GIFT OF 
 
 ASSOCIATED ELECTRICAL AND 
 MECHANICAL ENGINEERS 
 
 MECHANICS DEPARTMENT 
 
 399 
 
University of California 
 
THE 
 
 GAS TURBINE 
 
 PROGRESS IN THE DESIGN AND CONSTRUCTION 
 
 OF TURBINES OPERATED BY GASES 
 
 OF COMBUSTION 
 
 BY 
 
 HENRY HARRISON SUPLEE, B.Sc. 
 
 MEMBER OF THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS, MEMBER OF THE FRANKLIN 
 
 INSTITUTE, MEMBRE DE LA SOCIETE DES INGENIEURS CIVILS DE FRANCE, 
 
 MITGLIED DES VEREINES DEUTSCHER INGENIEURE. 
 
 AUTHOR OF 
 
 "THE MECHANICAL ENGINEER'S REFERENCE BOOK," ETC., ETC. 
 
 PHILADELPHIA 
 
 J. B. LIPPINCOTT COMPANY 
 
 1910 
 
A 1 
 
 Engineering 
 Library 
 
 COPYRIGHT, 1910 
 BY J. B. LIPPINCOTT COMPANY 
 
 Printed by J. B. Lippincott Company 
 The Washington Square Press, Philadelphia, U. S. A. 
 
PREFACE 
 
 THIS volume is intended to place in the hands of engi- 
 neers and experimenters such theoretical and practical data 
 as are now available in the solution of the problem of the 
 gas turbine. 
 
 At the present time such machines are yet in the experi- 
 mental stage, and it is still uncertain to what extent they 
 may become generally practicable. There has, however, been 
 expended much study and effort, both in the investigation of 
 the theoretical principles upon which the gas turbine depends, 
 and in the construction of machines intended to realize, more 
 or less effectively, the possibilities which have been indicated 
 /by such studies. 
 
 Much of the information contained in this book is included 
 in the transactions of learned societies, in the pages of peri- 
 odicals, and in the records of private experimenters, and it is 
 believed that by gathering in one volume the results of the 
 work of English, French, German, and other organizations, 
 the engineers and mechanics who are investigating the sub- 
 ject may be assisted by perceiving what has already been 
 accomplished, and thus avoid unnecessary repetition of 
 work which is already on record. 
 
 The gas turbine need not be a machine of exceedingly 
 high thermal efficiency in order to be available for many 
 purposes. The advantages accompanying a continuous turn- 
 ing effort, instead of the intermittent impulses of the recip- 
 rocating gas or gasoline motor, may, in many instances, over- 
 balance a somewhat lower fuel economy; while the reduction 
 in weight, consequent upon the attainment of a very high 
 rotative speed, may become of controlling importance. It is 
 therefore of the utmost desirability that all the conditions be 
 
 i 
 
PREFACE 
 
 taken into account, and it is for this purpose that the present 
 volume has been prepared, collecting together the relative 
 influence of the various elements of which the problem is 
 composed. 
 
 The author desires to acknowledge the assistance which 
 has been freely rendered to him in the preparation of this 
 volume. To the memory of his colleague in the Societe des 
 Ingenieurs Civils de France, M. Rene Armengaud, he records 
 his obligations for personal communications describing the 
 experimental work conducted in the laboratory at St. Denis; 
 and to M. Alfred Barbezat he desires to express his apprecia- 
 tion for the continuation of this most important work. To 
 M. Armand de Dax, Secretaire Administratif of the Societe 
 des Ingenieurs Civils de France, and to M. L. Sekutowicz, his 
 colleague in the Societe, he acknowledges the kind permission 
 to translate the important paper of the latter author; and to 
 Mr. Edgar Worthington, Secretary of the Institution of 
 Mechanical Engineers (London) as well as to the author, Mr. 
 R. M. Neilson, he is indebted for permission to reproduce the 
 paper of the latter, as well as the discussion which it elicited. 
 The writer also wishes to express his indebtedness to Dr. San- 
 ford A. Moss, Dr. Charles E. Lucke, Prof. Sidney A. Reeve, 
 Prof. Lionel S. Marks, and Dr. H. N. Davis, for valued sug- 
 gestions and assistance. 
 
 HENRY HARRISON SUPLEE. 
 
 NEW YORK, November, 1909 
 
CONTENTS 
 
 PAGE 
 
 INTRODUCTION 5 
 
 CHAPTER I. 
 
 HISTORICAL . 9 
 
 The Smoke Jack, the First Gas Turbine; Barber's Turbine; Ferni- 
 hough's Patent; Burdin's Experiments; Tournaire's Communication 
 to the French Academy; Bourne's Suggestions; Boulton's Patents; 
 the Stolze Hot- Air Turbine; Parsons's Patent; Laval, Lemale, Armen- 
 gaud; Later Papers. 
 
 CHAPTER II. 
 
 THE DISCUSSION BEFORE THE INSTITUTION OF MECHANICAL ENGINEERS. 28 
 Paper by R. M. Neilson, Discussing Various Available Cycles and 
 their Merits and Disadvantages; Discussions by Messrs. Henry 
 Davey, F. W. Burstall, James Atkinson, Col. R. E. B. Crompton, H. 
 M. Martin, Robert H. Smith, and Mr. Neilson. Communications 
 from Dugald Clerk, W. J. A. London, E. Kilburn Scott, and George 
 A. Wigley. 
 
 CHAPTER III. 
 
 THE DISCUSSION BEFORE THE SOCIETY OF CIVIL ENGINEERS OF FRANCE. 108 
 Paper by M. L. Sekutowicz, Treating of the Mechanical and Ther- 
 modynamic Efficiencies of the Gas Turbine, the Various Cycles and 
 the General Details of Construction. 
 
 CHAPTER IV. 
 
 THE DISCUSSION BEFORE THE SOCIETY OF CIVIL ENGINEERS OF FRANCE 
 
 (Continued) 219 
 
 Comments on the Paper of M. Sekutowicz by MM. Armengaud, Rey, 
 Hart, Bochet, and Letombe. 
 
 CHAPTER V. 
 
 ACTUAL BEHAVIOR OF GASES IN NOZZLES 222 
 
 Experiments of Dr. Charles E. Lucke; Table of Temperature Drop 
 with Compressed Air in the Laval Turbine; Trials of Turbines at St. 
 Denis; Desirability of Further Experiments. 
 
 iii 
 
iv CONTENTS 
 
 CHAPTER VI. 
 
 THE PRACTICAL WORK OF ARMENGAUD AND LEMALE 227 
 
 The Explosion Gas Turbine; the Combustion Turbine; the Original 
 Experimental Machine at St. Denis; Details of Combustion Chamber; 
 Structural Arrangement of Parts; the Rateau Polycellular Air Com- 
 pressor; the 300 Horse-Power Gas Turbine; Gas Turbines in Subma- 
 rine Torpedoes; Data and Results of Tests; the Karavodine Turbine. 
 
 CHAPTER VII. 
 
 GENERAL CONCLUSIONS 251 
 
 INDEX . 255 
 
THE GAS TURBINE 
 
 INTRODUCTION. 
 
 ALTHOUGH the gas turbine was one of the earliest forms 
 of combustion motor it failed to attain practical or com- 
 mercial importance, and it is only since the steam turbine 
 has reached its present commanding position that the possi- 
 bility of developing the gas turbine in similar manner has 
 been seriously considered. 
 
 The practical difficulties in the way of the realization 
 of a successful gas turbine are very great. The high tem- 
 peratures involved demand especial care in order that the 
 strength of the material may not be unduly affected. The 
 high rotative speeds required, if high efficiencies are to 
 be secured, render the mechanical problems connected with 
 centrifugal action more serious even than with the steam 
 turbine; while the doubt as to the action of hot gases in 
 diverging nozzles renders an important element in the 
 theory yet uncertain. 
 
 At the same time there has been going on, during the 
 past few years, in Europe and in America, some very effec- 
 tive experimental work upon the gas turbine; while the 
 theory has also been made the subject of elaborate study 
 by English, French, German, and American scientists. 
 
 Most of the information regarding this work is in a form 
 unavailable for the practical engineer and investigator. 
 The theoretical discussions are, for the most part, contained 
 in the transactions of professional societies; much of it in 
 languages other than English. The practical experiments 
 
 5 
 
GAS TURBINE. 
 
 are being conducted behind closed doors and reliable infor- 
 mation is not generally attainable in detailed form. 
 
 It has therefore been thought desirable to gather under 
 one cover the most important papers which have appeared 
 upon the subject of the gas turbine in England, France, 
 Germany, and Switzerland, together with some account 
 of the work in America, and to add to this such information 
 upon actual experimental machines as can be secured. 
 
 FIG. 1. Scheme of gas turbine with reciprocating compressor. 
 
 In the present state of the art this is all that can be done, 
 but it is believed that this will aid materially in the conduct 
 of subsequent work, and place in the hands of the gas-power 
 engineer a collection of material not generally accessible 
 or available in convenient form. 
 
 The general lines along which the plans of the various 
 gas turbines, now under experimental investigation, are con- 
 
THE GAS TURBINE. 
 
 structed, will be understood from the accompanying sche- 
 matic diagram. This is substantially as given by Dr. Sanford 
 A. Moss in connection with his thesis upon the gas turbine 
 presented to the faculty of Cornell University in 1903. 
 
 The fuel, in this case some form of liquid hydrocarbon, 
 is forced into a combustion chamber, together with the 
 proper amount of compressed air for its combustion. The 
 products of combustion are discharged through a diverg- 
 
 CombusElon Chamber 
 
 Air 
 Inlet 
 
 LJ 
 
 Compressor 
 
 Turbine & I 
 
 Exhaust 
 
 FIG. 2. Scheme of gas turbine with multiple wheels and rotary compressor. 
 
 ing nozzle upon the buckets of an impulse wheel which is 
 thus caused to rotate, a portion of the power developed 
 being used to drive the compressor, and the remainder being 
 available for external use. 
 
 Since one of the presumed advantages of the gas turbine 
 over the ordinary gas engine is the substitution of continu- 
 ous rotary motion for the reciprocating action of a piston 
 in a cylinder, it is undoubtedly desirable that a rotary com- 
 pressor be used, so that the reciprocating pump may be 
 
8 THE GAS TURBINE. 
 
 dispensed with. The general arrangement of such an appa- 
 ratus, including also a multistage turbine, is here given, as 
 devised by Mr. Rudolf Barkow (Fig. 2). 
 
 Here the air is drawn in and compressed by a rotary 
 compressor, mounted on a continuation of the turbine shaft. 
 The compressed air is delivered to the combustion chamber, 
 into which the liquid fuel is also injected, and the combustion 
 takes place under pressure, the gases and products of com- 
 bustion passing through the diverging nozzle to the buckets 
 and guide vanes of the turbine. 
 
 The extent to which these schematic forms have been 
 developed from the earliest beginnings, and the lines along 
 which theory and experiment have been pushed, will be 
 seen in the following pages. 
 
CHAPTER I. 
 
 * 
 
 HISTORICAL. 
 
 THE use of the expansive action of heat upon elastic gases 
 to operate a revolving wheel for the production of power 
 is by no means recent; in fact it antedates the employ- 
 ment of a piston reciprocating in a cylinder for the same 
 purpose. Considered in this broad sense there is no doubt 
 that the windmill is entitled to be called a gas turbine, since 
 the pressure of the moving air upon its sails can be traced 
 to 'the currents produced by changes of temperature in the 
 atmosphere. 
 
 Leaving aside the windmill, however, there can be little 
 question as to the claims of the mediaeval "Smokejack" 
 to be considered as a gas turbine. This machine, the origin 
 of which it is impossible to trace, has been attributed to 
 Leonardo da Vinci and illustrations of it are to be found 
 in his engineering sketch books. A somewhat later form is 
 shown in the illustration, this being taken from an engraving 
 in Bishop Wilkins's book " Mathematical Magick," published 
 in 1648, the present engraving and following description 
 being found in the edition of 1680. 
 
 After referring to the action of windmills and to Eolipiles, 
 the learned bishop continues: "But there is a better inven- 
 tion to this purpose mentioned in Cardan,* whereby a spit 
 may be turned (without the help of weights) by the motion 
 of the air that ascends the Chimney; and it may be useful 
 for the roasting of many or great joynts : for as the fire 
 must be increased according to the quantity of meat, so 
 the force of the instrument will be augmented proportionably 
 to the fire. In which contrivance there are these conveni- 
 ences above the Jacks of ordinary use : 
 
 * Cardan. De Variet. Rerum. I. 12, c. 58. 
 
10 
 
 THE GAS TURBINE. 
 
 "1. It makes little or no noise in the motion. 
 
 "2. It needs no winding up, but will constantly move of 
 itself, while there is any fire to rarifie the air. 
 
 "3. It is much cheaper than the other instruments that 
 are commonly used to this purpose. There being required 
 with it only a pair of sails, which must be placed in that 
 part of the Chimney where it begins to be straightened, and 
 one wheel, to the axis of which the spit line must be fastened, 
 according to the following Diagram. 
 
 FIG. 3. The smoke jack. The first gas turbine. From Bishop Wilkins's "Mathematical 
 
 Magick," 1680. 
 
 "The motion to these sails may likewise be serviceable 
 for sundry other purposes, besides the turning of a spit, 
 for the chiming of bells or other musical devices; and there 
 cannot be any more pleasant contrivance for continual and 
 
THE GAS TURBINE. 11 
 
 cheap music. It may be useful also for the reeling of yarn, 
 the rocking of a cradle with divers the like demestick occa- 
 sions. For (as was said before) any constant motion being 
 given, it is easie for an ingenious artificer to apply it unto 
 various services. 
 
 "These sails will always move both day and night, if 
 there is but any fire under them, and sometimes though 
 there be none. For if the air without be much colder than 
 that within the room, then must this which is more warm 
 and rarefied, naturally ascend through the chimney, to give 
 place unto the more condensed and heavy, which does 
 usually blow in at every chink or cranny, as experience 
 shews." 
 
 After the smoke jack, the next proposition for a gas 
 turbine appears to be that of Barber, who took out a British 
 patent in 1791, No. 1833, which seems like a very complete 
 .anticipation of nearly all the most recent developments in 
 this line. Barber's patent includes the distillation of the 
 gas from wood, coal, or oil, its delivery, with the proper 
 amount of air into a combustion chamber, and the discharge 
 of the products of combustion upon the buckets of a turbine 
 wheel. 
 
 He even went so far as to inject water into the combus- 
 tion chambers to reduce the temperature, the mixed steam 
 and gases acting upon the wheel. 
 
 The illustration, Fig. 4, gives an idea of Barber's patent. 
 
 The vessels marked 1, 1, are retorts for the production 
 of the gas to be used, these being intended for the distilla- 
 tion of coal, wood, etc., by means of an external flame. 
 When it is remembered that Murdock did not begin his 
 experimental investigations into the manufacture of coal 
 gas until 1792, one year after Barber's patent, and only 
 made his results public in 1797, it will be seen that Barber 
 was distinctly in advance of his time. The retorts shown 
 in Barber's drawing are in duplicate, for alternate charging 
 
12 
 
 THE GAS TURBINE. 
 
 FIG. 4. Barber's gas turbine, 1791. 
 
 and discharging, the' gas being delivered into a cooling 
 chamber B, from which it is drawn by one of the compres- 
 sing pumps C, D, and delivered to the receiver 4, from 
 which it passes to the triangular-shaped combustion cham- 
 
THE GAS TURBINE. 
 
 13 
 
 bers. The other compressing pump delivers air and vapor 
 of water into the combustion chamber, and the products 
 of combustion are discharged upon the buckets of the wheel 
 to effect its rotation. The drawing shows the reducing 
 gearing for operation of the compression pumps, the power 
 to be taken from the upper gear shaft. 
 
 It is evident that Barber's machine involved construc- 
 tive problems altogether unsolved in his time, but the appa- 
 ratus was surprisingly complete in its conception, including 
 combustion at constant pressure, with pumps for the supply 
 of air and fuel, together with the use of vapor of water for 
 the reduction of temperature, and a train of gear wheels 
 for the reduction of the speed. 
 
 FIG. 5. Fernihough's turbine, 1850. 
 
 Nothing seems to have been done for more than fifty 
 years after the patent of Barber, but in 1850 a mixed steam 
 and gas turbine was proposed by W. F. Fernihough, and 
 patented in Great Britain, No. 1328, of 1850. This appa- 
 ratus, Fig. 5, consisted of a chamber A, lined with refractory 
 material, and fitted with a grate B, on which the fuel was 
 
14 THE GAS TURBINE. 
 
 ignited. Air was supplied under pressure through E, while- 
 water was sprayed from above at H, and the mixture of 
 steam and the gases of combustion were delivered through 
 the nozzle I upon the wheel L, L. 
 
 In the mean time Burdin, in France, had proposed, in 
 1847, to make a hot-air turbine, using a multiple-wheel 
 rotary compressor to deliver air through a heating chamber 
 to a corresponding rotary motor. This plan was included 
 in the remarkable communication of Tournaire, presented 
 to the Academic des Sciences in 1853. The original memoir 
 of Tournaire is remarkable in many ways, both for the 
 breadth of its conception of the problem, and also because 
 it refers to " elastic fluid turbines," not limiting the action 
 to steam, but including hot air and gases, and thus distinctly 
 including the gas turbine. In view of the importance of 
 this communication it is here translated entire, from the 
 Compte Rendu des Seances de V Academic des Sciences of 
 March 28, 1853, pp. 588-593. 
 
 "APPLIED MECHANICS. Note upon multiple and succes- 
 sive-reaction turbine devices for the utilization of the motive 
 power developed by elastic fluids; by M. Tournaire, Ingenieur 
 des Mines. Commission: MM. Poncelet, Lame, Morin, 
 Combes, Seguier: 
 
 " Numerous attempts have been made to cause the vapor 
 of water or other gaseous substances to act by reaction upon 
 the blades or passages of rotative apparatus similar to tur- 
 bines or other hydraulic wheels; but down to the present 
 time these inventions have not been crowned with practical 
 success. The economical application of the principle of 
 reaction to machines operated by elastic fluids would never- 
 theless be of a very high degree of interest, since the moving 
 portions would thereby be reduced to very small dimensions, 
 and, in the great majority of cases, the transmission of the 
 motion would be lightened and simplified. In a word, such 
 machines would enable the same advantages to be realized 
 
THE GAS TURBINE. 15 
 
 as are found with hydraulic turbines compared with water- 
 wheels of large diameter. 
 
 " Elastic fluids acquire enormous velocities, even under 
 the influence of comparatively low pressures. In order to 
 utilize these pressures advantageously upon simple wheels 
 analogous to hydraulic turbines, it would be necessary to per- 
 mit a rotative motion of extraordinary rapidity, and to use 
 extremely minute orifices, even for a large expenditure of fluid. 
 These difficulties may be avoided by causing the steam or 
 gas to lose its pressure, either in a gradual and continuous 
 manner, or by successive fractions, making it react several 
 times upon the blades of conveniently arranged turbines. 
 
 1 'We must attribute the origin of the researches which 
 we have made upon this subject to the communications 
 which M. Burdin, Ingenieur en Chef des Mines, and membre 
 Correspondant de I'lnstitut, has had the courtesy to make 
 to us, and which go back to the close of 1847. M. Burdin, 
 who was then engaged upon a machine operated by hot 
 air, desired to discharge the compressed and heated fluid 
 upon a series of turbines fixed upon the same axis. Each 
 one of these wheels was placed in a closed chamber, the air 
 to be delivered through injector nozzles and discharged 
 at a very low velocity. The author proposed to compress 
 the cold air by means of a series of blowers arranged in a 
 similar manner. This idea of employing a number of suc- 
 cessive turbines in order to utilize the tension of the fluid 
 a number of times seemed to us a simple and fertile one; 
 we perceived in it the means of applying the principle of 
 reaction to steam and air engines. 
 
 "Since the differences in pressure, as used in steam 
 engines, are considerable, it became evident that a large 
 number of turbines would be required to give a sufficient 
 reduction in the velocity of the fluid jet. The lightness and 
 small dimensions of the moving parts permits of very high 
 rotative speeds compared with those of ordinary engines. 
 
16 THE GAS TURBINE. 
 
 " Notwithstanding the multiplicity of parts, it is essential 
 that the apparatus should be simple in its action, susceptible 
 of a high degree of precision, and that adjustments and 
 repairs should be readily made. We believe that we have 
 fulfilled these essential conditions by means of the following 
 arrangements : 
 
 "A machine is composed of several independent motor 
 axes, connected by means of pinions to a single wheel for 
 the transmission of the motion. Each of these axes carries 
 several turbines; these receive. and discharge the fluid at 
 the same distance from the axis. 
 
 " Between two turbines is placed a fixed ring of guide 
 blades. The guides receive the discharge from one reaction 
 wheel and give to it a direction and velocity suitable to 
 act upon the following wheel. Each of these systems of 
 fixed and moving organs is to be enclosed in a cylindrical 
 case. The guide blades will form portions of rings or annular 
 pieces placed in the fixed cylinder, and these should be fitted 
 very exactly the one to the other. The turbines should also 
 have the form of rings, and should be fitted to a sleeve 
 attached to the shaft. Projections fitting into grooves 
 secure the guides to the cylindrical case, and fasten the 
 turbines to the shaft. The first set of guides, which act 
 simply as injector nozzles, may be made in one solid piece, 
 carrying the journal of the shaft. Nothing could be easier 
 than to erect or dismount such an apparatus. In order to 
 transmit the motion it is necessary that the shaft should 
 pass through the end of the cylindrical case through an 
 opening fitted with a tight packing; such a single stuffing 
 box will answer for each series of reaction wheels. 
 
 " After having acted upon the turbines on the first shaft, 
 and thus parted with more or less of its elasticity, the fluid 
 is caused to act upon the turbines of the second series, and 
 so on. For this purpose large openings connect the end of 
 each case with the beginning of the one which follows. 
 
THE GAS TURBINE. 17 
 
 These cases and passages may form portions of the same 
 casting. Since the steam or gas expands in proportion to 
 its passage through the blades of the turbines and the 
 guides, it is necessary that these blades should offer pas- 
 sages of continually increasing size, and the last portions 
 of the apparatus will have much greater dimensions than 
 the first. 
 
 u As in the case of hydraulic reaction wheels, the last 
 turbine on each shaft should discharge the fluid with a very 
 low velocity. At its flow from the other turbines the fluid 
 should have a velocity best adapted to its entrance into the 
 passages between the guide blades. The motive power 
 developed by these wheels will be produced, in great part, 
 not by the extinction of the actual velocity of the fluid, but 
 from the differences in pressures in entering and leaving 
 the blades. This difference in pressures will produce a great 
 excess in the relative velocity of discharge over the relative 
 velocity of entrance, and, in order that this effect may be 
 obtained, it will suffice, by reason of the continuity of the 
 motion, for the orifices of discharge of the passages to be 
 of smaller area than the entrance orifices; this corresponds, 
 in general, to the arrangement in most hydraulic turbines. 
 Considered with regard to the relative velocity of rotation, 
 the velocity of flow through the passages of our turbines 
 will be much greater than in the passages in ordinary reac- 
 tion wheels, and, in consequence, they will be capable of 
 utilizing a much greater proportion of motive power. 
 
 u As is the case in all kinds of machinery, there are many 
 causes tending to dimmish the useful effect of our apparatus, 
 and to render it lower than the theoretical effect. 
 
 "One portion of the fluid will escape through the clear- 
 ance intervals which must be left between the fixed and 
 moving portions, and will have no effect upon the turbines 
 and will not be directed by the guide blades. There will be 
 produced shocks and eddies at the entrance and discharge 
 2 
 
18 THE GAS TURBINE. 
 
 of the buckets. The considerable friction, due to the narrow- 
 ness of the passages, will absorb a considerable portion of 
 the theoretical work. 
 
 "All these injurious effects are produced in hydraulic 
 turbines, some with an intensity almost equal in degree, 
 others, such as the frictional resistances, to a much less 
 extent. These reaction wheels are, nevertheless, excellent 
 machines. In order that our steam or hot-air machines 
 should equal them in respect to the effective power utilized, 
 a very perfect construction will be necessary, which it will 
 perhaps be difficult to attain, because of the small size of 
 the parts. But if we consider the results obtained with 
 piston engines operated by steam we see that we may make 
 a large allowance for losses before our turbines fail to give 
 equally good results. Many of the causes of loss inherent 
 in the use of pistons and cylinders will be avoided. Thus, 
 the cooling effect due to radiation from the exterior walls 
 in contact with the surrounding medium will become negli- 
 gible, since our cylindrical casings offer a very small mass 
 and volume, traversed by a very large flow of heat. 
 
 "In order that the application of our principles may be 
 successfully applied to engines operated by elastic fluids 
 it is necessary that great care and a very high degree of pre- 
 cision be given to the construction and erection of the parts, 
 and that the dimensions and curves of the blades be care- 
 fully studied. 
 
 "It is necessary that the teeth of the gear wheels, which 
 are operated at very high speeds, should run with great 
 smoothness, without shock or vibrations; the helicoidal 
 system of gearing of White will probably be found desirable. 
 The shafts should also be held by outside collars in order 
 that the metallic stuffing boxes may not be subjected to 
 heavy pressures. The journals will receive the pressure 
 parallel to the axis; this, however, will be small, on account 
 of the small dimensions of the turbines. 
 
THE GAS TURBINE. 19 
 
 11 As for the regulators of the flow of the fluid, their 
 functions will be performed by two slides or valves, one 
 placed in the pipe connecting the engine to the generator, 
 and the other in the opening through which the exhaust is 
 discharged into the atmosphere. 
 
 "The principal advantage offered by the motors which 
 we propose lies in the extreme lightness and small size which 
 they offer. This is a point upon which we believe it unneces- 
 sary to insist at length. The present engines are too heavy 
 and cumbersome, and are yet incapable of application to 
 many purposes which are still accomplished by the physical 
 effort of man. Without doubt the realization of our pro- 
 jects would extend widely the domain of mechanical power. 
 
 11 Applied to steam motors we believe that our multiple 
 turbines would permit a reduction in the dimensions of the 
 reservoirs or generators of the fluid; because, the consump- 
 tion of the motive material being continuous, the ebullition 
 will be effected very regularly in the boiler, and there will 
 be much less danger of the entrainment of a large proportion 
 of water. If hot air be substituted for steam, as we may 
 hope from the beautiful and fertile experiments of Ericsson, 
 our turbines will replace, very happily, the enormous cylin- 
 ders and pistons used by the Swedish engineer to receive 
 the action of the compressed air. It remains to be seen if 
 similar rotative apparatus may not be usefully employed 
 for the compression of cold air. In case of success a complete 
 mechanical revolution will be effected not only with regard 
 to the quantity of combustible consumed but also in the 
 matter, not less important, of the masses and volumes 
 which enter into machine construction." 
 
 It seems surprising that the clearly expressed ideas of 
 Tournaire failed of immediate realization, especially as 
 they were passed in review under the eyes of such a com- 
 mittee of mechanical specialists as Morin, Lame, and Ponce- 
 
20 THE GAS TURBINE. 
 
 let; but it is probable that constructive difficulties, the 
 extent of which was fully realized by Tournaire, stood in 
 the way. His work seemed to have been almost entirely 
 overlooked until recently, but there is no doubt that he 
 fully grasped the problem, as the text of his communication 
 to the French Academy shows. 
 
 At the present time the term " elastic-fluid" turbine 
 appears in nearly all patent specifications for such machines, 
 their scope not being limited to steam alone. Tournaire 
 not only used this very expression, but also foresaw the 
 application of the multiple-turbine principle to pressure 
 blowers as well. 
 
 He further saw that high fuel economy, while probably 
 attainable with the turbine, was not the only advantage, 
 but that material reduction in weight and in bulk might 
 also be attained, points which to-day are of even more 
 importance than they were fifty years ago. 
 
 An interesting forecast of the practicability of the gas 
 turbine appears in the fifth edition of Bourne's large treatise 
 on the steam engine, published in 1861. Discussing the 
 advantages of superheated steam, Mr. Bourne says: 
 
 " Steam of a high temperature will, therefore, be more 
 economical in its use than steam of a lower temperature, 
 and surcharged steam being much hotter than common 
 steam is consequently more advantageous. After all, how- 
 ever, the temperatures which it is possible to use with any 
 kind of steam in an engine are too low to render any very 
 important measure of economy possible by their instrumen- 
 tality. We are, therefore, driven to consider the applicability 
 of other agents, the most suitable of which appears to be air, 
 and this brings us back to the point from whence we started 
 at the commencement of the present chapter. Small meas- 
 ures of improvement are worth very little consideration 
 when great and important steps of progress are apparently 
 within our reach, and to us it appears quite clear that the prod- 
 
THE GAS TURBINE. 
 
 21 
 
 ucts of combustion may be employed to produce motive power, 
 not through the instrumentality of a cylinder and piston, but 
 rather by means of a turbine or an instrument like a smoke 
 jack or Barker's mill, and which may be made to work in 
 water or some other liquid. In this way very high tempera- 
 tures may be dealt with, and it is only by employing very 
 high temperatures that any very great step of improvement 
 is to be attained." 
 
 FIG. 6. Boulton's multiple jet system 
 
 In 1864 the problem of combustion at constant pressure, 
 in connection with the operation of a gas turbine, was inves- 
 tigated by M. P. W. Boulton, and his British patent, No. 
 1636 of 1864, contains some points of interest, in the light 
 of what has been done since. He realized that the high 
 velocity of the jet of gases issuing from the nozzle offered 
 a practical difficulty, and proposed to remedy this by the 
 use of successive induced jets of increasing volume and 
 consequently lower velocity. This is shown in Fig. 4, the 
 gases being delivered through the nozzle A, inducing a cur- 
 rent in B, and this again in C. The turbine is represented 
 at D, operated by the increased volume of fluid at the 
 reduced velocity. 
 
22 
 
 THE GAS TURBINE. 
 
 Another method proposed by Boulton for maintaining 
 combustion at constant pressure is shown in Fig. 7. The 
 gas is burned at A, in a chamber C, under water, the prod- 
 ucts of combustion passing up through the water between 
 the baffle plates E, E, and the mixed gases and steam being 
 delivered to the turbine from the top of the chamber B. 
 
 FIG. 7. Boulton's constant-pressure combustion chamber. 
 
 The idea of combustion at constant pressure to furnish 
 an elastic fluid composed of hot air and products of com- 
 bustion for use in a turbine appears to have occupied the 
 attention of a number of engineers from 1870 onward. John 
 Bourne, the well-known British engineer and writer on the 
 steam engine, took out two patents, one in 1869 and the 
 other in 1870, relating to the combustion of coal dust for 
 the production of gases for use in a turbine. His plans 
 included the dilution of the gases with air and with the vapor 
 of water, and involved the use of high pressures, up to 1,000 
 pounds per square inch. Bourne's patents refer entirely 
 

FIG. 9. The Stolze hot-air turbine. 
 
THE GAS TURBINE. 
 
 23 
 
24 THE GAS TURBINE. 
 
 to the production of the working fluid, and do not give any 
 details of the turbine which he proposed to use. 
 
 Another British patent of about the same time is that 
 of James Anderson, this including the combustion of a mix- 
 ture of gas and air in the combination chamber or channel, 
 the gases resulting from the combustion being led into a 
 reaction turbine. He also proposed to make the combina- 
 tion chamber in the arms of the turbine itself. 
 
 It does not appear that any of these plans were ever put 
 into actual operation. In 1872, however, we find that Dr. 
 F. Stolze, of Charlottenburg, near Berlin, applied for a 
 patent from the Prussian Government for a so-called "fire 
 turbine, " this practically being the same as the machine 
 experimented upon by Burdin in 1847 and described by 
 Tournaire in his communication to the French Academy 
 of Sciences. The general scheme of the Stolze turbine is 
 shown in Fig. 8, there being a multiple turbine compressor 
 and a multiple power turbine on the same shaft, the com- 
 pressed air being passed through a heating chamber and 
 thus deriving energy from the heat of the fuel before pass- 
 ing to the power turbine. The exterior of the Stolze turbine 
 is shown in Fig. 9, this representing his experimental 
 machine at Charlottenburg. 
 
 The early work of the Hon. C. A. Parsons is generally 
 supposed to have related wholly to the steam turbine, but 
 in his original patent of 1884 (British Patent No. 6735) the 
 following reference to the gas turbine occurs : 
 
 " Motors, according to my invention, are applicable to a 
 variety of purposes, and if such an apparatus be driven, it 
 becomes a pump arid can be used for actuating a fluid column 
 or producing pressure in a fluid. Such a fluid pressure- 
 producer can be combined with a multiple motor, according 
 to my invention, to obtain motive power from fuel or com- 
 bustible gases of any kind. For this purpose I employ the 
 pressure-producer to force air or combustible gases into a 
 
THE GAS TURBINE. 
 
 25 
 
 furnace into which there may or may not be introduced 
 other fuel (liquid or solid). From the furnace the products 
 of combustion can be led in a heated state to the multiple 
 motor which they actuate. 'Conveniently, the pressure-pro- 
 ducer and multiple motor can be mounted on the same shaft, 
 the former to be driven by the latter; but I do not confine 
 myself to this arrangement of parts. In some cases I employ 
 water or other fluid to cool the blades, either by conduction 
 of heat through their roots or by other suitable arrangement 
 to effect their protection." 
 
 FIG. 10. Combustion nozzles of De Laval gas turbine, 1893. 
 
 In 1893 De Laval proposed to deliver compressed air 
 into a combustion chamber into which a liquid fuel was 
 sprayed, the products of combustion being directed upon 
 the blades of a wheel similar to that of the steam turbine 
 known by his name. The general arrangement is shown 
 in Fig. 10. The compressed air enters at a and the sprayed 
 combustible at b, the combustion taking place in the space 
 B. At c provision is made for an injection of water if neces- 
 sary, the gaseous products passing through the nozzle C to 
 the wheel D. 
 
26 THE GAS TURBINE. 
 
 The first patent of M. Charles Lemale was taken out in 
 1901, followed in 1903 by a more complete development of 
 the combustion chamber and expansion nozzle. M. Lemale, 
 in conjunction with the late M. Rene Armengaud, continued 
 to experiment with the gas turbine, under the auspices of 
 the Societe des Turbomoteurs, and the results of this work 
 will be given hereafter at length. 
 
 In the United States Dr. Sanford A. Moss published, in 
 1903, a discussion of the subject of the gas turbine, in the 
 form of a thesis presented to the faculty of Cornell Univer- 
 sity, this containing an examination of the thermodynamics 
 of the gas turbine and a brief account of some experimental 
 work. 
 
 The question has been discussed from a theoretical view- 
 point by Mr. R. M. Neilson in a paper presented before the 
 Institution of Mechanical Engineers in October, 1904, which 
 with the discussion it evoked will be given in a following 
 chapter. 
 
 It was also very fully examined by members of the Societe 
 des Ingenieurs Civils de France in consequence of an important 
 paper by M. L. Sekutowicz, presented at the session of Feb- 
 ruary 2, 1906, the discussion being taken up by MM. J. 
 Deschamps, Rene Armengaud, Jean Rey, G. Hart, L. 
 Letombe, and A. Bochet. M. Armengaud presented an 
 important paper upon the subject before the Mechanical 
 Section of the International Engineering Congress at Liege 
 in June, 1905, this having been revised for publication in 
 Cassier's Magazine for January, 1907. 
 
 In the Schweizerische Bauzeitung for August 27, 1904 
 there appeared an analysis of the action of the Armengaud 
 and Lemale gas turbine by Alfred Barbezat, while two papers 
 by Dr. Charles E. Lucke in the Engineering Magazine of 
 April, 1905, and August, 1906, and one by Professor Sidney 
 A. Reeve in the same magazine for June, 1905, formed cur- 
 rent contributions to the theory of the subject. 
 
THE GAS TURBINE. 27 
 
 An elaborate investigation of the practicability of the 
 gas turbine was published in the Zeitschrift fur das Gesamte 
 Turbinenwesen by A. Baumann, of Zwickau, this appearing 
 in the issues between December 15, 1905, and May 20, 1906. 
 Several pamphlets upon the subject have appeared in Ger- 
 many, among which may be mentioned : Studien zur Frage 
 der Gas-Turbine (Studies upon the Question of the Gas Tur- 
 bine), by Rudolf Barkow; Ein Praktisch Brauchbare Gas- 
 Turbine (A Practical, Useful Gas Turbine), by Dr. Richard 
 Wegener; and Die Aussischten der Gas Turbine (The Out- 
 look for the Gas Turbine), by Felix Langen. 
 
 The most important work from a theoretical point of 
 view is given in the discussions before the Institution of 
 Mechanical Engineers, in London, and the Society of Civil 
 Engineers of France, and these are given practically entire, 
 followed by abstracts of other papers, and as much informa- 
 tion concerning actual machines as can at present be made 
 public. 
 
CHAPTER II. 
 
 THE DISCUSSION BEFORE THE INSTITUTION OF 
 MECHANICAL ENGINEERS. 
 
 ON October 21, 1904, Mr. R. M. Neilson, Associate Mem- 
 ber of the Institution of Mechanical Engineers read before 
 the Institution at its house in London, a paper entitled: 
 "A Scientific Investigation into the Possibilities of Gas 
 Turbines." By the kind permission of the Council of the 
 Institution this paper is here given entire, together with 
 the discussion which it elicited from the membership, this 
 forming one of the most important contributions to the 
 question which has yet appeared in England. 
 
 In examining this paper and the discussion upon it, it 
 must be remembered that at the time of its presentation, 
 1904, the investigations of Dr. Charles E. Lucke upon tem- 
 peratures and pressures in free expansion of hot gases in 
 nozzles had not yet been made public, nor had the work of 
 Professor Rateau in the construction of turbine air com- 
 pressors of high efficiency been completed. 
 
 A SCIENTIFIC INVESTIGATION INTO THE POSSI- 
 BILITIES OF GAS TURBINES 
 
 BY MR. R. M. NEILSON 
 
 Associate Member, Institution of Mechanical Engineers. 
 
 A prophecy expressed frequently in engineering circles at 
 the present day is that turbines actuated by hot gases other 
 than steam will eventually come to the front as prime mov- 
 ers. The idea of employing hot gases (other than steam) 
 to drive turbines is by no means new; but the success of the 
 steam turbine has recently brought the question into prom- 
 
 28 
 
_ THE GAS TURBINE. _ 29 
 
 inence. Although the subject is interesting and important, 
 and although many minds seem to be considering it, there 
 appears to be hardly any literature on the subject, except 
 that which is found in patent records. 
 
 There is no doubt that many persons speak of the advan- 
 tages of gas turbines without duly considering the difficul- 
 ties to be encountered. There are probably many others 
 who have valuable ideas on the subject, supported in some 
 cases by experimental data, but who are apt to let their 
 thoughts run in a groove and to consider (rightly or 
 wrongly) that the only possible solution of the gas turbine 
 problem lies in the particular direction in which they are 
 working. 
 
 This Paper is written with the object of expressing and 
 and comparing as concisely as possible the advantages and 
 possibilities of gas turbines worked on different cycles, and 
 the difficulties to be overcome to make these turbines a suc- 
 cess. A further and more important object is to draw opin- 
 ions from other engineers who have studied the question, 
 and especially from those who have conducted experiments. 
 If these objects be obtained, even in an imperfect manner, 
 the author believes that a foundation of knowledge will 
 be obtained and placed on record, which will be of consider- 
 able use to engineers who may be endeavoring or about to 
 endeavor to produce practical machines. 
 
 Carnot's formula for the efficiency of an ideal heat engine 
 
 
 is well known, but its real meaning is sometimes forgotten; 
 and it may not be out of place here to put in a reminder 
 that, in Carnot's cycle, all the heat is put in at tempera- 
 ture T 1 and all the heat withdrawn at temperature T 2 . An 
 increase in the range of temperature does not necessarily 
 cause a thermodynamic gain, and it is possible largely to 
 
30 THE GAS TURBINE. 
 
 increase the range of temperature (as for example by super- 
 heating steam before use in a steam engine) without ther- 
 modynamically increasing the efficiency by more than a, 
 small percentage. 
 
 The greatest possible efficiency of a gas engine (recipro- 
 cating or turbine) working on Carnot's cycle between the 
 limits of temperature 1600 C. (2912 F.) and 17 C., will be 
 found to be: 
 
 (1600 + 273)- (17 + 273) nQ 
 1600 + 273 
 
 If the gas engine be an explosion motor with compression 
 to 60 pounds per square inch above atmosphere, combustion 
 at constant volume, and expansion to atmospheric pressure, 
 the greatest possible efficiency between the same limits of 
 temperature is only 0.50; and, in the engine work on the 
 ordinary Otto cycle with the same compression and between 
 the same limits of temperature, the greatest possible efficiency 
 is only 0.37. 
 
 Efforts must therefore be made not so much to get the 
 maximum and minimum temperatures respectively as high 
 and as low as possible, but to get the mean temperature at 
 which heat is given to the gas and the mean temperature at 
 which heat is withdrawn from it respectively as high and as 
 low as possible. Of these two temperatures the lower one 
 is usually by far the more important. An ideal gas engine 
 working on Carnot's cycle between the limits of temperature 
 2000 C. (3632 F.) absolute and 300 C. (572 F.) absolute 
 will lose as much by an increase of 100 C. to the lower 
 temperature as it will by a decrease of 500 C. from the 
 higher temperature. 
 
 Coming now to discuss more particularly gas turbines, 
 there are four cycles on which it seems to the author that 
 these could be worked with the possibility of good results. 
 Two of these are what Mr. Dugald Clerk designates Type 2 
 
THE GAS TURBINE. 
 
 31 
 
 and Type 3.* The author will call them respectively 
 Cycle I and Cycle II. 
 
 It has not been considered worth while to discuss the Car- 
 not cycle at length, but a few remarks are made about it 
 towards the end of the Paper (page 74). 
 
 Y B, B B 2 r 
 
 O A 
 
 FIG. 11. Pressure-volume diagram. 
 
 A pressure-volume diagram of an engine working on 
 Cycle I is shown in Fig. 11, and an entropy-temperature 
 diagram in Fig. 12. 
 
 V--- 
 
 b ; a /id 
 
 FIG. 12. Entropy-temperature diagram. 
 
 The working fluid is compressed adiabatically from A to 
 B. Heat is then supplied by combustion at constant pres- 
 sure from B to C; the gas expands adiabatically from C to D, 
 and the fluid is then cooled at constant pressure from D to 
 A. Reciprocating gas engines have been worked on this 
 cycle by Brayton and others, but have never come into com- 
 mon use. (The Diesel engine may be considered to belong 
 
 * "The Gas and Oil Engine," by Dugald Clerk (Longmans & Co.), Chap- 
 ter III. 
 
32 THE GAS TURBINE. 
 
 to this class, although no decided constant-pressure line is 
 discernible on indicator diagrams taken from the engine.) 
 One great difficulty that has been experienced in working 
 reciprocating engines on this cycle is that of getting com- 
 plete combustion during the period B C without the charge 
 occasionally firing back. If the air and fuel are brought into 
 contact only on entering the cylinder, it is difficult to get good 
 combustion during the period B C. If, on the other hand, the 
 air and fuel are previously mixed together, it is difficult 
 to prevent occasional firing back. Of course the chamber 
 in which the air and fuel are mixed may be made strong 
 enough to stand explosions; but any back firing upsets the 
 regular working of the engine and is otherwise objection- 
 able. 
 
 It has been proposed for gas turbines to cause air and fuel 
 to unite in a nozzle, which thereafter diverges, the idea 
 being that the air and fuel will combine on meeting each 
 other, and the hot products of combustion will then acquire 
 a high velocity in the divergent nozzle with which velocity 
 they will enter the turbine buckets. The results of a trial 
 of such a scheme would be interesting. The author doubts 
 if the combustion would be quick enough to give a good 
 efficiency. If, however a combustion chamber of ample 
 size were provided in which the burning gases could rest a 
 short interval before passing to the turbine, better results 
 could, in the author's opinion, be expected. The air and 
 fuel would be separately pumped into the chamber from 
 which the products of combustion would flow continuously 
 and uniformly by one or more passages into the turbine. 
 
 At any rate the difficulties should not be as great with 
 turbines working on this cycle as with reciprocating engines, 
 as the latter have to receive the hot gases intermittently, 
 while the turbine receives a continuous flow. This is an 
 important point as regards controlling the flame. With an 
 engine of the Brayton type the fuel has to be ignited in the 
 
THE GAS TURBINE. 33 
 
 cylinder for every working stroke, and the supply of gas to 
 the flame has to be cut off for every working stroke. With 
 a turbine the fuel and air could be supplied at a constant 
 velocity to the flame and a steady flame maintained without 
 interruptions. This is important, because, if a mixture of 
 air and fuel be always supplied to the flame with a velocity 
 greater than the velocity or propagation of the flame, there 
 can of course be no firing back, and this result can be ob- 
 tained without the use of a wire-gauze screen. The main- 
 taining of this velocity of supply to the flame above the re- 
 quired minimum when starting and stopping the motor, 
 and when running at low powers, is of course a problem 
 to be considered, and some consideration is given to it later 
 on (pages 77 and 78). The strength of the mixture of air 
 and fuel should be kept constant. The power of the tur- 
 bine can be varied by other means, which will be referred to 
 later (pages 77 and 78). It must be noted that if the air 
 and fuel are compressed adiabatically to a sufficient extent, 
 which depends on the nature of the fuel, combustion will 
 occur immediately the two are brought into contact with 
 each other. It is therefore necessary in such cases to keep 
 the air and fuel apart until the instant when combustion is 
 desired. It must also be noted that with a turbine there 
 will be no hot waste gases mixed with the fresh air and gas to 
 be compressed. 
 
 This cycle allows of a fairly high ideal efficiency being 
 obtained with a moderate maximum temperature. Now a 
 moderate maximum temperature is of the utmost impor- 
 tance in the case of a turbine of the Parsons type. A Par- 
 sons turbine with steel blades could probably be designed 
 without any great difficulty to stand a temperature of about 
 700 C. (1292 F.) without any water jacketing or cooling 
 devices of any sort (except for the bearings). With temper- 
 atures above this, the blades would need to be cooled. 
 This would necessitate a radical alteration in design. The 
 3 
 
34 THE GAS TURBINE. 
 
 question of designing a turbine to stand high temperatures 
 will be considered later on. It is only desired here to point 
 out that great difficulties with a certain class of turbine are 
 avoided by keeping the maximum temperature moderate. 
 The cycle under consideration may therefore have great 
 advantages for turbines. 
 
 It had better be stated here that the author has made 
 several assumptions with regard to the working fluid or 
 fluids. These assumptions are as follows: 
 
 1. That the specific heats of gases dealt with are con- 
 stant at all temperatures and pressures, and are as follows:- 
 
 Specific heat at constant pressure or Kp =0.238. 
 Specific heat at constant volume or Kv=0.17 
 
 2. That weight per cubic foot of gases dealt with = 
 0.0777 pounds at a pressure of 15 pounds per square inch 
 absolute and a temperature of 17 C. 
 
 3. That PF = a constant for all pressures and temper- 
 atures. 
 
 4. That PF = a constant for isothermal expansion and 
 compression at all temperatures and pressures. 
 
 5. That combustion produces no change of volume ex- 
 cept that due to change of temperature. 
 
 Some of these assumptions will probably be appreciably 
 inaccurate in certain cases ; but it seemed advisable to sacri- 
 fice something for simplicity and uniformity. As regards 
 the variability of the specific heats, it has . been thought 
 better to assume constancy until more knowledge on the 
 subject has been obtained and a scale of change (if any) 
 has been agreed upon. 
 
 Pressures have been reckoned in pounds per square inch, 
 and temperatures have generally been reckoned on the 
 Centigrade scale, although for convenience the corresponding 
 readings on the Fahrenheit scale have also been given. The 
 numbers on the diagrams representing pressure and temper- 
 ature are all representative of absolute pressures in pounds 
 
THE GAS TURBINE. 35 
 
 per square inch, and temperatures on the absolute Centi- 
 grade scale. 
 
