are two 
 distinct changes; first, the temperature is raised, and second, the 
 entropy is increased. In Fig. 56 the increase in entropy is repre- 
 sented by ab, and the rise in temperature by Ic, so that the new con- 
 dition of the water on the diagram is represented by the point c. 
 Now let a little more heat be added and the temperature and entropy 
 still further increased, as by cd and de, so that the new condition is 
 now represented by the point e. We can at once see that when the 
 additions of heat are infinitely small and continuous, i. e., when 
 heat is added proportional to the rise in temperature, the points 
 representing the condition of the water at each instant will, when 
 plotted on the diagram, lie on a curve aA. 
 
 Let B be a point on the curve whose absolute temperature is T B . 
 Then the area OaBC represents the quantity of heat supplied be- 
 tween the limits of temperature T and T B , and if a is the heat 
 supplied for each degree of rise in temperature, i. e., the mean spe- 
 
 ( T B 
 
 cific heat of water, then the area OaBC = a dT. The area of 
 
 J T o 
 
 any element, as ef, is Td<j>, and the whole area OaBC is equal also to 
 j B Td<f>, and we have 
 
 [ B dT, or 
 
 or finally 
 
 ^L 1 * B / O \ 
 
 Since the point B may be any point on the curve, and the tem- 
 perature T } the temperature at any such point, we may drop the 
 subscript B and write for the entropy of water at any temperature 
 
 T ' T (4) 
 
 where T is the absolute temperature of the water, and T is the 
 absolute temperature of the point taken as the zero of entropy, in 
 this case the absolute temperature corresponding to 32 Fah. 
 
MEASUREMENT OF THE QUALITY OF STEAM 105 
 
 The Entropy of Steam. Again referring to Fig. 56,, suppose that 
 at B the pound of water is completely evaporated, and dry steam 
 formed having the same temperature as the water from which it was 
 formed. Remembering that if heat is added at constant tempera- 
 ture, the increase in entropy is proportional to the quantity of heat 
 added, we see that the 
 
 Increase of entropy from B= ,, , . H eat added m ( 5) 
 
 Absolute temp, at which added 
 
 The change is an isothermal one as the temperature has not 
 changed and the increase of entropy BB' is laid off from B on the 
 horizontal line corresponding to the absolute temperature of B. The 
 distance of B' from the axis of OT measures the value of the en- 
 tropy of one pound of dry steam at the absolute temperature T^. 
 Referring to equation (5) the heat added is the latent heat of steam 
 at the absolute temperature TB, and calling this L B , we have 
 
 BB'=?-. (6) 
 
 * B 
 
 Remembering that B is any point, we have, dropping the sub- 
 scripts, for dry steam at any absolute temperature T, 
 
 * = .log^ + 4- ' (7) 
 
 where L is the latent heat corresponding to that temperature and 
 a is the mean specific heat of water at that temperature. Plotting 
 the curve as given by equation (7) we get A' a', Fig. 56. Comparing 
 equations (?) and (4), we see that the entropy curve for dry steam 
 is formed by laying off, from each point on the curve for water, the 
 
 value of -= , and since the latent heat of steam decreases as the 
 
 temperature rises, the value of ^ will decrease as we ascend the 
 
 scale of temperature and the curves will approach each other. 
 
 Mixture of Steam and Water. Again referring to the condi- 
 tion of the water at the point B, Fig. 56, suppose that only a frac- 
 tion, x, is evaporated. We would have a mixture of steam and water, 
 and the heat added at the point B would be xL, and the increase of 
 entropy would be BB", less than BB'. It will be readily seen that 
 the entropy in this case is given by the equation 
 
106 EXPERIMENTAL ENGINEERING 
 
 It is evident that this is the general equation, for if we make 
 x=0, we get equation (4) and if we make x=l, we get equation 
 (7) . Any mixture between the limits of all water and all steam can 
 be represented by (8) by giving the proper value to x, which is the 
 dryness fraction, or quality of the steam. 
 
 Isothermal Lines on Diagram. Isothermal changes are made at 
 constant temperature, and are plotted on the diagram by straight- 
 lines parallel to the line 0, from which absolute temperatures are 
 measured. Such a change occurs when steam at boiler pressure is 
 being admitted to the cylinder; also, when the exhaust is being 
 pushed out at the pressure corresponding to the vacuum in 
 condenser. 
 
 Adiabatic Lines on Diagram. When no heat is received or given 
 out during an operation, it is called an adiabatic operation. Ex- 
 pansion and compression occurring in a cylinder without gain or 
 loss of heat are adiabatic and are represented on the diagram by 
 vertical lines. Drawing a line on the diagram to represent such an 
 expansion, it will be noted that the temperature falls and some of 
 the steam is condensed, but the total quantity of heat contained has 
 remained constant. 
 
 If cylinder walls could be constructed of non-conducting material, 
 expansion and compression would in practice be adiabatic. But 
 during expansion, as the temperature falls, heat is received from the 
 cylinder walls and during compression heat is given up to them. 
 
 Temperature-Entropy Diagram for an Ideal Engine. In Fig. 5? 
 A"B"C"D shows the cycle of operations in an ideal engine. The 
 cycle begins at A", the position of the point A" corresponding to 
 the temperature and entropy of a pound of water just ready to 
 form steam. The isothermal A"B" is the admission line. As the 
 steam is generated, it is admitted to the engine and follows the pis- 
 ton to the point of cut-off, represented by B". It then expands with- 
 out gain or loss of heat, in other words adiabatically, to the end of 
 the stroke, where the pressure and corresponding temperature has 
 been reduced to that of the condenser. This expansion line is rep- 
 resented by B"C". On the return stroke the steam is swept com- 
 pletely out of the cylinder into the condenser. This operation is 
 represented by the isothermal G"D. The point D then represents 
 
MEASUREMENT OF THE QUALITY or STEAM 
 
 107 
 
 the temperature and entropy of a pound of water in the condenser. 
 This water is returned to the boiler and there receives heat, bringing 
 it back to the condition at the beginning of the cycle. This part of 
 the operation is represented by DA". 
 
 FIG. 57. 
 
 Temperature-Entropy Diagram for a Real Engine. In the actual 
 engine the conditions are somewhat different. There is a slight loss 
 of pressure and temperature between boiler and throttle valve, and 
 a further loss between throttle and engine. The conditions at 
 throttle and engine are represented by the lines A 'B f and AB, 
 respectively. The cycle of the ideal engine should therefore actually 
 begin at A, instead of at A". 
 
 At the beginning of the stroke there is a certain mass of steam 
 contained in the clearance volume of the cylinder, the amount of 
 this depending upon the point of exhaust closure. The cycle then 
 begins with a mixture of this steam and the water that is received 
 from the boiler. For one pound of the mixture, the cycle in the real 
 engine will therefore not begin at A, but at a, a point representing 
 the same temperature as A, but greater entropy. 
 
108 EXPERIMENTAL ENGINEERING 
 
 In the real engine, the point of cut-off will not fall at B, but at 1), 
 and the real admission line is ab. The heat lost by initial condensa- 
 tion is represented by the area between aB, ah, and the full length 
 of the ordinates through & and B. 
 
 From & to &' the steam in expanding gives up heat to the cylinder 
 walls, and there is a corresponding loss of entropy. At b f the tem- 
 perature of the steam has fallen to the temperature of the walls and 
 in expanding further to c, there is a gain in entropy due to the ab- 
 sorption of heat from this scource. At c the exhaust opens and the 
 steam is permitted to expand into the condenser, with falling tem- 
 perature and partial condensation, the entropy falling to d, which 
 represents the condition at the end of the stroke. From d to d' the 
 steam is being swept into the condenser on the return stroke, this 
 line approaching very closely to an isothermal. The fact that this 
 line does not coincide with DC indicates that there are resistances 
 between the engine and condenser, the temperature being slightly 
 higher in the cylinder than in the condenser. 
 
 At d' the exhaust closes, that portion of the steam received from 
 the boiler at the beginning of the cycle, having been returned to the 
 condenser, and the weight contained in the original clearance vol- 
 ume having been retained in the cylinder. This latter, it will be 
 understood, is in the form of steam, which is then compressed, its 
 temperature rising, and since there is a loss of heat to the walls of 
 the cylinder, there is a slight loss of entropy. That part of the 
 diagram from d r to a then represents the line for one pound of a 
 mixture of steam and water, consisting of that portion which is 
 returned to the condenser, and thence to the boiler, combined with 
 that portion which remains in the cylinder as steam at the point of 
 exhaust closure. The area between d'a and DA represents the loss 
 due to clearance. 
 
 For further information regarding the practical use of the tem- 
 perature-entropy diagram, the student is referred to text-books on 
 that subject. It is of particular value in investigating the ther- 
 modynamic changes that occur in a steam turbine. By tapping into 
 the various passages in a turbine and determining the temperature 
 and quality of the steam, it is possible to construct the diagram, and 
 from it to obtain all necessary data concerning the economical 
 performance. 
 
MEASUREMENT OF THE QUALITY OF STEAM 109 
 
 In studying the design of reciprocating engines, the diagram is 
 used as a means of analyzing indicator diagrams, in order to dis- 
 cover thermodynamic defects that may exist. 
 
 For further study of the subject the student is referred to more 
 extended works treating of it and its application. The following 
 works treating of it and its practical application are in the library 
 of the Head of Department of Marine Engineering and Naval Con- 
 struction : The Energy Chart, Sankey; The Entropy-Temperature 
 Analysis of Steam Engine Efficiencies, Reeve ; The Steam Turbine, 
 Thomas; Experimental Engineering, Pullen; The Steam Engine, 
 Creighton; Steam Tables and Diagrams, Marks and Davis. The 
 last-named volume has diagrams giving the relation between pres- 
 sures, qualities, total heats, volumes and entropies of wet, saturated 
 and superheated steam. They will be found particularly useful in 
 solving problems involving wet or superheated steam. 
 
CHAPTEE V. 
 MEASUREMENT OF THE RATE OF FLOW OF WATER. 
 
 This subject becomes of special importance to the naval engineer 
 in testing the output of pumping machinery, or in the calibration 
 of instruments used in such tests. 
 
 Water Meters. The best known type of water meter is the 
 one in which the water flowing through actuates a train of wheels, 
 thus communicating motion to an index on a dial which registers 
 the volume of water that flows through. Meters of this type are 
 made by numerous manufacturers and are more widely used than 
 any other on account of their convenience. They are fairly accu- 
 rate when used for indicating the flow of water under ordinary 
 conditions, but they must be carefully calibrated and the proper 
 correction applied to the readings. 
 
 The Worthington Water Meter. This instrument is of the 
 above type. It may be simply described as a reversed duplex pump, 
 in which all the water passing through it is used to impart motion 
 to one or the other of the two pump pistons. The water is dis- 
 tributed to the two pump cylinders by plain slide valves, each of 
 which is operated by the movement of the opposite piston. The 
 pistons are connected up so that each movement registers on the 
 counter, and the reading indicates the volume of water that has 
 passed through, independent of any clock mechanism such as is 
 necessary in other types of meters. The reading is shown on a 
 series of dials, the first indicating tenths up to one cubic foot, the 
 second cubic feet up to ten, the third tens of cubic feet up to one 
 hundred, etc. The largest size meter of this kind registers up to 
 one million cubic feet. 
 
 The Worthington Company has recently brought out a second 
 form of meter in which the casing contains a turbine motor, which, 
 while permitting the water to flow through it with but little resist- 
 ance, imparts the necessary movement to the registering mechanism. 
 The method of reading is similar to that above described. 
 
MEASUREMENT OF THE KATE OF FLOW OF WATER 111 
 
 The Keystone Water Meter, as manufactured by the Pittsburg 
 Meter Company, is shown in Fig. 58. There is a flat disc, D, hav- 
 ing a ball and socket bearing, B, which is adapted to oscillate in a 
 chamber, in which each of the upper and lower faces, F and G, 
 approximates in shape the frustum of a cone, the extreme confining 
 wall, W, having a globular shape. The disc has a single slot, pro- 
 jecting radially from the boss, which embraces a fixed metallic 
 
 w. 
 
 FIG. 58. 
 
 diaphragm, II, set within and crosswise on one side of the chamber. 
 The disc is thus prevented from rotating and an axial pin, P, mov- 
 ing around the cone surface, C, causes the disc at all times to take 
 an inclined position, in contact with the upper and lower cone 
 surfaces of the chamber. As the water passes through, the disc 
 oscillates, dividing the chamber into a succession of sub-compart- 
 ments, or measuring spaces. The oscillation gives a movement to 
 P around C, and thus operates the adjoining crank, K, and gives 
 motion to the train of wheels and register on plate L. 
 
112 
 
 EXPERIMENTAL ENGINEERING 
 
 To determine the rate of flow with either of the meters as above 
 described, it is necessary to take readings and divide the difference 
 by the elapsed time between them. 
 
 Meters as regularly constructed do not work well when installed 
 in piping on board ship. This is due to the irregular service and 
 to the effect of hot water in injuring the mechanism or causing it 
 to register incorrectly on account of expansion of certain of the 
 parts. They are generally used for measuring the quantity of water 
 taken on board from outside sources. For this purpose they are 
 very suitable. 
 
 FIG. 59. 
 
 In calibrating a meter, the water should be at approximately the 
 same temperature as in the service for which the correction is to 
 be applied. 
 
 The Venturi Meter. This instrument, though of a different 
 type, is, like the preceding, used to determine the volume of water 
 flowing through a pipe. Its theory is based on the discovery made 
 more than a hundred years ago by an Italian philosopher named 
 Venturi, that when water flows through a contraction in a pipe, 
 the pressure is less in the contracted section than in the larger 
 section on either side. 
 
MEASUREMENT OF THE RATE OF FLOW OF WATER 113 
 
 If at any point in a pipe through which water is flowing a small 
 pipe connection is made and the small pipe is carried up vertically, 
 water will rise freely in the small pipe due to the pressure in the 
 large pipe. Suppose the flow stopped and let H=ihe height to 
 which the water will rise under this condition. Then if /i = the 
 height to which the water will rise when it is flowing with velocity 
 V, the relation between H, Ji, and V is expressed by the formula 
 
 V* = 2g(H-h). 
 
 Fig. 59 represents a longitudinal section through the Venturi 
 tube. An annular chamber, A, communicates through a number 
 of holes with the large section of the tube, and a second similar 
 chamber, B f communicates with the contracted section. These 
 chambers, A and E, have pipes, C and D, connecting them with the 
 gage for registering the rate of flow. 
 
 Let lii an d Ii 2 represent the height in feet of the columns of 
 water that would rise freely in C and D respectively, assuming that 
 C and D are carried high enough and opened to the atmosphere at 
 the upper end. Let a and 2 = the areas of section at A and B, 
 respectively, and v 1 and v 2 the corresponding velocities. Let 
 H = ihe height to which the water would rise when it is not flow- 
 ing, and () = the volume discharged per second, then 
 
 v, 2 = 2g(H-h i ) andv 2 2 = 2g(H 
 Assuming no loss of energy between A and 
 
 , 2 . 
 
 Inserting in this expression the values of v and v 2 , in terms of 
 Q, a-L and a 2) and solving for Q, we have 
 
 Q a 1 o, 2 
 
 /2g(h 1 -h 2 ) 
 V a^-a,* 
 
 This may be called the theoretical discharge. Dividing this ex- 
 pression by ! gives the velocity v and dividing it by a 2 gives the 
 velocity v 2 . Owing to frictional resistances to the flow of water, 
 
114 
 
 EXPERIMENTAL ENGINEERING 
 
 there is an actual loss of energy between A and B, so that this 
 expression must be multiplied by a coefficient, thus 
 
 at'-o,' 
 
 The value of c has been determined by experiment to lie between 
 0.95 and 0.99. 
 
 Fig. 60 shows the recording apparatus on. a 
 form of the Venturi meter, adapted for labor- 
 atory use, in which the difference between h l and 
 h 2 shows directly on a scale that may be gradu- 
 ated to read directly the rate of flow. This 
 reading may be in cubic feet per hour, gallons 
 per minute, or as may be most convenient. 
 
 It should be noted that while this instrument 
 shows the rate of flow, it does not of itself 
 record the quantity of water that has been 
 delivered through the tube. To do this it is 
 necessary to combine it with a recording mech- 
 anism. In some forms of the instrument this 
 mechanism actuates a counter in which the 
 speed of movement is governed by the rate of 
 flow. In other forms the rate of flow is re- 
 corded by a pen in a continuous line on a paper 
 that is moved by a clock mechanism. 
 
 The Venturi meter is generally used for 
 measuring the discharge through large water 
 mains. For this purpose it is said to be more 
 reliable than a self-contained meter, where the 
 parts are liable to wear and thus increase the 
 error of the readings. 
 
 Pitot 's Tube. The Pitot tube is a simple 
 and, in? its improved form, reliable instrument 
 for determining the velocity of a current from 
 indications of its pressure. It was first used for 
 this purpose by Pitot in 1730. In its simplest 
 form it consists of a vertical glass tube with a 
 FIG. 60. right-angle bend at the lower end placed so that 
 
MEASUREMENT OF THE BATE OF FLOW OF WATER 115 
 
 its mouth will point toward the direction of flow. The impulse 
 pressure of the flowing water is balanced by a column of water 
 raised in the other leg of the tube above the general level of the 
 stream. Pitot, for his use, enlarged the mouth of the tube to a 
 funnel or bell shape, as shown in Fig. 61. This, however, causes 
 the liquid to rise a height li, which is about 1^ times the true height 
 or head due to the velocity. This form of entrance is additionally 
 objectionable because it interferes with the current, and the velocity 
 in front of the mouth is not the same as the velocity of the unob- 
 structed stream. 
 
 FIG. 61. Pitot Tube with Bell Mouth. 
 
 Fig. 62 shows the improved form of orifice employed by Darcy 
 and Bazin, in which the tube is drawn out very small with the great 
 benefit of interfering little if any with the natural velocity of the 
 stream; and also that the reduced size of opening tends to check 
 oscillations of the column of water in the tube, instead of encourag- 
 ing them as is the case of a tube provided with a bell-shaped mouth- 
 piece. In the drawn-out form, shown in Fig. 62, Darcy found that 
 when it was placed as shown at P 19 the height h was almost exactly 
 
 when placed as shown at P 2 , having the plane of the orifice parallel 
 to the direction of flow, the water rose practically level with the sur- 
 face of the stream; and when turned with the mouth down stream 
 
116 
 
 EXPERIMENTAL ENGINEERING 
 
 like P 3 , the water sinks to a depth h 1 ', which is nearly the same 
 amount as the rise in case it is headed up stream like P^. If 7i' be 
 the rise in the column P 1 above the surface of the water and h" be 
 the depression in the column P s below the surface, the pressure 
 
 head h will equal ^ nearly. This form of tube is made use of 
 2 
 
 in the pitometer described on page 119. Ti - 
 
 Darcy's Improved Pitot Tube. An objection to employing a 
 simple tube like Eig. 62 is the difficulty of reading the height h 
 direct from the surface of the stream. This is overcome in Darcy's 
 
 FIG. 62. Improved Form of Tube employed by Darcy and Bazin. 
 
 improved form of Pitot tube, by means of which readings can be 
 made above the surface of the stream, or the instrument may be 
 entirely removed from the water for that purpose. 
 
 Fig. 63 illustrates the leading features of the Darcy instrument. 
 Two Pitot tubes, HE and JG, made of copper, have openings at 
 right angles to each other. In making velocity measurements the 
 instrument is so held that the opening in the end of the tube HE 
 is presented against the current while that of JG is downward. 
 The space between the tubes is filled with a solid piece of wood or 
 metal for strengthening the tubes and holding them in place. The 
 upper ends of the tubes are made of glass mounted on a wooden 
 support WW f which in turn is supported by the clamp Q and the 
 guide bracket K which surround the standard BE. Each glass tube 
 
MEASUREMENT or THE KATE OF FLOW OF WATER 117 
 
 C 
 
 FIG. 63. Darcy's Improved Form of Pitot Tube. 
 
118 EXPERIMENTAL ENGINEERING 
 
 is provided with a vernier C for reading the height of the columns 
 on a vertical scale 8. By means of the handle A and hook L the 
 instrument can be raised or lowered to any desired depth and held 
 at the desired elevation by adjustment of the clamp Q, and when 
 thus supported the main body of the instrument acting as a rudder, 
 swings itself around the standard BB to a position parallel with 
 the current and presents the opening of the tube EH up stream. 
 
 Connection of the copper tubes to the glass tubes is made through 
 a two-way cock D which can be operated by cords MM. The glass 
 tubes are connected at their upper ends by a brass fitting which is 
 provided with a stop cock P, the outlet of which is provided with a 
 piece of flexible tubing N with a mouthpiece. Having adjusted 
 the instrument to the desired depth, with the cocks D and P open, 
 water then rises in the tube EH to greater or less extent above the 
 surface of the stream, while it rises in the tube JG to the same level 
 as that of the stream. If then a little air be sucked out of tube N 
 and the cock P closed, water will rise in both glass tubes an amount 
 equal to their respective differences from atmospheric pressure and 
 will stand with the same relative difference between their levels as 
 they first had in the lower part of the instrument before any air was 
 exhausted. The cock D is then closed, preserving the relative height 
 of the columns, and the difference is easily read off with the instru- 
 ment in place, or by removing it from the stream. 
 
 In using this instrument for obtaining the mean velocity of a 
 stream, a number of readings have to be taken to obtain average 
 velocity at any point, as the velocity of a stream varies at different 
 points of its cross section. For determining the mean velocity of 
 a section, the mean of averages of different points of the cross sec- 
 tion must be taken. The mean velocity of a stream is quickly found 
 by one accustomed to using the instrument, once the cross section 
 is established. It is not well suited for measuring very slow ve- 
 locities. A difficulty which should be guarded against is the liability 
 of obtaining too small readings of the column connected with the 
 tube GJ, due to dirt gathering in the opening at G. This defect 
 can be guarded against by occasional examination of the instru- 
 ment before exhausting air by the tube N. Using the instrument 
 in clear water rarely gives trouble, but in all work of importance it 
 
MEASUREMENT OF THE KATE OF FLOW OF WATER 119 
 
 should be calibrated in water of known velocity to obtain its mean 
 variation of readings from the formula, 
 
 Darcy's instrument is used for measuring the rate of flow in an 
 open stream. Its principle, though, can be applied in an instrument 
 for measuring the rate of flow through pipes. This, follows from 
 an examination of Fig. 62. All three of the tubes there shown are 
 subject to the static pressure of the fluid in the pipe. 'P^ in addition 
 is subject to the pressure due to the rate of flow, while in P 3 the 
 static pressure is diminished by approximately a like amount. 
 
 The Pitometer. This instrument is extensively used by water 
 works engineers to determine the rate of flow through different 
 water mains. The instrument shows the varying velocity of flow 
 in a water main by the deflections of a colored liquid in a glass U 
 tube. Its accuracy depends upon four things, viz., first, the proper 
 setting of the instrument in the main; second, the use of the cor- 
 rect specific gravity of measuring liquid; third, the removal of air 
 from the connections; fourth, the use of the proper decimal or 
 coefficient belonging to the pipe where the pitometer is used, for 
 as is well known, water runs more slowly near the wall of the pipe 
 than at the center; hence to get the correct average velocity from 
 which flow is figured, we must know what per cent it is of the ob- 
 served center velocity. This average velocity always bears the same 
 relation to the center velocity in any one pipe no matter what the 
 rate of flow is, and, as we have found, this varies from about 70 
 to 90%, according to the age of the pipe or the roughness of its 
 interior surface. 
 
 Fig. 64 shows the pitometer in its simplest form. It consists of 
 the Rod Meter, containing the two Pitot tubes with nozzles for in- 
 sertion in the pipe, the U tube for measuring the difference in level 
 and a street connection, permanently attached to the water main for 
 inserting the instrument. 
 
 The Rod Meter consists of a brass "sheath" of flat oval cress 
 section containing two ^-inch brass tubes, each terminating in a 
 curved phosphor-bronze orifice at the lower end. At the upper end 
 of each tube is clamped a finger, which engages the notch marked 
 
120 
 
 EXPERIMENTAL ENGINEERING 
 
 FIG. 64. " Street Connection," Rod Meter and Manometer in Place. 
 
MEASUREMENT OF THE RATE OF FLOW OF WATER 121 
 
 " shut " of a loose sleeve in such a way that when the tube is re- 
 volved to one position the orifices will be turned " in " ready for 
 insertion through the 1-inch pipe tap. Turning the fingers to en- 
 gage the other notches marked " open " the orifices will be turned 
 " out " in a line parallel to the flat side of the oval sheath. This 
 latter is the position of the orifices when in use, and care should 
 be taken to see that the finger clamps are so adjusted that when in 
 this position the orifices will be in alignment. 
 
 The " IT " Tube is a. glass tube about four feet in length, bent 
 in the middle into U shape so as to bring the two legs near together. 
 The top of each leg is connected by rubber tubing and the metallic 
 tubes to the orifices of the meter. This U tube is filled for about 
 half its height with a carbon tetrachloride mixture of definite 
 specific gravity. When the orifices are brought into a current of 
 water the velocity causes the liquid to rise in one leg and fall in the 
 other. The vertical distance between the tops of the liquid in the 
 two legs constitutes the " deflection " by means of which the velocity 
 is known. 
 
 Street Connections. In order to provide a convenient method by 
 which the meter may be repeatedly used upon a water main laid 
 beneath a street pavement, a street connection, shown in Fig. 65, 
 is set at each point, where observations are required and the pave- 
 ment restored. This street connection becomes a part of the pipe 
 system and affords immediate means of access at all times. 
 
 The Traverse. The accuracy of pitometer work depends very 
 largely on the traverse and the determination of the pipe coefficient 
 or decimal. 
 
 The pitometer measures velocity only at that point in the pipe 
 at which the orifices are placed. If, while the discharge of a pipe 
 is constant, the orifices of the pitometer be moved slowly along 
 the diameter, it will be noticed that the velocities vary at each point 
 gradually increasing from the inner surface toward the center. 
 
 To determine the quantity of water being discharged, a traverse 
 is first made of the pipe at the gaging point, and from this traverse 
 the pipe ' coefficient (mean velocity divided by center velocity) is 
 obtained. The orifices are then left at the center and the mean 
 velocity at any rate of flow may always be found by multiplying the 
 center velocity by this coefficient. 
 
122 EXPERIMENTAL ENGINEERING 
 
 A sufficient number of traverse velocities should be taken to locate 
 definitely a smooth curve. It may be found that the points are not 
 falling on a single curve. This is due to a varying center velocity. 
 The curve should consist only of points which are taken with the 
 same center velocity. This may be accomplished by returning the 
 orifices quickly to the center for a check reading. 
 
 In the practical use of this instrument a recording device is 
 arranged to register continuously the deflection in the IT tube. 
 
 FIG. 65. Pitometer Street Connection. 
 
 Tables are employed giving the velocity in feet per second for every 
 possible amount of deflection of the liquid, in the various sizes of 
 pipe in which the instrument is to be used. 
 
 Weirs. This method is used to measure the flow of water in 
 aqueducts, ditches, and other open streams, where the size of the 
 stream, flow of water, and other circumstances, permit the con- 
 struction of a dam and weir. For measuring the flow of water 
 discharged from a pipe, this method is also sometimes employed, 
 using a tank of the form described on page 125. 
 
 A weir for measuring the flow of water is a notch in the top of 
 the vertical side of a vessel or reservoir, through which water flows. 
 
MEASUREMENT OF THE BATE OF FLOW OF WATER 123 
 
 The notch is generally rectangular, the lower edge of the rectangle 
 being truly horizontal, and its sides vertical. The lower edge of 
 the rectangle is called the " crest " of the weir. Fig. 66 shows an 
 outline of the usual form of weir, in which the vertical edges of the 
 notch are sufficiently removed from the sides of the reservoir or 
 
 FIG. 66. 
 
 feeding canal, so that the sides of the stream may be fully con- 
 tracted. This is called a weir with end contractions. In another 
 form, not so often used, the edges of the notch are coincident with 
 the sides of the stream. 
 
 In taking accurate observations of the rate of discharge by means 
 of a weir, it is necessary that the inner edge of the notch shall be 
 a definite angular corner, so that the water in flowing out may touch 
 
 FIG. 67. 
 
 the crest only in a line, thus ensuring complete contraction. In 
 precise observations a thin metal plate should be used for a crest, 
 as shown in Fig. 67. Where extreme accuracy is not important, it 
 may be sufficient to have the crest formed by a plank of smooth 
 hard wood with its inner corner cut to a sharp right angle and its 
 
124 EXPERIMENTAL ENGINEERING 
 
 outer edge bevelled. In weirs with end contractions, the vertical 
 edges should be made in the same manner, while for those without 
 end contractions, the sides of the feeding canal should be smooth 
 and prolonged a slight distance beyond the crest. The distance 
 from the crest to the bottom of the feeding canal, or reservoir, 
 should be at least three times the head of water on the crest. For a 
 weir with end contractions, a similar distance should exist between 
 the vertical edges of the notch and the sides of the feeding canal. 
 Let Ji = ihe head of water on the crest = the vertical height of the 
 plane of the level surface taken well back of the weir, above the edge 
 over which the water flows. Let b = the breadth of the crest in feet, 
 and Q = the discharge in cubic feet per second. Then theoretically, 
 
 Practically the rate of flow is not so great as this. Extensive 
 experiments were conducted by Professor Francis in 1854, who de- 
 duced the following formulas for determining the rate of flow : 
 
 For weirs with full contraction, 
 
 For weirs with one end contraction suppressed, 
 
 = 3.33(&-0.1/i)P. 
 For weirs with both end contractions suppressed, 
 
 The value of Q thus obtained is sufficiently accurate for ordinary 
 purposes. If extreme accuracy is desired it is best to consult tables 
 giving the number of cubic feet of water that will flow over a weir 
 1 inch wide and of any given depth. The figure thus found is 
 multiplied by the width of the weir in inches. 
 
 The value of Ji should not be less than 0.1 foot, and it rarely 
 exceeds 1.5 feet. The least value of b in practice is about 0.5 foot, 
 and it does not often exceed 20 feet. 
 
 The Hook Gage. The value of li must be determined with pre- 
 cision in order to avoid error in the computed discharge. Obser- 
 vations of its value are sometimes made by setting up a stake 6 feet 
 or more behind the crest of the weir. The level of the water is 
 first marked on the stake when the supply is shut off and the water 
 
MEASUREMENT OF THE BATE OF FLOW OF WATER 125 
 
 just on the point of flowing over the crest. With the water flowing 
 the level is again marked on the stake and the distance between 
 marks is measured. More reliable observations are taken by means 
 of the hook gage. A bent hook, with pointed end upward is in- 
 serted in the lower end of a rod sliding vertically in fixed supports, 
 the amount of vertical motion being determined by the readings of 
 a vernier. The vernier can be set to read 0.000 when the point of 
 the hook is at the level of the water when it just reaches the crest 
 
 INLET FOR 
 WATER 
 
 FIG. 68. 
 
 of the weir. When the water is flowing over the crest the rod is 
 raised by a tangent screw until the point of the hook is at the water 
 level. With the water flowing, before the point pierces the surface, 
 a slight ripple or protuberance will be seen to rise above it. The 
 hook should then be carefully lowered until this ripple is barely 
 perceptible, when the point will be at the true water level. The 
 scale and vernier reading then indicates the value of li. 
 
 Tank for Weir Apparatus. Fig. 68 shows a tank for use in 
 measuring the flow of water from a pipe, by means of a weir. This 
 9 
 
126 
 
 EXPERIMENTAL ENGINEERING 
 
 tank and weir can often be constructed for use in cases where an 
 accurate meter is not obtainable. In constructing such a tank it 
 should be sufficiently long to permit the water to rise to a level and 
 flow smoothly for a distance of at least 6 feet to the weir. The 
 water enters at the end opposite the weir and flows under and over 
 a series of diaphragms that are so placed as to remove the disturb- 
 ing effect of the stream of water pouring in. 
 
 This tank and its operation have been described by Professor 
 Spangler of the University of Pennsvlvania. as follows : 
 
 First level up with water in tank from lower edge of weir to 
 exact point of hook gage and mark on edge of tank. When run- 
 ning, raise hook until a slight hill of water just shows over point, 
 but point does not show. Mark this on edge of tank for height of 
 
 
 FIG. 69. 
 
 water over weir. The difference between these points is Ji. The 
 vertical edges of weir must be exactly at right angles to the hori- 
 zontal edge and the horizontal edge exactly level. The horizontal 
 edge must be square at side within tank, over which water flows. 
 In Fig. 69 if the water flows across the side from a to &, the edge 
 b may be of any shape, but a must be square. In flowing over the 
 edge a the water will turn upward and it will be possible to insert 
 a finger nail at &. 
 
 The hook gage used with this tank is made of -J-inch wire in- 
 serted in the end of a wooden batten. The end of the wire is turned 
 up to form a hook and the batten is made adjustable to clamp in a 
 vertical position when held by a beam lying across the tank. 
 
 The Right-Angled Triangular Weir. In 1858 Professor James 
 Thomson, of Queen's College, Belfast, suggested the use of the 
 
MEASUREMENT OF THE BATE OF FLOW OF WATER 127 
 
 right-angled triangular notch with its apex pointing downward, to 
 take the place of rectangular weirs, because the latter were not 
 adapted to the measurement of small and variable quantities of 
 water. Another advantage of the triangular notch given by Prof. 
 Thomson is, that the quantity of water flowing becomes a function 
 of only one variable, viz., the head of water ; while in the rectangular 
 notch it is a function of at least two quantities : the head and the 
 horizontal width. The application of the triangular weir at the 
 present time is limited, but for some kinds of laboratory testing 
 and for measuring small flows in irrigation work, it is very con- 
 venient. 
 
 --b-- 
 
 FIG. 70. Triangular Notched Weir. 
 
 In Fig. 70 let 
 
 7t:=head of water, 
 
 1) = width of notch at level of the water, 
 
 then the area of the notch filled is %bh; but in a right-angled 
 triangle 
 
 b = 2h f 
 therefore, 
 
 Theoretically the velocity of a body falling through a vertical 
 distance h is, 
 
 V2p, 
 
 where g is the acceleration due to gravity and the theoretical dis- 
 charge is, 
 
 Thus the discharge is proportional to h*. This reasoning was sub- 
 stantiated by more rigorous mathematical development later. 
 
128 EXPERIMENTAL ENGINEERING 
 
 Now it remained to determine by experiment whether there was 
 some constant which multiplied by 7i* would give the quantity of 
 water discharged for all values of \, that is, 
 
 q = ch*. 
 
 After extended trials, with h varying from 2 to 4 inches, the follow- 
 ing expression was arrived at, 
 
 where 
 
 q cubic feet discharged per minute, and 
 
 h = head measured vertically in inches from the still-water level 
 of the pool down to the vertex of the notch. 
 
 It is the present custom to measure all heads of water in feet, 
 and weir discharges are usually measured in cubic feet per second, 
 thus 
 
 The coefficient reduced to give the result in these units will change 
 the equation to, 
 
 Miner's Inch. Though not used in steam engineering ex- 
 perimental work, this unit is much used in certain parts of the 
 country and is here explained for the information of the student. 
 It is roughly defined as the quantity of water that will flow from 
 a vertical standard orifice 1 inch square, when the head on the 
 center of the orifice is 6J inches. This gives a rate of flow of about 
 1.5 cubic feet per minute, which may be taken as the mean value 
 of the miner's inch. Owing to the fact that, when water is bought 
 for mining or irrigation purposes, a much larger quantity than one 
 miner's inch is required, orifices much larger than 1 square inch 
 in area, and of various sizes are used in measuring it. This leads 
 to considerable variation in the mean value of the standard as used 
 in various localities. The actual value ranges from 1.20 to 1.76 
 cubic feet per minute. 
 
CHAPTER VI. 
 MEASUREMENT OF THE RATE OF FLOW OF AIR AND STEAM. 
 
 Anemometer. This instrument is used in measuring the velocity 
 of a current of air. In steam engineering work it is sometimes em- 
 ployed in boiler tests for determining the approximate quantity of 
 air supplied for combustion. 
 
 FIG. 71. 
 
 The usual form of such instrument is shown in Fig. 71. It 
 consists of a fan wheel with dial for registering the number of 
 linear feet of air that passes the fan. The index can in the usual 
 form of construction be set to zero and the registering mechanism 
 can be thrown in or out of gear at will. 
 
 To take an observation, the index is set to zero and the instru- 
 ment is placed in the air current with the fan wheel facing the cur- 
 rent. The disconnector is withdrawn and the observation com- 
 
130 EXPERIMENTAL ENGINEERING 
 
 menced. At the end of the period of observation, touch the dis- 
 connector and throw the movement out of action, the result may 
 then be read off. Commencing with the highest dial, note carefully 
 what figure the index has actually passed, and add to this the figure 
 last passed by the index of each succeeding dial, ending with the 
 large hand. The reading represents the number of linear feet of 
 air passed during the observation. This multiplied by the area in 
 square feet of the cross section of air passage, will give the number 
 of cubic feet of air registered. 
 
 In case the index cannot be set to zero, it should be noted that the 
 reading of the instrument must be taken both before and after the 
 observation. The difference will then give the number of feet of 
 air that have passed. 
 
 TUB*. /fS'r TH/Ct^. TMM/? TUB " JJ//W. 
 
 FIG. 72. 
 
 Air measurements obtained through the use of the anemometer 
 cannot be relied upon, since the velocities at different points in a 
 cross section of an air pipe are not uniform. If taken at the center 
 of the pipe a correction is sometimes applied to obtain the mean 
 velocity through the pipe. This method is also inexact since air is 
 an elastic fluid and does not always move in lines parallel to the axis 
 of the pipe. This is particularly true where the velocities are low. 
 
 Pitot's Tube for Measuring Low Air Velocities. Taylor's Method. 
 In a paper by Naval Constructor D. W. Taylor, U. S. N., read 
 before the Society of Naval Architects and Marine Engineers in 
 1905, the following method of measuring the volume of air de- 
 livered by ventilating fans was proposed. This method is now 
 used for testing ventilating sets and forced draft fans for the Navy. 
 
 A special form of Pitot's tube is used as shown in Fig. 72. 
 
 The impact pressure is taken on the tapered end of this tube and 
 
MEASUREMENT OF BATE OF FLOW OF AIR AND STEAM 131 
 
 communicated through the center tube to the right angle branch at 
 the opposite end. The static pressure is communicated through the 
 slots on sides, and the annular tube to the first right angle branch. 
 
 A nest of nine of these tubes is made up so that it exactly fits in a 
 slot cut 2 inches wide and 13 inches long in the pipe on which the 
 test is to be made. The tubes can be adjusted as to depth, depend- 
 ing upon the size of the pipe under test. Stops are placed so that 
 when the tubes are swung out against them, the ends will come at 
 a distance from the center such that each tube represents an area 
 equal to one-ninth of the whole. The arrangement is that shown in 
 Fig. 73 where the cross section is divided into nine equal zones, 
 with one of the tubes in each zone. 
 