 Referring to Fig. 12 (page 31), the heat absorbed by the 
 fluid is represented in this figure by the area aBCd, and the 
 heat abstracted or discarded by the area aADd. The heat 
 converted into work is represented by the area ABCDj and 
 consequently, if E represents the ideal efficiency of an engine 
 working on this cycle, 
 
 area ABCD 
 
 E = 
 
 area aBCd 
 
 AT * u i * ^ AB DC V r ^ 
 
 Now, as it can be proved* that ~ a ^~ == ~^r~ == where pqr 
 
 is any ordinate cutting the lines ad, AD, and BC, which are 
 all constant-pressure lines, 
 
 v AB DC 
 therefore tf____. (1) 
 
 Let t represent the temperature before compression. 
 Let Z c represent the temperature at the end of compression. 
 Let T represent the temperature at the end of combustion. 
 Let T l represent the temperature at the end of adiabatic 
 expansion. 
 
 * Since all vertical lines represent adiabatic expansion, therefore, by the 
 laws of adiabatic expansion, 
 
 7-1 
 
 temp, at AT press, at A~\ 
 temp. at B |_ press, at B J 
 
 where 7= 
 
 Kv 
 
 7-1 
 
 Similarly temp, at q _ I" press, at q "I y 
 
 temp, at r [_ press, at r J 
 
 But press, at A = press, at q, since AqD is a constant-pressure line; and press. 
 at B = press, at r, since BrC is a constant-pressure line, 
 
 therefore temp, at A_ temp, at q 
 
 temp, at B temp, at r 
 
 therefore AB = DC _qr 
 
 aB dC pr 
 
36 THE GAS TURBINE. 
 
 Then, from equation (1) and referring to Fig. 12, 
 
 This can be proved quite well without any entropy- 
 temperature diagram.* The diagram, however, shows the 
 efficiency better. 
 
 It is important to consider the amount of negative work 
 done and the ratio of this to the total or gross work. The 
 negative work is the work of compressing the gas and de- 
 livering it in its compressed state. It is true that with 
 some engines there is no work of delivery. In a reciprocat- 
 ing gas engine in which the gas is compressed in the motor 
 cylinder, the only negative work (ideally) is that of com- 
 pressing the charge; and, even when a separate cylinder is 
 used for the compression, the work of delivering might be 
 avoided. With a turbine, however, the fluid cannot be com- 
 pressed in the motor; and, whatever arrangement is adopted, 
 the compressed fluid will have to be delivered after compres- 
 sion. The author has, therefore, considered it better in all 
 cases to include in the negative work the amount required 
 to deliver the compressed gas. The motor proper of course 
 gets the benefit of this work. 
 
 In Fig. 11 (page 31) the work to compress the gas is 
 represented by the area AbB, and the work to deliver it in 
 compressed state by the area yYBb. The total negative 
 work is therefore represented by the area yYBA. The gross 
 work of the motor is represented by the area yYCD, of which 
 the part yYBb represents the work done before expansion, 
 and the part bBCD the work done during expansion. By 
 deducting the negative work from the gross work the net 
 work is obtained; this is represented by the area ABCD. 
 This net work is the same as that represented on the en- 
 tropy-temperature diagram, Fig. 12 (page 31), by the area 
 ABCD. 
 
 * See "The Gas and Oil Engine," by Dugald Clerk, pages 46-48. 
 
THE GAS TURBINE. 37 
 
 Cycle I, Case 1. 
 
 If the gas is required to be used in a Parsons turbine with- 
 out cooling arrangements the maximum temperature must 
 not exceed 700 C. (1292 F.). A case with this maximum 
 temperature will now be considered : 
 
 In all cases 
 
 Let t and p represent respectively absolute temperature C. and absolute pres- 
 sure pounds per square inch before compression. 
 
 Let t c and p represent respectively absolute temperature C. and absolute pres- 
 sure pounds per square inch after compression. 
 
 Let T and P represent respectively absolute temperature C. and absolute pres- 
 sure pounds per square inch after combustion. 
 
 Let T l and P, represent respectively absolute temperature C. and absolute 
 pressure pounds per square inch after expansion to atmospheric pressure. 
 
 Let v represent one cubic foot of the fluid at temperature t and pressure p. 
 
 Let v c , V and V, represent the volume of the same at t c , p c ', T, P, and T lt 
 P 1 respectively. 
 
 Suppose that in all cases t = 17 C. (290 absolute C.) and 
 the corresponding pressure = 15 pounds absolute. First 
 by compressing to 42 pounds absolute: t c will then be 389 
 absolute C. This compression is shown by the line AB on 
 the pressure-volume diagram, Fig. 13 (page 38), and on the 
 entropy-temperature diagram, Fig. 14. 
 
 Let heat be supplied and the gas expand at constant pres- 
 sure along the line BC till the temperature is 973 absolute 
 C. Let the gas expand adiabatically along the line CD till 
 the pressure falls to 15 pounds absolute. The fluid is then 
 exhausted into atmosphere, and as the new charge is taken 
 at the same pressure and at temperature t, it can be assumed 
 that the discharged gas is cooled at constant pressure and 
 used over again. Both diagrams can therefore be com- 
 pleted by the constant-pressure line DA. 
 
 The heat absorbed by the fluid is represented by the area 
 aBCd in Fig. 14, and the heat rejected by the area aADd. 
 The heat converted into work is represented by the area 
 ABCD and 
 
 area ABCD ^ c - * = 389 - 290 = 99 
 ~ t c 389 ~389 
 
38 
 
 THE GAS TURBINE. 
 
 The negative work is represented in Fig. 13 by the area 
 yYBA, the gross work by the area yYCD, and the net work 
 by the area ABCD, 
 
 negative work _ area yYBA 
 gross work area yYCD 
 
 therefore 
 
 0.4. 
 
 f 
 
 \ 2 3 
 
 Volume 
 
 FIG. 13. Cycle I, Case 1. Pressure-volume diagram. 
 
 The expansion line is carried right down to atmosphere. 
 It should be possible in practice without difficulty to do this 
 very nearly in a turbine, although the volume at D is 2} 
 times the volume at A. In dealing with large volumes and 
 
 C T-973 
 
 T-725 
 
 \\\\\\\\\\\\\\\\\\\\\\\ 
 
 FIG. 14. Cycle I, Case 1. Entropy-temperature diagram. 
 
 small pressures there is an immense difference between tur- 
 bines and reciprocating engines. Reciprocating engines re- 
 quire large cylinders. These large cylinders, besides being 
 objectionable on account of bulk and cost, necessitate great 
 frictional losses. The low pressure dealt with is of little 
 import as regards friction, which will be nearly the same 
 whether the pressure is 13 pounds below atmosphere or 
 13 pounds above atmosphere. With a turbine, however. 
 
THE GAS TURBINE. 39 
 
 the large volume of the fluid does not necessitate such a 
 bulky machine. Moreover in a turbine the friction depends 
 on the pressure. With high pressures the friction is great, 
 with low pressures very small. (In marine propulsion by 
 steam turbines it is not considered worth while uncoupling 
 the reversing turbines when the vessel is going ahead. 
 These turbines are allowed to rotate (above their normal 
 speed) in the low pressure which exists at the exhaust ends 
 of the main low-pressure turbines. 
 
 Cycle I, Case 2. 
 
 700 C. (1292 F.) must not, however, be considered as 
 the limiting temperature for gas turbines. Much higher 
 temperature can be employed if water-cooling or other 
 cooling arrangements be used. Mr. Parsons has circulated 
 steam for heating purposes through passages formed in the 
 rings supporting the fixed blades of his radial-flow steam 
 turbines.* Water could as easily be circulated, and there 
 should be no great difficulty in passing the water also through 
 the rings supporting the moving blades. 
 
 It has been proposed by Mr. Parsons and others to circu- 
 late water or other cooling fluid through the actual blades 
 of a turbine, these being formed hollow. It has also been 
 proposed to keep the blades of a single-wheel turbine cool 
 by causing the actuating fluid to act only at one point of the 
 circumference of the wheel, while a cooling fluid is projected 
 onto the blades at another point. 
 
 By the employment of cooling devices a turbine might 
 possibly be made to stand a temperature of 1500 C. 
 (2732 F.) or even 2000 C. (3632 F.). 2000 C. is a very 
 high temperature, and there would be great difficulty in de- 
 vising and constructing cooling arrangements which would 
 keep the blades in good working order when acted on 
 
 * " The Steam Turbine," by R. M. Neilson (Longmans and Co.) , pp. 43-45. 
 
40 THE GAS TURBINE. 
 
 by gas at a temperature approaching this. Let it be as- 
 sumed, however, that 2000 C. is allowable for the maximum 
 temperature; then, if the same compression is kept as in 
 Case 1, the ideal pressure- volume and entropy-temperature 
 diagrams will be as shown in Figs. 15 and 16. In these 
 Figs, the line CD has been reproduced from Figs. 13 and 14 
 (page 38), and is shown in dotted lines in order that the 
 two cases may be readily compared. 
 
 5-845 
 5 V 
 
 FIG. 15. Cycle I, Case 2. Pressure-volume diagram. 
 
 Referring to Fig. 16 the heat absorbed by the fluid is rep- 
 resented by the area aBEf, the heat rejected by the area 
 aAFf, and the heat converted into work by the area ABEF. 
 
 Therefore P 1 ~~~ tc * 
 
 area aBEj 
 
 389-290 
 389 
 
 0.25. 
 
 The increase in the maximum temperature has, therefore, 
 added nothing to the efficiency, and this will always be the 
 case if the initial temperature and pressure are unchanged 
 and compression is made to the same amount. That is to 
 say, as long as the constant pressure lines are started from 
 the same points, A and B, they can be extended any dis- 
 tance to the right and connected by any adiabatic line; 
 E will remain unchanged. In Fig. 16 the additional area 
 dCEf is divided by the line DF in the same ratio as the orig- 
 inal area aBCd is divided by the line AD. 
 
 The negative work (in Case 2) is represented in Fig. 15 
 by the area yYBA', it is the same as in the last case. The 
 
THE GAS TURBINE. 
 
 41 
 
 gross work is represented by the area yYEF, and the net 
 work by the area ABEF, 
 
 Therefore 
 
 negative work area yYBA 
 
 ^ = 0.171. 
 
 gross work area yYEF 
 The ratio of negative work to gross work has, therefore, 
 
 been very considerably diminished. 
 
 E T-2273 
 
 TH695 
 
 a 
 
 FIG. 16. Cycle I, Case 2. Entropy-temperature diagram. 
 
 Cycle I, Case 3. 
 
 In Case 1 it was necessary to have a low compression be- 
 cause a high compression with a maximum temperature of 
 only 700 C. (1292 F.) would have given an impractically 
 high value to the ratio of negative work to gross work. In 
 
42 
 
 THE GAS TURBINE. 
 
 fact this ratio was high even with the low compression 
 adopted. 
 
 With the maximum temperature raised to 2000 C. 
 (3632 F.), however, a much higher compression can be 
 adopted. Suppose a compression to 300 pounds per square 
 inch absolute is adopted. This will make t c 682.5 absolute 
 C. (1260.5 F.). The pressure-volume and entropy-tem- 
 perature diagrams will then be as shown in Figs. 17 and 18. 
 
 FIG. 17. Cycle I, Case 3. Pressuie-volume diagram. 
 
 Referring to Fig. 18 it is seen that 
 
 area AGHK 
 
 E 
 
 area aGHk 
 L-t 682.5-290 
 
 682.5 
 
 = 0.58 
 
 which is much better than (more than double) that in Cases 
 1 and 2. There is, however, the inconvenience of a high com- 
 pression, and compared with Case 1 more heat is likely to be 
 lost through radiation owing to the higher average temper- 
 ature. This question of radiation will be more or less impor- 
 tant according to the type of turbine. 
 
 The negative work is represented in Fig. 17 by the area 
 
THE GAS TURBINE. 
 
 43 
 
 zZGA, the gross work by the area zZHK, and the net work 
 by the area AGHK. 
 
 ,, . negative work area zZGA 
 
 Therefore . = , 7 w =0.3. 
 
 gross work 
 
 2213 
 
 =6825 
 
 FIG. 18. Cycle I, Case 3. Entropy-temperature diagram. 
 
 Cycle I, Case 4. 
 
 It will be interesting to find what efficiency can be ob- 
 tained with a maximum temperature of 2000 C. (3632 F.) 
 by increasing the compression till the ratio of negative work 
 to gross work is 0.4 the same as in Case 1. This ratio will 
 be attained when t c = 909 absolute C. (1668 F.), which 
 corresponds to a pressure of 818 pounds absolute. 
 
 Then 
 
 
 
 E= 909 
 
44 
 
 THE GAS TURBINE. 
 
 The pressure-volume and entropy-temperature diagrams 
 for this case are given in Figs. 19 and 20. 
 
 The line BG is shown dotted on Fig. 20 to allow Case 4 
 to be compared with Case 1. 
 
 Volume 
 FIG. 19. Cycle I, Case 4. Pressure- volume diagram. 
 
 The sharp corner at M would likely be rounded off in 
 practice. This would reduce the efficiency slightly. It 
 would also, however, reduce the maximum temperature, and 
 for this reason it might be advantageous in some cases to 
 round off the corner intentionally. 
 
THE GAS TURBINE. 
 
 45 
 
 In every case it has been assumed that the compression is 
 adiabatic; it is usually important that it should be at least 
 nearly so. If, for example, in Figs. 11 and 12 (page 31) 
 the compression, instead of being along the adiabatic AB 
 had been along the line AB l} which is below the adiabatic 
 line, that is, if heat had been allowed to escape during the 
 
 M T*2273 
 
 fc-909 
 
 T.-725 
 
 a 
 
 FIG. 20. Cycle I, Case 4. Entropy-temperature diagram. 
 
 compression, the heat absorbed by the fluid for the same 
 value of T would have been increased by the area b^B^Ba in 
 Fig. 12, while the heat converted into work would have 
 been increased only by the relatively small area AB^B. 
 E would, therefore, have been reduced. 
 
 If on the other hand the compression had been along the 
 line AB 2 , which is above the adiabatic, that is, if heat had 
 
 
46 THE GAS TURBINE. 
 
 been put into the fluid during compression, the heat ab- 
 sorbed and the heat converted into work would both have 
 been reduced by the same amount, namely, the area ABB 2 . 
 E would, therefore, obviously be reduced in this case also, 
 assuming that the heat put into the fluid during compression 
 is obtained by the combustion of fuel. 
 
 If, however, the heat put into the fluid during compression 
 is obtained for nothing if, for example, it is heat that would 
 otherwise be radiated away or carried away by convexion 
 the effect on E is not obvious. 
 
 A compression along the line AB 2 , Figs. 11 and 12, will 
 give a higher value to E than a compression along the line 
 AB, if the heat absorbed during the compression AB 2 is got 
 for nothing, and if the two cases are otherwise the same; 
 but a compression along the line AB 2 produces a higher ratio 
 of negative work to gross work. This will be clear from Fig. 
 11. Now with this ratio of negative work to gross work, 
 a still higher efficiency could be obtained by keeping the com- 
 pression adiabatic and continuing it further. A hot com- 
 pression, such as along the line AB 2J when the heat is got for 
 nothing, may be advantageous in a few cases, viz., T 1 is low 
 compared with t c ] but generally such a compression will be 
 harmful. 
 
 It is, in general, disadvantageous to heat the air or fuel 
 before compression, no matter what be the source of heat. 
 
 If gas is allowed to enter a water-cooled turbine at a high 
 temperature, such as 2000 C. (3632 F.), there will neces- 
 sarily be a great amount of heat carried away by the water. 
 In a reciprocating engine the metal surface with which the 
 gas comes into contact is very small compared with that in a 
 multiple-expansion turbine; and in a reciprocating engine 
 the bulk of the gas may expand and fall from its maximum 
 temperature to the temperature at exhaust without ever 
 coming near a metal surface. In a multiple-expansion tur- 
 bine, on the other hand, every particle of gas must practi- 
 
THE GAS TURBINE. 47 
 
 cally slide along a metal surface immediately it comes to the 
 first ring of blades. With turbines employing gas which 
 enters the turbine casing at such a temperature, the heat 
 lost through the walls and carried away by the water must 
 necessarily be very great indeed. It is true that the metal 
 surface in contact with the gas can be allowed to be at a 
 much higher temperature than the inside of the cylinder 
 walls of a reciprocating engine; but, in spite of this, the 
 heat lost through the walls and carried away by the cooling 
 water (or other cooling medium) will probably be much 
 greater with a turbine actuated by gas entering the turbine 
 casing at about 2000 C. than in a reciprocating engine in 
 which the maximum temperature is 2000 C. This loss of 
 heat will cause the actual work done by the engine to be very 
 much below the ideal. This is not only important in itself, 
 but, as will be explained subsequently (pages 50 and 51), 
 it prevents useful employment of a high ratio of negative 
 work to gross work. The question of utilizing this lost heat 
 will be discussed later on pages 59 to 66. 
 
 Cycle I, Case 3a. 
 
 Instead of employing cooling arrangements for the metal, 
 some or all of the available heat energy of the gas can be 
 converted into kinetic energy before causing it to act on the 
 turbine, so that the latter is not exposed to an unduly high 
 temperature. This can be done by allowing the gas, when 
 at the maximum temperature, to expand in a divergent 
 nozzle till its temperature falls to a degree that the turbine 
 can stand. More than one nozzle can be employed, but, to 
 reduce the radiation losses, the nozzles should be large and 
 few in number. 
 
 Suppose that the gas is compressed adiabatically to 300 
 pounds absolute, and then is heated at constant pressure to 
 a temperature of 2273 absolute C. (4132 F.), as in Case 3. 
 If now the gas be allowed to expand in a suitable nozzle, 
 
48 THE GAS TURBINE. 
 
 adiabatic expansion can be obtained; and if this be continued 
 till the pressure falls to 15 pounds absolute the temperature 
 will be 966 absolute C. (693 C.). This is just below the 
 temperature which was fixed on as a maximum for a turbine 
 without artificial cooling. The entropy-temperature diagram 
 will be the same as in Case 3, Fig. 18 (page 43) , and E 
 will therefore be the same, namely 0.58. The ratio of nega- 
 tive work to gross work will also be the same as in Case 3, 
 namely 0.3. 
 
 Referring to the pressure-volume diagram for Cycle I 
 Case 3, Fig. 17 (page 42), the area zZHK represents the 
 kinetic energy of the gas leaving the nozzle, which kinetic 
 energy equals 33,840 foot-pounds. This is for a quantity 
 of gas which measures 1 cubic foot at A. The velocity is 
 5290 feet per second. 
 
 For the sake of comparison it may be advantageous to 
 mention the velocities of the steam jets employed in De 
 Laval steam turbines. If saturated steam at 50 pounds, 
 absolute pressure is expanded adiabatically to a pressure of 
 0.6 pounds absolute, which corresponds to a temperature 
 of 85 F., and its heat energy turned into kinetic energy, 
 the velocity acquired works out at 3690 feet per second. 
 If saturated steam at 300 pounds absolute pressure were 
 treated similarly, the velocity would be 4380 feet per 
 second. The velocities actually obtained in practice must 
 be somewhat less than these figures, owing to friction in 
 the nozzles. 
 
 To get the best results from a fluid velocity such as 5290 
 feet per second would require, with a single turbine wheel, a 
 vane speed which cannot be obtained at present for want of 
 a sufficiently strong and light material the stresses pro- 
 duced by centrifugal force are too great. This difficulty is 
 experienced with De Laval turbines. The obvious way out 
 of the difficulty is to employ several wheels in series, the gas 
 passing through the several wheels with diminishing velocity, 
 
THE GAS TURBINE. 49 
 
 but with nearly constant pressure. This has been done in 
 steam turbines. 
 
 With the same object of reducing the vane speed, a device 
 has been proposed whereby the nozzles are mounted on a 
 wheel which rotates in the opposite direction to the wheel 
 carrying the vanes. If the two wheels rotate at the same 
 speed (in opposite directions) this speed will be half of that 
 of the single wheel if the nozzles were stationary. The cen- 
 trifugal force is, therefore, only one-fourth of what it would 
 otherwise be. 
 
 The frictional losses in the nozzles of a gas turbine will 
 probably be less than those in the nozzles of a steam turbine 
 for the same velocity of exit from the nozzle. 
 
 Cycle I, Case 4a. 
 
 If one tries to work to the same entropy-temperature 
 diagram as in Case 4, Fig. 20 (page 45), but employs a 
 divergent nozzle, as in Case 3a, to reduce the maximum 
 temperature to 700 C., so that the gas can be used in a tur- 
 bine without cooling arrangements, T l in this case will be 
 725 absolute C. (452 C.). It is not necessary, therefore, to 
 perform all the adiabatic expansion in a divergent nozzle, but 
 a portion of it can be performed in the turbine. If the fluid 
 is expanded in the nozzle only till its temperature falls to 
 700 C. (1292 F.), the pressure will then be 42 pounds abso- 
 lute; so that 27 pounds can be dropped in the turbine. 
 
 Referring to the pressure-volume diagram for Cycle I, 
 Case 4, Fig. 19 (page 44), the line qQ is drawn to represent 
 the pressure at which the gas leaves the nozzle. The kinetic 
 energy of the gas leaving the nozzle is represented by the 
 area XMQq. It can be ascertained that this amounts to 
 33,660 foot-pounds (for one cubic foot of gas measured at A), 
 and the velocity works out at 5280 feet per second. E will 
 be the same as in Case 4, and so will the ratio of negative 
 work to gross work. 
 
50 THE GAS TURBINE. 
 
 It seems to the author that an engine working on this 
 cycle, according to Case 3a or Case 4a, or between these, has 
 good prospects. The ideal efficiency is high from 0.58 to 
 0.68. How near one could approach this efficiency in prac- 
 tice would depend, of course, both on the losses in the motor 
 proper and on the losses in the pump. 
 
 The losses in the motor proper may be taken to include 
 the losses in the combustion chamber, if such is employed, 
 and in the nozzles. The motor losses will then consist of : 
 
 1. Loss of heat by radiation and conduction. 
 
 2. Fluid friction. 
 
 3. Friction in turbine bearings. 
 
 4. Loss due to incomplete expansion. 
 
 The first loss will be large, but should be less than in re- 
 ciprocating engines, owing to the higher velocities employed 
 and to the higher temperatures allowable in the metal. 
 
 The second loss will be considerable, but much less than in 
 turbines using saturated steam. It has been found by ex- 
 periment that hot dry air causes much less friction than wet 
 steam. (The steam is always wet in a De Laval turbine 
 casing, unless it enters the nozzles with a large amount of 
 superheat.) 
 
 The third loss will be trifling and the fourth loss should be 
 moderate. The discharge of heat with the exhaust gases is 
 here only considered as a loss in so far as it exceeds that of an 
 ideal engine. 
 
 It is. difficult to estimate the pump losses. Rotary com- 
 pressors on the turbine principle seem to have been em- 
 ployed up to only about 80 pounds pressure. Whether or no 
 they are suitable for high pressures is a point which it is very 
 desirable to ascertain. One would be inclined to believe that 
 the fluid frictional losses with such machines would be very 
 great if attempts were made to obtain high pressures. It 
 by no means follows, however, that a fairly efficient rotary 
 air compressor cannot be devised. 
 
THE GAS TURBINE. 51 
 
 A reciprocating compressor always has the disadvantage 
 that the air when drawn in becomes heated by contact with 
 the hot metal surfaces before compression commences. This 
 evil is reduced by compounding. It is an evil which occurs 
 to a serious extent with reciprocating gas engines working 
 on the Otto cycle. 
 
 With a reciprocating compressor it will be difficult to 
 avoid the necessity of jacketing the cylinder if high compres- 
 sions are employed. This will bring the compression curve 
 below the adiabatic and reduce the efficiency as before ex- 
 plained. 
 
 In any case, whatever be the nature of the pump, there is 
 bound to be a certain amount of heat passed through the 
 walls of the pump cylinders or casing. If this loss be made 
 up by friction or impact within the pump, the compression 
 may be along an adiabatic curve, but the loss will still have 
 to be considered. 
 
 The ratio of negative work to gross work (in the particular 
 cases here referred to) is somewhat high 0.3 to 0.4. In 
 the case of a turbine one need not fear the increase in the 
 bulk of the engine due to this high ratio; for the bulk of the 
 turbine will probably be very small for the power. Fric- 
 tional and other losses become, however, of much greater 
 importance when the ratio is high. To show this forcibly, 
 consider an extreme case. Suppose that the ratio of nega- 
 tive work to gross work in an ideal engine is 0.5, or, in sim- 
 pler language, suppose the pump requires half the gross 
 power of the machine, there being no friction. If now the 
 machine is not ideal, and if the mechanical efficiency of the 
 pump is only f and that of the motor proper only f , no useful 
 work whatever will be got out of the machine all the work 
 will be absorbed by friction. For, if the power of the motor 
 proper, including that spent on friction, is 100, the pump 
 will require 50, and as its efficiency is , it will take 75. This 
 is exactly what the motor will give out after deducting 
 
52 
 
 THE GAS TURBINE. 
 
 friction. There will, therefore, be no power got out of the 
 machine. When there is a high ratio of negative work to 
 gross work, success will, therefore, be dependent largely on 
 the efficiency of the pump. Unless the pump is at least 
 fairly efficient, success cannot be expected. In the Diesel 
 engine the bulk of the air is compressed to about 500 pounds 
 per square inch, and the air which carries the oil into the 
 cylinder is compressed from 100 pounds to 200 pounds 
 higher.* It would be interesting to know with what effi- 
 ciency the air is compressed in the Diesel engine. 
 
 [53 
 0256 
 
 Volume 
 
 FIG. 21 Cycle II, Case 1. Pressure-volume diagram. 
 
 Otto cycle reciprocating engines having ideal efficiencies 
 of 0.4 to 0.45 have given practical efficiencies of half that 
 amount. By practical efficiency is meant ratio of brake 
 horse-power to thermal units in gas consumed, calculated on 
 the higher calorific value. When the ideal efficiency is 
 
 *"The Diesel Engine," by H. Ade Clark. Proceedings, Inst. Mech. 
 Engrs., 1903, Part 3, page 395. 
 
THE GAS TURBINE. 
 
 53 
 
 increased above 0.45, the ratio of practical efficiency to ideal 
 efficiency usually falls below 0.5 the greater the ideal effi- 
 iency, the greater are the losses. With a turbine the losses 
 ought also to increase when the ideal efficiency is increased, 
 but whether to the same extent as with an Otto engine it is 
 
 T T=2273 
 
 Tf855 
 
 frsocfi 
 
 ^=290 
 
 FIG. 22. Cycle II, Case 1 Temperature-entropy diagram. 
 
 difficult to say. When considering high compressions, it is 
 well to note that the Diesel engine, with a high compression 
 and an incomplete expansion, has given some of the highest 
 practical efficiencies yet attained. The compression should 
 not cause the same trouble in starting a turbine as in starting 
 a reciprocating engine, as with a turbine it should be practi- 
 cable to arrange that at every instant the gross work is 
 
54 THE GAS TURBINE. 
 
 greater than the negative work. With a reciprocating 
 engine having a single cylinder working on the Otto cycle 
 there are, of course, periods when the negative work exceeds 
 the gross work. 
 
 Cycle II, Case 1. 
 
 With regard to explosion turbine engines, suppose that 
 the fluid is compressed adiabatically to, say, 101 pounds per 
 square inch absolute, that is to a temperature of 500 abso- 
 lute C. (932 F.). Let it now be heated at constant volume 
 by explosion, and let there be a mixture of such a strength 
 that the temperature will rise to 2000 C. (2273 absolute 
 C.). The pressure will then be 459 pounds absolute. If the 
 gas is now allowed to expand adiabatically till its pressure 
 is atmospheric (when its temperature will be 855 absolute 
 C.), and then cooled at that pressure till it resumes its 
 original state, the pressure-volume and entropy-temperature 
 diagrams will be as shown in Figs. 21 and 22 (pages 52 and 53). 
 
 In Fig. 22 the heat supplied to the fluid is represented by 
 the area aRTs, the heat rejected by the area aASs, and the 
 heat converted into work by the area ARTS. 
 
 area ARTS 
 Therefore 
 
 The negative work can be compared with the gross 
 work in Fig. 21. The ratio of negative work to gross work 
 
 area v VRA 
 
 Cycle II, Case 1, very nearly resembles common practice 
 to-day with reciprocating explosion engines. The expansion 
 is, however, continued to atmospheric pressure. This as a 
 rule is not desirable in a reciprocating engine, on account of 
 the extra length required to be given to the engine cylinder, 
 which not only increases the loss by friction but increases 
 
THE GAS TURBINE. 55 
 
 the loss of heat by the expanding gas and, if the same 
 length of stroke is employed for drawing in the fresh charge, 
 increases the heating of the charge before compression. 
 The case, however, is very different with turbines; and there 
 seems no good reason why with these the adiabatic expan- 
 sion should not be carried practically to atmospheric pres- 
 sure. 
 
 In practice the maximum pressure and the average maxi- 
 mum temperature throughout the gas would be less than the 
 values here indicated, owing to radiation losses. 
 
 Cycle II, Case la. 
 
 The gas could not be allowed into an uncooled turbine at 
 the maximum temperature in Cycle II, Case 1; but, if the 
 expansion was performed wholly or nearly wholly in a 
 divergent nozzle, the temperature of exit from the nozzle 
 would be sufficiently low to allow of the gas entering an 
 uncooled turbine. 
 
 For example, if the gas at the maximum temperature of 
 2273 absolute C. (4123 F.) and the maximum pressure 
 of 459 pounds absolute were expanded in a perfect divergent 
 nozzle till the temperature fell to 700 C. (973 absolute 
 C.), which was fixed on as the maximum allowable temper- 
 ature in an uncooled turbine, the mean pressure on leaving 
 the nozzle would be 23.5 pounds absolute. The kinetic 
 energy of the gas (1 cubic foot at A) on leaving the nozzle 
 would be represented by the area VRTQ& in Fig. 21, and 
 would amount to 20,500 foot-pounds. The mean velocity 
 (the square root of the mean square) would be 4120 feet per 
 second. 
 
 On comparing Cases 1 and la of Cycle II by reference to 
 the Table (page 75) with Cases 2, 3, 3a, 4 and 4a of Cycle I, 
 which have the same maximum temperature, it will be found 
 that the efficiency is very much greater than Cycle I, Case 2; 
 is nearly as great as Cycle I, Cases 3 and 3a; and is con- 
 
56 THE GAS TURBINE. 
 
 siderably below Cycle I, Cases 4 and 4a. The ratio of nega- 
 tive work to gross work is, however, greater than in Cycle I, 
 Case 2, and less than in Cases 3, 3a, 4 and 4a of Cycle I. 
 
 There are two objections to the use for turbines of a cycle 
 such as Cycle II, and these objections must be set against the 
 advantage which turbines would possess over reciprocating 
 explosion motors, in being able to make better use of the tail 
 end of the pressure-volume diagram. 
 
 One objection is that explosions at constant volume have 
 to take place intermittently, while a turbine desires a contin- 
 uous supply of fluid. If the supply is not continuous the 
 power of the turbine is less than it would otherwise be for a 
 given size of machine; and the initial cost, the bulk and 
 most important the loss by friction are greater in propor- 
 tion to the power developed than they would otherwise be. 
 
 The other objection is that the fluid must leave the explo- 
 sion chamber at varying pressure. This necessitates, unless 
 special means' are provided to prevent it, the fluid entering 
 the turbine casing either at varying pressure or at varying 
 velocity, which of course is objectionable, as the speed of ro- 
 tation of the turbine cannot, during the period of a cycle, be 
 made to vary correspondingly. 
 
 The second objection might be met by employing in a par- 
 allel flow turbine of the De Laval type long radial blades, 
 and causing the nozzles to be altered in position according to 
 the pressure, so as to direct the gas onto the outer ends of 
 the blades at low pressures. The difficulty could also be met 
 by an arrangement of reciprocating engine combined with a 
 turbine, the gas being first expanded in the reciprocating 
 engine to a certain pressure and then passed on to the tur- 
 bine to complete its expansion. If several reciprocating cyl- 
 inders were employed, the first objection also would be got 
 over, but it is true that with such a combination some of the 
 most important advantages of the turbine would be lost. 
 The idea is, however, in the author's opinion, worthy of con- 
 
THE GAS TURBINE. 57 
 
 sideration. Reciprocating steam engines have been success- 
 fully combined with steam turbines in this manner.* 
 
 Cycle II, Case 2. 
 
 An explosion engine, in which a very high compression 
 pressure is employed, will now be considered. If compres- 
 sion be carried to 818 pounds absolute as in Cycle I, Case 4, 
 one obtains with a maximum temperature of 2000 C. 
 (3632 F.) a maximum pressure of 2045 pounds absolute 
 and a very high ratio of negative work to gross work. If a 
 much lower compression namely 417 pounds absolute is 
 adopted, this will give a temperature of compression of 750 
 absolute C. (1382 F.). Working on the same cycle as in 
 the last case and arranging the explosive mixture to give a 
 maximum temperature of 2000 C. (2273 absolute C.), a 
 maximum pressure of 1265 pounds absolute is obtained, 
 and the pressure-volume and the entropy-temperature dia- 
 grams will be as shown in Figs. 23 and 24 (page 58). 
 
 Referring to Fig. 24, 
 
 , 
 =U.DO. 
 
 area aUW w 
 Referring to Fig. 23, 
 
 negative work _ area ^ 
 gross work area u 1 U l UWW l 
 
 E is the same as in Cycle I, Case 4, and the ratio of negative 
 work to gross work is also the same. The compression is 
 lower than in Cycle I, Case 4, but the maximum pressure, is 
 very much higher. 
 
 The excessively high maximum pressure is an objection 
 to this case. 
 
 * See Paper by Professor Rateau read before the North of England Insti- 
 tute of Mining and Mechanical Engineers at Newcastle-on-Tyne, Dec. 13, 1902; 
 or Paper by the same author read at the Chicago Meeting of the Institution of 
 Mechanical Engineers, Proceedings 1904, Part 3 (page 737). 
 
58 
 
 THE GAS TURBINE. 
 
 T-2273 
 
 FIG 24. Cycle II, Case 2. Temperature- 
 entropy diagram. 
 
 Volume 
 
 FIG. 23.-Cycle II Case 2. Pressure- 
 volume diagram. 
 
THE GAS TURBINE. 59 
 
 Cycle II, Case 2a. 
 
 If the expansion took place in an ideal divergent nozzle as 
 before till the temperature fell to 700 C. (973 absolute C.), 
 the gas would still have a pressure of 70 pounds absolute, 
 while the mean velocity of exit from the nozzle would be 
 4300 feet per second. If the gas were expanded in the 
 nozzle down to 25 pounds absolute, the temperature would 
 then be 741 absolute C. (1366 F.), and the mean velocity 
 of the gas leaving the nozzle would be 4830 feet per second. 
 
 Cycle III, Case 1. 
 
 It has been proposed, when a water-jacket is employed, to 
 utilize the heat passed into the jacket water by causing this 
 heat to generate steam from the water. This steam could 
 then receive further heat from the products of combustion, 
 which would therefore be reduced in temperature, while the 
 steam would be superheated. The steam and products of 
 combustion could then expand adiabatically, doing work in 
 the same or in separate turbines. The carrying out of this 
 idea would affect the efficiency in the several cases consid- 
 ered of Cycle I. Cooling arrangements are not required in 
 Cycle I, Case 1, so this case need not be further considered. 
 In Cycle I, Case 2, let it be supposed that the combustion 
 chamber is jacketed and that the jacket water is heated 
 and converted into steam by heat taken from the products of 
 combustion, which have their temperature thus lowered from 
 2000 C. to 700 C., that is, to the temperature at which they 
 can safely be allowed into an uncooled turbine, the steam 
 being superheated up to 700 C. Let this be called Cycle 
 III, Case 1. 
 
 Referring to Fig. 16 (page 41), the heat in the products 
 of combustion which is converted into work is now repre- 
 sented by the area ABCD instead of by the area ABEF. 
 The heat represented by the area dCEf has, however, been 
 
60 THE GAS TURBINE. 
 
 employed in heating water and generating and superheating 
 steam. The fraction of this heat which is converted into 
 work will not now be as great as in the original scheme 
 of working. That is to say, the net work got out of the heat 
 put into the water and steam will be less than the area 
 4 ' DCEF. By transferring heat to the water and steam from 
 the gas, E is therefore reduced. There must, however, in any 
 case, as already mentioned (page 30), be lost in practice a 
 large amount of heat when the products of combustion enter 
 the turbine casing at a temperature such as 2000 C., and, 
 by adopting this combined steam and gas scheme, a much 
 higher practical efficiency may possibly be attained than 
 would otherwise be possible. As the net work ideally is less 
 than in Cycle I, Case 2, and as the negative work is not less 
 (and may be greater by the amount of work required to 
 pump the water into the jacket if under pressure), the ratio 
 of negative work to gross work is increased. In Case 2 of 
 Cycle I, the ratio of negative work to gross work is low, and 
 it will, therefore, be allowable to increase this ratio. 
 
 Cycle III, Case 2. 
 
 Case 3 of Cycle I could be modified in the same way by 
 reducing the temperature of the products of combustion from 
 2000 C. to 700 C., and by employing the heat so given up 
 in heating water and generating and superheating steam. 
 The steam could be generated at 300 pounds pressure ab- 
 solute (the same pressure as the products of combustion) 
 and superheated to 700 C. at this pressure. The steam and 
 gas could then be expanded adiabatically in the same or 
 separate turbines. As in the previous case, E would be 
 reduced, and the ratio of negative work to gross work 
 increased. As in the previous case, the practical efficiency 
 might also be largely increased. 
 
 The pressure-volume and entropy-temperature diagrams 
 for the gas in this case (called Cycle III, Case 2) are shown 
 
THE GAS TURBINE. 
 
 61 
 
 in Figs. 25 and 26 respectively (pages 61 and 62). The gas is 
 compressed along the line AG as in Cycle I, Case 3, till its 
 pressure is 300 pounds absolute and its temperature is 
 409.5 C. (682.5 absolute C.). It is then heated by com- 
 bustion at constant pressure along the line GH as in Cycle I 
 Case 3, till its temperature is 2000 C. (2273 absolute C.). 
 Heat is now withdrawn from the gas at constant pressure and 
 transformed to the water and steam, the temperature of the 
 gas falling along the line HH lt to 700 C. (973 absolute C.) 
 at H r The heat transferred from the gas to the water is 
 
 -30O 
 LV- 0-392 
 
 K 
 
 Volume 
 FIG. 25. Cycle III, Case 2. Pressure-volume diagram gas. 
 
 represented, Fig. 26, by the area kJH^Hk. The gas now ex- 
 pands adiabatically along the line H^K^ till the pressure 
 is 15 pounds absolute, when the temperature will be 140 C. 
 (413 absolute C.). The contraction of the gas at constant 
 pressure along the line K^A completes the cycle. Dotted 
 lines have been placed on Figs. 25 and 26 to illustrate Cycle I 
 Case 3, where this differs from the present cycle. The two 
 cycles can thus be compared. 
 
 Pressure-volume and entropy-temperature diagrams for 
 the water are shown in Figs. 27 and 28 (pages 63-65). Re- 
 ferring to Fig. 28, the water is heated at a constant pressure 
 of 300 pounds per square inch absolute along the line fc from 
 
62 
 
 THE GAS TURBINE. 
 
 100.6 C. (373.6 absolute C.) to 214 C. (487 absolute C.), 
 which is the boiling point at this pressure. The water is 
 now converted into steam, this process being represented by 
 the line eg-, and the steam is superheated at constant pressure 
 as represented by the line gd, till its temperature is 700 C. 
 
 T-2273 
 
 t-Z90 
 
 a 
 
 FIG. 26. Cycle III, Case 2. Entropy-temperature diagram gas. 
 
 (973 absolute C.). The steam is then expanded adiabati- 
 cally along the line de till it falls to 15 pounds absolute 
 pressure, its temperature then being 184 C. (457 absolute 
 C.). The steam is now exhausted and cools along the line 
 eh. At h it is saturated, its temperature being 100.6 C. 
 (373.6 absolute C.), and thereafter it condenses along the 
 line hf and is compressed to its initial state. 
 
THE GAS TURBINE. 
 
 63 
 
 Fig. 27 shows the work done by the steam in its generation, 
 superheating and adiabatic expansion. The work done in 
 forcing the water into the chamber at 300 pounds pressure is 
 not shown in Fig. 27 and is negligible in the present inves- 
 tigation. 
 
 The heat required to raise the water from 373.6 absolute 
 C. to 487 absolute C., is represented in Fig. 28 by the area 
 tjcc v The area c 1 cgg 1 represents the latent heat of steam 
 at a pressure of 300 pounds absolute (the temperature being 
 487 absolute C.), and the area g l gde 1 represents the heat re- 
 
 jrarr 
 
 Volume 
 FIG. 27. Cycle III, Case 2. Pressure-volume diagram steam. 
 
 quired to superheat the steam from 487 absolute C. to 973 
 absolute C. The area fjhh^ represents the latent heat of 
 steam at a pressure of 15 pounds absolute, and the area 
 h l hee l represents the heat required to superheat this steam 
 from 373.6 absolute C. to 457 absolute C. 
 
 Comparing this case with Case 3, of Cycle I, it is found 
 that the total heat absorbed is the same in both cases, being 
 represented by the area aGHk in Fig. 26. The portion of 
 this heat which is converted into work in Case 3, Cycle I, is 
 represented by the area AGHK, while the corresponding 
 portion in the present case is represented by the sum of the 
 areas AGH^K^ Fig. 26, and fcgdeh, Fig. 28. This sum is less 
 
64 THE GAS TURBINE. 
 
 than the area AGHK, and E in this case is only 0.33 as com- 
 pared with 0.58 in Case 3 of Cycle I. The fall in the value 
 of E is due to the relatively low efficiency of the steam por- 
 tion which has an ideal efficiency of only 0.28. (For a 
 steam engine this is really not low.) 
 
 The feed-water has been taken at a temperature corre- 
 sponding to atmospheric boiling-point. It has been assumed 
 that the steam is exhausted into the atmosphere, and is not 
 condensed for use over again. It would, therefore, be neces- 
 sary, in order to follow the cycle, to heat the feed-water to 
 100 C. It should not be difficult to approximately accom- 
 plish this by utilizing the heat of the exhausting gases. 
 By heating the feed-water still more, the efficiency could 
 be improved; but the improvement would be slight (less 
 than in an ordinary steam engine) and the feed-water would 
 have to be under pressure. As, moreover, any increase of 
 exhaust or back pressure is a serious matter with a turbine, 
 and as feed-water heaters must to a certain extent affect 
 this back pressure, any prospect of gain by heating the feed- 
 water beyond 100 C. need not be considered. 
 
 The gross work in the present case is represented, Figs. 
 25 and 27, by the area zZH^ + the area acde. This is less 
 than the gross work in Cycle I, Case 3, which is represented 
 by the area zZHK. The negative work in Cycle I, Case 3, 
 was represented by the area zZGA. In the present case it 
 is also represented by this area, neglecting the work of pump- 
 ing the water into the jacket. The ratio of negative work 
 to gross work in the present case is 0.41 as compared with 
 0.3 in Cycle I, Case 3. This ratio (0.41) is rather high. It 
 will, however, probably not be so objectionable in the pres- 
 ent case as the ratio 0.40 in Case 4 of Cycle III, as the real 
 efficiency in practice will come nearer to the ideal in this case 
 than in Case 4 of Cycle III. In the present case the ratio 
 could be reduced by lowering the compression. This would 
 reduce E. 
 
THE GAS TURBINE. 
 
 65 
 
 As the mass of the water employed is not the same as the 
 mass of the air and fuel, the scale for entropy in Fig. 28 
 has been made different from that in the other entropy- 
 temperature diagrams, so that in all these diagrams areas 
 represent quantities of heat to the same scale. In all the 
 pressure-volume diagrams the scales are the same except in 
 Fig. 29 (page 66), which will be referred to hereafter. It 
 might be mentioned here that all the numerical results given 
 in this Paper have been obtained by calculation and not by 
 scaling the diagrams. 
 
 973 d 
 
 373-6 
 
 FIG. 28 Cycle III, Case 2. Entropy-temperature diagram steam. 
 
 It will be seen that it has been assumed that the gas and 
 steam expand adiabatically separate from each other. The 
 adiabatic curve of the one is different from that of the other, 
 as the specific heats are different; and, while the gas falls to 
 a temperature of 140 C. (413 absolute C.), the steam falls 
 only to 184 C. (457 absolute C.). This will be correct if the 
 steam and gas are not mixed. It is much simpler to consider 
 this case than to consider the case where the gases are 
 intimately mixed. In this latter case the diagram Fig. 26 
 would be altered, and it could not so easily be seen where the 
 loss of efficiency came in. In practice, however, it will prob- 
 ably be found convenient to mix the gases. This will alter 
 the diagrams and the efficiency somewhat; but what has 
 been considered gives a good idea of the general effect of the 
 5 
 
66 
 
 THE GAS TURBINE. 
 
 employment of steam in conjunction with gas. If the steam 
 and gas are not mixed, a condenser could be employed for the 
 former. The steam could then be expanded to a much lower 
 temperature and pressure, and the efficiency would be con- 
 siderably raised. 
 
 Cycle II could be modified by combining steam with the 
 gas, in the same way as Cycle I was modified. A case of this 
 nature has not been worked out; but Case 1 of Cycle II 
 could probably be modified in this way. Case 2 of Cycle II 
 could not be so treated on account of the high ratio of nega- 
 tive work to gross work that would occur. 
 
 10 
 
 12 
 
 14 
 
 FIG. 29. Pressure-volume diagram. The horizontal scale is half that of the other diagrams. 
 
 One might try to improve on all these cycles, by extending 
 the adiabatic expansion line of the gas below atmosphere, 
 instead of stopping it at atmospheric pressure. It would, 
 of course, be necessary to compress the fluid back again to 
 atmospheric pressure; but, if this compression were isother- 
 mal or between the isothermal and adiabatic, there would 
 be an increase of efficiency. Carnot's cycle is in fact being 
 approached in the lower part of the diagram. 
 
 Figs. 29 and 30 are respectively pressure-volume and 
 entropy-temperature diagrams of Cycle I, Case 3, modified 
 by continuing the adiabatic expansion to a pressure of 2 
 pounds per square inch absolute. The scale for volumes in 
 
THE GAS TURBINE. 
 
 67 
 
 Fig. 29 has, for convenience, been made half that of the other 
 diagrams. Kb represents the addition to the adiabatic line of 
 expansion, and be represents isothermal compression of the 
 gas from 2 pounds absolute at b to 15 pounds absolute at c. 
 There should be no difficulty in a turbine in extending the 
 expansion from K to b. There may be difficulty, however, in 
 
 H T2273 
 
 /r-682-, 
 
 (-290 
 
 Ti543 
 
 1 
 
 Vv\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\V\ 
 
 FIG. 30. Entropy-temperature diagram. 
 
 getting isothermal compression from b to atmospheric pres- 
 sure at c. As the volume at b is 14 times the initial volume 
 it will be desirable to get the fluid discharged as quickly as 
 possible. A rotary compressor will probably be best for this 
 purpose. A compression, sufficiently near to the isothermal 
 and sufficiently remote from the adiabatic to raise the effi- 
 ciency appreciably, should be obtainable. 
 
68 THE GAS TURBINE. 
 
 The temperature at b is 270 C. (543 absolute C.), and if 
 the compression were isothermal, this would of course be the 
 temperature all along the line be. The gases could be passed 
 through or around water-cooled tubes to keep down the tem- 
 perature during compression. With the gas at a temperature 
 of 543 absolute C. it would not do to spray water into it, 
 unless sufficient water were sprayed to cool the gas below the 
 boiling-point of the water, which is 326 absolute C. at this 
 pressure. 
 
 If compression takes place along the isothermal line be, a 
 net amount of work will be gained, represented by the area 
 Kbc. The gas will be discharged into the atmosphere at c, 
 the volume at discharge being 1.874 of the original volume 
 (at A). Even if the compression is not isothermal, an 
 amount of work may be gained which will wipe out the .extra 
 losses in the machine, provide for pumping out the cooling 
 water, and perhaps leave a margin of net gain. 
 