 FIG. 73. 
 
 It will be noted that tubes 1, 2 and 9 are in a vertical line, while 
 tubes 3, 5 and 7 are in a line inclined 30 to the left and tubes 4, 6 
 and 8 in a line inclined 30 to the right, as shown further in 
 Fig. 73. The depth of setting for each tube in any given diameter 
 of pipe is taken from a diagram furnished with the apparatus. 
 
 The Manometer. For measuring the flow with the tubes as thus 
 arranged a special form of manometer is employed, as shown in 
 Fig. 74. Two inverted cans are placed in a trough of water. Each 
 of these cans is divided into nine compartments, each having the 
 same cross sectional area, and a tube passes through the bottom of 
 
132 
 
 EXPERIMENTAL ENGINEERING 
 
 the trough above the surface of the water into each compartment. 
 Connecting the nine impact pressure nozzles with the nine tubes 
 under one can and the static pressure nozzles with the tubes under 
 
 FIG. 74. 
 
 the other can we have a direct means of measuring the mean dif- 
 ference between impact and static pressure over the whole cross 
 section of air duct. A sliding weight on a scale beam enables the 
 
MEASUREMENT OF KATE OF FLOW OF Am AND STEAM 133 
 
 two cans to be balanced and the pressure read off in pounds per 
 square foot. 
 
 Method of Calculating the Rate of Flow. The velocity in the 
 duct is given by the formula 
 
 where v is the velocity in feet per second, W is the weight per cubic 
 foot and (Pi p 2 ) is the difference in pressures shown by the tubes. 
 
 For ordinary work it is sufficient to take 17=. 0807, which is the 
 approximate value for dry air at an atmospheric pressure corre- 
 sponding to normal barometer reading of 29.92 inches. 
 
 For more exact observations W is given by the formula 
 
 w= .080723 _ fr-0.378e 
 
 l + .0020389(* + 32) X 29.921 ' 
 
 where t is the temperature in degrees F., b is the barometer read- 
 ing, and e is the pressure due to the vapor in the air in inches of 
 mercury. Tables and curves for applying these corrections have 
 been prepared and are issued in pamphlet form by the Bureau of 
 Construction and Repair, Navy Department. 
 
 Having obtained the mean velocity of the air, this multiplied by 
 the area of cross section of the duct and by 60 gives the number of 
 cubic feet discharged per minute, which is the result desired. 
 
 Use of a Single Pilot's Tube. A single Pitot's tube placed in the 
 center of the duct is sometimes employed, and a correction applied 
 to give the mean velocity, but with low velocities, results thus ob- 
 tained are not reliable. A single tube may be employed where the 
 rate of flow continues during a considerable period of time, long 
 enough to permit shifting the position of .the tube and thus obtain a 
 series of observations, the mean of which gives the mean velocity in 
 the pipe. 
 
 Measuring the Rate of Flow of Steam. 
 
 Where steam is discharged through nozzles, the velocity is given 
 very closely by the following empirical formula which was proposed 
 by Lord Napier: 
 
 Flow in pounds per second = absolute initial pressure X area in 
 square inches + 70. 
 
134 EXPERIMENTAL ENGINEERING 
 
 This rule is applicable where the terminal pressure does not ex- 
 ceed 58% of the initial pressure. Within these limits any variation 
 in the terminal pressure does not affect the rate of flow. 
 
 When compared with results obtained, by measuring the water 
 consumption on the trials of the Curtis marine turbines on the 
 IT. S. S. North Dakota, this formula gave results within 1 to 3%. 
 The results obtained through measurement may have been in error 
 to that extent. 
 
 Steam Meters. 
 
 The direct measure of the efficiency of a steam engine is the 
 weight of steam used by it in a given time to produce a given power. 
 In engine tests the steam is usually condensed and weighed, but the 
 same purpose is accomplished much more easily if a reliable steam 
 meter can be employed. 
 
 Steam meters may be conveniently grouped in two general classes, 
 which, for lack of more suitable names, may be designated as series 
 meters, and shunt meters. 
 
 The series meter is an integral part of the piping, the entire mass 
 of fluid to be measured passing through the apparatus. The St. 
 John's and Venturi meters are examples of this class. In the 
 former the volume of fluid passing is determined by the rise and fall 
 of a weighted plug valve and in the latter the velocity of flow is 
 determined by the well-known principles of the Venturi tube. Both 
 are indicating instruments and show only the rate of flow. 
 
 The Sarco Steam Meter. This is another example of the series 
 meter. If steam be allowed to expand from a vessel in which there 
 is a pressure p^ through an orifice into a lower pressure p 2 the 
 theoretical weight of steam flowing out per minute when p., is 
 nearly equal to /> is given by the equation 
 
 where W is the weight of steam issuing per minute in pounds; A 
 is the area of the orifice in square inches; p^ and p 2 the pressures 
 in pounds per square inch (absolute) ; and v the specific volume of 
 the steam at pressure p^ in cubic feet per pound. 
 
MEASUREMENT OF KATE OF FLOW OF AIR AND STEAM 135 
 
 This principle is applied in the Sarco meter, where a disc, 1 in 
 Figs. 75 and 76, ^ inch thick, held in position by the bolts sur- 
 rounding it, is placed between two flanges at a point in the pipe-line 
 where it is desired to measure the weight of steam passing. 
 
 This disc has a bore slightly smaller than that of the pipe, causing 
 the steam to be throttled to the extent of a drop in pressure of about 
 | pound per square inch, or less than would be visible on an ordinary 
 pressure gage. 
 
 FIG. 76. 
 
 The difference in pressure on the two sides, which is the medium 
 through which the Sarco meter determines the flow of steam, is 
 conveyed to the instrument through copper pipes of J-inch (or 
 larger) internal diameter, which connect with holes drilled in the 
 disc, and communicate with the high and low pressure sides. 
 
 For vertical pipes the outlets of these connecting pipes are 
 counter-sunk (Fig. 76) in such a way that no water can collect in 
 the bore which points upwards, as this would cause an unequal head 
 on the two sides of the meter. The connecting pipes, before being 
 led to the instrument, are carried along horizontally in the same 
 plane for a few feet with the same object of preventing a difference 
 in head of water on the two sides which might result from unequal 
 
136 
 
 EXPERIMENTAL ENGINEERING 
 
 condensation. The pipes between the disc and meter are always 
 kept filled with water. 
 
 The Recorder. The high-pressure side of the throttle disc is con- 
 
 FIG. 77. 
 
 nected through a three-way cock or valve 7 (Fig. 77) to the pipe 9; 
 this leads into a mercury reservoir 10, which in turn connects to a 
 hollow cone 12 through the tube 11. 
 
 The cone 12 is suspended by means of springs 26, and has a con- 
 
MEASUREMENT OF RATE OF FLOW OF AIR AND STEAM 137 
 
 nection 13 at its upper end. This gives access to the low-pressure 
 side of the disc through a special port and the tube 8. Thus it will be 
 seen that the higher pressure, acting through the water in the con- 
 necting tubes upon the mercury in 10 will tend to drive this out 
 along tube 11, thus causing the cone 12 to sink. On the other hand, 
 the lower pressure from the other side of the disc will press on the 
 surface of the mercury through 8 and 13, attempting to force it 
 back into 10. The difference between the two pressures will de- 
 termine the position of the cone 12. Its movements are recorded on 
 a chart 17 by means of a pen gear 16, operated through levers 14 
 and 15. 
 
 The chart is driven by clockwork; it is arranged for 24 hours, and- 
 is calibrated directly in pounds of steam per second. The total con- 
 sumption over any period may be directly obtained from this by an 
 ordinary planimeter. The charts and throttle disc to correspond are 
 varied for each different size of pipe, and arranged so as to permit 
 of the maximum flow of steam likely at the point in question to fall 
 within the range provided. 
 
 Where it is desired to have a record of the total flow of steam over 
 a period, an integrator or totaliser is fitted. 
 
 This consists of a disc 20, driven by a separate clock movement, 
 and a friction gear 18 suspended from the pen lever 15 and moving 
 up and down with it on the surface of the disc 20. A friction wheel 
 19 is driven by the disc, and, by means of a worm gear, causes the 
 pointer 22 to move around dial 21. The speed of the pointer will 
 then depend upon the position of 19 in respect of the periphery of 
 disc 20. The dials are calibrated in pounds of steam, and are read 
 once in 24 hours. 
 
 As accurate results could only be obtained with the instrument 
 as so far described where the steam pressure is constant, automatic 
 compensators are used where fluctuations exceeding 5 per cent have 
 to be dealt with. 
 
 These regulators consist of an oil chamber 30 which communi- 
 cates through small holes with a piston 31. The piston rod 32 is 
 held down by a strong spring 33 connecting to a segment lever 34. 
 
 The regulator is put under pressure through tube 28 (connected 
 at 27) and controlled by the valve 29. When piston 31 is forced 
 
138 
 
 EXPERIMENTAL ENGINEERING 
 
 upwards, the spring 33 extends, and this causes lever 34 to swing 
 downwards, thus shifting the fulcrum of the pen lever and so auto- 
 matically correcting the chart and integrator readings. 
 
 A simple form of indicating instrument is constructed on the 
 same principle as the recorders, for use as a load meter. Attached 
 to the front of a steam boiler with the throttle disc inserted in the 
 main steam pipe, immediately behind the boiler stop valve, these 
 little meters do excellent service in showing instantly the slightest 
 change in the load, thus enabling firemen to adapt their firing and 
 prevent the otherwise inevitable loss of pressure usually the first 
 indication of a change. 
 
 Shunt Meters. In the shunt meter only a portion of the steam 
 to be measured is diverted through the apparatus, the velocity of 
 flow through the shunt being an indication of that in the main pipe. 
 In this class one or more small openings ^ inch or less in diameter 
 suffice for attaching the apparatus to the pipe. One instrument 
 suitably calibrated may answer for any size of pipe. The Pitot tube 
 forms the basic principle of practically all meters of this class. 
 
 Trailing Set 
 
 FIG. 78. 
 
 The General Electric Company's Steam and Air Flow Meters. 
 
 The principle governing the action of the flow meter is a modifica- 
 tion of that of the Pitot tube. A brass nozzle plug, Fig. 78, screwed 
 into the pipe at the point where the flow is to be measured, carries 
 two sets of openings : a leading set, facing the direction of flow and 
 extending diametrically across the pipe; and a trailing set, con- 
 sisting of two openings at 90 and one at 180 to the direction of 
 flow. The impingement of the steam against the leading openings 
 sets up in them a pressure equal to the static pressure plus the 
 pressure due to the velocity head, while the trailing set is acted on 
 by the static pressure less that due to the velocity. The difference 
 
MEASUREMENT OF KATE OF FLOW OF AIR AND STEAM 139 
 
 in these values is a measure of the velocity, and for constant tem- 
 perature and pressure, gives the rate of flow. The pressures exist- 
 ing in the two sets of openings are communicated through separate 
 longitudinal tubes to the outer end of the plug and from there by 
 J-inch iron pipes to the meter. 
 
 Recording Steam Flow Meter. The Eecording Steam Plow 
 Meter, Fig. 79, is a curve drawing instrument, accurately calibrated 
 to record the total rate of steam flow in pounds, per hour in any 
 diameter pipe at any condition of pressure, temperature or moisture. 
 
 In this meter there are two cylindrical hollow cups filled to about 
 half their height with mercury and joined together at the bottom 
 by a hollow tube. This " U " tube is supported on, and free to move 
 as a balance about, a set of knife edges. The two pressures obtained 
 by the nozzle plug are communicated to the cups by flexible steel 
 tubing, whereupon the difference in pressure is equalized by a rising 
 of mercury in the left-hand cup and a falling in the right-hand 
 cup. Due to the displacement of the mercury, the beam carrying 
 the cups tilts on the knife edges until the moment of the counter 
 weights on the extreme right of the meter exactly balances the 
 moment caused by the displacement of the mercury in the left-hand 
 cup. 
 
 The motion of the beam is multiplied by levers and is registered 
 by a pen. The time element of the meter consists of an eight-day 
 clock driving a drum and feeding paper at the rate of 1 inch per 
 hour. Charts are supplied in sizes to 'measure a flow of from 2000 
 to 240,000 pounds per hour, and of sufficient length to last one 
 month. The rate of flow can be read at any instant or the average 
 rate of flow calculated for a given time. 
 
 Automatic Pressure Correction Device. The meter is adapted to 
 any condition of pipe diameter, pressure, superheat or moisture by 
 a hand adjustment of a correction weight on a graduated arm. A 
 chart supplied with the meter shows the correct position for any 
 existing condition. 
 
 Eeferring again to Fig. 79, if the pressure in the steam main 
 varies more than 10 pounds from normal, compensation is neces- 
 sary for the error thus introduced. An automatic pressure cor- 
 rection device, consisting of a hollow spring, similar to the pressure 
 
140 
 
 EXPERIMENTAL ENGINEERING 
 
MEASUREMENT or KATE OF FLOW OF AIR AND STEAM 141 
 
 TbHo/e L /'/) Mozz/ef>/tf$ 
 
 SHQ/C/IT/NG FLOW METE ft 
 
 TYfff 
 f&t/re A 
 
 FIG. 80. 
 
 10 
 
142 EXPERIMENTAL ENGINEERING 
 
 FIG. 81. Indicating flow meter, type I, form F for measuring steam 
 
 flow. 
 
MEASUREMENT or RATE OF FLOW OF AIR AND STEAM 143 
 
 spring in a steam gage, is connected so as to be influenced by the 
 static pressure of the steam at the point where the flow is being 
 measured. Any variation of the static pressure causes the spring 
 to expand or contract, and this movement actuates the small cor- 
 rection counter weight and affects the movement of the pen in such 
 a manner that the recorded rate of flow is correct. 
 
 Indicating Steam Flow Meters. The instrument shown in Figs. 
 80 and 81 will meet general requirements where an indicating rather 
 than a recording instrument is required. 
 
 The meter consists of an iron casting, cored out to form a " U " 
 tube, and partially filled with mercury. The difference in pressures, 
 as transmitted from the nozzle plug causes a difference in the 
 mercury levels, and the displacement of the mercury actuates a 
 pulley by means of a small float suspended by a silk cord. The 
 pulley moves a small " U " magnet on the end of the shaft next to 
 the dial in proportion to the change in level of the mercury in the 
 " U " tube. The indicating needle is mounted in a separate cylin- 
 drical casing. The pivoted end consists of a bar magnet, free to 
 turn in the same plane as the magnet on the inside of the meter. 
 The mutual attraction of the two magnets keeps them parallel; a 
 packed joint to transmit the motion of the pulley to the indicating 
 needle is thus eliminated. 
 
 Proper adjustments for the existing conditions of pipe diameter, 
 pressure and temperature are readily made by setting the graduated 
 cylinder which actuates the rack carrying the pointer. When these 
 settings are made, the rack is rotated by hand until the pointer 
 coincides with the indicating needle. The point on the calibrated 
 dial at the intersection of the needle and pointer gives the true in- 
 stantaneous rate of flow in pounds per hour per square inch pipe 
 area. 
 
 Air Flow Meter. The Indicating Air Flow Meter is identical in 
 principle and method of operation with the Indicating Steam Flow 
 Meter, except that water is used in the " U " tube as a working fluid 
 and the chart dial is calibrated to read in cubic feet free air per 
 minute at 70 Fah. per square inch pipe area. 
 
CHAPTEE VII. 
 MEASUREMENT OF POWER. 
 
 Power is the term used to denote how fast work is done, or 
 energy transferred. Technically it is denned as the rate of doing 
 work per unit of time. 
 
 Horse Power. Taking the foot pound as the unit of work, the 
 horse power has come to be recognized as the unit of power. One 
 horse power is equivalent to 33,000 foot pounds of work done per 
 minute. 
 
 Indicated Horse Power is the term used to denote the work done 
 by the steam on the piston of a steam engine and takes its name 
 from the indicator. 
 
 Brake Horse Power or Shaft Horse Power is the power trans- 
 mitted to the shaft and is less than the indicated horse power by an 
 amount equal to the friction of the moving parts. In an ordinary 
 steam engine the brake horse power is about 85% of the indicated 
 horse power. In a marine steam turbine the shaft horse power is 
 usually assumed to be 92% of the indicated horse power that would 
 be developed in a reciprocating engine that would transmit to the 
 shaft the given shaft horse power. 
 
 Metric Horse Power. The horse power, cheval vapeur, used in 
 countries that employ the metric system, is 75 kilogrammeters of 
 work per second. This is approximately equivalent to 32548-foot 
 pounds of work per minute, or .9863 horse power in our units. 
 
 Efficiency of a Machine. In any mechanical apparatus the effi- 
 ciency is equal to the ratio 
 
 Useful work done by machine 
 Total energy received by machine* 
 
 A large portion of experimental engineering work is devoted to 
 the determination of this ratio for different machines and resolves 
 itself into the measurement of power and the calibration of the 
 instruments employed. 
 
MEASUREMENT OF POWER 145 
 
 The Steam Engine Indicator, used for determining the power de- 
 veloped in the cylinders of reciprocating engines, is described in 
 text-books on the steam engine, together with the methods em- 
 ployed in taking cards and in calculating the power developed and 
 the distribution of steam. We will therefore take up only the errors 
 that are likely to be met with and their methods of correction. 
 
 These errors may be classified as follows: 
 
 (1) Errors due to incorrect calibration of indicator piston 
 springs. Springs that are calibrated cold will show errors when hot. 
 
 (2) Badly fitting pistons. A piston that will not fall through 
 the cylinder by its own weight is too tight and will produce errors 
 due to friction. The normal friction of an indicator piston should 
 be so small that there will not be even the thickness of a hair be- 
 tween the lines obtained from rising and falling steam pressures 
 in the tests to determine the hot scale of the springs. At the same 
 time if the piston is loose enough to leak sufficient steam to cause 
 any back pressure, it cannot indicate correctly. 
 
 (3) Errors due to inertia of piston and pencil mechanism. A 
 distorted diagram is produced, due to vibration of the pencil. These 
 errors are now reduced to a minimum by making the mechanism as 
 light as possible. 
 
 (4) Errors due to variable tension on cord. Paper drum inertia. 
 With the crosshead of the engine moving away from the indi- 
 cator, the cord unwinds and moves the drum against the action 
 of the drum spring. On the return stroke the drum spring moves 
 the drum, winding up the cord. With the engine standing still the 
 tension on the cord depends on the amount of movement that has 
 taken place against the drum spring, in other words on the position 
 of the engine piston. With the engine in operation this tension is 
 influenced by the inertia of the moving drum, and for one particular 
 speed, the tension in cord throughout stroke will be approximately 
 constant. The drum spring should be long so that the tension due 
 to its compression will vary as little as possible throughout the 
 stroke. The drum should be light to reduce inertia and the cord 
 should be of such material as will stretch as little as possible, so as 
 not to distort the indicator card. 
 
 (5) Errors due to incorrect indicator motion. The mechanism 
 
146 EXPERIMENTAL ENGINEERING 
 
 for reducing the motion of the engine crosshead and imparting 
 this reduced motion to the paper drum should be so constructed 
 that the reduced motion will be exactly similar to that of the engine 
 crosshead. If not, the card will be distorted. Various forms of 
 reducing motions are in use,, many of which are only approximately 
 correct. Some, for the sake of their simplicity, have been adopted 
 that are very incorrect. In engine tests the indicator motion should 
 be analyzed to determine its accuracy in reproducing on a reduced 
 scale the exact motion of the engine piston. 
 
 Calibration of Indicator Springs. The manufacturers of recent 
 high-grade indicators now calibrate their springs hot, so that the 
 errors shown by them are small in comparison with what they 
 were before the importance of so calibrating them was understood. 
 A spring that registers correctly to scale at one temperature will not 
 be correct at a temperature considerably above or below it. In the 
 ordinary form of indicator, the spring is subjected to the heat of 
 the steam, which increases with the pressure. For accurate work it 
 is necessary to have a table of corrections, to be applied to the pres- 
 sures obtained from the cards, in order to obtain the real pressures. 
 Springs that have been accurately calibrated will change after a cer- 
 tain period of time, so that it will be found necessary to recalibrate 
 them. 
 
 In all forms of indicator spring testing apparatus, the indicator 
 is connected to a reservoir of steam in which the pressure may be 
 varied at will, provision being made for accurately measuring such 
 pressure. The earlier form of such apparatus, which was developed 
 by engineer officers of the U. S. Navy, measured the pressure by 
 means of a mercury column. While much work of great value was 
 done with this apparatus, it was unhandy, owing to the great height 
 of the mercury column, which it was necessary to build into the 
 wall of a tall building. In later forms of such apparatus, a scale 
 beam is employed to measure the pressure. 
 
 Professor Carpenter's Indicator and Gage Testing Apparatus. 
 This is shown in Fig. 82 and consists of a weighing scale mounted 
 on a heavy cast iron box connected to the steam reservoir. A 
 cylinder with a piston of -| square inch area is connected to this 
 box in such a way that the piston balances the beam of the scale if 
 there is no pressure in box. 
 
MEASUREMENT OF POWER 147 
 
 The box is tapped in different places for the connecting of indi- 
 cators or gages. In making a test, steam is admitted and the pres- 
 sure weighed by the scale. The " friction of rest " between the pis- 
 ton and the cylinder in which it works is intended to be overcome 
 by slightly rotating the piston with the finger, by means of the pro- 
 jections provided for that purpose. Accurate results can only be 
 obtained after considerable practice. 
 
 Professor Cooley's Indicator Testing Apparatus. This is shown 
 in Fig. 83 with a section through cylinder and connections shown 
 
 FIG. 82. 
 
 separately in Fig. 84. The indicator is attached to the indicator 
 cock, mounted at A, above a cylinder B, in which is a rotating 
 plunger C. This plunger is given a rapid rotary motion by means 
 of a belt from a small electric motor, on the wheel D. The wheel 
 D is mounted by a ball bearing and ball and socket joint on the step 
 E, which is placed on the platform of a specially constructed plat- 
 form scale. As pressure is applied in cylinder, the rotation of 
 plunger, eliminates the friction of rest between plunger and cylinder, 
 so that the amount of the pressure can be accurately weighed with 
 the scale beam. The wheel F serves to raise or lower the cylinder 
 slightly, thus equalizing the wear on plunger. 
 
 The cylinder B is in connection with a manifold under the table 
 through pipe G. H is a connection to the exhaust, on which is a 
 
148 
 
 EXPERIMENTAL ENGINEERING 
 
 valve I, and a smaller by-pass valve J. Connecting to manifold 
 under table are valves K, admitting steam; L f admitting air pres- 
 sure; and M, admitting water pressure. A fourth connection, not 
 shown, goes to an air pump for placing the manifold under a 
 vacuum. 
 
 The grooved cup above wheel D serves to catch any liquid that 
 
 FIG. 83. 
 
 leaks past plunger. Such liquid is carried off by the connections 
 shown, and by a rubber tube, clear of the apparatus. 
 
 There are three beams on the scale. N is for increments of 5 
 pounds. is graduated to measure fractional increments of 5 
 pounds, and P, at the left, measures descending increments, and is 
 chiefly used in calibrating springs at pressures below the atmosphere. 
 The counterpoise Q serves to balance the apparatus when no pres- 
 sure is on. 
 
MEASUREMENT OF POWER 
 
 149 
 
 FIG. 84. 
 
150 EXPERIMENTAL ENGINEERING 
 
 Method of Operation. The indicator spring being in place and 
 all ready for testing, start up the motor causing plunger to rotate 
 in cylinder. With all steam off and the exhaust valve open to the 
 atmosphere, place all poises at and balance the scale by means of 
 the counterpoise Q. Turn steam on and get indicator thoroughly 
 heated. Then, by means of the indicator cock, turn off the steam 
 and draw the atmospheric line. This will be the data line from 
 which all measurements must be taken. Open indicator cock, place 
 5 pounds on the scale beam, and by means of valve K, regulate the 
 pressure of steam so as to just a little more than balance the scale 
 beam. Then by means of valve 7, and by-pass 3 ' , the scales can be 
 balanced perfectly. 
 
 When the scales have been balanced with the 5 pounds on beam, 
 take a small wooden stick and tap lightly the cylinder of indicator. 
 At the same time pull the drum cord, holding the pencil point 
 against paper. Be sure that the scale is kept in perfect balance 
 while this is being done. The light tapping of indicator cylinder is 
 to overcome the friction of rest between indicator piston and 
 cylinder. 
 
 By increasing the weights on scale by such increments as may be 
 desired and taking the record on indicator card in this manner for 
 each increase of pressure, the complete record is obtained. 
 
 This method of testing eliminates the effect of friction of in- 
 dicator piston on the test cards. In a well constructed, properly 
 proportioned indicator, this should be so small as to be negligible. 
 To determine the effect of friction, a test line is drawn with the 
 steam pressure ascending and another with it descending, both with 
 the same setting of the scale. Half the distance between these lines 
 is the friction. 
 
 Method of Calculating Results. The area of plunger is square 
 inch. The pressure per square inch on plunger, and therefore on 
 indicator piston, will be twice the weight shown by scale beams. 
 Calculating this and measuring and dividing by the height of test 
 line above atmospheric line, we have the true scale of the spring. 
 
 Testing for Pressures Below Atmosphere. The same apparatus 
 may be used for tests below atmospheric pressure. The weight of 
 plunger (7 and attached parts is such that even a perfect vacuum in 
 
MEASUREMENT OF POWER 151 
 
 cylinder B would not support it. A vacuum pump is connected to 
 the manifold, with which the cylinder B is placed under a vacuum. 
 By means of beam P } on the scale, increments of pressure below the 
 atmosphere, are removed and tests made as before. Beam P is 
 graduated to 6 pounds, corresponding to 12 pounds below the atmos- 
 phere in cylinder. This is about as great a vacuum as an ordinary 
 vacuum pump will produce and the test is seldom carried further 
 in practice. 
 
 The Hospitalier-Carpentier Manograph for indicating internal 
 combustion engines and high-speed steam engines overcomes the in- 
 
 ACETYLENE 
 BURNER 
 
 PRESSURE 
 
 7W ' OF TRIPOD STAND 
 
 GENERAL APPEARANCE. 
 
 FIG. 85. 
 
 herent difficulties of the ordinary piston type of indicator which is 
 not suitable for very high-speed engines on account of the inertia 
 of its moving parts. The springs in an indicator also have a tend- 
 ency to break, due to the almost instantaneous combustion pressures 
 produced in the cylinders of internal combustion engines. 
 
 The manograph is shown in Fig. 85 with sectional views in Fig. 
 86. It substitutes a beam of light for the pencil of the ordinary 
 indicator and this beam traces an indicator card on a ground glass 
 screen. The light is projected from an acetylene burner furnished 
 with the instrument, reflected by a mirror which is pivoted in two 
 distinct planes at right angles to each other. Movement in one 
 
152 
 
 EXPERIMENTAL ENGINEERING 
 
MEASUKEMENT OF POWER 153 
 
 plane is obtained from the crankshaft through a flexible connection 
 to the mirror mechanism, thus producing horizontal motion of the 
 beam of light on the screen. A pipe connects the combustion cham- 
 ber of the cylinder with a small circular diaphragm, so arranged 
 that deflections in the diaphragm caused by pressure in the cylinder 
 act upon the mechanism,, so as to rotate the mirror in the other 
 plane and produce vertical movement in the beam of light. The 
 diaphragm corresponds to an indicator spring, and may be changed 
 to suit any desired pressure. The card outlined on the screen is 
 correct as regards pressures in the cylinder, but it must be cor- 
 rected for the angularity of connecting rod, which is not the case 
 with the ordinary steam engine indicator. 
 
 The mechanism for moving the mirror is arranged so that the 
 card may be brought to synchronism with the strokes of any piston. 
 Pipes leading to all the cylinders of a multicylinder engine may be 
 used in succession by means of a several-way cock, thus making it 
 possible to indicate the whole engine with one instrument. 
 
 When desired, the screen may be removed and a plate holder and 
 photographic plates substituted, in order to obtain a permanent 
 record. The principle use, however, is in connection with the 
 screen, since the observer is enabled to note immediately the effect 
 upon the card of various changes in the adjustment of the engine 
 such as the time of ignition, the position of throttle, the richness of 
 mixture, etc. This feature makes it particularly valuable for in- 
 struction purposes. 
 
 Ripper's Mean Pressure Indicator. This instrument, shown in 
 Fig. 87, consists mainly of a composition valve box, having pipe 
 connections to both ends of the engine cylinder, and fitted with 
 two dial pressure gages. By the automatic action of the valves in 
 the valve box one of these gages is always in communication with 
 the driving or impelling steam pressure acting on the piston, while 
 the other receives the back pressure acting against the piston. The 
 gages are fitted with fine regulation valves, and these are throttled 
 until the pointers are practically steady. They will then indicate 
 the exact mean of the varying pressures they are subject to, and the 
 difference between the readings of the two gages is the total mean 
 effective pressure acting on the piston. 
 
154 
 
 EXPERIMENTAL ENGINE: RING 
 
 The details of the instrument are shown in Fig. 88. Connection 
 is made to the cylinder ends by means of the pipes A and B. The 
 pipe C is connected to the gage showing the mean forward pressure 
 and the pipe D to the gage showing the mean back pressure. 
 
 FIG. 87. 
 
 The constant communication of the forward pressure gage with 
 the forward pressure side of the piston is secured by means of the 
 ball valve M, which is pushed forward at the end of every stroke by 
 the greater pressure of the incoming steam, thereby making com- 
 
MEASUREMENT OF POWER 
 
 155 
 
 municatlon between the forward pressure gage and the high pressure 
 side of the piston and closing the communication to the other side 
 of the piston. Similarly,, the double beat valve N, under the action 
 of the steam, makes the communication between the back pressure 
 gage and the back pressure side of the piston. 8 and T are dirt 
 collecting pockets, which may be cleared by removing the plugs V 
 and W. 
 
 w 
 
 FIG. 88. 
 
 The taps at J and K are for blowing through the instrument when 
 required, and the plugs J and K are for filling the gage syphons 
 with water. 
 
 The mean pressures are obtained by throttling a fine adjustment 
 valve fitted to each gage, but in addition, there are throttling cocks 
 G and H fitted on the instrument for the purpose of retaining the 
 water in the gage syphon so as to keep the gages cool. It is found 
 that the water will not disappear from the syphons so long as these 
 
156 
 
 EXPERIMENTAL ENGINEERING 
 
 cocks G and H are sufficiently throttled to prevent the range of 
 pressures in the syphon from exceeding about 10 pounds. By touch- 
 ing the syphon tubes C and D it is easy to ascertain whether they 
 remain charged with water. 
 
 Since the mean pressures are obtained with this instrument by 
 throttling, they are on a time base. They may be converted to the 
 means on a distance base, as given by the indicator diagram, by 
 
 
 @ 
 
 
 
 
 
 /W7? 0>4/?0 
 
 ItlfPCPS HCiH POCSlUtt IKSICtWI) 
 
 HM.Destroyer'FLYING FISH'. \ 
 
 
 
 
 'DATE. 
 
 
 
 TIME 
 REVpMIN.' 
 
 35-Q 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 ^C- 
 
 
 
 
 I.H.P: 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 5 
 
 13 -a. 
 
 Total M.E.P. 
 L.p'5into 
 
 /7 ./ o Jj? 
 
 
 
 
 
 
 4 
 
 I 
 
 1 
 
 
 1 
 
 
 ' ' !? ' ' '' ^ 
 
 I 
 
 LJL 
 
 2JO 
 
 / PT 4 s 
 
 
 
 
 /! 
 
 
 2,0 
 
 
 
 
 - 
 
 
 4|0_ 
 
 
 1 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 .A^ 
 
 
 
 
 JfnfeirJ 
 
 
 
 
 ^ 
 
 ?^y" 
 
 
 
 f 
 
 
 
 r- 
 
 
 
 
 lie 
 
 US 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 <? 
 
 
 
 
 - W/?^ 
 
 p 
 
 JWi. 
 
 R 
 
 "^ 
 
 
 
 
 9 
 
 
 
 
 
 p 
 
 =iT^gh^ 
 j 
 
 i .i.hu.i.iffii.i.iii.u.ili*'" " f ^ 0iur "> HS 
 
 
 II 
 
 
 J 
 
 FIG. 89. 
 
 adding a percentage depending upon the average point of cut off 
 of the engines, as follows : 
 
 Point of cut off 0.3 0.4 0.5 0.6 0.7 0.8 0.9 
 Add to m. e. p. 1/70 1/30 1/25 1/30 1/40 1/50 1/100 
 
 This correction may be conveniently applied by adding the per- 
 centage to the total mean effective pressure referred to the low 
 pressure, as illustrated in Fig. 89. 
 
 The inventor claims for this instrument that it gives for ordi- 
 nary engines, a close approximation to the results obtained with 
 
MEASUREMENT OF POWER 157 
 
 an indicator. In engines working under a high degree of compres- 
 sion,, such as locomotives and compound non-condensing engines, an 
 error is introduced owing to the reversal of the valves taking place 
 too early in the stroke. In these cases he recommends standardizing 
 the instrument against an ordinary indicator for different positions 
 of the link, and using the corrections thus arrived at, 
 
 Ripper's Power Board. This affords a means of determining 
 the indicated horse power, from the readings of the gages on the 
 mean pressure recorder, by simple mechanical manipulation, with- 
 out calculations, and enables the engineer to see almost at a glance, 
 what the engines are doing and how the power is distributed among 
 the cylinders. 
 
 Eef erring to Fig. 89, there are on the board a number of grooves, 
 one for each cylinder of the engine. Each groove has a marker, 
 with pointer to indicate the mean effective pressure in that cylinder, 
 and, for all except the low pressure cylinder, there is an additional 
 pointer to show the equivalent mean effective pressure on the low 
 pressure piston. The relative positions of markers show at once the 
 relative amount of work done in the various cylinders. By noting 
 on a pad, arranged on the end of board, the various mean effective 
 pressures referred to low pressure cylinder, adding them, and cor- 
 recting, the total mean effective pressure referred to the low pres- 
 sure cylinder is obtained. 
 
 A specially constructed horse power slide rule is fitted on the 
 lower part of the board. By setting the lower slide to the total mean 
 effective pressure referred to the low pressure cylinder, and the 
 upper slide to the speed in revolutions per minute, the horse power 
 may be at once read off. 
 
 Explanation of the Power Board. The upper part of the board 
 serves merely to facilitate reduction of the mean effective pressure 
 in all cylinders to its equivalent referred to the low pressure cylin- 
 der. Adding to the mean pressure in the low pressure cylinder the 
 results thus obtained for the H. P. and I. P. cylinders, the total 
 mean effective pressure referred to the low pressure cylinder is 
 obtained. 
 
 It will be seen from inspection of the lower part of the board that 
 it is a special form of the duplex slide rule, adopted for solving the 
 11 
 
 \ 
 
158 EXPERIMENTAL ENGINEERING 
 
 equation I. H. P. = mean effective pressure x revs, x const, which 
 
 P X 7? 
 
 may be written in the form I. H. P. = = . In order to limit 
 
 Ki 
 
 the size of the instrument, the various scales are shortened, such 
 portion only being furnished for each of them as is required for 
 the particular engine for which it is to be used. 
 
 The lower fixed pointer, corresponds to the fixed pointer of a 
 slide rule. The pointer on the M. E. P. scale may be placed at a 
 distance from its zero equal to log Ic so that with the M. E. P. 
 scale properly set the horizontal distance between these two pointers 
 equals log P log k. Now setting the " Revolutions " scale, so as to 
 bring r opposite the second pointer, it is obvious that the horizontal 
 distance between on this scale and the fixed pointer represents the 
 logarithm of the I. H. P. In order to obtain a direct reading of 
 this value, a third scale of horse power is used, which is read oppo- 
 site a pointer on the revolutions scale. In order to make the in- 
 strument more compact, this pointer is placed some distance to 
 the right of the on the revolutions scale, and the pointer on the 
 M. E. P. scale is found to the left of the true value of c, these 
 changes being compensated for by a corresponding shift in the posi- 
 tion of the horse power scale. 
 
 Though devised for use in connection with the mean pressure 
 indicator, this power board may be used with the ordinary engine 
 indicator and a slide rule of this type may be constructed for use 
 in calculating horse powers from brake or torsionmeter readings 
 of any particular engine, where the calculations are made from the 
 general formula 
 
 r>X D 
 
 Horse power = - -. 
 k 
 
 Hudson's Horse Power Scale. This is a compound slide rule 
 for horse power, similar to that described above, but in place of the 
 pointer on the lower fixed part of the instrument, as shown in Fig. 
 89, there is a scale of cylinder diameters, which makes the instru- 
 ment adapted to any engine within the limits of its construction. 
 
 Another form of logarithmic scale, adapted to the computation 
 of horse power, which may be readily constructed by anyone pro- 
 vided with the necessary drawing materials, is shown in Fig. 90. 
 
MEASUREMENT OF POWER 
 
 159 
 
 -I7O 
 '-'/SO 
 
 -/so 
 
 -/20 
 -110 
 
 -35OO 
 -3000 
 -2500 
 -2000 
 -/7SO 
 -/fOO 
 -/ZSO 
 
 500 
 400 
 350 
 300 
 250 
 ZOO 
 /7S 
 /39 
 /Z3 
 /OO 
 
 Zf 
 
 r 
 
 9 
 
 
 /o 
 
 H 
 FIG. 90. 
 
160 EXPERIMENTAL ENGINEERING 
 
 Writing the expression for horse power in the form 
 
 I. H. P.= P * R , or I.H.P.xfc='Pxfl. 
 We have 
 
 log I. H. P. + log k= log P + log R. 
 
 In Fig. 90 assume a base line AB and erect two perpendicu- 
 lars, AC and BD. On one of these, as AC, construct a scale to 
 represent various values of log P. On the other, construct a simi- 
 lar scale to represent various values of log R. Assume any two 
 values of P and R, such that AE = log P and BF=\og R. Draw 
 EF and bisect at G. Drop a perpendicular to the point H on AB. 
 Then 
 
 rj r_ AE+BF _ logP+logff _ log I. H. P. + log A; 
 22 2 
 
 Lay off jETJ=-f . It is obvious then that G lies at a point 
 a 
 
 such that Jg- logLH ' P '. By using a drawing scale of 1:2, 
 
 with the point J as the zero point, JG may be graduated to repre- 
 sent the I. H. P. on a logarithmic scale, such that for any two 
 values of P and R, falling on their respective scales at E and F, 
 the corresponding I. H. P. lies on the scale JG, at a point G where 
 this scale intersects with the mid point of the straight line joining 
 E and F. 
 