 In Fig. 30 the heat absorbed by the fluid is represented by 
 the area aGHk, the heat rejected by the area aAcbk, and the 
 heat converted into work by the area AGHbc. As the heat 
 absorbed remains unchanged, while the heat converted into 
 work is increased by the area Kbc, E is of course increased. 
 
 __area AGHbc 
 ~ area AGHk 
 
 This enlarging of the diagram of course affects the ratio of 
 negative work to gross work. Referring to Fig. 29 (page 
 66), 
 
 gross work = area zZHKebc 
 negative work = area zZGA + area Keb 
 net work = area AGHbc 
 
 negative work _ area zZGA + area Keb 
 gross work area zZHKebc 
 
 In the free piston explosion engines, which were at one 
 time in fairly common use, the best known of which is the 
 
THE GAS TURBINE. 69 
 
 Otto and Langen, the expansion was carried to a pressure 
 considerably below the atmosphere. The compression to at- 
 mospheric pressure which followed must have been between 
 the isothermal and the adiabatic. 
 
 If this continuation of the adiabatic expansion below at- 
 mospheric pressure is not found to be advisable to the extent 
 that has just been described, it may be found advisable to a 
 less extent. If it is found advisable in any case, it is more 
 likely to be so in a case in which the high pressure of the gases 
 after combustion is reduced to a low pressure in divergent 
 nozzles, before the gas is allowed into the turbine casing, 
 than in a case in which the whole fall of pressure takes 
 place in the turbine casing. In the former case very high 
 vane speeds are necessary, and the friction between the 
 rotating parts and the fluid in the casing is an extremely 
 important matter. The reduction of the pressure within the 
 turbine casing from atmospheric pressure (or above that) 
 to one-quarter or one-eighth of that amount may therefore 
 very much reduce the frictional losses. It is true that the 
 rotary pump, if such is employed for completing the cycle, 
 has to deliver at atmospheric pressure, but the rotating 
 parts of the pump can revolve at a much lower speed, and 
 the friction will therefore be of much less consequence. 
 
 With such high speed turbines there is another question to 
 be considered. It has been stated in discussing Cases 3a and 
 4a of Cycle I, and la and 2a of Cycle II, that the velocity 
 of the gases escaping from the divergent nozzles would be 
 over 4000 feet per second, if the heat energy converted 
 into kinetic energy was as mentioned. The author is not, 
 however, aware of any results of experiments having been 
 published in which velocities of these amounts were obtained, 
 when the pressure of the medium into which the divergent 
 nozzle discharged was atmospheric. It is supposed by some 
 that there is a maximum limit to the velocity of a gas leav- 
 ing a divergent nozzle and escaping into a given medium 
 
70 THE GAS TURBINE. 
 
 which is at a given pressure, etc., and that this limit velocity 
 is dependent on the pressure in the medium into which the 
 nozzle discharges, and is less when the pressure in this medium 
 is greater, and vice versa. That is to say, it is supposed by 
 some that, after a certain velocity of discharge has been 
 attained, no increase in the initial temperature or pressure 
 will increase this velocity; but a reduction of the pressure in 
 the medium may do so. The author does not express any 
 opinion himself on this point, but if it should be found 
 that the reduction of the pressure inside a turbine casing 
 below atmospheric pressure enables the heat energy of the 
 gas to be more effectively converted into kinetic energy, this 
 will be a further argument in favor of so reducing the pres- 
 sure. Whether or not there is an advantage to be gained 
 remains to be proved, but there is at any rate a possibility 
 of gain by thus extending the expansion and it is a possibility 
 which, in the author's opinion, should not be ignored. In 
 dealing with large volumes and small pressures there is, as 
 already mentioned, an immense difference between turbines 
 and reciprocating engines. 
 
 Cycle IV. 
 
 The fourth cycle which will be considered in this Paper is 
 one in which a high ideal efficiency can be obtained with a 
 low compression, and without having an abnormally high 
 ratio of negative work to gross work. 
 
 Figs. 31 and 32 are, respectively, pressure-volume and 
 entropy-temperature diagrams for an engine working on 
 this cycle. In explaining the cycle it is best to start at E l . 
 At this point the temperature of the fluid is 1592 C. (1865 
 absolute C.), and the pressure is 30 pounds absolute. 
 
 Let the fluid be heated by combustion at constant pressure 
 along the line E l C l till the temperature reaches 2000 C. 
 (2273 absolute C.). Now let the gas expand adiabatically 
 from C 1 to D l till the pressure is atmospheric. The tempera- 
 
THE GAS TURBINE. 71 
 
 ture will then be 1592 C. (1865 absolute C.). Now let the 
 gas pass through a regenerating chamber and be cooled at a 
 constant pressure from D l to F l till the temperature is 80 C. 
 (353 absolute C.)- The gas escapes at F l into atmosphere, 
 and thereafter cools at constant pressure to 17 C. (290 
 absolute C.) at A. A new charge is taken at A and com- 
 pressed adiabatically to B 1 where the pressure is 30 pounds 
 absolute and the temperature 80 C. (353 absolute C.). 
 The fluid is now passed through the regenerating chamber, 
 and is heated at constant pressure along the line B 1 E\ 
 taking back the heat given up by the last charge. This 
 will raise its temperature to 1592 C. (1865 absolute C.) and 
 place the fluid in the condition it was at the start. 
 
 FIG. 31. Cycle IV. Pressure- volume diagram. 
 
 Referring to Fig. 31, the gross work is represented by the 
 area g l G l C l D l , the negative work by the area g l G 1 B l A ) and 
 the net work by the area AB*C 1 D 1 . 
 
 ,, . negative work area Q 1 G 1 B 1 A 
 Therefore _ = - _^=o. 16 (0.1553). 
 
 gross work area g*G l l D l 
 
 The heat absorbed by the fluid (other than that obtained 
 in the regenerator from a previous charge) is represented in 
 Fig. 32 by the area eE l C l d. The heat rejected (other than 
 that given to the regenerator) is represented by the area 
 aAF*f. The heat converted into work is represented by the 
 difference of these two areas. 
 
 = area aB l C*d area aB l E l e area aAF l f 
 =area eE^d 
 
72 
 
 THE GAS TURBINE. 
 
 Therefore the area AB 1 C 1 D 1 represents the heat converted 
 into work 
 
 area 
 
 Therefore 
 
 area eE l C l d 
 
 0.84. 
 
 The ideal efficiency is high; but the highest actual effi- 
 ciency which could practically be obtained would be very 
 
 fr=2273 
 
 T,-I865 
 
 \\V\\\\\\\\\\\\\\\\\\\\\\\\\ 
 
 a -f e <z 
 
 FIG. 32. Cycle IV. Entropy-temperature diagram. 
 
 much below this. Besides the losses in the motor proper and 
 in the pump, there would be a very great loss in the regen- 
 erator. It would not be practicable to reduce the tempera- 
 ture of the exhausting gases in the regenerator to 80 C., or 
 to raise the temperature of the fresh gases in the regenerator 
 to 1592 C. 
 
THE GAS TURBINE. 73 
 
 If the losses in the regenerator and in the passages leading 
 to it and from it amounted to 50 per cent, of the heat which 
 is ideally given to or taken from the regenerator, these losses 
 would have to be made up by extra heat given to the fluid 
 by combustion, and the efficiency would fall to 0.3. This 
 does not take into account the losses in the motor proper 
 and in the pump. The heat losses in the motor would 
 probably be very great. This cycle may, however, when 
 used with turbines, give results sufficiently good to justify 
 its use. It certainly seems to promise better results with 
 turbines than with reciprocating engines, on account of the 
 lower frictional losses that might be expected with turbines. 
 With reciprocating engines the large volumes and the low 
 pressure of the fluid would cause extremely high percen- 
 tage losses in friction. 
 
 The cycle has the disadvantage that gas at a very high 
 temperature has to be conveyed from the regenerator to 
 the turbine. This practically makes it absolutely neces- 
 sary to have the turbine quite close to the regenerator. It 
 would seem to be expedient to build the regenerator of brick- 
 work and to erect the turbine in this brick-work. This would 
 very much limit the usefulness of the cycle, as it would not 
 be quite feasible in many cases to have either a regenerator 
 of the nature required at the place where power is wanted, 
 or to transmit the power from a place suitable for holding 
 the regenerator. Nevertheless there will be cases in which 
 it will be quite practicable to build a regenerator beside the 
 turbine, and this cycle therefore seems to be worthy of con- 
 sideration. A rotary pump driven from the turbine spindle 
 could easily be used for compressing the gases, thus simpli- 
 fying the mechanical moving parts. 
 
 The several cases can be compared in the Table (page 
 75). It will be seen that a high ideal efficiency is, as a 
 rule, accompanied by a high ratio of negative work to gross 
 work. Cycle 4 is, however, an exception to the rule. The 
 
74 THE GAS TURBINE. 
 
 cycle has the highest ideal efficiency and the lowest ratio of 
 negative work to gross work. As has been already pointed 
 out, however, the efficiency which could be actually looked 
 for with this cycle would be very much below the ideal, and 
 the cycle has other objections, as already stated. 
 
 Cycle I, Case 1, has a high ratio of negative work to gross 
 work, although the efficiency is the lowest. This is because 
 all the heat is supplied to the gas at a comparatively low 
 temperature. 
 
 Engineers interested in any particular cycle can work out 
 other cases for themselves if they consider it necessary, but 
 it is suggested that after a careful perusal of this Paper the 
 effect of any change can be guessed at with fair accuracy. 
 It might be possible to use the exhaust gases from a turbine 
 working according to Cycle I, Cases 1 and 3, or Cycle II, 
 Case 1, to heat the fluid after compression and so to save 
 fuel. Consider Fig. 14 (page 38). The heat supplied to 
 the fluid is represented by the area aBCd. Part of this heat 
 might be obtained from the hot exhaust gases and the 
 efficiency of the cycle thus raised, as will be clear from the 
 description just given of Cycle IV. With Cycle I, Cases 1 
 and 3, and Cycle II, Case 1, there should be no necessity 
 to use a regenerator of brick or such like refractory material. 
 The exhaust gases could be passed through tubes and the 
 fresh air passed over the outside surfaces of the tubes, or 
 some equivalent construction might be employed. 
 
 Many other cycles or modifications of cycles might have 
 been investigated; but the author had considered it inadvis- 
 able to burden the Paper with them. 
 
 As regards the Carnot cycle, an engine working on this 
 cycle would have the same value for E as one working on 
 Cycle I for the same values of t and t c ', and, if the isothermal 
 expansion were carried far enough, it would (under ideal 
 conditions) do the same work per cycle for the same amount 
 of fluid, and have the same ratio of negative work to gross 
 
THE GAS TURBINE. 
 
 75 
 
 work. The maximum volume of the fluid would, however, 
 be very much greater; and although this is not such a serious 
 matter with a turbine as with a reciprocating engine it is 
 nevertheless not a condition to be accepted without due 
 recompense. 
 
 TABLE Comparing the Several Cycles and Cases. 
 
 Cycle. 
 
 Case. 
 
 Compression. 
 
 Maximum 
 Temp, 
 absolute 
 C. 
 
 Maximum 
 pressure 
 Ibs. 
 per sq. in. 
 absolute. 
 
 Ideal 
 efficiency. 
 (*) 
 
 Ratio of 
 negative 
 work 
 to gross 
 work. 
 
 Temp, 
 absolute 
 C. 
 
 Pressure 
 Ibs. 
 per sq. in. 
 absolute. 
 
 I 
 
 1 
 
 389 
 
 42 
 
 973 
 
 42 
 
 0.25 
 
 0.40 
 
 I 
 
 2 
 
 389 
 
 42 
 
 2273 
 
 42 
 
 0.25 
 
 0.17 
 
 I 
 
 3&3a 
 
 682.5 
 
 300 
 
 2273 
 
 300 
 
 0.58 
 
 0.30 
 
 I 
 
 4 & 4a 
 
 909 
 
 818 
 
 2273 
 
 818 
 
 0.68 
 
 0.40 
 
 II 
 
 1 & la 
 
 500 
 
 101 
 
 2273 
 
 459 
 
 0.55 
 
 0.23 
 
 II 
 
 2&2a 
 
 750 
 
 417 
 
 2273 
 
 1265 
 
 0.68 
 
 0.38 
 
 III 
 
 2 
 
 682.5 
 
 300 
 
 2273 
 
 300 
 
 0.33 
 
 0.41 
 
 IV 
 
 
 
 353 
 
 30 
 
 2273 
 
 30 
 
 0.84 
 
 0.16 
 
 If an engine working on Cycle I has T l =t c , then an engine 
 working on the Carnot cycle with the same values of t and t c 
 would require, in order to do the same work, to commence 
 its adiabatic expansion at the point where the engine working 
 on Cycle I leaves off. If p and P 1 on the Cycle I engine are 
 atmospheric pressure, then, on the Carnot cycle engine, the 
 whole of the adiabatic expansion would take place below 
 atmospheric pressure. It is interesting to compare the Car- 
 not cycle with other cycles, but it hardly seems useful in the 
 present investigation to devote any more space to this com- 
 parison. 
 
 Although the practical efficiency of an engine is usually of 
 great importance, there are many occasions on which a very 
 poor efficiency will be tolerated if other conditions are satis- 
 factory. Small gas engines using lighting gas, costing 2s. to 
 3s. per 1000 cubic feet, are employed in great numbers, al- 
 though the fuel cost per brake horse-power hour is very high. 
 
76 THE GAS TURBINE. 
 
 Small electro-motors are extensively used consuming cur- 
 rent which costs over 2d. per Board-of-Trade Unit (kilo- 
 watt-hour). The fuel cost and the energy cost, respectively, 
 in the two cases are high; but the user prefers to put up 
 with this, rather than employ a power plant which has a 
 higher initial cost or which requires more attention or is 
 generally more inconvenient. 
 
 If, therefore, small gas turbines could be sold at a low 
 price, and if they required little attention and did not 
 readily get out of order, they might be in great demand, even 
 although the gas consumption per brake horse-power hour 
 was high. With the average user of a small engine, produc- 
 ing say 100 brake horse-power hours per week, a reduction 
 of 10 in the initial cost is of more consequence than a re- 
 duction of 2 cubic feet per brake horse-power hour in gas 
 consumption. The same user would no doubt be quite 
 willing to allow an additional 5 cubic feet of gas per brake 
 horse-power hour if he were saved trouble and anxiety and 
 small expenses in the working of the engine. 
 
 To produce a gas turbine cheaply it seems necessary 
 to avoid reciprocating parts entirely and to be content with 
 a low compression. Cycle I, Case 1, appears to lend itself 
 to cheapness of construction and simplicity, but it might 
 be advisable to reduce the compression at the expense of 
 efficiency. A rotary pump could undertake the compres- 
 sion. 
 
 In many cases the vibration of a reciprocating engine is 
 extremely objectionable; and a motor that ran with practi- 
 cally no vibration would be popular, even if its initial cost 
 were greater and it were more extravagant with fuel. In 
 motor cars, for example, oil or spirit explosion engines are 
 used for their lightness and compactness; but the vibration 
 they cause is objectionable. If a satisfactory turbine were 
 obtainable, there is no doubt that motor-car builders would 
 eagerly buy it and install it on their cars, even if the cost 
 
THE GAS TURBINE. 77 
 
 were greater and the efficiency less than the many arrange- 
 ments of explosion reciprocating engines now in use. 
 
 One might mention many other uses to which gas tur- 
 bines could advantageously be put, if they were obtainable 
 as fairly efficient and reliable machines. In many factories 
 and engineering works, electric motors fed by current from 
 a central station are used to drive individual machines or 
 groups of machines, in order to save the losses and inconven- 
 iences produced by driving by belts or ropes. Gas engines 
 (reciprocating) have been used to a limited extent for the 
 same purpose; but the foundations required for these and the 
 vibration caused by them have prevented their extensive 
 use. If a gas turbine were obtainable which could be set 
 down anywhere like an electric motor, it would serve 
 splendidly for this purpose, and, in order to displace electric 
 driving, it would only require to possess an efficiency greater 
 than the efficiency of the central station engine multiplied 
 by the efficiencies of dynamo, mains, and motor. 
 
 Suction gas producers are coming into extensive use, and 
 by their means gas engines can take the place of steam en- 
 gines where they otherwise would not. Whatever objections 
 there might be to the supply of gas to gas engines from these 
 suction producers, these objections should be less rather than 
 greater with turbines than with reciprocating engines. 
 
 The power of a gas turbine could be effectively varied with 
 an insignificant variation of speed, by cutting out one or 
 more of the nozzles or admission ports which admit the fluid 
 to the turbine buckets or first set of buckets when several 
 are employed in series, the fluid being passed through the 
 acting ports or nozzles at a uniform pressure and with a 
 uniform velocity. To enable the acting nozzles or admis- 
 sion ports, no matter how many may be in use, to deliver 
 the fluid always in the same uniform manner, it will be 
 necessary, if one flame supplies all the nozzles or admission 
 ports, to control the fuel and air supplies to the flame in con- 
 
78 THE GAS TURBINE. 
 
 junction with the mechanism which cuts off the nozzles or 
 admission ports. It will, however, be usually advisable to 
 supply the air and fuel to the flame at a constant velocity 
 which, although it could be done when one flame supplied 
 a varying number of nozzles, would involve complications; 
 and it may therefore be found expedient to have a separate 
 flame and a separate combustion chamber for each nozzle. 
 This would involve the necessity of igniting or extinguishing 
 a flame for each change of power, and, although this could 
 be done automatically, it is objectionable. Much careful 
 consideration will be required to determine which is the least 
 evil to put up with and which course had best be taken. 
 
 The author hopes that this comparison of the several 
 cycles treated will be of some use in showing what may be 
 expected from each, and which will be best suited for a motor 
 which is required to work under given conditions and which 
 it is important should have given characteristics. One 
 cannot, however, compare the several cycles and estimate 
 the actual efficiencies, &c., which might be expected in prac- 
 tice in anything like so satisfactory a manner as could be 
 desired, without having more information obtained by ex- 
 periment on the following three points : 
 
 1. Losses in pneumatic compression to high pressures. 
 
 (a) With reciprocating compressors. 
 (6) With rotary compressors, 
 (c) With a combination of reciprocating and ro- 
 tary compressors. 
 
 2. Expansion of hot gases in divergent nozzles. 
 
 3. Radiation losses and transference of heat from gases 
 to metals at high temperatures. 
 
 It would immensely aid the solving of the gas turbine 
 problem if a thorough set of experiments on these three 
 points were made and the results published. This would 
 naturally cost a considerable amount of money; but the 
 information obtained by the engineering world would be 
 
THE GAS TURBINE. 79 
 
 very good value at the price. Money has been spent and is 
 being spent by engineering and scientific societies on inves- 
 tigations which, while no doubt interesting and instructive, 
 are not of so far-reaching importance as experiments which 
 would materially aid in the production of a successful gas 
 turbine. 
 
 Discussion. 
 
 The PRESIDENT said that the author of the Paper had 
 been for a long time associated with Mr. Dugald Clerk. It 
 was rather an exception for the Council of the Institution to 
 accept a Paper based upon theoretical considerations alone, 
 the rule being to reject Papers unless they were founded 
 upon something that had actually been accomplished in a 
 concrete form. But he thought the large audience before 
 him that evening justified the Council in having admitted the 
 present Paper for theoretical discussion. With the American 
 Society of Mechanical Engineers no formal vote of thanks 
 was passed for a Paper. Thus time was saved for the techni- 
 cal discussion. The procedure of this Institution had latterly 
 followed the same course, and the appreciation of the present 
 Paper might now be indicated by a round of applause. 
 
 Mr. NEILSON said there were many methods of working 
 and devices suggested in the Paper which had been pre- 
 viously proposed, some of them over and over again. It was 
 extremely difficult in many cases to ascertain to whom the 
 credit was really due for those devices or methods of work- 
 ing, and therefore he had in most cases refrained from giving 
 acknowledgment to any one. He mentioned that fact in 
 case any one should feel aggrieved at not being given the 
 credit of something stated in the Paper. He wished to em- 
 phasize what he had said at the beginning of the Paper 
 that its most important object was to draw opinions from 
 other engineers who had studied the question, and especially 
 from those who had conducted experiments. 
 
80 
 
 THE GAS TURBINE. 
 
 Mr. HENRY DAVEY, Member of Council, said he had 
 made some experiments, but they had been on a small scale, 
 and, with a view to starting the discussion, he would just 
 state what his experience had been. First of all he would 
 like to tender his thanks to the author for bringing the Paper 
 before the Institution. It dealt with an important subject. 
 When a Paper was brought before engineers, and the theory 
 of the subject was put before them so lucidly as it had been 
 put by the author, it encouraged them to look more minutely 
 into the matter, and to see if some practical good might not 
 come out of it. The Paper was valuable as indicating the 
 
 Non-Return 
 
 \y**uy fluid 
 
 
 ira/ve ^ 
 
 ^~] 
 
 t 
 f 
 
 
 1 
 
 
 Air 
 
 E^ 
 
 
 _ : 
 
 ^ 
 
 
 1 
 
 nnr 
 
 
 zn: 
 
 
 Gas- 
 
 Hi 
 
 \ 
 
 
 4 \ 
 
 \ /&* 
 
 
 % 
 
 '/ , 
 
 \ 
 
 / 
 
 % 
 
 
 I 
 
 W//A 
 
 I 
 
 w// 
 
 & 
 
 FIG. 33. Experimental working fluid producer. 
 
 direction in which the possibilities of the problem lay. The 
 first part of the problem was that of producing in a practical 
 way the working fluid; then the temperature difficulties 
 followed, and they must be succeeded by the mode of appli- 
 cation to the turbine. 
 
 It was from the point of view of producing the working 
 fluid that he had made two or three experiments, on quite 
 a small scale. They were undertaken with an apparatus, 
 shown in Fig. 33, based on Cycle III, Case 2 (page 60). 
 A was a combustion chamber lined with fire-brick, and 
 above it was a small steel boiler consisting of a shell contain- 
 ing the water, having fire-tubes B extending to the smoke- 
 
THE GAS TURBINE. 81 
 
 box or vessel C. A non-return valve opened into the smoke- 
 box. When steam was raised above the pressure in the 
 furnace, it passed through the valve and mixed with the 
 products of combustion. The object of the apparatus was 
 to produce a mixture of steam and products of combustion, 
 the steam being highly superheated. The furnace was fed 
 by means of gas and air pumps, of the relative capacities for 
 delivering a burning mixture. The gas and air were deliv- 
 ered in separate pipes and came together just inside the 
 furnace. To start the apparatus it was necessary to get the 
 fire-brick lining into a red-hot state, so that it might main- 
 tain combustion; then the air and gas pumps were put to 
 work. He experimented with this apparatus on a small 
 scale, but he found many practical difficulties. He com- 
 menced with low pressures, intending to go on step by step 
 to higher and higher pressures; but what happened was that, 
 if the gases ceased to burn in the furnaces from any acci- 
 dental cause during one or two strokes of the pump, then a 
 big explosion frequently ensued. It seemed to him that the 
 system promised well, if the practical difficulties of keeping 
 the combustion constant could be overcome. If the diver- 
 gent nozzle was as efficient as one might imagine, it might 
 be that a turbine worked with a fluid produced from an 
 apparatus of that kind, at moderately low pressure, would 
 give a good efficiency; but the loss of efficiency in the air 
 and gas pumps would be considerable. His pumps were 
 reciprocating; and the loss would be probably much more 
 considerable if they were rotary ones. That would tell very 
 much against the scheme, as it was all a question of what 
 useful work could be got out of a given quantity of gas. 
 Instead of bringing air and gas from the furnace into the 
 combustion chamber to be burnt with a constant flame, he 
 altered the apparatus as shown at D. The gas and air united 
 together in the pipe e and pushed forward the products of 
 combustion of the last charge, forming a kind of cartridge 
 
 6 
 
82 THE GAS TURBINE. 
 
 in the pipe. Then the supply was cut off, and a sparking 
 plug at once fired the cartridge. With a moderate sized 
 brick-lined chamber A, a fairly constant pressure could be 
 kept up. The cartridge might be small and fired very fre- 
 quently, or three or four of them might be used and fired 
 intermittently, to avoid intermittent impulses on the tur- 
 bine. He had not got beyond what might be termed labora- 
 tory experiments, but had succeeded in getting the thing 
 to work fairly at a low pressure. He had not used the work- 
 ing fluids to propel a turbine, but he had taken them, in the 
 particular case he mentioned, into a reciprocating engine. 
 It was desirable to know what were the losses due to com- 
 pression both with reciprocating and rotary pumps, and 
 then came the very great difficulties of using high tempera- 
 tures in the turbine itself. With regard to the divergent 
 nozzle, there was, as far as he knew, scarcely any practical 
 information available. It was one of those subjects which 
 required to be thrashed out experimentally, and one of the 
 most important points in connection with the development 
 of the gas turbine. 
 
 Professor F. W. BURSTALL said it was a source of great 
 gratification to him to be able to attend and he would not 
 say discuss the Paper, because it was not a Paper which 
 admitted of a great deal of discussion speak on a subject 
 which was one of immense possibilities and potentialities. 
 It might not be for the present generation to inherit the gas 
 turbine, but it was probable that it would come to subse- 
 quent generations. If he might say one word of criticism in 
 regard to the Paper, he thought perhaps the author was a 
 little optimistic on the subject. He knew very well the 
 enormous difficulties which stood in the way of a commer- 
 cial gas turbine. From the days of Watt it had taken nearly 
 120 years to develop a comparatively simple thing like the 
 steam turbine, and, when it was considered that the gas 
 
THE GAS TURBINE. 83 
 
 turbine had not got to turn the energy of the fluid into 
 kinetic energy on the shaft, but had to compress the fluid, 
 to ignite the fluid, and deal with a fluid which was infinitely 
 more troublesome to deal with than steam, it would be 3een 
 what difficulties had to be overcome. He did not for a mo- 
 ment say that a gas turbine could not be made; perhaps he 
 would plead guilty to having schemed a good many himself. 
 There was no difficulty in making one; the difficulty was to 
 make a turbine which would produce any useful work at the 
 engine shaft, and for a reason which be would very soon 
 make clear. In the Otto cycle engine the compression was 
 produced almost entirely in the motor cylinder itself, and 
 therefore any heat which escaped from the charge of air and 
 gas into the walls of the cylinder was only there for a short 
 time, if at all. When the charge exploded, the cylinder walls 
 were heated, and cooled on exhaust, but at any rate they 
 were probably very much above the temperature of the com- 
 pression charge. It was clear, he thought, that no turbine 
 could ever be made to work on that principle. The com- 
 pression of the air and gas must take place outside the tur- 
 bine casing. Therefore one had to contemplate the possibil- 
 ity of economically producing that compression. 
 
 The author had alluded to the rotary compressor. The 
 rotary compressor, as far as he knew it and that was only 
 up to about 7 pounds or 8 pounds per square inch was 
 singularly inefficient, and he felt convinced that any rotary 
 compressor at present known would take the whole power 
 of the engine to compress the air and gas to begin with, 
 because its efficiency was probably not more than 0.4; and, 
 when it was considered that the negative work was 0.4, it 
 would be found there was about 30 per cent, of the work of 
 the engine absorbed in compression. Referring to the exper- 
 iment of Diesel, by taking a portion of the air during com- 
 pression and compressing it in an independent cylinder, a 
 very much higher efficiency had been obtained. On that 
 
84 THE GAS TURBINE. . 
 
 point he could answer the question with regard to the effi- 
 ciency which Diesel got in his compression cylinder. The 
 pressures were not correctly stated in the Paper, though 
 perhaps they were the pressures found by Mr. Ade Clark. 
 In his own Diesel engine the compression in the cylinder 
 was 35 atmospheres, and the compression on the blast used 
 for spraying the oil was 57 atmospheres. The efficiency in 
 the small compressor was not high, but then that small 
 compressor took such a very small charge that it did not 
 affect the general efficiency of the Diesel engine, or hardly 
 appreciably. 
 
 When dealing with the steam engine, one was dealing 
 with a fluid that could be carried over very large distances 
 without very much loss of heat. A 5 per cent, liquefaction 
 in a pound of dry steam would liberate quite a considerable 
 quantity of heat. In the gas engine, however, those condi- 
 tions were reversed. Any fall of temperature of the gas 
 came entirely from the internal energy of the gas, and there- 
 fore an explosion charge could not be trusted to remain in 
 contact with the metallic surface for any length of time 
 without an excessive loss of heat. From that, he inferred 
 that any form of Parsons turbine was not the most desirable 
 form for working with gases. The cooling in passing through 
 the comparatively small passages would be so immensely 
 rapid as to prevent anything like -a reasonable efficiency 
 being obtained. In the reciprocating engine there was a 
 rather different set of conditions. There was a large volume 
 of gas in a cylindrical vessel, and only the outer portion of it 
 became cooled at all, the centre core remaining always hot, 
 there being not sufficient time to pass its heat to the outer 
 walls. Hence the losses from cooling in a reciprocating 
 engine were not serious. 
 
 With regard to the possibility and he was afraid it was 
 the only possibility in the last instance of using the diver- 
 gent nozzle, the author had clearly indicated what were the 
 
THE GAS TURBINE. 85 
 
 limiting conditions for that divergent nozzle, namely, that 
 a velocity was required of about 5200 feet per second in 
 order to turn the heat energy of the charge into the kinetic 
 energy of the turbine vanes. Probably, at present, that was 
 not possible, but he saw no reason whatever to suppose that, 
 with the advances in metallurgical science, engineers would 
 not be in possession of a material which would stand the 
 high stresses and which might probably be lighter. It 
 seemed to him by no means impossible to get that, and, 
 granting the fact of a material, then the turbine was mate- 
 rialized almost at once. It was a matter of mechanical dif- 
 ficulties. As he had stated before, there was no reason to 
 suppose that it would be anything like economical, owing 
 to the very high ratio of the negative work, which did not 
 occur with the reciprocating engine. 
 
 The questions which the author raised with regard to 
 data were of course extremely important, particularly the 
 questions relating to the reciprocating compressor and the 
 expansion of hot gases in the divergent nozzle. The author 
 would probably know very well that any one of those exper- 
 iments was a most serious matter to undertake, and to get 
 results which were in any way commensurate with practical 
 work meant conducting the experiments on a large scale. 
 When experiments such as those were conducted on a large 
 scale the expenses were apt to be enormous, and therefore 
 one could hardly expect any private individual to give his 
 information for the good of the world at large. If the experi- 
 ments were made, no doubt they would be made by some 
 individual who, naturally, was seeking his own advance- 
 ment and who wished to produce a commercial article. 
 Whether such a thing was possible he did not know, but he 
 felt very strongly there was no possibility of advance in that 
 direction until more efficient compressors were obtained. 
 That was the real gist of the matter. In the last five years 
 he had had on the average from six to eight patents for gas 
 
86 THE GAS TURBINE. 
 
 turbines put before him, and in every case he had discovered 
 that they were perfectly vague on the subject of the nega- 
 tive work. If the author only succeeded in persuading engi- 
 neers that that was the real rock on which they split in devis- 
 ing a gas turbine, then he thought the object of the Paper 
 would be thoroughly accomplished. He thanked the author 
 for having brought such a subject before the Institution, 
 even if it were only in the form of a scientific investigation. 
 
 Mr. JAMES ATKINSON said the author had given a most 
 excellent Paper on the theory of gas turbines, and as a mat- 
 ter of fact the gas turbine simply existed now in theory. 
 Whether it would remain so or not was a question that might 
 take a good deal of deciding. With regard to the tempera- 
 ture which the author assumed could be put into a turbine 
 of the Parsons type, the author had taken it as 700 C. 
 (1292 F.). Professor Burstall had spoken about the prob- 
 abilities of a metal which would stand those temperatures 
 and the scouring action which must take place in a turbine, 
 but he, Mr. Atkinson, thought such a metal was a long way 
 from being discovered. Any metal that had iron in its com- 
 position commenced to oxidize at about 400 C. (752 F.), 
 or 500 less than the temperature given by the author. If 
 that temperature existed in the presence of free oxygen, 
 metal would begin to oxidize, and the scouring action which 
 took place in the turbine would very soon wash away the 
 blades. He thought the author might have made a slight 
 mistake in talking about superheated steam in steam-engines 
 and the efficiency being only a small percentage. As a mat- 
 ter of fact, he thought superheating in the steam engine 
 was more efficient than the actual steam working in the 
 steam engine. In a turbine, superheating the steam in- 
 creased the efficiency very largely, and it would increase the 
 efficiency of a steam engine equally as well, if it were not for 
 the fact that superheated steam went a very small way 
 through a steam engine, and in a triple-expansion engine it 
 
THE GAS TURBINE. 87 
 
 very rarely went beyond the high-pressure cylinder, if as far 
 as that. He expressed his admiration for the care, atten- 
 tion, and trouble the author had taken in producing his cal- 
 culations, which would be of very great use to engineers who 
 had to deal with the question. The Paper practically did for 
 the gas turbine what Mr. Dugald Clerk did for the gas 
 engine many years ago. 
 
 Lt.-Colonel R. E. B. CROMPTON, C.B., thought that the 
 two great difficulties which stood in the way of producing a 
 gas turbine had been fully stated by Mr. Davey and Pro- 
 fessor Burstall, the latter of whom had rounded up the sub- 
 ject so completely that it left little for others to say. There 
 was one point, however, which might perhaps be mentioned, 
 although he was uncertain whether it could be fairly dis- 
 cussed under the question of gas turbines. He had recently 
 been present at the Electrical Congress at St. Louis, and had 
 taken part in the discussions which followed two interesting 
 Papers, read by Mr. Emmett, of the General Electric Co., 
 of Schenectady, on the Curtis form of steam turbine, and by 
 Mr. Hodgkinson, of the Westinghouse Co., of Pittsburg, on 
 the Parsons form of steam turbine. Several speakers ap- 
 peared to think that one method of obtaining the maximum 
 efficiency in electrical energy from the energy in the coal 
 would be by using the gas producer and gas engine to deal 
 with the higher temperatures, and by utilizing the energy 
 remaining in the jacket-cooling water and in the exhaust of 
 the gas engine by raising steam in a suitable boiler and util- 
 izing this steam in a low-pressure steam turbine. In this 
 way it seemed possible to increase the present highest prac- 
 tical efficiency obtainable with the gas engine from the 
 figure of about 28 per cent, up to possibly 38 per cent. He 
 thought that there were no practical difficulties in the way 
 of the mechanical engineer in carrying out this development. 
 What would then remain to be considered would be whether 
 the economical advantage in increased efficiency would not 
 
88 THE GAS TURBINE. 
 
 be more than counterbalanced by the extra first cost and 
 afterwards by the maintenance cost of the extra plant that 
 would have to be added to obtain this increased duty from 
 the fuel. It appeared to him that, in many cases where fuel 
 was dear, it was quite probable that the answer would be 
 favorable, and that therefore such a proposition ought to 
 be seriously taken in hand by those interested in the ques- 
 tion. Speaking as a power engineer, he thought that this 
 combined use of gas engines and steam turbines was likely 
 to be of more immediate practical importance than the 
 remote possibility that the mechanical difficulties connected 
 with the gas turbine would be successfully surmounted. 
 
 Mr. H. M. MARTIN said that Professor Burstall had men- 
 tioned, in connection with rotary air compressors, 0.4 as the 
 highest efficiency he had come across. In a turbine compres- 
 sor built for the Farnley Ironworks, in which the rotary air 
 compressor was driven by a steam turbine, the combined effi- 
 ciency as determined by Professor Goodman was 61 per cent. 
 That was to say, the " air horse-power " was 61 per cent, of that 
 theoretically due from the steam. This meant that the effi- 
 ciency of the compressor per se was something like 80 per cent. 
 
 With respect to experiments on divergent nozzles, it 
 might be stated that some had been made by Professor 
 Stodola, who found that with a pressure drop of from 10J 
 to YO atmospheres, the loss was 20 per cent. The experiments 
 were not entirely satisfactory, as the measurements were 
 made by means of a small tube passing centrally down the noz- 
 zle, which acted as a sort of Pitot gauge. In fact, Professor 
 Stodola concluded that in the absence of the resistance caused 
 by this tube the loss would have been about 15 per cent.* 
 
 * The use of a thermo-couple would seem to afford a means of measuring 
 the loss in a divergent nozzle with the least possible disturbance of the flow. 
 With such an instrument the distribution of temperature throughout the nozzle 
 could be determined, and this known, together with the weight discharged per 
 second, the efficiency of the nozzle could be readily calculated by elementary 
 thermodynamics. 
 
THE GAS TURBINE. 
 
 89 
 
 There was another possibility with respect to gas tur- 
 bines which he did not know whether it would prove prac- 
 ticable to work out. Consider the well-known water ejector, 
 such as indicated in Fig. 34. High-pressure water entering 
 through a suitable- nozzle into a combining cone drew in and 
 
 FIG. 34. Water ejector. 
 
 carried with it low-pressure water to form a combined jet. 
 The efficiency claimed for the appliance was over 90 per 
 cent. He did not know whether it would be possible to work 
 an air ejector on the same lines. The method would be to 
 let a jet of products of combustion at high pressure develop 
 
 FIG. 35.- 
 
 Zero 
 -Divergent nozzle, showing varying pressure. 
 
 its full kinetic energy in a divergent nozzle. This would then 
 be utilized to draw in, say, four or five times its weight of 
 air by an ejector, and the combined jet having a correspond- 
 ingly reduced velocity and temperature would be utilized 
 to drive the turbine. Some experiments by Professor Stodola 
 went to show that the necessary suction could be obtained 
 
90 THE GAS TURBINE. 
 
 with an air-jet. Thus he found that if a divergent nozzle 
 designed for a high drop of pressure was used for a smaller 
 drop, the distribution of pressures in the nozzle was such as 
 indicated in Fig. 35 (page 89). Obviously, if holes were 
 put through the walls of the nozzle in the region of low pres- 
 sure, air would be drawn in, and Professor Stodola had 
 attempted to construct an ejector somewhat on these lines, 
 but so far had failed to obtain an efficient apparatus. 
 
 Professor ROBERT H. SMITH said that much useful and 
 interesting information, especially upon the outflow of high- 
 pressure fluids through varying shapes of nozzles, was to be 
 found in the book by Professor Stodola on " Steam Tur- 
 bines."* As the author in his Paper referred to a theory 
 that more than a certain critical velocity could not be ob- 
 tained in such nozzles, he might mention that Professor 
 Stodola's experiments showed velocities in the nozzles which 
 were considerably higher than the critical velocity calcu- 
 lated on the assumption which the author referred to; but 
 when the critical velocity was passed, extremely interesting 
 phenomena occurred, which had been suggested by the dia- 
 gram shown by Mr. Martin, Fig. 35 (page 89). The pres- 
 sure oscillated, not in time, but in distance along the axis of 
 the nozzle. Great waves, steady waves, of pressure existed. 
 The most interesting practical point resulting from these 
 experiments was that such waves of pressure did not serve 
 any useful purpose, and that the nozzle should be designed 
 so that they did not arise. The nozzle could be designed 
 for each discharge pressure so that the waves could be 
 
 * "Die Dampfturbinen," by Dr. A. Stodola, of Zurich, 1903, published by 
 J. Springer. Berlin. See pages 17-37. English translation by Dr. L. C. Lowen- 
 stein, published by D. van Nostrand Co., New York, and Archibald Constable 
 & Co., London. 
 
 Other recently published books on the subject are " Le Turbine a Vapore 
 et a Gaz," by Ing. Guiseppe Belluzzo, published by U. Hoepli, Milan, 1905; 
 and "Dampfturbinen, Entwickelung, Systeme, Bau u. Verwendung," by Wil- 
 helm Gentsch, published by Williams and Norgate, London, 1905. 
 
THE GAS TURBINE. 91 
 
 smoothed down and made to disappear. That, he thought, 
 was the most useful result of Professor Stodola's very long 
 and exceedingly interesting series ot experiments last year. 
 
 The difficulty to which Mr. Atkinson had alluded seemed 
 to be the greatest difficulty to be feared in gas turbines, 
 because the fluid resulting from combustion must be diluted 
 very greatly in order to keep down pressure and temperature 
 within practical limits, and the only cheap way of diluting 
 it was with an excess of air. There would be, therefore, a 
 very large excess of extremely hot air containing large quan- 
 tities of free oxygen passing through the narrow nozzles and 
 wheel buckets, and he fancied these would be rapidly burned 
 away. He was surprised to hear the statement that air com- 
 pressing could not be done at a greater efficiency than 40 
 per cent. He knew many reciprocating air compressors in 
 which it was possible to easily attain over 60 per cent, effi- 
 ciency. Mr. Henry Davey had sketched an apparatus he had 
 made upon a small scale, Fig. 33 (page 80). He, Professor 
 Smith, had spent two years in experimenting upon a similar 
 apparatus on a larger scale, which was designed for 600 
 pounds working pressure and 45 indicated horse-power, and 
 he had worked it at 130 pounds; but it was not for the com- 
 bustion of gas, but for that of oil. The difficulty had been, 
 so far, one similar to that which Mr. Davey referred to, 
 namely, the occasional sudden extinction of the flame and the 
 difficulty of controlling the flame, leading, in his apparatus, 
 to an inconvenient rush of the water over into the fire-tubes. 
 
 In the Paper there was a very interesting and important 
 remark (page 29) relating to Carnot's formula for the 
 efficiency of an ideal heat engine. This formula was not 
 intended for practical application to working plant, but was 
 framed only for the purpose of proving a theoretical prop- 
 osition, a proposition of immense value in the science of 
 thermodynamics. It was not a measure of efficiency appli- 
 cable to any heat engine which it was possible to construct 
 and work. The upper and lower limits of temperature might 
 
92 THE GAS TURBINE. 
 
 be greatly changed without to any extent changing the ther- 
 modynamic efficiency. It was an historical fact that every 
 engine that had been made and worked successfully had cut 
 down its upper temperature limit before it had obtained 
 practical success. 
 
 He was struck by the remark (page 76) that "With the 
 average user of a small engine, producing, say, 100 B.H.P.- 
 hours per week, a reduction of 10 in the initial cost is of 
 more consequence than a reduction of 2 cubic feet per B.H.P.- 
 hour in gas consumption." It occurred to him that that 
 was rather an extravagant estimate of the value of low 
 initial cost. It was usually his fate to argue in favor of 
 greater consideration of initial cost. Considering that 2 
 cubic feet per B.H.P. at 100 B.H.P.-hours per week was 
 200 cubic feet of gas per week, and taking gas at 2s. 6d. per 
 thousand, that meant 26s. a year; 26s. was 13 per cent, upon 
 10, and 13 per cent, was rather a high percentage to take 
 in the comparison. Perhaps a saving of 3 cubic feet of gas 
 per B.H.P. -hour in an engine of this small size working 
 with so bad a load factor might be more fairly comparable 
 with the advantage of 10 decrease in capital cost. On 
 page 76 the author suggested the use of gas turbines upon 
 motors for certain reasons which he gave, but surely the 
 difficulty of using any sort of gas engine on the motor was 
 not in the engine itself but in the weight and bulk of the 
 vessels that had to be carried to hold the gas. Some years 
 ago he worked that out, and he found it was practically 
 impossible, even with the high pressures that were used in 
 weldless steel gas cylinders, to carry any reasonable bulk of 
 gas in a motor car. Lower down on the same page the author 
 made a very useful remark, pointing out that in order to dis- 
 place electric driving the turbine would only require to possess 
 an efficiency greater than the efficiency of the central station 
 engine multiplied by the efficiencies of dynamo, mains, and 
 motor. If those four efficiencies were multiplied together a 
 very low compound efficiency would be obtained in the end. 
 
THE GAS TURBINE. 93 
 
 Mr. NEILSON, in reply to the discussion, stated that Pro- 
 fessor Bin-stall's remarks on rotary compressors and their 
 efficiencies had been already replied to by Mr. Martin (page 
 88). It by no means followed that an efficient rotary com- 
 pressor could not be used, although the turbine compressors 
 were not successful at high pressures. Turbine compressors 
 had been used up to about 80 pounds pressure or so, that 
 was, compressors arranged like a converse steam turbine. 
 It was quite reasonable to suppose that such compressors 
 would not be efficient at high pressures, as they had a great 
 number of stages, and the air was somewhat roughly handled 
 at each stage. He thought there was great room for inves- 
 tigation in the matter of trying to find an efficient rotary 
 compressor on another principle. Professor Burstall had 
 stated that a gaseous charge could not be allowed to remain 
 in contact with the metal for long without a great transfer- 
 ence of heat, and that gas was very different from steam in 
 that respect. In steam turbines with superheated steam 
 there was practically a gas at the high pressure end of the 
 turbine. Superheating had very much improved the effici- 
 ency of steam turbines, and it would be reasonable to sup- 
 pose that no great loss of heat occurred at the high pressure 
 end of the turbine. He believed experiments had been made 
 at the Manchester Technical School with a Parsons turbine, 
 and the temperatures taken at different stages of the expan- 
 sion, but he had not seen the figures. Those experiments 
 were with steam slightly superheated.* 
 
 Professor Burstall had referred to the possibility of get- 
 ting a material some day that would stand the high stresses 
 caused by high speeds. During the last few years there had 
 been very great improvement in that respect. He did not 
 think that a De Laval turbine, such as the 300-H.P. now 
 
 * The author subsequently saw the figures through the courtesy of Mr. 
 Mellanby; but the superheat was so slight that he considered the figures, 
 although interesting, were of no value in the present instance. 
 
94 THE GAS TURBINE. 
 
 built, could have been built 20 years ago, as there was no 
 material then to stand the enormous stresses. It had to 
 be remembered that the stresses varied pretty nearly as the 
 square of the velocity, so that, if the velocity was doubled 
 there was four times the stress. The experiments mentioned 
 at the end of the Paper certainly would cost a great deal of 
 money. Those with divergent nozzles would perhaps entail 
 least expense, as such experiments would not require a great 
 amount of apparatus. Professor Stodola's experiments were 
 exceedingly interesting, but it was impossible to take the 
 results as being very exact. Professor Stodola admitted 
 that his measuring apparatus was of such a nature as would 
 be apt to distort the results. For example, in his nozzle he 
 had a measuring tube which was very small, but still of a 
 diameter that was appreciable considering the minimum 
 diameter of the nozzle. Experiments on the effect of wind 
 pressure on different bodies had been made and they were 
 described before the Institution of Civil Engineers* by Dr. 
 T. E. Stanton. The experimenters discovered that in order 
 to get reliable results the models they put in the air duct had 
 to be very small compared with the dimensions of the duct. 
 That was with a comparatively small velocity of air. In Pro- 
 fessor Stodola's experiments the velocity of the fluid was enor- 
 mous, and the effect of any body in the centre of the jet would 
 presumably make a much greater difference. Therefore, while 
 Professor Stodola's experiments were interesting, further 
 investigations were required in connection with the matter. 
 With regard to the fluctuation of pressure mentioned by 
 Mr. Martin and Professor Smith as being found by Professor 
 Stodola, the steam in passing through certain nozzles fell in 
 pressure more than it should do, and then fluctuated up and 
 down in regular beats before it remained steady at the final 
 pressure. Fig. 36 illustrated the oscillations of pressure, 
 which were like those of a spring. In the diagram, heights 
 
 * Proceedings Institution of Civil Engineers, 1903, Vol. clvi, page 78. 
 
THE GAS TURBINE. 
 
 95 
 
 represented pressure, and horizontal distances represented 
 measurements along the nozzle. He did not think Professor 
 Stodola was the first to discover that. It illustrated what 
 he had said earlier, that it was very difficult in many cases 
 to tell to whom credit was due. The fact was mentioned in 
 a Paper* which Mr. Hodgkinson read in America some few 
 years ago, but even he did not seem to be the first to discover 
 the phenomenon. 
 
 Measurements 
 a.?or?g nozzfe 
 
 FIG. 36. Fluctuating pressure of steam passing through nozzle. 
 
 Mr. Atkinson had referred to the statement about super- 
 heating (page 30). Of course, superheating did increase 
 the efficiency, not owing to thermodynamic reasons, but be- 
 cause of the reduction in friction and for other reasons which 
 he need not go further into. Mr. Martin alluded to the pro- 
 posal to draw in an extra quantity of air to mix with the 
 products of combustion in order to cool down the tempera- 
 ture of the explosion and get a more reasonable temperature, 
 and he mentioned the proportion as 4 pounds of air to 1 
 pound of products of combustion. The lower temperature 
 suited the turbine, but it had to be remembered that in a 
 practical turbine the extraction of the available energy of the 
 
 * Proceedings, Engineers' Society of Western Pennsylvania, Nov., 1900. 
 
96 THE GAS TURBINE. 
 
 fluid was never complete, and that with this air-dilution 
 scheme, instead of discharging 1 pound of fluid with a certain 
 amount of unrecovered energy, it would be necessary to dis- 
 charge 5 pounds, thus throwing away five times the available 
 energy that would be otherwise thrown away. If water 
 instead of air were used to dilute the products of combus- 
 tion, the mass of fluid would not be increased to the same 
 extent. A somewhat similar proposal had been made as 
 regards steam turbines, namely, to dilute the steam from a 
 divergent nozzle with water or other fluid in order to bring 
 down the velocity so as to enable the velocity to be utilized 
 conveniently in the turbine wheel. This would allow the 
 wheel to run at a lower speed. Instead, however, of dis- 
 charging 1 pound of fluid with a certain final velocity, the 
 scheme would involve the discharging of, say, 5 pounds or 
 6 pounds. 
 
 Professor Smith had referred to the question of the rela- 
 tive advantage of low initial cost and low working expenses 
 (page 92). The case mentioned by Professor Smith he had 
 worked out, but he forgot what the saving amounted to as a 
 percentage of the extra capital. Taking it as 13 per cent. 
 (as given by Professor Smith), he thought that with a small 
 turbine the reduction in initial cost would be of more im- 
 portance than the 13 per cent, saving. It was not like put- 
 ting money into a building. No one could say how long the 
 engine was going to be up-to-date, and no one knew how 
 long it would be before another engine was required. More- 
 over, small users generally wanted money at the time. He 
 thanked the Institution for the vote of thanks that had been 
 accorded to him. 
 
 Communications. 
 
 Mr. EDWARD BUTLER wrote that perhaps the greatest 
 obstacle in the way of the production of an efficient gas tur- 
 bine namely, "the very low reactionary value and destruc- 
 
THE GAS TURBINE. 97 
 
 tive property of a highly heated gas, as compared with steam 
 of the same pressure" could be removed by a combination 
 of apparatus in which gas and air would be supplied to a 
 mixing chamber communicating with a series of long explo- 
 sion chambers with siphon formations for receiving water. 
 The modus operandi being for the series of explosion cham- 
 bers to be charged with explosive mixture and fired in rota- 
 tion, the effect of the explosions would force out the water 
 contained in the well or siphon formation of each tube in suc- 
 cession and this water could be directed on to the wheel of a 
 water turbine. Succeeding charges of gas mixture would 
 enter automatically from the mixing chamber through non- 
 return valves immediately after each explosion. The only 
 operating mechanism required would be means for quickly 
 filling the siphons with water and for igniting the mixture 
 in the series of five or six explosion tubes, by which an almost 
 continuous stream of water would be directed onto the tur- 
 bine wheel at high velocity, thus avoiding all the trouble 
 that would follow from any method of employing a turbine 
 wheel actuated direct by the hot rarefied gases of combus- 
 tion. He offered this solution of a gas power turbine, after 
 having had some experience in the working of a steam tur- 
 bine worked somewhat on these lines, namely, with water 
 accelerated by a steam injector and boiler. 
 
 Mr. DUGALD CLERK wrote that he had read the Paper 
 with much interest, and he quite sympathized with the 
 author in his effort to clear up in a preliminary way the many 
 abstract points which required consideration before the 
 practical problem of the gas turbine could be attacked with 
 any chance of success. In view of the wonderful results 
 obtained by the Hon. C. A. Parsons and his imitators with 
 the steam turbine, it was very natural that the attention 
 of engineers should be called to the problem of the gas tur- 
 bine. Belonging to the older school of engineer inventors 
 and having become somewhat mentally fatigued by the 
 7 
 
98 THE GAS TURBINE. 
 
 numerous difficulties experienced with cylinder and piston 
 gas engines in every stage of their progress, it might be that 
 the writer was less likely to take a hopeful view of the gas 
 turbine problem than a younger engineer, whose life and 
 practical engineering difficulties were still before him. 
 Whether this were so or not, he must confess that his view 
 of the future of the gas turbine was not so hopeful as the 
 author's. The difficulties appeared to be even greater than 
 the author had apprehended. Mr. Neilson rightly laid a 
 certain stress on the relatively low efficiencies possible in tur- 
 bine compressing plants; but he did not think he had laid 
 sufficient stress upon the similarly low efficiency of expan- 
 sion in steam turbines of existing types. No doubt the 
 author had clearly in his mind the fact that there were spe- 
 cial advantages in the steam turbine which counterbalanced 
 the disadvantages of relatively inefficient expansion. The 
 steam turbine had, to begin with, the great advantage of 
 practical freedom from losses due to initial condensation. 
 It had the further great advantage of expansion to a much 
 lower pressure point than could ever be effectively attained 
 with a reciprocating steam engine. These two advantages, 
 in the case of steam, undoubtedly enabled the Parsons tur- 
 bine to reach figures of economical steam consumption which 
 were not touched at all by ordinary reciprocating steam 
 engines, and were very rarely reached even in special recip- 
 rocating steam engines using high superheats. These ad- 
 vantages of no initial condensation and largely extended 
 expansion were advantages special to steam; they were not 
 advantages which would be found in using a turbine con- 
 struction even for the purpose of expanding high pressure 
 cool gases, such as compressed atmospheric air. In an ordi- 
 nary gas engine cylinder, the efficiency of compression and 
 expansion of the gaseous contents of the cylinder (without 
 combustion at all) was so high that practically no difference 
 could be detected between the compression and expansion 
 
THE GAS TURBINE. 99 
 
 lines of the air within the cylinder during successive com- 
 pression and expansion strokes. Practically, what one might 
 call the efficiency of gaseous compression and expansion in 
 an ordinary gas engine cylinder was certainly not less than 
 99 per cent. This, undoubtedly, enabled engines to be oper- 
 ated economically with large proportionate values of nega- 
 tive work. If, however, the efficiency of compression had 
 been, say, 70 per cent. a figure higher than any turbine air- 
 compressor could give at present and the efficiency of ex- 
 pansion 80 per cent. a figure higher than any existing tur- 
 bine could give operated by expanded compressed air even 
 then the united efficiencies would only amount to 56 per 
 cent. That is, assuming the gases to be heated for the ex- 
 pansion period, the loss to be made good to bring the dia- 
 gram up to unity would be 44 per cent. ; that is, if the volume 
 of the gas were increased by heating from 56 to 100, then the 
 expansion would only be sufficient to produce a diagram 
 which would keep the engine running if it were quite friction- 
 less. In the writer's view, therefore, the problem to be faced 
 required not only the invention of a turbine air compressor 
 compressing, say, to 200 pounds per square inch, with an 
 efficiency of something nearly 90 per cent., but it also re- 
 quired the invention of a turbine motor engine which would 
 give a like efficiency of expansion of the gases so compressed. 
 High efficiencies of compression and expansion of gases, such 
 as air, were undoubtedly theoretically possible; but he knew 
 of no turbine yet in existence which would give any efficien- 
 cies such as he had indicated. Until these efficiencies were 
 obtained, it seemed to the writer impossible to design any 
 gas turbine having a chance of success. 
 
 The author had very accurately made clear the point 
 that, in some types of turbine, low temperature, and there- 
 fore large negative work, was necessary for working condi- 
 tions. He himself could not help thinking that, although 
 the high temperatures he proposed 2000 C. (3632 F.) 
 
100 THE GAS TURBINE. 
 
 and so forth certainly diminished negative work, they yet 
 introduced for the inventor a much worse condition from 
 the points of view of durability of blades, and heat losses if 
 the blades were kept cool. There seemed to be no possibility 
 whatever of working a turbine at temperatures approaching 
 2000 C. The De Laval type was proposed by the author 
 to enable temperatures to be reduced while the energy was 
 maintained in the shape of velocity of the moving gases. 
 He could not help thinking that any scheme which allowed 
 of such reduction would also cool the expanding gases by the 
 very conditions required in the constructions of the expand- 
 ing nozzle. He did not see any immediate future for a gas 
 turbine, except in possibly utilizing the exhaust gases from 
 reciprocating engines, which at present were liberated under 
 considerable pressures. He feared that the attack of the 
 problem of an efficient turbine air expander was more 
 within the province of the practical inventor than the scien- 
 tific investigator. He would be much interested to hear 
 any ideas Mr. Neilson might have in the direction of pro- 
 ducing efficient turbine compressors and expanders. 
 
 He hoped that the author would not be discouraged at 
 the somewhat pessimistic tone of these remarks. He had 
 often thought over this subject, and had not as yet seen any 
 hope of getting as good results, either as to economy or dura- 
 bility, with gas turbines as were obtained with reciprocating 
 engines. 
 
 Mr. W. J. A. LONDON wrote that, in studying the' Paper, 
 one could not help feeling impressed by the thoroughness 
 with which the author had dealt with the various possible 
 cycles to which engineers were to look in the future gas tur- 
 bine. The most interesting point, however, in the writer's 
 opinion, was the way in which the author's investigations 
 showed the difficulties to be met with in the successful de- 
 signing of such machines. It was a pity that he had not 
 
THE GAS TURBINE/ ; 
 
 attempted to consider from a more practical standpoint 
 several of the cycles set forth; had he done this, the writer 
 could not help feeling confident that he would have omitted 
 several of the cycles referred to. 
 
 On page 32 a reference was made to a combustion 
 chamber where the burning gases could rest before entering 
 the turbine. He thought this would be a great source of 
 loss, because this chamber would have to be cooled, and, to 
 allow the gases to remain in a chamber with cooled walls, a 
 great amount of efficiency would naturally be lost. Further, 
 on the same page, the author referred to a mixture of air 
 and fuel being always supplied to the flame with a velocity 
 greater than that of the propagation of the flame. This, in 
 the writer's opinion, would be a very difficult thing to do, 
 especially at the high outgoing velocities attending the ex- 
 panding gas. On page 39 the author referred to the Par- 
 sons turbine with steel blades for working at about 700 C. 
 (1292 F.). This temperature was very high, and it was 
 doubtful whether any blades with sharp edges could be 
 made to withstand it for any length of time. 
 
 In Cycle I, Case 2 (page 39), the author proposed to 
 circulate water in the revolving part. If he would consider 
 the difficulties to be met with in circulating water in the 
 revolving chamber, the writer thought he would find that it 
 would be a more difficult problem than he considered. The 
 water would be thrown against the outer walls, which was 
 the point to be desired, but whether the circulation could be 
 arranged without seriously interfering with the running bal- 
 ance of the machine was another question. Further, even 
 if the rotating drum were cool, this would not effectively cool 
 the blades, and to do this as the author suggested, by making 
 the blades hollow, would require very large blades with 
 internal pipe arrangements for ensuring circulation. With 
 his device he claimed that 2000 C. was allowable; this cor- 
 responded to 3632 F., or above the melting-point of steel, 
 
KJ2 1 THE GAS TURBINE. 
 
 so that it was hardly likely that any sharp edges could be 
 expected to remain on the blades. For mechanical reasons 
 he did not think one should look for the gas turbine on the 
 Parsons principle, but more on the lines of the De Laval, as 
 set forth in Cycle I, Case 3a (page 48), where the gas was 
 expanded in nozzles, and the temperature reduced before 
 the working fluid came into contact with the moving blades. 
 
 The author referred (page 56) to a combination of 
 reciprocating engines and turbines working with gas on the 
 principle now being adopted in connection with steam 
 engines and turbines. It would be interesting to learn what 
 gain this system had over reciprocating engines. The idea 
 in the steam combined system was that the reciprocating 
 engines were perhaps more economical at lower speeds than 
 the normal. The steam turbine working non-condensing 
 had great difficulty in competing with the reciprocating 
 engine, but when working condensing the conditions were 
 different, and the turbine was more capable of taking care 
 of what might be called the tail end of the expansion curve. 
 These conditions when working condensing did not come 
 into account when working non-condensing, as in the case 
 of a gas turbine, so that it was doubtful whether the com- 
 bined system would be even as economical as a gas engine of 
 the ordinary reciprocating type. 
 
 The system which the writer thought offered the greatest 
 possibilities was on the principle of Cycle III, Case 1 (page 
 59), where the steam jet was inserted into the gas, thus 
 utilizing the expansion of the superheated steam and reduc- 
 ing the temperature of the working fluid to within practical 
 limits. The ideal efficiency of 0.33 and the ratio of negative 
 work of gross work of 0.41 did not look as satisfactory as 
 some of the other cases set forth, but it was undoubtedly 
 more practical, and it seemed more than likely that the solu- 
 tion of the difficulty would be found in this direction. Re- 
 ferring to the author's remarks (page 69) on the velocity 
 
THE GAS TURBINE. 103 
 
 of gases issuing from diverging nozzles, there must be a 
 limiting velocity for the flow of gas through an orifice, this 
 velocity being at a point where the weight discharged multi- 
 plied by the velocity due to difference in pressure was a 
 maximum. He thought this point had been experimentally 
 investigated, and it had been found that the point of maxi- 
 mum discharge was when the ratio of internal to external 
 pressure varied between 0.5 and 0.6 according to the nature 
 of the gas. This pressure of 0.5 to 0.6 initial pressure would 
 be at the throat, or, in other words, at the smallest part of 
 the nozzle, and the velocity had been found to be somewhere 
 about 1500 feet per second. However much the difference 
 of pressure was increased, this law remained good, and no 
 more gas would pass through the orifice, nor would the 
 velocity at the throat increase. The terminal velocity, how- 
 ever, would be greater in proportion to the drop in pressure. 
 
 Mr. E. KILBURN SCOTT wrote that he thought the author 
 struck the keynote of the present position when he suggested 
 (page 78) that a thorough set of experiments should be 
 made and the results published. The information so ob- 
 tained by the engineering world would be extremely valu- 
 able, whatever the price, as it would save much time and 
 overlapping if the several fundamental considerations in- 
 volved could be investigated straight away. It was essen- 
 tially a matter for some public body, and he suggested that 
 the Institution of Mechanical Engineers should undertake 
 the work. The tendency of the times was well shown by the 
 working arrangement between the General Electric Co. of 
 America and the Allgemeine Elektricitats Gesellschaft of 
 Berlin, whereby experimental data, drawings, etc., were now 
 the common property of both firms. This arrangement was 
 come to with the idea of cutting down expenses, and at the 
 same time facilitating the progress of experimental research 
 and of new designs. 
 
104 
 
 THE GAS TURBINE. 
 
 The high temperatures and high velocities involved in 
 working turbines with gas, and the fact that the hot gases 
 came so intimately into contact with the metal parts, called 
 for an investigation as to the most suitable metal to employ. 
 
 Separate blades, caulked in as with the Parsons turbine, 
 would hardly do for gas turbines, as this construction was not 
 even satisfactory at the temperature of highly superheated 
 
 PRINCIPLES OF TWO GERMAN ARRANGEMENTS OF TURBINES. 
 
 ft 
 
 1 I I 
 
 <Steam or Ga*s. 
 FIG. 37. Principle of Bucholz turbine. 
 
 This <Sfej/7t or G<ss 
 space gr<3dua//y 
 contracts 
 
 FIG. 38. Principle of Patschke turbine. 
 
 steam. New turbines were, however, coming forward 
 which had not so delicate a constitution. The Buchol/,* 
 for example (Fig. 37) had a series of metal discs with 
 holes drilled through them at an angle, and against which 
 the working medium impinged; and the ingenious Patschke 
 turbine (Fig. 38) had the rotating part simply in the form of 
 a drum with a number of holes drilled in the rim. With such 
 
 * "Electrical Review," January 8th, 1904, page 66. 
 
THE GAS TURBINE. 105 
 
 solid construction, exceedingly high velocities could be at- 
 tained, and temperature had very little effect, whilst these 
 two designs had the further advantage of being reversible. 
 
 The question of water-supply to the heated metal was 
 mainly a matter of mechanical design, and he thought it was 
 less difficult to arrange water-circulation in the interior of the 
 rotating part of a turbine than it was to get it into the pistons 
 and exhaust valves of a large slow-running power gas engine. 
 
 In comparing gas turbine with electric motor driving, it 
 must be remembered that practically all turbines ran at 
 speeds which were unsuitable for driving shafting and ma- 
 chine tools, and that the latter were par excellence the very 
 things for which the electric motor was most suitable, by 
 reason of the ease with which the motor could be started 
 and stopped or its speed varied, and also moved about for 
 use with portable tools. A pipe containing gas could never 
 compete with an electric cable for such work. As a matter 
 of fact, the author need not worry as to uses for gas tur- 
 bines; only let it be clearly demonstrated that they were 
 commercial, and they would become the favorite prime- 
 mover for driving electric generators, and in that direction 
 alone there was scope enough. Any one prophesying two 
 years ago that the Curtis turbine would attain its present 
 position in the United States would have been thought a 
 dreamer, yet the writer had recently seen five stations 
 equipped with 8000 horse-power Curtis sets running with- 
 out a hitch. So far as power station work in the United 
 States was concerned, steam turbines were preferred to 
 reciprocating steam engines, and he believed that one day 
 gas turbines would come to the front just as rapidly. 
 
 Mr. GEORGE A. WIGLEY wrote that he thought it was 
 remarkable that during the discussion no direct reference 
 was made to oil turbines or engines, with the exception of a 
 few remarks on the Diesel engine. One of the first problems 
 
106 THE GAS TURBINE. 
 
 to be considered with regard to a gas turbine was the pro- 
 duction of the gas, and if it were necessary to supply a gas- 
 producer even of the modern suction type, a pump for com- 
 pressing this gas and a mixing chamber, it was useless to 
 compare such a plant with an electric motor, irrespective of 
 maintenance costs, but if a mineral oil were substituted for 
 the gas the problem immediately assumed a different aspect. 
 In Cycle III, Case 2 (page 61), the author clearly 
 pointed out the advantages to be obtained by the use of 
 water or steam with the gas, and this would, of course, hold 
 good with oil as fuel, the ratio of negative work to gross work 
 being decreased, but it was this question of negative work 
 which, in the writer's opinion, was most important. In oil 
 (mineral oil) they had what was practically a gas compressed 
 to liquefaction all ready to hand in a very convenient form, 
 and in such a condition that, when intimately mixed with the 
 proper amount of oxygen, it formed a gas which was ready 
 to give up a great proportion of the work which would be or 
 had been necessary to compress it to liquefaction. So that 
 they had what was to all intents and purposes a gas already 
 compressed; and as this gas had had work done upon it far 
 greater in amount than was necessary for the required im- 
 mediate object, it followed that the oil, when fired, gave up 
 this work, and not only reduced the negative work given out 
 by the turbine to the extent of that required to compress 
 the gas to the necessary pressure, but also reduced the excess 
 work required to liquefy it, less that corresponding to the 
 latent heat of vaporization. The result was that the nega- 
 tive work calculated for gas was greatly reduced if oil were 
 used, the extra work being given in the convenient form of 
 heat and kinetic energy of the particles. 
 
 Mr. NEILSON, in reply to the written communications, 
 wrote that Mr. London had raised the question as to whether 
 if the burning gases were allowed to rest a short interval in a 
 
THE GAS TURBINE. 107 
 
 combustion chamber before being taken to the turbine, there 
 would not be a great loss of heat. He (the author) agreed 
 that this question deserved consideration, but he did not 
 think that there need be anything like as great a loss of heat 
 to the walls of this combustion chamber as to the walls of the 
 combustion chamber of a reciprocating engine, which re- 
 quired to be kept much cooler than would be necessary in 
 the case of a turbine. As regards the last remarks made by 
 Mr. London, it must be noted that the highest velocity ob- 
 tainable when a gas issued from a simple orifice was not 
 necessarily the highest velocity that could be obtained 
 when a divergent nozzle was placed on the delivery side of 
 the orifice. 
 
 With regard to supplying the air and gas at a velocity 
 higher than the velocity of the propagation of the flame, 
 mentioned by Mr. London, he did not think that this was 
 really a difficult problem with the gas turbine. With a re- 
 ciprocating engine it was complicated, but with a gas tur- 
 bine having a continuous flow of gas it should not be difficult. 
 With a mixture of a given gas and air in constant propor- 
 tion, the flame would have a given rate of propagation; and, 
 if care were taken to supply the air and gas to the combustion 
 chamber at a greater velocity, there should be no firing back. 
 Mr. London had referred to the proposal to circulate water 
 in the turbine. It was said in the Paper that this had been 
 proposed, but he (the author) did not propose it himself, 
 and he had not much faith in it. 
 
 He agreed with Mr. Kilburn Scott as to the benefit to be 
 derived from experiments made by a public body. 
 
CHAPTER III. 
 
 THE DISCUSSION BEFORE THE SOCIETY OF CIVIL ENGINEERS 
 
 OF FRANCE. 
 
 ONE of the most complete papers upon the subject of 
 the gas turbine which have yet appeared was presented 
 before the Society of Civil Engineers of France, on February 
 2, 1906, by M. L. Sekutowicz, this covering a thermody- 
 namic treatment of the question as well as a study of the 
 gas turbine from a mechanical point of view. This paper 
 elicited an animated discussion which included the views 
 of a number of eminent specialists who had devoted atten- 
 tion to the subject, the whole appearing in the Memoires 
 of the Society for February 1906. 
 
 After a brief historical introduction, M. Sekutowicz 
 proceeds with his thermodynamic study as follows : * 
 
 As is the case with all heat motors, the gas turbine in- 
 volves the action of a given mass of a determinate fluid, dur- 
 ing a unit of time, between certain limits of pressure and 
 temperature. 
 
 In order that this action may be subjected to computa- 
 tion it is necessary that each of its phases be considered as 
 simple; that is, characterized by the constancy of some one 
 of the independent variables upon which the state of the 
 fluid depends at each instant, such as: volume, pressure, 
 temperature, and entropy. 
 
 The phase of compression may be adiabatic or isothermic. 
 The introduction of the heat may he isothermic (as in the 
 Carnot cycle and its derivatives), or isobaric (combustion 
 under constant pressure), or isopleric (combustion under 
 constant volume: explosion). The expansion will be adia- 
 batic. Finally the rejection of the heat at the lower tem- 
 
 * Mem. Soc. Ing. Civ. de France, Feb. 1906, p. 204 et seq. 
 108 
 
THE GAS TURBINE. 109 
 
 perature, can hardly be other than isobaric or isothermic 
 (the Carnot cycle). 
 
 The first two phases are the most important, for the last 
 two can be only partially realized in practice. 
 
 In fact the expansion, in the case of the gas turbine, 
 will be almost rigorously adiabatic, while the rejection of 
 heat can only take place at constant pressure to a regenera- 
 tor; or, what amounts to the same thing as regards the cycle, 
 by the non-closure of the cycle. 
 
 We then have six principal variants. Besides these we 
 may examine the cycles of Stirling and of Ericsson, in which 
 the adiabatics of Carnot are replaced by the isodiabatics and 
 the regenerative cycles derived therefrom. Finally we have 
 to study the effects of the injection of water or of cold gases. 
 
 Heat motors are too frequently studied with regard to 
 the extreme temperature limits of the cycle. There results 
 a misleading comparison with the Carnot cycle, or at least 
 a comparison which fails to indicate the actual development. 
 What really concerns us is the value of the thermal efficiency 
 />, the mechanical efficiency >?, and the total useful effect 
 represented by their product. 
 
 Now, as we shall see, these efficiencies are not always 
 limited by the extreme limits of temperature. Other factors, 
 no less important, must be considered. In the present case 
 these are: the limits of pressure, the ratio of the work of 
 compression to the useful work, and the quantity of useful 
 work produced by a kilogramme of air. 
 
 The lower temperature limit is usually that of the atmos- 
 phere, and may be taken as about 300 C. absolute. The 
 final temperature of the expansion, however, will generally 
 be much higher. This value is most important, since it is 
 the temperature of the gases delivered upon the rotating 
 metallic portion of the turbine. 
 
 We will assume that the turbine wheel can be so con- 
 structed as to stand a temperature of 700 C. absolute, 
 
110 THE GAS TURBINE. 
 
 without injury. This fact has been fully demonstrated in 
 practice. In all that follows, therefore, we shall consider 
 the temperature at the end of the expansion as being 700, 
 except when examining the influence of variations of this 
 factor. 
 
 The upper temperature limit is determined largely by 
 the heat resistance of the refractory material used, not only 
 for the lining of the combustion chamber, but also for the 
 construction of the nozzle in which the expansion takes place. 
 There are now available such substances as carborundum, 
 which are capable of resisting the highest temperatures 
 developed. 
 
 Under these circumstances the maximum temperature 
 is limited by the following conditions: 
 
 1. It must be such that, with the degree of expansion 
 attainable, the final temperature of expansion shall not 
 exceed 700 C. absolute. 
 
 2. It must be attainable by the combustion of ordinary 
 fuels with a sufficient quantity of air to insure a complete 
 combustion. 
 
 As is well known, combustion under constant volume 
 produces a much greater elevation of temperature than is 
 caused by combustion at constant pressure. Besides this, 
 when the compression is adiabatic the temperature of the 
 gas is raised to a greater or less degree before combustion, 
 this effect being added to the temperature of combustion. 
 For a compression of 30 atmospheres the temperature will 
 reach 800 C. absolute. 
 
 Thus, illuminating gas requires 5.5 times its volume of 
 air in order to enable perfect combustion to be effected. 
 If, in practice, we assume that 6 volumes of air are required, 
 1 kilogramme of the mixture will evolve 574 calories. Under 
 these conditions, starting from the ordinary atmospheric 
 temperature, the combustion, if conducted at constant 
 volume, would produce a temperature of 2450 C., absolute. 
 
THE GAS TURBINE. Ill 
 
 If the combustion takes place under constant pressure the 
 temperature would be about 2000 absolute. With acety- 
 lene this limit may be extended. With other gases slightly 
 different results are obtained, as shown hereafter. 
 
 The upper limit of pressure is determined almost wholly 
 by practical considerations. As we shall see, it is without 
 direct influence on the velocity of discharge. If the com- 
 pression is isothermic it may be increased without increas- 
 ing the ratio of the work of compression to the useful effect. 
 We may consider compressions of 40 to 60 atmospheres 
 (600 to 900 pounds per square inch) as entirely admissible, 
 both with respect to the compressor and the combustion 
 chamber. 
 
 The lower limit of pressure will be that of the atmos- 
 phere if the exhaust is made into the open air, or it may 
 be a very low pressure, approaching a perfect vacuum if 
 the discharge is made into a space provided with an air 
 pump. 
 
 The ratio of the work of compression to the useful effect 
 
 C7 c 
 
 =- plays an important part in the gas turbine, because 
 cr u 
 
 the compressor, being necessarily distinct from the turbine, 
 its mechanical efficiency T} C (which includes that of the 
 transmission mechanism by which it is driven) has a very 
 marked influence upon the efficiency of the entire machine. 
 When this ratio is very high the importance and bulk of the 
 compressor occasions some practical inconveniences. A 
 ratio approaching unity is practically prohibitory. 
 
 Finally, the quantity of useful work produced per kilo- 
 gramme of gas gives a measure of the specific power of the 
 machine. 
 
 These are the principal elements which form the criterion 
 in the discussion which follows. But, before commencing 
 this discussion it remains for us to indicate the hypothesis 
 and the numerical data upon which it is based. In this 
 
112 THE GAS TURBINE. 
 
 connection it may be noted that all the computations have 
 been made with the slide rule, this giving a degree of pre- 
 cision quite within the limits of error of the premises. 
 
 We begin with the simple and well-known laws of ther- 
 modynamics as used by M. Witz in his classical labors upon 
 the gas engine. As a first approximation we assume the 
 specific heat as constant, the value for hot air at constant 
 pressure C P being taken at the usual value 0.2375, and the 
 specific heat at constant volume c v being taken as 0.1686, 
 so that their ratio is 
 
 r -^-i.4i. 
 
 c v 
 
 The specific constant of air is taken at 29.3. We neglect 
 the contraction, which may attain 5 per cent. 
 
 For the vapor of water we adopt the value C P = OAS. 
 It is only in special cases that we shall take into account 
 the variation of specific heat with the temperature, using 
 the linear formula of M. Lechatelier, C = a+bT. 
 
 We shall retain for adiabatic expansion the formula of 
 Laplace or Poisson:pv y = constant. 
 
 The modern exponential formulas are very interesting, 
 but their use would burden this discussion to such an extent 
 as to render comparisons impossible. 
 
 After having cleared away the general discussion we 
 shall return to the special modifications to which our results 
 must be submitted if we desire to follow the laws of gases 
 to a higher degree of precision. 
 
 The Mechanical Efficiency of the Gas Turbine. 
 
 We have to consider two distinct machines the turbine 
 and the compressor, each having its own efficiency. More 
 or less of the heat energy which is transformed into work 
 in the turbine, with the particular efficiency of this machine, 
 is expended in driving the compressor, and the available 
 power is only the difference between the two values. 
 
THE GAS TURBINE. 
 
 113 
 
 Thus, let Q be the quantity of heat furnished by the com- 
 bustion of a kilogramme of gas, q the heat rejected with the 
 exhaust, and p the total thermal efficiency, equal by defini- 
 
 tion, to 
 
 Q q 
 Q 
 
 If all the losses are reduced to losses of a 
 
 thermal order there will be produced in work : 
 
 c a a 
 
 FIG. 39. Useful work and work of compression. 
 
 Now let ^c be the theoretical work of compression per 
 kilogramme of air (this being computed hereafter for each 
 case), and let TJ C be the mechanical efficiency of the com- 
 
 cy^ft 
 
 pressor, defined in such a manner that is the quantity 
 
 of work delivered to the shaft of the compressor to compress 
 one kilogramme of air.* 
 
 On the other hand each kilogramme of air produces in 
 the motor turbine a quantity of " indicated" work, equal, 
 by definition, to the sum of the net available work on the 
 shaft and all the passive resistances (friction, etc.). This 
 
 * In the case of isothermal compression the theoretical amount of work 
 may be computed by the law of Mariotte. If this law is not exactly followed 
 and if the gas is slightly heated by the compression the corresponding energy 
 is included in the mechanical losses and in the value of n c . 
 
 8 
 
114 
 
 THE GAS TURBINE. 
 
 work ^T, which we shall compute for each case, is equal to the 
 useful work 'tfu, denned above, increased by the work of com- 
 pression ^c, as we shall see by examining the diagram, Fig. 39. 
 
 i.oo 
 
 080 
 
 0.60 
 
 040 
 
 020 
 
 Tc 
 TT 
 
 2.0 
 
 1.8 
 
 1.6 
 1.4 
 12 
 
 1.0 
 
 0^ 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 / 
 
 Of 
 
 
 
 
 7 
 
 0.4 
 02 
 
 
 
 / 
 
 ' 
 
 
 
 / 
 
 
 ^x 
 
 x 
 
 
 1 
 
 0.4 
 
 0.6 
 
 0.8 
 
 1.0Q= 
 
 FIG. 40. Values of efficiency in terms of tempera- 
 ture ratio. 
 
 0.2 0.4 0.6 0.8 1 I 
 FIG. 41. Values of tem- 
 perature ratio in terms of 
 efficiency. 
 
 The indicated power furnished by the motor turbine, 
 per kilogramme of air, is: 
 
 *& indicated = ^u + ^c. (1) 
 
 If we call the mechanical efficiency of the turbine yt the 
 effective work on the shaft will be: 
 
 ^ effective = yt^u + C"c) . (2) 
 
 It follows that the effective work available upon the com- 
 mon shaft of the turbine and the compressor, supposing 
 them to be connected thus as one machine, will be: 
 
 1^ 
 
 *C" net work = 
 
 (3) 
 
 If the compressor be driven through any intermediate 
 transmission the efficiency of this transmission should be 
 included in T) C . 
 
THE GAS TURBINE. 115 
 
 The mechanical efficiency of the two machines together 
 will then be: 
 
 net work / IX^c , > 
 
 The mechanical efficiency then disappears when 
 
 For example, taking 7)t = r} c the efficiency will be zero for 
 
 which gives the following values: 
 
 7^ = ^ = 0.5 0.6 0.7 0.8 
 ^ = 0.34 0.56 0.96 1.78 
 
 We shall see later on that under the actual conditions 
 of turbine construction yt will have a value of about 0.7. 
 
 As for the efficiency of the compressor i) c , this will be 
 for improved reciprocating machines 0.8 to 0.9. 
 
 Since it is necessary, however, to introduce a speed- 
 reduction transmission, i) c will be reduced to 0.75 to 0.85. 
 
 If the compressor is made of the multicellular turbine 
 type, permitting direct connection, it is possible that the 
 efficiency will be in the neighborhood of 0.6 to 0.7. 
 
 If we take rj t = rj c = 0.7 we see that the mechanical effi- 
 ciency will be totally annulled for 
 
 < TTc = c Cw, about. 
 We have in general 
 
 7^=0.700-0.729^- 
 
 This shows the fundamental importance of the ratio of the 
 work of compression to the useful work. 
 
116 THE GAS TURBINE. 
 
 In order to establish our ideas in this respect, we may 
 consider theoretically that this ratio will lie somewhere 
 between 0.2 and 0.4, which will cause the total mechanical 
 efficiency to range between 0.4 and 0.6. As we shall find 
 the thermal efficiency to lie between 0.4 and 0.6 we see that 
 the total useful efficiency will be from 0.16 to 0.36. 
 
 We shall now pass to the discussion of the various cycles 
 applicable to the gas turbine. 
 
 I. 
 A. Cycles Using the Isothermic Introduction of Heat. 
 
 The typical cycle of this group is that of Carnot. Diesel 
 has sought to use this by realizing isothermal combustion 
 in his motor. This result can be obtained in a gas tur- 
 bine only by causing the combustion to be continued in the 
 expansion nozzle, or by causing the expansion to take place 
 in several stages with successive interheaters. This last 
 solution, however, would only be an approximative one. 
 
 The Carnot Cycle. 
 
 The kilogramme of gas under consideration is compressed 
 from p to p t maintaining at the same time the initial tem- 
 perature T . This isothermal compression absorbs a quan- 
 tity of work ^i, given by the equation: 
 
 c i = RT log hyp 
 
 The gas is now compressed adiabatically from p 1 to p 2 . 
 The temperature passes from T to T 2 and the work absorbed 
 by the compression is 
 
 We also have: 
 
THE GAS TURBINE. 
 
 117 
 
 We then introduce the quantity of heat Q upon the 
 isothermal CD, at the temperature T 2 , during which period 
 the pressure falls from p 2 to p r 
 
 We then have: 
 
 We know that the thermal efficiency of the Carnot cycle 
 
 is equal to: 
 
 T 
 
 FIG. 42. The Carnot cycle. 
 
 and that the useful work is: 
 
 We also have: 
 
 ,ART 2 . 
 
 Po Ps 
 
 The properties of the cycle depend only upon the tern- 
 
 /v\ 
 
 perature of combustion T 2 , the total compression ratio , 
 
 Po 
 and the introduction of the heat Q. The temperature of the 
 
 exhaust is that of the atmosphere, about 300 C. absolute. 
 
118 
 
 THE GAS TURBINE. 
 
 The thermal efficiency, which depends only upon T 2 , 
 may reach very high theoretical values, but to attain these 
 involves the use of excessively high compressions; thus: 
 
 Temperature of combustion T 2 
 
 300 
 
 600 
 
 900 
 
 1200 
 
 1500 
 
 1800 
 
 2100 
 
 Thermal efficiency p 
 
 o 
 
 050 
 
 066 
 
 075 
 
 080 
 
 083 
 
 086 
 
 Adiabatic compression ratio 
 
 1 
 
 11 
 
 46 
 
 128 
 
 282 
 
 525 
 
 913 
 
 We see that it is impossible to pass a thermal efficiency 
 of 0.66 without being obliged to have recourse to excessive 
 compressions, since it is necessary to multiply the adiabatic 
 compression ratio, which we shall calculate. 
 
 2500 
 
 FIG. 43. Efficiency and ratio of adiabatic compression. Carnot cycle. 
 
 This latter : is a function of . For T 2 = 900 degrees, 
 
 Pi 1 2 
 
 and Q=300 calories, we have: r = 0.333, and 
 
 J- > 
 
 * = 120. 
 Po 
 
THE GAS TURBINE. 119 
 
 The ratio of the work of compression to the useful work 
 is given by: 
 
 whence: 
 
 2 _ i 
 T 
 
 1 
 
 In our particular case we have: 
 
 We have seen above that the mechanical efficiency dis- 
 appears as the ratio between the work of compression and 
 the useful work approaches unity. As a matter of fact we 
 cannot even admit sufficiently high values for the thermal 
 efficiency and for the temperature T 2 , since the latter, 
 resulting from the adiabatic compression, cannot exceed 
 700 degrees C. absolute, whether we employ a reciprocating 
 piston compressor or a rotary turbine compressor. 
 
 The Carnot cycle is therefore not adapted to the gas 
 turbine, since the high thermal efficiency which can be real- 
 ized by its use is obtainable only by the employment of 
 very high compressions and enormous masses of gas. In 
 consequence, the compressor, which is admittedly the weak 
 point of the gas turbine, assumes an excessive importance, 
 and the mechanical losses would absorb all the useful work. 
 
 The Diesel Cycle. 
 
 Theoretically the cycle of Diesel differs from the Carnot 
 cycle by the substitution of a wholly adiabatic compression 
 for the two successive compressions, isothermal and adia- 
 
120 
 
 THE GAS TURBINE. 
 
 batic, of Carnot. The rejection of heat to the cooling 
 medium is thus produced by the non-closure of the cycle : 
 Here again the isothermal expansion is defined by: 
 
 Q 
 
 ART* 
 
 Pa 
 
 and a considerable degree of expansion is required to enable 
 a sufficient quantity of heat Q, to be introduced. 
 
 FIG. 44. The Diesel cycle. 
 
 Even if we admit that the temperature of combustion, 
 obtained at the end of the adiabatic compression, may 
 attain 800 degrees (corresponding to a compression of 35 
 atmospheres), we have, for: 
 
 Q = 100 200 300 calories 
 
 Ps 
 
 37 220 
 
 We cannot therefore exceed an introduction of 200 calories, 
 and even at this figure there would no longer be an adiabatic 
 expansion. 
 
 The maximum temperature of the cycle should therefore 
 occur at the beginning of the combustion, and should be 
 superior to that produced by the compression, and thus 
 the curve of combustion should keep above the isothermal. 
 
 In any case this cycle is not adapted to the gas turbine. 
 
THE GAS TURBINE. 
 
 121 
 
 Partial Isothermal Cycles. Some writers, Barkow among 
 others, have suggested that the combustion should be 
 started under constant pressure, and completed isother- 
 mically. We shall examine this solution later on, but it is 
 difficult of realization in turbines, and offers no especial 
 advantages. 
 
 B. Cycles Using the Isobaric Introduction of Heat. 
 
 Combustion under Constant Pressure. With combustion 
 at constant pressure the temperature of the gas is raised. 
 It is preceded by a compression which may be either adia- 
 batic or isothermic. In the first case the compression is 
 not accompanied by the transfer of any heat to the cooling 
 medium, but it involves the expenditure of a greater amount 
 of work. A complete computation is necessary to show 
 which of the two systems should be adopted. The follow- 
 ing table will serve as a basis for the calculations : 
 
 Compression ratio. 
 
 5 
 
 10 
 
 15 
 
 20 
 
 25 
 
 30 
 
 40 
 
 60 
 
 80 
 
 100 
 
 Final temperature of adiabatic compression . . 
 
 479 
 
 585 
 
 658 
 
 716 
 
 764 
 
 804 
 
 875 
 
 990 
 
 1040 
 
 1150 
 
 Equivalent in calories of work ( adiabatic 
 
 42 
 
 68 
 
 85 
 
 99 
 
 110 
 
 120 
 
 136 
 
 164 
 
 176 
 
 203 
 
 of compression , 1 isothermal . . 
 
 33 
 
 48 
 
 56 
 
 62 
 
 67 
 
 71 
 
 76 
 
 85 
 
 90 
 
 95 
 
 Ratio of the two efforts 
 
 0.78 
 
 070 
 
 0.66 
 
 0.62 
 
 0.61 
 
 0.59 
 
 0.56 
 
 0.52 
 
 0.51 
 
 0.48 
 
 
 These figures are calculated upon the assumption of an 
 initial temperature of 300 degrees C. absolute, and result 
 in the following considerations: 
 
 The work absorbed by the compressor consists of the 
 compression, properly so-called, which is given, in the case 
 when operating at constant temperature, by the formula 
 
 and the work necessary to drive the compressed air into 
 the reservoir at the pressure p^ that is p l v^ p V Q . 
 
122 
 
 THE GAS TURBINE. 
 
 But in this case the second term is zero, and we have 
 
 IL 
 
 5 cr 
 
 150 
 
 100 
 
 50 
 
 10 
 
 ZO 
 
 30 
 
 40 50 
 
 60 
 
 70 80 
 
 Pressures 
 
 FIG. 45. Equivalent in calories of the work of compression per kilogramme of air. 
 
 In the case of the adiabatic compression the first term 
 has a value EC V (T 1 T ), and the second (p l v l p Q v ) is equal 
 to R(T l T Q ), whence: 
 
 and since 
 we have: 
 
 E(C P c v ) 
 
 From this it follows that: 
 
 * lo 
 
 '(IT- 1 
 
THE GAS TURBINE. 123 
 
 This ratio tends to approach unity for infinitely small 
 compressions. It decreases rapidly as the compression 
 ratio increases, and falls to 0.5 for a compression of about 80. 
 
 The power absorbed by the compressor is therefore less, 
 for the same compression, with the isothermal method than 
 with the adiabatic. This difference is still more marked 
 if, for any reason, the gas which has been compressed adia- 
 batically is allowed to return to its initial temperature. 
 Thus a compression of 20 will drop to about 8.4. This fact 
 renders adiabatic compression inadmissible for the ordinary 
 applications of compressed air. In the case of the gas tur- 
 bine, however, the sensible heat of the adiabatically com- 
 pressed gas is not lost, and a fuller discussion of the subject 
 becomes necessary. 
 
 Adiabatic Compression. 
 
 During the compression the pressure passes from p to 
 Pi and the temperature from T to T r 
 
 T 
 We have: fe- 
 
 Po \T 
 
 The introduction of Q calories, at the constant pressure 
 p v raises the temperature to T 2 , the temperature of combus- 
 tion, and 
 
 Q = C P (T 2 -T l ). 
 
 The mixture then expands adiabatically from T 2 to T 3 , and 
 
 r, 
 
 We see that: 
 
 The quantity of heat rejected to the cooling medium is 
 equal to the heat carried off by the exhaust gases, that is: 
 
124 
 
 THE GAS TURBINE. 
 
 The thermal efficiency p will then be : 
 
 _Q-q_ _T 
 
 p - Q 
 
 We therefore obtain the same efficiency as in a Carnot 
 cycle having the same ratio of adiabatic compression, but 
 without having the necessity for the preliminary isothermal 
 compression. 
 
 FIG. 46. Cycle of isobaric combustion with adiabatic compression. 
 
 The total compression is therefore much lower, but the 
 upper temperature of the cycle is much higher, a matter 
 which offers no inconvenience. 
 
 The ratio of the work of compression to the useful work 
 is given by 
 
 TO^T" l ~^ Cv ~^' 
 
 It is easy to see that this ratio is constant if we give the 
 temperature T 3 at the end of the expansion a fixed value, 
 
 rwi rji 
 
 for we have 7^ = 7^, and consequently : 
 
 ^3 -* 
 
THE GAS TURBINE. 
 
 125 
 
 This ratio attains greater value, therefore, as the tem- 
 perature T 3 has a higher value. Since, however, for con- 
 structive reasons, T 3 cannot be allowed to exceed 700 degrees 
 C., we have: 
 
 C7V. 1 
 
 = 0.75 
 
 ^u 700_ 
 300 
 
 The corresponding value of the total mechanical effici- 
 ency T), given by the equation ^=0.700 0.729^, is there- 
 fore only about 0.15. 
 
 The properties of the most advantageous family of cycles 
 are given below: 
 
 Ratio of compression 
 
 5 
 
 10 
 
 15 
 
 20 
 
 30 
 
 Final temperature of compression T t 
 
 480 
 
 585 
 
 658 
 
 716 
 
 804 
 
 Final temperature of combustion T 2 
 
 1120 
 
 1400 
 
 1540 
 
 1670 
 
 1876 
 
 Final temperature of expansion T t . . 
 
 700 
 
 700 
 
 700 
 
 700 
 
 700 
 
 Heat introduced (calories) Q 
 
 149 
 
 188 
 
 205 
 
 227 
 
 255 
 
 Heat lost in the exhaust q 
 
 92 
 
 92 
 
 92 
 
 92 
 
 92 
 
 Thermal efficiency p 
 
 37 
 
 49 
 
 054 
 
 58 
 
 063 
 
 Equivalent A^c of work of compression . 
 Equivalent A*Cu of useful work . . 
 
 42 
 
 56 
 
 68 
 94 
 
 85 
 112 
 
 99 
 134 
 
 120 
 162 
 
 Total useful effect py 
 
 06 
 
 08 
 
 086 
 
 092 
 
 10 
 
 Equivalent An*Gu of net mech. work .... 
 Consumption of air per H.P. hour, kg 
 
 Ratio of powers - - 
 
 8.4 
 75 
 
 7 1 
 
 14.1 
 
 45 
 
 7 1 
 
 16.8 
 37 
 
 7 1 
 
 20 
 32 
 
 7 1 
 
 24.3 
 26 
 
 7 i 
 
 
 
 
 
 
 
 These results, plotted in the diagram, are not very 
 encouraging. An adiabatic compression of 20 gives a final 
 temperature of 716, which should not be exceeded.* 
 
 The useful effect does not exceed 9 per cent., and the mass 
 of gas required to produce a unit of work is considerable. 
 
 * That is, if the action is truly adiabatic, without any artificial cooling of 
 the parts in contact with the gas. If this is not the case it is impossible to cal- 
 culate accurately the results which may be attained. 
 
126 
 
 THE GAS TURBINE. 
 
 
 
 
 ~ 
 
 CO **- 
 
 s 
 
 CO 
 
 8 
 
 O 
 
 1O 
 
 d 
 
 % 
 
 d d 
 
 
 
 
 00 
 
 P s ^ 
 
 s 
 
 1 
 
 o 
 
 1O 
 
 d 
 
 8 S 
 
 d d 
 
 
 
 o 
 
 <O nn ^ 1O co 
 
 3 $ > d 
 
 
 
 1 
 
 o 
 
 d 
 
 CO 00 
 
 d d 
 
 
 
 o 
 "t 
 
 | g 8 g | 
 
 S 
 
 
 
 rH 
 CO 
 
 o 
 
 d 
 
 00 IO 
 C^J O5 
 
 d d 
 
 
 s . 
 
 3 
 
 QQ JO C<J r H CD 
 
 00 05 g 
 
 g 
 
 a 
 
 CO 
 
 o 
 
 d 
 
 i o 
 
 d ^ 
 
 
 i 
 
 1C 
 
 o o "^ 
 
 t- w OS CO g 
 
 S 
 
 s 
 
 o 
 
 1 
 
 CO 
 
 1 rH 
 
 O 
 
 
 i 
 
 s 
 
 rH ^ ^ O 
 
 S 
 
 a 
 
 d 
 
 d 
 
 S? c^ 
 
 "^ rH 
 
 O 
 
 
 H 
 
 2 
 
 J5 <M G* * o 
 
 s 
 
 rH 
 
 O5 
 
 S 
 
 d 
 
 O CO 
 
 d 
 
 
 -e 
 
 O 
 
 lO , lO 
 CO ^ G^l CO rt< 
 
 S * d 
 
 00 
 
 rH 
 r-t 
 
 d 
 
 % 
 
 d 
 
 OQ 
 
 d ^ 
 
 
 i 
 
 rd 
 
 1C 
 
 1 g? s 
 
 rH ^ O 
 
 co 1 
 
 g 
 
 d 
 
 d 
 
 rH . 
 
 d ^ 
 
 
 g 
 
 
 
 
 
 
 
 
 
 g 
 
 
 
 
 
 
 
 
 
 | 
 
 
 
 
 
 
 
 
 
 
 
 
 h| 
 
 
 
 
 
 
 
 
 
 'i 
 
 
 
 .0 
 
 
 
 
 
 
 S 
 
 
 
 
 02 
 
 
 
 
 
 
 
 &J5 
 
 
 
 
 
 
 
 
 
 
 Compression ratic 
 
 Temperature of combustion T 2 . . 
 Heat introduced (calories) Q . . . 
 Heat lost in the exhaust q 
 Heat lost in the compression Q ' 
 Thermal efficiency p 
 
 Equivalent A""Cc of work of com] 
 
 Equivalent A ^u of useful work 
 
 .2 
 
 i 
 
 Mechanical efficiency rj 
 
 g;^ a 
 
 -feife; 
 
 1 | 
 
 ! 1 
 
 3 1 
 ^ ( 
 
 
THE GAS TURBINE. 
 