 An instrument constructed in this manner may be adapted 
 for general use by making the scale of I. H. P. adjustable in a 
 vertical line. The value of HJ, or Je, corresponding to the particu- 
 lar cylinder of the engine for which it is desired to use the instru- 
 ment is then read downward from J, and the scale of I. H. P. set 
 with the proper value of k falling at H. The instrument is then 
 ready for use. When constructed for use in calculating the indi- 
 cated horse power of the main engines, the various values of log fc 
 may, for convenience, be laid off and marked for each end of each 
 cylinder, instead of in the form of a graduated scale. In order to 
 make the diagram more compact, the base line may also be moved 
 upward, corresponding to the minimum values of P and R that will 
 be obtained in practice. 
 
MEASUREMENT OF POWER 161 
 
 The diagram in Fig. 90 is laid down for a value of A; = 16. 8, 
 corresponding to a cylinder 25 inches in diameter and 2-ft. stroke. 
 
 Brake Horse Power. This term is generally applied to the horse 
 power actually transmitted to the shaft and is so-called from the 
 brake which absorbs and measures the power transmitted. Shaft 
 Horse Power is equivalent to the brake horse power, but it is 
 measured in transmission without being absorbed by the apparatus 
 used. This term is used by marine engineers to designate the 
 power delivered to the propeller. The power of a marine turbine, 
 measured by the torsion meter,, is shaft horse power. 
 
 Absorption Dynamometers. 
 
 The Prony Brake. This is the most common form of absorp- 
 tion dynamometer. Its construction is simple and it is usually 
 extemporised when needed for the test of an engine. 
 
 Eef erring to Fig. 91 two wooden beams are clamped together, 
 
 T 
 
 \G 
 
 FIG. 91. 
 
 on a friction wheel carried by the shaft A, by bolts (7(7. DD are 
 stops to prevent the beam turning through more than a small 
 angle when the shaft is rotated. The long end of the lever G rests 
 on a platform scale, while the short end, F f may or may not carry 
 a hook on which weights are hung to balance the overhanging 
 weight of the long end. 
 
 Suppose the shaft turning to the left. If the bolts CO be 
 tightened, there is increased friction on the brake wheel with 
 corresponding increased pressure at G. If the weight of the beam 
 is exactly balanced at F, the pressure at G multiplied by the length 
 
162 EXPERIMENTAL ENGINEERING 
 
 of the arm a is the moment which must be overcome in turning the 
 shaft. In one revolution the work done = Pxax27r and the brake 
 horse power, 
 
 TJ TJ T> ^ a XPXR 
 
 33000 ' 
 
 where a is the length of the arm in feet, P is the pressure at G in 
 pounds, and R is the number of revolutions per minute. 
 
 The Water Brake. The friction wheel and band of the Prony 
 brake may be replaced by a runner and casing similar to the parts 
 of similar name in a centrifugal pump. The blades on the runner 
 are specially designed to give an inefficient pump and when water 
 is admitted, the power of the engine is dissipated in churning the 
 water, and this effect is greatly increased by restricting the flow. 
 The casing is balanced on bearings which permit a slight rotary 
 movement. An arm on the casing has its end resting on a plat- 
 form scale and the tendency of the casing to rotate is measured by 
 the pressure on the scale. The power is calculated in the same 
 manner as for the Prony brake. 
 
 Small water brakes are in use in the laboratory for measuring 
 the power of small engines. They are manufactured for testing 
 large steam turbines and motors up to a capacity of several thousand 
 horse power. 
 
 Electric Brake. A dynamo may be fitted to be used as a brake. 
 The armature shaft is connected to the engine under test. The 
 field is mounted on bearings so that it can revolve freely through 
 a small angle about the armature shaft and has a brake arm at- 
 tached to it. When the dynamo is operated the power exerted 
 tends to rotate the armature and is measured as in the case of the 
 Prony brake. 
 
 Electric Horse Power. When an engine is used to drive a dy- 
 namo in regular service the output of the dynamo is a measure of 
 the work done, and is obtained by reading the voltmeter and 
 ammeter. Volts X amperes = watts. Since 1 horse power = 746 
 watts, we have 
 
 ^ , . , volts X amperes , . x 
 
 Electric horse power = l - . (1) 
 
MEASUREMENT OF POWER 
 
 163 
 
 Remembering that the efficiency of a machine = 
 useful work done 
 
 total energy received, 
 
 we have 
 
 Efficiency of the dynamo= gectric horge power 
 
 snait horse power 
 
 (2) 
 
 Therefore dividing (1) by the efficiency of the dynamo, which 
 should be known as the result of previous tests, we may obtain the 
 shaft horse power. By further dividing the result thus obtained 
 by the known efficiency of the engine, we obtain a close approxima- 
 tion to the indicated horse power. 
 
 This may be better understood by reference to the accompany- 
 ing diagram, Fig. 92, in which 
 
 T) TT T> 
 
 EI = Efficiency of engine = ' ' ' , 
 
 or 
 
 E 2 = Efficiency of dynamo = 
 
 E. H. P. 
 I.H. P. ' 
 
 E. H. P. 
 
 E. H. P. 
 
 I.H.P.= 
 
 Engine 
 
 Shaft 
 
 Dynamo 
 E. H. P. 
 
 I.H.P. 
 
 B. H. P. 
 
 FIG. 92. 
 
 
 
 Transmission Dynamometers. 
 
 Under this head is classed a wide variety of instruments used 
 for the determination of the power transmitted by a shaft while 
 employed in doing the work for which it was designed. No work 
 is absorbed by such instruments other than that necessary to move 
 them. 
 
 Belt Dynamometers, 'Where power is transmitted by a belt if 
 it is arranged to measure the tension on each side of the belt,, the 
 
164 
 
 EXPERIMENTAL ENGINEERING 
 
 difference in such tensions multiplied by the speed of the belt gives 
 the work done. Owing to the loss of power due to the stiffness of 
 the belt, and the uncertainty due to slipping, such dynamometers 
 have not been used extensively. 
 
 The Kenerson Transmission Dynamometer. This instrument, 
 the invention of Prof. Kenerson, of Brown University, is used for 
 measuring the power transmitted by a shaft. Pig. 93 shows a 
 
 1*o Gage 
 
 FIG. 93. Dynamometer Shown in Section. 
 
 sectional view of the instrument, while Figs. 94 and 95 show the 
 instrument as applied to the driving shaft of an automobile. The 
 lettering of corresponding parts in Figs. 93 and 95 is the same. 
 
 The couplings A and B, each keyed to its respective shaft, are 
 held together loosely by the stud bolts C. The holes in the flange A 
 are larger than the studs C, so that these studs have no part in 
 transmitting power from one shaft to the other. The power is 
 transmitted from A to B through the agency of the latches L, four 
 of which are arranged around the circumference of the flange B. 
 These latches are mounted and are free to turn on the studs E. 
 The two fingers of the latches engage the studs F on the flange A. 
 On the ends of each latch are knife edges parallel to the stud about 
 
MEASUREMENT OF POWER 
 
 165 
 
 s 
 
166 
 
 EXPERIMENTAL ENGINEERING 
 
 which the latch turns. For either direction of rotation of the flange 
 A the latches L, which are in effect double bell-crank levers, will 
 exert a pressure on the disc G, tending to force it axially along the 
 
 FIG. 96. 
 
 hub of the coupling B, and this pressure, it will be seen, is pro- 
 portional to the torque. 
 
 Between the end-thrust ball, or roller, bearings MM, is held 
 the stationary ring S, which is the weighing member. is a thrust- 
 collar screwed on the hub of B, and 'P is its check nut, which is 
 
MEASUREMENT OF POWER 
 
 167 
 
 ordinarily pinned to the hub when in position. The stationary 
 member S, in the form of a ring surrounding the shaft, is pre- 
 vented from rotating by fastening to some fixed object the attached 
 arm shown in the view (Fig. 94) of the assembled instrument. In 
 this ring is an annular cavity covered by a thin, flexible copper 
 diaphragm D, against which the ball-race of one of the thrust-bear- 
 
 PIG. 97. 
 
 ings presses. The edge of this ball-race is slightly chamfered to 
 allow some motion to the diaphragm. The cavity is filled with a 
 fluid, such as oil, and connected by means of a tube to a gage. 
 The oil pressure measured by the gage is proportional to the pres- 
 sure between the thrust-bearings, which in turn is proportional to 
 the torque. 
 
 The Emerson Power Scale. This instrument is made to go on a 
 belt-driven shaft next to the loose-belt pulley. The two studs CC, 
 shown in Figs. 96 and 97, engage with the spokes of the loose 
 
168 
 
 EXPERIMENTAL ENGINEERING 
 
 pulley and cause the plate E to run with it. E revolves freely ex- 
 cept for the bolt G which may be pushed out to engage with it by 
 means of the slide H. When this is done the rim of the apparatus 
 is made to turn and through the system of levers, shown in Fig. 96, 
 this motion is transmitted to the shaft. These levers are a part of 
 a weighing mechanism, very similar in construction to that con- 
 tained in standard platform scales. When the shaft is in opera- 
 tion the torque is registered by the weighing mechanism. This, 
 multiplied by the revolutions per minute and the constant of the 
 instrument, gives the power. 
 
 The revolutions of the shaft are registered on the dial shown in 
 Fig. 96. 
 
 As used in the laboratory in connection with an explosion en- 
 gine, the apparatus is not in any way secured to the line of shaft- 
 ing. The line of shafting projects a short distance into the shaft of 
 the power scale to act as a support and a guide. The shaft of the 
 power scale ends in a sleeve that fits over the end of the line shaft. 
 
 The power is worked out by knowing the radius of the first 
 transmission of the power and the revolutions. For the special 
 apparatus in the laboratory it is as follows : 
 
 'Circumference of point of application of power. . 6 feet. 
 
 Revolutions a 
 
 Weight read on scale b 
 
 6xax& Circ. X Wt. X Rev. T TT T> 
 
 33000 
 
 33000 
 
 The table below gives the capacity at 100 revolutions: 
 
 Hole, 
 Inches. 
 
 Greatest 
 Diameter. 
 
 Weight. 
 
 Capacity. 
 
 Circle of 
 Graduation. 
 
 No. 2 
 
 26 inches. 
 
 260 Ibs. 
 
 27 h. p. 
 
 6 ft. 
 
 An allowance must be made for the reading of the power scale 
 when running with no resistance at each number of revolutions. 
 This must be tested and a table made. The net pressure is the dif- 
 ference between that under power and light. 
 
MEASUREMENT OF POWER 
 
 169 
 
 Torsionmeters. When power is transmitted through an elastic 
 shaft, within its elastic limit, the shaft is twisted through an angle, 
 the amount of such distortion depending on the size of the shaft 
 and the power transmitted. In any given shaft there is a fixed 
 proportion between the amount of twist and the applied moment of 
 rotation. Having calibrated the shaft, it remains to provide an 
 instrument to measure the amount of its distortion when running. 
 From this and the revolutions of the shaft the power is calculated. 
 
 FIG. 98. 
 
 Calibration of Shafting. Previous to being placed in position in 
 the ship, the shaft should be calibrated. This should be done by 
 coupling the various lengths together, including the length over 
 which the torsion is to be measured, placing it in bearings and pro- 
 ceeding in the following manner: One end is clamped rigidly so 
 that it cannot turn, and to the other is attached a lever to the end 
 
170 EXPERIMENTAL ENGINEERING 
 
 of which weights are attached. A pointer is also attached to the 
 free end of the shaft, about 6 or 8 feet in length and stiff enough 
 so that it will not bend under its own weight. A paper scale is 
 placed under the end of the pointer to register its movement in 
 inches. 
 Eeferring to Fig. 98 let 
 
 c movement of the pointer in inches under the influence of 
 
 the added weight W, applied to arm a. 
 r length of pointer in inches. 
 
 I length of shaft being twisted, exclusive of couplings. 
 L = length of portion of shaft on which torsionmeter is 
 
 applied. 
 
 R radius of arc on which torsionmeter readings are taken. 
 C = reading of torsionmeter corresponding to movement of 
 
 pointer c. 
 Extending the lines of the figure as shown, we have 
 
 / C 1 r 
 
 = -s- and x- -r=- 
 
 from which 
 
 and 
 
 C LXR c L_ JR_ 
 
 c ' IXr ' X I X r 
 
 the angle of torsion of length L of the shaft, expressed in circular 
 measure. 
 
 The moment M, caused by the pull W on an arm of length a 
 Wxa. From the results of the calibration experiments, a curve 
 is constructed, the ordinates giving the values of M in foot pounds 
 and the abscissae the corresponding readings of the torsionmeter. 
 From this curve the moment, corresponding to any reading of the 
 torsionmeter, is obtained. 
 
 The shaft horse power (S. H. P.) is equal to 
 
 2 X TT x M X Revs. _ M X Revs. 
 33000 ~ 5252 ' 
 
MEASUREMENT OF POWER 171 
 
 Horse Power from the Torsionmeter Without Calibration of 
 Shafting. From applied mechanics we have the angle of torsion 
 for a shaft within the elastic limit, 
 
 K/f v T fi 
 
 $~W T~~I = ~rc\ > wnere #=the angle of torsion meas- 
 LJ x 1 X d ooU 
 
 ured in degrees. 
 
 M= the twisting moment in inch pounds. 
 L length of shaft in inches. 
 
 E = modulus of elasticity of the material of the shaft. 
 d = diameter of shaft in inches. 
 
 I moment of inertia of the section about the axis= ^ 
 
 Xd 3 for solid shafts or^f ~ 7 1 ) where d = out- 
 16 \ d ) 
 
 side diameter and d i = inside diameter of shaft, both 
 in inches. 
 
 It is more convenient to express M in foot pounds. Making this 
 change, we have 
 
 ,,-_ ir _ 7r fi 
 
 360xl2x ~ 16 X 
 
 for solid shafts, or 
 
 M- * 
 
 ' 16 
 
 for hollow shafts. 
 
 Substituting these values of M in the above expression for horse 
 power, we have for solid shafts 
 M xRevs. 
 
 33000 ~ 360 X 16x12x33000 XL' 
 
 For high-grade steel, such as is used for the shafting of naval 
 vessels, E is approximately 11,750,000. 
 Substituting this and evaluting, we have 
 
 Q TT p_<2 4 X0X Revs. 
 
 for solid shafts and similarly 
 
 S H P --(^ 4 
 
 3.13XL 
 
 for hollow shafts, the terms being named as before. 
 
172 
 
 EXPERIMENTAL ENGINEERING 
 
 In any particular installation of a torsion meter, these formulae 
 are applicable, the expression in each case taking the form 
 
 S. H. P. 
 
 where K is a constant and C is the torsionmeter reading. This ex- 
 pression is general and applies in all cases where C is proportionate 
 to the angle of twist, whether it be measured in degrees, or in 
 linear units on an arc of given radius. 
 
 FIG. 99. 
 
 Results obtained by calculation from the modulus of elasticity 
 approximate closely to those obtained by calibration, their accuracy 
 depending of course on the correctness of the figure used for the 
 modulus. 
 
 The Navy Department requires the shafting of its vessels to be 
 calibrated before installation. This is more necessary with hollow 
 than with solid shafting, since for the formula for hollow shafting 
 to be correct, the hole must be centered with exactness. 
 
 The Denny-Johnson Torsiometer. Two bronze wheels, A and 
 B, Fig. 99, are placed on the shaft at a definite and known distance 
 apart, the distance being as great as possible. On each of these 
 wheels a permanent magnet, with a sharp chisel-shaped edge is 
 
MEASUREMENT OF POWER 173 
 
 fixed radially, at the periphery of the wheel, and with the sharp 
 edge parallel to the shaft. Under one of these wheels, soft iron 
 sector cores are placed, wound with a series of coils of very thin 
 wire, so fine indeed that each coil, with its dividing wall in one of 
 the sectors, only occupies a space of 0.02 inch. Under the other 
 wheel similar sectors, or inductors, as they are called, are fitted, 
 but in this case the coils are further apart, viz., 0.2 inch. 
 
 In the lower part of Fig. 99 is shown the arrangement of recording 
 boxes. For each shaft this consists of two circles, fitted with con- 
 tact studs and movable contact arms; the one marked E is con- 
 nected with the inductor at the wheel next the propeller and has 
 six contacts. That is to say, each of the six coils, spaced 0.2 inch 
 apart, is connected to one of the studs. The other one, marked F, 
 has thirteen contact studs, connected to the thirteen coils, spaced 
 0.02 inch apart, on the inductor at the turbine end of the shaft. 
 
 The magnets are so placed that when passing the inductors, cur- 
 rents of opposite potentiality are induced. These are carried to a 
 telephone receiver, and when either magnet passes an inductor a 
 click is heard. If they pass simultaneously, the two currents tend 
 to neutralize one another and the single click is greatly diminished, 
 or vanishes altogether. There are two resistance boxes, not shown 
 in the figure, for throwing into series with the differential wind- 
 ings of the telephone receiver and the two inductor circuits, which 
 circuits must be accurately balanced before absolute silence can be 
 obtained in the receiver. 
 
 To take a reading after the instrument is once set, and the 
 resistance of the two telephone circuits is adjusted, all that is neces- 
 sary is to turn the movable arm on scale F round the various studs 
 until there is silence in the telephone, when the amount of torsion is 
 immediately read of? the scale. If no such position be found, it 
 means that the shaft is being twisted more than is covered by this 
 scale ; the arm E is then turned to the first contact, and the arm F 
 is again swept round the circuits. If silence be still not obtained, 
 the arm E is turned to the second contact, and so on, the combined 
 range of the scales being altogether 1.24 inches, which is more than 
 sufficient to measure the maximum torque usually obtained. From 
 torsion experiments on the shaft, made previous to its being fitted 
 12 
 
174 EXPERIMENTAL ENGINEERING 
 
 in the ship, the factor by which this reading is to be multiplied is 
 obtained, and the power gotten by a simple multiplication. 
 
 The instrument is made for one, two, three, and four shafts; 
 a group of scales and resistances, as described above, being required 
 for each shaft, all mounted on one panel. By means of the contact 
 arms and studs the various shafts are thrown into circuit with their 
 receivers and readings taken from all the shafts in a very short time. 
 The sound in the receiver at usual revolutions is so distinct that 
 even an untrained observer, after a few minutes' practice, can get 
 perfectly accurate results, and powers transmitted can be obtained 
 from moment to moment. The instruments are mounted in a room 
 in some quiet part of the ship where the observer will be free from 
 interruption. 
 
 Instrument for Low Speeds. The above instrument is not suited 
 for very low revolutions, as the induced current becomes too weak 
 to make a distinct sound in the telephone; but it is suitable down 
 to about 100 revolutions per minute. For lower revolutions a 
 somewhat similar arrangement has been used in which the two 
 wheels keyed on the shaft were on insulating material, and each had 
 a contact point arranged at its periphery in such manner that the 
 point made a momentary contact with a metallic tongue or brush 
 once in each revolution. The contact points were connected to the 
 shaft, and the metal brushes to a battery and a telephone receiver. 
 The method adopted was to first adjust the brushes so that both 
 made contact with the points simultaneously when the shaft was 
 revolving, but transmitting no power. When transmitting power 
 the shaft was of course subject to torsion and thus the brushes were 
 put out of simultaneous contact. One of the brushes was then 
 moved arouad its disc concentrically, until simultaneous contact was 
 once more established. The amount of this shift gave a measure of 
 the torque on the shaft, and to ascertain the correct amount of this 
 shift the telephone receiver was placed to the ear, no sound being 
 heard except when both brushes were in contact with the respective 
 contact points. 
 
 This apparatus, though more simple than that above described, 
 is not applicable on shafting running at high speed, since it is im- 
 possible to obtain the make and break with certainty. Experiments 
 
MEASUREMENT OP POWER 
 
 175 
 
176 EXPERIMENTAL ENGINEERING 
 
 with it on high speed shafting led to the perfection of the Torsio- 
 meter, above described. 
 
 The Denny-Johnson Torsiometer is largely used on the earlier 
 turbine-driven vessels. It is accurate and satisfactory when care is 
 taken with its installation. The numerous electrical connections 
 sometimes give trouble, chiefly through short circuiting with salt 
 water. 
 
 The Bevis-Gibson Flashlight Torsionmeter. Eeferring to Fig. 
 100, the shaft has two discs clamped on it in positions as far apart 
 as practicable. These discs are each pierced, around a circum- 
 ference of the same radius, with twelve radial slots. Behind each 
 of the discs there is a fixed support above the shaft. On one of 
 these a lamp is placed and on the other there is an eye piece, with 
 shutter, mounted on a movable " finder." Both lamp and eye 
 piece are at the same distance from the center of the shaft. 
 
 When the shaft is at rest, with the finder in the zero position, the 
 lamp, similarly placed slots in the two discs, and the eye piece, 
 are in line, and the light from the lamp is seen by the observer 
 through the eye piece. If the shaft were turned rapidly, without 
 torsion, the light would be seen as a continuous ray. 
 
 When the shaft is transmitting power, the relative position of the 
 two discs is changed, and the lamp and the eye piece are no longer 
 in line with similarly placed slots, so that the ray is cut off. To 
 find it again it is necessary to move the finder along a graduated 
 scale, fitted with a vernier for close reading. The amount of move- 
 ment required to bring the ray again into coincidence is propor- 
 tionate to the angle of torque. 
 
 Fig. 101 shows the principle of operation and the method of pick- 
 ing up the deflected ray of light by moving the finder. 
 
 The flash light torsionmeter is free from complicated mechanical 
 parts, which in some other torsionmeters introduce error through 
 lost motion. There may be considerable variation in the readings, 
 due to the breadth of the beam of light as seen through the eye 
 piece, but in the hands of a careful observer very accurate results 
 can be obtained with it. The edge of the beam of light should be 
 caught, first on one side, then on the other, the mean of the two 
 giving the desired reading. 
 
MEASUREMENT OF POWER 
 
 177 
 
 The instrument does not give a permanent record and it is not 
 always possible to find for it the amount of space in the fore and 
 aft direction that it requires. 
 
 The Fottinger Torsionmeter. Fig. 102 shows the working parts 
 of this instrument. The ring B with two arms and the ring D at 
 the after end of the steel sleeve C are both clamped firmly to the 
 shaft. On the forward end of the sleeve is a ring A similar to B. 
 There is a clearance between this ring and the shaft, and it is cen- 
 tered on B by means of radial centering rods having knife edges 
 on each end. The sleeve C and ring A act as a rigid mass and turn 
 
 Find* 
 
 2 PLAN. 
 
 Shaft. Transmitting Power 
 Light obscured. 
 
 3 PLAN, 
 
 Shaft, Transmitting Power 
 
 " Finder" moved through an angle equal to amount of Torque 
 Light visible 
 
 FIG. 101. 
 
 with that part of the shaft to which D is clamped. When a tor- 
 sional stress is applied to the shaft the ring B twists relative to 
 A. Through a flexible link E, a bell crank H pivoted on A, and a 
 link J connected to the magnalium ring N which floats on the sleeve 
 C, this relative twist is magnified and changed to a fore-and-aft 
 motion. A shoe P is brought into contact with a flange on the 
 magnalium ring, and the fore-and-aft motion of the ring is again 
 magnified and indicated by a pointer Q on a scale T. 
 
 The constant for the machine is found by calibrating the shaft 
 and torsionmeter in the shop previous to installation in the ship. 
 The after end of the shaft is bolted to a fixed flange. The forward 
 end rests in a roller bearing and on the forward flange is bolted a 
 double lever. A known twisting moment is applied to this lever 
 
178 
 
 EXPERIMENTAL ENGINEERING 
 
MEASUREMENT OF POWER 
 
 179 
 
 and the corresponding torsionmeter readings are observed. From 
 the slope of the line plotted between torsionmeter readings and 
 torque a constant is obtained such that 
 
 in which H. P. = shaft horse power, R = revolutions of the shaft 
 per minute, and F= torsionmeter reading. 
 
 This torsionmeter requires but little space and gives a direct in- 
 dication of the torque. A later form is designed to give a perma- 
 nent record. It is open to the objection that the numerous levers 
 and joints introduce lost motion with corresponding errors of ob- 
 servation. 
 
 FIG. 103. Torsionmeter Mounted Complete on Shaft. 
 
 The Hopkinson-Thring Torsionmeter. The principal of this 
 apparatus is a differential one, and consists in the observation of 
 the twist between two points on the shaft by means of two beams 
 of light projected on to a scale from a fixed and a movable mirror. 
 The beam projected on the scale by the fixed mirror is taken as the 
 zero point, while the beam projected by the movable mirror indi- 
 cates the amount of torque on the shaft. Both mirrors revolve with 
 the shaft, but even at moderate speeds the reflections appear as 
 continuous lines of light across the scale. 
 
 The torsionmeter is shown in Fig. 103, mounted complete on a 
 
180 EXPERIMENTAL ENGINEERING 
 
 shaft, and the scale box in Fig. 104, while a diagrammatic arrange- 
 ment of the complete apparatus is shown in end elevation and plan 
 in Figs. 105 and 106 respectively. A collar A, clamped to the 
 shaft of which the torque has to be measured, is provided with a 
 flange projecting at right angles to the shaft and an extension 
 (Fig. 106.) 
 
 A sleeve B (Fig. 106), provided with a similar flange and ex- 
 tension at one end, is clamped at its further end on to the shaft in 
 such a manner that its flange is close to that on the collar A, while 
 its entension overlaps that of the collar A, on which it is supported 
 to keep it concentric. Both the collar and sleeve are quite rigid, 
 and it is obvious that when the shaft is twisted by the transmission 
 
 FIG. 104. Scale Box with Lamp, for Torsionmeter. 
 
 of power, the flange on the sleeve B will move relatively to that on 
 the collar A, the movement being equal to that between the two 
 parts of the shaft on which these fittings are clamped. This move- 
 ment is made visible by a system of torque mirrors mounted be- 
 tween the two flanges, which reflects a beam of light, projected from 
 a lantern, on to a scale divided in a suitable manner on ground 
 glass. 
 
 This system of torque mirrors consists of a mounting, pivoted 
 top and bottom on one or other of the flanges, in which two mirrors' 
 are arranged back to back. This mounting is provided with an 
 arm, the end of which is connected by a flat spring to an adjustable 
 stop on the other flange. Any relative movement of the two flanges 
 
MEASUREMENT OF POWER 
 
 181 
 
 TORQUE MIRROR 
 
 LAMP 
 
 GLASS 
 
 -ZERO M'RRORS 
 
 FIG. 105. 
 
 LLAR A 
 
 FIG. 106 
 
182 EXPERIMENTAL ENGINEERING 
 
 will turn the torque mirror and thereby cause the beam of light to 
 move on the scale, the deflection produced being directly propor- 
 tional to the torque applied to the shaft. Hence, if the rigidity of 
 the material and the number of revolutions per minute are known, 
 the H. P. transmitted can be readily calculated. 
 
 With the arrangement described, a reflection will be received from 
 each mirror at every half revolution of the shaft; but where the 
 torque varies during a revolution (as with reciprocating engines), 
 a second system of mirrors may be arranged at right angles to the 
 first system, so that four readings can be taken during one revolu- 
 tion ; or, if two scales are used, eight readings can be taken. 
 
 Fig. 105 shows how the beam of light reflected by the mirror 
 when in its highest position passes through the upper part of the 
 scale; while the second reflection will occur when the mirror is in 
 the position occupied by the zero mirror, the beam of light passing 
 through the lower part of the scale. The position of the torque 
 mirror in Fig. 106 is such that the reflected beam strikes the scale 
 to the right of the zero line, but when the shaft has made a further 
 half revolution, the reflected beam from the other mirror will strike 
 the scale to the left of the zero line. Obviously the deflection on 
 both sides should be equal. 
 
 The fixed mirror is attached to one of the flanges (in Fig. 106 to 
 the flange of the sleeve B). This must be adjusted so that the beam 
 of light reflected from it is received at the same point on the scale as 
 those from the movable mirrors when there is no torque on the 
 shaft. To facilitate the erection and adjustment of the apparatus, 
 the box containing the scale and carrying the lamp is fitted with 
 trunnions, so that it can be inclined as required. 
 
 If the position of the apparatus becomes altered relatively to the 
 scale owing to the warming up of the shaft or from other causes, 
 this is indicated immediately to the observer by an alteration in the 
 position of the zero as reflected by the fixed mirror. Hence, the 
 zero can be adjusted by moving the scale so that its zero coincides 
 with the reflection from the fixed mirror. It will be obvious that it 
 is not necessary to move the scale, as the mean of the two readings 
 will be the same. It will readily be understood that a movement of 
 the torque mirrors can only occur through a relative movement of 
 
MEASUREMENT OF POWER 
 
 183 
 
 the two flanges, so that vibration of the shaft or of the ship will not 
 influence the readings. 
 
 The constant of the instrument, viz., the factor which, when 
 multiplied or divided into the product of the torsionmeter reading 
 and the revolutions, gives the horse power, may be calculated within 
 2 or 3 per cent, if the section of shaft within the instrument is 
 uniform. A direct calibration of the shaft with the instrument in 
 position is recommended before the former is put into the ship. 
 This is effected readily by applying a known twisting couple. 
 
 This instrument is inexpensive and the absence of complication 
 makes it fairly accurate, but there is no permanent record. Con- 
 siderable transverse space is required for the scales, which must be 
 darkened to observe the readings. It is widely used on vessels of 
 the British ISTavv. 
 
 FIG. 107. 
 
 The Metten Torsionmeter. This instrument was designed by 
 Chief Engineer Metten of the Wm. Cramp & Sons Ship and 
 Engine Building Company of Philadelphia. The principal parts 
 are shown in Fig. 107. A is a sleeve, fitting loosely on the shaft 
 at the end shown, and rigidly attached to the shaft at the opposite 
 end. This sleeve is made in halves for convenience in assembling. 
 Two extension arms, Y, are attached, forming part of the sleeve. 
 One of these is for attaching to the mechanism, while the other 
 serves merely to balance the apparatus. B is a disc securely fixed 
 to the shaft, close to the arm Y, as shown. Pivoted on this disc, 
 at Z, is a very light frame, F f carrying two knives, K, for making 
 
184 EXPERIMENTAL ENGINEERING 
 
 the record. At the apex of frame F is a crank pin, S, with flat con- 
 necting link, L, to a pin, P, at the upper end of Y. The pin P is 
 eccentric and can be rotated while making the zero adjustment of 
 the instrument. The clamp, C, serves to hold P rigidly in position 
 after the adjustment is made. is a grease cup for lubricating 
 the pivot bearings at Z.. 
 
 When the shaft is transmitting power, P and Y tend to rotate 
 slightly with reference to each other. The frame, F, is caused to 
 swing about Z, the proportions of the frame being such that the 
 knives, K, are each given a movement fifteen times that of the pin 8. 
 One knife moves from the zero line toward the center of the shaft, 
 while the other moves outward. A strip of paper is held in a station- 
 ary clamp in front of the disc, so that the knives in passing make 
 cuts in the edge of the paper. The distance between these cuts will 
 be thirty times the linear movement of S, and is a direct measure 
 of the torque in shaft. 
 
 The constant of the instrument is determined in a manner similar 
 to that used for other torsionmeters. 
 
 Fig. 108 shows the table for holding the paper on which records 
 are made with this torsionmeter. The table proper, 3, is on a 
 frame, II, secured by a hinge 5, to the fixed base I. The clamp, 8, 
 holds the paper firmly with its edge parallel to the disc of torsion- 
 meter. The set screw, 6, enables the distance of edge of paper 
 from face of disc to be adjusted, without altering the setting of 
 the paper in the clamp. 
 
 Torsionmeters Compared with the Indicator. The shaft horse 
 power, obtained by the use of the torsionmeter, can be compared 
 with the indicated horse power by multiplying the latter by a factor 
 representing the efficiency of the engine. The factor 0.92 is used 
 in ordinary calculations. There is no means of measuring directly 
 the indicated horse power of a turbine, though an approximate re- 
 sult can be obtained by observing the fall in pressure through suc- 
 cessive expansions, by means of gages attached to the turbine stages, 
 and taking corresponding observations to determine the quality 
 of the steam. The results of such observations are entered on a 
 temperature-entropy chart, thus completing a diagram representing 
 the cycle of heat operations in the turbine. This work requires con- 
 siderable time and labor. 
 
MEASUREMENT OF POWER 
 
 185 
 
 FIG. 108. Card Table for Metten Torsionmeter. 
 
186 EXPERIMENTAL ENGINEERING 
 
 Torsion Readings for Reciprocating Engines. Perfect results 
 can be obtained with the torsionmeter only where the turning mo- 
 ment is uniform,, as it is in turbine driven machinery. In a re- 
 ciprocating engine, due to the irregular twisting moment, it is 
 difficult of application, though good results have been obtained by 
 fitting as many as six sets of instruments at regular intervals 
 around the periphery of the discs placed on shafts. By taking 
 the mean of the six readings, fairly accurate results are obtained. 
 The apparatus has been thus fitted for experimental purposes on a 
 few vessels, and the accuracy of the torsionmeter thus demonstrated. 
 The arrangement is too cumbersome for regular application on a 
 vessel driven by reciprocating engines. 
 
 Curves of Chest Pressures and Horse Power. For a vessel in com- 
 mission curves are sometimes constructed showing the observed 
 shaft horse power for given values of the steam-chest pressure and 
 vacuum. Such curves cannot be depended upon absolutely for es- 
 timating horse power since the revolutions will vary, depending upon 
 conditions of the ship's bottom and the surface of propellers, state 
 of the sea, etc. Such curves are of great value, however, in indi- 
 cating the condition of the turbine blading. Any considerable in- 
 jury to the blading would be marked by a corresponding decrease in 
 efficiency which would be indicated on examination of the curves. 
 
CHAPTER VIII. 
 TESTING MATERIALS OF CONSTRUCTION. 
 
 Strength of Materials. In all engineering work of whatever 
 character, the question of the strength of the materials employed is 
 of great importance. In designing the machinery of a war vessel 
 the naval engineer is confronted with the necessity of reducing 
 weights as much as possible, but he must, while doing this, make 
 the machinery of sufficient strength to stand any load that may be 
 put upon it in an emergency. The materials employed must be of 
 the highest grade as regards strength, and of such uniform character 
 as to permit the reduction of the factor of safety as much as pos- 
 sible. Specifications for material are closely drawn and much of 
 the work of the engineer lies in the inspection of such material to 
 see that it conforms to specifications. 
 
 Stress is the uniformly distributed force applied to a material. 
 It is of three different kinds : - longitudinal, divided into tension 
 and compression; transverse, divided into shearing and bending; 
 and, twisting or torsional. Stress is measured by the number of 
 pounds of load per square inch of cross section. 
 
 Strain is the distortion of a material resulting from the ap- 
 plication of stress and follows in its classification the different kinds 
 of stress above enumerated. In elastic material it is proportional to 
 the stress. 
 
 Elasticity is that property of a material that causes it to return 
 to its original form when the forces acting upon it have been re- 
 moved. This property is possessed to a limited extent only, by 
 most materials, and if the deformation, or strain, exceeds a certain 
 amount, the material will not again regain its original form. 
 
 Elastic Limit is the critical point beyond which the material 
 cannot be strained without a permanent distortion or set. This 
 point, when gradually approached, in most materials is indicated by 
 an increase in the increment of strain due to a constant increment 
 of stress. 
 
188 EXPERIMENTAL ENGINEERING 
 
 Rigidity or Stiffness is the property by means of which, bodies 
 resist change of form. 
 
 Coefficient of Ultimate Strength is the number of pounds per 
 square inch required for rupture, and is obtained by calculation 
 from the original area and the maximum load. 
 
 The Coefficient of Strength at the Elastic Limit is the number of 
 pounds per square inch acting on the material when the elastic 
 limit has been reached. 
 
 Percentage of Elongation. The elongation is the total relative 
 strain; or the amount that a piece stretches before rupture. It is 
 usually expressed as a percentage of the original length of the test 
 piece. 
 
 Reduction of Area of Cross Section. After fracture of a test 
 piece, the area at the point of fracture is measured and the reduction 
 in area is expressed as a percentage of the original area. 
 
 Modulus of Elasticity. This is the number expressing the ratio 
 of the stress per square inch to the deformation per inch accom- 
 panying that stress, within the elastic limit. It is denoted by E. 
 
 Modulus of Resilience. This is the amount of work, in foot 
 pounds, done in deforming a cubic inch of the specimen up to 
 elastic limit. It is equal to one-half the stress per square inch at 
 elastic limit multiplied by the total elongation in feet per inch of 
 length up to elastic limit; or to the square of the stress at elastic 
 limit divided by twice the value of E. 
 
 Breaking Load and Maximum Load. As the load on a test piece 
 under tension is gradually increased, the piece will be seen to visibly 
 diminish in diameter, this diminution being at first uniformly 
 distributed along the whole length of the piece. When the maxi- 
 mum load is reached, this reduction in area assumes a maximum 
 amount at one point, where a distinct thinning takes place. The 
 area then becomes rapidly reduced at this point, and unless the 
 total load is reduced, the load per unit cross section becomes greatly 
 increased, and the piece will no longer be able to support the maxi- 
 mum load. By careful manipulation, it is possible to so reduce the 
 load that the scale beam is kept balanced up to the point of actual 
 rupture, where the total load is called the 'breaking load. 
 
 The Safe Load must always be less than the load at the elastic 
 
TESTING MATERIALS OP CONSTRUCTION 
 
 189 
 
 limit. It is usually taken as a certain fraction of the ultimate or 
 breaking load. 
 
 The Factor of Safety is the ratio of breaking load to safe load. 
 
 The Strain Diagram. In testing a piece of material, if we lay 
 off the strain on the horizontal axis, to a scale that is readily 
 appreciable to the eye, and the corresponding loads as ordinates to 
 
 FIG. 109. 
 
 a convenient scale, as 3000 or 5000 pounds per square inch, a curve 
 drawn through the points thus plotted gives the strain diagram. 
 Such a curve is drawn autographically on the testing machine and 
 is shown in Fig. 109. The strain is represented by distances parallel 
 to OX, the load as a certain number of pounds per inch parallel to 
 OY. From to A this diagram is a straight line, showing that the 
 strain is proportional to the stress. At A there is a sudden increase 
 in strain without a "marked increase in load, shown by the curved 
 line from A to B. The point A is often spoken of as the yield point, 
 13 
 
190 EXPERIMENTAL ENGINEERING 
 
 and marks the elastic limit. In curves for ductile material, taken 
 autographically, this sudden increase of strain is usually accom- 
 panied by an apparent reduction of stress, as shown by the curve 
 from B to C. This is probably due to the fact that the increase in 
 strain is so great that the scale beam falls until the stress is in- 
 creased. The curve then continues to rise, reaching its maximum 
 height at D, which indicates the point of maximum load. From 
 D the load is reduced to E, which indicates the breaking load. 
 
 The Stress-Strain Diagram. This diagram is sometimes errone- 
 Dusly called the stress-strain diagram. The ordinates do not indi- 
 cate stress, but load. By observing and plotting the area of cross 
 section of the test piece and dividing the load by it, the amount of 
 stress can be calculated and the true stress-strain diagram can be 
 thus constructed. 
 
 Testing Machines. 
 
 Machines for the testing of materials may be said, in general, 
 to consist of (1) a power system, by which stress is applied to 
 the specimen, and (2) a weighing system, by which the stress ap- 
 plied is measured. The power system may consist of a train of 
 gears or an hydraulic cylinder, either of which may be operated by 
 power or by hand. The weighing system usually consists of a sys- 
 tem of levers and a poise by which the stress is balanced as on an 
 ordinary pair of scales. 
 
 The form of testing machine in general use is that in which the 
 load is applied through a train of gears and screws operated by 
 power, and the stress is measured by a system of levers. It is of 
 the vertical type, in which the tensional and compressional speci- 
 mens are held in a vertical position. The vertical screws connecting 
 with the power system operate a movable head, to which the lower 
 end of the tension specimen is connected. The upper end of the 
 specimen is connected to the upper head, which is a part of the 
 weighing system and rigidly connected with the table of the ma- 
 chine, the latter resting in turn on the lever system. In compres- 
 sional tests the specimen is placed directly between the movable 
 head and the table. For shearing and bending tests the specimen is 
 held between supports resting on the table and the stress is applied 
 *s for compressional tests. 
 
TESTING MATERIALS OF CONSTRUCTION 191 
 
 Kiehle Screw Power Testing Machines. Fig. 110 shows a Biehle 
 testing machine of 100,000 pounds capacity, as installed in the 
 laboratory. It is of the automatic and autographic type and is 
 described by the makers as follows : 
 
 The Straining Mechanism. The top head is supported by two 
 cast-iron columns, which rest on the weighing table. This table 
 in turn rests upon a series of eight hardened steel knife edges in 
 the main levers, these levers resting on steels which are fitted in 
 cast-iron bearings on the cover plate. Beneath this and bolted to 
 it is the cast-iron box, containing the main gears, and to which the 
 bracket supporting the beam and lever stands is attached. 
 
 Through holes in the table two pulling screws pass up and reach 
 nearly to the top head, running through two brass nuts in the pull- 
 ing head, which is raised or lowered according to which direction 
 the screws revolve. In the top head and pulling head are cut rec- 
 tangular holes, sloping on two sides. For tensile tests, the ends of 
 the specimen are held by hardened steel wedges, or " grips/' which 
 fit into these openings in the heads. Owing to their wedge shape, 
 the grips hold the specimen more firmly as the strain on it increases. 
 
 In compression tests the specimen is crushed between the two 
 crushing tools, one being attached to the under side of the pulling 
 head, and the other resting centrally on the table. 
 
 In bending tests, one V-shaped tool is attached to the lower 
 side of the pulling head, the other two V-shaped tools resting on the 
 table equidistant from the first, and as far from it as the operator 
 wishes. The specimen rests on the two tools placed on the table 
 and is bent or broken by the third tool pressing down upon it. In 
 shearing tests a similar arrangement is used except that in place of 
 the three V-shaped tools, the shearing tool, Fig. Ill, is employed. 
 The block which carries the knives and specimen rests on the table 
 of the testing machine, and the pulling head, carrying a crushing 
 tool, forces the upper knife through the specimen. The lower block 
 is cast iron, with a V-groove, in which the specimen is placed. The 
 two lower knives are exactly one inch apart and are held in the block 
 with a wedge, by which they are brought to the correct position. 
 The upper knife, which is movable and guided by the block, is one 
 inch in width so that it just fills the space between the two lower 
 knives, as it is forced through the specimen. 
 
110. 
 
194 
 
 EXPERIMENTAL ENGINEERING 
 
 The table is kept from jumping by the recoil bolts which pass 
 through its four corners, running up from the cover plate, and 
 are secured by nuts with thick rubber washers, which serve to take 
 up the shock occasioned by the breaking of large specimens. 
 
 The Weighing Apparatus. The table rests wholly upon the 
 main levers, the recoil bolts passing through it loosely, and any pres- 
 sure on it is transmitted directly through the levers to the beam. 
 All contacts, by which the load is transmitted to the beam, are in the 
 form of hardened steel surfaces resting on knife edges, by which the 
 friction is eliminated. A heavy weight hanging at the back end of 
 
 RIEHLE 
 
 ,- 
 
 FIG. 111. 
 
 the beam serves to counterbalance the weight of the other end, and 
 a counterpoise above this, with screw and hand wheel for adjust- 
 ment, permits the balancing of the beam when the poise is at zero. 
 
 When making a test, it is desirable to have the weight registered 
 on the beam simultaneously with the increase of the load on the 
 specimen, the equipoise of the beam being maintained as nearly as 
 possible. The poise should be first placed at zero and the beam bal- 
 anced, then as the load is added the poise should be advanced at 
 such a rate that the beam vibrates freely between the upper and 
 lower bars at the smaller and further end. The load on the speci- 
 men is only weighed accurately when the beam is in equipoise. 
 When the beam is against the upper bar the load is greater than the 
 
TESTING MATERIALS OF CONSTRUCTION 195 
 
 poise indicates,, and it is impossible to know just how much greater 
 until the beam is again balanced by the outward movement of the 
 poise. 
 
 The Screw and Vernier Beams. The beam is graduated in 1000 
 pound marks, but from the dial at the operating end of the beam, 
 subdivisions of 100 pounds and 10 pounds can be read. One revo- 
 lution of the screw moves the poise forward 1000 pounds. 
 
 The poise is attached to a carriage running along the top of the 
 beam on rollers, and is propelled by a screw sunk in the top of the 
 beam. The screw revolving, carries the poise out or back, as the 
 case may be, or the nut can be released from the screw and the poise 
 run back quickly by hand. 
 
 The Driving Mechanism. The two pulling screws pass down 
 through long bearings in the cover plate and to their lower ends are 
 keyed the main gear wheels. Between these gears and the cover 
 plate are ball bearings to take the thrust and reduce friction. Both 
 these wheels are driven by the same pinion on a vertical shaft, 
 which is driven through the large bevel wheel shown in cut. This 
 in turn is driven by a large bevel pinion, which receives its motion 
 through a series of gears from the pulley shaft down in cut. As 
 installed in the laboratory, this pulley is driven by a reversible elec- 
 tric motor. 
 
 The lever to right of pulley throws the motor in direct gear for 
 quick speeds. The small hand wheel to left of pulley throws in the 
 back gear, which can only be operated when the lever is vertical, 
 disconnecting the direct gear. 
 
 The lever to left of hand wheel operates to give two speeds, and 
 two speeds are given by the tumbling ball lever at base of machine. 
 There are thus seen to be eight speeds for the pulling head, which, 
 for a speed of 200 revolutions per minute for the pulley shaft, are 
 as follows : 
 
 Speed per min. B^ekGear. Lever. Tumbling Ball. 
 
 ^ inch Back Slow Slow 
 
 inch Back Fast Slow 
 
 1. inch Back Slow Fast 
 
 2 inch Direct Slow Slow 
 
 1| inches Back Fast Fast 
 
 2 inches Direct Fast Slow 
 
 3 inches Direct Slow Fast 
 10 inches Direct Fast Fast 
 
196 EXPERIMENTAL ENGINEERING 
 
 The fastest speed is used only for adjustment of the pulling head 
 to the desired position. If speeds slower than 1/10 inch are de- 
 sired, they can be obtained by stopping the motor and operating 
 the pulley shaft by hand. 
 
 The Automatic Apparatus. On the beam stand of the ma- 
 chine is the attachment for operating the poise automatically. A 
 small belt runs from the hub of the driving pulley. This sheave 
 turns a cast-iron disc which is arranged to give motion to either of 
 a pair of fiber-rimmed wheels placed near the periphery, on oppo- 
 site sides of the disc. These wheels are on the same spindle which 
 is connected by a belt to the wheel for operating beam screw. A pair 
 of electro-magnets are so placed that when current is turned on one 
 of them, an armature is attracted and one of the fiber-covered wheels 
 is brought against disc. With current on the other magnet the 
 other wheel is brought in contact. The end of beam has electrical 
 contact points, arranged so that when the beam rises the circuit is 
 completed that 'brings the wheel into action for forcing the poise 
 outward. When the beam falls this contact is broken and the poise 
 remains stationary until the lower electrical contact is made and the 
 other wheel is brought into action, forcing the poise backward. The 
 fiber-covered wheels may be shifted back and forth across the face 
 of the disc, thus giving variation to the speed of the poise. 
 
 The Autographic Apparatus. The extensometer for autographic- 
 ally recording the stretch or compression of a specimen is shown 
 attached to bracket fixed to the back of one of the columns. The 
 arm of extensometer carrying fingers is swung into position for 
 test, and adjusted vertically. The upper finger of extensometer is 
 placed under top spring clamp. The lower finger rests upon bottom 
 spring clamp, and moves with it as specimen stretches. The motion 
 of stretch is converted into circular motion and transferred by gear- 
 ing and a sliding shaft with universal joints to the paper drum 
 attached to beam stand above beam. The motion as transferred to 
 the paper is multiplied five times. A vertical screw in front of drum 
 carries the pencil attached to nut. This screw has a finer pitch than 
 beam screw to reduce load line on diagram, and moves simultane- 
 ously with the beam screw. The rotary motion of drum, due to 
 stretch of specimen, and the vertical movement of pencil, due to 
 load, form the strain diagram. 
 
TESTING MATERIALS OF CONSTRUCTION 197 
 
 The fingers of extensometer should be swung to one side just 
 before the specimen breaks, in order to prevent injury to the in- 
 strument. 
 
 The Riehle-Kenerson Extensometer. This extensometer, shown 
 in Fig. 112,, was designed by Prof. Kenerson, of Brown University, 
 and "is arranged to draw the strain diagram of a tension specimen 
 to just beyond the elastic limit with about 150 magnification, per- 
 mitting the modulus to be obtained direct from the diagram. It 
 can also be used as a micrometer extensometer, reading to .0001 
 inch. 
 
 The instrument is attached to the specimen by means of set 
 screws, three in the upper and two in the lower clamp. The upper 
 clamp supports the mechanism of the instrument, including the 
 cylinder on which the diagram sheet is placed, the weight to which 
 the pencil is attached, and the micrometer drums. The two set 
 screws in lower clamp are placed exactly opposite on the specimen 
 so as to suspend this clamp between them. On one side of this 
 clamp is a column with pivot and screw for adjustment. The other 
 side supports the spindle carrying micrometer drums. This spindle 
 rests on a pivot, but its upper end is threaded and passes through 
 a nut in upper clamp. It carries a grooved drum, around which is 
 wound a cord for supporting the weight to which pencil is attached. 
 As the specimen is stretched this weight descends and causes the 
 spindle to revolve in its nut, thus indicating the amount of the 
 extension directly on the graduated wheel and vernier. 
 
 A cord is attached to the poise on the beam of the testing machine 
 and through a reducing motion turns the cylinder. As the speci- 
 men stretches the pencil moves downward and the combined move- 
 ment produces the diagram. The diagram card is 6 inches long 
 by 5 inches high, and with a magnification of 150, this permits a 
 specimen to stretch .03 inch. When used as a micrometer extenso- 
 meter, a stretch of 1 inch can be noted. 
 
 This instrument produces a card similar to that made by the 
 autographic apparatus on the testing machine, but on a greatly en- 
 larged scale. It is only used for work up to the elastic limit. 
 
 Practical Use of the Testing Machine. In all specifications for 
 machinery, the material is required to meet certain requirements 
 
TESTING MATERIALS OF CONSTRUCTION 199 
 
 as to strength and elasticity. One or more test pieces must -be cut 
 from every heat, to be tested by the inspector before such material 
 can be passed. It is usual to fix the minimum requirements as to 
 (1) coefficient of ultimate strength, (2) coefficient of strength at 
 the elastic limit, (3) percentage of elongation, (4) percentage of 
 reduction of area. 
 
 The usual method of conducting a test is, after balancing th 
 beam at zero load, to gradually apply the load, keeping the beam 
 balanced by running the poise out by hand. At the yield point, the 
 beam suddenly drops. The load at this point is recorded and from 
 it the coefficient of strength at the elastic limit is calculated. The 
 test is then carried on to the breaking point. The maximum load is 
 thus obtained and from it the coefficient of ultimate strength is cal- 
 culated. The specimen is then removed from the machine and the 
 two parts having been put together, the total elongation is measured 
 directly and the diameter at the break taken. From these the per- 
 centages of elongation and reduction of area are calculated. 
 
 If, as in some cages, the modulus of elasticity is also required, an 
 extensometer must be used to determine the amount of elongation 
 at the elastic limit. 
 
 Test Record. 
 
 Date: 
 
 Specimen of : Tested in : 
 
 Original dimensions : L : d= : area= 
 
 Final dimensions : L : d : area= 
 
 (1) Coefficient of strength at elastic limits 
 
 (2) Coefficient of ultimate strength = 
 
 (3) Percentage of elongation in 8 inches 
 
 (4) Percentage of reduction of area 
 
 (5) Modulus of elasticity 
 
 (l\ Load at elastic limit 
 
 (2) = 
 
 Original area of specimen ' 
 
 Maximum load 
 Original area of specimen 9 
 
 i<*\ _ Final length -original length v1ftn . 
 Original length J 
 
200 EXPERIMENTAL ENGINEERING 
 
 / .v _ Original area final area - 
 Original area 
 
 elongation at elastic limit ' 
 
 Compression Tests. Materials are tested in compression to de- 
 termine their crushing strength and strength to resist bending; 
 also at times their elastic limits, and, if ductile, their plastic limits. 
 Short specimens, those whose length is less than five diameters, 
 usually fail by crushing or flowing. Long specimens usually fail 
 by bending toward the side of least resistance. 
 
 Materials may be divided into two general classes, in accordance 
 with their behavior when subjected, in short specimens, to compres- 
 sional stress. In the first classification are the plastic materials, 
 such as wrought iron, soft steel, copper, the alloys, etc., which fail by 
 flowing. After the elastic limit is passed, further compression re- 
 sults in an increase in the cross sectional area under a continually 
 increasing load. For such materials there are two fixed points inde- 
 pendent of shape of specimen, viz., the elastic limit and the plastic 
 limit. It has been found that the elastic limit of these materials is 
 nearly the same as their elastic limit in tension, and for this reason, 
 and in view of the difficulty of measuring the deformation of short 
 specimens, compressional tests upon them are seldom made. 
 
 The second classification embraces the brittle materials, such as 
 stone, brick, wood, cement, cast iron, etc., which fail by crushing, 
 due to the shearing on definite angles. With these materials the 
 ultimate strength is easily determined. 
 
 In general it may be said that compressional tests are very rarely 
 required for materials used in naval engineering. For soft steel and 
 steel of high grade such as is used in building machinery the charac- 
 ter of the material is sufficiently well determined by tensile tests. 
 
 Cross Bending Tests are used to determine, in case of brittle 
 materials, like cast iron, the modulus of rupture and the resilience, 
 and in ductile materials like wrought iron and soft steel, the elastic 
 limit and modulus of elasticity. Other tests on springs, rails, rail 
 joints, etc., determine the stiffness, i. e., the deflection at given 
 loads, the elastic limit, and in some cases, the ultimate strength. 
 
TESTING MATERIALS OF CONSTRUCTION 
 
 201 
 
 For naval purposes, cross bending tests are frequently required 
 on cast iron. They are usually limited to the determination of the 
 coefficient of ultimate strength. 
 
 Shearing Tests are not prescribed in the Specifications for the 
 Inspection of Engineering Material in the Navy. If the tensile 
 properties of a metal are known, or can be ascertained, it is suffi- 
 cient for all practical purposes to assume the shearing strength to 
 bear a definite ratio to the tensile strength. This applies to material 
 under a direct shear, as in a riveted joint. Shearing also takes place 
 in a bar subjected to torsional stress, acting in a rotary direction 
 around the axis. Such stresses are best investigated in a torsional 
 testing machine. 
 
 DEPARTMENT OF MARINE ENGINEERING AND NAVAL CONSTRUC- 
 TION, U. S. NAVAL ACADEMY. 
 
 TESTS OP "ELEPHANT BRAND" PHOSPHOR BRONZE, MADE WITH THE RIEHLE 
 
 SCREW POWER TESTING MACHINE. BRANDS A, B, S, AND F 2 , 
 
 CAST IN GREEN SAND. BRANDS X AND Y, ROLLED. 
 
 TENSION. 
 
 Finished rods. 
 
 Original length inches 
 
 2. 
 
 2. 
 
 2.358 
 .977 
 .875 
 .75 
 .601 
 15,630 
 24,360 
 17.9 
 19.87 
 20,840 
 32,480 
 
 2. 
 
 .965 
 .965 
 .732 
 .732 
 
 22,160 
 
 o.o 
 o.o 
 
 30,273 
 
 2- 
 2.14 
 .930 
 .906 
 .680 
 .645 
 18,160 
 21,230 
 7.0 
 5.15 
 26,706 
 31,221 
 
 2. 
 
 2.718 
 .798 
 .437 
 .5 
 .15 
 21,920 
 35,700 
 35.9 
 70.0 
 43,840 
 71,400 
 
 10. 
 11.188 
 .798 
 .422 
 .5 
 .14 
 21,522 
 29,840 
 11.88* 
 72.0 
 43,044 
 59,680 
 
 Final length inches 
 
 .... 2.45 
 
 
 977 
 
 
 844 
 
 
 75 
 
 
 .56 
 
 Load at elastic limit Ibs 
 
 ... 14 680 
 
 
 . . . 25 230 
 
 
 . .. 22 5 
 
 
 . . 95 3 
 
 Coef of elastic strength 
 
 . . . . 19 573 
 
 
 ... 33 640 
 
 
 
 * In 10 inches. 
 Brand 
 
 COMPRE 
 
 ISSION. 
 
 B F 2 
 Round Square 
 1. 1. 
 .622 .945 
 .786 1. 
 1.11 1.05 
 37.8 5.5 
 41.22 5.0 
 
 S 
 Round 
 1. 
 .655 
 .786 
 1.05 
 34.5 
 33.59 
 
 S 
 Square 
 1. 
 .745 
 1. 
 1.21 
 25.5 
 21.0 
 
 
 Square 
 
 
 
 
 687 
 
 
 1 
 
 
 .. . 1 41 
 
 
 31 3 
 
 Per cent increased area. .. 
 
 41.0 
 
 Specimens were still within the plastic limit and showed no signs of rupture 
 under a load of 100,000 pounds, the limiting load of the machine. 
 
202 EXPERIMENTAL ENGINEERING 
 
 Standard Test Pieces. Bureau of Steam Engineering Standards. 
 
 The specifications for steel and iron prescribe that " Test pieces 
 from blooms, rolled bars, forgings, and castings are to have a length 
 of 2 inches between measuring points and an area of cross section 
 of 0.2 square inch, diameter 0.505 inch." Test pieces for composi- 
 tion castings have a length of 2 inches between measuring points 
 and an area of cross section of 1 square inch. 
 
 Full-size bars and rods, within the capacity of the testing ma- 
 chine, may be used as tensile test pieces. 
 
 Test Pieces from Plates. When heats are rolled into plates of 
 varying thickness, the test pieces shall be taken from plates not less 
 than 0.3-inch thick. The standard width of tensile test pieces from 
 plates and boiler tubes will be 1J inches, the thickness the same as 
 the plate or tube, and the length between measuring points 8 inches. 
 
 All tensile test pieces shall be uniform in cross section between 
 measuring points. A variation of 5% above or below in area is 
 allowed. 
 
 The Bureau of Ordnance uses two standard test pieces. Fig. 113 
 shows the standard used for blooms, forgings, bars, and castings. 
 Fig. 114 shows the standard for plate material. It will be noted 
 that these standards also meet the specifications of the Bureau of 
 Steam Engineering for most material. 
 
 Richie-Miller Torsional Testing Machine. This machine, shown 
 in Fig. 115, consists of a head for applying the stress to the speci- 
 men, and a carriage for carrying the weighing mechanism and 
 the head which receives and transmits the load to the weighing 
 mechanism. The twisting head is operated through worm gearing 
 and spur gearing to reduce the motion of the head. The worm shaft 
 is connected to the motor by a silent chain drive. The motor is 
 reversed by a controller which, by means of resistance box, gives 
 six variations of speed in either direction, varying from 50% to 
 the normal speed of the motor. There is a lever, seen just above 
 spur wheel on head, for operating the friction clutch to start or stop 
 the machine. The weighing end can be adjusted to suit different 
 lengths of specimens up to 5 feet, by means of a rack and pinion 
 operated by hand wheel. 
 
 The head on weighing end has a set of toggle grips for holding 
 
TESTING MATERIALS OF CONSTRUCTION 
 
 203 
 
 the specimen, and is supported by a parallel system of levers that 
 completely neutralize any load except the torsion or twisting load 
 exerted on the specimen. A similar grip arrangement holds the 
 other end of specimen in the head that applies the load. These 
 grips are designed to take round or hexagon specimens from J inch 
 to 1J inches in diameter. 
 
 The beam is similar to that for the testing machine, previously 
 described, but the poise is moved by hand wheel only. 
 
 10 THREADS PER INCH 
 
 FIG. 113. 
 
 
 -s --q-1%0- 
 
 zy 
 
 \ 
 
 
 f.9 
 
 o 
 
 
 
 i . t *" 
 
 
 
 J 
 
 
 
 1 
 
 
 
 Y 
 
 -* 
 
 J*PJ7 
 
 Ift" + 
 
 
 OF SAME THICKNESS ASTHE PLATE 
 
 FIG. 114. 
 
 The angle of torsion of the specimen, or the number of turns 
 required to break the specimen, is noted on the graduated ring at- 
 tached to the twisting head. 
 
 The Torsion Indicator. In reading the angle of torsion on the 
 head of the machine, it is difficult to eliminate errors due to the 
 slipping of the specimen in the grips. To avoid this difficulty, it is 
 necessary to use an indicator attached directly to the specimen. The 
 one in use in the laboratory consists of a disc, graduated on its outer 
 edge in degrees, clamped on one end of the specimen, and a pointer 
 clamped on the other end, the finger of the pointer extending paral- 
 
204 
 
 EXPERIMENTAL ENGINEERING 
 
TESTING MATERIALS OF CONSTRUCTION 205 
 
 lei to the specimen, up to the graduations on the face of the disc. 
 The movement of the pointer over the graduated disc indicates the 
 angle of torsion. 
 
 With this form of indicator, a different pointer is required for 
 each length of specimen. The Eiehle Torsion Indicator avoids this 
 difficulty and is suitable for specimens of any length. A light gear 
 wheel is attached to each end of the specimen by 3 set screws. Each 
 of these drives a pinion moving a pointer over a dial. The dial is 
 graduated and when the gear wheel is turned, the angle is indicated 
 to within ^ degree on the dial. With a specimen under test the 
 reading of both dials is taken and the reading on the dial near the 
 weighing end is subtracted from the reading of the dial near the 
 power end. 
 
 Cements. 
 
 Cements are used only to a limited extent on board a ship, but 
 their use is so extensive in other branches of engineering allied 
 to marine engineering, that some knowledge of them and the method 
 of testing them is of importance to the naval engineer. 
 
 Hydraulic Cement may be broadly defined as a material which, 
 when pulverized and mixed with water into a paste, acquires the 
 property of setting and hardening under water. In engineering 
 work, four classes of cement are generally recognized (1) Portland 
 cement, (2) natural cement, (3) Pozzuolana cement, (4) mixed or 
 blended cement. 
 
 Portland Cement is an artificial cement, manufactured from 
 selected materials, commonly limestone and clay, with other mate- 
 rials varying with each brand. The mixture is calcined until the 
 materials begin to run together and the resulting clinker is ground 
 to a fine powder. It is named from its resemblance in color to the 
 famous Portland limestone in England. 
 
 Natural Cement is the product resulting from the burning and 
 subsequent pulverization of an impure grade of limestone, con- 
 taining lime and clay, the heat of burning being insufficient to cause 
 vitrefaction. It is sometimes called Roman cement, from a fancied 
 resemblance to Eoman mortar. In the United States it is also 
 widely known as Eosendale cement, from the town in the State of 
 New York, where much of it is produced. 
 
206 EXPERIMENTAL ENGINEERING 
 
 Pozzuolana or Puzzolan Cement is obtained by grinding together 
 an intimate mixture of slaked lime and blast furnace slag or vol- 
 canic lava. It is not burned, the hydraulic ingredients being present 
 only as a mechanical mixture. It is so named from a town in Italy 
 where much of it is made and where there is a large deposit of the 
 particular form of lava used. 
 
 Mixed Cements cover a large variety of products, made by com- 
 bining the other forms of cement, or mixing them with an inert 
 material. " Improved Cements " are naturals, containing from 10 
 to 30 per cent of Portland clinker. Sand cements are made by finely 
 grinding a mixture of Portland cement and sand in varying propor- 
 tions dependent upon the use to which they are to be applied. 
 
 Neat Cement is cement mixed with water only. 
 
 Concrete is a mixture of cement with sand, stone, or pebbles, and 
 water. 
 
 Reenforced Concrete contains steel bars or other shapes to give 
 added strength. 
 
 Portland cement is the variety commonly employed and when 
 the term cement is used without qualification, this is what is gen- 
 erally understood. The commercial article is usually a mixture 
 containing more or less sand. For this reason the specifications for 
 cement are closely drawn in order to obtain material of the quality 
 desired. In ordinary use, one part is mixed with from -J to 5 parts 
 of sand according to the purpose for which it is to be employed. 
 The cleaner and sharper the sand, the stronger will be the resulting 
 mixture. 
 
 Natural cement is distinguished in use by its lighter weight, 
 quicker set, and lower strength in the earlier stages of hardening. 
 It is not adapted for use on board ship on account of its lower 
 strength. 
 
 Salt water should never be used in mixing cement, on account of 
 the action of the salt in producing disintegration. 
 
 TJses of Cement on Board Ship. On board a naval vessel, the 
 bilges and double bottoms are covered with cement to protect the 
 plating and make them easier to keep clean. Inaccessible pockets 
 in bilges are frequently filled with cement. The mixture used is 1 
 part best Portland cement to 2 parts sand. 
 
TESTING MATERIALS OF CONSTRUCTION 207 
 
 The walls of trimming tanks and tanks for the storage of fresh 
 water are covered with cement to prevent corrosion. For this pur- 
 pose a thin wash of neat cement is applied with a paint brush. 
 
 In old boilers, with leaky shell seams that give trouble,, cement 
 is sometimes used to advantage. When thus used the surfaces 
 must be quite clean and the cement in the form of a thin wash, 
 mixed neat, is applied on the inside of the boiler. Considerable care 
 must be exercised in order to stop a leak by this method, as water 
 will in many cases work its way along the seam from a point some 
 distance removed from the leak. When thus used on a steel or iron 
 surface exposed to heat, but little trouble is experienced with the 
 cement cracking, since its coefficient of expansion is approximately 
 the same as that of the metal. 
 
 Cement Testing. The usual tests applied to cement are those for 
 specific gravity, tensile strength, fineness, constancy of volume, and 
 time of initial and final set. It is not usual to make compression 
 tests in America. The results in tension vary directly with those in 
 compression, so that the tensile strength is a satisfactory index of 
 the value of the cement in compression. 
 
 The Standard Cement Specifications of the American Society for 
 Testing Materials provide that the acceptance or rejection of cement 
 shall be based on the following requirements : 
 
 Specific Gravity. The specific gravity of the cement thoroughly dried 
 at 100 C., shall be for Portland cement not less than 3.10, and for 
 natural cement not less than 2.8. 
 
 Fineness. Portland cement shall leave by weight a residue of not 
 more than 8# on the No. 100 and 25$ on the No. 200 sieve. Natural 
 cement shall leave not more than 10# on the No. 100 and 30# on the 
 No. 200 sieve. 
 
 Time of Setting. Portland cement shall develop initial set in not 
 less than 30 minutes and final or hard set in not less than one hour 
 nor more than 10 hours. Natural cement shall develop initial set in 
 not less than 10 minutes and hard set in not less than 30 minutes nor 
 more than 3 hours. Setting of the samples under test should take place 
 in moist air under as uniform conditions as possible. A sudden change 
 or range of temperature in the room in which the tests are made, a 
 very dry or humid atmosphere, and other irregularities, vitally affect 
 the rate of setting. 
 
 Tensile Strength. The minimum requirements for tensile strength 
 for briquettes one inch square in cross section shall be within the 
 
208 EXPERIMENTAL ENGINEERING 
 
 following limits, and shall show no retrogression in strength within 
 the periods specified: * 
 
 Portland Cement, Neat. 
 
 Strength. 
 Lbs. 
 
 24 hours in moist air 150-200 * 
 
 7 days, 1 day in moist air, 6 days in water. . . 450-550 
 28 days, 1 day in moist air, 27 days in water. . 550-650 
 
 One Part Portland, Cement, Three Parts Standard Sand. 
 1 days, 1 day in moist air, 6 days in water. . 150-200 
 28 days, 1 day in moist air, 27 days in water . . 200-300 
 
 Natural Cement, Neat. 
 
 Strength. 
 Lbs. 
 
 24 hours in moist air 50-100 
 
 7 days, 1 day in moist air, 6 days in water.. 100-200 
 28 days, 1 day in moist air, 27 days in water. . 200-300 
 
 One Part Natural Cement, Two Parts Standard Sand. 
 
 1 days, 1 day in moist air, 6 days in water. . 25- 75 
 
 28 days, 1 day in moist air, 27 days in water. . 75-150 
 
 Constancy of Volume. Pats of neat cement about 3 inches in diameter, 
 y 2 inch thick at center, tapering to a thin edge, shall be kept in moist 
 air for a period of 24 hours. 
 
 (a) A pat is then kept in air at normal temperature and" observed 
 at intervals for at least 28 days. 
 
 (b) Another pat is kept in water maintained as near 70 Fah., as 
 practicable, and observed at intervals for at least 28 days. 
 
 (c) In the test of Portland cement only, a third pat is exposed in 
 any convenient way in an atmosphere of steam above boiling water, 
 in a loosely closed vessel for five hours. 
 
 These pats, to satisfactorily pass the requirements, shall remain 
 firm and hard and show no signs of distortion, checking, cracking, or 
 disintegrating. 
 
 Gilmore Needles. The times for initial and hard set are made 
 respectively with a needle 1/12 inch in diameter, weighted to % lb., 
 and 1/24 inch in diameter, weighted to 1 Ib. The moment when the 
 
 * For example, the minimum requirement should be some specified 
 value within the limits of 150 and 200 Ibs., and so on for each period 
 stated. 
 
TESTING MATERIALS OF CONSTRUCTION 
 
 209 
 
 coarse needle fails to sink into the cement, when held lightly in a 
 vertical position between the fingers, the point resting upon the speci- 
 men, is called the time of initial setting, and similarly, the time when 
 the fine needle will no longer penetrate, is the moment of final or hard 
 setting. 
 
 Tensile Tests. Fig. 116 shows the standard form of tensile 
 briquette in use in the United States. It is 1 inch thick, giving an 
 area of cross section of 1 square inch at the smallest part where the 
 break occurs. 
 
 
 / \ 
 
 / \ 
 
 
 / \ J 
 
 L / \ / 
 
 I / 
 
 FIG. 116. 
 
 The samples to be tested are thoroughly mixed and the specified 
 quantity of water added. The mixture is then worked up into a 
 plastic mass and moulded in brass briquette moulds. It is usual to 
 do this work by hand where many specimens are to be tested, the 
 experience of the operator giving uniformity of results. A briquette 
 making machine is shown in Fig. 117,, in which a hammer of fixed 
 weight is made to fall on a disc placed over the mould containing 
 the briquette. The number of strokes is fixed and the blows are 
 stopped automatically after a definite number. The machine is 
 useful to secure briquettes of uniform density particularly where 
 but few tests are made and the work is divided among several 
 operators. 
 14 
 
210 
 
 EXPERIMENTAL ENGINEERING 
 
 Fairbanks' Improved Cement Testing Machine. The construc- 
 tion of this machine is shown in Fig. 118. Its operation is de- 
 scribed as follows : 
 
 Hang the cup F on the end of the beam D. See that the poise R 
 is at the zero mark and balance the beam by turning the ball L. Fill 
 the hopper B with fine shot. Place the briquette in the clamps 
 N-N, using great care to avoid eccentricity. Tighten the hand wheel 
 P until the indicators are in line. By means of the hook lever Y 
 the worm is now engaged with the gear. The shot valve is then 
 
 FIG. 117. 
 
 opened, allowing the shot to run into the bucket, and the crank is 
 turned with sufficient speed to hold the beam in equilibrium until 
 the briquette is broken. At the point where the spout joins the res- 
 ervoir will be noticed a small valve, by which the flow of shot may 
 be regulated. 
 
 When the briquette breaks, the beam D will drop and automati- 
 cally close the valve J. Then remove the cup, with its contents, 
 hanging the counterpoise G in its place. Hang the cup F on the 
 hook under the large ball E, and weigh the shot, using the poise R 
 on the graduated beam D and the weights H on the counterpoise G 
 The result will give the number of pounds required to break the 
 specimen. 
 
TESTING MATERIALS OF CONSTRUCTION 211 
 
 FIG. 118. 
 
CHAPTER IX. 
 ENGINE LUBRICATION. 
 
 Open Systems. Until recently all engine bearings were lubri- 
 cated by gravity through tubes, the oil being fed into the tubes 
 through wicks. Many engines built in the last few years have been 
 fitted with the sight feed system, in which the flow of oil to each 
 tube is regulated by a valve, a portion of the tube being made of 
 glass to permit visual regulation of the flow. 
 
 Much attention has been given recently in our navy to economy 
 in the consumption of all expendable stores, including oil. One of 
 the results has been to direct attention to the superiority of the 
 wick feed system. By varying the number of strands in the wick 
 and the number of wicks, the quantity of oil can be regulated to a 
 nicety. Sight feed lubrication cannot be closely regulated except 
 by the most careful attention which it is difficult to give where 
 there are a large number of bearings and impossible in many out 
 of the way locations. 
 
 Forced Lubrication. In late machinery designs forced lubrica- 
 tion has been fitted. The oil is sent to the bearings under pressure 
 from a pump. It then drains to a receptacle in the crank pit or 
 elsewhere low down in the ship, where it is taken up by the pump 
 suction. At a convenient location in the system it passes through 
 a cooler built on the plan of a surface condenser, where the heat 
 from the bearings is given up to the cooling water. 
 
 This system permits the same oil to be used over and over again, 
 the only new oil required being such as is necessary to make up 
 losses from leakage and from wear. It is necessary from time to 
 time to pump the oil into tanks and allow the water and sediment 
 to settle, thus removing it from the system, but this can be done at 
 such times as will not interfere with the operation of the engines. 
 This system is very economical in the consumption of oil, but a 
 greater advantage is the great reduction in friction and wear in 
 the bearings. 
 
ENGINE LUBRICATION 213 
 
 Engine Lubricants. 
 
 Practically all oils used for lubricating purposes at the present 
 time in the United States have as their base a heavy mineral oil, 
 which is the last of the series of distillates from crude petroleum. 
 Various other constituents are added by oil manufacturers to 
 obtain a compounded oil that will suit the particular requirements 
 for which it is intended. The oils that are used on board ship 
 are as follows : 
 
 Oils for Wick or Gravity Feed. These are compounded oils of 
 high lubricating value sufficiently fluid to feed well through wicks. 
 Such oils are usually purchased under their trade names, the name 
 carrying the manufacturers' guarantee of the good quality and uni- 
 formity of the oil. Oils of this character, however excellent the re- 
 sults they may give individually are not likely to do well when two 
 or more brands are mixed. When changing from one brand to 
 another all oil cups and other receptacles must be thoroughly 
 cleaned before starting the new oil. 
 
 These oils are used on engine pins and bearings where the feed 
 is by gravity. Also in receptacles such as thrust bearings where the 
 shaft runs in a bath of oil. In such bearings the oil usually shows 
 a tendency to form a lather, this being characteristic of mixed oils 
 where any water is present. This lather is itself an excellent lubri- 
 cant and as it cannot form in a hot bearing its presence indicates 
 that the bearing is working well. 
 
 Oils for Forced Lubrication must be pure mineral oils on account 
 of the tendency of mixed oils to lather and choke the small pipes 
 and passages. Such oils are also sold under trade names that are 
 accepted as guarantees of their uniform character. Although all 
 such oils are supposed to be pure mineral oils, it will be found the 
 best policy not to begin to use a different brand without first clean- 
 ing out the system. 
 
 Cylinder Oils must not vaporize or carbonize at temperatures 
 that obtain in the engine cylinders. They must therefore have a 
 high flash point. When used in the cylinders a considerable per- 
 centage passes over into the condensers and thence into the boilers. 
 In order to produce the minimum amount of injury to the boilers 
 thev must contain no acid. These conditions are best met in a 
 
214 EXPERIMENTAL ENGINEERING 
 
 heavy mineral oil with no animal or vegetable constituents and it 
 is usual to specify that cylinder oils must be pure mineral oils. 
 
 Cylinder oil was considered necessary with horizontal engines. 
 On account of the injury to boilers and particularly to water-tube 
 boilers, resulting from carrying it over into condensers, it became 
 necessary to reduce the quantity as much as possible. It has been 
 found practicable to entirely dispense with it in vertical engines, 
 and the only cylinder oil now employed is used to swab on piston 
 rods in case they should run warm in stuffing boxes. 
 
 Ice Machine Oils must not vaporize at the temperatures that 
 exist in the compressor cylinders. Otherwise the oil vapors would 
 pass over and condense in the pipes of the system, thus preventing 
 its efficient operation. On the other hand in a machine of the cold- 
 air type the same oil used as a lubricant for the expander cylinder 
 must not solidify at the low temperatures existing in that cylinder. 
 A cylinder oil might be satisfactory in the compressor cylinder of 
 a compression machine, but if used in the expander cylinder of an 
 air machine would be wholly unsatisfactory. It is best when prac- 
 ticable to obtain ice machine oils of a brand that is known to be 
 satisfactory. 
 
 Vaseline, etc. Heavy refined petroleum is used for coating pol- 
 ished surfaces that are liable to rust when not in use. Engine 
 cylinders when they are to stand idle for several days must be 
 wiped out and coated in this manner. 
 
 Grease cups are little used for engine lubrication at the present 
 time. They are still fitted on hoisting and other machinery subject 
 to irregular use, where when the bearing begins to warm the grease 
 melts and runs down. Vaseline may be used in grease cups. Albany 
 grease is a proprietary preparation usually supplied for this purpose. 
 Tallow, which was formerly employed, is but little used at present 
 on account of the acid which it is very liable to contain attacking 
 the polished surface of the pins. 
 
 Testing of Lubricants. 
 
 Determinations Required. The following particulars are re- 
 quired in a complete test of a lubricant, as made in a laboratory : 
 
ENGINE LUBRICATION 215 
 
 (1) Its composition, including the detection of any adulterant 
 that may be present. 
 