 127 
 
 100 
 
 050 
 
 
 
 seful effect 
 
 - d -LtprH.Phour_ 
 
 2000 
 
 1000 
 
 11 K 8 
 
 10 
 
 15 
 
 20% 
 
 PIG. 47. Cycles with adiabatic compression and isobaric combustion. Exhaust discharged 
 
 at 700 degrees C. 
 
 Isothermal Compression. 
 
 The isothermal compression from p to p lt effected at a 
 temperature T , absorbs a quantity of work of which the 
 equivalent in heat is: 
 
128 
 
 THE GAS TURBINE. 
 
 The introduction of Q calories, at the constant pressure 
 p lt raises the temperature to T 2 , and: 
 
 The adiabatic expansion from p l to p Q brings the tempera- 
 ture to T 3 , and 
 
 /PoV 
 2 \pj 
 
 The exhaust gases carry off with them heat equal in 
 calories to: 
 
 Reheating 
 
 Cooling 
 
 FIG. 48. Cycle with isobaric combustion and isothermal compression. 
 
 The thermal efficiency will therefore be: 
 
 C f (T t - T,) -C P (T 3 - T,)ART log hyp-' 
 
 c p (:r 2 -r ) 
 
 The table on page 125 gives the results of computations 
 for a series of cycles having the same temperature of exhaust, 
 7^3 = 700 C. 
 
 These results show that it is important to use as high a 
 compression as possible. In this respect the limits are 
 those imposed by the values for the temperatures of com- 
 
THE GAS TURBINE. 129 
 
 bustion and of the introduction of the heat, these becoming 
 excessive when the total compression passes 60. 
 
 In this case the results may be ameliorated by admitting 
 a lower temperature for the exhaust. For example, let us 
 take a compression of 80. Let the temperature of the ex- 
 haust be placed at 600 absolute, instead of 700. 
 
 The temperature of combustion will then be 2150. We 
 also find that the temperature Q of introduction of the 
 
 C7 c QQ 
 
 heat will be 440, the efficiency 0.635, the ratio c 
 
 0.325, whence 77 =0.46, and consequently py =0.29; while 
 we should have had 0.28 with a compression of 40, and an 
 exhaust temperature of 700 absolute. 
 
 There would be no theoretical advantage in thus lower- 
 ing the temperature of the exhaust, if it did not have the 
 practical value of aiding in the use of successive pressure 
 stages in the turbine. By the use of this system, however, 
 we may carry the expansion down to 700 in the nozzles of 
 the first turbine, and continue it from 700 to 600 in the 
 guide blades of a second group of revolving wheels. 
 
 A series of practical tests would be required to deter- 
 mine the exact limit from which high compressions could be 
 utilized in this manner. This limit will be found to depend 
 both upon the true law governing the expansion and upon 
 the temperature of combustion actually attainable*. 
 
 Below this limit it will not be found advantageous to 
 attempt to utilize high compressions by the reduction of 
 the temperature of the exhaust, and it will be a better en- 
 deavor to increase the introduction of the heat and the 
 maximum temperature. 
 
 It does not seem to be a practical impossibility to realize 
 higher compression ratios than 50. The construction of the 
 compressors and the combustion chambers for such pressures 
 seems practicable. 
 
 * The values for 7, and for the specific heats, may differ in practice from 
 those which have been taken in the computations. 
 
 9 
 
130 
 
 THE GAS TURBINE. 
 
 But the actual pressure may be modified by causing the 
 exhaust gases to be discharged at a lower pressure. This 
 point will be considered more fully hereafter, and it is now 
 only necessary to mention that the lower pressure p may 
 be made less than atmospheric pressure (such as J, }, or -J- 
 atmosphere) without in any way changing the preceding 
 arguments and calculations. The upper pressure limit p l 
 will then be reduced to J, J, or i of its former value. The 
 
 sr\ 
 
 ratio will not be changed, and the efficiencies will not be 
 
 Po 
 
 modified. 
 
 /*> 
 
 T,; 
 
 o v 
 
 FIG. 49. Cycle with isobaric combustion and isothermal expansion. 
 
 We may mention here a modification suggested by 
 Barkow among others, in which the combustion, commenced 
 under constant pressure, is completed in the course of an 
 isothermal expansion. 
 
 Let us suppose the adiabatic expansion is the same as 
 in the preceding case, and the final temperature 700 C. 
 absolute. It follows that the upper temperature limit will 
 be the same, and consequently also the quantity of heat Q, 
 introduced under constant pressure. But we may introduce 
 a supplementary quantity of heat K, during the isothermal 
 expansion. Let / be the ratio of this expansion, we will 
 then have: K = ART 2 log hyp (/). 
 
_ THE GAS TURBINE. _ 131 
 
 The total compression will be /-times greater than 
 before, so that the work of compression will be increased 
 by a supplementary amount 
 
 There will then be a gain of work equal to: 
 AR(T 2 -T.} log hyp (*) or ^~^K. 
 
 1 2 
 
 The quantity of heat introduced at constant temperature 
 is then utilized with a thermal efficiency equal to that of the 
 Carnot cycle, so that the efficiency of the entire cycle is im- 
 proved. It is necessary, however, to give a considerable 
 value to /, in order that K may obtain any importance. 
 A complete computation shows that it would be better, 
 so far as the total useful effect is concerned, to utilize all 
 the compression available to raise to a maximum the intro- 
 duction of heat under constant pressure. 
 
 Discussion of Comparative Efficiencies. 
 
 We may now make a definite comparison of the two 
 modes of compression. 
 
 It will be seen at once, by an inspection of the diagram, 
 that the thermal efficiency p is slightly greater when we 
 use adiabatic compression. But since, with this system we 
 cannot exceed in practice a compression ratio of 20, the 
 maximum value for p is 0.58; while when the compression 
 is isothermal, we may carry the compression as high as 60, 
 which gives for p the maximum value 0.63. 
 
 The superiority of isothermal compression is still more 
 marked from the point of view of the mechanical efficiency. 
 While this remains constant whatever the compression in the 
 adiabatic system, it increases with the compression in the iso- 
 thermal system, and attains values ranging from double to 
 triple those realized in the first case. The same is practically 
 true with regard to the total efficiency py, which, according 
 to our hypotheses, appears to have a limit of about 0.30. 
 
132 
 
 THE GAS TURBINE. 
 
 It will, therefore, be necessary to expend 2120 calories 
 per effective horse-power hour, delivered on the shaft, which 
 corresponds to a consumption of 212 grammes of hydro- 
 carbon fuel, having a calorific value of 10,000 calories (lower 
 calorific value). The Diesel and the Banki motors have a 
 consumption of 180 to 250 grammes. Gas engines operating 
 with blast-furnace gas require at a minimum about 2000 
 calories per effective horse-power. 
 
 The fuel consumption of the gas turbine is therefore 
 comparable with that of the best motors known. The weak 
 point appears in the fact that the effective power absorbed 
 by the compressor is equal to about 85 per cent, of the net 
 effective power practically available on the shaft. 
 
 ^It should be noted that if, by reason of the defective 
 arrangement of the compressor, the heat developed during 
 the compression is not immediately absorbed by the injection 
 of water and by cooling the walls, but only disappears in the 
 flow of the gas between the compressor and the turbine, the 
 efficiency will be much less than in the two preceding cases./ 
 
 We have, in fact, the following results: the ratio of the 
 work of compression to the useful work will be constant 
 whatever the degree of compression for any given tempera- 
 ture of exhaust. For an exhaust temperature of 700 C. 
 absolute, this ratio is 0.75, whence y =0.17. As for the ther- 
 mal efficiency, it will vary as shown in the following table : 
 
 Ratio of compression. 
 
 5 
 
 10 
 
 15 
 
 20 
 
 30 
 
 40 
 
 60 
 
 80 
 
 100 
 
 Heat introduced Q 
 
 195 
 42 
 61 
 0.31 
 0.053 
 
 254 
 68 
 94 
 0.37 
 0.063 
 
 292 
 
 85 
 115 
 0.39 
 0.067 
 
 328 
 90 
 137 
 0.44 
 0.071 
 
 375 
 120 
 163 
 0.44 
 0.074 
 
 415 
 136 
 187 
 0.45 
 0.077 
 
 480 
 164 
 224 
 0.47 
 
 0.080 
 
 522 
 
 176 
 254 
 0.49 
 0.083 
 
 563 
 202 
 
 271 
 0.49 
 0.080 
 
 Heat lost in compression, Q' 
 Equivalent ^L^w, useful work 
 Thermal efficiency p 
 
 Total useful effect pn 
 
 
 The results would not be as low as indicated in the table 
 if the compression were effected in several stages, because 
 the cooling of the gas between the cylinders reduces the 
 
THE GAS TURBINE. 
 
 133 
 
 expenditure of work. It might be possible to obtain a 
 satisfactory result if the compression were divided into a 
 number of stages, with complete inter-cooling. 
 
 Cycles with Isopleric Introduction of Heat. 
 (Cycles for Explosion Motors.) 
 
 Without discussing, for the moment, whether or not this 
 method is practically applicable to the gas turbine we may 
 examine the efficiencies which it is theoretically capable of 
 realizing. 
 
 The introduction of heat at a constant volume causes 
 an increase of pressure, and produces a greater rise of tem- 
 perature than if a constant-pressure system is employed. 
 
 Adiabatic Compression. 
 
 If the compression is effected adiabatically the final 
 temperature of the explosion will be very high, and the 
 introduction of heat per kilogramme of gas cannot be very 
 great. In fact we are obliged to require excessive final 
 
 Ratio of compression 
 Po 
 
 i 
 
 5 
 
 10 
 
 15 
 
 20 
 
 Final temperature compression !7\ 
 
 300 
 980 
 115 
 0.200 
 
 23 
 
 
 
 0.70 
 0.140 
 
 
 
 3.27 
 
 16 
 39.5 
 
 480 
 1530 
 188 
 0.505 
 42 
 96 
 
 0.44 
 
 0.38 
 0,192 
 
 1.66 
 
 15.4 
 
 36.5 
 17.4 
 
 585 
 1930 
 230 
 0.595 
 
 68 
 138 
 
 0.50 
 
 0.34 
 0.203 
 
 2.1 
 
 32.7 
 
 47.0 
 13.6 
 
 655 
 
 2190 
 260 
 0.645 
 
 85 
 168 
 
 0.52 
 
 0.32 
 0.210 
 
 2.3 
 
 49 
 
 54.0 
 11.8 
 
 717 
 2300 
 270 
 0.654 
 99 
 178 
 
 0.55 
 
 0.30 
 0.200 
 
 2.6 
 
 65 
 
 54.0 
 11.8 
 
 Final temperature combustion T 2 
 
 Heat introduced Q (calories) 
 
 Thermal efficiency p 
 
 Equivalent A e Cc of work of compression . 
 Equivalent A^Tu of useful work . ... 
 
 Ratio ~ 
 
 TO* 
 Mechanical efficiency f) 
 
 Total useful effect pri . . . . 
 
 Nc 
 Ratio of power . . .... 
 
 Ratio of pressures . . 
 
 Po 
 Net available mechanical work 77 TO* 
 Consumption of air, kg. per H.P. hour. . . 
 
134 
 
 THE GAS TURBINE. 
 
 *fe 
 
 I 
 loo" 
 
 050 
 
 i 
 
 % 
 
 2000 
 
 1000 
 
 10 
 
 20 
 
 FIG. 50. Cycle with adiabatic compression and isopleric combustion. Exhaust escaping 
 at 700 degrees C. absolute. 
 
 explosion pressures in order to attain the temperature of 
 700 at the end of the expansion. It is therefore impractica- 
 ble to exceed a compression ratio of 15, which leads to an 
 explosion pressure of 49 atmospheres. Under these condi- 
 tions, themselves difficult to realize, the thermal efficiency 
 
THE GAS TURBINE. 135 
 
 p will be about 0.64, with a mechanical efficiency of 0.33 
 and a useful effect of 0.21, as shown in the table on page 133, 
 which gives, as before, the results of a series of cycles, for 
 an exhaust temperature of 700 C. absolute. 
 
 Explosion turbines with adiabatic compression have, 
 therefore, a low efficiency, the total useful effect not exceed- 
 ing 20 per cent. Increase of initial compression has but a 
 slight influence, so that such machines are of interest only 
 for small powers, or in cases in which the consumption of 
 fuel is a secondary consideration. 
 
 Isothermal Compression. 
 
 In this case we introduce a quantity of heat Q for each 
 kilogramme of gas and have: 
 
 The pressure becomes 
 
 The adiabatic expansion brings the temperature down 
 to T 3 . 
 
 We then have: 
 
 ^2 
 
 and T. = 
 
 The heat discharged to the cooling medium is composed 
 of the calories rejected with the exhaust: C P (T 3 T ) and 
 the heat subtracted during the isothermal compression: 
 
 RT Q log hyp M^J. The thermal efficiency will then be: 
 
 c v (T 2 -T )-C p (T s - T )RT log hypf^- 1 
 P = V ^o 
 
136 
 
 THE GAS TURBINE. 
 
 o ' v 
 
 FIG. 51. Cycle with isopleric combustion and isothermal compression. 
 
 If, as before, we take the exhaust temperature at 700 C. 
 ., we have the corresponding family of cycles as follows: 
 
 Compression ratio 
 
 1 
 
 s 
 
 10 
 
 15 
 
 20 
 
 Temperature of compression T 2 
 
 980 
 
 1820 
 
 2420 
 
 2850 
 
 3210 
 
 Heat introduced. Q (calories) 
 
 115 
 
 255 
 
 355 
 
 430 
 
 490 
 
 Heat lost in the exhaust q 
 
 92 
 
 92 
 
 92 
 
 92 
 
 92 
 
 Heat lost in the compression 
 
 
 
 33 
 
 48 
 
 56 
 
 62 
 
 Thermal efficiency p 
 
 019 
 
 051 
 
 061 
 
 066 
 
 068 
 
 Equivalent A ^c of work of compression . 
 Equivalent A^u of useful work 
 
 
 23 
 
 33 
 130 
 
 48 
 215 
 
 56 
 
 282 
 
 62 
 336 
 
 r> * ^ 
 
 Katio . 
 
 
 
 025 
 
 0.22 
 
 020 
 
 019 
 
 ^u 
 Mechanical efficiency y . . 
 
 070 
 
 052 
 
 0.54 
 
 0.55 
 
 056 
 
 Total useful effect prj .... 
 
 13 
 
 0265 
 
 033 
 
 0365 
 
 038 
 
 Nc 
 Ratio of powers . 
 
 o 
 
 068 
 
 058 
 
 052 
 
 049 
 
 Ratio ^ 
 
 327 
 
 6.1 
 
 8.1 
 
 9.5 
 
 10.7 
 
 Pi 
 Ratio ^ 2 .. 
 
 3.27 
 
 30.4 
 
 80.7 
 
 143 
 
 214 
 
 Po 
 
 Net available mechanical work ^'Cw 
 Consumption of air, kg. per H.P. hour . . . 
 
 16 
 39.5 
 
 68 
 9.4 
 
 116 
 5.5 
 
 156 
 4.1 
 
 188 
 3.4 
 
_ THE GAS TURBINE. _ _ 137 
 
 The ratio of the work of compression to the useful work 
 is low, which is a great advantage. But with even a com- 
 pression ratio of 10 the final pressure passes 80 atmospheres, 
 a limit very difficult to handle. The total useful effect then 
 reaches 0.38, while with isothermal compression followed 
 by combustion at constant pressure the limit is 0.31. 
 
 If the exhaust is discharged into a space having a reduced 
 pressure it becomes practicable to use higher pressure ratios. 
 Thus, with a compression of 3 and an introduction of 430 
 calories, the maximum pressure reaches 28.5 atmospheres. 
 If we allow this to escape into a space having a pressure of 
 i atmosphere we get a total expansion of 143 times, this 
 being necessary to reduce the temperature of the gas from 
 2850 to 700 absolute, which gives a useful effect of 0.365, 
 while the power of the compressor will be reduced to about 
 one-half the net available power. These results are very 
 encouraging. Unfortunately, it is not easy to construct a 
 satisfactory explosion turbine, the operative portions of 
 the explosion chamber being unable to resist the very high 
 temperatures developed. 
 
 It may readily be shown that the efficiency becomes less 
 favorable if the temperature of the exhaust is made lower 
 than 700 degrees. 
 
 Isopleric Combustion Cycles without Compression. 
 
 If the gas is not compressed before the explosion the 
 efficiency is low, as we have already seen. We may investi- 
 gate the manner in which it varies if the temperature T 3 of 
 the exhaust is varied. 
 
 We have: 
 
 T 7 , _ HP rp _ rp 
 
 3 
 
 It is easy to see that efficiency will be a minimum when 
 T , and increases with the increase of T 3 above its 
 
 minimum value T . 
 
138 
 
 THE GAS TURBINE. 
 
 For 
 
 We have 
 and hence 
 
 600 
 
 0.16 
 
 0.11 
 
 800 
 
 0.23 
 
 0.16 
 
 1000 
 0.27 
 0.17. 
 
 ^=0.04 
 
 We see that the highest efficiency corresponds to the 
 highest temperature of the exhaust admissible, with regard 
 to the endurance of the metallic turbine wheel. Under the 
 most favorable conditions the total efficiency cannot be 
 expected to surpass 14 per cent. 
 
 II. 
 
 Cycles with Expansion Prolonged below Atmospheric 
 Pressure. 
 
 In the cycles thus far examined the gas has been ex- 
 panded down to the pressure of the atmosphere and rejected 
 at a temperature dependent upon the conditions of opera- 
 tion; chosen, however, as high as possible, with respect to 
 
 FIG. 52. Cycle with prolonged expansion. 
 
 the endurance of the turbine wheel. There is, however, 
 nothing to prevent the arrangement of the parts in such a 
 manner as to cause a part of the process to be conducted 
 at a pressure below that of the atmosphere; as has already 
 been done in the so-called "atmospheric" gas engines, using 
 a free piston. 
 
THE GAS TURBINE. 
 
 139 
 
 In the case of the turbine this may be effected by the 
 use of an air pump. When, however, the large dimensions 
 are considered, a piston pump is seen to be unsuited for 
 this purpose. It is necessary, therefore, to use multicellular 
 turbine machines similar to those already designed for com- 
 pressors. For a reduction of the pressure to i atmosphere 
 it will not be found necessary to use more than five turbine 
 wheels. In order, however, to reduce the amount of work 
 absorbed by this machine it is necessary that it should be 
 
 Air 
 
 Exhaust 
 Discharge 
 
 FIG, 53. Turbine with exhaust under reduced pressure without regenerator. 
 
 operated at a constant temperature, and it is desirable that 
 the temperature of the exhaust gases should be brought 
 down to 300 C. absolute, before these gases enter the suction 
 blower, because the work absorbed by the machine, R T\og (/), 
 is proportional to the absolute temperature of the gases. 
 
 This result may be obtained by cooling the gases by 
 means of an abundant injection of water, in connection 
 with the use of a sort of barometric condenser ( Fig. 53), 
 or by the use of a tubular refrigerator with circulating 
 water (Fig. 54). 
 
140 
 
 THE GAS TURBINE. 
 
 It is evident that this latter method may be used in con- 
 nection with some system of regeneration. The gases may 
 also be cooled in the vaporizer of sulphurous-acid gas, 
 forming part of a refrigerating machine, as we shall see here- 
 after (Fig. 55). 
 
 Water 
 
 FIG. 54. Turbine with exhaust under reduced pressure, without regenerator, with inter- 
 cooler. 
 
 s& 
 
 ^ 
 
 Turbine 
 
 1 
 
 LJ 
 
 
 
 
 Exhauster Turbine 
 
 /" 
 
 ^ 
 
 ^SOz Liquid 
 
 
 , S0 2 Gas 
 
 Gases 
 
 FIG. 55. Turbine with exhaust at reduced pressure, with recovery of waste heat by a 
 sulphur dioxide turbine. 
 
 The system of exhausting at low pressure enables a 
 regenerator to be employed, but the construction of the 
 regenerator is not very easy, because the transmission of 
 heat is not very active at low pressures. 
 
 There are three methods in which the system may be 
 employed. 
 
 The temperature of the exhaust may be brought below 
 700 degrees by the use of some one of the cycles already dis- 
 cussed, this plan permitting the use of a multistage turbine, 
 
THE GAS TURBINE. 141 
 
 although at the expense of a certain increase in the work 
 of compression. The actual balance of power can only be 
 definitely determined by examining each case by itself. 
 Still we have already seen, that, in general, if we attempt 
 to increase the efficiency y by the use of a multistage tur- 
 bine, the total effect will be improved to a greater extent 
 if we increase the amount of heat supplied rather than by 
 lowering the temperature of the exhaust. 
 
 The second application of the system which we are con- 
 sidering consists in increasing the expansion ratio, and 
 utilizing this increase to allow a corresponding increase in 
 the amount of heat supplied, while maintaining the tempera- 
 ture of the exhaust at 700 degrees. This method presents 
 especial advantages in cases in which the efficiency is limited 
 by consideration of the maximum pressure of the cycle; 
 which is notably true in the explosion cycles. 
 
 Finally, we may utilize the low-pressure exhaust in a 
 manner which avoids the use of a piston compressor. We 
 shall see that multicellular turbine compressors are not well 
 adapted for the production of very high pressures. Under 
 such conditions their efficiency is materially reduced by 
 reason of the friction of the latter wheels of the series 
 in the gas or air of high density. It is therefore better to 
 arrange a turbine compressor to deliver the gas into the 
 combustion chamber at a pressure, say, of six atmospheres, 
 and follow the gas turbine by an exhaust blower, reducing 
 the exhaust pressure to J atmosphere, than it is to employ a 
 single compressor operating at a pressure of 36 atmospheres. 
 
 In addition, the power turbine will operate with less 
 frictional resistance by reason of the lower pressure. 
 
 It seems as if some such arrangement as this is necessary 
 if the piston compressor is to be entirely eliminated in gas 
 turbine design. 
 
 From a thermodynamic point of view there should be 
 no difference between the operation with exhaust at low 
 
142 
 
 THE GAS TURBINE. 
 
 pressure or at atmospheric pressure, provided the ratio of 
 the two extreme pressures is the same in both cases; and 
 provided that the two compressors operate isothermically 
 and at the same temperature T , in both cases. 
 
 III. 
 Cycles Using Heat Regenerators. 
 
 It is understood that it is possible to employ cycles 
 having the same efficiency as that of Carnot between the 
 same limits of temperature, by replacing the adiabatics of 
 the Carnot cycle by two isodiabatics. The two simplest 
 solutions of this problem are those of Stirling and of Erics- 
 son, but the first of these involves reheating under constant 
 volume, and is not applicable to our case. 
 
 The Ericsson Cycle. 
 
 In the Ericsson cycle, on the contrary, the exchanges 
 of heat are made under constant pressure: the two isodia- 
 batics are isobarics. 
 
 A fo 
 
 FIG. 56. Ericsson cycle. 
 
 The gas is compressed along AB at the constant tempera- 
 ture T m , it is reheated under constant pressure (p^ along 
 BCj by means of a regenerator, which raises the tempera- 
 ture to T 2 . The heat furnished by the fuel is introduced 
 
THE GAS TURBINE. 143 
 
 along CD at the constant temperature T 2 , during which 
 the pressure falls to p . Finally the gas is cooled in the 
 regenerator, from T 2 to T , at the constant pressure p . 
 
 Unfortunately this cycle cannot be realized in practice 
 any more than can the Carnot cycle. 
 
 Independently of the practical difficulty of obtaining 
 an isothermal combustion in the expansion nozzle, we en- 
 counter the impossibility of introducing large quantities 
 of heat without using excessively high compressions, for 
 we have: 
 
 Pi = e ARJ r 2 
 Po 
 
 T 
 The thermal efficiency p = l TTT cannot exceed 0.57, 
 
 * 2 
 
 since the gases are discharged upon the turbine wheel at a 
 temperature T 2 . 
 
 The work of compression "TTc is given by: 
 
 Ur o loghyp(f) or RT.-^r 
 
 \PQ/ A-tt 1 3 
 
 , 
 whence 
 
 We then have ,0 = 0.57 and = 0.75 
 
 (ju 
 
 also )?=0.15 and ^^=0.15X0.57=0.086 
 
 We see, therefore, that the Ericsson cycle is neither 
 practicable nor advantageous. It is possible, however, to 
 apply the principle of regeneration to other cycles, and as 
 we shall see, with advantageous results. 
 
 In general, the method of regeneration is available only 
 for cycles using isothermal compression, and especially those 
 in which the combustion takes place under constant pressure. 
 
144 
 
 THE GAS TURBINE. 
 
 Cycles Employing Isobaric Introduction of Heat. 
 
 As large a proportion as possible of the heat contained 
 in the exhaust gases should be recovered by passing these 
 hot gases through a system of tubes by means of which 
 they heat the compressed gas on its way to the combustion 
 chamber. We see that with a regenerating surface of 
 infinitely great extent we might recover all the heat, if the 
 compressed gas to be heated left the compressor at the ordi- 
 nary temperature (say about 300 C. absolute). This 
 involves an isothermal compression, while if the compres- 
 sion is adiabatic the exhaust gases cannot be cooled below 
 the final temperature of compression. 
 
 O V 
 
 FIG. 57. Cycle with isobaric combustion and isothermal compression. 
 
 The compression is accompanied by a consumption of 
 heat: 
 
 We then introduce by regeneration K calories under 
 the constant pressure p iy and the temperature passes from 
 T to TV We have 
 
 K = C P (T,-T.}. 
 
 The fuel, furnishing Q calories, raises the temperature 
 from T 1 to T 2 : 
 
THE GAS TURBINE. 145 
 
 The adiabatic expansion from p^T 2 to p T 3 gives: 
 
 y-l 
 
 If the regeneration could be complete the gas would 
 enter the regenerator at T 3 and leave it at T , the surround- 
 ing temperature, leaving behind it C P (T 3 T ) calories. 
 But in reality the temperature of the gases is not reduced 
 to T , besides which the compressed gas cannot acquire all 
 the heat units thus gathered, because of the losses by radi- 
 ation, conductivity, etc. If we call the total efficiency of 
 the operation //, we have : 
 
 For example, if the gases leave the turbine at 700 degrees 
 they contain 92 calories per kilogramme which are recover- 
 able, and we have K =92 /*. 
 
 Here the quantity of heat introduced in the cycle is 
 (K + Q), the quantity given up in cooling during the com- 
 pression is Q' ', and that which is discharged with the exhaust 
 is equal to 
 
 The quantity of heat converted into useful work is there- 
 
 fore equal to: 
 
 (K+Q)-Q'-C P (T 3 -T ). 
 
 The actual amount of heat abstracted from the fuel 
 being Q, we then have: 
 
 (K+Q)-Q'-C P (T,-T a ) 
 
 p = nr 
 
 If we maintain a standard temperature of the exhaust, 
 say 700 C. absolute, the value (K+Q) of the total heat 
 introduced is' equal, for each compression ratio, to that 
 which has been computed for cycles without regeneration. 
 It follows that the useful work obtained per kilogramme of air 
 
 10 
 
146 
 
 THE GAS TURBINE. 
 
 is*s 1 1 
 
 OOOO OO 
 
 '^^fr^ t^-cocoo "^ 
 
 'iSlii coi>-i>-oo co 
 
 dddd dddd 
 
 8? 
 
 H CO 
 
 $ 
 
 d d 
 
 coo 
 ioco 
 
 oooo 
 
 
 d d 
 
 THi-IOD COO-^QO 232^S? 
 
 ^^^ sss sill 
 
 w d o 
 
 II 
 
 OTM 
 
 0000 
 
 dd 
 
 
 
 i ico 
 
 33 
 
 O 
 
 Ils3 1 
 
 1 
 
 dddd 
 
 oooo 
 
 dddd 
 
 
 
 
 s\+ 8 
 
 H^ 
 
 Temperature of combusti 
 Total introduction of hea 
 
 il 
 
 O 02 
 
 OiCM OCOO5(N 
 COOS Tt^COOi 
 
 Si58 8S8 8SJ28 
 
 o'drH dddr-5 dodi-H 
 
 II II II II II II 11 II II I! II 
 
 5L^^- =t 5J_ 5L =L =L . =L . 
 
 I 
 
 i 3 
 5.1 
 
 ** 
 
 A 
 A 
 
 f 
 
 Equivalent 
 Equivalent 
 
 Ratio =- 
 
 echanical efficiency 
 
 r-* 
 
 1 
 
 1 S. 
 1.* 
 
 <D .. 
 
 S.b 
 
 -s * 
 S-s 
 
 15 o 
 
 "! 
 
 cs a 
 
 Eq 
 Co 
 
THE GAS TURBINE. 
 
 147 
 
 is not increased by the use of the regenerator, and the ratio 
 of the work of compression to the useful work is not changed. 
 The table on page 146 gives the results for various ratios 
 of compression, and for an exhaust temperature of 700 G. 
 absolute. 'i 
 
 FIG. 58. Cycles with isothermal compression and isobaric combustion with regenerator. 
 
 Exhaust at 700 C. 
 
 The principal advantage of the regenerator is that 1 it 
 permits the attainment of the same total useful effect with 
 much lower compression ratios. Thus, when the coefficient 
 of regeneration, /*, reaches 0.75 a total useful effect of 0.30 
 is secured with a compression ratio of 25, while without 
 regeneration a compression of 60 would be required to secure 
 the same efficiency. 
 
148 THE GAS TURBINE. 
 
 For a given compression, a regeneration of 50 per cent. 
 (/* = 0.5) improves the total useful effect py from 10 to 30 per 
 cent, and a regeneration of 75 per cent, improves the total 
 useful effect from 15 to 30 per cent., according to the com- 
 pression; the influence of the regenerator being greater with 
 low compressions than with the higher compression ratios. 
 
 The use of the regenerator does not extend the limit of 
 maximum total useful effect except when the compression 
 ratio exceeds 80. In this case we are limited by the calorific 
 value of the fuels, which control the maximum amount of 
 heat introduced. The use of the regenerator permits this 
 maximum to be extended. 
 
 Summing up, with a coefficient of regeneration of 0.75, 
 which seems reasonable, we have theoretically, for a com- 
 pression ratio of 25, the possibility of obtaining an effective 
 horse-power hour with 2100 calories; and w r ith a compres- 
 sion ratio of 60 we may produce a horse-power hour with 
 about 1800 calories. These are very encouraging results. 
 
 The following table (page 149) gives the quantity of 
 heat transmitted to the regenerator to obtain a given 
 amount of available mechanical work. 
 
 It is seen that high compressions have the advantage 
 of reducing, to a large degree, the quantity of heat to be 
 regenerated. Thus, with a value /*=0.75 it is necessary to 
 transmit to the regenerator 510 calories per net effective 
 horse-power if the compression ratio is 25, and only 290 
 calories if the compression ratio is 60. 
 
 The transmission of heat per effective horse-power in 
 the condensers of steam engines ranges from 3180 to 6360 
 calories per horse-power hour, according as the engine con- 
 sumes 5 or 10 kilogrammes of steam per horse-power, while 
 the condensing surface ranges from 0.10 to 0.30 square 
 metre per horse-power. 
 
 The exhaust gases of the turbine enter the regenerator 
 at a temperature of 700 C. absolute, and should leave it at 
 
THE GAS TURBINE. 
 
 149 
 
 
 
 * I 
 
 -ft -ft -ft -ft -ft -ft 
 
 II II II II II II 
 
 i- 1 P P H- 1 O O 
 
 8 S S 8 a S 
 
 CO tO H- 1 
 
 00* OS 
 
 I" 1 p p 
 
 CO CO OS 
 
 to co os 
 
 o 
 to 
 
 pop 
 
 bo bs '^ 
 o o o 
 
 (- H- O 
 
 bi M ^ 
 
 *- OS I 
 
 o o o 
 
 H- p O, 
 
 8 8 g 
 
 h- p p 
 
 o bo en 
 
 OO I ' rf^ 
 
 pop 
 
 rf^ OO tO 
 OO Oi rf^- 
 
 pop 
 
 CO -<l 4^- 
 
 OS to 00 
 
 O O P 
 
 4^ io to 
 CO CO 1 
 
 P P 
 ^ OS 
 
 CO O 
 
 oo 
 
 pop 
 
 *<$$ ^r CO 
 
 pop 
 
 CO tO J-J 
 
 o co en 
 
 pop 
 en 4^. to 
 Co <o **-J 
 
 O O O 
 
 000 
 
 pop 
 
 rf^ OO tO 
 
$50 _ THE GAS TURBINE. _ 
 
 400 absolute to realize a regenerative efficiency of 75 per 
 cent. The compressed air enters at 300 and leaves at 600. 
 With a systematic circulation the drop in temperature will 
 be 100 from the entrance to the exit, and the mean tempera- 
 ture drop 
 
 D-d 
 
 log 
 
 is thus 100 degrees. 
 
 We find in steam superheaters a heat transmission of 10 
 to 15 calories per hour, per square metre, per degree, or in 
 the present case, 1000 to 1500 calories. 
 We then have: 
 
 510 
 1000 to 1500 
 
 square metres, or 
 
 0.34 to 0.51 
 
 = 0.20 to 1.30 
 
 1000 to 1500 
 
 square metres per net effective horse-power. 
 
 Thus, to realize a regenerative efficiency of 75 per cent, 
 with a compression ratio of 60, it will suffice to have about 
 the same area of heat transmitting surface as would be 
 given to the condenser surface for a steam engine of the 
 same effective power. 
 
 Regeneration in Cycles Using the Isopleric Introduction 
 of Heat. 
 
 This case may be treated in the same manner as the pre- 
 ceding. It is necessary, however, to note that in practice 
 the regenerated heat can be introduced only at constant 
 pressure and does not act to raise the final pressure. The 
 latter will, therefore, not be so high for a given initial com- 
 pression as in cases in which there is no regeneration, 
 and this diminishes somewhat the advantages of the explo- 
 sion cycle. 
 
THE GAS TURBINE. 
 
 151 
 
 The following results are deduced from the computation: 
 
 Q 
 
 1 i 
 
 which give for 
 ing table. 
 
 = 0.5 and T 3 =700 the results in the follow- 
 
 Compression ratio 
 
 5 
 
 10 
 
 is 
 
 20 
 
 T6mp6ra.tu.r6 of explosion T>, 
 
 1540 
 
 2030 
 
 2380 
 
 2680 
 
 
 16 
 
 41 
 
 73 
 
 110 
 
 Heat furnished by combustible Q 
 
 177 
 
 260 
 
 319 
 
 370 
 
 Thermal efficiency p 
 
 0.55 
 
 0.64 
 
 0.68 
 
 0.71 
 
 Eouivalent A ""GYt of useful work .... 
 
 98 
 
 166 
 
 217 
 
 262 
 
 Ratio ^ 
 
 033 
 
 029 
 
 0.26 
 
 0.24 
 
 ^u ' ' 
 IVlechanical efficiency f] ... 
 
 046 
 
 049 
 
 0.51 
 
 0.53 
 
 Total useful effect py 
 
 025 
 
 031 
 
 035 
 
 038 
 
 Equivalent Arj^u of net mechanical work TJ^U . . 
 Consumption of air kg per H P hour eff 
 
 45 
 14 10 
 
 81 
 790 
 
 111 
 5 70 
 
 139 
 
 460 
 
 Ratio of calories regenerated to effective work . . . 
 
 1.00 
 
 0.57 
 
 0.42 
 
 0.33 
 
 We thus obtain the same results as with combustion at 
 constant pressure, but with compressions only about one- 
 half as great. The absolute maximum of useful effect is not 
 increased, since we are limited by the consideration of the 
 pressure and temperature of explosion to compression 
 ratios only about one-half as great. 
 
 IV. 
 
 Cycles Involving the Injection of Water, Steam, 
 or Cool Gases. 
 
 We have already seen that it is very desirable to be able 
 to reduce the amount of gas to be compressed to realize a 
 given amount of work. If, to fix our ideas upon this matter, 
 we assume compression ratios above 80 to be excessive, we 
 cannot introduce more than 450 calories per kilogramme of 
 
152 THE GAS TURBINE. 
 
 gas, while there are certain combustible mixtures which 
 readily furnish from 550 to 600 calories. We are therefore 
 obliged to dilute these latter, and thus increase the volume 
 of gas to be compressed some 20 to 30 per cent. This incon- 
 venience becomes aggravated with lower compressions. 
 
 This fact has led to investigations as to whether we may 
 not use the rich combustible mixtures without dilution by 
 using certain artifices to limit either the temperature of 
 combustion or the terminal temperature. 
 
 Limitations of the Temperature of Combustion. 
 
 The external cooling of the combustion chamber is en- 
 tirely inconvenient. The calories thus abstracted take no part 
 in the development of power. It would be simpler and more 
 economical to reduce the amount of combustible introduced. 
 
 ^To steam 
 
 \v 
 
 >*-Water injected 
 FIG. 59. Combustion chamber for gas and steam turbines. 
 
 But, if the heat abstracted can be used to vaporize water, 
 and if the steam thus produced is delivered, either to a 
 separate turbine; or, by a separate nozzle, to the main gas 
 turbine; or into the expansion nozzle of the gas; or, finally, 
 into the combustion chamber itself, this heat will partake 
 in the development of power according to a cycle more or 
 less effective, and the loss will be reduced. 
 
_ THE GAS TURBINE. _ 153 
 
 Suppose, for instance, that the steam thus produced 
 is utilized in a separate turbine, which may be either con- 
 nected to a condenser or exhaust into the atmosphere. 
 The combustion chamber of the gas turbine will then act 
 as the furnace of the steam boiler for the separate turbine. 
 Leaving aside, for the moment, the complication of this 
 arrangement, and assuming that we vaporize the water 
 in the generator to a pressure of 20 atmospheres and super- 
 heat the steam to a temperature of 700 degrees absolute, 
 by means of the calories derived from the walls of the com- 
 bustion chamber we obtain a temperature of ebullition of 
 488 degrees absolute. 
 
 The heat contained in a kilogramme of water will be: 
 
 calories. 
 
 If the exhaust is discharged into the air, the temperature 
 will be 373 absolute, while if a condenser is used the tem- 
 perature will be about 320. 
 
 The thermal efficiency of the steam portion of the system 
 wilF be: 
 
 Exhausting into atmosphere: 
 
 _ ^700 - ^73 _ 843 - 637 
 ~ " 
 
 843 
 
 ^=0.172 
 
 Exhausting into a condenser: 
 
 
 - 700 
 
 ^ = 0.185. 
 
 In steam turbines the mechanical efficiency y is about 
 0.70. The result obtained when operated with a condenser 
 corresponds to a consumption of steam of about 4 kilo- 
 grammes per effective horse-power (8.8 pounds). Now a gas 
 turbine, without regeneration or water injection and with a 
 compression ratio of 10, gives a total efficiency ^=0.18. 
 
154 
 
 THE GAS TURBINE. 
 
 The arrangement which we have been discussing is 
 therefore without interest as regards efficiency unless we 
 adopt compressions higher than 10. The only advantage 
 lies in the reduction in the importance of the compressor. 
 
 If the steam, produced at the expense of the heat devel- 
 oped in the combustion chamber, is delivered upon the 
 wheel of the gas turbine through separate nozzles, the effici- 
 ency will be the same as above, and the same conclusions 
 
 Combustible 
 
 To the ? 
 turbine 
 
 injected 
 FIG. 60. Combustion chamber for mixed turbine taking steam from jacket. 
 
 Water injected 
 Combustible 
 
 FIG. 61. Combustion chamber for mixed turbine with independent water injection. 
 
 follow. The same is true if the steam is mingled with the 
 burned gases in the expansion nozzles of the gas turbine 
 itself; and with this arrangement certain precautions, im- 
 portant from a kinetic point of view, are necessary, as will 
 be seen hereafter. 
 
 Finally, if the steam produced at the expense of the heat 
 in the combustion chamber is delivered into the combustion 
 chamber itself, the result will be the same as if the water 
 were delivered directly into the combustion chamber in 
 
THE GAS TURBINE. 155 
 
 the liquid form. This is the arrangement which we shall 
 now examine (Fig. 60). 
 
 Let x be the weight of water injected per kilogramme 
 of gas burned, and let p be the pressure in the combustion 
 chamber. The tension p l of the steam is found from the 
 law of the mixture of gases and vapors, and is equal to : 
 
 P ~ 
 
 in which R and R l are the specific constants of air and of 
 the vapor of water. Supplying these constants, we have: 
 
 i = 46.8 a; 
 
 P ~ P 29.3+46.8z* 
 
 Let 6 be the temperature of ebullition which corresponds 
 to this pressure p 1 . 
 
 The heat absorbed by the vaporization of 1 kilogramme 
 of water injected at 0, into the combustion chamber, is 
 given by: 
 
 A=g+r=606.5+0.305(0-273). 
 
 The steam produced is also superheated, and if we 
 represent the mean value of the specific heat of this steam, 
 superheated between the temperatures of 6 and 7 7 2 , by Cp&^ 21 
 the superheating will absorb 
 
 CpoT 2 (T 2 6) calories. 
 
 The total amount of heat absorbed by 1 kilogramme 
 of steam may readily be calculated by assuming 0.48 as the 
 mean value of the specific heat of steam, and by using the 
 formula of Lorenz 
 
 which gives: 
 
 /T 
 \ a 
 
 with a = 0.43 and b = 36Xl0 5 . 
 
156 
 
 THE GAS TURBINE. 
 
 ; 
 
 If we take the value of 7- the same for the superheated 
 steam as for the gas, we may calculate the temperature at 
 the end of the expansion T s and the corresponding heat of 
 the steam X Tz , from whence the thermal efficiency p of the 
 steam, considered separately, will be: 
 
 It will be observed that X T2 and X Tz are dependent upon 
 the ratio x, or the proportion of water to gas, by weight. 
 The lower this ratio is the more the tension of the steam is 
 reduced with relation to the pressure of combustion p. If 
 the computations are made it will be found that the results 
 differ very little from those corresponding to the case of 
 saturated steam without the presence of any air (in which 
 x = infinity) at least when the temperatures are relatively 
 high, as in the case which we are considering. 
 
 The following table gives the results: 
 
 Absolute pressure of combustion p 
 
 s 
 
 10 
 
 15 
 
 20 
 
 25 
 
 30 
 
 40 
 
 Temperature of combustion T 2 
 
 1120 
 
 1305 
 
 1533 
 
 1680 
 
 1780 
 
 1880 
 
 2050 
 
 Temperature of ebullition 
 
 425 
 
 453 
 
 472 
 
 488 
 
 498 
 
 503 
 
 523 
 
 Heat of vapor Aj^ 
 
 990 
 
 1130 
 
 1230 
 
 1310 
 
 1390 
 
 1490 
 
 1580 
 
 Heat of vapor Ay 3 . . . . 
 
 790 
 
 790 
 
 790 
 
 790 
 
 790 
 
 790 
 
 790 
 
 < for the vapor . . 
 Thermal efficiency p . < , A . 
 1 for the gas 
 
 ( for the vapor . . 
 Total efficiency ?/>... < , ., 
 1 for the gas 
 
 0.20 
 0.34 
 0.14 
 0.11 
 
 0.30 
 0.43 
 0.21 
 0.16 
 
 0.36 
 0.47 
 0.25 
 
 0.18 
 
 0.40 
 0.52 
 0.28 
 0.22 
 
 0.43 
 0.55 
 0.30 
 0.25 
 
 0.47 
 0.57 
 0.33 
 0.26 
 
 0.50 
 0.60 
 0.35 
 0.28 
 
 This table is computed on the assumption that x = in- 
 finity, T 3 = 700 absolute, and that the exhaust is discharged 
 against atmospheric pressure. If the exhaust pressure is 
 reduced the figures will be modified. 
 
 It will be seen that the thermal efficiency of the cycle 
 of the vapor is lower than that for the gas, but if we consider 
 the total efficiency T^O, taking the efficiency of the turbine 
 
THE GAS TURBINE. 157 
 
 at 0.7, and that of the compressor (including its transmis- 
 sion) also at 0.7, these results are reversed. This follows 
 because the work of the compressor is reduced by the use 
 of the steam. 
 
 We conclude from this analysis, that the injection of 
 water is more advantageous than the introduction of an 
 excess of air for combustion, above all because it permits 
 a material reduction in the dimensions of the compressor. 
 
 It may be desirable to consider whether or not there is 
 any risk of the dissociation of the water under the conditions 
 of temperature and pressure existing in the combustion 
 chamber. In all probability there is no danger of such 
 action, since dissociation does not begin, at atmospheric 
 pressure, until a temperature of 1300 degrees C. absolute, 
 and the tension of dissociation does not reach a value of 0.5 
 until 2100 C. At the pressures under consideration there 
 can therefore be no appreciable dissociation, and there can 
 be still less during the expansion, for the drop in tempera- 
 ture with the pressure is very rapid. 
 
 Injection into the Combustion Chamber, of Steam 
 Produced in a Regenerator. 
 
 It has been proposed to replace the introduction of an 
 excess of air in the combustion chamber by an injection 
 of steam. If this steam is produced by the combustion of 
 fuel under a boiler the result will be the same as in the case 
 of the injection of water which we have just examined. 
 This arrangement, however, would be accompanied with 
 the heat losses involved in the use of a separate boiler, 
 together with the mechanical complications accompanying 
 it, besides which it would be necessary to inject much more 
 steam to produce the same effect. 
 
 Assuming, as before, that x = infinity, the total amount 
 of heat absorbed by the injection of a given weight of water 
 
158 
 
 THE GAS TURBINE. 
 
 in the liquid state is 1.5 to 2.5 times greater than if it is 
 injected in the form of steam at 6 degrees. 
 
 The injection of steam into the combustion chamber is 
 of interest only when the steam is generated in some form 
 of regenerator, heated by the exhaust gases of the turbine. 
 We will examine this case, always assuming that the pres- 
 sure of the steam in the mixture is the same as that of the 
 pressure of combustion. 
 
 10 
 
 10 20 30 
 
 FIG. 62. Efficiencies for mixed turbines. 
 
 The exhaust being at the temperature of 700 degrees 
 absolute, each kilogramme of exhaust gases represents 92 
 calories. Each kilogramme of water carries 790 calories, 
 but only 153 calories (the sensible heat) can be regenerated 
 without condensation, if the exhaust takes place at atmos- 
 pheric pressure. The result will be improved if the exhaust 
 
THE GAS TURBINE. 159 
 
 occurs at reduced pressure, and if we take into account the 
 fact that the pressure of the vapor is lower than that of 
 the surroundings into which it is discharged. 
 
 If, then, x kilogrammes of water are mingled with 1 kilo- 
 gramme of burnt gases we may regenerate 92 + 153x calories, 
 and the effective recuperation will be /*(92 + 153z), which 
 gives a vaporization of a weight of water x: 
 
 92 
 
 The weight of steam which may be injected is thus well 
 denned and distinctly limited. If 6 is the temperature of 
 ebullition corresponding to the pressure of the vapor of 
 water in the mixture, each kilogramme of steam injected will 
 absorb a quantity of heat equal to CpoT-t (^2 ^)- This 
 quantity is computed below, assuming for simplification 
 that 6 is equal to the temperature of ebullition at the pres- 
 sure p (page 160). 
 
 The injection of water absorbing ? calories per kilo- 
 gramme of gas burned, it is possible to increase the amount 
 of heat introduced by an equal amount without modifying 
 the temperatures T 2 and T 3 , provided the calorific power 
 of the combustible will permit it. 
 
 It will thus be found, if we take the same efficiency for 
 the regenerator (0.75, for example), that the total useful 
 effect obtained differs very little from that secured by the 
 use of a regenerator heating the compressed air. Never- 
 theless the actual consumption of air per effective horse- 
 power is less, a fact which has a distinct practical advantage. 
 
 This method is especially applicable when the nature 
 of the combustible permits the introduction of a large 
 amount of heat, and when the exhaust is discharged at a 
 reduced pressure. 
 
160 
 
 THE GAS TURBINE. 
 
 3as?s 
 
 T* rH 1C ^ rH 
 
 O O 
 
 s a 
 
 s3sci|l 
 
 
 
 o "* 
 
 <M X" t- 
 
 Qi OO QQ 
 
 IP rH rH 
 
 45 
 
 CO C<j t> ^ OO 
 
 
 
 ^D OO 
 iC <>l 
 
 o o 
 
 ^s^ilS 
 
 8 - 
 
 o *"" 
 
 8 S S a a ^ ^ ^ 
 
 gs 33^8*3^23 
 
 QO S 
 3^ w^ ^T 1 
 
 rH 01 
 
 S 8 
 
 3 > O ^ 
 
 ^ 01 W ^ W ^. ^ 
 
 0:1 rH .-: O O 
 
 O rH 
 
 cQoaoQoacntHcKQQ 
 
 QlQ)Q)QjQ>Q>Q)Qj . . . . 
 
 C g 'C *C 'C "C 'C 'C 
 
 ,OaOOOOQQ 
 
 1 li s g"s 
 
 S 6 
 
THE GAS TURBINE. 161 
 
 The regenerator may be made in the form of a boiler 
 similar to the Serpollet flash boiler, or of the type proposed 
 by Colonel Renard. 
 
 The gases enter the regenerator at a temperature of 700 
 degrees, and leave it at about 400 degrees absolute. The 
 water, raised from a temperature of zero to 450 or 500 
 degrees absolute, will be vaporized at this latter tempera- 
 ture, at a pressure of about 5.30 atmospheres. It is easy 
 to compute that the mean drop in temperature will be about 
 100 degrees in the boiler, and 75 degrees in the regenerator, 
 corresponding to a heat transmission of 3000 and 1500 
 calories respectively per square metre per hour. This will 
 require about 0.0366 square metre of surface per kilogramme 
 of air consumed per hour in the turbine, or about 0.16 square 
 metre per horse-power delivered on the shaft, the consump- 
 tion per horse-power hour being 4.25 kilogrammes of air, 
 and 0.51 kilogramme of water, for a combustion pressure 
 of 30 atmospheres. 
 
 The necessary heating surface will therefore be of the 
 same order of magnitude as that of the condenser of an ordi- 
 nary marine engine, but probably greater than that of a re- 
 generator for a gas turbine using a regeneration of gas to gas. 
 
 Practically, vaporization under pressures exceeding 
 30 atmospheres may appear to offer certain difficulties. 
 This method of regeneration, however, becomes very simple 
 if the exhaust is discharged at reduced pressure. The pres- 
 sure of combustion, for example, being from 5 to 10 atmos- 
 pheres, and the exhaust pressure | atmosphere. The regen- 
 erator-boiler should be operated at pressure ranging only 
 from 5 to 10 atmospheres. The drop in temperature would 
 be materially increased, which would facilitate the trans- 
 mission of heat. At the same time, the efficiency of the cycle 
 would be increased. 
 
 Under such a system, using a producer of the Gardie 
 type, operating under 5 to 10 atmospheres pressure, the 
 
 . 11 
 
162 THE GAS TURBINE. 
 
 loss of the sensible heat of the gas could be avoided, and 
 the proportion of steam or water injected increased. 
 
 It may be noted that in the case of compressors using 
 water injection, the vapor produced from the injected 
 water is evolved with the gaseous mass, and permits an 
 increase in the amount of heat introduced, thus improving 
 the efficiency. 
 
 The Use of Large Injections of Water in Connection with a Very 
 Rich Fuel. Turbines Using Liquid Oxygen. 
 
 As a matter of curiosity it may be noted that if pure 
 oxygen be used in the combustion, the total weight of gas 
 burned would be only about one-fourth that otherwise 
 required; and therefore, the introduction of heat being 
 quadrupled, might reach 2000 calories per kilogramme. 
 The injection of water into the combustion chamber might 
 then be materially increased. Such a mixed turbine would 
 require a much smaller compressor, consuming much less 
 power, or if liquid oxygen were used a small centrifugal 
 pump operating at high pressure would replace the air 
 compressor. 
 
 A machine of this kind would require three such pumps; 
 one for the liquid oxygen, one for the liquid fuel, and the 
 third for the water. A tubular heater, heated by the exhaust 
 gases, would heat the water and vaporize the liquid oxygen, 
 the only other elements required being the combustion 
 chamber and the turbine wheel. 
 
 The temperature of combustion would be the same as 
 before, but the temperature of the exhaust would be materi- 
 ally lowered by reason of the calories absorbed by the vapor- 
 ization of the oxygen, so that the thermal efficiency should 
 be at least equal to that computed above. 
 
 With regard to the mechanical efficiency >?, this, neglect- 
 ing the work absorbed by the pumps, would be above 0.70, 
 because of the absence of the compressor. The total useful 
 
THE GAS TURBINE. 163 
 
 effect, pi], would therefore be 0.70 or 0.75 times 0.70, or 
 about 50 per cent. Although the amount of work available 
 would thus be very high, the velocity of discharge of the 
 mixture would be much greater than in the ordinary case, 
 and the mechanical efficiency of the turbine would be lower. 
 
 Such a machine, however, would be extremely light. 
 It is true that it would be necessary to carry 4 kilogrammes 
 of liquid oxygen and 5 kilogrammes of water for every kilo- 
 gramme of petrol, but for certain applications the final 
 result would be very favorable. 
 
 While this application of the gas turbine is yet within 
 the domain of scientific curiosities, it is by no means an 
 absurdity. M. Cailletet has not hesitated to propose a 1 
 similar combination, using piston engines, for the design 
 of extremely light and powerful motors for aerial or sub- 
 marine navigation. 
 
 * * v 
 Limitations of the Temperature of Expansion. Injection of 
 
 Water, Steam, or Cool Gases after Expansion. 
 
 If the exterior of the expansion nozzle is cooled, the 
 expansion is no longer adiabatic* and cannot be subjected 
 to computation. All the energy thus abstracted, however,. 
 is evidently lost. 
 
 The same is not the case if the expanding gases are 
 cooled by an injection of water, since the vapor thus formed 
 is added to the fluid mass. Nevertheless, at a temperature 
 of 700 degrees, and at atmospheric pressure, about -^ of the 
 calories absorbed by the injection are lost and absorbed by 
 the vaporization properly so-called. 
 
 We are therefore led to consider the injection of steam. 
 If the velocity of the steam is lower than that of the current 
 of gases, there is caused, as we shall see, an important loss 
 of energy. Let us then assume that the two currents have 
 the same velocity. In order to accomplish this, it is neces- 
 sary that the vapor be generated at a pressure higher than 
 
164 THE GAS TURBINE. 
 
 that of the gas in the combustion chamber. We will pass 
 over this difficulty. In order to obtain a better result than 
 is secured by the direct injection of water the steam must 
 be regenerated by the use of waste heat. Under these con- 
 ditions, and assuming that the expanded steam is still 
 saturated, dry, or slightly superheated, and calling x the 
 weight of this steam delivered for each kilogramme of air, 
 calculated as heretofore, we may complete the temperature 
 7y of the expanded gas. 
 
 Cp(ZY-700) = 0.48(77-373)3 
 
 167-180* 
 whence '* -0.24-0.48z* 
 
 We may then calculate the new temperature of combustion, 
 the new introduction of heat, and the new efficiency: 
 
 With a coefficient of regeneration of 0.75 we may inject 
 12 to 13 per cent, of water, and permit a final temperature 
 of expansion of the gas of 800 degrees absolute instead of 
 700 C. 
 
 It is thus seen that for a given compression ratio, the 
 useful effect is slightly lower than that obtained by injecting 
 the steam before the expansion. In practice the injection 
 of steam after the expansion, offers considerable difficulties 
 of a kinetic order. 
 
 We will now consider the injection of cool .gases. 
 
 Injection of Cool Gases at Low Velocities. 
 
 Stodola has shown in the following manner that a mix- 
 ture of two currents of gases having two different velocities 
 V)i and w 2 results in a material loss of kinetic energy. 
 
 There are two cases to be considered. The first corre- 
 sponds to the use of a mixing chamber so formed as to per- 
 mit the operation to be effected without raising the pressure. 
 
THE GAS TURBINE. 165 
 
 The second case corresponds to the use of a cylindrical 
 chamber, which leads to an elevation in the final pressure. 
 If we call dP x the force acting axially upon an element 
 dm, and call 77^ 77 2 , and n = IJ l +II 2 the flow by weight, the 
 theorem of quantities of motion gives: 
 
 ndt fn^dt n 2 dt \ 
 w [ -lWi H w 2 } = 2dtdP 3 
 Q \ Q a 
 
 g 
 
 from which we get: 
 
 IJw=II l w 1 
 
 This is the formula for impact of non-elastic bodies, 
 and the loss of energy is: 
 
 _ 1/77! 2 77 2 2 \ 1/7 
 Z = - -^ X + 2 iu 2 ) -s 
 2\ 2 J 2 
 
 w . 
 g g 
 
 If we call ^3 and the respective temperatures of the 
 two gaseous currents before mixture, and T 3 ' the tempera- 
 ture after mixture, we have: 
 
 ~ = n 2 (T,' 
 
 These three relations enable us to compute the tempera- 
 tures and the efficiency. 
 
 Let us take the extreme case in which the gas is 
 injected cold and without velocity. We have : 
 
 w 2 = 0; and w = -~ w \ 
 
 nji 2 wf 
 
 whence -JT20* 
 
 The ratio of the lost energy to the amount of energy 
 available in the gaseous current before the mixture will then 
 be, for this particular case* 
 
 For example, if # 2 = 1 kilogramme, and it is desired to 
 reduce the temperature to T' 3 = 700 degrees by injecting 
 
166 THE GAS TURBINE. 
 
 U 2 kilogrammes of air without velocity at 300 degrees abso- 
 lute (or #=300), the limiting case corresponding to T 3 = T' 3 
 
 will be attained when -r = 95 77 2 . 
 
 If x be the thermal equivalent of the kinetic energy of 
 1 kilogramme of burned gas before the mixture, we have: 
 
 -r=-ffXf whence jfx-=Q5U 2 , and U = ~ 1 
 A 11 11 yo 
 
 For example, 
 
 for /=100 200 300 400 calories 
 
 we have 77 2 >0.05 1.10 2.15 3.20 calories 
 
 and e>0.05 0.52 0.68 0.76 calories. 
 
 There is, therefore, a considerable loss in the kinetic 
 energy of the gaseous current when the latter attains a con- 
 siderable value. 
 
 Suppose that we are using a cylindrical mixing chamber. 
 The pressure beyond the zone of mixture will then be higher 
 than that in front of it. Professor Stodola, who has ex- 
 amined this question, finds that there may be two solutions, 
 and that the velocity of the mixture may have two distinct 
 values. One of these corresponds to a simple mixture, 
 with a loss of kinetic energy and a relatively moderate rise 
 in temperature. The other corresponds to a velocity greater 
 than that of sound and involves the existence of a shock 
 of compression of which we shall speak hereafter. However 
 the mixture may be effected there is no more advantageous 
 result to be expected than in the preceding case, and there 
 is nothing to be deduced from the idea other than the results 
 involved in progressive mixtures in successive chambers. 
 
 Injection of Cold Gases at the Same Velocity as the 
 Principal Current. 
 
 Suppose now that the cold gases are given, by the 
 use of a blower or similar apparatus, a velocity equal to 
 that of the principal current, and that the mixing chamber 
 is of such a shape that there is no increase in pressure. In 
 this case there is theoretically no loss of energy. 
 
THE GAS TURBINE. 167 
 
 It is, however, necessary to expend, in driving the blower, 
 an amount of energy equal to the necessary kinetic energy. 
 The question then presents itself as follows: Is it more 
 advantageous to compress all the air required and deliver it 
 at once to the combustion chamber, or to compress only a por- 
 tion of it to the pressure of combustion, and to cause the re- 
 mainder to be delivered by a blower to the mixing chamber? 
 
 Let us suppose, for example, that we have a combustible 
 capable of permitting an introduction of about 520 calories 
 per kilogramme of mixture. If we compress to 10 kilo- 
 grammes per square centimetre, we can introduce only about 
 260 calories per kilogramme to exhaust at 700 degrees. If 
 we do introduce 520 calories we shall have a temperature 
 of combustion of 2500 degrees and an exhaust of 1270 
 degrees. The kinetic energy will be equivalent to 290 
 calories per kilogramme of gas. In order to bring the tem- 
 perature of 1270 to 700 it will be necessary to mix with 
 each kilogramme of burned gases 1.32 kilogramme of air, 
 and the kinetic energy to be imparted to this cold air will 
 be equivalent to 1.32x290=410 calories. We then have 
 at the outlet of the mixing chamber a kinetic energy equiva- 
 lent to 290+410=700 calories for 2.30 kilogrammes of 
 mixture. We will get in work on the shaft 0.7 X700 =490. 
 Now, the work required for the compressor will be 48 calories 
 and for the blower 410, a total of 458. There will therefore 
 each have to give an efficiency of 0.94 in order that they 
 should not absorb more power than the turbine itself pro- 
 duces. The injection of cold gases is therefore wholly 
 impracticable. 
 
 V. 
 
 Combination Cycles. The Adaptation of a Second 
 Engine to Utilize the Waste Heat. 
 
 It has been suggested that the waste heat discharged 
 by a gas turbine should be utilized to operate a second tur- 
 bine, employing sulphurous acid gas, for instance, or even 
 vapor of water. We have already shown that the exhaust 
 
168 THE GAS TURBINE. 
 
 gases should be discharged at a temperature of about 700 
 degrees C., absolute, and that it is not advantageous to 
 lower this temperature by diminishing the amount of heat 
 introduced. 
 
 The heat abstracted from the gas during compression 
 may be carried off by water circulating about the compres- 
 sion cylinders and through the inter-coolers. The amount 
 of heat thus withdrawn compares in importance with that 
 escaping with the exhaust gases, but its temperature is 
 much lower. Theoretically, the temperature of the jacket 
 water should not materially exceed that of the atmosphere, 
 and in no case should it be higher than 50 to 100 degrees C. 
 It is therefore necessary to resort to some substance having 
 a low boiling point, such as sulphurous acid, in order to 
 utilize this heat in a secondary engine. 
 
 We have at our disposal from 150 to 200 calories per 
 kilogramme of gas burned. If we use a steam turbine as 
 the secondary motor, operating at a pressure of 20 kilo- 
 grammes per square centimetre, superheating to 700 degrees 
 absolute, and operating condensing, the thermal efficiency 
 will be: 
 
 The total useful effect will then be: 
 ^=0.70X0.265 = 0.185. 
 
 Now if the vapor of water had been used directly with 
 the gases of combustion it would have given a useful effect 
 of about 0.30. 
 
 It has been proposed to replace the vapor of water by 
 a gas which is readily liquefiable, such as sulphurous acid. 
 According to Professor Josse an indicated horse-power may 
 be obtained in the secondary motor with a consumption 
 
 * This result agrees with practice, since it corresponds to a consumption 
 of 4 kg. (8.8 pounds) of steam per horse-power hour, and consumptions below 
 4.6 kg. have already been obtained. 
 
THE GAS TURBINE. 169 
 
 of 7800 calories, or even with 5000 calories when operating 
 at a pressure of 25 atmospheres (90 degrees C.) in the 
 condenser. 
 
 This last figure gives a thermal efficiency of 0.127, or a 
 net useful effect of 0.7 X0.127, or 0.089. 
 
 This result might be materially improved if we could 
 permit the superheating of the sulphurous acid gas without 
 causing corrosion upon the parts of the turbine with which 
 it came in contact. 
 
 In any case a secondary turbine would permit a recupera- 
 tion of 150 to 200 calories X0.09 = 14 to 18 calories, if we 
 use sulphurous acid, or 92x0.185 = 17 calories, if we use 
 water, admitting a coefficient of recuperation /* equal to 
 unity. Taking /i=0.75, we get work equal to about 13 
 calories per kilogramme of gas. Now the net mechanical 
 effort y^u realizable per kilogramme of gas, with or without 
 recuperation, varies between 25 calories and 200 calories 
 when the pressure of combustion varies from 5 to 100 
 atmospheres. 
 
 The amount of work recoverable by the use of a second- 
 ary machine is therefore not of sufficient importance to 
 warrant the complication of a separate machine to secure it.* 
 
 VI. 
 
 Conclusions from the Thermodynamic Study of the Gas Turbine. 
 
 Method of Development to be Adopted. Probable Efficiency. 
 
 Probable Divergence Between Theory and Practice. 
 
 The study of the gas turbine from a thermodynamic 
 point of view does not appear to reveal any combination 
 capable of giving results greatly differing from those already 
 obtained from the latest improved gas engines. The high 
 thermal efficiencies theoretically probable seem to be offset 
 by the low mechanical efficiency. But this latter is capable 
 
 * In practice the relative importance of the work recovered might be a 
 little greater, since all losses of energy have the effect of increasing the amount 
 of heat in the exhaust gases, and of such leaks we have taken no account. 
 
170 
 
 THE GAS TURBINE. 
 
 of improvement, so that there remains a margin for progress 
 which is encouraging for the future. 
 
 The analysis which we have undertaken may be reviewed 
 as follows: 
 
 1. Combustion under constant volume, as compared 
 with combustion under constant pressure, shows, for the 
 same initial pressure, a better efficiency, while at the same 
 
 030 
 
 1000 2000 8000 
 
 FIG. 63. Efficiencies for various combustion temperatures. 
 
 time it permits the use of a less important compressor. 
 Nevertheless, the absolute value of the efficiency is not 
 greater, because we are more promptly limited by the maxi- 
 mum limit of permissible temperature of combustion T 2 . 
 This is true either for the specific power, or for the consump- 
 tion of air per horse-power hour. 
 
THE GAS TURBINE. 171 
 
 This method is advantageous, therefore, only from the 
 point of view of the necessary compression ratio. This is 
 a matter for consideration if we limit ourselves to the use 
 of rotary compressors. In practice the mechanical efficiency 
 is low because of the kinetic losses due to irregularities of 
 flow under varying operation, besides the inconveniences 
 attending an explosion machine. As a matter of fact, the 
 explosion turbine is applicable only to very small powers, 
 and for machines of light weight, in which the efficiency is 
 a matter of secondary importance, and preliminary compres- 
 sion is undesirable. 
 
 2. Isothermal combustion involves excessively high 
 ratios of compression, and is otherwise not practically 
 realizable. 
 
 3. It follows that the best method available for the gas 
 turbine corresponds to that for the gas engine; namely, 
 combustion under constant pressure, with a preliminary 
 isothermal compression. 
 
 4. If the ratio of the extremes of pressure has a given 
 value, it is immaterial whether these pressures are high or 
 low, in an absolute sense. This point is of interest in con- 
 nection with the question of exhausting at low pressure, 
 and with the use of multiple rotary compressors. 
 
 5. The temperature of the exhaust should be as high 
 as practicable, with regard to the maintenance of the revolv- 
 ing wheels. It is deceptive to attempt to lower it by pro- 
 longing the expansion by the use of an air pump. 
 
 6. The best method of saving the heat escaping in the 
 exhaust is by a simple tubular regenerator transferring the 
 heat from outgoing to incoming gas. Regeneration by means 
 of a steam boiler is worthy of consideration only for very 
 rich combustibles, and the best plan then is to deliver the 
 steam into a combustion chamber. 
 
 It may now be asked what important relations may be 
 established practically between the above theoretical deduc- 
 
172 THE GAS TURBINE. 
 
 tions and the practical results attainable with such machines 
 as may actually be constructed. 
 
 It is probable that the practical cycles will differ from 
 the theoretical ones in the gas turbine much as they do in 
 the gas engine, but to a less extent. 
 
 Thus, as concerns the compression', there are two differ- 
 ences between theory and practice in effecting isothernal 
 compression. One is the increase in the work of compression; 
 the other, the elevation in temperature of the compressed 
 gas, reducing the value of Q, and consequently the specific 
 power. The results in practice lie between those computed 
 for isothermal compression and those corresponding to 
 adiabatic compression. But, as we have seen, the difference 
 is not very great, and we have taken a sufficiently low value 
 for the mechanical efficiency >? c to cover any discrepancy 
 on this account. 
 
 With respect to the combustion, there are more import- 
 ant divergences between theory and practice, which must 
 be taken into account. Thus, we have assumed that the 
 reaction is effected in surroundings w T hich are strictly adia- 
 batic. This cannot be effected in practice, and notwith- 
 standing all our precautions a loss of heat will occur. 
 
 The combustion will also be incomplete, hence there will 
 be a loss of a portion of the combustible, or a partial dissocia- 
 tion, this being less probable under pressures of 30 to 40 
 atmospheres. 
 
 These three, causes have one and the same result, an 
 increase in the weight of combustible consumed per horse- 
 power hour. This, however, does not affect the general 
 development, especially if the fuel is in the liquid or solid 
 state, since the additional amount of fuel required does not 
 affect the work of compression. 
 
 The specific heat of the products of combustion differs 
 materially from that of air, and varies with the temperature; 
 and it is probable that the value taken for the temperature 
 
THE GAS TURBINE. 173 
 
 of combustion is greater than the real value. It follows 
 that the introduction of heat is more limited than in our 
 -calculations, at least in the case* in which this limit is fixed 
 by -the calorific value of the combustible. But since this 
 limit is not likely to be reached in practice the only effect 
 resulting from the disagreement between theory and prac- 
 tice is to reduce the amount of air required for dilution. 
 
 Finally, the expansion is not strictly adiabatic, and 
 the form of the expansion curve is not precisely that which 
 corresponds to the relation 
 
 pv l ' 4l = constant. 
 
 Thus, there is a loss of heat which may be small, but 
 can never be strictly zero. Besides, the gas is heated to a 
 certain extent by friction. The true law of the expansion 
 can be determined only by experiment. In any case the 
 exponent 7 in the formula will differ from 1.41 because we 
 are dealing with gases other than air and because the ratio 
 
 C 
 '- is not constant when the temperature varies between 
 
 C 
 
 very wide limits. 
 
 Even taking into account the variability of the specific 
 heats with the temperatures, M. Vermand has shown that 
 the law of Poisson is expressed practically by the relation: 
 
 p v y = constant 
 
 when 7-=i-f 
 
 in which a =0.162, so that for air 7- = 1.441. 
 
 If we use data obtained from certain trials of gas engines, 
 we are led to accept for 7- values such as 1.3 to 1.2. 
 
 The corresponding results differ materially from those 
 which we have computed above. 
 
 Thus, to obtain the theoretical temperature of 700 
 degrees for the exhaust, with a combustion pressure of 30 
 atmospheres, it is necessary to produce a combustion tern- 
 
174 
 
 THE GAS TURBINE. 
 
 perature of 1880 degrees, introducing 375 calories per kilo- 
 gramme of air. 
 
 These are the results obtained above for 7- = 1.4 (giving 
 <o = 0.57). 
 
 If now, we take 7- = 1.3, or /--1.2 we shall have: 
 
 snce - 
 
 '--=0.23 -=0.17 
 
 r r 
 
 7 7 2 =1526 7 7 3 = 1253 
 
 9QK /^) 9*30 
 
 _-./*) v^ ^Ov/ 
 
 ^? = 0.45 ^ = 0.29. 
 
 The efficiency is influenced, as we see, to a remarkable 
 extent by the variation of 7-; and the diagram (Fig. 64) shows 
 
 FIG. 64. Variations of exponent of expansion. 
 
 that it varies almost in proportion to 7- 1, at least in the 
 case under consideration, in which the cycle uses combus- 
 tion at constant pressure and isothermal compression.* 
 
 * The thermal efficiency becomes zero for > = 1, for we have =* = (30)= 1 
 and Q then becomes zero. 
 
THE GAS TURBINE. 175 
 
 It may be of interest to note the following values for 
 
 C 
 
 F = -, at a temperature of zero, and at atmospheric pres- 
 c 
 
 sure, for the gases named: 
 
 H,0,N, Air, CO 1.41 
 
 H 2 O 1.34 
 
 CO 2 1.29. 
 
 In engines utilizing the explosion of gases behind a 
 piston, the value of the exponent f has been deduced from 
 the form of the expansion curve, and the figures thus ob- 
 tained range from 1.3 to 1.6. In such machines, however, 
 the action of the walls of the cylinder play an important 
 part, while in the diverging nozzle of a gas turbine this 
 action is reduced to a minimum, because a continuous flow 
 is maintained. 
 
 However this may be, the true law of expansion in the 
 nozzle of a turbine constitutes the principal unknown 
 practical element which presents itself in the gas turbine. 
 It has even been maintained that 7- may become equal to 1, 
 and that the expansion may be accompanied by no drop in 
 temperature, and hence be incapable of producing any use- 
 ful effect.* 
 
 Some experimenters have not been able to find the drop 
 in temperature by thermometric observations, and have 
 attributed this fact to the heat developed by the friction 
 of the gases upon the thermometer. We cannot go into this 
 objection at length. The expansion doubtless follows the 
 formula of Poisson, pv y = constant, but the true value of 
 the exponent 7-, and consequently of the efficiency, can be 
 determined only by experiment. 
 
 We may note here a final reason for the discrepancies 
 in our calculations between theory and practice. This is 
 
 * See Charles E. Lucke, Ph.D., Practical Investigations in the Gas 
 Turbine Problem; Engineering Magazine, April, 1905. The Gas Turbine, 
 Engineering Magazine, August, 1906. 
 
176 THE GAS TURBINE. 
 
 the relative inexactness of the simple physical laws which 
 we have accepted as relating to the substances under con- 
 sideration: the laws of Mariotte, of Gay Lussac, of the 
 constancy of specific heats, etc. In all standard works there 
 may be found formulas which are more precise than those 
 which we have used, and these may be substituted for the 
 more simple laws. The greater degree of precision thus 
 obtained is of minor interest, since the inevitable uncertain- 
 ties of the question render any such excessive precision 
 illusory. 
 
 Influence of the Nature of the Combustible. 
 
 Before leaving the thermodynamic study of the gas 
 turbine it is desirable to examine whether the nature of the 
 fuel available may have any important influence upon the 
 possible efficiency. 
 
 In order to use the most advantageous cycles it is desira- 
 ble, from what we have already seen, to be able to introduce 
 from 375 to 415 calories, if we do not inject any steam, and 
 from 470 to 524, if steam injection is to be used; the pres- 
 sure ranging from 30 to 40 atmospheres. 
 
 Now, even using lean gases, such as that made in the 
 Dowson producer, or the waste gases from blast furnaces, 
 with a heating value of 800 calories per cubic metre (about 
 90 B.T.U. per cubic foot), it is possible to introduce about 
 460 calories per kilogramme of mixture; while with the 
 richer gases, such as illuminating gas, acetylene, etc., we 
 may get from 500 to 600 calories. The nature of the com- 
 bustible has, therefore, a minor influence from this point 
 of view. The constitution of the burned gases varies but 
 little for the different combustibles, so that the specific 
 heats are not greatly different. The cycles calculated upon 
 the actual composition of the mixtures will therefore agree 
 fairly well with those which we have based upon the prop- 
 erties of air. 
 
THE GAS TURBINE. 177 
 
 It also follows that the total weight of gas to be com- 
 pressed varies but little, and the same is true of the work 
 required for compression. When a liquid fuel is used a 
 greater amount of air is required per kilogramme of combus- 
 tible than with a gaseous fuel. 
 
 It is an error to assume, as has sometimes been done, 
 that gaseous fuels are less easily employed than liquid fuels. 
 This may be the case for motors of the Diesel type because 
 the intermittent action brings in the important question 
 of ignition. Apart from the necessity for two separate 
 compressors, however, the compression is not more trouble- 
 some when a gaseous fuel is employed. Since the combus- 
 tion is continuous in the case of the gas turbine, the question 
 of ignition is of secondary importance. 
 
 In the accompanying table the computations have been 
 made according to the stated compositions of the various 
 gaseous mixtures, taking the data calculated by M. Vermand. 
 
 It will be seen that C P varies about 20 per cent., and 
 f about 1 per cent., in passing from one mixture to another. 
 
 The composition of the burned gases varies but slightly. 
 Nitrogen predominates, being 62 to 74 per cent., followed 
 by carbon dioxide, 12 to 33 per cent. Oxygen appears to 
 be present in very small quantities, so that oxidation of 
 metallic parts need hardly be feared. 
 
 The number of cubic metres of air and of gas to be com- 
 pressed to correspond to the introduction of an amount 
 of heat equal to 100 calories into the cycle, gives an idea 
 of the necessary capacity for the compressor, or compres- 
 sors. As this quantity varies from 148 to 180 litres, a differ- 
 ence of 22 per cent., this is not an element in which a serious 
 error need enter. 
 
 In like manner the total weight of gas to be compressed 
 
 per 100 calories introduced forms a measure of the total 
 
 power required for the compression. The extreme limits 
 
 are 0.17 and 0.22 kilogramme, a difference of 30 per cent. 
 
 12 
 
178 
 
 THE GAS TURBINE. 
 
 
 s 
 
 Sa 
 
 0) 
 
 
 
 
 
 1! 
 
 fcH 
 
 I! 
 
 a CO CO O rH ' 
 
 g CN co ^ co ; 
 
 gj rH rH rH rH ] 
 
 III 
 
 C30 rH CO GO CO 
 
 
 
 
 
 
 o|3 
 
 O O O O O 
 
 
 
 j 
 
 
 
 S3 
 
 
 
 
 
 
 
 o c fc" 
 
 
 
 
 'Si 
 
 Q) 
 
 CO 1C Tf* t 
 
 
 
 
 
 
 
 CM O O CO 
 
 Ijl 
 
 
 
 
 gas burned 
 
 * 
 
 p ic p p 
 co cd csi "t 
 
 t^ O CO 1>- 
 
 111 
 
 ^ 1C O C^l <M 
 
 vj; l^ rH (M O5 O5 
 
 ^ rH <M <M rH rH 
 
 O O C5 O 
 
 
 
 o 
 
 
 
 4H M n-i 
 
 
 
 
 
 
 M 
 
 O 1C O 05 . 
 
 S"!^ a S 
 
 1C Ttl CO CO Tt< 
 
 
 
 | 
 
 8 
 
 CN CO CO TH 
 i 1 CM CO rH 
 
 |l|ir| 
 
 ^ GO GO i> I>- t s> * 
 
 " 
 
 
 
 c 
 
 I 
 
 w 
 
 ^ 05" rH CO '. 
 
 IP 1=1 
 
 CNJ 00 (M ^ 
 
 1 rH CO rjn o 
 C> O 
 
 
 
 
 
 <M CO rH CO 
 
 o^ -g 
 
 iS^hi 
 
 CO CO Tf TH rH 
 1 t^ rH CO !> t^ 
 
 
 
 
 
 rH rH rH rH 
 
 I s 1 
 
 O O C5 O O 
 
 
 
 1 
 
 o 
 
 1C -* 1 CM <M 
 
 1C CO rH O 
 CO CO CO CO 
 
 o o o o '. 
 
 Ipf! 
 
 . rH O5 rH CO CO 
 
 oc O^ *O "^ O^ O5 
 
 ^ o o o o o 
 
 
 
 
 
 
 
 ^ > 
 
 
 
 
 H 
 
 
 
 *- S 0! 
 
 
 
 
 
 
 r* rH CO CO 1C 1C 
 O I"- CO 1C CM rH 
 
 ii-f^ii 
 
 . CO CO O rH 
 be C^ CO rH CO 
 
 
 
 
 
 
 SP.2 o ^ 
 
 ^ rH rH rH rH 
 
 
 
 
 
 i 
 
 &i s s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ! : . . 
 
 
 
 
 
 
 
 
 
 
 
 
 ! 
 
 i 
 
 " f-t 
 
 rH 
 
 6 
 
 S-H 
 
 
 
 
 . 
 
 -? $) 
 
 *3 
 
 >^ ti) 
 
 
 
 > 
 
 
 '* rH -3 I S 
 
 x 
 
 a 
 
 N * pM 
 
 "3 ' r-J CO 
 
 . JH O (M 
 
 
 
 c 
 
 i 
 
 *o *S3 + 
 
 "S 
 
 ^ * CO r- + 
 
 
 
 
 
 
 
 > . .S 02 
 
 
 
 f 
 
 
 50 1 + *. ' 
 
 o 
 
 a 
 
 'o o .g 
 
 
 
 ] 
 
 ! 
 
 |2 ||i 
 
 1 
 
 |2|^1 
 
 
 
 1 
 
 i 
 
 > i 1 + 
 
 8 
 
 c 3 "^ 1 + i 
 
 
 
 
 
 -S tjo S a) ~ 
 
 
 ^^ Vin f^ O /C 
 
 
 
 
 
 I II ~ 
 
 
 fl jj ^ 
 
 
 
 
 
 B P -^ &VP 
 
 r2 O ^ O "S 
 
 P^ Q PQ <j PH 
 
 
 llltl 
 
 PH Q pQ ^ PH 
 
 
THE GAS TURBINE. 
 
 179 
 
 The lean gases are less advantageous in this respect than 
 the richer mixtures. 
 
 We may therefore reach the conclusion that the nature 
 of the combustible (liquid or gaseous) is not a matter of 
 great importance from the point of view of the total probable 
 efficiency py, but that the lean mixtures are less advanta- 
 geous than rich mixtures because of the unfavorable influ- 
 ence upon the mechanical efficiency when the work of com- 
 pression acquires great importance. 
 
 The Gas Turbine from a Mechanical Viewpoint. 
 Flow of Gases Through Nozzles. 
 
 According to the principle of the conservation of energy, 
 the velocity of discharge w 2 is given if we neglect the initial 
 velocity w l of the gas, by the relation: 
 
 u\ 
 
 if the evolution undergone by the gas corresponds to an 
 isothermal compression, followed by an isobaric combustion. 
 For the group of cycles exhausting at 700 degrees, we 
 have: 
 
 Compression ratio. 
 
 5 
 
 10 
 
 15 
 
 20 
 
 25 
 
 30 
 
 40 
 
 60 
 
 80 
 
 100 
 
 Q g= calories . . 
 
 103 
 
 162 
 
 200 
 
 236 
 
 258 
 
 283 
 
 323 
 
 388 
 
 430 
 
 471 
 
 w 2 = metres per second 
 
 927 
 
 1162 
 
 1288 
 
 1400 
 
 1466 
 
 1533 
 
 1637 
 
 1794 
 
 1892 
 
 1975 
 
 These velocities exceed those of steam in ordinary 
 steam turbines when compression ratios of 15 to 20 are 
 exceeded. For this reason, under similar structural con- 
 ditions for the revolving disc, the efficiency of the disc itself 
 will be less. Fortunately other factors act in the opposite 
 sense, and in favor of the gas turbine. 
 
 The power delivered per unit of terminal section of a 
 nozzle is greater in the case of the gas turbine. If the gas 
 
180 
 
 THE GAS TURBINE. 
 
 escapes at a temperature of 700 degrees C. absolute, and 
 at atmospheric pressure, the weight of a cubic metre is 
 about 0.5 kilogramme. The discharge, by weight, will 
 then be as follows: 
 
 for w 2 = 1000 m / s 1500 m / s 2000 m / s 
 
 the values 1.8 kg. 2.7 kg. 3.6 kg. 
 
 per square millimetre of cross-section. 
 
 Calories 
 
 eooo, 
 
 1000 
 
 o 
 
 FIG. 65. Velocity of discharge of 
 steam from nozzles. Cycle with isobaric 
 combustion; exhaust at 700 degrees. 
 
 
 
 10 
 
 FIG. 66. Energy in calories de- 
 livered per square millimetre of termi- 
 nal cross-section of nozzle. 
 
 The power delivered in the form of kinetic energy per 
 square millimetre of section will then be equivalent to the 
 following quantities of heat, if we consider the same group 
 of cycle as before: 
 
THE GAS TURBINE. 
 
 181 
 
 Pressure of combustion. 
 
 5 
 
 10 
 
 20 
 
 30 
 
 40 
 
 60 
 
 80 
 
 100 
 
 Discharge in weight kg . ... 
 
 16 
 
 2.1 
 
 25 
 
 2.75 
 
 295 
 
 320 
 
 34 
 
 36 
 
 Power delivered, calories 
 
 165 
 
 340 
 
 590 
 
 780 
 
 950 
 
 1240 
 
 1460 
 
 1700 
 
 If the exhaust takes place under a pressure reduced to 
 - atmosphere the figures for the power delivered will be 
 
 reduced in the same proportion. 
 
 In a steam turbine operating with a pressure of 10 
 atmospheres and against a pressure of yV atmosphere in the 
 condenser, the power delivered per square millimetre of 
 nozzle is equivalent to only 60 calories. 
 
 The Formula of Saint Venant. 
 
 If a gas flows through a passage without friction, the 
 initial and final pressures being aj t and <o 2 , and the velocities 
 Wi and w 2 , the adiabatic flow is represented by the formula 
 of Saint Venant: 
 
 2 2 ^\ 
 
 Wi I r 
 
 r / ^-F^' 
 
 JiO 2 
 
 jf-w* r 
 
 (3) 
 
 or " 2g r l 1 (^ 1 -^). (4) 
 
 Velocity of a Gas Flowing from a Reservoir through a Nozzle. 
 
 If we assume the velocity of approach w l as negligible, 
 we have at any point of the nozzle corresponding to a pres- 
 sure p x , a velocity w x given by the equation: 
 
 w x = 
 
 and since 
 
 (5) 
 
 
182 
 
 THE GAS TURBINE. 
 
 we have 
 
 Vr 
 ^J/s-y 
 ri vJt\<t>i/ 
 
 4fe<fc&t& 
 
 FIG. 67. Diagram of nozzle velocities. 
 
 (6) 
 
 f | 
 
 FIG. 68. Diagram of nozzle sections. 
 
 
 We readily find that the minimum value of s x (or of ) 
 
 w x ' 
 
 corresponds to a pressure a> m , given by: 
 
 (7) 
 
 whence 
 
 (8) 
 
 / 
 
 IT /o r /^mXY 
 
 /7 = S m -t /2^ f- rf--) 
 V * f + lXttt/ 
 
 and 
 
 For air we have 7- = 1.4 and hence: 
 
 (9) 
 
 whence 
 
THE GAS TURBINE. 183 
 
 It can be demonstrated that p m cannot fall below this 
 value, and that w m cannot exceed the velocity of sound.* 
 
 To release the air without loss of energy, down to the 
 pressure of the atmosphere; or, more generally, down to any 
 given pressure p 2 , we must therefore use: if we have w 2 > 
 0.529^, a converging nozzle; if we have w 2 < 0.529^, a 
 converging-diverging nozzle. The latter case is the only 
 one to be considered for an air turbine, in which we always 
 have to 2 = l, and w 1 >1.9 kilogrammes. 
 
 Length and Final Section of Nozzle. 
 
 In practice the diverging portion of the nozzle is made 
 in the form of a cone of an angle of about 10 degrees, in 
 order to avoid the breaking of the vein, which cannot follow 
 the walls if a greater angle is used. The final section s 2 , 
 and hence the length, of the nozzle, will then be determined 
 by 
 
 (10) 
 
 V m 
 
 1 ,~, (11) 
 
 Since, as for air, we have a> m = 0.529^, we have 
 
 y-i 
 
 5 2 ( r\ Kc>r\ co i\ y /I (0.529) y . /ir>x 
 
 - ji = (0.529- J ^ / '-^r- (1^) 
 
 Sm ( j -y -^ 
 
 * The velocity of sound in a gas of which the absolute density is D is 
 given by the formula of Newton: 
 
 si 
 
 E, being the coefficient of elasticity of the gas, has for its value X p. We 
 then have V 
 
184 
 
 THE GAS TURBINE. 
 
 Thus, for example, we have, for 
 
 " 
 
 Sm 
 
 20; 
 
 2.91. 
 
 The diagram (Fig. 69) shows these results. It will be 
 seen that the ratio of cross-sections for a given expansion 
 is less in the case of air than for steam. The nozzle will, 
 therefore, be shorter. 
 
 10 ZO 30 40 
 
 Fio. 69. Ratio of nozzle sections for air and saturated steam. 
 
 This relation depends wholly on and not on the tem- 
 perature. 
 
 Velocity in the Neck of the Nozzle. 
 
 The velocity w m in the neck of the nozzle depends upon 
 the absolute temperature lt and not upon the relation of 
 
THE GAS TURBINE. 185 
 
 the pressures, as is shown in equation (8), in which we may 
 replace aj^^ by R6 l 
 
 r -4-[ R6 i ( 13 ) 
 
 Thus we have for: 
 
 7^ = 1000 1500 2000 2500 
 to m = 484 593 685 765. 
 
 metres per second. 
 
 Velocity of Discharge. 
 
 If, in equation (5), we note that: 
 
 i/Ji 
 
 0, 
 
 we have: 
 
 Vf 1 
 ^-i4C;) Y -'> 
 
 which demonstrates the correctness of our original result 
 based upon the principle of the conservation of energy.* 
 
 Influence of the Lower Pressure. 
 
 We have already seen that the velocity in the neck of 
 the nozzle is entirely independent of the expansion ratio, 
 and depends wholly upon the temperature of the gas before 
 the expansion. Thus, in the case of air, the pressure in the 
 neck of the nozzle is 
 
 0.529^. 
 
 Beyond the neck the expansion continues and the velocity 
 increases regularly, while at the same time the pressure falls. 
 
 * Equations (3) to (14) have been taken from Stodola's treatise on the 
 steam turbine; also figures 70 and 71. 
 
186 THE GAS TURBINE. 
 
 If the cross-section increases as the square of the distance 
 from the neck (a conical nozzle) the pressure varies accord- 
 ing to a law which we may determine by taking the value 
 of the velocity at each point (which is dependent upon the 
 section), calculating the resulting variation in kinetic energy 
 (from the neck to the point under consideration), and thence 
 obtaining the temperature and the pressure for the given 
 point, according to the law of adiabatic expansion. 
 
 If the angle of opening is given it will then be possible 
 to determine a definite pressure for the terminal section 
 as a function of the length of the nozzle. 
 
 The question arises : What will be the result if the medium 
 into which the gas is discharged from the nozzle has a differ- 
 ent pressure from that at the end of the nozzle? 
 
 The experiments of Professor Stodola upon steam have 
 shown that if the pressure is lower than that of the exhaust 
 it will produce sound waves, the pressure varying accord- 
 ing to a sinusoidal curve in the discharge chamber. 
 
 Emden has calculated for air, and Prandtl for the vapor 
 of water, the corresponding wave lengths. 
 
 The formula, of the form: 
 
 iteL 
 
 fV>m* ,\/> 
 
 "h-- 1 ) 
 
 in which c is the velocity of sound in the discharge chamber, 
 and w m the velocity of the fluid at its discharge, shows that 
 the waves can be produced only when the velocity of dis- 
 charge is greater than that of sound (Fig. 70). Professor 
 Stodola admits that the fluid leaving the nozzle expands at 
 once to the pressure of the surrounding medium, this causing 
 the transformation of an excess of the potential energy of 
 the exhaust fluid into living force. It is this excess which 
 causes the sound waves and which is transformed into heat 
 by friction and eddies. 
 
 If the pressure of the discharge chamber is greater than 
 that corresponding to the terminal section of the discharge 
 
THE GAS TURBINE. 
 
 187 
 
 nozzle a sudden shock will be produced, causing a rebound 
 in the pressure curve, followed by strong waves (Fig. 71). 
 This rebound may even force its way back into the nozzle, 
 as shown in curve D. 
 
 Absolute Pressure 
 
 vcw<wNS 
 
 20 10 10 20 30 40 50 60% 
 FIG. 70. Pressure waves in discharge nozzles. 
 
 If the counter pressure be increased until it reaches the 
 same order of magnitude as the original pressure, the re- 
 bound will penetrate further and further back into the 
 nozzle, and may reach as far as the neck. This condition 
 will be attended with a great loss of energy. 
 
 The theory of shock has been studied by Lord Rayleigh, 
 Weber, Grashof, Lorenz, Prandtl and Proell, Stodola, and 
 others. These researches, of much theoretical interest, lead 
 to the following practical conclusions: 
 
 In order to secure adiabatic expansion from p l T l to p 3 T 3 , 
 it is necessary to use a nozzle of well-defined length. 
 If too short the expansion will be incomplete, and the living 
 force produced will not utilize all the available energy. If it 
 
188 
 
 THE GAS TURBINE. 
 
 is too long, a shock will be produced, accompanied by a loss of 
 kinetic energy. The best length of nozzle can be determined 
 only by experiment. In practice the pressure may be varied 
 until the best result is produced. It will, therefore, be seen 
 that it is difficult to obtain a good efficiency with an explosion 
 turbine, in which p l and T l are continually varying, and in 
 which the nozzle will be alternately either too short or too long. 
 
 1*6 
 
 Absolute Pressure 
 
 FIG. 71. Oscillations in discharge nozzles. 
 
 It is also evident that any method of regulating a turbine 
 by varying the maximum pressure must be accompanied by 
 a loss of kinetic energy. It is possible that a compensating 
 action may be obtained by varying the temperature and the 
 maximum pressure. However this may be, the best method 
 for regulating a turbine is by the variation of the admission. 
 
 Influence of Friction. 
 
 If, according to Stodola, we designate the successive 
 states of the gas in two distinct sections of the fluid vein by 
 the indices 1 and 2, and let Q 8 be the quantity of heat dissi- 
 
_ THE GAS TURBINE. _ 189 
 
 pated by radiation and conductivity during the period in 
 which the gas is passing from one of these sections to the 
 other, we have: 
 
 which gives the formula of Zeuner: 
 
 From this, taking the particular case of adiabatic flow, 
 we have: 
 
 and since ^ = constant +CT 
 
 w 2 C 
 
 If, upon an infinitely small element of the gaseous mass we 
 consider the influence of the walls to act in such a manner as 
 to add or subtract a quantity of heat dQ, and, on the other 
 hand, there is disengaged by friction a quantity of heat dR, 
 we have: 
 
 If we thus introduce the idea of frictional resistance, we 
 find that the formula of Saint Venant thus generalized be- 
 comes, in the case of adiabatic flow (E = 0, Q 8 = 0) 
 
 As Professor Stodola has remarked, an adiabatic flow 
 with friction is not adiabatic in the same sense as a flow 
 without friction. 
 
 The work of friction is not wholly lost, for a portion is 
 transformed into heat and acts to reheat the fluid, thus in- 
 creasing the amount of heat transformed into work in the 
 course of the expansion. 
 
190 
 
 THE GAS TURBINE. 
 
 In the entropy diagram, the successive states of the fluid 
 during the flow being represented by the curve A i} A 2 , the 
 area A l A 2 A" A" represents the total work of friction, 
 (R), while the effective loss of kinetic energy is only the area 
 
 A i A 2 A" A" . 
 
 T 
 
 \ 
 
 \Ai 
 
 Bi Ti Ci 
 
 A," Az 6 
 
 Fio. 72. Entropy diagram of friction losses in nozzles. 
 
 Messrs. Delaporte, Lewicki, and Stodola have each given 
 the figures resulting from their investigations with steam, 
 showing the actual losses of kinetic energy in a diverging 
 nozzle to be from 5 to 15 per cent.* 
 
 The actual loss of kinetic energy may be expressed by 
 
 in a cylindrical tube. For a conical nozzle it is necessary 
 to proceed by integration. 
 
 * Delaporte: Nozzle of 6 to 9 mm. on 50 mm., discharging into the atmos- 
 phere, loss 5 per cent. 
 