 (2) Its specific gravity. 
 
 (3) Its viscosity. 
 
 (4) Any tendency to gum. 
 
 (5) The temperatures of flashing, ignition, and solidification. 
 
 (6) The detection of any acid that it may contain. 
 
 (7) The coefficient of friction. 
 
 (8) Its durability and heat removing power. 
 
 (9) The presence of any grit or other foreign matter. 
 
 The composition and adulteration of a lubricant can only be 
 properly determined by a chemical analysis. 
 
 The specific gravity is determined by the ordinary methods in 
 use in the physical laboratory. Special care should be observed to 
 thoroughly clean all instruments with benzine after using for the 
 testing of oil. 
 
 Viscosity of oil is closely related but not proportional to its 
 density. It is also closely related, and in many cases inversely pro- 
 portional, to its lubricating properties. The viscosity varies accord- 
 ing to the temperature, but not in the same proportion for different 
 oils, hence tests for viscosity should be made with the temperatures 
 the same as those at which the oil is to be used. The less the vis- 
 cosity, consistent with the pressure to be used, the less the friction 
 
 The viscosity test is considered of great value in determining the 
 lubricating qualities of oils. By it alone we could probably deter- 
 mine the lubricating qualities to such an extent that a good oil need 
 not be rejected nor a bad oil accepted. There are, however, no set 
 standards for the determination of viscosity and the results are to 
 be considered as comparative rather than absolute. 
 
 There are several methods of determining the viscosity. It is 
 usual to take the viscosity as inversely proportional to the flow 
 through a standard nozzle, while maintained at a constant, or con- 
 stantly diminishing head, and constant temperature. A comparison 
 is made with water or with some well-known oil, as sperm, lard, or 
 rape-seed, taken as a standard, under the same conditions of pres- 
 sure and temperature. 
 
216 
 
 EXPERIMENTAL ENGINEERING 
 
 Engler's Viscosimeter. Engler's Viscosimeter, shown in Fig. 
 119, consists of a chamber holding the oil to be tested, a water bath, 
 a flask graduated so as to receive 200 cc. of the oil, thermometers 
 and the opening through which the heated oil flows out upon the 
 withdrawal of the plug. 
 
 FIG. 119. 
 
 In using this instrument the viscosity of an oil is stated in 
 seconds required for 200 cc. of the oil to run into the flask. Heat 
 can be applied to the water bath, the viscosity being determined at 
 any temperature required up to 100 C. Any temperature up to 
 360 C. can be secured by filling the water bath with paraffine 
 instead of water. 
 
 Engler recommends that all viscosities be compared with water, 
 thus: If water requires fifty-two (52) seconds for delivery of 
 200 cc. into the receiving flask, and the same amount of an oil 
 
ENGINE LUBRICATION 217 
 
 under examination requires 130 seconds, the ratio is determined by 
 
 130 
 
 =2.50, the oil thus having a viscosity of 2.5 times that of 
 o2i 
 
 water. 
 
 The bath is filled with a suitable liquid to a height roughly cor- 
 responding with the point of the gage in the oil cylinder. Water 
 answers well for temperatures up to 90 C. or 200 F., and for 
 higher temperatures a heavy mineral oil may be used. The liquid 
 having been brought to the required temperature, the oil to be 
 tested, previously brought to the same temperature, is poured into 
 the cylinder, until the level of the liquid just reaches the point of 
 the gage. A narrow-necked flask holding 200 cc. to a point marked 
 on the neck, is placed beneath the jet in a vessel containing a liquid 
 of the same temperature as the oil. The rod valve is then raised, a 
 stop-watch at the same time started, and the number of seconds 
 occupied in the outflow of 200 cc. noted. 
 
 It is of the greatest importance that the oil cylinder should be 
 filled exactly to the point of the gage, after inserting the ther- 
 mometer, and that the standard temperature should be precisely 
 maintained during the experiment, a difference of -J degree F. 
 making an appreciable alteration in the viscosity of some oils. 
 It is also essential that the oil should be quite free from dirt or 
 other suspended matter, and from globules of water, as the jet 
 may be otherwise partially obstructed. If the oil cylinder requires 
 to be wiped out, paper rather than cloth should be employed, as 
 filaments of the latter may be left adhering. When oils are being 
 tested at temperatures much above that of the laboratory, a gas 
 flame is applied to the copper heating tube, and the agitator kept 
 in gentle motion throughout the experiment. 
 
 The Boverton-Redwood Viscosimeter. This is shown in Fig. 120. 
 A is a central vessel containing the oil which flows out through 
 the standard orifice at B. C is an enveloping vessel around A, 
 through which hot water is circulated, passing out at the cock D. 
 E is a framework carrying vanes for circulating the water and a 
 thermometer F for measuring its temperature. G is a thermometer 
 for measuring the temperature of the oil. H is a handle for moving 
 the framework E. J is a water leg to which a lamp is applied in 
 
218 
 
 EXPERIMENTAL ENGINEERING 
 
 heating up the oil. K-K are thumb screws in the feet of the sup- 
 porting tripod, by which the instrument is levelled. L is a spirit 
 level which is placed over A and the apparatus levelled by its guid- 
 ance before operations are commenced. 
 
 FIG. 120. 
 
 In operation the standard quantity of oil is placed in A and 
 brought to the standard temperature. The orifice, which has been 
 plugged, is then opened and the time required for the oil to flow 
 through is noted. 
 
ENGINE LUBRICATION 219 
 
 Gumming or Drying is a conversion of the oil into a resin by 
 oxidation, and occurs on exposure of the oil to the air. In linseed 
 and the drying oils it occurs very rapidly, and in the mineral oils 
 very slowly. 
 
 The usual method of testing for this property is by use of a 
 slightly inclined plane of metal or glass. A small, but fixed, quan- 
 tity of the oil is started at the upper edge of the plane and the time 
 required to reach the bottom is taken as a measure of its gumming 
 properties. Comparison is made with a standard oil. This test 
 has but little value in determining the quality of an oil. 
 
 A better test is made with the standard oil testing machine. 
 Fresh oil is applied, a run made, and the friction noted. After the 
 bearing has been exposed to the air for a time, a second run is made, 
 and the increase of friction is noted. In this case also, comparison 
 must be made with some standard oil. 
 
 The Flash Test. The effect of heat is to increase the fluidity 
 of oils and to lessen the viscosity. The temperature at which oils 
 flash, ignite, boil, or congeal is often of importance in the determi- 
 nation of the suitability of an oil for some special purpose. 
 
 The temperature at which inflammable vapors are given off is the 
 flashing point, and should in all cases be known for inflammable 
 oils that are to be stored on board ship. The test is made in two 
 ways. 
 
 The Open Cup. The oil to be tested is placed in an open cup 
 of watch glass form, which rests on a sand or water bath. Heat 
 is applied to the bath- and as the oil becomes heated a lighted match 
 is passed at intervals of a few seconds over the surface of the oil, at 
 a distance of about half an inch from it. At the instant of flashing, 
 the temperature of the bath is noted, which is the " flash point." 
 
 Improvised apparatus, such as this, is sometimes employed for 
 tests, where standard instruments cannot be obtained. The results 
 are subject to considerable variation, due to differences in the 
 method of applying the match. 
 
 The New York State Board of Health Flash Tester. This is 
 shown in Fig. 121 and is on the open cup principle. A light mica 
 cover rests on the cup, having two openings, one for the passage of a 
 thermometer, and the other for the application of the match. While 
 
220 
 
 EXPERIMENTAL ENGINEERING 
 
 not sufficiently heavy to confine the vapors, this cover serves to give 
 uniformity to the results obtained. The thermometer in the oil 
 measures its temperature directly. 
 
 The Cleveland Flash Point Tester. Fig. 122 shows the Cleve- 
 land Flash Point Tester. The oil is contained in a central vessel, 
 
 FIG. 121. 
 
 which is surrounded by a bath of steam. The oil is completely 
 covered and the vapors that are given off pass down through a 
 central tube, to the mouth of which the match is applied. A ther- 
 mometer in the oil gives the temperature of flashing. A second 
 thermometer in the surrounding jacket gives the temperature of 
 the bath. If the temperature of flashing is above that obtainable 
 with steam, the jacket may be filled with sand and heated by the 
 
ENGINE LUBRICATION 
 
 221 
 
 application of a Bunsen burner. For ordinary purposes, steam is 
 better, since the heat thus obtained is more uniform. This appara- 
 tus is largely used for making determinations for heavy oils. 
 
 Other apparatus for making the flash test is described in 
 Chapter X. 
 
 E- 
 
 FIG. 122. 
 
 The Burning Point is determined by heating the oil to such a 
 temperature, that when the match is applied as for the flash test, 
 the whole of the oil will take fire. The reading of the thermometer 
 just before the match is applied is the burning point. The appara- 
 tus used is the open cup flash point testing apparatus with ther- 
 mometer directly in the oil to be tested. 
 
222 EXPERIMENTAL ENGINEERING 
 
 Evaporation. Mineral oil will lose weight by evaporation, which 
 may be determined by placing a given weight in a watch glass 
 and exposing to the heat of a water bath for a given time, as twelve 
 hours. The loss denotes the existence of volatile vapors, and should 
 not exceed 5 per cent in good oil. Other oils often gain weight 
 under these conditions by the absorption of oxygen. 
 
 Cold Tests are made to determine the behavior of oils and greases 
 at low temperatures. The sample to be tested is placed in a test 
 tube, in which is inserted a thermometer. The tube is then packed 
 in a freezing mixture, composed of small particles of ice mixed 
 with salt, with provision for draining off the water. After the 
 sample has congealed, the tube is removed from the freezing mix- 
 ture, and the oil is stirred gently with the thermometer. The 
 temperature indicated when the oil is melting is the chill point. 
 
 Acid Tests. The ordinary test for the presence of acid is to 
 observe the effect on blue litmus paper. This is a qualitative test 
 only, and is not very satisfactory. For this test a sample of the oil 
 should be sent to the chemical laboratory. 
 
 Oil Testing Machines. The coefficient of friction, the dura- 
 bility, and the heat removing power of the oil, are determined by 
 the use of oil testing machines. These are of various designs, the 
 form used in several U. S. Navy laboratories being shown in Fig. 
 123. 
 
 The main journal of the machine rests on the four large rollers, 
 shown in the figure. This reduces friction and prevents the heating 
 of the shaft, which would affect the results of temperature tests. 
 Ball thrust collar bearings prevent motion along the direction of 
 the axis of the journal, and take any thrust in this direction that 
 would cause friction. 
 
 The bearing for use in the tests, rests on the top of the journal, 
 and fits in a cap, to which the yoke frame is attached. This is con- 
 nected by a system of links and levers, through hardened steel sur- 
 faces and knife edge supports, to the weighing beam, seen on the 
 right. A turn buckle, concealed under the frame of the machine, 
 enables the load on bearing to be varied by the operator as may be 
 desired. The amount of such load is weighed \)j the beam. 
 
 The diameter of journal is 6 inches. The bearing is 4 inches 
 
EXPERIMENTAL ENGINEERING 
 
 long by 2% inches wide. This gives a projected area of 10 square 
 inches, but after determining the direction in which the machine is 
 to run, the leading edge of this brass is shaped down in a horizontal 
 plane so as to reduce this area by one square inch, removing the 
 sharp edge to permit ready access of oil and bringing the projected 
 area of bearing to 9 square inches. 
 
 In the center at top of cap is a hole for thermometer well. This 
 is carried into the cap brass to within J inch of the bearing sur- 
 face, and a J-inch gas pipe nipple is screwed in to seal the joint 
 between cap and brass. A standard thermometer is provided for 
 use in the well in making tests. The space in well around ther- 
 mometer is filled with some good lubricating oil in order to make 
 contact between the brass and the bulb of thermometer. 
 
 In making a test, after the machine is in operation, the pres- 
 sure is applied by setting out the poise on the pressure beam to 
 the desired point and then tightening the turn buckle until the 
 beam is balanced. Before the first test is made, the total weight of 
 the parts whose weight is sustained by the bearing must be deter- 
 mined, and this amount, which is approximately 500 pounds, must 
 always be added to the pressure given by the beam, in order to ob- 
 tain the total pressure on the bearing. When the poise is at 1000 
 pounds the total pressure on the bearing will be 1000 plus 500, or 
 1500 pounds, approximately. 
 
 Methods of Conducting Tests. No standard method of conduct- 
 ing a test of lubricating oils has been generally adopted. In gen- 
 eral the object of a test is to determine the coefficient of friction. 
 The usual method is to run the machine at a given speed and under 
 a given pressure on bearing. Under these conditions, if the load 
 is not excessive, and a sufficient quantity of lubricant is supplied, 
 the temperature will rise to a certain point and remain approxi- 
 mately constant. Four methods of observation are then practiced : 
 First, to note the temperature at the end of a given time; second, to 
 note the time necessary to reach a given temperature ; third, to note 
 the temperature to which the bearing rises and at which it remains 
 constant for a given time; fourth, to maintain the bearing at a 
 given temperature by cooling it. The third method is commonly 
 used because it is easier to carry out and more nearly approximates 
 the running conditions of a bearing. 
 
ENGINE LUBRICATION 225 
 
 To conduct a test according to the third method: Clean the 
 journal, bearing, and pad with gasoline in order to remove all traces 
 of lubricating oil. Soak the pad with the sample oil and also apply 
 it freely to the journal and bearing. Eun the machine a few min- 
 utes, until the bearing is well lubricated, then apply the desired 
 pressure by means of the turn buckle as has been described. The 
 machine must be run at a constant speed, or if there is a variation 
 of speed it must be slight, and readings of the revolutions per minute 
 must be taken in order to obtain the average. Keep a log of the 
 test, taking readings of the revolutions, temperature, and friction, 
 every five minutes, and continue until the temperature remains con- 
 stant for thirty minutes. Record the following data : Date, kind of 
 lubricant, total pressure on bearing, pressure in pounds per square 
 inch of projected area, friction load, revolutions per minute, velocity 
 in feet per minute, duration of test, temperature of room, tempera- 
 ture of bearing at start, temperature of bearing when constant. Cal- 
 culate the coefficient of friction and make it a part of this record. 
 To calculate the coefficient of friction, divide the friction load, as 
 obtained from the friction beam, by the total pressure in pounds on 
 the bearing. 
 
 Navy Department Method. The following method of testing 
 oil has been evolved by Mr. C. A. Webb, in charge of the testing 
 laboratory, Department of Steam Engineering, Navy Yard, New 
 York, where a machine of this description has been in use for sev- 
 eral years. The results and conditions of the test are embodied in 
 specifications for lubricating oil for marine machinery, issued by 
 the Bureau of Supplies and Accounts, Navy Department. The 
 object of the test is to determine the wearing quality of the oil, 
 although the coefficient of friction is also obtained. 
 
 The accessories with the machine, necessary for this test, are, a 
 pair of laboratory scales, preferably in glass door case, a small copper 
 cup of 4 ounces' capacity to hold the oil, and an orange wood stick 
 about J inch diameter by 6 inches long, shaved to a thin blade for 
 half its length. The oil pad and water-cooling device for the jour- 
 nal are not used in this test. 
 
 The directions for making a test, as given by Mr. Webb are as 
 follows : 
 
 15 
 
226 EXPERIMENTAL ENGINEERING 
 
 (1) Clean the bronze bearing and journal with gasoline, using a 
 cloth to wipe them dry. 
 
 (2) Weigh about half an ounce of the oil to be tested, together 
 with the oil stick, in cup, and call it Weight of Oil at Start. 
 
 (3) Apply a small amount of this oil to the bearing and journal; 
 place bearing in place on journal and apply the full load. 
 
 (4) Place the thermometer for taking temperature of bearing in 
 its place, and after reading thermometers, counter, and time, start 
 the machine. 
 
 (5) At the instant the machine is started a drop of oil from the 
 cup should be applied to the journal on the advancing side, where 
 the bearing touches the journal, and spreading the oil so as to make 
 a uniform lubrication. All oil used on test will be applied so, pass- 
 ing the stick backwards and forwards along the face of the bearing, 
 spreading the oil evenly. If the pull in pounds increases in spite 
 of the spreading, apply fresh oil from the cup until the pull falls 
 back or becomes less. The pull in pounds should fall to normal in 
 about 20 minutes from start to test. Keeping the poise on friction 
 beam just balanced while the oil is uniformly spread, the proper 
 amount of oil can be easily found ; for, as the temperature increases 
 the pull decreases, until the oil has reached its best working tem- 
 perature. As long as the pull does not increase with a uniform 
 lubrication by spreading with the stick, it does not require fresh oil, 
 but as soon as the pull overbalances, a little fresh oil will bring it 
 back to balance. Keep poise advancing and in balance while spread- 
 ing the oil. The first half hour will in most cases finish the applying 
 of fresh oil, and the last hour and a half will require constant 
 spreading. The temperature and pull will reach the normal read- 
 ing for the oil under test in the first hour, and remain so to the end, 
 as a rule. 
 
 (6) At the end of the two hours, read the counter and weigh the 
 oil cup with stick and remaining oil. The difference in weight in 
 grains Troy gives the weight of oil used for the test. The reading 
 of counter will give the revolutions for the two hours, from which 
 we get the average revolutions per minue, and work up the results. 
 
 Circumference of journal in feet x R. P. M. x load not less 
 
 Troy grains of oil than 325,000. 
 
 pounds 
 
 Coefficient of friction = 
 
 Total load 
 
ENGINE LUBRICATION 
 
 227 
 
 Before starting the machine, the weight beam is adjusted to 
 exactly counterbalance the weight of framing and the friction beam 
 arms are also brought to equipoise. A load of 2700 pounds is then 
 added to the brass by the weight beam scales, equalling 300 pounds 
 per square inch of projected area. This is the standard load for all 
 these tests. A starting load of about 100 pounds is also put on that 
 friction beam which is behind the direction of rotation of top of 
 journal. This is subject to further adjustment during the tests. 
 
 The number 325,000 in the above formula is the coefficient of 
 performance, which must be reached by an oil in order for it to 
 be accepted. It does not indicate the foot pounds of work done per 
 grain of oil, but is simply an arbitrary standard. The coefficient of 
 friction is not specified in the requirements for oil, but it indirectly 
 governs the limit of temperature and the quantity of oil used. 
 
 Log of Test. 
 
 Sample: 
 
 OIL FRICTION TEST. 
 Date: 
 
 Time 
 
 Temp. 
 
 Load Speed 
 
 Friction 
 
 Air 
 
 Bearing 
 
 Total 
 
 Sq.In. 
 
 Counter 
 
 Total Ft. 
 
 B. P. M. 
 
 Pull Lbs. 
 
 Coef. 
 
 
 
 
 
 
 
 
 
 
 
 Specification = 
 
 Circum. journal in feet x R. P. M x load 
 
 Troy grains oil 
 
 > 325,000. 
 
 Pull in pounds 
 
 Coeftcient fnct on = 
 
 Total load 
 
 Tests of Oil on Board Ship. Attempts to improvise a testing 
 machine on board ship will be found unsatisfactory. Tests have 
 been made utilizing a lathe with mandrel to improvise a testing 
 machine. The bearing surface thus used must necessarily be small 
 and it will be found difficult to apply and measure a sufficient load 
 for the test. Such apparatus can only give an approximate means 
 of comparison with another oil of known good quality. 
 
 A sample of the oil under examination is sometimes used to lubri- 
 cate the bearings of some part of the auxiliary machinery that is in 
 
228 EXPERIMENTAL ENGINEERING 
 
 constant use. This is not very satisfactory since it will be difficult 
 to completely remove all particles of the old oil before beginning 
 with the sample. A mixture of two different oils will frequently 
 give very different results from those obtained with either one of 
 them when used alone. An instance is on record where a mixture 
 of two of the best known engine oils, made by different manu- 
 facturers,, when used in a thrust bearing, caused a heavy deposit of 
 paraffine, with consequent heating of the bearing. Either oil, when 
 used alone, was perfectly satisfactory. 
 
 The only certain test of a lubricating oil is to use it for the 
 lubrication of a bearing under service conditions and the test 
 should be continued long enough to be conclusive. In a testing 
 machine the bearing is set up close and the load applied is con- 
 stant. In service there is clearance in the bearing and the load is 
 irregularly applied, in extreme cases coming in the form of heavy 
 blows. 
 
 Tests in an oil testing machine on shore are of value in indi- 
 cating the probable character of a new brand of oil, but the final 
 test must be that of actual service. In beginning the use of a new 
 brand care must be taken to carefully clean out all the old oil. 
 
 In general it may be said that no lubricating oil should be pur- 
 chased for use on board ship that has not been carefully tested, or 
 the character of which is not guaranteed by a well-known brand, 
 sold by a reputable dealer. Such brands are sold all over the world 
 and it is poor economy to save a few cents on the gallon in the 
 purchase of an oil that is of inferior quality. When using an in- 
 ferior oil there is danger of overheating the bearings, causing dam- 
 age that will, besides crippling the ship, result in a cost for repairs 
 many times exceeding the amount saved in the purchase of the oil. 
 
CHAPTEE X. 
 THE SELECTION AND TESTING OF FUEL. 
 
 Fuel Economy. Economy in fuel consumption is influenced in 
 two ways. (1) By practicing the greatest possible economy in 
 burning the fuel. (2) By selecting fuel that is best adapted to the 
 conditions under which it is to be burned and that will develop the 
 greatest amount of heat from a given quantity of the fuel. For 
 commercial uses the last consideration is stated in another way, the 
 desire being to select a fuel that will produce the greatest quantity 
 of heat at the lowest cost. This consideration also holds in select- 
 ing ji fuel for naval consumption, but it is not of paramount im- 
 portance, since considerations of military efficiency often lead to the 
 selection of a more expensive fuel on account of its higher heating 
 value. 
 
 Selecting Coal. The following qualities in coal govern its selec- 
 tion for use as a fuel : 
 
 (1) Percentage of Moisture. The moisture present in coal rep- 
 resents a dead loss on account of its weight. Contracts for coal 
 should be based on the net weight of dry coal. 
 
 (2) Percentage of Volatile Matter. Boilers must have a large 
 combustion chamber space in order to efficiently burn coal con- 
 taining a high percentage of volatile matter. Such coal must also 
 be handled differently in firing. Where the coal is fired by hand 
 there should be no great variation in the percentage of volatile 
 matter, since it will in that case be necessary to train the firemen 
 anew with each variable lot of coal received. 
 
 (3) Percentage of Ash. This is of importance as it affects the 
 heating value and there is a further direct loss in cleaning fires. 
 If the ash is fusible, making heavy clinkers, these choke the grate 
 and prevent efficient combustion. 
 
 (4) The Size of the Coal. Lump coal as a rule burns better 
 than fine slack coal, though if slack coal cokes readily it will burn 
 
230 EXPERIMENTAL ENGINEERING 
 
 efficiently. Some of the best American steaming coals come with a 
 very small proportion of lump. Foreign coals should as a rule be 
 avoided when there is not a large proportion of lump. 
 
 (5) Heating Value. This is the most important quality of all. 
 It is stated as the number of B. T. IPs per pound of dry coal, or 
 combustible. The term dry coal is used to signify that from which 
 all moisture has been driven off, or the net weight, deducting for 
 the percentage of moisture. The term combustible is used to desig- 
 nate the net weight, deducting for the percentage of ash. 
 
 For naval use the coal having the highest heating value should 
 as a rule be selected, since the bad qualities usually increase in in- 
 verse proportion to the heating value, and military necessity impels 
 us to fix upon a high grade of coal and require all contractors to 
 furnish coal of this grade. This may also be the most economical 
 policy in point of cost, since with an unknown brand of coal of 
 mferior quality, the firemen who are unaccustomed to it, are 
 likely unknowingly to waste enough to more than make up for the 
 difference in cost. 
 
 For commercial purposes it is now becoming the general practice 
 to contract for coal on the basis of its heating value per unit of 
 cost. Where a low price can be thus obtained it is often possible in 
 a stationary plant to make what arrangements are necessary to burn 
 such coal for a long period of time, the arrangements and the con- 
 tract being made with a view to producing the required power at the 
 lowest price during the time for which the contract is to run. 
 
 Testing Coal. Whatever may be the policy with regard to the 
 quality of fuel that is to be used, it becomes of necessity to test the 
 coal that is under consideration, either practically, by burning it 
 under service conditions, or by careful tests in a properly equipped 
 laboratory. For a test on shipboard, the only practicable procedure 
 is to obtain as large a sample as possible and burn it on the grate 
 of a boiler. The grate should be clean to start the test and the 
 sample should be large enough to make a thorough test under full 
 service conditions. 
 
 Sampling Coal for Test. The object in taking a sample of coal 
 is to obtain a small portion which represents as nearly as possible 
 the entire lot of coal which is under consideration. 
 
THE SELECTION AND TESTING OF FUEL 231 
 
 The original sample should preferably be collected in a large 
 receptacle with cover attached, by taking small shovelsful from 
 many parts of the car,, barge, or vessel as it is being unloaded, or 
 from as nearly all parts of a pile as possible, care being taken in 
 all cases to secure practically the same amounts from the top, 
 middle, and bottom of the pile. If sampling for a service test, the 
 amount taken should be at least sufficient for a day's steaming 
 under service conditions. If for a laboratory test, the first sample 
 should amount to 500 pounds, or more, preferably 1000 to 2000 
 pounds. A separate sample should be taken from each 1000 tons 
 or less delivered. The gross sample thus collected should contain 
 the same proportion of lump and fine coal as exists in the whole 
 shipment. It should be protected from the weather, in order to 
 avoid gain or loss in moisture and should be immediately quartered 
 down to a smaller sample, according to the following method : 
 
 The large lumps of coal and impurities should be broken down 
 on a clean, hard, dry floor, with a suitable maul or sledge. The 
 coal should be thoroughly mixed by shovelling it over and over and 
 formed in a conical pile. The pile should then be quartered, using 
 a shovel or board to separate the four quarters. Two opposite 
 quarters should then be rejected and the remaining two broken 
 down to a smaller size, mixed and reformed in a conical pile and 
 quartered as before. This process should be continued until the 
 lumps are J inch in size, or smaller, and a one or two-quart final 
 sample remains. All of this final sample should immediately be 
 placed in one or more glass or metal cans and sealed air-tight. The 
 outside of the can should be plainly labelled and a corresponding 
 description placed inside the can. 
 
 In preparing a sample the work should be carried on as rapidly 
 as possible in order to avoid loss of moisture through contact with 
 the air. 
 
 Test for Moisture. A portion is accurately weighed into an 
 oven and dried for one hour at a temperature of about 105 F. 
 Then reweighing, the difference gives the percentage of moisture. 
 
 Test for Volatile Matter. A portion of dry coal should be 
 weighed into a flask and heated to incandescence for about fifteen 
 minutes. This will drive off the volatile matter and on reweigh- 
 
232 EXPERIMENTAL ENGINEERING 
 
 ing, the difference gives the percentage of volatile matter. This 
 test is sometimes carried out on an open tray, but this requires a 
 very careful adjustment of temperature. If not hot enough some 
 of the volatile matter will remain and if too hot some of the 
 carbon will be consumed. In computing the percentage of volatile 
 matter the calculation must be based on the weight of that portion 
 of the sample of coal which is used, that is before either the moist- 
 ure or the volatile matter is driven off. 
 
 Test for Ash. A portion of the sample is weighed on a platinum 
 dish, then heated in the open air until all combustible matter is 
 burned. The weight of the residue, compared with that of the 
 portion used gives the percentage of ash. The value thus obtained 
 is the actual net value and is lower than it is possible to obtain 
 when burning the coal on a grate, owing to the sifting through of 
 fine particles of combustible with the ash. The loss due to this 
 cause may, with careless firing, be very large and every effort 
 should be made to reduce it as low as possible. 
 
 Measurement of the Heating Value of Fuels. 
 
 The methods of calculating the theoretical heating value of a 
 fuel, its evaporative power, the amount of air required to burn it, 
 the temperature of the furnace that it is possible to obtain and the 
 method of conducting boiler tests have been given in other text- 
 books. 
 
 This work will include under this heading only such apparatus 
 as is used in the experimental determination of the heating power 
 of a fuel. 
 
 Mahler's Calorimeter. This instrument is perhaps the best 
 known of the bomb calorimeters, in which a sample of the fuel under 
 test is burned in a bomb immersed in water and the amount of the 
 heat of combustion is measured by the rise in temperature of the 
 water. 
 
 Principle of the Apparatus. The combustible is placed in a 
 closed bomb, made strong enough to resist h^avy pressure. Oxygen 
 is introduced under pressure and the gases of combustion are con- 
 fined in the bomb. The bomb is immersed in the water of the cal- 
 orimeter and the combustion is started by an ignition contrivance. 
 
THE SELECTION AND TESTING OF FUEL 233 
 
 On account of the large quantity of oxygen the combustible burns 
 completely and almost instantaneously. Its heat is given off and 
 transmitted to the water and to the various parts of the apparatus, 
 such losses as occur being easily estimated in all calorimetric opera- 
 tions. Owing to the rapidity of the operation most of the correc- 
 tions that would be made in a physical laboratory become negligible ; 
 for example, that which takes account of the evaporation of the 
 water. 
 
 Description of the Apparatus. The apparatus is shown in Fig. 
 124, in which A is the water jacket, B the bomb of enamelled steel, 
 the platinum tray for holding the fuel, D the calorimeter, E an 
 electrode, F a piece of fine iron wire for priming, G the support 
 for agitator, K the mechanism of agitator, L the lever for operat- 
 ing, M a pressure gage for oxygen, the flask of oxygen, P an 
 electric battery, 8 the agitator, T the thermometer, Z a clamp for 
 holding the bomb while removing or replacing the cap. 
 
 The bomb is of forged steel of the best quality. It is of about 
 650 cc. capacity, with walls 8 mm. in thickness. This capacity is' 
 such as to assure in all cases perfect combustion of the fuel by a 
 considerable excess of oxygen. It also enables the bomb to be used 
 for experiments on gases and gas mixtures containing as much as 
 70% of inert matter, where it is necessary to take a large quantity 
 if a rise in temperature is to be obtained sufficiently great for satis- 
 factory results. 
 
 The bomb is nickel-plated on the outside. On the inside a coat of 
 enamel preserves it against the action of nitric acid, which is al- 
 ways formed during the combustion. The bomb is closed by a cap, 
 screwed down on a lead gasket. This cap has a valve in its center, 
 with screwed nozzle for connecting to the flask of oxygen. It is also 
 pierced by a platinum electrode, well insulated, prolonged on the 
 inside by a platinum rod E. A second platinum rod, fixed to the 
 cap, sustains the platinum tray C, which carries the sample of fuel 
 under test, 
 
 A spiral of very fine iron wire connects C with E, coming in con- 
 tact with the fuel when the bomb is charged. When the current is 
 turned on this wire is heated to redness, then burns in the atmos- 
 phere of oxygen, igniting the fuel. 
 
N 
 
THE SELECTION AND TESTING OF FUEL 235 
 
 The agitator consists of vanes, carried on a central rod with spiral 
 thread passing through a fixed nut. It is operated by a lever which 
 permits the operator to systematically stir the water in the calo- 
 rimeter, thus ensuring an even temperature. 
 
 The valve on flask does not have a fine enough adjustment to 
 permit gradual introduction of the oxygen. A second valve, not 
 shown in the figure, having a very fine adjustment, is placed in the 
 connection to tank for this purpose. The figure shows a flask for 
 oxygen containing about 1000 liters. It is ordinarily supplied in 
 this manner at about 120 atmospheres pressure. Since the pres- 
 sure convenient for the combustion of one gram of coal is only 
 about 25 atmospheres, there is thus a provision for about 60 tests. 
 
 A high-grade thermometer reading to 1/50 C., an electric bat- 
 tery of 12 volts and 2 amperes capacity, and a stop-watch complete 
 the apparatus. 
 
 Determination of the Calorific Value. The following method 
 of procedure for determining the calorific value of a solid or liquid 
 combustible is that given by the inventor of the apparatus. 
 
 One gram of the fuel is weighed and placed in the tray C. The 
 small iron wire F of a known weight is adjusted in contact with 
 the fuel and serves as a primer. After having introduced all in the 
 bomb, it is placed in the clamp Z and the cap is screwed on hard 
 by means of a heavy hexagonal wrench. 
 
 The valve on cap is then opened, the second valve for fine ad- 
 justment having first been closed. The valve on flask is opened 
 and then, very slowly, the adjusting valve, until the gage indicates 
 25 atmospheres. After having closed all valves the tube is then 
 disconnected. 
 
 The bomb, thus prepared, is placed in the calorimeter D. The 
 thermometer T and agitator 8 are placed in position and a meas- 
 ured quantity of water, sufficient to completely cover the bomb, is 
 poured in. This quantity will be about 2200 cc., which is the 
 amount used by M. Mahler in his experiments. The water is 
 stirred for some minutes, in order to let the whole system arrive at 
 an even temperature, then observations are commenced. 
 
 The temperature is noted from minute to minute for 5 minutes, 
 in order to fix the rate of variation of the thermometer before igni- 
 
236 EXPERIMENTAL ENGINEERING 
 
 tion. At the end of the fifth minute contact is made and the fuel 
 fired by means of the battery connected to the electrode E and to a 
 point on the valve. Ignition takes place immediately. 
 
 The temperature is noted half a minute after the contact is made, 
 then at the end of a minute, and the observations are continued from 
 minute to minute up to the point where the thermometer commences 
 to fall regularly. This is the maximum. 
 
 The observations are then continued for five more minutes in 
 order to fix the rate of variation of the thermometer after it reaches 
 the maximum. 
 
 The principal data for the calculations are then at hand, in- 
 cluding data for the correction for loss of heat by radiation from 
 the calorimeter. This correction is made according to the following 
 rules, true between large limits, even where the amount of con- 
 tained water is not more than half the water equivalent of the 
 calorimeter. 
 
 (1) The rate of decrease of temperature, observed after reaching 
 the maximum, represents the rate of loss of heat from the calorim- 
 eter before reaching the maximum, provided the fall in temperature 
 is not greater than 1 C. per minute. 
 
 (2) If the fall in temperature per minute is greater than 1, but 
 less than 2 C., the figure representing the rate of decrease, when 
 diminished by 0.005, gives the desired correction. 
 
 The two preceding paragraphs cover all cases. It is possible also, 
 and that without altering the accuracy of the experiment, to con- 
 sider the variation during the first half of the minute following the 
 ignition as that which exists at the minimum temperature. 
 
 During the whole experiment, the observer should continually 
 operate the agitator. 
 
 When the observations are ended, the valve on the bomb is first 
 opened, then the bomb itself. The bomb will contain the ordi- 
 nary products of combustion, composed principally of carbonic acid 
 gas and water, a considerable quantity of free oxygen, and an appre- 
 ciable quantity of nitric acid formed during the combustion from 
 such nitrogen as was present in the bomb at atmospheric pressure 
 before it was charged with oxygen. 
 
 The interior of the bomb is washed with a small quantity of water 
 
THE SELECTION AND TESTING OF FUEL 237 
 
 to remove the liquid acid formed during the explosion. The amount 
 of nitric acid is then determined by a simple chemical analysis, and 
 the calorific value, h, is determined from the formula : 
 
 fc = r(l + a)(P + P')-(230p + 1600p'), wnere 
 r =: the rise in temperature. 
 
 a = the loss of temperature during the experiment. 
 P = the weight of water in the calorimeter. 
 P' = the water equivalent of the calorimeter. 
 p the weight of nitric acid. 
 p' = the weight of the iron ignition wire. 
 230 = the heat of formation of one gram of nitric acid. 
 1600 = the heat of combustion of one gram of iron. 
 
 In making a test of coal no separate account is taken of the 
 quantity of sulphuric acid, resulting from the oxidation of the small 
 quantity of sulphur present in the sample, such acid being treated 
 as nitric acid. The error is negligible in ordinary work. It may be 
 noted that the sulphur being entirely oxidized and transformed into 
 sulphuric acid, the bomb gives a means of evaluating it. For this 
 purpose, in order to give a sufficient quantity for satisfactory opera- 
 tion, it will be better to burn 2 grams under 30 atmospheres, with- 
 out taking readings of the thermometer. 
 
 If desired, account may be taken of the heat generated by the 
 formation of sulphuric acid, which is 0.73 calories per gram of acid. 
 
 In testing a substance containing but little hydrogen, coke for 
 example, so little water of combustion is formed that the quan- 
 tity is insufficient to dissolve the acid. It is then best to place in 
 the bottom of the bomb a few cc. of water, which must be taken 
 into account in making the calculations. 
 
 The procedure is the same for a liquid as for a solid. If the 
 liquid gives off vapors it is well to weigh the sample in a closed vial, 
 having thin points through which is passed the film of iron wire. 
 At the moment of introducing the vial into the bomb, care should 
 be taken to break these points in order to bring the oxygen into con- 
 tact with the liquid. 
 
 M. Mahler has also used the apparatus for the determination 
 of the calorific value of various gases. After having exhausted the 
 bomb and measured the pressure remaining, the gas is introduced 
 
238 EXPERIMENTAL ENGINEERING 
 
 for the first time. The bomb is then exhausted a second time, after 
 which it is filled with the gas at barometric pressure and at the tem- 
 perature of the laboratory. The oxygen is then added and the pro- 
 cedure is carried on in the same manner as for solid and liquid fuels. 
 
 The determination of the calorific value of gases offers a special 
 difficulty. If diluted with too great a quantity of oxygen, the mix- 
 ture will not be combustible. For illuminating gas, 5 atmospheres 
 will be sufficient. For producer gas, half an atmosphere only should 
 be used, measured. on a mercurial manometer. 
 
 Determination of the Water Equivalent of the System. In 
 order to determine the value of P f , the term representing in water 
 the exact equivalent of the system, the simplest method is to per- 
 form a double experiment as follows: 
 
 Burn in the bomb a known weight, one gram for example, of a 
 combustible or fixed composition, such as fuel oil, and with 2300 
 grams of water in the calorimeter. Then burn the same weight of 
 the same combustible with only 2100 grams of water in the 
 calorimeter. 
 
 There will then be two equations between which the heat of com- 
 bustion of the fuel may be eliminated and the value of the water 
 equivalent may be deduced. 
 
 Example. The following example of the work of the apparatus 
 is given by M. Mahler : 
 
 The fuel under test is a sample of colza oil. An approximate 
 analysis gave : 
 
 Carbon 77.182 
 
 Hydrogen 11.711 
 
 Oxygen and Nitrogen 11.107 
 
 100.000 
 
 Weight of sample tested, 1 gram. 
 
 Water in calorimeter, 2200 grams. 
 
 Water equivalent of the bomb and accessories, previously deter- 
 mined, 481 grams. 
 
 The apparatus being prepared as above directed, a little time is 
 allowed to elapse for the temperature to equalize, then the stop 
 watch is started and the temperatures are noted as below. 
 
THE SELECTION AND TESTING or FUEL 239 
 
 Preliminary Period. 
 
 minutes ................................... 10.23 
 
 1 " ................................... 10.23 
 
 2 " ................................... 10.24 
 
 3 " ......................... ......... 10.24 
 
 4 " ................................... 10.25 
 
 5 " ................................... 10.25 
 
 5 
 
 The combustible is then fired. 
 
 Period of Combustion, 
 
 5y 2 minutes ................................. 10.80 
 
 6 minutes ................................... 12.90 
 
 7 " ................................... 13.79 
 
 8 " ................................... 13.84 (max.). 
 
 Final Period. 
 
 9 minutes .................................. 13.82 
 
 10 " .................................. 13.81 
 
 11 " .................................. 13.80 
 
 12 " .................................. 13.79 
 
 13 " .................................. 13.78 
 
 5 
 
 No further readings of the thermometer are taken. 
 The change in temperature has been 13.84-10.25 = 3.59. 
 Corrections. The apparatus has lost during the minutes (7, 8) 
 (6, 7) a quantity of heat measured by 
 
 13.84-13.78 X 2 = 0.012(1)X2 = 0.024. 
 
 During the half minute (5J, 6) it has lost a quantity of heat 
 represented by 
 
 (0.012-0.005) xi = 0.0035. 
 
 And during the half minute (5, 5J) it gained 
 10.25-10.23 
 
240 EXPERIMENTAL ENGINEERING 
 
 Finally, the loss during the minute (5, 6) is 
 
 0.0035-0.002 = 0.0015. 
 To sum up, the loss during the whole experiment has been 
 
 0.024 + 0.0015 = 0.0255, 
 
 a quantity which should be added to the 3. 59 already found. 
 
 The corrected rise in temperature is then 3. 6 15, neglecting the 
 ten thousandths. 
 
 The quantity of heat observed is therefore 
 
 (2200 + 481) X3.615 = 9691.8 calories. 
 
 In order to obtain the final result we subtract from this figure 
 
 (1) The heat of formation of 0.13 gram of nitric acid, deter- 
 mined volumetrically, 0.13x230 = 29.9 cal. 
 
 (2) The heat of combustion of 0.025 gram of iron wire, 
 0.025x1600 = 40.0 cal. 
 
 Amount to be deducted, 69.9 cal. 
 
 The final result is then 9691.8-69.9 = 9621.9 calories. 
 Or, for a kilogram of oil, 9621.9 kilo-calories. 
 To transform this result into B. T. U. per pound of oil, multiply 
 by 1.8. 
 
 9621.9x1.8 = 17,319.42 B. T. U. 
 
 Applying the formula h = 14,500 (C + 4.28(71-0/8) ) we obtain 
 for h the theoretical value 17,597. 
 
 The dimensions of this apparatus are such that it is possible to 
 so regulate the conditions of a test as to cancel the minor correc- 
 tions. If a is the correction due to the loss of heat during the 
 operation, it will be seen from inspection of the equation on page 
 237, that it will not be necessary to take into account such correc- 
 tions if 
 
 since the equation then becomes h = r(P + P'). 
 
 Since the value of p depends chiefly on the quantity of nitrogen 
 contained in the bomb before charging with oxygen, it will, when 
 testing solid or liquid fuel, be practically constant, p' is variable 
 at the pleasure of the operator within small limits, but mav also be 
 
THE SELECTION AND TESTING OF FUEL 
 
 241 
 
 given a constant value. It is possible then to so regulate the values 
 of a and P that the above equation will be sufficiently true in all 
 ordinary operations. This has been done by M. Mahler with the 
 apparatus under his own direction. For example, the calorific value 
 of colza oil under examination was found to be 9621.9 calories. 
 Using the values 3.59 and 481 which were found for r and P' 
 respectively, we have 
 
 r(P + P') =:3.59(2200 + 481) =9624, 
 which approximates very closely to the exact result obtained. 
 
 FIG. 125. 
 
 This approximate method is that usually followed in testing 
 a solid or liquid fuel. If somewhat greater accuracy is desired, the 
 value, p = Q.l3, found in the above experiment may be used with the 
 observed values of a, p' and P. In testing a gas, where the propor- 
 tion of nitrogen is large, it will be necessary to evaluate the quan- 
 tity of nitric acid. 
 
 Carpenter's Improved Coal Calorimeter. In construction this 
 instrument may be compared to a thermometer, in the bulb of 
 which combustion takes place, the heat of combustion being ab- 
 sorbed by the liquid contained in the bulb, and the amount of this 
 heat being proportional to the expansion of the liquid. The ex- 
 pansion is measured by the rise of the water level in an open glass 
 16 
 
242 EXPERIMENTAL ENGINEERING 
 
 tube, from which the number of B. T. IPs absorbed by the liquid 
 is determined. As set up for use, the complete apparatus is shown 
 in Fig. 125 and consists of the calorimeter proper, a flask of oxygen, 
 a pressure reducer for oxygen, and the electric firing connections. 
 Description. Eeferring to Fig. 126 : The combustion chamber, 
 a, is supplied with oxygen through tube b, the products of com- 
 bustion being conducted through spiral tube c, c. The tube ends 
 in a hose nipple d, from which a hose connection is made to a 
 small chamber e, attached to the outer case, and fitted with a 
 manometer / for measuring the pressure of combustion. A plug 
 g, with pin hole, is attached to the chamber for the discharge of 
 gases. Surrounding the combustion chamber a f is a large closed 
 chamber h, filled with water and connected to the open glass tube i. 
 The level of water in i is measured on the scale /. Above the water 
 chamber is a diaphragm k which, by means of the screw I, is used 
 to adjust the zero level to any desired point in the open glass tube i. 
 Glass is fitted at m, n, and o, so that the process of combustion may 
 be observed. A screw plug p, is removed when filling or emptying 
 the water chamber. The plug q, which stops up the bottom of the 
 combustion chamber, carries a dish r, in which the fuel for com- 
 bustion is placed, and two vertical adjustable insulated wires s, s, 
 the upper ends of which are joined by a thin platinum wire. 
 These wires are connected to an electric battery, or circuit, the 
 current of which is used for firing the fuel. A silver mirror on 
 top of the plug deflects any radiating heat. The plug itself is so 
 constructed and fitted that little or no heat is transferred to the 
 outside, but practically all heat is rapidly transmitted to the sur- 
 rounding water. The instrument is supported on strips of felting 
 u and v f and fits into a nickel-plated case, the inside of which is 
 polished to reduce radiation. Eight inches water pressure has been 
 found sufficient for combustion. 
 
 The sample of coal to be tested weighs about 1J grams and is 
 burned in the combustion chamber in an atmosphere of oxygen, 
 kept at a constant pressure of 8 inches of water. Oxygen is re- 
 ceived in the laboratory in heavy flasks under about sixty atmos- 
 pheres' pressure. This pressure, in the ordinary form of the ap- 
 paratus, is reduced by throttling to the low pressure used for com- 
 
THE SELECTION AND TESTING OF FUEL 
 
 243 
 
 <P 
 
 FIG. 126. 
 
244 
 
 EXPERIMENTAL ENGINEERING 
 
 bustion. The reduction is so great, however, as to make perfect 
 regulation with a simple throttle valve very difficult. It has been 
 found of great advantage to introduce a pressure reducer between 
 the flask and the combustion chamber. This apparatus, as- designed 
 by Lieutenant-Commander F. D. Karns, U. S. N., and built at 
 
 o 
 
 the Naval Academy, is shown in Fig. 127, and is described as fol- 
 lows: 
 
 Pressure Reducer for Oxygen. Eeferring to Fig. 127: A is 
 a cylindrical chamber open at the top and closed at the bottom 
 by a cap R. B is a 3-inch pipe lowered into A, with its lower end 
 resting on three studs P, as shown. The upper end of B is closed 
 
THE SELECTION AND TESTING OF FUEL 245 
 
 by the cap C. Studs Q serve to keep B upright. Two J-inch pipes, 
 D and E, enter the lower end of A, pass up through the inside of B } 
 with their open ends terminating with J inch of cap C. Water 
 fills the chamber A about two-thirds full and, with valve open, 
 rises to the same level m, ra', m, in A and B, the level being at all 
 points under atmospheric pressure. Oxygen from the flask under 
 pressure is admitted through D to the clearance space above the 
 water level in B, creating a pressure with valve closed that forces 
 some of the water out of B, to the level n', thereby causing the 
 level in A to rise to the level n, n, where a balance is established 
 against the pressure of oxygen in B. The rise of the water is 
 noted in the gage glass, and the corresponding pressure in inches 
 of water is read off on the attached scale F. The pressure in B is 
 controlled by means of the throttling valve H and maintained at 
 12 inches water pressure, which is found to be that best suited to 
 the test requirements. From B the oxygen is conveyed through 
 the pipe E and valve L, where it is further throttled and kept at a 
 pressure of 8 inches of water as it discharges to the combustion 
 chamber of the calorimeter. Valve is for the purpose of placing 
 B under atmospheric pressure for marking off the zero water level. 
 It also serves to permit the air in B to escape, so that the clearance 
 space may be completely filled with pure oxygen before beginning 
 a test. 
 
 The scale F is graduated to read in inches water pressure, the 
 value corresponding to any observed water level in the gage glass 
 being determined as follows: 
 
 Referring to Fig. 127 : 
 
 Let A^ = area of cross section in A (annular space). 
 " A 2 = area of cross section in B (excluding small pipes). 
 " h^ amount water level rises in A. 
 " h 2 = amount water level falls in B. 
 
 7 .. . A 
 
 Then h 2 X A 2 = h^x A 19 and Ji 2 = 
 Head, in inches, =h i -\-h 2 - 
 
 A, 
 
 A 2 
 
 since A and A 2 are both known constants. h is measured by the 
 rise of the water in the gage glass. For purposes of graduating 
 
246 EXPERIMENTAL ENGINEERING 
 
 the scale, the "head" is marked off and indicated for heights h lf 
 increasing by one quarter of an inch. Assume the simplest case, 
 in which A 1 = A 2 . Then head = 2h l : > and if the actual rise of the 
 water is one inch, the corresponding water pressure is two inches, 
 and would be so marked on the scale. 
 
 Method of Conducting a Test. (1) Preparation of the Sample. 
 Select an accurate sample by a system of quartering, reduce it 
 to powder, place in a dry asbestos cup of known weight and weigh 
 accurately. 
 
 (2) Adjustment of Gas Pressure. Open valve and adjust 
 zero point of scale F to water level in gage glass. Exhaust air from 
 clearance space of B, taking care that no water enters the small 
 pipes, and then close 0. Admit oxygen from flask through pipe 
 D and valve H until water level rises to 12 on scale F, and keep it 
 there by closing or manipulating H. During this time valve L is 
 kept closed. 
 
 (3) Firing the Charge. Introduce coal sample into calorimeter, 
 raise platinum wire above coal, make battery connection, and as 
 soon as heat from wire causes water in glass tube to begin to rise, 
 turn on oxygen gas (by opening valve L at the reducer), and fire 
 charge by pulling heated wire down. The instant coal is lighted, 
 break electric connection and note first scale reading and time. 
 The gas pressure in the combustion chamber must be kept constant 
 at eight inches during the test, and this is done by properly throt- 
 tling the valve L, noting the pressure indicated by the manometer 
 attached to the discharge side of the calorimeter. 
 
 (4) Actual Scale Reading. Watch the combustion, which usu- 
 ally requires about 10 minutes for each gram of coal, and when 
 completed note second scale reading and time. The difference be- 
 tween second and first scale readings is the " actual " scale reading. 
 
 (5) Correction for Radiation. Allow the calorimeter to stand 
 for a time equal to, and under the same conditions as, that of 
 combustion, except that the oxygen gas is shut off, and note third 
 scale reading and time. 
 
 (6) Corrected Scale Reading. The difference between second 
 and third scale readings is the correction for radiation, and must 
 be added to the " actual " scale reading to get the " corrected " 
 reading. 
 
THE SELECTION AND TESTING OF FUEL 247 
 
 (7) To Find Calorific Value. From the calibration curve find 
 the heat value of the sample in B. T. U. corresponding to the " cor- 
 rected " scale reading, and divide the B. T. U. thus obtained by the 
 weight of the sample in pounds. The result will be the calorific 
 value in B. T. U. per pound of coal. 
 
 (8) To Determine the Ash. Weigh the cup in which combus- 
 tion took place, with its contents. From this weight subtract the 
 known weight of the cup, and the difference will be the weight of 
 the ash in the sample. 
 
 General Instructions. (1) To prepare for another test, remove 
 calorimeter from outside case and immerse in cold water for a few 
 minutes, care being taken to prevent any water entering tubes or 
 combustion chamber. 
 
 (2) It is important that the water in the calorimeter should be 
 free from air and that the oxygen gas be supplied at a constant 
 pressure. 
 
 (3) In burning coals having a large percentage of volatile mat- 
 ter, the water resulting from the combustion may affect the rate of 
 flow of the burned gases, and thus change the reading of the ma- 
 nometer. In this case the result of the test is of uncertain value. 
 
 (4) The temperature of the calorimeter at the beginning of a 
 test should be a few degrees above that of the surrounding atmos- 
 phere. 
 
 (5) Complete combustion will always be obtained when asbestos 
 cups are used. 
 
 (6) An asbestos cup is made by wrapping a piece of sheet 
 asbestos about the end of a small cylinder, and using a weak glue 
 to hold it in cup form. The cup is then heated to a white heat 
 in order to remove any combustible matter. The cup when weighed 
 must be dry. 
 
 (7) Oxygen for combustion is usually purchased in condensed 
 form in heavy steel flasks, but in case the supply should at any 
 time become exhausted, it may be made by heating a mixture of 
 about equal parts of dioxide of manganese and chlorate of potash 
 placed in a closed retort. 
 
 (8) Electric current for heating the platinum wire may be 
 taken from a dry battery or reduced from a lighting circuit. 
 
248 EXPERIMENTAL ENGINEERING 
 
 (9) The calorimeter will give good results only when all the 
 conditions under which the calibration was made are maintained 
 during the test. 
 
 Calibrating the Calorimeter. (1) Preparation of Sample. 
 Reduce some charcoal from sugar or soft coal to powder, fill a 
 porcelain or clay crucible one-third full, cover it tightly, and heat 
 it by means of a blast lamp or a forge fire for half an hour. When 
 cold, grind it in a mortar to very fine powder. Repeat this opera- 
 tion for other samples to be tested. 
 
 (2) To Free the Water in the Calorimeter from Air. Connect 
 the glass tube opening by rubber hose with a smaller vessel filled 
 with water. Boil the water in the calorimeter, using a burner and 
 protecting the calorimeter by a thin sheet of asbestos paper. All 
 air and steam will pass to the smaller vessel, which must be kept 
 boiling until the calorimeter has cooled off. Then remove rubber 
 connection and insert glass tube, taking care that no air is trapped. 
 The instrument is now ready for calibration. 
 
 (3) Testing the Sample. Follow the instructions already given 
 under " Method of Conducting the Test." The difference between 
 the weight of cup and sample, and the weight of cup and ash, is 
 the weight of pure carbon burned. Multiply the weight of pure 
 carbon burned by 14,600 and obtain the number of heat units in 
 the sample. Find this for several weights. 
 
 (4) To Construct the Curve of Calibration. With the "cor- 
 rected" scale readings as ordinates, and the corresponding heat 
 values of the several samples burned as abscissa, make points in 
 crossing and through these points draw a fair curve. All the points 
 should lie on a straight line whose origin is at zero. 
 
 The calibration curve, shown in Fig. 128, was determined as the 
 result of a series of tests made in the Engineering Laboratory. 
 
 The Parr Standard Calorimeter. This instrument, devised by 
 Professor S. W. Parr, of the University of Illinois, is a bomb 
 calorimeter in which the fuel under test is placed in the bomb to- 
 gether with chemicals containing the constituents necessary for its 
 complete combustion and for the further absorption by chemical 
 combination of the gases given off in the process of combustion. 
 
THE SELECTION AND TESTING OF FUEL 
 
 249 
 
 Slg 
 
 NO CM 
 
250 
 
 EXPERIMENTAL ENGINEERING 
 
 The instrument is shown in Fig. 129. A is a can in which, for 
 the test, two liters of water are placed. B and C are the outer and 
 inner walls of the containing vessel, divided by an air space be- 
 tween them. To further prevent loss of heat by radiation, they are 
 made of non-conducting material. 
 
 FIG. 129. 
 
 Combustion takes place in the bomb D. This rests on a pivot and 
 is rotated at a speed of about 100 revolutions per minute by means 
 of a small motor. This is for the purpose of ensuring a uniform 
 temperature of the water, which is thus stirred by means of the 
 vanes shown in the figure. 
 
 Two forms of ignition are used. As first designed, the charge 
 
THE SELECTION AND TESTING OF FUEL 
 
 251 
 
 -K 
 
 B 
 
 was ignited by dropping into the bomb a small piece of red-hot 
 wire. The bomb as fitted later for electrical ignition is shown in 
 section in Fig. 130. A is the shell of the bomb. B is the tube 
 whose lower end fits gas and water tight over A. 
 F is a cap for securing B. C is the bottom of 
 the bomb, secured gas and water tight by the plug 
 D. G is the ignition wire, which passes through 
 tube with insulated gas-tight fittings to contact 
 point at K. 
 
 Operation. This instrument is used for soft 
 coal, anthracite or coke, and for oil fuels. For 
 each of these different classes the charge of fuel 
 and chemicals is made up in different proportions, 
 as directed, and the proper constant, as determined 
 by the inventor, is applied for the calculation of 
 the desired result. 
 
 To prepare the cartridge for filling, dry all the 
 parts perfectly inside and out; see that the inner 
 bottom C with gasket is properly seated, screw on 
 the outer bell E, then with the spanner wheel screw 
 up -firmly the outer bottom D and place on a sheet 
 of white paper. 
 
 For coals under test, the sample is in all cases 
 prepared by grinding in a mortar and passing 
 through a 100-mesh sieve. Coals containing more 
 than 2 or 3 per cent of water should be dried be- 
 fore testing. In such cases the exact charge of the 
 commercially dry coal is weighed out and dried for 
 an hour at a temperature of 220 -230 F., then 
 transferred to the cartridge. One-half gram con- FIG. 130. 
 stitutes the charge for all varieties of coal. 
 
 The apparatus as supplied by the manufacturers includes the 
 various chemicals that are to be used for testing the different 
 varieties of fuel, as well as the instruments necessary for measuring 
 or weighing them. 
 
 Procedure for Soft Coals. One full measure of sodium per- 
 oxide, from the bottle labelled " Chemical/' is put in the cartridge 
 and the charge of coal added ; then, one gram of finely ground chem- 
 
 Hllll 
 
252 EXPERIMENTAL ENGINEERING 
 
 ically pure chlorate of potash, from the bottle labelled " Chlorate 
 Mixture." It is well to have the latter follow the coal and thus 
 clean out the vessel in which it was weighed. 
 
 The stem and top B, Fig. 130, with the terminals HI, having a 
 loop of fine ignition wire extending about an inch below, are put in 
 position, and the cap F screwed firmly in place. Shake vigorously 
 to thoroughly mix the contents. When the mixing is complete, tap 
 the cartridge lightly to settle the contents and to shake all the ma- 
 terial from the upper part of the cylinder. Put on the spring clips 
 with vanes. The cartridge is now put in place, the can with water 
 being already in position. Adjust the cover. Insert the thermometer 
 so that the lower end of bulb will be about midway towards the 
 bottom of the can, place the pulley on the stem and connect with the 
 motor. The cartridge should turn towards the right, so as to deflect 
 the currents downward. 
 
 After about three minutes, the reading of the thermometer may 
 be taken, giving the initial temperature of the water. This should, 
 for good work, be about 3 Fah. below the temperature of the atmos- 
 phere. The current is then turned on, igniting the charge. Com- 
 bustion will be indicated by a rapid rising of the temperature of the 
 water, which should reach its maximum in from four to five min- 
 utes. During the combustion the revolving of the cartridge must 
 be kept up continuously. 
 
 Calculation of Results. The initial temperature is subtracted 
 from the maximum temperature to give the rise in temperature 
 of the water. From this is subtracted the correction factor for 
 the heat of the wire and chemical, as indicated on the small bottle 
 of chlorate mixture. The remainder is multiplied by 3115, and the 
 product thus obtained is the number of B. T. U. in a pound of the 
 fuel. 
 
 The factor 3115 is deduced as follows: The water used, plus 
 the water equivalent of the calorimeter is 2134 grams. In the reac- 
 tion that takes place at the time of combustion, 73% of the heat 
 produced is due to the combustion of the coal, and 27% is due to the 
 heat of combination of CO 2 and H 2 with the chemical. If now 
 ^ gram of coal causes 2134 grams of water to rise r degrees, and if 
 only 73% of this is due to combustion, then 0.73x2134x2x 
 r=rise in temperature that will result from the combustion of one 
 
THE SELECTION AND TESTING OF FUEL 253 
 
 gram of the coal. 0.73x2134x2 = 3115. From this we see that 
 one gram of the coal will raise 3115 grams of water through r 
 degrees, or one pound of the coal will raise 3115 pounds of water 
 through r degrees. 
 
 After using, dismantle, and thoroughly clean the instrument. 
 Eemove the thermometer, pulley, and cover; then take out the can 
 and contents entire, so that the lifting out of the cartridge will not 
 drip water into the dry parts of the instrument. Remove the spring 
 clips and unscrew the ends. It is better to loosen the bottom D, 
 Fig. 130, and unscrew the entire bell E for cleaning. The fused 
 mass is easily driven out at the bottom by aid of a short metal rod, 
 or it may be dissolved out by immersing the cartridge and contents 
 in hot water. The cartridge and ends, when rinsed clean and thor- 
 oughly dried, will be ready for another test. 
 
 Procedure for Anthracite or Coke. The method is the same as 
 for soft coal, except that instead of 1 gram of " Chlorate Mix- 
 ture," H grams of a persulphate mixture, containing 2 parts potas- 
 sium persulphate (KS0 4 ) to 1 part ammonium persulphate 
 (NHSOt), from the bottle labelled " Special Chemical " is substi- 
 tuted. With this exception, the procedure and method of calculat- 
 ing results is the same as before. The correction factor for the 
 " Special Chemical " and fine wire is marked on the label of the 
 " Special Chemical " bottle. 
 
 Procedure for Oil Fuels. Prepare a small light 15 cc. weighing 
 flask with perforated cork and dropping tube with common rubber 
 bulb cap. Fill the flask about f full of oil and weigh carefully. 
 Place in the cartridge J measure of ordinary " Chemical" (sodium 
 peroxide) and 1^ grams, exactly weighed out, of the " Special 
 Chemical" (persulphate mixture). Then, by means of the drop- 
 ping tube, add about 0.3 gram (25-30 drops) of oil. Add one 
 measure of ordinary " Chemical." Screw on the top, shake well, 
 place in the calorimeter, and ignite as usual. Weigh back the flask 
 carefully and determine the amount of oil taken by difference. 
 Compute the result by formula instead of using a factor, thus : 
 Correcting as before for the wire and " Special Chemical," let 
 r=the corrected rise in temperature; then 
 
 rX 0.73x2134 -D m TT j -i 
 
 , ., . - =B. T. U. per pound of oil. 
 
 wt. oi oil in grams 
 
254 
 
 EXPERIMENTAL ENGINEERING 
 
 Junker's Calorimeter. This is an apparatus especially designed 
 for gaseous fuels. It is shown in elevation and section in Figs. 
 131 and 132. The principle of its action is based on the heating of 
 a current of air passing around a jet of burning gas whose heating 
 value is to be determined. A combustion chamber A has a burner 
 
 FIG. 131. 
 
 at the center which is connected to the source of supply through an 
 accurate gas meter. The combustion chamber is surrounded by a 
 water jacket BB, this jacket being again surrounded by a closed 
 annular air space C, through which the air cannot circulate, and 
 fitted for the purpose of preventing loss by radiation. The nozzle 
 D in the center of the container E is connected to any water supply ; 
 F is an overflow pipe, the discharge from which should be visible, 
 
THE SELECTION AND TESTING OF FUEL 
 
 255 
 
 but need not be measured, as this water does not form part of the 
 quantity heated. 
 
 FIG. 132. 
 
 The cold water enters the container at D, passes down through 
 G, H, and leaves the jacket at L, while the combustion gases enter 
 at M, through a series of small vertical tubes T, in the water jacket, 
 pass down and leave at N. The gas and water move in opposite 
 
256 EXPERIMENTAL ENGINEERING 
 
 directions, so that all the heat is transferred to the water, and the 
 waste gases leave the apparatus approximately at atmospheric 
 pressure. The temperature of the water entering and leaving can 
 be read by thermometers and 'P. A third thermometer gives the 
 temperature of the products of combustion from the gas jet, and a 
 fourth that of the gas in the meter Q. The water produced by the 
 combustion of the hydrogen in the burning jet and the oxygen of 
 the air forming steam is condensed in the calorimeter, thus giving 
 up its latent heat. This condensed water is a factor in the determi- 
 nation of the higher or lower heating value of the gas, consequently 
 it is measured by running it into the graduate K. The cocks 
 should be opened for a short interval before a test in order to clear 
 the pipes of air and obtain water only. 
 
 Operation. Water is run into D until it passes through the 
 jacket and flows out at the discharge L. The gas is lighted at 
 the burner, inserted in the calorimeter, and the size of the flame 
 or quantity of gas passing to it is carefully adjusted by the gas 
 supply stop cock. The air for combustion is supplied from the 
 open bottom, as shown by the arrows. The products of combustion 
 move to the upper part of the calorimeter, and descend through 
 a number of small vertical copper tubes, finally emerging at N, 
 and go to waste through U. The heat in the products of com- 
 bustion is absorbed by the water which circulates round the tubes 
 as it passes through the calorimeter. 
 
 The cock V (Fig. 131) regulates the quantity of circulating water 
 which is measured by the graduated glass X, and by varying the 
 opening of V before getting underway for the test the difference 
 of temperature between inlet and outlet is approximately deter- 
 mined at the beginning. This adjustment is made so that the 
 circulating water during the passage of the gases absorbs all the 
 heat given out, and the products of combustion at release are prac- 
 tically at the temperature of the atmosphere. 
 
 It is necessary that the flow of circulating water should be, as 
 nearly as possible, constant. In order to accomplish this, the water 
 entering at D supplies the tank with a little more than is necessary, 
 and this surplus overflows into the space W in the tank and thence 
 into the drain pipe. Similarly an overflow is provided at Y. By 
 
THE SELECTION AND TESTING OF FUEL 257 
 
 this means a constant head, is maintained, and a constant rate of 
 flow obtained through the calorimeter. The water, in its passage 
 at entrance, passes through the thermometer 0, from which the 
 initial temperature is obtained and the final temperature by the 
 thermometer P. 
 
 Having obtained the pounds of circulating water per minute, the 
 rise in temperature during the passage through the calorimeter, and 
 the number of cubic feet of gas used per minute, the calorific value 
 h f of the fuel is obtained from the formula 
 
 where W pounds of circulating water measured, 
 
 T difference in degrees F. of the initial and final tem- 
 
 perature of the circulating water, 
 (7 = cubic feet of gas used. 
 
 For oils a special burner is used. 
 
 Example. The weight of the circulating water used was seven 
 pounds per minute, the rise in temperature of the circulating w r ater 
 was 30 F., and 0.333 cubic feet of gas was used per minute, then 
 
 calorific value = ~ =630.6 B. T. IT. per cubic foot. 
 
 This result is usually termed the gross calorific value, or the 
 higher heating value. In the calorimeter the water produced in 
 combustion is condensed and carried off as already explained 
 through pipe R. In this operation it gives up its latent heat, and 
 is measured in the cooling water, but in most industrial operations, 
 such as the performance of a gas engine, the water formed during 
 combustion passes off as steam, and carries with it the latent heat 
 of that steam. The lower heating value, or what is sometimes 
 called the available calorific value is equal to the gross value minus 
 the latent heat of steam formed per cubic foot of gas at the tem- 
 perature it leaves the calorimeter. As the temperature of the 
 calorimeter is about 50 F., the total heat per pound is equal to 
 1147- (50-32) =zll29 B. T. U. per pound of water. Hence 
 the lower heating value is obtained by deducting this quantity of 
 heat from the higher heating value for every pound of water pro- 
 duced by the combustion of one pound of gas. 
 17 
 
258 
 
 EXPERIMENTAL ENGINEERING 
 
 w 
 
 o 
 
 S| 
 
 H ^ 
 
 Q o 
 j 
 
 2 
 
 
 rt <^ 
 
 Q fa 
 
 cc 
 PQ 
 O 
 
 02 
 
 9l S 
 ola 
 
 
 2'o o 5o 
 
 ^ D jQ O .O iO 
 
 g-s| . 
 
 Hi 2 
 PJ 
 
 
 o'd do do do 
 
 Oo OO OO OO 
 
 83 fe$ S &g 
 
 j pJBpUBlg 
 
 H* 
 
 10 
 
 S r? 8 S 3 S 
 
 1 1 i i | i 
 
 CQ '^89X JO 
 
 UONJPUOO 
 
 2 
 
 So 8 S 8 
 
 ,_, 
 
 3 
 
 ^- I ,r: ^ 
 
 ja^BAV jo seqoui 
 k SB) jo ojnssajjj 
 
 2 
 
 ? -* II 
 
 i 
 
 ri 1 s 55 
 
 rH 
 
 S S J 8 
 
 iO IO tO US 
 
 O fr* 
 
 5 - 
 
 H _ 
 
 
 rH 
 
 8 
 
 si si 8 a 
 
 ^ 
 ' 
 
 I i o 
 
 * m 
 
 05 
 
 8 S ff 
 3 S S S 
 
 ! * 1 
 
 s ~ 
 
 CO 
 
 S. S-'SJ 
 
 H 
 UJOOH 
 
 - 
 
 S 8 5 
 
 S S S S 
 
 Wno^oueaejgia 
 
 O 
 
 r; r; r; ^ 
 
 *W *no 
 
 
 
 
 *^J *no 
 
 , 
 
 1111 
 
 jnoq aad 
 '^j 'no 'SBJC) jo a^BE 
 
 
 
 S S S S 
 
 rH C-J ei CO 
 
 **Sa,I, JO 9^BQ[ 
 
 <N 
 
 oT o o 
 ^ ^ rt 
 
 ^S9i JO 'OK 
 
 rH 
 
 i 1 frl CO "* 
 
 t 8 
 
THE SELECTION AND TESTING OF FUEL 259 
 
 There is a difference of opinion among scientists as to whether, in 
 a calorimeter, the higher or lower heating value should be taken. 
 In France the higher heating value is used, while in Germany the 
 lower value is always taken, consequently in any reports of tests 
 made, the condition employed should always be stated. 
 
 The accompanying table from an efficiency test of a 125-horse- 
 power gas engine by C. EL Robertson shows the method of tabu- 
 lating and computing results by means of a Junker calorimeter. 
 
 Testing Fuel Oil. 
 
 Navy Specifications for Fuel Oil. (1) Fuel oil shall be a petro- 
 leum oil of best quality, free from grit, acid, fibrous, and other 
 foreign matter likely to clog or injure the burners or valves. 
 
 (2) The unit of quantity to be a gallon of 231 cubic inches at a 
 standard temperature -of 60 F. For every variation of temperature 
 of 10 F. from the standard, one-half of \% shall be added or 
 deducted from the measured or gaged quantity for correction. 
 
 (3) Sulphur in the oil must not exceed three-fourths of 1% by 
 weight. 
 
 (4) Free water and sediment in the oil shown by gasoline test 
 must not exceed 1% by volume. The gasoline test for moisture 
 shall be made as follows : 
 
 (5) Samples taken at random should be thoroughly shaken and 
 mixed; then 50 cc. of sample placed in a 100 cm. graduated glass 
 cylinder. An equal quantity of not less than 68 gasoline should be 
 mixed with the sample in the graduated glass, the combined mix- 
 ture of oil and gasoline should then be thoroughly shaken and 
 allowed to stand not less than six hours, when percentage of water 
 and sediment will be taken by the inspector. 
 
 (6) The oil shall not flash at a temperature less than 200 
 F., the test to be made by the Abel or Pensky-Martin closed-cup 
 method. 
 
 (7) The oil shall have a gravity Baume not greater than 30 at 
 60 F., determined from an average sample. 
 
 (8) The oil shall flow freely and in a continuous stream through 
 a ^-inch circular hole under a 2-foot head at a temperature of 40 
 F. Oil that fails to flow freely and in the required quantities to 
 
260 
 
 EXPERIMENTAL ENGINEERING 
 
THE SELECTION AND TESTING OF FUEL 261 
 
 pump suctions, pass through pumps, piping, and burners at a 
 temperature of 40 F. may be rejected. 
 
 (9) The oil shall have a calorific value not under 144,000 B. T. 
 IT. per gallon, to be taken from an average sample or samples of 
 the product as delivered, or made before delivery. In determining 
 this value, the bomb calorimeter shall be used to determine the 
 B. T. U. per pound, from which the B. T. U. per gallon shall be 
 calculated by using 8.331 pounds of distilled water per gallon at 
 60 F. and the specific gravity of the oil as determined by the 
 Baume test at same temperature. Should the conditions be such 
 that oil of the required calorific value could not be obtained, oil of 
 a less calorific value may be accepted at a reduction of 1% for each 
 1000 B. T. U. or fraction exceeding one-half thereof; but no oil 
 having less than 135,000 B. T. U. per gallon will be accepted. 
 
 Each vessel that burns oil and all oil supply stations are pro- 
 vided with a portable test outfit as illustrated in Fig. 133. This 
 has instruments for determining the flash point, percentage of 
 water and sediment, and specific gravity. Samples are sent to a 
 laboratory for the determination of the percentage of sulphur and 
 the heating value. 
 
 The following instructions cover the use of the standard test 
 outfit : 
 
 Instructions for Testing Fuel Oil for IT. S. Navy. 
 
 The Pensky-Martens Apparatus for Flash Point: Description. 
 
 Referring to Fig. 134. E is the oil container, which is placed in a 
 metal heating vessel H, provided with a mantle L in order to pro- 
 tect H from loss of heat by radiation. The oil cup E is closed by a 
 tightly fitting lid (shown in plan 2). Through the center of the 
 lid passes a shaft carrying the stirring arrangement, which is 
 worked through a flexible connection by means of the handle J. 
 In another opening of the cover is fixed a thermometer. The lid 
 is perforated with several orifices, which are left open or covered, 
 as the case may be, by a sliding cover. This can be rotated by turn- 
 ing the vertical spindle with milled head G. By turning G, an 
 opening of the slide can be made to coincide with an orifice in the 
 cover, and simultaneously a jet of flame P, from a very small spirit 
 
262 
 
 EXPERIMENTAL ENGINEERING 
 
 lamp, is tilted on the surface of the oil. This contrivance is shown 
 on a larger scale in plan 2. 
 
 Operation. All water which is contained in the oil must be re- 
 moved before testing for flash point by filtering it through one of 
 the small felt niters and funnels contained in the outfit. When 
 
 FIG. 134. 
 
 the sample is prepared for test, the oil cup is filled up to the mark, 
 the cover is fixed and the oil heated rapidly until its temperature 
 reaches a point about 50 F. below the expected flash point. The 
 wire gage screen, shown in Fig. 134, is then placed in position and 
 the rate of rise in temperature is thus reduced to about 5 a minute. 
 Handle J is turned slowly and continuously for stirring. From 
 
THE SELECTION AND TESTING OF FUEL 263 
 
 time to time the milled head G is turned,, opening the shutter at 
 top of the cup and tilting into it the flame P. This is done at in- 
 tervals of 5 F. rise in temperature until near the probable flash 
 point when the intervals are made 2 F. When the flash point is 
 reached there will be a slight explosion when the flame is tilted 
 down. 
 
 A sample can only be used for one test,, since the more volatile 
 products are driven off and subsequent tests would show a higher 
 flash point. 
 
 The Fire Test. This is the temperature at which the oil will 
 give off vapors which when ignited will burn continuously. It is 
 maae by continuing to heat the oil after the flash point is estab- 
 lished. In a closed testing apparatus the cover is removed for this 
 test, the thermometer remaining in place. The flame is extin- 
 guished by putting on the cover. 
 
 Test for Determining Water and Sediment. The outfit contains 
 a graduated glass cylinder, as shown, having a small stem at bottom 
 of 3 cc. capacity. Fifty cc. of the oil under test is placed in the 
 cylinder and an equal quantity of gasoline or kerosene added. 
 The whole is then shaken thoroughly and allowed to stand for at 
 least two hours. All the oil and sediment will then be found to 
 have settled in the narrow stem where it can be measured. If 
 bubbles are found to be adhering to the glass they are removed with 
 a thin wire agitator. 
 
 Each cc. of water and sediment found in the stem will represent 
 2 per cent in the sample tested. 
 
 Specific Gravity. A sample of the oil is placed in the graduated 
 glass jar, shown in Fig. 133, and a hydrometer slowly sunk into it. 
 Care must be taken riot to plunge the hydrometer deeper than it 
 will float as this will make an accurate reading impossible. After 
 reading the hydrometer it is removed and the temperature taken. 
 By means of a correction table the specific gravity is reduced to 
 60 F., which is the standard for comparison. 
 
 Commercially the specific gravity of an oil in the United States 
 is usually given according to the Baume scale, an arbitrary stand- 
 ard whose value at various points is as follows. The weight per 
 gallon is given at 60 F. : 
 
264 EXPERIMENTAL ENGINEERING 
 
 10 Baume = s. g. 1.000 = 8.331 pounds per gal. 
 
 15 " = " .967 = 8.056 " " " 
 
 20 " = " .936 = 7.798 " " " 
 
 25 " = " .907 = 7.556 " " " 
 
 30 " = " .880 = 7.331 " " " 
 
 35 " = " .854 = 7.115 " " " 
 
 40 " = " .830 = 6.915 " " " 
 
 45 " = " .807 = 6.723 " " " 
 
 50 " = " .785 = 6.540 " " " 
 
 55 " = " .765 = 6.373 " " " 
 
 60 '" = " .745 = 6.206 " " " 
 
 Other Oils for Burning. In addition to fuel oil, there is carried 
 on board our naval vessels kerosene or mineral oil for use in lamps 
 and as fuel in some of the internal combustion engines for motor 
 boats, and gasoline which is the fuel most commonly employed in 
 motor-boat engines. Lard oil, after it had been supplanted as a 
 lubricant by compounded oils, was for several years carried for 
 use in oil lamps. This has now been wholly dispensed with. 
 Hand lamps in the engineer department are fitted to burn vaclite, 
 a very heavy paraffine oil which is solid at ordinary temperatures. 
 This material is usually purchased under its proprietary name. 
 Other lamps are now fitted to burn kerosene. 
 