 Lewicki: Nozzle of 6 to 7 mm. on 30 mm., ratio of pressures 6.86, loss 8 
 per cent. 
 
 Stodola: Nozzle of 12 mm. on 150 mm., losses 10 to 20 per cent. 
 
THE GAS TURBINE. 191 
 
 Classification of Gas Turbines. Efficiencies of Revolving Discs. 
 Losses in the Blades. 
 
 The turbine may be either axial or radial, and composed 
 of one or several discs. 
 
 The characteristic feature, however, in all cases, is the 
 difference in pressure before and beyond the blades. If 
 this difference is zero, that is to say, if the pressure in the 
 space comprised between the guide blades and the revolv- 
 ing wheel is the same as at the discharge of the revolving 
 buckets, we have an impulse turbine. In turbines of this class 
 the transformation of energy takes place in the guide nozzles, 
 and the revolving wheel utilizes the force thus developed. 
 
 If, on the contrary, a difference of pressure exists we 
 have to deal with a reaction turbine. 
 
 The reaction turbine, having a degree of reaction of 1 : 2, 
 operates with a velocity 40 per cent, greater than that 
 necessary for an impulse turbine. For the same tangential 
 velocity it therefore requires double the number of revolving 
 wheels. 
 
 For these reasons, and especially because the impulse 
 turbine lends itself to the use of a single expansion and does 
 not require the gases to be discharged upon the revolving 
 wheel until after they have been cooled by their expansion, 
 the reaction type has not been applied to the gas turbine. 
 
 The impulse turbine with several wheels may be con- 
 structed either with velocity stages or with pressure stages. 
 
 The advantages of using multiple discs lies in the fact 
 that the same hydraulic efficiency is obtained with one- 
 third the tangential velocity if two revolving discs are used 
 instead of one, and so on. 
 
 Unfortunately it is not yet possible to determine in ad- 
 vance the efficiency of the revolving discs of future gas tur- 
 bines. We know what the hydraulic efficiency should be, 
 neglecting the resistance of friction, but it is impossible. 
 
192 THE GAS TURBINE. 
 
 without previous experiments, to determine what the fric- 
 tion of the hot gases, more or less expanded, will be upon the 
 fixed and moving blades. 
 
 It seems as if the friction should be less than in the case 
 of the steam turbine. If so, the number of revolving discs 
 might be increased, thus increasing both the total efficiency 
 and the hydraulic efficiency, properly so called. 
 
 In steam turbines of the impulse type the present ten- 
 dency is to multiply the number of pressure stages and to re- 
 duce the number of revolving discs in each stage. It would 
 not be easy to carry this method very far with the gas tur- 
 bine since it would be necessary to have a final temperature 
 of less than 700 degrees absolute, in order to avoid having 
 too high temperatures in the upper pressure stages. This, 
 as we have already seen, would necessarily be accompanied 
 by a reduction in the thermal efficiency. 
 
 There are, therefore, in the case of the gas turbine, cer- 
 tain contradictory influences, of which the final effect upon 
 the mechanical efficiency can be determined only by experi- 
 ment. Nevertheless, everything indicates that the mechan- 
 ical efficiency of the gas turbine should be at least equal to 
 that of the steam turbine. 
 
 Friction of Discs Revolving in Air. 
 
 In existing steam turbines, the friction of the revolving 
 discs is the cause of a certain loss of energy. This loss is less 
 than was formerly believed, but it is nevertheless of impor- 
 tance. We shall see that this friction loss will be rela- 
 tively less in the gas turbine, even when the pressure of the 
 fluid is equal to that of the atmosphere. At the present 
 time we have no means of knowing the magnitude of this re- 
 sistance when the machine is operating under load. The 
 investigations of Odell, Lewicki, Stodola, and others, have 
 enabled us to determine the amount of power absorbed by 
 friction when operating without load. 
 
_ THE GAS TURBINE. _ 193 
 
 It should be noted that this power is from two to four 
 times greater for a bladed disc revolving in the atmosphere 
 than when it is enclosed in a chamber, sealed against all 
 ventilation. Under such conditions, a bladed disc absorbs 
 but little more work than a flat disc. The work of friction 
 is proportional to the cube of the number of revolutions and 
 may be represented by 
 
 in which a is a coefficient, D the diameter of the disc, u 
 the tangential velocity, and d the specific gravity of the 
 gas. All other things being equal there is absorbed in 
 saturated steam 1.3 times the amount of work that is 
 absorbed in air, a point which is favorable to the gas turbine. 
 
 Under atmospheric pressure, steam superheated to 300 
 degrees Centigrade offers the same frictional resistance as air 
 at the ordinary temperature. 
 
 Lewicki has found that at 300 degrees, and under a very 
 low pressure, the amount of work absorbed is a little less 
 than in air at atmospheric pressure. 
 
 In the case of a gas turbine, however, with an exhaust of 
 600 degrees absolute, the density of the air is only 
 
 and we conclude that the loss would probably be less than 
 that observed in the steam turbine. 
 
 There is another factor which should reduce still further 
 the relative importance of this loss. This is the relatively 
 small size of the discs of the gas turbine. We have already 
 seen that the total section required for the passage of the 
 gas, for the same amount of work, is less than in the case of 
 the condensing steam engine. 
 
 In consequence it may be possible, at least in the case of 
 very large turbines, to reduce the diameter of the revolving 
 13 
 
194 THE GAS TURBINE. 
 
 wheels. For small units the diameter is controlled by the 
 tangential velocity and by the necessary number of revolu- 
 tions, and this may not be reduced. On the contrary, the 
 higher velocity of the flow of air may lead to larger wheel 
 diameters. 
 
 Regulation of Gas Turbines. 
 
 The question of speed regulation, a matter of great in- 
 dustrial importance, does not yet appear to have been made 
 the object of any important effort. Let us see how the regula- 
 tion of a turbine with a single nozzle may be effected. There 
 are two cases to be considered, that in which the compressor 
 is not mechanically connected to the turbine wheel, and that 
 in which a direct mechanical or electrical connection exists 
 between these two elements. 
 
 In the first case: The relation 
 
 shows that the power delivered by a nozzle is not influenced 
 very strongly by a variation in the temperature before ex- 
 pansion 6 t (which is the temperature of combustion), if 
 the ratio of pressures is not varied. 
 
 In fact the kinetic energy of a unit of mass is propor- 
 tional to t but the density of the gas at discharge is propor- 
 
 tional to a-; the pressure oj 2 not varying. The velocity w 2 
 a i 
 
 being proportional to V Q l the power developed, propor- 
 tional to ^~, is proportional to -W - 1 X O lf or to yO v 
 
 If we vary the ratio of pressures without modifying O l 
 we obtain a variation of 6 2 proportional to ( 2 J 7~. From 
 this the density varies as ^-, in supposing that w 2 is fixed 
 
THE GAS TURBINE. 195 
 
 and that w^ alone varies. The flow in weight is then pro- 
 portional to 
 
 , . mm? . 
 
 and the product ^- is proportional to 
 
 If, therefore, we act solely on the temperature of combus- 
 tion, we may reduce the power in the ratio of V O v and the 
 temperature of the exhaust will be reduced when operat- 
 ing at light loads. The efficiency will not be materially 
 affected (it will be rigorously unchanged if the compression is 
 adiabatic). 
 
 If we act solely on the pressure of combustion the tempera- 
 ture of the exhaust will be increased as the power is reduced, 
 a condition which is entirely inadmissible, since it would 
 lead to higher exhaust temperatures than 700 degrees at 
 light loads. If, on the contrary, we arrange that the ex- 
 haust temperature shall not exceed 700 degrees at light 
 loads, we shall have a lower exhaust temperature under full 
 load, and the efficiency will be reduced. 
 
 This latter inconvenience may be avoided and a more 
 energetic regulation effected by causing the temperature and 
 the pressure to vary at the same time, in such a manner as to 
 keep the final temperature 2 constant, although this method 
 will lower the thermal efficiency at light loads. 
 
 Under these conditions the density does not vary; the 
 velocity varies as 
 
 iu 
 
196 THE GAS TURBINE. 
 
 and the energy developed varies as 
 
 If the reduction in pressure is produced by a valve 
 placed between the combustion chamber and the compressor, 
 the efficiency will be reduced to an inadmissible extent. 
 In such a case it will be advisable to use an independent 
 compressor, with variable compression. 
 
 The variations in flow resulting from variations of 6 l or 
 
 of - or from both of these elements at the same time, will 
 
 cause the nozzles to operate under different conditions from 
 those for which they were designed, when the machine is 
 under light loads. This will reduce their hydraulic efficiency ; 
 on the contrary, the efficiency of the revolving discs will be 
 increased. Another method applicable to the single nozzle 
 is based on the hit-or-miss system. So, far as efficiency is 
 concerned, this is probably the best method, since the vol- 
 ume of the combustion chamber is so small compared to the 
 flow of gas that the transition periods of low hydraulic 
 efficiency would be of negligible duration. On the other 
 hand, some method of re-ignition would be required, unless 
 the compression and temperature were sufficient to insure 
 automatic ignition. 
 
 If the compressor is mechanically or electrically connected 
 to the turbine, it may be included in the regulation system 
 by causing the excess production to be accumulated in a 
 reservoir during the periods when the demand for the tur- 
 bine is below normal. 
 
 Let us now consider turbines with multiple nozzles. In 
 such cases it will be necessary to determine, according 
 to the conditions of each particular machine, whether it is 
 better to modify the discharge of a single nozzle or to act 
 proportionally upon them all. 
 
THE GAS TURBINE. 197 
 
 The true solution appears to be that employed in the 
 Curtis steam turbine, by shutting off a certain number of 
 nozzles without affecting the others. This method should 
 be accompanied by the use of a fly-wheel (or accumulator 
 of energy) if great regularity of speed is required, or if there 
 are not many nozzles. 
 
 It will be necessary to control the gas and combustible at 
 each nozzle by two valves. The regulator will act, through 
 an auxiliary motor, directly upon the air valves, and upon 
 the valves controlling the combustible by means of a small 
 piston, acting only when the pressure in the combustion 
 chamber reaches a certain value; its movement being, to a 
 certain extent, proportional to this pressure. 
 
 However all these points may be, it is evident that the 
 question of the regulation of the gas turbine should not 
 present any serious difficulties. 
 
 Details of Qas=Turbine Construction. 
 Air Compressors. 
 
 We have seen that it is necessary to have as high a ratio 
 of pressures as possible in order to attain a high efficiency, 
 both thermal and mechanical. 
 
 It appears that the gases which are delivered into the 
 combustion chamber should, therefore, be compressed to 
 about 40 atmospheres ; or else that a lower pressure, say about 
 10 atmospheres, be used in connection with the maintenance 
 of a pressure of about J atmosphere in the exhaust space. 
 In the second case it is evident that we should have to use a 
 second compressor to raise the pressure of the exhaust gases 
 from J atmosphere up to atmospheric pressure. 
 
 In order that the work required for this operation should 
 not exceed that required to produce the higher initial pres- 
 sure of 40 atmospheres it is necessary that the absolute tem- 
 perature of the exhaust gases should be kept down to a 
 minimum value T Q which corresponds to the temperature 
 
198 THE GAS TURBINE. 
 
 of the atmosphere. This is a difficult matter, and can be 
 effected only by using regenerators or tubular coolers of 
 large size; or by using injections of water, involving the 
 employment of a wet air pump, or possibly a barometric 
 condenser and dry air pump. 
 
 With these reservations the two systems may be con- 
 sidered as equivalent, but that which involves the employ- 
 ment of an air pump is principally of interest so far as it 
 bears upon the practicability of replacing reciprocating com- 
 pressors with turbine air compressors. 
 
 In practice it appears to be difficult to attain higher de- 
 grees of compression than 20 to 25 atmospheres with the 
 turbine compressor. To produce such pressures it is neces- 
 sary to have a large number of revolving wheels, rotating at 
 high velocities; and as these must be in a fairly dense me- 
 dium the mechanical losses resulting from the friction of the 
 wheels in the compressed air must have an injurious influ- 
 ence upon the efficiency of the machine. 
 
 If it is desired to avoid the use of piston compressors 
 it appears to be necessary to subdivide the operation in the 
 manner indicated hereafter. If, on the contrary, com- 
 pressors of the reciprocating type are permissible, there is 
 no difficulty in attaining pressures as high as 50 atmospheres. 
 These machines, as we shall see, have an excellent efficiency, 
 which is a most important feature in the present instance. 
 At the same time they add greatly to the bulk of the appara- 
 tus, and take from the gas turbine many of the advantages 
 possessed by the steam turbine. 
 
 In cases in which the bulk of a large compressor is a 
 matter of importance it may be found desirable to have the 
 reciprocatory compressor preceded by a turbine compressor, 
 or by an ordinary rotary compressor. 
 
 For example, we may use a turbine compressor to com- 
 press the air to a pressure of 4 atmospheres, which involves 
 an amount of work equivalent to only 30 calories per kilo- 
 
THE GAS TURBINE. 199 
 
 gramme, and then complete the compression by the use of a 
 piston compressor, which would raise the pressure from 4 to 
 40 kilogrammes per square centimetre (4 to 40 atmospheres) 
 with an expenditure equal to 46 calories. 
 
 The mechanical efficiency of the turbine compressor being 
 assumed to be equal to 0.70 and that of the reciprocating 
 compressor to 0.85, the combination would have an effi- 
 ciency of 0.59, but the bulk of the piston compressor would 
 be reduced to one-fourth that otherwise necessary, since the 
 volume of air to be handled is reduced to one-fourth. 
 We then have as available the following three plans : 
 A piston compressor, compressing the air directly to 40 
 atmospheres; a turbine compressor, compressing the air to 
 4 atmospheres, followed by a piston compressor raising it to 
 40 atmospheres; or a turbine compressor, compressing 
 the air to 10 atmospheres, and a cooler beyond the gas tur- 
 bine, followed by a turbine air pump taking the burned gases 
 at J atmosphere and discharging them into the atmosphere. 
 
 Ordinary Piston Compressors. 
 
 An excellent study of machines of this type was given 
 by M. Barbet in the Genie Civil from May to August, 1895. 
 We may here note at once that compressors with liquid 
 pistons, such as those of Sommellier, Dubois, and Frangois, 
 and others, are too bulky to be considered in connection with 
 the gas turbine, and we are, therefore, obliged to resort to 
 compressors of the Colladon type, with simple injection of 
 water; or of the Mekarski type, using several stages of 
 compressors, with intercoolers. 
 
 M. Mekarski has constructed a number of vertical com- 
 pressors, operating at 100 to 150 revolutions and compressing 
 air to about 60 atmospheres, taking 0.065 kilogramme of 
 air per revolution. These machines have four single-acting 
 cylinders, arranged in tandem pairs, having automatic 
 valves, and provided with water injection in the low pres- 
 
200 THE GAS TURBINE. 
 
 sure cjdinders, the high-pressure cylinders being jacketed. 
 The volume efficiency was about 67 per cent. The effi- 
 ciency in work may be computed as follows : The indicated 
 power of the engine operating the compressor was 97 horse- 
 power, and there was compressed 390 kilogrammes of air 
 per hour. 
 
 Theoretically, an isothermal compression to 60 atmos- 
 pheres would require: 
 
 42,308X390 
 
 270,000 
 
 61.1 h.p. 
 
 The efficiency of the combined air compressor and steam 
 
 AT i 
 
 engine, was, therefore, -7^-= 0.63. 
 
 y / 
 
 If we assume the efficiencies of the compressor and the 
 steam engine to be the same, their value would be 1/0.63 = 
 0.79. This shows that even with a piston compressor of 
 such a moderate size as 100 horse-power a mechanical 
 efficiency of nearly 0.8 may be obtained. Under these condi- 
 tions one horse-power delivered to the shaft of the compressor 
 would furnish 5 kilogrammes of air compressed to 60 atmos- 
 pheres. 
 
 Taking compressors of large size we may cite the vertical 
 triple expansion machines of 2000 horse-power, built by the 
 Creusot Works for the Paris Compressed Air Company. 
 The air, in these machines, is raised to 7 atmospheres, the 
 ratio thus being 1 : 7. The air cylinders are double acting, 
 and are arranged tandem with the steam cylinders. The 
 compression is effected in two stages, with a pressure of 
 2.7 kilogrammes per square centimetre (38.4 pounds per 
 square inch) in the intermediate reservoir. The air cylin- 
 ders have no water jackets, but water injection is used in the 
 cylinders, in the valve chests, and in the intermediate reser- 
 voir. 
 
 Operating at 50 revolutions per minute, these machines 
 handle 20,720 cubic metres of air, or 25,400 kilogrammes per 
 hour, with a volumetric efficiency of about 0.80. 
 
THE GAS TURBINE. 201 
 
 According to tests conducted by M. Bourdon and by 
 Professor Gutermuth, the expenditure of one horse-power 
 indicated on the pistons of the steam engine compresses 
 12.7 kilogrammes of air to the ratio of 1:7. Other tests give 
 a combined efficiency for the engines and compressors of 
 about 0.90. 
 
 In these machines of the Paris Compressed Air Company 
 the curve of compression corresponds closely to the equation 
 
 J p^ 1 ' 3 = constant 
 
 and the air enters at a temperature of -f 5 C. and leaves at 
 + 25, while an adiabatic compression would give a final 
 temperature of +200. The cooling by water injection thus 
 appears to be very effective. 
 
 As a final example of a compressor operating at slow 
 speed we may cite a Strnad machine, of which tests by 
 Professor Gutermuth upon one of 300 horse-power gave the 
 following results : 
 
 Ratio of compression 1 to 7.1 
 
 Number of revolutions 50 to 70 
 
 _, . theoretical work of compression 
 
 Ratio: . ,. c : = 0.73 
 
 indicated power of engine 
 
 ~ t . indicated work of compression 
 
 Ratio: r r -5 : - . = 0.84. 
 
 indicated work of engine 
 
 It follows that the mechanical efficiency of the compres- 
 sor, assuming it equal to that of the engine would be 1/0.84 
 = 0.92, but since the elevation of temperature has the effect 
 of raising the value of the ratio 
 
 theoretical work of compression 
 indicated work of compression 
 
 to a value 0.87, we have in reality an efficiency 0.87x0.92 
 = 0.80. The Strnad compressor is a two-stage machine with 
 two cylinders in tandem with the steam cylinders of a com- 
 pound engine. It is characterized by an injection of water 
 
202 THE GAS TURBINE. 
 
 under pressure in the cylinders, and by the use of Corliss 
 valves. The consumption of water is 2.55 litres per cubic 
 metre of air. With this machine one horse-power com- 
 presses 12 kilogrammes of air to a ratio of 1:7. 
 
 The air compressors at the Billancourt Works of the 
 General Omnibus Company of Paris have an indicated 
 power of 875 horse-power, and compress 3420 kilogrammes 
 of air per minute to a pressure of 80 kilogrammes per square 
 centimetre (1137.6 pounds per square inch), making 32 
 revolutions per minute. The compression is effected in 
 three stages: 4 kg., 24.5 kg., and 80 kg. 
 
 There are injected two kilogrammes of water for every 
 kilogramme of air in the low-pressure cylinders. With the 
 external air at 18 C. (64.4 F.) the temperature of the com- 
 pressed air is 60 C. (140 F.). 
 
 We have: 
 
 = 0.53 to 0.55 
 
 ^i of the steam engines 
 Whence the efficiency is 0.73 to 0.74. 
 
 We see, therefore, that the older, slow-running com- 
 pressors which we have examined give values which may 
 be taken as : 
 
 0.70 to 0.80 for machines of 100 to 1000 h.p. 
 0.80 to 0.90 for units of 2000 h.p. 
 
 Modern High=Speed Compressors. 
 
 During the past few years the necessity for providing 
 machines adapted for direct connection to electric motors 
 has led designers to produce high-speed air compressors, 
 capable of being operated at speeds of 500 to 600 revolu- 
 tions per minute for sizes of 100 horse-power. 
 
 At the Diisseldorf exposition there was shown by Messrs. 
 Pokorny and Wittekind, of Frankfort, a compound tandem 
 compressor with Koster valve gear, driven by an electric 
 motor of 70 horse-power, running at 550 revolutions. 
 
THE GAS TURBINE. 203 
 
 The stroke of pistons was 150 mm. and the diameters of 
 cylinders 300 and 190 mm. respectively. This machine 
 compressed about 750 kilogrammes of air per hour to a pres- 
 sure of 7 atmospheres. Data concerning tests upon modern 
 high speed compressors will be found in the papers of 
 Richter, Lebrecht, Biel, etc.* 
 
 Machines of this type offer advantages so far as reduc- 
 tion in dimensions are concerned, especially if preceded by a 
 turbine compressor. If the latter machine is used to com- 
 press the air to a pressure of 4 atmospheres absolute, the 
 small machine described above would be available to com- 
 press 3000 kilogrammes of air from 4 to 28 atmospheres. 
 This would be sufficient for a gas turbine of 600 horse- 
 power. 
 
 Rotary Compressors. 
 
 Rotary compressors have not, up to the present time, 
 been very successful. They are, however, used to some ex- 
 tent for moderate and small powers. 
 
 The Compagnie Generate Electrique of Nancy has devel- 
 oped the Hult rotary steam engine, the principle of which is 
 applicable to the compression of air. 
 
 Messrs. Siemens and Halske have produced a rotary 
 compressor, adapted only to small sizes, and, being without 
 clearance space and capable of direct connection to electric 
 motors, may be used for compressions of a ratio 1:3, or to 
 produce a vacuum as low as 1.5 mm. of mercury. It would 
 be interesting to know the efficiency of these machines, as 
 they might be applicable for use with an initial compression 
 of 4 atmospheres and an exhaust pressure of T V atmosphere, 
 to give an expansion ratio of 1 : 40. These machines would 
 be adapted only to small gas turbines. 
 
 * Richter: Thermische Untersuchung an Kompressoren, Zeitschr. d Ver. 
 deutscher Ing., July, 1905. Lebrecht: Versuche mit raschlaufende gas-com- 
 pressoren, Zeitschr. d. Ver. deutscher Ing., 1905. Biel: Zeitschr. d. Ver. deut- 
 scher Ing., 1905, p. 540. 
 
204 THE GAS TURBINE. 
 
 Turbine Compressors. 
 
 In 1902 M. Rateau directed attention to the properties 
 of high-speed blowers, showing such a machine connected 
 directly to a steam turbine making 10,000 to 20,000 revolu- 
 tions per minute. This blower, with an efficiency of about 
 0.60, gave a compression ratio of 1 to 1.5. By coupling a 
 series of such blowers the pressure may be increased in 
 a geometric proportion, giving with four such blowers a 
 pressure of (1.5) 4 , or about 5 atmospheres absolute. 
 
 Theoretically, the efficiency of such an arrangement is 
 equal to the efficiency of a single wheel. Nevertheless, 
 it does not appear to be practicable to use this machine 
 for pressures much higher than 5 atmospheres, since the 
 frictional resistance in atmospheres of higher densities will 
 increase to an extent which causes the efficiency to fall off 
 materially. 
 
 If a machine of this type is used to lower the exhaust 
 pressure of a turbine, the above inconvenience does not 
 appear, and a greater number of wheels will not be required 
 to produce the desired reduction in pressure. At the same 
 time the apparatus will have to be much larger for the same 
 discharge of gases by weight. 
 
 For example, if it is desired to produce a pressure of i 
 atmosphere, the first wheel will produce a reduction of only 
 -- = 0.1 atmosphere. 
 
 It follows that n wheels will give a final pressure equal 
 to 0.2 X (1.5) n , and we should have 0.2 X (1.5) n = l, whence 
 (1.5) n = 5, and n = 4. It is understood that this assumes 
 that the exhaust gases have been cooled to the temperature 
 of the air before entering the apparatus. 
 
 If this arrangement is employed it is very desirable that 
 the multicellular blower should be connected directly to the 
 gas turbine; but since the tangential velocity of the blower 
 should reach 260 metres per second to obtain the pressure 
 
THE GAS TURBINE. 205 
 
 ratios given above, a high rotative speed becomes necessary 
 (20,000 revolutions in a machine of 200 horse-power com- 
 pressing air from 1 to 5 atmospheres). 
 
 If the gas turbine is designed for driving dynamos, 
 etc., it should make not more than 1000 to 2000 revolutions. 
 In such cases it would probably be advisable to have a sep- 
 arate turbine to drive the blowers. 
 
 Besides the investigations of M. Rateau several other 
 attempts have been made to utilize the turbine compressor. 
 
 In England experiments have been made with the 
 Parsons turbine, while similar attempts have been made by 
 the General Electric Company to utilize the Curtis turbine. 
 No practical results, however, have yet been made public. 
 
 The idea of Burdin and Tournaire has, therefore, not yet 
 entered into practical operation, but it appears to be on the 
 eve of successful application. 
 
 Thermal Regenerators. 
 
 The application of alternating regenerators, necessary 
 for use with piston engines, does not appear to have given 
 practical results. The question appears in a different 
 light, however, when considered in connection with the gas 
 turbine. 
 
 In the latter case the gases issue from the turbine in a 
 continuous manner, with a fairly high velocity, which is a 
 great advantage in connection with the transmission of heat. 
 
 The gases to be heated are also delivered in a continuous 
 manner by the compressor, and in view of the high pres- 
 sures under consideration the losses involved in producing 
 a rapid circulation in a system of tubes need not be very 
 serious. 
 
 The burned gases discharged by a gas turbine would be 
 much cleaner than those dealt with in the case of super- 
 heaters attached to steam boilers, or than those discharged 
 from reciprocating gas engines; the continuous combustion 
 
206 THE GAS TURBINE. 
 
 under pressure being that which leaves a minimum amount 
 of residue. We have also seen that these gases have a very 
 slight oxidizing action. These conditions are all favorable 
 to the use of the regenerator. 
 
 Regeneration from Gas to Gas. 
 
 Devices intended to heat the compressed air by means of 
 the heat abstracted from the exhaust gases, are analogous to 
 steam superheaters and are operated under similar temper- 
 ature conditions. They are less subject to oxidation, the 
 gases being poor in oxygen, and the temperatures are more 
 uniform. 
 
 The nature of the metal used is immaterial, so far as the 
 transmission of heat from gas to gas is concerned. Tubes 
 either of iron or steel may be employed. Copper or alumi- 
 num tubes are of interest only as regards increased durabil- 
 ity. High circulation velocities are essential, and the coun- 
 ter-current principle should be adopted, taking the necessary 
 structural precautions against injury from expansion and 
 contraction. 
 
 Regeneration by Steam. 
 
 When the regeneration of the waste heat is effected 
 by the aid of steam, the latter is generated in a boiler of the 
 instantaneous flash type, heated by the exhaust gases. 
 Such a boiler includes a reheater, followed by the boiler 
 proper, and by a superheater, all with a systematic circula- 
 tion. These three portions, however, are not necessarily 
 clearly defined. We may use, for example, a boiler of the 
 Serpollet type, or one of the system proposed by Col. Renard.* 
 
 The water, entering at one end, travels continuously 
 through until it reaches the other extremity in the state 
 of superheated steam. Here, as before, the gases should 
 circulate very rapidly. 
 
 * See Ge"nie Civil, 1905. 
 
THE GAS TURBINE. 207 
 
 Production of the Combustible Mixture. 
 
 The combustible employed may be solid, liquid, or 
 gaseous, but it is improbable that a turbine employing a 
 solid combustible will soon be realized. It would un- 
 doubtedly be very difficult to inject a solid combustible into 
 a combustion chamber against a pressure of 30 to 40 atmos- 
 pheres, and the ashes discharged with the burned gases 
 would cause excessive wear upon the blades of the turbine. 
 
 The gas turbine is, therefore, not adapted for the direct 
 use of solid combustibles, and, for a long time, at least, 
 the use of a gas producer must be considered necessary. 
 
 Gaseous Fuel. 
 
 As we have already seen, the lean gases are almost as 
 advantageous as the richer gases. In the case of Dowson 
 gas the air compressor must deliver a volume of air equal 
 to 1.2 times that of the gas. This involves the use of two 
 compressors of different power and capacity. 
 
 It may be suggested that the Gardie system is available 
 for this purpose, in view of the fact that it utilizes the sen- 
 sible heat of the gases which leaves the producer at about 
 700 degrees and which is ordinarily lost; but the operation of 
 a gas producer under a pressure of more than 10 atmospheres 
 appears to be a difficult problem. It would thus be neces- 
 sary to operate the producer under a moderate pressure, a 
 method practicable only when the turbine is operated with a 
 low-pressure exhaust. The principal advantage of this sys- 
 tem appears to lie in the avoidance of any wasting of the gas. 
 
 If we use the waste gases from blast furnaces it will also 
 be necessary to have two compressors of different dimen- 
 sions, the volumes being in the proportion of 0.7 to 0.9 of air 
 to 1 of gas. 
 
 It would be difficult to use a regenerator in this case, for 
 it is well known that the large amount of dust contained in 
 
208 THE GAS TURBINE. 
 
 furnace gases renders their use difficult under steam boilers. 
 This difficulty, however, is not impossible of removal, and 
 the use of the regenerator would enable improved efficiency 
 to be secured. 
 
 With very rich gases, such as illuminating gas or acety- 
 lene, a very small compressor would be required for the gas. 
 The relations of the volumes to be compressed would be, 
 respectively, as 1 to 8 and as 1 to 20. The total value of 
 the work of compression, however, is not much less than in 
 the case of lean gases. 
 
 With liquid combustibles a very small pump, consuming 
 but little power, is required, but the air compressor demands 
 about as much power as the two pumps together in the case 
 'of gaseous fuel. 
 
 Liquid combustibles are used, either by volatilization 
 in a carburetor, similar to those employed for gasoline 
 motors, or by direct injection into the combustion chamber 
 in connection with some sort of pulverizer or atomizer. It 
 may also be found advantageous to heat or volatilize the 
 liquid fuel by means of the heat of the exhaust gases. 
 
 In turbines using an injection of water or steam, the 
 injection may be made either in the combustion chamber or 
 before entering it. We are inclined to favor the latter 
 solution, especially in the case of steam. 
 
 The Combustion Chamber. 
 
 The dimensions of the combustion chamber should be 
 such that the reaction may be completed before the gases 
 leaving the zone of combustion penetrate the nozzle. 
 
 According as the maximum temperature T l varies from 
 1000 degrees to 2500 degrees absolute, the velocity in the 
 throat of the nozzle varies from 500 to 800 metres per 
 second. If we take a temperature of 2000 C. absolute, 
 we have a velocity of 685 metres per second. For the 
 terminal part of the section of the nozzle we have, for a 
 
THE GAS TURBINE. 209 
 
 compression ratio of 25 to 30, and a temperature of 700 at 
 the end of the expansion, a velocity of 1500 metres per 
 second. What, then, should be the length of the combus- 
 tion chamber, under these conditions? Taking a tempera- 
 ture of 2000 absolute, and a pressure of 25 atmospheres, the 
 
 density will be: (25 to 30) X 2000 = 8 ' 75 to 10 ' 5 times that 
 of the exhaust gases; let us say 10 times. If, then, the 
 section of the combustion chamber is equal to that of the 
 lower terminal section of the nozzle we should have a 
 velocity one-tenth as great, or about 150 metres per 
 second. 
 
 It is generally admitted that the velocity of propagation 
 of flame, at atmospheric pressure, is only about 1 to 2 metres 
 per second. It would therefore be necessary to give the 
 combustion chamber a section of about 100 times that of the 
 end of the nozzle, in order that the combustion may be 
 absolutely complete. The best dimensions can be deter- 
 mined only by experience. It is hardly probable that such 
 a size as indicated above would be necessary, for with the 
 temperatures and pressures under consideration the com- 
 bustible would ignite spontaneously, and hence the velocity 
 of propagation of the flame would become almost infinite. 
 
 In practice it seems as if a cross-section 10 times as great 
 as that of the terminal section of the nozzle ought to be 
 sufficient. This .would give a velocity in the chamber of 
 about iV that of the discharge, or about 38 metres per second 
 This velocity would be considered normal in ordinary gas 
 mains, operating under atmospheric pressure. 
 
 The length of the combustion chamber may be fixed 
 at 5 to 10 times its diameter, but there is no special rule 
 governing this dimension. 
 
 For example, suppose we take a nozzle having a terminal 
 section equivalent to that of a circle 10 mm. in diameter, 
 and make the diameter of the combustion chamber 50 mm., 
 
 14 
 
210 THE GAS TURBINE. 
 
 we get a velocity in the latter of: 
 
 38 m. per sec. 
 
 % - =15.2 m. per sec. 
 Z.o 
 
 If we give the chamber a length of 300 mm., the gas will 
 be in it for about: 
 
 0.30 m. 
 
 = 0.02 second. 
 
 15.2 m. per sec. 
 
 This corresponds to the duration of the combustion in a 
 gasoline motor making 1500 revolutions per minute. 
 
 Nozzles for Injection and Expansion. 
 
 In the case under consideration there are many reasons in 
 favor of partial injection. 
 
 The injection nozzles, or ring of fixed blades, being sub- 
 jected to the high temperature of combustion, it is necessary 
 in their construction to provide a material possessing special 
 properties, including resistance to heat and to chemical reac- 
 tions, also to repeated expansion and contraction, and finally 
 possessing considerable mechanical strength. 
 
 These conditions are well filled by the material carborun- 
 dum, with the exception of the fact that precautions 
 must be taken with respect to its rather high conductivity 
 for heat. 
 
 The Combustion Chamber and the Nozzles for the Characteristic 
 Portion of the Gas Turbine. 
 
 The simplest arrangement, applicable to small turbines, 
 consists of the combination of a combustion chamber and 
 single diverging nozzle, the latter preferably of rectangular 
 cross-section. The terminal portion of the nozzle may be 
 made of metal, since the gases at this point will be cooled 
 to below 800. The end of the nozzle may be divided by 
 thin partitions, thus forming a portion of the distributing 
 ring. 
 
THE GAS TURBINE. 
 
 211 
 
 Construction of the Revolving Wheels. 
 
 The construction of the revolving wheels of the gas 
 turbine does not differ materially from that already fol- 
 lowed for the steam turbine. The arrangement of a single 
 disc may be adopted or that of multiple discs; the latter is 
 preferable, since the velocity of the fluid will be at least as 
 high as in the steam turbine. 
 
 The practical design depends upon the linear speed, 
 and the moderate velocity of 120 to 200 metres per second at 
 the perimeter of the wheels may be maintained. 
 
 It must be considered that the temperature of 300 to 
 400 Centigrade is somewhat higher than ordinarily exists 
 in the steam turbine, and it is desirable to select materials of 
 which the strength is reduced as little as possible at these 
 temperatures. No serious difficulties need be apprehended 
 on this score. Steam turbines of the Laval type have been 
 operated with steam superheated to 600 absolute and with 
 air heated to 700 absolute, without any difficulty.* 
 
 Tests upon nickel steel have given the following results: 
 
 Absolute temperature. 
 
 300 
 
 500 
 
 600 
 
 700 
 
 Breaking load (kg. per cm. 2 ) 
 
 81 
 
 91 
 
 92 
 
 73 
 
 Elastic limit (kg per cm 2 ) . . . . 
 
 70 
 
 60 
 
 54 
 
 40 
 
 Elongation per cent 
 
 107 
 
 87 
 
 83 
 
 70 
 
 Reduction of section, per cent 
 
 608 
 
 60 
 
 608 
 
 70 
 
 
 
 
 
 
 We may expect that the light metal blades, when sub- 
 jected to the oxidizing action of the gases at these tempera- 
 tures will wear somewhat more rapidly than the blades of 
 the steam turbine. In the latter machine, however, it is 
 admitted that the principal cause of wear is due to the action 
 of particles of moisture. In the case of superheated steam, 
 and consequently with the gases of the gas turbine, the wear 
 
 * See Lewicki: Zeitschr. d. Ver. deutscher Ing., 1905. 
 
212 
 
 THE GAS TURBINE. 
 
 of the blades is negligible. The burned gases contain but a 
 small proportion of free oxygen (5 to 10 per cent.), and 
 nickel steel is especially resistant to oxidizing influence. 
 
 (C) Cycle with regeneration 
 of heat 71-0.75 
 
 (A)Cyc!e without regeneration 
 of heat 
 
 FIG. 73. Diagrams showing the distribution of heat from the combustion of 1 kilo- 
 gramme of gas. Water-injection not included.. Isothermal compression; isobaric combus- 
 tion at a pressure of 40 atmospheres. 
 
 General Design of a Gas Turbine. 
 
 Let us How apply the preceding theoretical calculations 
 to three practical examples. 
 
 For this purpose we will select a system employing iso- 
 thermal compression, with combustion under constant pres- 
 sure; the turbine being intended to develop 150 horse-power, 
 and thus capable of driving a dynamo of 100 kilowatts. 
 
THE GAS TURBINE. 
 
 213 
 
 Ulip p ^S P P 
 
 CO 
 
 cn en 
 
 p. 
 3 B r 
 
214 THE GAS TURBINE. 
 
 The pressure ratio will be taken at 40, and this may be 
 attained either with the exhaust at atmospheric pressure or 
 with exhaust at | atmosphere; in this latter case the pressure 
 of combustion will be 9 atmospheres above perfect vacuum. 
 
 We will consider the following cases : 
 
 A. Illuminating oil with a lower calorific value of 10,000 
 
 calories, without regeneration. 
 
 B. Same data, with steam regeneration, with /*=0.75. 
 
 C. Same data, with regeneration of gas to gas, /*=0.75. 
 The results of the computations are given in the preceding 
 
 table. The figures must be understood as having only a 
 relative value, owing to the numerous hypotheses which 
 have been made in deriving them. Their comparative 
 value, however, remains unaffected. We see, for example, 
 that the injection of steam permits the reduction in capacity 
 and power of the compressor by about one-third, but that 
 this increases the velocity of discharge about 15 per cent, 
 to the detriment of the mechanical efficiency of the machine. 
 The diagrams on page 212 will be of interest, as showing 
 the distribution of heat and work. 
 
 Experimental Researches Necessary to Determine Precise Data 
 for Gas Turbine Calculations. 
 
 We are now led to consider what experimental data are 
 to be determined in order that we may substitute more or 
 less precise computations for the hypothetical ones which 
 have thus far been employed in our investigations. 
 
 Above all things it is desirable that a study should be 
 made of the work absorbed in the compression and of the 
 final temperature of the compressed gases. There appears to 
 be no experimental difficulty in making these determina- 
 tions. It is especially desirable that these studies should be 
 made with rotary compressors, since the subject of piston 
 compressors has already been well investigated. 
 
THE GAS TURBINE. 215 
 
 Further, it is desirable that the phenomena of combustion 
 should be studied, especially to determine: 
 
 1. The limits to the amount of combustible and air 
 which can be used when attaining a perfect combustion at 
 different pressures. 
 
 2. The real temperatures of combustion thus attained. 
 
 3. The conditions of temperature and pressure which 
 will insure the self-ignition of the mixture. 
 
 4. The time required to enable a perfect combustion to 
 be realized, and thus the volume of the combustion chamber 
 to be determined. 
 
 5. The influence of these various elements upon the veloc- 
 ity of the gaseous current. 
 
 6. The influence of a greater or less quantity of vapor 
 of water in the mixture. This should include the extent to 
 which any portion of the water is dissociated, and under 
 what conditions. 
 
 It is not difficult to see what methods might be employed 
 to carry out such a programme. The present electrical 
 appliances for determining high temperatures are already 
 available for such a purpose. 
 
 It would then be desirable to investigate the subject of 
 expansion. It is important to determine the extent to 
 which the temperature is lowered by expansion. This is the 
 most delicate part of all the investigations. 
 
 While it is comparatively easy to measure the tem- 
 perature before expansion and the pressures before and 
 beyond the nozzle, it is much more difficult to determine 
 the true temperature of a gas leaving a nozzle at a high 
 velocity. 
 
 The thermometric element placed in the midst of the 
 gaseous current which is flowing at a velocity of 1200 to 1800 
 metres per second, takes a temperature higher than that of 
 the gas, for the heat disengaged by friction is not dissipated 
 by radiation. In every case it is necessary that the thermo- 
 metric element should have a higher temperature than that 
 
216 THE GAS TURBINE. 
 
 of the surrounding medium in order that the heat thus dis- 
 engaged may be removed and a stable condition secured. 
 
 Since we are ignorant of the value of the velocity (this 
 involving a knowledge of the temperature of the gas), it is 
 hardly practicable to be able to make a correction. It is, 
 therefore, necessary to compute the temperature or the veloc- 
 ity by a series of successive approximations. 
 
 It seems preferable to leave aside any attempts to meas- 
 ure the final temperature of expansion, and to determine 
 rather the kinetic energy of the gases leaving the nozzle. 
 This may be done by measuring directly the push of the 
 gaseous current upon a turbine blade, as has been done by 
 Delaporte, Rateau, Stodola, and others. 
 
 Having thus determined the velocity of discharge w 21 
 we may deduce the temperature of the discharging gases 
 
 from the formula: 
 
 ?/> 2 
 
 We can then deduce a value of ? which takes into account 
 the friction in the jet. Experiments of this kind will enable 
 the best form of nozzle to be determined, as has already been 
 done with the steam turbine, and especially to permit 
 the question whether the form of constant acceleration pro- 
 posed by Proell is preferable to the ordinary conical form. 
 
 Finally it is desirable to make an experimental determi- 
 nation of the value of the coefficient of transmission of heat 
 from one gas to another through the walls of an assemblage 
 of tubes, in order to aid in determining the dimensions of 
 heat regenerators. 
 
 The Future of the Gas Turbine. 
 
 Having now discussed the question of the construction 
 of the gas turbine we may take up the subject of the future 
 in store for machines of this kind. 
 
 Their great theoretical interest has been apparent to all 
 those who have examined these questions since the period 
 
THE GAS TURBINE. 217 
 
 when the success of the turbine of Laval demonstrated the 
 value of the pressure type of steam turbine. This cele- 
 brated inventor follows the thought of Burdin and Tournaire, 
 and suggested very early the idea of constructing a gas tur- 
 bine. Many years have now passed, however, without the 
 practical realization of this idea. 
 
 Other investigators have taken up the same idea, 
 but thus far their efforts' have not reached commercial suc- 
 cess, while during the same period the steam turbine has 
 emerged from the experimental workshop and acquired its 
 well-known position among heat engines. 
 
 This should offer no reason for surprise, when we con- 
 sider the multiplicity of technical difficulties which present 
 themselves in the realization of a practical gas turbine. 
 The success of the steam turbine, however, has elicited 
 investigations of the greatest interest which lead us to 
 approach the construction of a gas turbine without hesita- 
 tion. Some investigations are yet required to enable the 
 determination of the conditions of combustion and the exact 
 laws governing the expansion. When these have been com- 
 pleted we will be in possession of all the data necessary for 
 the turbine itself without guesswork. 
 
 Rotary compressors, multicellular blowers, turbine com- 
 pressors of the Parsons, Curtis, and other types, are rela- 
 tively further from a definite, practical solution, but every- 
 thing leads us to believe that no material delay will occur in 
 this direction. 
 
 We may thus expect to see commercially produced, a 
 gas turbine, uniting in a certain degree the advantages of the 
 gas engine and the steam turbine. 
 
 Without overlooking the inconvenience resulting from 
 the presence of a compressor distinct from the motor itself, 
 the gravity of this objection may be exaggerated. 
 
 If we are willing to accept the piston compressor (or use 
 the alternative of the reduction of exhaust pressure below 
 atmosphere) the gas turbine presents the same advantages 
 
218 THE GAS TURBINE. 
 
 of moderate bulk and weight which have made the success of 
 the steam turbine. 
 
 The thermal efficiency of the new machine will be supe- 
 rior to that of the gas engine, but the lower mechanical 
 efficiency of the gas turbine will reduce the total useful 
 effect to about the same order as that of the Diesel motor; 
 while motors using blast furnace gases should give an effec- 
 tive horse-power with an expenditure of 2000 calories. 
 
 It does not appear that any sensational invention 
 can modify these results materially in the future. It is only 
 by continual improvements in structural details that the 
 mechanical efficiency may be increased by the reduction of 
 mechanical losses. 
 
 The gas turbine will not be a universal panacea, neither 
 will it dethrone the steam turbine. When we have to deal 
 with the combustion of ordinary coal, nothing can surpass 
 the steam boiler. 
 
 But for other combustibles, petrol, various hydrocarbons, 
 alcohol, producer gas, furnace gases, etc., direct combustion 
 is advantageous. It permits the avoidance of many important 
 losses, and removes many operative objections and dangers. 
 
 The utilization of blast-furnace gases, coke-oven gases, 
 etc., presents in itself an important field for the gas turbine, 
 which may well replace the bulky engines now in use. 
 
 The gas turbine also appears to be as well adapted to the 
 driving of dynamos and alternators as is the steam turbine. 
 The same is true as regards the propulsion of ships. 
 
 It is also possible that the development of the gas tur- 
 bine will permit the realization of motors of excessively light 
 weight for use in aerial navigation. 
 
 We may thus predict for the gas turbine an extensive 
 field of application, and it is altogether possible that 
 practical experience will enable many special advantages 
 to be developed, as so often has been the case in connection 
 with the appearance of new and improved appliances. 
 
CHAPTER IV. 
 
 THE DISCUSSION BEFORE THE FRENCH SOCIETY OF CIVIL 
 ENGINEERS. (Continued.) 
 
 THE paper of M. Sekutowicz, which has been given in full 
 in the preceding chapter, naturally elicited an animated dis- 
 cussion which will be found in the memoirs of the Societe.* 
 
 M. Rene Armengaud gave an account of his own experi- 
 mental researches made at St. Denis in connection with M. 
 Lemale, and these will be discussed at length in a following 
 chapter. 
 
 M. Jean Rey discussed especially the problem of the com- 
 pressor, showing the importance of the development of a satis- 
 factory rotary or turbine compressor. To use a reciprocating 
 compressor would be to deprive the gas turbine of most of 
 the advantages to be gained over the ordinary gas engine. 
 
 Passing to the turbine compressor, M. Rey described the 
 multiple turbine compressor of Rateau, as installed in the 
 mines at Bethune, and constructed by Sautter, Harle & Co. 
 
 In this machine there are four sets of turbine wheels 
 arranged in series, revolving at 4500 revolutions per 
 minute. The first set draws in the air at atmospheric 
 pressure, and raises it to 1.7 kg. per square centimetre abso- 
 lute (24 pounds per square inch). The second set increases 
 the pressure to 2.9 kg. (41 pounds); the third to 4.9 kg., and 
 the fourth to a final pressure of 7.2 kilogrammes absolute 
 per square centimetre (102.4 pounds per square inch). 
 
 This compressor has a capacity of 1 kilogramme of free 
 air per second; it has attained a capacity of 1.25 kilogramme, 
 and the pressure has been pushed up to 8.2 kilogrammes 
 absolute, or 7.2 kilogrammes above atmospheric pressure, or 
 
 * Memoires et Compte Rendu des Travaux de la Societe des Ingenieurs 
 Civils de France: May, 1906. Mm. Armengaud, Rey, Hart, Letombe, Bochet, 
 Deschamps. 
 
 219 
 
220 THE GAS TURBINE. 
 
 about 100 pounds per square inch over and above atmos- 
 pheric pressure. 
 
 The efficiencies of the various sections differ, attaining 70 
 per cent, for the first, and 55 per cent, for the fourth; the 
 mean efficiency of the entire machine being about 63 per cent. 
 
 M. Rey does not consider it practicable to construct such 
 compressors to produce pressures of 30, 40, or 50 kilogrammes 
 per square centimetre, as required by M, Sekutowicz, so 
 that it would be necessary to supplement it by a small piston 
 compressor. 
 
 M. Rey computes the practical efficiency of a turbine 
 by calculating the energy absorbed in the compression of 1 
 kilogramme of air, as well as the energy developed by a 
 kilogramme of burned gases upon the wheel, and his compu- 
 tation shows these two amounts to be about equal, so that 
 there would be no power available for external use. This, 
 however, hardly seems correct, since we have the energy 
 furnished by the burned fuel added to that contained in the 
 compressed air, and their sum should be considered. The 
 practical operation of the turbine of Armengaud and Lemale 
 also furnishes a refutation of the theoretical calculations 
 of M. Rey, since it has developed 500 horse-power, only 
 about one-half of which was required to operate the Rateau 
 compressor by which it was served. 
 