 Specifications for Mineral Oil (Kerosene). (1) Samples of 
 each lot, taken at random, will be tested photometrically after 
 burning one hour in lamps fitted with No. 1 hinge burners, Marcy's 
 patent, the standard employed being a standard Hoefner lamp. 
 After burning five hours longer, the lamps will be again tested to 
 determine any change in the intensity of the light. The flame 
 must be of at least 6 candlepower and must show no material change 
 in intensity during the five-hour interval. 
 
 (2) The samples must show a flash test of not less than 115 F. 
 and a fire test of not less than 140 F. The flash and fire tests are 
 to be conducted in a closed tester of the " Tagliabue " type. 
 
 (3) The oil will be tested for the presence of a free acid. Litmus 
 paper immersed in the oil for five hours must remain unchanged. 
 
 (4) The specific gravity must not be greater than 0.793 at a 
 temperature of 60 F. ; to be purchased and inspected by weight. 
 
 (5) The oil must burn steadily and clearly, in a suitable lamp, 
 
THE SELECTION AND TESTING OF FUEL 265 
 
 without smoking and with a minimum incrustation of the wick, 
 for a period of at least seventy-two hours. 
 
 The Tagliabue apparatus for taking the flash and fire test under 
 these specifications is similar to the New York State Board of 
 Health Tester shown in Fig. 121, except that the cup has a metal 
 cover with a hole to which a match is applied in making a test. 
 A slide over this opening is kept closed except when applying the 
 match. 
 
 Specifications for Gasoline for Use in II. S. Navy Motor 
 Launches. (1) To be a high grade, refined, gasoline, free from all 
 impurities. 
 
 (2) Inspection. Before acceptance the gasoline will be in- 
 spected. Samples of each lot will be taken at random; these 
 samples will be well mixed in a clean, closed vessel, and a sample 
 for test taken from this mixture. 
 
 (3) Test. 100 cc. will be taken as a test sample. This amount 
 will be distilled in an Engler apparatus at a rate of not less than 
 10 cc. per minute. 
 
 (a) Boiling point must not be above 135 F. 
 
 (b) Not less than 10% shall distill over below 150 F. 
 
 (c) At least 50% of the sample must distill over below 200 F. 
 
 (d) One hundred per cent must distill over below 310 F. 
 
 (e) Not less than 96% of the liquid will be recovered from the 
 distillation. 
 
 (4) Five cubic centimeters of the sample when poured over a 
 sheet of white paper shall evaporate completely without leaving any 
 stain. 
 
 (5) Apparatus. The apparatus used for distillation and 
 method of conducting the test shall be as follows : The apparatus 
 shall consist of a 4-ounce Engler flask with outlet high on neck. 
 The top of the thermometer bulb shall be opposite the bottom of the 
 outlet tube. The condenser shall be a standard 20-inch Liebig 
 type of condenser. The boiling point will be the temperature 
 shown by the thermometer when the first drop of the condensed 
 liquid falls from the end of the condenser into the receiving flask. 
 The distillation shall be pushed to completion, at which time the 
 bottom of the flask will be dry. The end point at this time will be 
 indicated by a small flask or puff of smoke. 
 
CHAPTEE XI. 
 FLUE GAS ANALYSIS. 
 
 When coal is burned in the furnace of a boiler, the products of 
 combustion are, for the most part, carbon dioxide (C0 2 ), carbon 
 monoxide (CO), free oxygen, and nitrogen. A small percentage of 
 steam (H 2 0) is also present, resulting from the combustion of such 
 portion of hydrogen as is in the coal. If it were possible to supply 
 air in such quantity that the carbon in the coal would be completely 
 burned to C0 2 , and no more, and, furthermore, if it were possible 
 to completely utilize all this air, combustion of maximum efficiency 
 would be obtained. The C0 2 in the resulting products of combus- 
 tion would then be about 21% and there would be no free oxygen 
 or carbon monoxide present. In practice, in order to obtain the 
 best possible results, a much larger quantity of air must be supplied, 
 and as this quantity is increased the percentage of C0 2 given off is 
 reduced. The efficiency of combustion is correspondingly reduced 
 on account of the great quantity of heat lost with the additional 
 volume of gases through the chimney. If, on the other hand, an 
 insufficient quantity of air should be supplied, the percentage of CO 
 will be increased and the percentage of C0 2 correspondingly re- 
 duced, showing a loss of efficiency from incomplete combustion. 
 
 The percentage of C0 2 present in the flue gases is thus seen 
 to measure approximately the efficiency of combustion. Where 
 21% CO 2 represents perfect efficiency, 15% C0 2 represents a loss 
 of 12%' in efficiency of combustion. Fifteen per cent is about the 
 maximum proportion of C0 2 that it is possible to obtain, the usual 
 figure running down to 7% or 8%, representing a loss of 20% to 
 21% in efficiency. It becomes evident then that an apparatus for 
 measuring the percentage of C0 2 in the products of combustion, 
 affords a valuable means of checking losses from this cause. 
 
 The Orsatt-Muencke Apparatus. This is shown in Fig. 135 in 
 which B is a measuring tube, surrounded by a water jacket and 
 connected to a levelling bottle K. The upper end of B is connected 
 
FLUE GAS ANALYSIS 
 
 267 
 
 by a pipe to a nozzle F from which a hose connection leads to the 
 sampling tube for obtaining the gas that is to be tested. Three 
 U-shaped reagent bottles A, A', and A" are each connected at one 
 end by a short rubber tube with stop cock to the pipe connecting 
 B with F. The other end is open to the atmosphere. 
 
 FIG. 135. 
 
 The reagent bottles are about half filled with reagents as follows : 
 
 (1) Bottle A with potassic hydrate to absorb C0 2 . This is a 
 solution in distilled water of from 3 to 5% of white caustic potash, 
 which comes in sticks. 
 
 (2) Bottle A! with potassic pyrogallate to absorb the free 
 oxygen. This reagent is prepared by mixing a strong solution of 
 
268 EXPERIMENTAL ENGINEERING 
 
 potassic hydrate with 3% solution of pyrogallic acid. It will take 
 about five minutes to absorb the oxygen. 
 
 (3) Bottle A" with cuprous chloride solution in concentrated 
 hydrochloric acid to absorb the CO. This solution keeps best in a 
 bottle containing a few pieces of clean copper wire. The absorp- 
 tion of CO takes longer than that of 0. 
 
 Operation. The cocks on the reagent bottles having been closed 
 and the cock on F opened, the bottle K is partially filled with 
 water, the water being allowed to run in and fill the measuring 
 tube B. By lowering K, gas is drawn in through F filling B. 
 Eaising and lowering K several times the apparatus is cleared of 
 air. Finally 100 cc. of gas is drawn into B and the cock on F 
 closed. Then open the cock on A and allow the reagent in it to 
 absorb the C0 2 by raising and lowering K alternately several times. 
 The last time the reagent must be allowed to fill the leg next to 
 the measuring tube completely. In order to be sure that the 
 absorption of C0 2 is complete, the test should be repeated until the 
 last readings of the measuring tube agree within 0.1%. The dif- 
 ference between the last reading and the 100 cc. originally drawn 
 into B gives the volume of C0 2 absorbed which represents the 
 percentage of C0 2 in the gas. 
 
 Similarly using reagent A' the is absorbed, after which using 
 A" the CO is eliminated. The volume in each case represents the 
 percentage present in the original 100 cc. For these gases the last 
 readings of the measuring tube must agree exactly. 
 
 The remainder, after the C0 2 , and CO are absorbed, is classed 
 as nitrogen. When transferring the gas during these operations 
 care must be taken not to allow any of the reagents to get into 
 the measuring tube, as the water in the bottle must then be changed. 
 A small quantity of water running into the reagent bottles during 
 a transfer will do no harm. 
 
 A complete analysis will require about twenty minutes. Care 
 must be taken to have the water in levelling bottle and reagents 
 at the approximate temperature of the room, so that while the 
 analysis is in progress there will be no change in the temperature 
 of the gas and the volumes absorbed will correctly represent the 
 percentages of each constituent by volume. 
 
FLUE GAS ANALYSIS 269 
 
 Calculations from Results of Analysis. The weight of dry gases 
 per pound of fuel as fired, is calculated from the analyses of the 
 gases and the fuel and from that the number of pounds of dry air 
 passing through the furnace per pound of fuel is found. An ap- 
 proximate heat balance can then be compiled, showing the losses in 
 the heating value of the fuel due to the several causes. The heat 
 units utilized are obtained from the results of the evaporative test 
 of the boiler. In order to make such results comparative, the com- 
 putations are based on the pound of combustible. 
 
 Example. The gas analysis shows C0 2 , 12.7; 0, 5.7; CO, 0.5, 
 and 1\, 81.1 by difference. The fuel analysis shows C, 83.5; H, 
 4.8; 0, 3.2; N", 1.2; S, 0.5; moisture, 1.5; and ash 5.3. The 
 calorific value of one pound of dry coal was found to be 14,580 
 B. T. U. by coal calorimeter. 
 
 The analysis of the coal, referred to combustible, i. e. f coal less 
 the ash, would then be, in per cent, 88.2 C (83.5x100-^-94.7), 
 5.1 H, 3.4 0, 1.2 N", 0.5 S, and 1.6 moisture. The calorific value 
 of one pound of dry combustible was found to be 15,640 B. T. U. 
 
 By Avogadro's hypothesis, the weight of a gaseous compound is 
 equal to its molecular weight, referred to hydrogen as unity. The 
 chemical equivalent by volume of CO and C0 2 is 2, and of and N 
 is 1, referred to hydrogen as unity, or, the molecular weights of 
 H, and N are twice that of their atomic weights. The weight 
 of dry gases will, therefore, be the percentage of each gas found in 
 the analysis multiplied by its molecular weight, or 
 Poundsof drygas= %C0 2 x 44+ %0 2 X 32 + %CO X 28 + %N 2 X 28. 
 
 The pounds of dry gas, per pound of carbon, will then be this 
 amount divided by the product of the atomic weight of carbon and 
 the sum of the percentages of the carbon bearing gases. This is 
 deduced as follows : 
 
 Weight of gases containing carbon = C0 2 X 44 + CO x28. 
 
 Since -fa of the C0 2 and -f- of the CO is carbon, therefore, 
 3xC0 2 x44 3xCOx28 
 
 Letting the symbols represent the percentages, by volume, 
 Pounds of dry gas per pound of carbon burned 
 
 12(C0 2 + CO) 3(C0 2 + CO) 
 
270 EXPEEIMENTAL ENGINEERING 
 
 Substituting the percentage values from the gas analysis given 
 above, we get, 
 
 Dry gas per pound of carbon 
 
 The number of pounds of dry gas per pound of combustible 
 = pounds of gas per pound of carbon multiplied by the percentage 
 of carbon (in decimals) in the combustible. 
 
 Or, in this case, 
 
 Dry gas per pound of combustible = 19. lx. 882 = 16. 85 pounds. 
 
 The number of pounds of dry gas per pound of coal as fired 
 = pounds of gas per pound of carbon multiplied by percentage of 
 carbon (in decimals) in the coal. 
 
 The gas analysis accounts for the carbon only, and we must, 
 therefore, add the H 2 in the gases, formed by evaporating the 
 moisture and by burning the H in the coal. The latter we find 
 on the same principle as above, and the former, from the coal 
 analysis. Hence, 
 
 48x9 
 
 H 2 from hydrogen in coal = ' =.52 pound. 
 
 oo.5 
 
 H 2 from moisture in coal = . 053 x. 835 = .04 pound. 
 
 .56 pound. 
 
 Or, the total weight of gases per pound of carbon=19.1 + .56 
 = 19.66 pounds; per pound of combustible = 16.85 + .56 = 17.41 
 pounds ; and per pound of coal as fired = 19.66 X .835 = 16.42 pounds. 
 
 The quantity of air which passed through the furnace for each 
 pound of combustible is, therefore, 17.4 1 = 16.4 pounds. The 
 air per pound of coal is only approximately one pound less than 
 the gas per pound of coal. 
 
 Loss by Heat of Gases. Suppose that the temperature of the 
 gases in the uptake or smoke pipe was 590, and that of the ex- 
 ternal air, 75 F. For all practical purposes, and in view of the 
 approximate results which can be obtained by the present state 
 of the art of gas analysis, the average specific heat of the dry gases 
 may be taken as .24, and of the dry gases including the H 2 0, as 
 .246. 
 
FLUE GAS ANALYSIS 271 
 
 The rise in temperature is 515 F. As there were 17.41 pounds 
 of gases for each pound of combustible, the sensible heat loss was, 
 
 17.41 X. 246x515 = 2204 heat units per pound of combustible (1). 
 
 Loss Due to Latent Heat in H 2 0. The loss of sensible heat in 
 the steam gas has been accounted for in the above calculation, but, 
 in addition, there is the loss of heat rendered latent by changing 
 the H 2 0, formed from the H and H 2 in the coal, from water into 
 steam. The latent heat of one pound of steam under atmospheric 
 pressure is 965.8. It was found above that .56 pound of H 2 gas 
 was evolved from the coal. The loss is, therefore, 
 
 .06x965.8 = 541 heat units per pound of combustible (2). 
 
 Loss Due to Incomplete Combustion. As we have found before, 
 the weight of the carbon in the gases is 12(C0 2 -1-CO). The 
 perfect or complete combustion of this total carbon would have 
 given 12(C0 2 + CO) x 14,600 heat unitsa. But the combustion 
 of the carbon, as shown by the gas analysis, was only partially 
 complete, and the heat generated was, therefore, only 12C0 2 X 14,600 
 + 1200x4400 units = 6. The difference between the two will 
 give the loss in heat units due to the incomplete combustion, or, 
 Loss = a b = 1200x10,200 heat units, or in per cent of a, 
 
 1200x10,200x100 00x10,200x100 , . , 
 
 12(C0 2 + CO) -CO. + CO ~ PCT P Und f Carb n - 
 
 And per pound of combustible, 
 
 T _ 00x10,200 %C in combustible 
 C0 2 + C0 X 100 
 
 Substituting values from the above gas and chemical analyses, 
 we get loss due to incomplete combustion, per pound of combustible 
 
 (3). 
 
 Suppose that the results of the evaporative test of the boiler 
 gave 11.6 pounds as the equivalent evaporation from and at 212 F. 
 per pound of combustible. Then, 
 
 11.6x965.8 = 11,203 heat units absorbed by boiler (4). 
 
272 EXPERIMENTAL ENGINEERING 
 
 From this and the losses computed above, we can make up a 
 heat balance which will show the approximate distribution of the 
 heating value of one pound of the combustible. 
 
 Calorific value of the combustible 15,640 heat units. 
 
 Absorbed by the boiler 11,203 
 
 Loss due to sensible heat in waste 
 
 gases 2,206 
 
 Loss due to latent heat in steam 
 
 gas , 541 
 
 Loss due to incomplete combustion. . 341 
 Other losses, due to radiation, etc., 
 
 by difference 1,349 
 
 15,640 
 
 These values are frequently expressed in per cent of the calo- 
 rific value of the combustible. 
 
 Apparatus for Determining C0 2 Alone. A complete analysis of 
 the flue gas affords most valuable information and is made a part 
 of all complete boiler tests. For the daily routine in a boiler plant 
 complete analyses are unnecessary and such work may be restricted 
 to the determination of the percentage of C0 2 alone. This in itself 
 affords a complete index to the character of the firing and discloses 
 any waste due to an excessive or insufficient supply of air. 
 
 The Hays C0 2 Apparatus. This is a modified and simplified 
 form of the apparatus described on page 267. It is used for the 
 determination of C0 2 alone, and is supplied to many of the vessels 
 of the navy. It is shown in Fig. 136 and the method of operation 
 is described by the manufacturer as follows : 
 
 Eeferring to Fig. 136. 
 
 Sample of gas is first measured in the burette A. It is then 
 passed into the pipette B where it comes in contact with an ab- 
 sorbent solution. It is next returned to the burette and remeas- 
 ured. The shrinkage represents the percentage of the gas absorbed. 
 
 The burette A is surrounded by a water jacket Al and suspended 
 on piano-wire springs. A leveling bottle, C, is connected with 
 bottom of burette and is filled with water, brine, or solution of 
 glycerine and water as preferred. Absorbent liquid is introduced 
 
FLUE GAS ANALYSIS 
 
 273 
 
 into B through the funnel E. Pipette holds sufficient potash for 500 
 C0 2 determinations. 
 
 To operate the instrument hang upon any convenient nail at T. 
 Remove stopper from leveling bottle and also stopper 8. Open 
 
 FIG. 136. 
 
 pinch cock K and raise leveling bottle until liquid entirely fills the 
 burette. Return leveling bottle to its compartment; connect hose 
 N with a piece of gas pipe and insert same into breeching or other 
 point from which it is desired to draw the gas to be analyzed. Open 
 pinch cock Q and pump the aspirator bulb. Gas enters the burette 
 and bubbles out through the leveling bottle. Next open pinch cock 
 18 
 
274 EXPERIMENTAL ENGINEERING 
 
 K and slowly raise leveling bottle until the liquid reaches the zero 
 mark on the burette. Gas is passed back and forth between burette 
 and pipette by opening pinch cock L and manipulating the leveling 
 bottle. 
 
 Any absorbable gas can be determined with this instrument, by 
 changing the solutions in the absorption pipette. 
 
 In determining and CO in boiler furnace practice it is usually 
 desired to know the averages of these gases from a large number of 
 samples. The residue after each C0 2 absorption is discharged into 
 the sampling bottle furnished with the instrument. One analysis 
 for and CO accordingly gives the averages for these gases on all 
 the samples submitted to C0 2 absorption. 
 
 An analysis for C0 2 can be made every two minutes with this 
 instrument and the average and CO determined in 10 minutes on 
 any desired number of gas samples. 
 
 The introduction of this apparatus in our naval vessels resulted 
 in an immediate increase in the efficiency of firing with a marked 
 decrease in coal consumption. 
 
 The Sarco Automatic CO, Recorder. Eef erring to Fig. 137, 
 the power required to drive this apparatus is derived from the main 
 flue or chimney in the following manner : A water tank A, of an- 
 nular section is filled with water, forming a seal for the gas tank C. 
 This gas tank is balanced by a weight D, suspended by a cord, pass- 
 ing over the pulley E. F is a tube which passes under A and up 
 through the inner compartment to the nozzle G, which is above the 
 level of the water. This tube is connected to the main flue or 
 chimney at H, through ordinary tubing, and the draught by ex- 
 hausting the gas tank 0, causes the same to sink downward into A. 
 
 This motion continues until C reaches its lowest point, when the 
 stop J strikes a lever, automatically opening the valve /. This 
 admits air to C, which then rises until the stop J 2 strikes the lever, 
 closing valve I, when the whole operation is repeated. This move- 
 ment continues, furnishing the power necessary for the operation 
 of the apparatus, so long as the communication with the flue or 
 chimney, through H, is uninterrupted. A pinch valve L affords a 
 means of regulating the speed of operation. 
 
FLUE GAS ANALYSIS 
 
 275 
 
 FIG. 137. 
 
276 
 
 EXPERIMENTAL ENGINEERING 
 
 The Gas Pumps and Valves. The power thus obtained is utilized 
 to work two plungers (Fig. 138) attached to the side of the motor 
 and connected to the large pulley by wires running over two small 
 pulleys, as shown in Fig. 137. These plungers, or pumps, A^ and 
 A 2 , are themselves small gas tanks, which 
 dip into an oil seal contained in the tanks 
 B! and j? 2 , the suction tubes to A^ and A 2 
 going down through the center. To these 
 tubes are attached two sets of valves C 19 
 C 2 , and D lt D 2 , so constructed as to pre- 
 vent the return of any gases that may 
 enter through overcoming the small re- 
 sistance offered by the glycerine with 
 which they are partly filled. The two 
 plungers move up and down alternately, 
 one sucking gases into the valves while the 
 other is pressing the same out through the 
 outlet E and into the registering cabinet, 
 a diagram of which is shown in Fig. 139. 
 A continuous flow of gases is secured by 
 this arrangement, air bubbles showing at 
 the seal F, in case the flow should in any 
 way be interrupted. To the lower outlet 
 of the pump valves is attached an escape 
 G, through which the surplus flue gases, 
 not required for analysis, pass out into the 
 atmosphere. 
 
 The Analyzing and Recording Appara- 
 tus. Coming from the pumps, the flue 
 gases enter the registering cabinet (Fig. 
 
 139) at H, pass down tube I, and into vessels J" and J 2 . These ves- 
 sels are in communication with a bottle K through the tube L. 
 Bottle K contains a mixture of glycerine and water. It is attached 
 to the weight D (Fig. 137), and is carried up and down regularly 
 with it. 
 
 On the up stroke the liquid in K rises in tube L and seeks its 
 level in vessels J^ and J 2 , into which the flue gas is being pumped. 
 
 Dl 
 
FLUE GAS ANALYSIS 
 
 277 
 
 FIG. 139. 
 
278 EXPERIMENTAL ENGINEERING 
 
 As soon as the liquid covers the end of tube /, where it enters 
 J 19 the flow of gas into J 19 J 2 , is stopped. Part of the contained 
 gases then escape through the inner tube J 3 into the atmosphere 
 and when this outlet is sealed by the liquid rising further, exactly 
 100 cubic centimeters of flue gases are tapped in J ly J 2 . This 
 quantity is gradually forced through the small curved tube M and 
 brought into contact with a solution of caustic potash, with which 
 vessel N is filled to the mark ra s . The pressure of the gas on the 
 surface of the potash displaces same in N, forcing it up into vessel 
 0. The air which is thus displaced in passes under cylinder P, 
 which is suspended by a silk cord and accurately counterbalanced. 
 
 The slight pressure thus created causes lever Q to swing upwards, 
 carrying with it pen R, which is attached to the end of the lever. 
 This pen rests against a circular drum fitted with clockwork and 
 makes a continuous record on a chart that is fitted on the drum. 
 This chart is calibrated in terms of percentage of C0 2 , reading 
 from zero at the top to 20% at the bottom. Each chart, corre- 
 sponding to one complete revolution of the drum, gives a record for 
 24 hours. 
 
 The actuating stops, J l and / 2 , shown in Fig. 137, are so ad- 
 justed that the motion of the motor is reversed the moment that 
 the pen has completed its upward stroke. The sealing liquid then 
 recedes again, the potash falls back to its original level mark ra 3 , 
 and the remaining gas mixture is drawn out of vessels N, J 1? and 
 J 2 , and passes into the atmosphere. 
 
 As soon as the level of the sealing liquid has fallen below the gas 
 inlet from tube 7, a constant supply of fresh gas is continually 
 pumped through the instrument, until the returning seal again 
 bottles off a fresh portion for analysis in the same manner as above 
 described. This process can be repeated as rapidly as desired, the 
 result of each analysis being recorded on the chart by a vertical 
 line. The tops of the various lines form a continuous curve, show- 
 ing the percentage of CO, in the flue gases at any time during the 
 24 hours. 
 
 A facsimile of a chart made by this apparatus is shown in Fig. 
 140. 
 
FLUE GAS ANALYSIS 
 
 279 
 
 Application of the Apparatus. The recorder should be located 
 in the fire-room in such place as will be readily accessible but 
 where it will be free from injury while working the fires. The 
 gases to be analyzed should be taken from the uptake on the boiler 
 side of the damper and conveyed to the recorder through ordinary 
 f-inch gas piping. By running a system of branch pipes, tapping 
 the uptake of each boiler, and connecting through a main pipe to 
 the recorder, separate readings may be taken for all of the boilers 
 at will. A cock or valve should in such case be fitted on each 
 branch. 
 
 ^tTCo a record Average for 24 hrs /<f- % Date ssyz^/ef 
 
 H"TC in n x i BH~ DiniyY^^nYiDiixiLin i~n IBY- 
 
 FIG. 140. 
 
 Filter. In order that the gases may be free from impurities 
 when passing through the apparatus, a filter of special construction 
 is provided with each outfit, and should be inserted in the supply 
 pipe as close to the boiler as practicable. 
 
 The filter is filled with fine wood shavings, intercepted by a layer 
 of sawdust. The lid is provided with a glycerine seal. At the 
 bottom is attached a water separator in which any water that may 
 condense in the pipes will collect and may be drained off from time 
 to time. The separator is filled with glycerine to a given mark in 
 order to exclude the air. 
 
 Draught for Motive Power. This is obtained from the base of 
 smoke pipe, above damper, and is brought through a 1-inch pipe 
 to the recorder, where the connection is made by rubber tubing. 
 About |-inch draught pressure is all that is required for the opera- 
 tion of the apparatus. 
 
 The Uehling C0 2 Recorder. Eef erring back to Fig. 26 if a con- 
 stant vacuum of say 48 inches of water be maintained in chamber 
 C" and the two apertures A and B are of the same size and are 
 
280 EXPERIMENTAL ENGINEERING 
 
 maintained at the same temperature, the manometer p will show 
 about one half the vacuum maintained in C f due to the fact that 
 the apertures oppose equal resistance to the passage of the gas. 
 
 This relation will be maintained so long as the same volume of 
 gas flows through B as enters at A. If, however, a constituent of 
 the gas (C0 2 ) be continuously taken away or absorbed in passing 
 through chamber C the vacuum therein will be correspondingly in- 
 creased. This increase of vacuum in C, shown by the manometer p, 
 therefore correctly, indicates the volume of the gas to be deter- 
 mined. 
 
 In the complete instrument the suction is produced by a steam 
 aspirator and is automatically regulated to 48 inches of water. 
 Both apertures are kept at a constant temperature by the exhaust 
 steam from the aspirator and are protected by efficient niters. 
 The C0 2 is absorbed between the two apertures of a dilute solu- 
 tion of caustic soda and the space between these apertures is con- 
 nected with a manometer which is provided with a scale calibrated 
 in per cent of C0 2 . This manometer may be placed on or near 
 the boiler front showing continuously the percentage of C0 2 for 
 the information of the fire room force, or there may be a recording 
 gage similar to those shown in Chapter III, adjusted to give a 
 continuous record. 
 
 TIME FIRING DEVICES. 
 
 The object of these devices is to automatically give audible and 
 visual signals in each fire room at regular time intervals. The 
 audible signal calls attention to the visual signal, which indicates 
 the furnace that is to be fired. The time intervals can be varied at 
 will from 20 seconds to 9 minutes, thus ensuring regular firing 
 and working of the fires at all speeds from the lowest to the 
 highest. 
 
 Each ship is provided with a transmitter, located in the work- 
 ing engine room, and an indicator in each fire room. Fig. 141 
 shows, in elementary form, a typical arrangement for a battleship. 
 The transmitter contains the timer and suitable mechanism for set- 
 ting it for the desired time intervals, the mechanism for closing 
 the indicator circuits at these time intervals, and cut out switches 
 and fuses, so that any indicator may be readily cut out of circuit 
 
TIME FIRING DEVICES 
 
 when desired. The indicators are constructed to suit the number 
 of furnaces in the fire rooms in which they are installed. The 
 devices are connected to the ship's lighting circuits, operating at 
 125 volts on later vessels and 80 volts on those of earlier date. 
 
 These principles govern the design of all devices of this kind 
 that have been introduced in the service. 
 
 A comparison of the different devices is afforded by a brief 
 description of the mechanism for controlling the time interval in 
 transmitters, the method of operation of indicators, and an ele- 
 mentary wiring diagram of each type. 
 
 FIG. 141. 
 
 The Corey Time Firing Device. Old Type. This has been sup- 
 plied to a number of our ships. A diagram is shown in Fig. 142, 
 from which the operation will be understood. A small constant- 
 speed motor is geared to the wheel A and this, through the small 
 friction wheel B, operates wheel C. A and C are flat discs, while 
 B is carried in a frame by the screw shaft D, and its position is 
 adjusted by the thumb screw E. By changing the position of B 
 we change the relative speed of A and C. An index is provided to 
 indicate the firing interval, as regulated through B by the speed of 
 revolution of C. 
 
 Two contactors, F and G, are seen, which are closed and opened 
 alternately six times for each revolution of C. These make con- 
 nections to the coils of a pair of electromagnets, shown near the 
 bottom of the diagram. These in turn make and break the cir- 
 cuits for operating the starboard and port indicators respectively, 
 in the fire rooms. 
 
2S2 
 
 EXPERIMENTAL ENGINEERING 
 
 The receiver consists essentially of an electromagnet, the arma- 
 ture of which carries an escapement that operates a shaft, and this 
 in turn carries a dial with the signal numbers on it. It will be 
 
 noted that there are twelve displays for a complete revolution of 
 the shaft, the numbers from 1 to 6 being repeated. This gives a 
 better arrangement of the mechanism. 
 
TIME FIRING DEVICES 
 
 283 
 
 Three wires only are necesary in the conduits leading to each 
 fire room. One of these is to the electromagnet, one to the lamp, 
 and the third is a common return. 
 
 E 
 
 E 
 
 FIG. 143. 
 
 FIG. 144. 
 
 Corey Time Firing Device. New Type. Figs. 143 and 144 show 
 external views and the wiring diagram respectively of the trans- 
 mitter and indicator. The transmitter contains a shunt-wound 
 
284 EXPERIMENTAL ENGINEERING 
 
 motor, A, running at a constant speed of 1600 revolutions per 
 minute. Through suitable worm gears for reducing the speed, 
 this motor drives a gear wheel R, at a speed of one-fifth of a revo- 
 lution per minute. An arm D, carrying an insulated pin E at its 
 end, is rigidly secured to this gear wheel. On the shaft C is an 
 ami, carrying contact springs H, which can be set at any desired 
 angle by means of a pointer and handle (V, Fig. 143), located 
 on the front of the transmitter case, and secured to the shaft C. 
 Current is supplied to these contact springs through collector rings 
 and brushes F and G. 
 
 There are two electromagnetic relays L and M , which are alter- 
 nately energized when the pin E, carried by the revolving arm D , 
 makes contact between, first the movable contact springs H, and 
 afterwards the fixed contact springs K. This causes the rocker 
 armature N to alternately close the two sets of knife-blade switches 
 and P, reversing the direction of rotation of the motor, and 
 therefore of the arm D. These sets of switches and P also trans- 
 mit indications alternately to the starboard and port indicators. 
 
 Since the speed of revolution of the arm D, carrying pin E, is 
 constant, the desired time intervals are obtained by varying the 
 angle between the fixed contact springs K and the movable springs 
 H. 
 
 Two lamps, R and S, are provided in the transmitter, one show- 
 ing through a green glass when signals are sent to starboard fire 
 rooms and the other through a red glass when the port indicators 
 are operated. 
 
 Fused cut-out switches T are installed in the bottom of the trans- 
 mitter case. 
 
 The indicators with this device are similar to those used with 
 the earlier type of Corey device, except that each indicator requires 
 only two wires. 
 
 Kilroy Time Firing Device. Figs. 145 and 146 show the ex- 
 ternal appearance and wiring diagram, respectively, of the trans- 
 mitter and indicator of this device. 
 
 The transmitter contains three electromagnets, A, B and C, two 
 of which, B and C f are so arranged that they will alternately attract 
 a pivoted armature D, which operates an automatic switch G, thus 
 
TIME FIRING DEVICES 
 
 285 
 
 closing the current alternately through the winding of the magnets. 
 A condenser c is connected across the line to reduce the spark. 
 The armature D is geared to a thin copper disc (not shown) which 
 
 r REGULATOR ^1 
 
 I FOd KIUROY'S PATENT 
 I STOKING INDICATORS 
 I NO. 1=3 
 I EVEKHEDSVIOiOl.ES'-'"' | 
 t^ LONDON. ^> 
 
 o7 
 
 6 
 
 
 
 FIG. 145. 
 
 FIG. 146. 
 
 revolves in the field of the upper electromagnet A and serves as a 
 brake or retarder for the armature D. The different time intervals 
 are controlled by regulating the strength of the field of magnet A. 
 This is done by means of the rheostat R, the regulation of which 
 
286 EXPERIMENTAL ENGINEERING 
 
 is effected by means of the pointer V, located on the front cover 
 of the transmitter. There is a dial with suitable lettering for 
 setting the pointer for the desired time interval. 
 
 The magnet A is wired in parallel with magnets B and C, so that 
 the strength of the field of magnet A, in which the copper brake 
 disc revolves, and the pull of magnets B and C on armature D may 
 keep the same relation to each other. By this arrangement the 
 accuracy of the time intervals will not be affected by changes in 
 strength of the supply current. 
 
 On one side of the transmitter are located fuse blocks F and cut- 
 out switches S, through which current is supplied to the several in- 
 dicators. It will be noted that the automatic switch G is so 
 arranged that it permits current to pass through only half the 
 indicators at one time. Usually signals are sent alternately to the 
 starboard and port fire rooms. 
 
 Each indicator contains an electromagnet M , with two armatures, 
 N and 0, one of which strikes a gong P, and the other revolves a 
 disc U, by means of a ratchet and pawl mechanism L. The dial U 
 has numerals spaced at regular intervals around its outer edge, 
 which become visible when opposite the opening T in the front 
 cover of the indicator. 
 
 Sub-Target Gun Company's Time Firing Device. Figs. 147 and 
 148 show the external appearance and wiring diagram, respectively, 
 of the transmitter and indicator for this device. 
 
 The timer, A, consists of an electromagnet of special construc- 
 tion, with an oscillating armature (not shown) which serves to 
 alternately open and close the current through the magnet coils. 
 A spring attached to this armature pulls it away from the magnet 
 poles when the circuit is open. A small fan is geared to the arma- 
 ture and spring, the motion of which is retarded by the air resist- 
 ance, causing it to act as a brake to regulate the time required for 
 the armature to make one complete oscillation or beat. The device 
 is adjusted so that this time is three seconds. 
 
 At each beat of the armature, the switch B is first closed, then 
 opened, energizing a second magnet C. This transmits current to 
 one of the magnets D or E, geared to a spindle F, so as to cause it 
 to revolve alternately in opposite directions. A contactor block G 
 
TIME FIRING DEVICES 
 
 287 
 
 is threaded on this spindle, so that it will travel back and forth 
 between a pair of fixed contact springs H and a pair of adjustable 
 contact springs I. When the spindle is being driven by one of the 
 
 I 
 
 O 
 
 FIG. 147. 
 
 FIG. 148. 
 
 electromagnets,, D for example, the contactor will move toward the 
 fixed contact H a certain distance each time the magnet D is ener- 
 gized, i. e., every three seconds. When contact is made at H the 
 magnet C is energized, thus reversing the switch K, throwing mag- 
 
288 EXPERIMENTAL ENGINEERING 
 
 net D out and magnet E in circuit. This causes the contactor Q 
 to move at regular intervals in the other direction until contact is 
 made at I, when magnet E is cut out of circuit and D thrown in 
 again. 
 
 Since the contactor moves an equal distance every three 
 seconds, it is noted that the desired time intervals may be ob- 
 tained by regulating the distance between contacts H and I. This 
 is done by means of a small hand wheel M (Fig. 147) on the front 
 cover of the transmitter, which is geared to the movable block 
 holding the contact springs I. A dial and pointer N, on the front 
 of the transmitter, shows the time interval at which it is set. A 
 lamp L shows through an opening in the front of the transmitter 
 each time signals are sent to the fire rooms. 
 
 The magnet C, in addition to directing current to magnets D 
 and E, also operates switch P, closing the circuit through the indi- 
 cators each time contact is made at H or I. 
 
 The indicator contains a magnet Q, the armature of which re- 
 volves a dial R through a ratchet and pawl mechanism, not shown. 
 The dial carries the perforated numerals 1, 2, 3, etc., denoting the 
 furnaces in the fire rooms. The magnet Q also lights the lamps T, 
 located behind the dial R and illuminates the numerals ; also closes 
 the circuit through a magnet U and sounds a gong V each time a 
 signal is received. 
 
 General Electric Company's Time Firing Device. New Type. 
 The principle parts of the transmitter in this device are shown 
 diagrammatically in Fig. 149. 
 
 A shunt- wound motor, A, runs at a constant speed and drives, 
 through suitable reducing gears, a cam shaft, B, at a speed of 2 
 revolutions per minute. This shaft carries a cam C, which rotates 
 a ratchet wheel E, by means of a lever D and pawl F. Another 
 pawl, N } holds the ratchet wheel when pawl F is being lifted. 
 
 The ratchet wheel E carries a pin H which, at a certain position 
 of its revolution, strikes the arm of a retaining pawl I and disen- 
 gages it, releasing a rock shaft K and throwing contact arm L 
 against cam M, thus closing the circuit to the indicators. At the 
 same time arm P strikes arm Q and disengages pawls F and N. 
 Eatchet wheel E is then drawn back by spring Q until pin H 
 brings up against an adjustable stop R. 
 
TIME FIRING DEVICES 
 
 289 
 
 Just before cam C again begins to raise the lever D, the con- 
 tact arm L snaps off cam M, breaking the circuit to the indicators. 
 Then, as the lever D is lifted, the arm Q raises rocker arm P, thus 
 resetting the other end of same under the retaining pawl I. This 
 holds the contact arm L clear of cam M until pin II again strikes 
 pawl I. 
 
 It will thus be seen that since cam shaft B revolves at a constant 
 
 FIG. 149. 
 
 speed of 2 revolutions per minute, the ratchet wheel E will be re- 
 volved through an equal angle every thirty seconds, and the desired 
 time intervals are obtained by regulating the angular distance be- 
 tween retaining pawl I and adjustable stop R. This is done by 
 means of a dial and pointer T located on the outside of the trans- 
 mitter case. 
 
 It will be noted that the movement of rock shaft K causes a 
 pawl V to rotate a ratchet wheel W, which carries a number signal 
 
 J- y o 
 
 cylinder 0. This cylinder supports a number of contact blocks 
 19 
 
290 
 
 EXPERIMENTAL ENGINEERING 
 
 which, by making contact with the several springs Y, cause to be 
 indicated the number of the furnace to be fired. A slotted wheel 
 X with roller and arm serves to hold the cylinder in position. It 
 will be noted that the movement of the cylinder always occurs 
 when the circuit is open, so that the circuit is made and broken 
 only by the arm L on cam M . 
 
 FIG. 150. 
 
 FIG. 151. 
 
 Switches and fuses are located in the base of the transmitter. 
 
 The Indicator. This consists of a number of 5-candle-power 
 lamps and an electromagnet in a water-tight case. A glass-covered 
 opening in front of each lamp is covered by a piece of metal having 
 a number cut in silhouette, so that each lamp, when lighted, will 
 indicate the number of a furnace to be fired. The electromagnet 
 operates a gong located on the outside of the case, and attracts 
 attention to each change of signal. 
 
TIME FIRING DEVICES 291 
 
 This device has the advantage of eliminating all coils, with few 
 make and break contacts in the transmitter, and also the elimina- 
 tion of moving parts in the indicator. The disadvantages are the 
 large number of working mechanical parts in the transmitter, and 
 if there should be a failure of any of the lamps of the indicator 
 there would be a corresponding failure in the display of the num- 
 ber corresponding to such lamp. 
 
 An elementary wiring diagram of a transmitter and an indicator 
 are shown in Fig. 150. Outside views of these parts are shown in 
 Fig. 151. 
 
 The Lobitz Time Firing Device. This device was designed by 
 Machinist Henry Lobitz, U. S. N"., and the initial installation was 
 made and installed by the ship's force on board the U. S. S. 
 Minnesota. 
 
 Eef erring to Fig. 152, there is a small leather-bound friction 
 wheel, 2-| inches diameter, placed on a shaft driven from the main- 
 engine revolution-indicator gear. The wheel is driven through 
 worm gearing proportioned to give one revolution of the wheel for 
 two hundred revolutions of the main engine. 
 
 The friction wheel makes contact with the face of and drives 
 an 18-inch disc as shown. The friction wheel is loosely mounted 
 on its shaft, with a sliding feather for driving it and with an ad- 
 justing screw for varying its distance from the center of the disc. 
 The pressure between the face of disc and circumference of fric- 
 tion wheel can be varied by an adjusting nut at the center of the 
 disc. This nut has a ball-bearing under it to lessen friction. 
 
 Gong Signals. The circumference of the disc carries four lugs 
 spaced 90 apart. These lugs carry one, two, three and four teeth, 
 respectively. As the disc revolves, each tooth in turn engages the 
 contact maker, shown at the right in the figure, and closes a circuit 
 leading to an electric gong in each fire room. The gongs are thus 
 made to strike one, two, three and four times at regular intervals, 
 the length of the intervals depending upon, first, the speed of the 
 main engines, and, second, on the distance of the friction wheel 
 from the center of the disc. 
 
 The index plate at the right indicates the proper setting of the 
 friction wheel for any firing interval that may be desired. 
 
292 
 
 EXPERIMENTAL ENGINEERING 
 
TIME FIRING DEVICES 293 
 
 Visual Signals. On the shaft of the rotating disc there is a 6- 
 inch hard-rubber disc with, first, a copper ring, and, second, four 
 segments of a copper ring at equal intervals around its circum- 
 ference. 
 
 On the friction disc there is a carbon holder which rides over 
 the copper ring and the segment. Each segment is wired to a lamp 
 in a light box placed in each fire room, the light box having the 
 figures 1, 2, 3, 4 on it. When the gong strikes one, the figure one 
 is lighted up in the light box 2 gongs : No. 2, lights up ; and so on. 
 
QUESTIONS ON THE TEXT. 
 
 CHAPTER I. 
 
 1. What is experimental engineering? What tests are made in con- 
 nection with marine engineering work and for what purposes? Why 
 are numerous tests made of different samples from the same lot of 
 material? Does this lead to more accurate results, and if so, why? 
 
 2. Explain the value of accuracy in engineering calculations and the 
 necessity for discrimination in the effort to obtain accurate results. 
 
 3. What are direct and indirect measurements? Give examples. What 
 determines the degree of accuracy for which a quantity is measured? 
 Explain the limits of error in observations. 
 
 4. Explain the use of a series of observations in determining the most 
 probable value: (1) Where the observations have been made with care, 
 (2) where they are not all equally reliable. 
 
 5. Discuss sources of error, the different classes of errors and how to 
 avoid them. Distinguish between mistakes and errors of observation. 
 
 6. Explain the term significant figures and how to determine the 
 degree of accuracy of the various places of significant figures in an 
 observed quantity when its average deviation is known. 
 
 7. Give the rules for significant figures and illustrate by examples. 
 
 8. What procedure should be followed in computing the results of a 
 test? What details should be included in the report? 
 
 CHAPTER II. 
 
 9. What is a logarithmic scale, how constructed and what do the 
 figures on it represent? Explain the use of the logarithmic scale: (1) 
 in multiplication, (2) in division. If the pointer should fall off the 
 scale what is done? 
 
 10. Explain the construction of Sexton's omnimetre. How determine 
 the logarithm of a number by the use of this instrument? 
 
 11. Briefly describe the functions of the several circles of the omni- 
 metre and explain in detail how you would proceed to find the value of 
 a formula expressed in the form: 
 
 a x sin ft x C* X sec & X & 
 f x tan ff x vers sinTi 
 
 12. Explain the use of the omnimeter in solving problems of the form 
 x = Va a -|- ft 2 and # Vo 2 b 2 . 
 
 13. Explain the construction of Fuller's calculating instrument and of 
 Sperry's Pocket Calculator. For what classes of work are each of these 
 adapted? 
 
 14. Explain the use of the straight slide rule. How does it differ from 
 the single logarithmic scale and what is the advantage gained thereby? 
 
 15. Explain the construction of Thacher's calculating instrument. 
 What is the essential difference between this and the Fuller instrument 
 and wherein lie the advantages of each? 
 
QUESTIONS ON THE TEXT 295 
 
 16. Explain the construction and use of the duplex slide rule. 
 
 17. Make a line sketch of the Amsler polar planimeter. Letter and 
 name the essential parts. What is the zero circle? How is its value 
 obtained and how is it applied? What feature makes this instrument 
 adaptable for use in computing indicator ends? 
 
 18. Demonstrate the accuracy of the Amsler polar planimeter in meas- 
 uring areas outside the zero circle and show that the demonstration is 
 general. 
 
 19. Make a plan view of Coffin's averaging instrument, showing an 
 indicator card in position; explain how to adjust the card and the in- 
 strument prior to measuring the card and describe the method employed 
 in measuring (1) the area of the card, (2) its mean ordinate. 
 
 20. Demonstrate the theory of the Coffin averaging instrument and 
 show how the instrument is a special form of the Amsler planimeter. 
 
 CHAPTER III. 
 
 21. Sketch the Pratt and Whitney measuring machine. Letter and 
 name its essential parts. Describe in detail the method of adjusting the 
 machine to zero and measuring an object in the machine. 
 
 22. Show by line sketches the detail construction of a Bourdon gage 
 and a diaphragm gage. 
 
 23. Distinguish, illustrating by sketches, between single and double 
 tube Bourdon gages. Which is preferred and why? 
 
 24. How are the dials of pressure and vacuum gages graduated? What 
 is a compound gage and how is it graduated? 
 
 25. Sketch and explain the essential features of a recording pressure 
 gage. Show by an additional sketch the recording mechanism of an 
 air pressure gage. 
 
 26. Show by sketch the working principle of the Uehling differential 
 recorder. Letter the different parts and explain its operation. 
 
 27. Make a line sketch, showing the essential features of a standard 
 steam and vacuum gage testing apparatus. Explain in detail how to 
 conduct a test with this outfit. 
 
 28. How are gages tested on board ship? What is a manometer and 
 for what purpose is it used? How are manometer readings expressed? 
 
 29. What is the practical high temperature limit of an ordinary mer- 
 curial thermometer and why? What are the different types of pyrom- 
 eters? 
 
 30. Explain the principle on which the gas thermometer is con- 
 structed. Make line sketch, showing the construction of the Industrial 
 Thermograph. Explain its action and state the material used in filling 
 the bulbs for various ranges of temperature. 
 
 31. Make diagrammatic sketch and explain the underlying principle 
 in Uehling's pneumatic pyrometer. 
 
 31. Make diagrammatic sketch of all the parts in Uehling's pneumatic 
 pyrometer and explain its operation. 
 
 32. Make line sketch of Uehling's pneumatic pyrometer assembled 
 complete. Also sectional view of the bulb. Explain how the instrument 
 
296 EXPERIMENTAL ENGINEERING 
 
 is connected up for use and how it may be used for a whole battery of 
 boilers. 
 
 33. How does a mercurial pyrometer differ from a mercurial ther- 
 mometer and why? What are the advantages of and limitations on such 
 an instrument? Explain the construction of an expansion pyrometer. 
 What are its advantages and disadvantages? 
 
 34. Explain the calorimetric pyrometer. Write and explain the for- 
 mula on which it is based. What are its limitations? Explain the 
 thermo-electric pyrometer. What materials are used in making up the 
 elements for different ranges of temperature? 
 
 35. Make line sketch, showing the wiring arrangement in the resist- 
 ance pyrometer. Explain its principle of operation. 
 
 36. Show by line sketch the principle of the reflecting pyrometer. 
 What is the rule for its working distance from the hot body? 
 
 37. Sketch an engine counter. What are maneuvering indicators and 
 how fitted? Explain a differential counter gear for multiple shafts. 
 
 38. Make line sketch and explain the operation of a tachometer for 
 high speed engines. 
 
 39. Make line sketches and explain the transmitter and receiver in a 
 Hutchison marine tachometer. 
 
 40. Explain the principle of the Hopkins electric tachometer. What 
 fault is possessed by electric tachometers in general and how is it over- 
 come? Explain the McNab marine register and indicator. 
 
 41. Sketch and explain the operation of the Davison speed regulator. 
 
 42. What is the Taylor counter? The Bailey counter? Explain with 
 line sketch the precision tacograph and its use. 
 
 CHAPTER IV. 
 
 43. Define or explain the following thermodynamic terms: (1) heat, 
 (2) quantity of heat, (3) temperature, (4) absolute temperature, (5) 
 entropy, (6) boiling point, (7) British thermal unit (B. T. V.}, (8) 
 specific Ueat, (9) sensible heat, (10) latent heat, (11) total heat, (12) 
 Joule's equivalent, (13) saturated steam, (14) wet steam, (15) super- 
 heated steam, (16) quality of steam. 
 
 44. What are the three general classes of steam calorimeters? Give 
 an example of each class, and briefly state the principle of construction 
 and operation. 
 
 45. Sketch and explain Carpenter's throttling calorimeter, deriving 
 the general equation for its use. Explain with diagram the graphical 
 solution. 
 
 46. Explain the calibration method for the use of Carpenter's throt- 
 tling calorimeter and derive the equation. Discuss briefly the limita- 
 tions of the instrument. 
 
 47. Sketch Thomas' superheating steam calorimeter and connections. 
 Briefly describe its operation and derive equations for its use. 
 
 48. Sketch Carpenter's separating steam calorimeter and connections 
 and derive the equations used with it. 
 
 49. Derive equations for use in connection with the barrel calorimeter. 
 Describe the test, stating precautions. In all calorimeter experiments 
 what care must be exercised in sampling the steam? 
 
QUESTIONS ON THE TEXT 297 
 
 50. Discuss the temperature-entropy diagram for water, and show its 
 application to steam. Explain the difference between isothermal and 
 adiabatic lines on the diagram. 
 
 51. Discuss the temperature-entropy diagram (1) for an ideal engine, 
 (2) for a real engine. 
 
 CHAPTER V. 
 
 52. For what purposes are water meters employed by naval engineers? 
 Describe the Worthington water meter. How is the rate of flow deter- 
 mined? What are the disadvantages attending the use of water meters, 
 as usually constructed, on board ship? 
 
 63. Sketch and describe the Keystone water meter. 
 
 54. Make sketch in section of the Venturi meter and describe its 
 operation. Make line sketch of the recording apparatus. 
 
 55. Derive equations for use with the Venturi meter. 
 
 56. Explain with sketches the early forms of Pitot's tube used for 
 measuring the rate of flow of water. 
 
 57. Sketch and explain the operation of Darcy's improved form of 
 Pitot's tube for measuring the rate of flow of streams. 
 
 58. In the pitometer, sketch the street connection, rod meter and 
 manometer in place. Explain how this instrument is inserted and 
 adjusted for use in a water main. 
 
 59. Sketch the street connection for use with a pitometer and explain 
 the traverse. 
 
 60. What is a weir? Give general features of construction, stating 
 the points that must be observed in order to obtain accurate results. 
 
 61. Describe the construction and operation of a tank for weir ap- 
 paratus, illustrating with free-hand sketches. 
 
 62. Describe the construction, operation and use of the hook gage. 
 Explain with sketch the triangular notched weir and write the for- 
 mulas for its use. What are its advantages? Explain what is meant 
 by miner's inch. 
 
 CHAPTER VI. 
 
 63. Sketch an anemometer and explain how it should be used. To 
 what extent may its indications be relied upon? 
 
 64. Sketch the special form of Pitot's tube used in Taylor's method 
 of measuring low air velocities. How is this used to obtain the mean 
 velocity in a pipe of circular section? 
 
 65. Make line sketch of the manometer used in measuring low air 
 velocities by Taylor's method and explain its use. 
 
 66. Knowing the difference between impact and static pressures by 
 Pitot's tube in an air pipe, how proceed to calculate the rate of flow? 
 How may a single Pitot tube be used in tests of this character? 
 
 67. Give Napier's rule for the rate of discharge of steam through a 
 nozzle. What are its limitations? What is the direct measure of the 
 efficiency of a steam engine? How is this usually obtained in engine 
 tests? What are steam meters and of what value are they? 
 
 68. Into what two classes are steam meters divided? Sketch the pipe 
 connections for the Sarco steam meter in horizontal and vertical pipes, 
 
298 EXPERIMENTAL ENGINEERING 
 
 write the formula for the flow of steam under these conditions and 
 explain its operation. 
 
 69. Make line sketch of the recording mechanism in a Sarco steam 
 meter and explain its operation. 
 
 70. Distinguish between shunt and series steam meters. Sketch the 
 nozzle plug used in the General Electric Company's steam flow meter 
 and explain its use. 
 
 71. Explain with line sketch the General Electric Company's recording 
 flow meter with automatic pressure correction device. 
 
 72. Explain with line sketch the General Electric Company's indi- 
 cating flow meter. How does the air flow meter differ from the steam 
 flow meter? 
 
 CHAPTER VII. 
 
 73. Explain what is meant by (1) power, (2) Horse power, (3) indi- 
 cated horse power, (4) brake Horse power, (5) shaft horse power, (6) 
 metric horse power, (7) efficiency of a machine. What is the relation 
 between indicated horse power and shaft horse power in an ordinary 
 steam engine, and in a steam turbine? 
 
 74. To what errors is the steam engine indicator subject and how are 
 they corrected? 
 
 75. What is the object of calibrating indicator springs? What pre- 
 cautions should be observed? Make line sketch of Carpenter's indicator 
 and gage tester and briefly describe its operation. 
 
 76. Make a free-hand line sketch, showing the essential parts of 
 Cooley's indicator testing apparatus, and give a brief description of the 
 apparatus. 
 
 77. Describe fully the method used for calibrating an indicator spring, 
 using Cooley's tester. Illustrate with free-hand sketch. Explain how 
 the results are calculated. How are pressure tests below the atmosphere 
 made? 
 
 78. Sketch and explain the operation of the Hospitalier-Carpentier 
 monograph. For what class of engines is it more particularly adapted? 
 
 79. Make a free-hand sketch, in section, of Ripper's mean pressure 
 indicator, and explain the operation of the apparatus. 
 
 80. Make a line sketch of Ripper's power board and show the infor- 
 mation that may be obtained from it. Explain its construction, show- 
 ing that it is merely a special application of the slide rule. From the 
 same sketch explain Hudson's power scale. 
 
 81. Explain in detail how to construct a logarithmic diagram for 
 computing horse power. 
 
 82. Describe with sketch the Prony brake, explaining its use. De- 
 scribe the water brake. How may a dynamo be used as a brake? What 
 is electric horse power and how is it made use of to determine the 
 indicated horse power? What is a belt dynamometer? 
 
 83. Sketch in section the Kenerson transmission dynamometer and 
 explain its operation. 
 
 84. Make line sketch showing the working parts of the Emerson 
 power scale and explain its operation. 
 
QUESTIONS ON THE TEXT 299 
 
 85. Explain the principle of a torsionmeter for measuring horse 
 power. Draw diagram and show how a shaft is calibrated, deducing the 
 equations. 
 
 86. Show how to obtain the horse power with a torsionmeter without 
 calibration of shafting. Why is it necessary to calibrate? 
 
 87. Explain the operation of one of the following torsionmeters, illus- 
 trating with line sketches: Denny- Johnson, Fottinger, Hopkinson- 
 Thring, Metten. What are the advantages and disadvantages of the 
 instrument selected? 
 
 88. Discuss briefly torsionmeters compared with the indicator, torsion 
 readings for reciprocating engines, and curves of chest pressures and 
 horse power. 
 
 CHAPTER VIII. 
 
 89. Define or explain what is meant by (1) stress, (2) strain, (3) 
 elasticity, (4) elastic limit, (5) rigidity or stiffness, (6) coefficient of 
 ultimate strength, (7) coefficient of strength at the elastic limit, (8) 
 percentage of elongation, (9) reduction of area of cross section, (10) 
 modulus of elasticity, (11) modulus of resilience, (12) breaking load, 
 (13) maximum load, (14) safe load, (15) factor of safety. 
 
 90. Make a line sketch of a stress-strain diagram and explain the in- 
 formation that may be derived from it. What are the general features 
 of construction of a power testing machine? 
 
 91. Show by line sketches the relative positions of pulling screws, top 
 and pulling heads, weighing table, and test specimen, in using the 
 Riehle screw power testing machine for tests as follows: (1) tension, 
 (2) compression, (3) bending and shearing. 
 
 92. Make a line sketch showing the essential features of a Riehle 
 screw power testing machine. Describe briefly the automatic apparatus 
 for keeping the scale beam in balance. Describe the autographic ap- 
 paratus. For what purpose is an extensometer used? What is the yield 
 point f 
 
 93. Describe the behavior of the two general classes of metals, plastic 
 and brittle, when subjected in short specimens to compressional stress. 
 What information may be obtained in each case? For what purpose are 
 cross bending tests made? 
 
 94. Describe the practical use of a testing machine. What informa- 
 tion is obtained? What is placed on the test record? Sketch and de- 
 scribe the standard test pieces used in the navy. 
 
 95. Describe with line sketches the construction and operation of the 
 Riehle-Miller torsional testing machine. Show the details of the weigh- 
 ing mechanism. Explain the use of the torsion indicator. 
 
 96. Name and briefly describe the different varieties of cement. What 
 is concrete? Reenforced concrete? What are the uses of cement on 
 board ship? 
 
 97. Enumerate and briefly describe the several tests to which cement 
 is subjected. 
 
 98. Sketch a cement tensile test specimen and describe how these are 
 made. 
 
 99. Make line sketch of Fairbank's cement testing machine. Show 
 the details of the weighing mechanism. Describe the test of a tensile 
 specimen. 
 
300 EXPERIMENTAL ENGINEERING 
 
 CHAPTER IX. 
 
 100. In engine lubrication distinguish between the open system and 
 forced lubrication. What is the difference in the character of the oil 
 used in these two systems? 
 
 101. What are the essential characteristics of the various oils used 
 for lubricating purposes at the present time? Name and distinguish 
 between these various lubricants. 
 
 102. What are the principal determinations in making a complete test 
 of a lubricant? What value have these various tests? 
 
 103. Make line sketch and describe Engler's or the Boverton-Redwood 
 viscosimeter. 
 
 104. Describe the method of determining the following information in 
 regard to a lubricant: (1) flash point, (2) burning point, (3) chill 
 point, (4) loss by evaporation, (5) gumming properties. 
 
 105. Sketch and describe the New York State Board of Health or the 
 Cleveland flash point tester. 
 
 106. Give a brief description of an oil testing machine, illustrating 
 with sketches. State the four methods of observation in practice, for a 
 durability test. 
 
 107. Describe the preparations for and the method of conducting an 
 oil durability test by the Navy Department method, using an oil testing 
 machine. 
 
 108. What tests of oil can be made on board ship? What are the best 
 safeguards in purchasing oil and in using it? 
 
 CHAPTER X. 
 
 109. In what ways may fuel economy be influenced? What qualities 
 in coal govern its selection as a fuel? What tests are made on ship- 
 board? 
 
 110. How proceed in obtaining a sample of coal for test? Describe the 
 tests (1) for moisture, (2) for volatile matter, (3) for ash. 
 
 111. Make neat line sketch of Mahler's fuel calorimeter. Letter and 
 name the essential parts. 
 
 112. Describe in detail the construction of the bomb used with Mah- 
 ler's fuel calorimeter. What are its capacity and thickness? Why is it 
 made so large and heavy? State the principle of operation of the 
 calorimeter. 
 
 113. With the assistance of a rough free-hand sketch explain the 
 method of conducting a test for calorific value, using the Mahler 
 calorimeter. What is meant by the water equivalent and how is its 
 value determined? 
 
 114. Make a neat sketch of Carpenter's improved coal calorimeter 
 and connections. Letter and name the essential parts. 
 
 115. Draw a sectional view, showing the details of the pressure re- 
 ducer used in connection with Carpenter's coal calorimeter. Describe 
 precautions in using. 
 
 116. With the assistance of a free-hand sketch, describe the method of 
 preparing for and conducting a test for calorific value, using Carpenter's 
 coal calorimeter. 
 
QUESTIONS ON THE TEXT 301 
 
 117. How is the curve of calibration obtained for Carpenter's coal 
 calorimeter? What is its use? Briefly state the principle of the calo- 
 rimeter. What is meant by the corrected scale reading and how is it 
 obtained? 
 
 118. Sketch Parr's standard coal calorimeter and briefly describe its 
 operation and use. 
 
 119. Sketch Junker's calorimeter. Letter and name the essential 
 parts. Explain its operation in determining the heating value of a 
 gas. 
 
 120. State briefly the requirements of the navy specifications for fuel 
 oil. What instruments are contained in the navy portable test outfit 
 and what are their uses? 
 
 121. Sketch the Pensky-Marten apparatus for determining the flash 
 point and explain its operation. 
 
 122. What oils are used on naval vessels as fuel, other than fuel oil. 
 What are the requirements for each? 
 
 CHAPTER XL 
 
 123. What is the object of making analyses of smoke pipe gases? 
 Explain how the percentage of CO 2 is a measure of the efficiency of 
 combustion. What reagents are used in flue gas analysis? Explain the 
 use of each. How obtain a sample of the gas? 
 
 124. Sketch the Orsatt-Muencke apparatus and explain its operation 
 in flue gas analysis. State precautions to be observed. 
 
 125. Having given the results of the chemical analysis of a fuel and 
 the results of flue gas analysis, burning this fuel, show how to calculate 
 the pounds of dry gas per pound of combustible. 
 
 126. Enumerate the several items that go to make up the heat balance, 
 showing the distribution of the heating value of one pound of com- 
 bustible, and explain briefly how each item is calculated. 
 
 127. Sketch and describe the Hays CO 2 apparatus and explain its 
 operation. 
 
 128. Make a free-hand sketch of the analyzing and recording ap- 
 paratus in the Sarco CO 2 recorder, and explain the operation of the 
 apparatus. 
 
 129. Make diagrammatic sketch and explain the essential principle of 
 the Uehling CO 2 recorder. 
 
 130. Explain the use of time firing devices in promoting efficiency of 
 combustion. Explain with line sketch the layout of such an apparatus 
 on board ship. 
 
 131. Draw diagram of one of the following time firing devices and 
 explain its operation: Corey's, old type; Corey's, new type; Kilroy's; 
 Sub-Target Gun Company's; General Electric Company's, new type; 
 Lobitz. 
 
 20 
 
INDEX 
 
 PAGE 
 
 Absolute temperature 85 
 
 Absorption dynamometers 161 
 
 Adiabatic expansion 106 
 
 Air flow meter 143 
 
 velocities, Taylor's method 130 
 
 Albany grease 214 
 
 Amsler's planimeter, demonstration 33 
 
 description 31 
 
 Anemometer 129 
 
 Averaging instrument, Coffin's, demonstration 37 
 
 description 35 
 
 Bailey counter 82 
 
 Barrel calorimeter 101 
 
 Beaume scale for oils 264 
 
 Belt dynamometers 163 
 
 Bevis-Gibson flashlight torsionmeter. 176 
 
 Boiling point 86 
 
 Boverton-Redwood viscosimeter 217 
 
 Brake, electric 162 
 
 Brake horse power 144, 161 
 
 Brake, Prony 161 
 
 water 162 
 
 Breaking load 188 
 
 Bristol's recording gage 46 
 
 Calculating instrument, Fuller's 27 
 
 Thacher's 30 
 
 Calibration of indicator springs 146 
 
 shafting 169 
 
 Calorimeter, barrel 101 
 
 Carpenter's improved coal 241 
 
 calibration 248 
 
 operation 246 
 
 pressure reducer for 
 
 oxygen 244 
 
 separating steam 99 
 
 throttling 87 
 
 Junker's, description 254 
 
 operation 256 
 
 Mahler's fuel, description 232 
 
 operation 235 
 
 water equivalent 238 
 
 Parr's standard fuel, description 248 
 
 operation, anthracite or coke. . 253 
 
 oil fuels 253 
 
 soft coal 251 
 
 separating 99 
 
 steam 87 
 
 superheating 92 
 
 Thomas superheating 92 
 
 throttling ~. 87 
 
 calibration method 89 
 
 graphic solution 89 
 
 limitations . 92 
 
304 INDEX 
 
 PAGK 
 
 Calorimetric pyrometer 65 
 
 Carpenter's improved coal calorimeter, calibration 248 
 
 description 241 
 
 operation 246 
 
 pressure reducer for oxygen 244 
 
 improved separating steam calorimeter 99 
 
 indicator and gage testing apparatus 147 
 
 Cement, constancy of volume 208 
 
 fineness 207 
 
 hydraulic 205 
 
 mixed 206 
 
 natural 205 
 
 neat -. 206 
 
 Portland 205 
 
 Pozzuolana or Puzzolan 206 
 
 setting 208 
 
 specific gravity 207 
 
 tensile strength 207 
 
 tests 209 
 
 testing 207 
 
 testing machine 210 
 
 test piece 209 
 
 time of setting 207 
 
 uses on board ship 206 
 
 Chest pressure and horse power curves 186 
 
 Cleveland flash tester 220 
 
 CO 2 apparatus, Hays 272 
 
 Sarco automatic recorder, application 279 
 
 description 274 
 
 Uehling recorder 279 
 
 Coal, sampling 230 
 
 selecting 229 
 
 test for ash 232 
 
 moisture 231 
 
 heating value 232 
 
 volatile matter 231 
 
 testing on board ship 230 
 
 Coefficient of strength at elastic limit 188 
 
 ultimate strength 188 
 
 Coffin's averaging instrument, demonstration 37 
 
 description 35 
 
 Compression tests 200 
 
 Computations, preparations for 15 
 
 Concrete 206 
 
 reenforced 206 
 
 Cooley's indicator testing apparatus 148 
 
 Corey time firing device, new type 283 
 
 old type 281 
 
 Counter, Bailey 82 
 
 revolution 72 
 
 Taylor 82 
 
 Cross bending tests 200 
 
 Cylinder oil 213 
 
 Darcy's Pitot tube 116 
 
 Davison speed regulator 79 
 
 Deviations from true measurement. . 11 
 
INDEX 305 
 
 PAGE 
 
 Denny- Johnson torsiometer 172 
 
 Diagram, strain 189 
 
 stress-strain 190 
 
 Differential counter 73 
 
 Duplex slide rule 30 
 
 Dynamometers, absorption 161 
 
 belt 163 
 
 Kenerson transmission 164 
 
 transmission 163 
 
 transmission, Emerson 167 
 
 Economy of fuel 229 
 
 Efficiency of a machine 144 
 
 Elasticity 187 
 
 modulus 188 
 
 Elastic limit 187 
 
 coefficient of strength 188 
 
 Electric brake 162 
 
 horse power 162 
 
 tachometer, Hopkins 77 
 
 tachometer, Hutchison 75 
 
 Elongation, percentage 188 
 
 Emerson power scale 167 
 
 Engineering calculations 9 
 
 Engine lubrication, forced 212 
 
 open 212 
 
 Engler's viscosimeter 216 
 
 Entropy 85 
 
 diagram 102 
 
 of steam 105 
 
 steam and water 105 
 
 water 103 
 
 Error, limits of in observations 11 
 
 sources of 12 
 
 Errors classified 12 
 
 Expansion pyrometer 63 
 
 Experimental Engineering defined 9 
 
 Extensometer, Riehl-Kenerson 197 
 
 Factor of safety 189 
 
 Fairbanks cement testing machine 210 
 
 Figures, significant 13 
 
 rules for 14 
 
 Flashlight torsionmeter 176 
 
 Flash tester, Cleveland 220 
 
 New York State Board of Health 219 
 
 Pensky-Martens 261 
 
 Tagliabue 265 
 
 Flash test, open cup 219 
 
 Flue gas analysis 266 
 
 calculations 269 
 
 combustion losses 271 
 
 Formula for air velocities, Pitot's tube 133 
 
 flow of steam, Napier's 133 
 
 Fottinger torsionmeter 177 
 
 Fuel economy 229 
 
 Fuller's calculating instrument 27 
 
306 INDEX 
 
 PAGE 
 
 Gage, hook 124 
 
 tester 52 
 
 testing apparatus, Carpenter's 148 
 
 dead weight 51 
 
 Gages, Bourdon 44 
 
 compound 45 
 
 recording pressure 45 
 
 diaphragm 44 
 
 pressure 43 
 
 testing 51 
 
 testing vacuum 53 
 
 vacuum 43 
 
 Gasoline, specifications for 265 
 
 Gas thermometer 54 
 
 General Electric Company, steam meter 138 
 
 time firing device, new type 288 
 
 Hays COo apparatus 272 
 
 Heat, defined 85 
 
 Hook gage 124 
 
 Hopkins electric tachometer 77 
 
 Hopkinson-Thring torsionmeter 179 
 
 Horse power 144 
 
 electric 162 
 
 Hospitalier-Carpentier Manograph 151 
 
 Hudson's horse power scale 158 
 
 Hutchison marine tachometer 75 
 
 Hydraulic cement 205 
 
 Ice machine oil 214 
 
 Inch, Miner's 128 
 
 Indicated horse power 144 
 
 Indicator, errors of 145 
 
 maneuvering 72 
 
 Ripper's mean pressure 153 
 
 springs, calibration 146 
 
 testing apparatus, Carpenter's 147 
 
 Cooley's 148 
 
 torsion 203 
 
 Industrial Thermograph 57 
 
 Isothermal expansion 106 
 
 Joule's equivalent 86 
 
 Junker's calorimeter, description 254 
 
 operation 256 
 
 Kenerson transmission dynamometer 164 
 
 Keystone water meter Ill 
 
 Kilroy time firing device 284 
 
 Latent heat 86 
 
 Limits of error in observations 11 
 
 Load, breaking 188 
 
 maximum 188 
 
 safe 188 
 
 Lobitz time firing device 391 
 
 Logarithmic scale 17 
 
 horse power 158 
 
 McNab marine register 78 
 
 Mahler's fuel calorimeter, description 232 
 
 operation 235 
 
 water equivalent 238 
 
INDEX 307 
 
 PAGE 
 
 Maneuvering indicator 72 
 
 Manograph, Hospitalier-Carpentier 151 
 
 Manometers 53 
 
 Manometer for measuring air velocities 131 
 
 Maximum load 188 
 
 Mean pressure indicator, Ripper's 153 
 
 Measurements, true, deviations from 11 
 
 direct and indirect 11 
 
 Measuring machines 40 
 
 Mercurial pyrometer 60 
 
 Meter, Venturi 112 
 
 Meters, steam 134 
 
 steam, series 134 
 
 shunt 138 
 
 water 110 
 
 Metric horse power 144 
 
 Metten torsionmeter 183 
 
 Miner's inch 128 
 
 Mistakes 13 
 
 Modulus of elasticity 188 
 
 Modulus of resilience 188 
 
 Napier's formula 133 
 
 New York State Board of Health Flash Tester 219 
 
 Observations, arithmetical mean 12 
 
 weighted mean 12 
 
 Oil, cylinder 213 
 
 Oil fuel, Beaume scale 264 
 
 fire test 263 
 
 flash test 261 
 
 navy specifications 259 
 
 specific gravity 263 
 
 test for water and sediment 263 
 
 testing, navy portable outfit 261 
 
 Oil, ice machine 214 
 
 kerosene, specifications for 264 
 
 lard 264 
 
 lubricating, for forced lubrication 213 
 
 wick or gravity feed 213 
 
 mineral, specifications for 264 
 
 Vaclite 264 
 
 test, acid 222 
 
 burning point 221 
 
 cold 222 
 
 evaporation 222 
 
 Oil test, flash 219 
 
 gumming or drying 219 
 
 specific gravity 215 
 
 viscosity 215 
 
 Oil testing, determinations required 214 
 
 log 227 
 
 machine . 222 
 
 methods with machine 224 
 
 Navy Department method 225 
 
 on board ship 227 
 
 Omnimetre, Sexton's 18 
 
 Orsatt-Muencke Apparatus, description 266 
 
 operation 268 
 
308 INDEX 
 
 PAGE 
 
 Parr standard fuel calorimeter, description 248 
 
 operation, soft coal 251 
 
 anthracite or coke 253 
 
 oil fuels 253 
 
 Pensky-Martens flash tester 261 
 
 Percentage of elongation 188 
 
 Phosphor-bronze, tests of 201 
 
 Pitot's tube 144 
 
 for air 130 
 
 single, for air velocities 133 
 
 Pitometer 119 
 
 rod meter 119 
 
 street connection 121 
 
 traverse 121 
 
 U tube 121 
 
 Planimeters 31 
 
 Planimeter, Amsler's, demonstration 33 
 
 description 31 
 
 Coffin's averaging instrument, demonstration 37 
 
 description 35 
 
 Pneumatic pyrometer, Uehling's 57 
 
 Portable tachometer 75 
 
 Power board, Ripper's 157 
 
 Power, defined 144 
 
 Power scale, Emerson 167 
 
 Pratt and Whitney measuring machine 40 
 
 Precision tacograph 83 
 
 Prony brake 161 
 
 Pyrometers 54 
 
 Pyrometer, calorimetric 65 
 
 expansion 63 
 
 mercurial 60 
 
 reflecting 71 
 
 resistance 67 
 
 thermo-electric 66 
 
 Uehling's pneumatic 67 
 
 Quality of steam 87 
 
 Quantity of heat 85 
 
 Recorder, Uehling's differential pressure 49 
 
 Recording pressure gage 45 
 
 Reduction of area 188 
 
 Reflecting pyrometer 71 
 
 Reports, how made 16 
 
 Resilience, modulus 188 
 
 Resistance thermometer 67 
 
 Revolution counters 72 
 
 Revolution counter, differential 73 
 
 Riehle-Kenerson extensometer 197 
 
 RiehlS screw power testing machine 191 
 
 Rigidity 188 
 
 Ripper's mean pressure indicator 153 
 
 power board 157 
 
 Rod meter for pitometer 119 
 
 Rules for significant figures 14 
 
 Safe load 188 
 
 Safety, factor 189 
 
 Sampling coal 230 
 
INDEX 309 
 
 PAGE 
 
 Sarco automatic CO 2 recorder, application 279 
 
 description 274 
 
 steam meter 134 
 
 Saturated steam 86 
 
 Scale, logarithmic 17 
 
 Selecting coal 229 
 
 Sensible heat 86 
 
 Sexton's omnimetre 18 
 
 Shaft horse power 144 
 
 Shafting, calibration 169 
 
 Shearing tests 201 
 
 Significant figures 13 
 
 rules for 13 
 
 Slide rule, double scale 28 
 
 duplex 30 
 
 Puller's 27 
 
 horse power 158 
 
 Sexton's omnimetre 18 
 
 Sperry's pocket calculator 27 
 
 straight 28 
 
 Thacher's 30 
 
 Sources of error 12 
 
 Specific heat 86 
 
 Speed regulator, Davison 79 
 
 Sperry's pocket calculator 27 
 
 Steam calorimeters 87 
 
 flow meter, indicating 143 
 
 recording 139 
 
 meters 134 
 
 meter, General Electric Co 138 
 
 Sarco 134 
 
 sampling , 102 
 
 Stiffness 188 
 
 Strain 187 
 
 diagram 189 
 
 Strength of materials 187 
 
 Stress 187 
 
 Stress-strain diagram 190 
 
 Sub-Target Gun Co.'s time firing device 286 
 
 Superheated steam 87 
 
 Tacograph, precision 83 
 
 Tachometers 73 
 
 Tachometer, Hopkins electric 77 
 
 Hutchison marine 75 
 
 portable 75 
 
 Tagliabue flash tester 265 
 
 Tallow 214 
 
 Tank for weir apparatus 125 
 
 Taylor counter 82 
 
 method for low air velocities 130 
 
 Temperature 85 
 
 Temperature-entropy diagram 102 
 
 ideal engine 106 
 
 real engine 107 
 
 measurement 53 
 
 Testing machines 190 
 
310 INDEX 
 
 PAGE 
 
 Testing machine, autographic apparatus 196 
 
 automatic apparatus 196 
 
 cement 210 
 
 compression tests 200 
 
 cross bending tests 200 
 
 driving mechanism 195 
 
 oil 222 
 
 practical use 197 
 
 Riehl6-Miller, torsional 202 
 
 Richie" screw power 191 
 
 screw and vernier beams 195 
 
 shearing tests 201 
 
 straining mechanism 191 
 
 test record 199 
 
 weighing apparatus 194 
 
 pressure gages 51 
 
 vacuum gages 53 
 
 Test pieces, standard 202 
 
 Thacher's calculating instrument 30 
 
 Thermal unit 86 
 
 Thermo-electric pyrometer 66 
 
 Thermograph, Industrial 57 
 
 Thermometer, gas 54 
 
 high temperature 54 
 
 resistance 67 
 
 Thomas superheating calorimeter 92 
 
 Time firing devices, general description 280 
 
 device, Corey, new type 283 
 
 old type 281 
 
 General Electric Co., new type 288 
 
 Kilroy 284 
 
 Lobitz 291 
 
 Sub-Target Gun Co 286 
 
 Torsion indicator 203 
 
 readings on reciprocating engines 186 
 
 Torsional testing machine 202 
 
 Torsionmeters 169 
 
 Torsionmeter and indicator compared 184 
 
 Torsionmeter, Bevis-Gibson flashlight 176 
 
 Denny-Johnson 172 
 
 Fottinger 177 
 
 Hopkinson-Thring 179 
 
 horse power without calibration 171 
 
 Metten 183 
 
 Total heat 86 
 
 Transmission dynamometers 163 
 
 Uehling CO 2 recorder 279 
 
 differential pressure recorder 49 
 
 pneumatic pyrometer 57 
 
 Ultimate strength 188 
 
 Vaclite 264 
 
 Vaseline 214 
 
 Venturi meter 112 
 
 Viscosimeter, Boverton-Redwood 217 
 
 Engler's 216 
 
INDEX 311 
 
 PAGE 
 
 Water brake 162 
 
 meters 110 
 
 meter, Keystone Ill 
 
 Worthington 110 
 
 Webb, C. A., oil testing, method 225 
 
 Weirs 122 
 
 Weir, tank 125 
 
 triangular 126 
 
 Wet steam 86 
 
 Worthington water meter 110 
 
UNIVERSITY OF CALIFORNIA LIBRARY 
 This book is DUE on the last date stamped below. 
 
 17t947 
 
 LD 21-100m-12,'46(A2012sl6)4120 
 
i^UV r 
 
 /