 M. G. Hart called attention to the practical structural 
 difficulties attending the realization of an operative gas tur- 
 bine. In addition to the question of an efficient rotary 
 compressor for high pressures, there are several other ques- 
 tions to be settled. Among these he emphasized the high 
 rotative speeds to be realized, these bringing centrifugal 
 stresses upon the materials of which the resistance would 
 necessarily be reduced by the high temperatures. Even 
 if the difficulties attending the cooling of the rotating parts 
 are successfully overcome, there will be expansion and con- 
 traction stresses which must be taken into account. 
 
THE GAS TURBINE. 221 
 
 As regards the combustion chamber there are several 
 questions involved in its successful construction, although 
 M. Armengaud appeared to have adopted an effective design. 
 M. Hart suggested that several combustion chambers 
 arranged in series might be found more advantageous than 
 a single one of larger size, especially in connection with speed 
 regulation for light loads. 
 
 The practical solution of the gas turbine question, 
 according to M. Hart, appears to lie in the perfection of a 
 number of details, a result attainable only by means of ex- 
 haustive experimental investigations. 
 
 M. Bochet called attention to the fact that high degrees 
 of compression were necessary if high thermal efficiencies 
 were to be attained, citing the experience of the Diesel 
 motor, in which the temperature of compression is sufficient 
 to cause the ignition of the combustible. Such high com- 
 pressions, however, are as yet entirely beyond the powers of 
 the best turbine compressors, a fact which militates severely 
 against the success of the gas turbine so far as efficiency is 
 concerned. 
 
 M. L. Letombe compared the possibilities of the gas tur- 
 bine with the achieved performances of the piston gas en- 
 gine. He believed that the steam turbine had, in some 
 cases, been found preferable to the steam engine because of 
 its greater simplicity, but it seemed as if this point could not 
 be advanced for the gas turbine, because the latter machine, 
 at least so far as developed at present, was more complicated 
 than the reciprocating gas engine. 
 
 In closing the discussion, M. Sekutowicz reviewed the 
 criticisms which his paper had elicited, commenting upon the 
 influence which the variability in the specific heat of gases at 
 very high temperatures might have upon his computations, 
 and emphasizing the desirability of submitting the doubtful 
 points to the test of actual investigation in the mechanical 
 laboratory. 
 
CHAPTER V. 
 
 ACTUAL BEHAVIOR OF GASES IN NOZZLES. 
 
 ONE of the most essential elements in the success of the 
 gas turbine lies in the practicability of the conversion of the 
 original potential energy of the gases into kinetic energy in 
 the nozzle. The extent to which this can be accomplished 
 is yet a matter for discussion. 
 
 Experimental investigations upon the free expansion 
 of gases in nozzles, conducted by Dr. Charles E. Lucke, at 
 Columbia University, appear to show that the nozzle is a far 
 less efficient means for the conversion of energy than the 
 piston and cylinder. Referring to experiments made upon 
 the expansion of compressed air to show the extent of 
 temperature drop, Dr. Lucke says: 
 
 "Holding a thermometer in the stream of air issuing from 
 an open valve or nozzle on a compressed air main will show, 
 for even a pressure drop of 100 pounds per square inch, only 
 three or four degrees temperature change. This also may 
 be due to impact on the thermometer raising the temperature 
 of the moving gases by bringing them to rest on the bulb; 
 but again this will not account for the whole difference be- 
 tween what is observed and what would be were this free 
 expansion equivalent to balanced expansion. To eliminate 
 the errors of impact as much as possible, a thermal couple 
 stretched axially along the jet and made of fine wire has 
 been used by the author for a measurement of the tempera- 
 ture of the air when moving at the maximum velocity. The 
 maximum temperature drop for air under 100 pounds initial 
 pressure, expanding through a steam turbine nozzle into 
 atmosphere, is only 30 degrees F. This result is only 12 per 
 cent, of the temperature drop that would have resulted did 
 the air suffer balanced expansion without gain or loss of heat. 
 
 222 
 
THE GAS TURBINE. 223 
 
 "Another instance of the same lack of equivalence in re- 
 sults by free and balanced expansion is found in the experi- 
 ments of Tripler and Linde on the making of liquid air. In 
 this work air highly compressed (2000 to 3000 pounds per 
 square inch) is first cooled by water and then some of the 
 air freely expanded through a hole, the discharge passing 
 around the pipe feeding the hole. This was intended to cool 
 the air in the pipe lower than the critical temperature for 
 liquefaction under the high pressure used. The results 
 were enormously different from the case for balanced expan- 
 sion, the temperature drop through the nozzle being about 
 \ degree F. per atmosphere-pressure drop, according to one 
 report. More accurately the results for the Linde process 
 are shown in the following table, the initial pressure being 
 220 atmospheres. 
 
 Temperature approaching the Actual temperature drop 
 
 nozzle. through nozzle. 
 
 + 30 F. 35 F. 
 
 F. 65 F. 
 
 30 F. 80 F. 
 
 60 F. 96 F. 
 -100 F. 112 F. 
 
 150 F. 135 F. 
 
 "Unless, by an increase of knowledge of free expansion of 
 perfect gases, it becomes possible to produce results equiva- 
 lent to those obtained with balanced expansion, there cannot 
 be the same amount of heat transformed into work by the 
 gas turbine engine as by the cylinder-and-piston gas engine." 
 Later investigations made by Dr. Lucke in operating 
 a De Laval steam turbine with compressed air gave interest- 
 ing results, an abstract of which is here given. 
 
 "For convenience of operation the air was cold air, 
 whereas in the practical gas turbine the air would be hot 
 and possibly more or less mixed with steam, or possibly no 
 air at all, but carbon dioxide. In any event, the working 
 
224 THE GAS TURBINE. 
 
 fluid would be largely a perfect gas. The turbine used was 
 a De Laval standard 30 horse-power machine intended for 
 steam at 110 pounds pressure and having six nozzles. The 
 turbine wheel runs at 20,000 revolutions per minute, and the 
 power shaft 2000 revolutions. The air used for driving 
 the turbine was measured by a Westinghouse metre. The 
 tests were run on no load, because the compressor used was 
 not sufficiently large to supply the amount of air needed at 
 full load, or even at full speed without load. With each type 
 of nozzle three different initial pressures were used, each 
 with a different number of nozzles. Readings were taken of 
 the temperature of the air entering the turbine and the 
 temperature of the air in the exhaust chamber, with the 
 corresponding pressures. The nozzles fitted to this turbine 
 in holes Numbers 1 and 4 are 110 pounds pressure and 25 J 
 inches vacuum; in holes Numbers 3 and 6 for 110 pounds 
 pressure and 26.3 inches vacuum; and in holes Numbers 2 
 and 5 for 110 pounds steam pressure and atmospheric 
 exhaust. The results of the pressure-drop runs are given 
 in the following table, which also gives the theoretical 
 temperature-drop, assuming an adiabatic expansion of air 
 between the same pressures. 
 
 "From this it appears that the temperature-drop realized 
 varies from 4 to 18 per cent, of the theoretical or adiabatic 
 temperature-drop. The preceding results are given with re- 
 spect to speeds also, which varied from 520 to 1920 revo- 
 lutions per minute. To show according to what law this 
 complete process takes place, the exponent of the tempera- 
 ture ratio in the equation between pressure ratio and tem- 
 perature ratio, which for adiabatic expansion of air is .29, 
 was determined and found to lie between .1005 and .0380." 
 These results obtained by Dr. Lucke must be compared 
 with the practical ones secured by the engineers of the 
 Societe des Turbomoteurs at St. Denis and the experiments 
 made by M. Alfred Barbezat upon the small experimental 
 
THE GAS TURBINE. 
 
 225 
 
 8<>COCOCOGOGCGOOOGOOOGOCOGOGCCD?OGOCOCOGOOOOOO?OCOCOCO 
 CO H- I 'COGO^ICOOl-^I<COOfcOCOt > tO GO CO H- ' h- Ol I '^GOCOtO^lCSCOGO 
 
 ClOOlGOtOGOCOGCCCGOrfi.OCOGCOib-'CCtO 
 
 ascoco^co GO oc oo GO ox 
 
 OasCni ' CO O5 I CO CO GO 
 
 1 1 Hi iLLiiiiu 1 1 1 1 jj n 
 
 ; OiOO:i 'OH- 'OOStOOCOCO 
 
 T3 
 
 II 
 
 -a |_i (-1 t ' tO I i 
 
 tOOiH- '^lOCni ' 
 
 GOtOOi 
 
 t ' CO 4^ h- h- 1 
 
 s 
 
 I* 
 
 3 
 
 5' * 
 
 g ^ 
 
 g t9 
 
 I fe 1 
 
 II 
 
 I 
 
 O5 GO CO 'in CO 
 
 OiOOC/iCnOOOO 
 
 
 
 o 
 
 15 
 
226 THE GAS TURBINE. 
 
 turbine of Mm. Armengaud and Lemale show a much 
 greater drop. It is greatly to be desired that this whole 
 subject of free expansion in nozzles for steam, for air, and 
 for mixed gases, at various temperatures should be thor- 
 oughly investigated experimentally, and it might well occupy 
 the efforts of some of the highly equipped mechanical and 
 physical laboratories of the technical schools. 
 
CHAPTER VI. 
 
 THE PRACTICAL WORK OF ARMENGAUD AND LEMALE. 
 
 THE most complete account of the Armengaud and Le- 
 male turbine, the gas turbine which, by its practical perform- 
 ances has done the most to demonstrate that the gas tur- 
 bine is a reality, and not merely an academic discussion, is 
 contained in an article by the late M. Rene Armengaud, 
 published in Cassier's Magazine, and here reprinted.* 
 M. Armengaud reviews the principles of the gas turbine, 
 and describes some early devices, and then proceeds : 
 
 Heat motors in general service at the present time may 
 be grouped into the following classes : 
 
 1. Alternating steam motors (reciprocating steam en- 
 gines). 
 
 2. Alternating combustion motors (reciprocating gas 
 engines). 
 
 3. Continuous steam motors (steam turbines). 
 
 4. Continuous combustion motors (gas turbines). 
 
 Of these various machines the latest, and certainly the 
 least known, is that which appears to have a most interesting 
 future, the gas turbine, and it is this which I now propose 
 to discuss. 
 
 A successful gas turbine aims to combine the great advan- 
 tages of the gas engine, including the elimination of the 
 steam boiler and a high thermal efficiency, with the special 
 advantages of the steam turbine, i. e., simplicity of construc- 
 tion, lightness, and the greatly desired property of continu- 
 ous motion in one direction, with the accompanying features 
 of control and regulation. 
 
 The various plans which have been discussed for the de- 
 
 * The Gas Turbine. Practical results with actual operative machines in 
 France. By Rene Armengaud, Cassier's Magazine, January, 1907. 
 
 227 
 
228 
 
 THE GAS TURBINE. 
 
 sign of the gas turbine may be divided into three groups : hot- 
 air turbines, explosion turbines, and combustion turbines. 
 So far as the first group is concerned, I have not attempted 
 to make any investigations in this direction, believing this 
 system to offer few advantages. The only machine of this 
 kind of which I have any knowledge is that of Dr. Stolze, of 
 Charlottenburg, of which the following description is ab- 
 stracted from his patents. Air is compressed by means of a 
 helicoidial compressor to about 1J atmospheres. The air, 
 after having circulated about a furnace, expands, and is then 
 passed through a turbine attached to the same shaft as the 
 compressor. 
 
 FIG. 74.-<-General arrangement of explosion gas turbine. 
 
 In the case of the second group, the explosion turbines, 
 the air compressor is eliminated or greatly simplified. The 
 explosive mixture is formed in the same chamber in which 
 it is ignited, being either at atmospheric pressure or slightly 
 above, and by its expansion, consequent upon the explosion, 
 it acts upon the turbine wheel. The principle of such a 
 machine is shown in Fig. 74. The explosion chamber is 
 .closed at the back by a valve A held to its seat by a light 
 spring B, the chamber having an expanding nozzle opening 
 at C. The gas enters at small openings, as at E under the 
 seat of the valve, and the air is admitted at F, the mixture 
 being ignited electrically at H, and discharged through tho 
 
THE GAS TURBINE. 
 
 229 
 
 nozzle C upon the buckets of the turbine wheel T. The dis- 
 charged gases pass through an induced current nozzle G which 
 acts to reduce the temperature of the issuing gases and lower 
 the velocity of the jet as it acts upon the turbine wheel. 
 
 Such an apparatus, when properly proportioned, will make 
 about three explosions per second, and will continue to run 
 automatically after it has once been started. 
 
 o 0.05- o.i o 
 
 FIG. 75. Explosion turbine diagram. 
 
 Various theories have been advanced to explain the ac- 
 tion of this device. The most satisfactory explanation of 
 the periodic action is that of the sudden cooling of the cham- 
 ber after each explosion. This cooling causes a correspond- 
 ing drop in the pressure, followed by the opening of the valve 
 A and the aspiration of the air and gas, and as soon as the 
 explosive mixture reaches the igniter a fresh explosion 
 follows. In Fig. 75, the variations in pressure in the cham- 
 ber are shown as a function of time. The maximum effective 
 pressure ranges from 2 to 3 kilogrammes per square centi- 
 metre, or about 30 to 45 pounds per square inch, although 
 the theoretical pressure in such an open vessel should 
 reach 4 kilogrammes, so that the mixture of the gas and air 
 is probably imperfect. 
 
 Theoretically, the explosion turbine should have a certain 
 thermal advantage over the corresponding cycle for a com- 
 bustion turbine. The specific heat at constant volume be- 
 
230 
 
 THE GAS TURBINE. 
 
 ing lower than the specific heat at constant pressure, the 
 same quantity of heat acting upon the same mass of gas 
 should produce a higher temperature after the explosion 
 than after a combustion. Since, according to the principle 
 of Carnot, the efficiency is proportional to the maximum 
 temperature of the fluid before expansion, the explosion tur- 
 bine should be more efficient than the combustion turbine. 
 Unfortunately the high velocities of discharge of the gases, 
 and the variations in the pressure, render it impracticable to 
 realize more than a small fraction of the energy of the jet 
 upon the wheel. Thus, the theoretical efficiency of such a 
 
 FIG. 76. Combustion gas turbine. A, combustion chamber. B, fuel inlet. C, fuel sprayer. 
 E, expansion nozzle. F, turbine. 
 
 machine should be about 16 per cent., while the actual per- 
 formance does not exceed 3 to 4 per cent. In addition to 
 this defect there are operative difficulties with the springs 
 and valves, and the frequent breakages and delicate adjust- 
 ments have rendered experiments to improve the apparatus 
 unsatisfactory. 
 
 There remains, then, the combustion turbine, which, in 
 spite of the necessity for an air compressor, is greatly 
 to be preferred, especially for large units. This machine 
 consists in principle of a combustion chamber A Fig. 76. 
 supplied by a continuous current of compressed air, and also 
 by a continuous supply of liquid fuel, gasoline, petroleum, 
 
THE GAS TURBINE. 231 
 
 or the like, under pressure through a tube B, the mixture 
 being ignited at the start by a platinum wire C, the combus- 
 tion developing a constant temperature of about 1300 de- 
 grees C. in the chamber D. The fluid products of combustion 
 are then continuously discharged through a nozzle E, upon 
 the buckets of the turbine wheel F. 
 
 The principal defect in this apparatus in comparison with 
 the reciprocating gas engine is the necessity for a separate 
 air compressor, instead of having the compression of the 
 charge effected in the motor itself. This defect is partially 
 remedied by the diminution of the losses through the walls, 
 and by the possibility of an expansion which is practically 
 
 C B 
 
 FIG. 77. Diagram for combustion turbine. 
 
 adiabatic. The combustion is also more complete than is 
 possible in a working cylinder, and all the products of com- 
 bustion are utilized. 
 
 The action of a combustion turbine is graphically shown 
 in Fig. 77. In this diagram the area OABC represents the 
 energy required for the air compressor. The combustion of 
 the liquid fuel increases the volume from CB to CD. If any 
 vapor of water is introduced, this volume will be diminished 
 from CD to CD l} while at the same time its mass increases 
 the volume of CD^ to CE. The effective energy exerted by 
 the turbine will be represented by the area OFEC and that 
 available after the deduction of the work of compression will 
 be AFEB. 
 
232 THE GAS TURBINE. 
 
 In endeavoring to produce such a cycle in an actual 
 working machine, the following practical difficulties must 
 be overcome: 
 
 A gaseous fluid moving at a high velocity must be kept 
 constantly ignited, by a device which must not be affected 
 by the high temperature of the combustion chamber. 
 
 The mixture of the combustible and the air must be made 
 as perfect as possible. 
 
 The injurious action of the gaseous products at a high 
 temperature upon the parts of the apparatus, and upon the 
 turbine wheel itself, must be prevented. 
 
 For three years a machine complying with these condi- 
 tions has been running successfully in the shops of the So- 
 ciete des Turbomoteurs at Paris, this apparatus being the 
 Armengaud-Lemale turbine, of which some further descrip- 
 tion will be given. 
 
 The original machine was made from a De Laval steam 
 turbine of 25 horse-power, arranged to be operated with 
 compressed air instead of steam. The air was supplied at 
 any desired pressure from a high speed compressor, of which 
 the efficiency had been closely determined, while prolonga- 
 tions of the pipe which connected the compressor to the tur- 
 bine formed the combustion chamber. At the entrance of 
 each chamber the gasoline, mixed with the air, was ignited 
 by an incandescent platinum wire, this ignition being neces- 
 sary only at the starting of the operation, the combustion 
 being maintained continuously thereafter at constant pres- 
 sure. The combustion chambers were lined with refractory 
 material, and a temperature of about 1800 degrees C. was 
 produced. In order to reduce the temperature to practical 
 limits the chamber was cooled by the introduction of vapor 
 of water generated in a spiral imbedded in a portion of 
 the combustion chamber. The steam thus produced was 
 allowed to mingle with the gases of combustion before expan- 
 sion in such proportion that the temperature of the mixture 
 was about 400 degrees C. 
 
 
FIG. 89. The 30 horse-power experimental gas turbine of the Societe des Turbomoteurs. 
 
FIG. 90. The 300 horse-power gas turbine of Armengaud and Lemale connected to th 
 Rateau polycellular turbine compressor. 
 
THE GAS TURBINE. 233 
 
 Although this apparatus was necessarily crude and not 
 proportioned in such a manner as to give the best results, it 
 enabled the conditions essential for a good efficiency to be 
 determined. 
 
 Among the practical points thus determined were proofs 
 that it was entirely possible to maintain the combustion 
 chamber, turbine wheel, and fuel pulverizer in operative con- 
 dition. The experiments also showed it to be practicable to 
 maintain a very high temperature continuously in the actual 
 combustion chamber, and, by means of this high heat to 
 secure a perfect combustion of any combustible. The work 
 of compression having been carefully ascertained for the pur- 
 pose of deducting it from the brake power developed by 
 the entire machine, it appeared that even with this imperfect 
 apparatus the total power was about double that necessary 
 to drive the compressor. This result was attained with a 
 pressure of about 10 kilogrammes per square centimetre, and 
 a temperature of 400 degrees C. at the exhaust. 
 
 As has already been said, the excessively high tempera- 
 tures developed were reduced in the earlier experiments by 
 mixing a certain quantity of steam with the gases of com- 
 bustion before expansion. This method, while accomplish- 
 ing the result desired, also acted to lower the efficiency of the 
 turbine, doubtless because of the latent heat of vaporization 
 lost in the exhaust. In the diagram, (Fig. 78) the curves 
 show the manner in which the economical performance of this 
 machine varied, represented as a function of the upper pres- 
 sure and of the temperature of the exhaust gases. This dia- 
 gram has been computed upon a basis of 60 per cent, 
 efficiency of the turbine wheel, and 80 per cent, of the com- 
 pressor. For example, with a pressure of 30 kilogrammes 
 per square centimetre, and an exhaust temperature of 450 
 degrees C., an efficiency of 18 per cent, is obtained. 
 
 It thus appears that the efficiency depends both upon the 
 pressure and upon the temperature of the exhaust gases. 
 
234 
 
 THE GAS TURBINE. 
 
 In order, therefore, to obtain the best efficiency it is neces- 
 sary to prevent cooling the gases before expansion, either by 
 introducing steam into the combustion chamber, or other- 
 wise, and to effect the greatest possible reduction in temper- 
 ature in the expansion alone. 
 
 The difficulties accompanying the high temperatures 
 may be met in the case of the combustion chamber and other 
 fixed parts by the use of a water jacket and by the employ- 
 
 t 1750'C 
 t 
 
 SCO 
 
 iSo 
 
 I 5 to 20 30 
 
 FIG. 78. Gas turbine economy curves. 
 
 ment of a refractory lining, and the real difficulties are 
 reached only when it becomes necessary to provide for the 
 effect of the highly heated fluid upon the rotating metallic 
 wheel, already weakened by the heavy centrifugal stresses 
 to which it is necessarily subjected. 
 
 The most practical way of keeping the turbine wheel cool 
 is to follow the jet of hot gases by another jet of a low tem- 
 perature so that the buckets of the wheel pass successively 
 through alternate hot and cool zones, the average tempera- 
 
THE GAS TURBINE. 
 
 235 
 
 ture of the two jets being sufficiently low to prevent injury to 
 the metal. The low temperature jet found most practicable 
 is that of low pressure steam, and this is readily provided 
 from the water jacket and from a device arranged as a re- 
 generator in connection with the exhaust gases. 
 
 This arrangement, shown in Fig. 79, gives a general idea 
 of the system. The air from the compressor enters at D and 
 is mixed with the liquid fuel in the concentric nozzle EE and 
 
 ^ 
 
 Water 
 
 FIG. 79. Mixed gas and steam turbine. Air enters at D, fuel at F, the ignition is made 
 at G. The combustion chamber A is lined with carborundum. The nozzle H is water- 
 jacketed, and the hot water passes to the steam generator L, which is heated by the exhaust 
 gases from the turbine. The steam acts to propel and cool the wheel by the nozzle M . 
 
 ignited by the platinum wire at G. The combustion takes 
 place continuously at constant pressure in the chamber A, 
 and the products of combustion are discharged through the 
 expansion nozzle H upon the buckets of the turbine wheel 1. 
 The nozzle itself is protected by a water jacket C, the water 
 leaving the jacket at K. On the other side of the wheel 
 there is arranged a sort of flash steam generator L, this 
 being composed of a serpentine pipe of continually increasing 
 diameter, the water entering the small end at K, this en- 
 
236 
 
 THE GAS TURBINE. 
 
 trance forming a part of the discharge pipe from the water 
 jacket of the nozzle H. The steam generator L is placed in 
 the path of the exhaust gases leaving the turbine wheel, 
 and these highly heated gases furnish the heat necessary to 
 convert the water into steam, the vapor thus produced 
 being discharged through the nozzle M upon the turbine 
 wheel, thus acting both to aid in the propulsion and to 
 form a zone of comparatively low temperature to abstract 
 heat from the wheel. By this arrangement it is possible to 
 reduce the temperature of the wheel to practicable limits, 
 provided the temperature of the exhaust gases is sufficiently 
 high to produce enough steam. That is, the expansion of 
 
 FIG. 80. The Lemale combustion chamber. 
 
 the gases in the nozzle must not lower their pressure and 
 temperature so far as to keep down the volume of steam too 
 low. In such case it is always practicable to admit a small 
 quantity of superheated steam into the combustion chamber 
 and thus obtain the required temperature without affecting 
 the efficiency of the machine too much. 
 
 The general heat balance of a gas turbine using a steam 
 regenerator according to the above plan is shown in Fig. 81, 
 in which the total -quantity of energy produced by the fuel 
 is represented by the dimension X, and the various losses 
 indicated by the cross hatched portions in the body of the 
 diagram. The efficiency of the machine is then obtained 
 as the ratio Y : X, Y being composed of two parts, one of 
 
THE GAS TURBINE. 
 
 237 
 
 which is obtained from the action of the gases upon the wheel 
 and the other by the steam. 
 
 In this arrangement the expansion of the gases and the 
 steam occur in parallel, so to speak, this being clearly 
 indicated in the diagram. For constructive reasons, how- 
 ever, it is found convenient to adopt the previous system, 
 the steam produced in the regenerator being delivered into 
 
 FIG. 81. Heat balance diagram for mixed gas turbine. I, total energy developed by 
 the combustion of the fuel. II, kinetic energy available at the discharge of the nozzle. 
 Ill, energy developed on the turbine wheel. IV, energy developed less the power required 
 to drive the compressor. V, energy recovered in the steam. VI, energy available in the 
 expanding steam. VII, energy developed by the steam on the wheel. X, total energy con- 
 tained in the fuel. Y, total energy produced in indicated work, a, Radiation losses from 
 the combustion chamber, b, Loss in the nozzle, c, Compressor losses, d, Theoretical 
 work of compression, e, Radiation losses from the turbine. /, Losses in the exhaust 
 steam, g, Losses in the exhaust gases. 
 
 the same nozzle as that used for the gases, and this plan has 
 been adopted in our most recent turbine, even at some re- 
 duction in the thermal efficiency. 
 
 This machine, shown in several views, is of the same 
 general type as the Curtis steam turbine, and is capable of 
 delivering from 400 to 800 h. p., according to the compressor 
 capacity utilized. The turbine is operated at 4000 revolu- 
 
238 THE GAS TURBINE. 
 
 tions per minute, and the speed regulation is effected by a 
 throttling valve in the air admission pipe for small speed var- 
 iations, and by a change in the fuel supply for larger changes, 
 the regulating valves being controlled by a Hartung gover- 
 nor. There are three pumps attached to the machine, one 
 the air compressor, another for the fuel supply, and the 
 third for the water. 
 
 The combustion chamber is made of cast iron, lined with 
 carborundum, the cast iron being protected with a water 
 jacket. An elastic non-conducting lining is placed between 
 the carborundum and the outer shell, this providing a bed- 
 ding for the carborundum and also permitting a slight move- 
 ment for differences in expansion and contraction. The 
 extremity of the chamber and the nozzles are surrounded 
 by a jacket space in which the water and steam circulate, 
 the nozzles being of the expanding type similar to those of 
 the De Laval steam turbine, although the expansion is 
 completed in a shorter time. It is necessary that the expan- 
 sion should be effected in a single operation in order that the 
 temperature be sufficiently reduced before the gases reach 
 the wheel. 
 
 The gasoline or other liquid fuel is delivered into the 
 combustion chamber through a pulverizer, or atomizer, the 
 construction of which is shown in Fig. 86. This is arranged 
 with a reverse annular opening B delivering the fuel back- 
 ward against the incoming stream of air, the angle causing 
 the gasoline to form a sort of cone of minute particles, these 
 becoming ignited as soon as their decreasing velocity per- 
 mits. The preheating of the fuel also riders the ignition 
 easy. The atomizer is protected against the intense radiant 
 heat of the chamber wall by the current of air with which it 
 is continually surrounded. The igniting coil of platinum 
 wire D is protected by a steel cap (7, the electric current 
 entering by the central insulated rod E, the circuit being 
 completed through the machine itself. A pressure of 2 
 
FIG. 82. The Armengaud and Lemale gas turbine. A view of the 500 horse-power turbine 
 in the experimental laboratory at St. Denis. 
 
 FIG. 83. The Armengaud and Lemale gas turbine. This view and the preceding one 
 show the wheel casing, governor, and general arrangement. 
 
FIG. 84. The Armengaud and Lemale gas turbine. This view shows the combustion 
 chamber, air and fuel inlets and connections. 
 
 FIG. 85. The Armengaud and Lemale gas turbine. Another view of the combustion 
 chamber, with air and fuel connections. 
 
THE GAS TURBINE. 
 
 239 
 
 volts is found sufficient to render the platinum wire incandes- 
 cent. The atomizer is inserted into the combustion cham- 
 ber in such a manner that it can readily be removed for in- 
 spection and cleaning, this operation also giving complete 
 access to the chamber itself. 
 
 The turbine wheel is arranged to be cooled by water cir- 
 culation as shown in Fig. 87, this representing a section of 
 the rim and a portion of the disc. A and B are circular 
 channels in the body of the rim, these being supplied with 
 water by radial passages as at E. Small passages also per- 
 mit the water to enter into each blade of the turbine and the 
 
 FIG. 86. Section of pulverizer and igniter. 
 
 difference in specific gravity between the hot and cold water 
 is found to make an automatic circulation, in connection with 
 the centrifugal force due to the high velocity of rotation. 
 
 The air supply for the turbine is furnished by a polycel- 
 lular rotary compressor of the Rateau system. This impor- 
 tant adjunct to the gas turbine, shown in Fig. 88, is composed 
 of a number of turbine blowers arranged in series and espe- 
 cially designed to be operated at very high rotative speeds, 
 so that it may be directly connected to the gas turbine. 
 
 The importance of the compressor is second only to that of 
 the turbine itself, since it is of little value to possess a rotary 
 
240 
 
 THE GAS TURBINE. 
 
 combustion motor if a reciprocating compressor is a necessary 
 auxiliary. It is on this ground, more than almost any other, 
 that the design of a successful gas turbine has been consid- 
 ered problematical, and here, as in many other cases in the 
 
 FIG. 87. Section of wheel of gas turbine, showing passages for cool ing- water. 
 
 history of the development of a device, the progress of other 
 departments of work becomes essential to complete success. 
 The work of M. Rateau in the improvement of the steam 
 turbine is well known, and by the application of the expe- 
 rience thus gained, a machine for the supply of compressed 
 
-II 
 
 II 
 n 
 
 |t 
 
 l! 
 
 t- O 
 
THE GAS TURBINE. 241 
 
 air to the gas turbine has been produced, involving only ro- 
 tary motion, and thus capable of being driven directly by the 
 turbine, and having an efficiency sufficiently high to permit 
 a good performance of the combined apparatus. 
 
 The Rateau compressor is practically a reversal of the 
 steam turbine, and is composed of a number of elements 
 connected in series, so that the pressure is cumulative, the 
 action being similar to that of the multiple centrifugal 
 pumps which have been employed with such success for 
 delivering water against high heads. 
 
 Each element of this compressor consists of two parts, 
 the revolving wheel and the diffuser. The diffuser is ar- 
 ranged to provide discharge passages for the air, having 
 gradually increasing section for the flow, in order that the 
 velocity of the air as it leaves the wheel may be reduced 
 with the least possible loss, the kinetic energy being con- 
 verted into pressure. The length of the machine is such that 
 intermediate bearings have been introduced to provide sup- 
 port and stiffness to the rotating parts, and the whole design 
 of the compressor has been so carefully worked out that an 
 efficiency of 65 per cent, has been already attained, and pres- 
 ent experiments indicate that this performance will be sur- 
 passed. 
 
 In Fig. 88 a Rateau compressor of three sections is shown, 
 but larger machines have been constructed, and the pressures 
 attained naturally depend upon the number of sections. 
 Experiments have shown that in the first section the air is 
 compressed to 1.7 kilogrammes per square centimetre, or 
 about 24 pounds per square inch, absolute, the succeeding 
 pressures being 2.9 kilogrammes, 4.9 kilogrammes, and 7.2 
 kilogrammes per square centimetre, the latter corresponding 
 to 112 pounds per square inch above vacuum. 
 
 In a subsequent issue of Cassier's Magazine,* an article 
 was published containing a communication from M. Alfred 
 
 * Cassier's Magazine, April, 1908. 
 16 
 
242 THE GAS TURBINE. 
 
 Barbezat, who had been associated with M. Rene Armen- 
 gaud, and who continued in the work after the death of the 
 latter. 
 
 M. Barbezat reviewed the early work of M. Armen- 
 gaud, and then described the later progress as follows: 
 
 The general principle of the machine involves the delivery 
 of air under pressure into a pear-shaped chamber lined with 
 refractory material and provided with an expanding nozzle 
 through which a uniform flow of gases can be delivered upon 
 the blades of the wheel. In the centre of the air nozzle there 
 is arranged an axial tube, with a pulverizer at the inner end, 
 through which the fuel, in the form of gasoline, or similar 
 liquid hydrocarbon, is forced into the chamber. The electric 
 sparking device enables the fuel to be ignited on starting, 
 after which the high temperature of the chamber maintains 
 the combustion indefinitely. The high temperature pro- 
 duced by the combustion greatly increases the volume of the 
 air, and this, together with the gaseous products of the com- 
 bustion of the fuel, flows at a high velocity through the ex- 
 panding nozzle upon the blades of the wheel. 
 
 In dealing with such high temperatures, the temperature 
 of the combustion being about 1800 degrees C., the best re- 
 fractory lining for the combustion chamber has been found to 
 be carborundum, this being a product of the electric furnace, 
 and thus having already sustained even higher temperature 
 than those in the turbine combustion chamber. An elastic 
 backing of asbestos provides for the expansion of the carbor- 
 undum lining, and the nozzle through which the gases are 
 discharged upon the wheel is also made of carborundum. 
 
 In addition to the provision of a refractory lining, it has 
 been found necessary to surround the combustion chamber 
 with a water jacket in the form of a coil of pipe imbedded 
 in the metal of the chamber walls, much in the same manner 
 as such coils are used in the tuyeres of blast furnaces, and 
 the circulation of the water in the coils aids in keeping the 
 
THE GAS TURBINE. 243 
 
 temperature of the combustion chamber walls within practi- 
 cable limits. 
 
 After the water has circulated in the jacket tube it is 
 delivered, through small holes, into the gases just before 
 they enter the nozzle, and is there converted into steam, 
 this acting both to lower the temperature of the issuing gases 
 to a point where they will not injure the blades of the tur- 
 bine, and also itself being discharged upon the wheel with 
 the gases and forming a part of the jet, which is thus com- 
 posed of mingled gas, steam, and highly heated air. 
 
 In order to obtain the desired result of a machine involv- 
 ing only rotary motion, it is necessary that the compressed 
 air by which the combustion chamber is fed should be pro- 
 duced, not by a reciprocating piston compressor, but by 
 some form of rotary machine, preferably so arranged 
 that it can be coupled directly to the turbine itself. This 
 means that the complete gas turbine must also include a 
 rotary air compressor, and that such a compressor must have 
 a high efficiency in itself, otherwise it will produce such a 
 large proportion of negative work as to detract materially 
 from the efficiency of the combined machine, even though 
 the actual thermal efficiency of the turbine be high. 
 
 After a number of experiments upon single impeller tur- 
 bine air compressors, driven at high rotative speeds by De 
 Laval steam turbines, the services of Professor Rateau were 
 enlisted in the work, and a multiple turbine compressor, 
 designed by him especially for this work, was constructed at 
 the works of Brown, Boveri & Co., at Baden, Switzerland. 
 This machine is arranged in three sections and provided with 
 continuous cooling circulation, and, being thoroughly tested, 
 was found to be capable of delivering one cubic metre of air 
 per second at a pressure of 6 to 7 atmospheres, with an 
 efficiency ranging between 60 and 70 per cent. 
 
 The illustration (Fig. 90) shows the arrangement with this 
 compressor coupled directly to the large experimental tur- 
 
244 THE GAS TURBINE. 
 
 bine constructed by M. Armengaud, the turbine and the 
 compressor thus forming practically one machine. 
 
 In this arrangement the compressor was found to absorb 
 about one-half the total power developed by the turbine, the 
 machine, when running at about 4000 revolutions per minute 
 developing about 300 horse-power over and above the nega- 
 tive work absorbed by the compressor. At the present 
 time experiments are being made upon the thermal efficiency 
 of the machine, which is, as yet, not as high as that of the re- 
 ciprocating gas engine; but these tests are not yet completed, 
 and the results not available for publication. 
 
 During the past few months a practical application of 
 this turbine has been made in connection with the operation 
 of submarine torpedoes. 
 
 It is well known that in certain types of such machines 
 the motive power for the brief period which elapses between 
 the discharge and the contact with the target is derived 
 from a store of compressed air, and in some such torpedoes 
 the compressed air acts upon a turbine wheel similar to the 
 steam turbine. This principle has now been extended to the 
 use of the gas turbine, the compressed air from the reservoir 
 passing through a combustion chamber and the total 
 products of combustion together with the vapor of water 
 acting on the turbine, and its capacity thus increased over 
 that operated by compressed air alone. 
 
 The turbines made for this purpose develop 120 horse- 
 power at a speed of 1000 revolutions per minute, the 
 expansion ratio being 8.4. The weight of the turbine alone 
 is 73.16 kilogrammes, or about 1.3 pounds per horse-power. 
 Including the weight of the reservoir of compressed air, 
 together with the petrol and water for a discharge lasting 80 
 seconds, the total weight of the whole apparatus is about 
 295 kilogrammes, or a little less than 2.5 kilogrammes, or 
 5.5 pounds per horse-power. 
 
 Although the gas turbine is, therefore, still in the experi- 
 
THE GAS TURBINE. 
 
 245 
 
 mental stage, it has made material advances in the past 
 year, the 300 horse-power combined compressor and turbine 
 being an accomplished fact, and a number of 120 horse- 
 power machines of a special type being actually installed 
 in submarine torpedoes completed for active service. 
 When this rate of progress is compared with the time re- 
 quired to bring the reciprocating gas engine to its present 
 state of perfection, there appears to be reason for encourage- 
 ment and interest. 
 
 WITH 
 
 ' \ 
 
 WITHAIRX 
 
 GAS 
 
 \ 
 
 6000 10000 15000 20000 25000 
 
 FIG. 91. 
 
 The accompanying diagram (Fig. 91) gives the results of 
 practical tests of a Rateau multiple air compressor, as well 
 as a characteristic curve of the small experimental gas tur- 
 bine of M. Armengaud and Lemale, as communicated to the 
 author by M. Alfred Barbezat, who has been associated 
 with the late M. Rene Armengaud in much of his work. 
 
 Experiments with the large turbine and compressor have 
 shown the operative practicability of the machine, but the 
 
246 THE GAS TURBINE. 
 
 consumption of petrol (1200 to 1300 grammes per horse- 
 power hour) being too high for industrial purposes. Exper- 
 iments which we are not yet at liberty to publish, however, 
 indicate that the fuel consumption will be very materially 
 lowered. 
 
 The following data concerning gas turbines furnished by 
 the Societe des Turbomoteurs to the Creusot Works for use 
 in submarine torpedoes, show the extent to which the practi- 
 cal development of the gas turbine has already attained: 
 
 Power 120 horse-power 
 
 Speed 1500 revolutions per minute 
 
 Expansion ratio 8.4 to 1.4 atmospheres (1: 6) 
 
 Weight of turbine 73.16 kilogrammes (162 Ib.) 
 
 Weight of petrol 1.55 kilogrammes (3.4 Ib.) 
 
 Weight of water 11.00 kilogrammes (24.2) 
 
 Weight of air and reservoir, 32 + 177 = 209 kilogrammes (627.6 Ib.) 
 
 ' During the past year there has been built in Paris, by 
 M. Karavodine, an explosion gas turbine developing about 
 2 horse-power, and operating with regularity and success; 
 and from recent tests of this machine by M. Alfred Barbezat 
 we are able to give some quantitative data concerning its 
 performance. 
 
 The Karavodine explosion gas turbine tested by M. 
 Barbezat was provided with four explosion chambers, the 
 products of the explosions being directed through four 
 separate nozzles upon a single turbine wheel. This wheel 
 was of the De Laval type, about 6 inches in diameter (150 
 centimetres), carried upon a flexible shaft, and fitted with 
 a Prony brake. 
 
 The general construction of the explosion chambers is 
 shown in the illustration. The body of the chamber B is 
 composed of cast iron and provided with a water jacket A, 
 which does not extend all the way to the top, thus per- 
 mitting the portion nearest the discharge nozzle to become 
 heated. At the lower end there is provided an opening C 
 for the entrance of the fuel, either gas or hydrocarbon vapor; 
 
THE GAS TURBINE. 
 
 247 
 
 also, an opposite opening D for the entrance of air. These 
 openings are both provided with throttle valves, not shown 
 in the illustration, by means of which the proportions of 
 air and gas may be regulated. At E is an electric ignition 
 plug, and at F is a plate steel valve, opening inward, and 
 held to its seat by a spiral spring G, its lift being regulated 
 by a set-screw H. The discharge nozzle is shown at 7, and 
 also a portion of the perimeter of the turbine wheel. 
 
 FIG. 92. Combustion chamber of the Karavodine turbine. 
 
 In starting the machine the air opening D is closed by 
 its throttle valve and a blast of air is blown through C, the 
 explosive mixture being ignited by a spark at E. After 
 the first explosion the air entrance D is opened and a sort 
 of pulsometer action follows, thus: After each explosion 
 there follows a depression, or partial vacuum, which acts 
 to draw air and hydrocarbon vapor or gas into the chamber 
 B, lifting the valve F. This mixture is instantly ignited 
 by the spark at E, and another explosion follows, to be 
 again followed by another suction, and so on indefinitely. 
 
248 THE GAS TURBINE. 
 
 After a short time the upper part of the chamber B becomes 
 so hot that the igniter E may be shut off, the charge being 
 exploded by the heat of the chamber. The nozzle 7 is made 
 rather long, and it is found that the friction against the 
 walls and the inertia of its contents prevent any material 
 negative or back suction through it, so that the chamber 
 B is filled at every stroke almost entirely from the air and 
 gas openings below. 
 
 When the tension on the spring G and the lift of the 
 valve F are both carefully adjusted this simple device will 
 run for hours, without miss or interruption, the explosions 
 following each other so closely as to make practically a 
 continuous discharge upon the turbine wheel. 
 
 In order to investigate the action and pressure in this 
 explosion chamber, a special form of recording gauge was 
 made. The pressure in the chamber acted upon a thin 
 steel diaphragm, of which the deflections actuated a small 
 mirror, throwing a beam of light upon a rapidly-moving, 
 sensitive film. The result was the production of a curve 
 of the sine type, in which the ordinates represent pressure 
 and the abscissae show time. 
 
 In the diagram shown in the illustration the solid curved 
 line is made up from the average of a number of diagrams, 
 while the dotted line shows the one which deviated most 
 widely from the mean. During the period A E there was a 
 partial vacuum in the chamber, and the mixed charge was 
 drawn in. From E to D the pressure of the explosion oc- 
 curred, and the contents of the chamber were discharged 
 upon the wheel. The ignition began at B } and the force of 
 the explosion reached its maximum at C, while the period 
 A B includes the inertia action of the gases. The diagram 
 shows that a complete oscillation required 0.026 part of a 
 second, corresponding to between 38 and 39 explosions per 
 second. The mean pressure A F in the diagram was 1.139 
 kilogrammes per square centimetre (absolute), or about 
 
THE GAS TURBINE. 
 
 249 
 
 pounds per square inch, the maximum force of the explo- 
 sion being 1.345 kilogrammes, or about 19 pounds per square 
 inch. The lowest suction pressure was 0.890 kilogramme, or 
 12.6 pounds absolute, thus giving a negative pressure of 
 
 Alnv- 
 
 O.I 
 
 FIG. 93. Diagram of the explosion turbine. 
 
 about 2 pounds to draw the charge in, and a discharge pres- 
 sure of between 4 and 5 pounds on the wheel. 
 
 In the machine tested by M. Barbezat the volume of 
 one chamber was 230 cubic centimetres. Each nozzle was 
 3 metres long and 16 millimetres bore, slightly curved at 
 the end to conform to the shape of the wheel. The wheel 
 
250 THE GAS TURBINE. 
 
 itself was 150 millimetres in diameter, or 5.9 inches, and 
 made 10,000 revolutions per minute, corresponding to a 
 perimeter velocity of 78.5 metres, or about 258 feet per 
 second. 
 
 At the same time the above diagrams were taken the 
 amount of air drawn into the four chambers was measured 
 by a meter, and the quantity of gasoline consumed was 
 measured, while the power developed was determined by 
 the Prony brake. The data and results were as follows: 
 
 Air consumed per hour, 62.5 cubic metres = 80 kg. 
 
 Gasoline, 6.5 litres = 4.7 kg. 
 
 Length of brake arm, 46.4 centimetres. 
 
 Weight on brake, 248 grammes. 
 
 Speed, 10,000 revolutions per minute. 
 
 From these figures the brake power works out 1.6 horse- 
 power, and as the wheel and journal friction was determined 
 at 0.5 horse-power, the actual indicated power was 2.1 
 horse-power. This gives a fuel consumption of 2.24 kilo- 
 grammes of gasoline per horse-power hour, which is very 
 fair for such a small machine, being only about one-third 
 greater than that of the old Lenoir gas engine. 
 
 In considering the availability of such a machine there 
 are a number of considerations other than the mere fuel 
 consumption. The continuous turning effort is often most 
 desirable, and when it is considered that the wheel of this 
 machine was less than 6 inches in diameter, the possibilities 
 of such an apparatus may become evident. The absence of 
 a compressor and corresponding reduction in weight and 
 size give such a machine marked advantages over the 
 combustion turbine, in which the compressor is much larger 
 than the turbine itself, and even if the fuel consumption is 
 as high as indicated above. 
 
CHAPTER VII. 
 
 GENERAL CONCLUSIONS. 
 
 IN the previous chapters there has been shown broadly 
 the mathematical and thermodynamical principles upon 
 which the possibilities of the construction of a practicable 
 gas turbine may be based, together with some account of 
 the success which has attended the design and operation 
 of actual machines. Much remains to be done before the 
 gas turbine can be expected to enter the market in competi- 
 tion with existing gas engines of the reciprocating type, but 
 there are many active and energetic minds at work upon 
 this portion of the problem, and commercial results may 
 soon be expected to follow. 
 
 So far as predictions may be made at this stage of the 
 question, it seems as if the most immediate results are to 
 be anticipated from the so-called " mixed" turbine; the 
 type in which the injection of water for cooling purposes 
 causes the machine to partake of the combined nature of 
 the gas and the steam turbine. This is especially true of 
 the combustion turbine, in which a continuous combustion 
 in a closed chamber provides the gases under pressure to 
 act upon the wheel. The turbine of the explosion type, 
 notwithstanding its low thermal efficiency, appears to have 
 arrived at a practical stage already, and the machine con- 
 structed by Karavodine, and tested by Barbezat, has 
 demonstrated that a dry gas turbine of this type is an 
 operative machine already about as efficient as a steam 
 engine of the same capacity. 
 
 Apart from the question of thermal efficiency, the 
 development of the gas turbine depends to a large extent 
 upon other properties. 
 
 One of the principal difficulties with the reciprocating 
 
 251 
 
252 THE GAS TURBINE. 
 
 gas engine lies in the intermittent character of the impelling 
 forces upon the crank shaft, a defect which the multiplica- 
 tion of cylinders in engines designed for automobiles and 
 aeroplanes is intended to remedy as far as practicable with 
 machines of that type. The continuous rotary action of 
 the turbine is such an advantage as to outweigh to a large 
 degree its present lack of fuel economy. In like manner the 
 high rotative speed lends itself to a corresponding reduction 
 in weight per unit of power, a matter which closely con- 
 cerns the development of mechanical flight. In this matter, 
 as in the case of submarine propulsion, fuel economy is a 
 secondary consideration. The late Professor Langley, in 
 speaking of the engine of his flying machine, is reported to 
 have said that it might burn gold if necessary, so long as it 
 fulfilled all the other requirements of the problem. 
 
 The development of the rotary air compressor has an 
 important bearing upon the success of the combustion tur- 
 bine, and the work of Rateau in this respect has shown 
 what may be accomplished by concentration upon such a 
 question. The analysis of M. Sekutowicz shows the advan- 
 tages of a high degree of compression, and the high efficiency 
 of the Diesel motor is well known to have resulted largely 
 from the high compressions employed in that most econom- 
 ical heat engine. 
 
 What is needed for the further development of the gas 
 turbine, then, is the experimental determination of the 
 data which mathematical analysis has shown to be lacking; 
 data concerning the behavior of gases in diverging nozzles, 
 concerning the action of highly heated gases upon the 
 resistance of materials of construction, data concerning the 
 velocity of efflux from nozzles, data upon the practicability 
 of maintaining extremely high rotating velocities in prac- 
 tical work. 
 
 Here is ample work for the engineering laboratories of 
 technical schools; work which can be conducted with exist- 
 
THE GAS TURBINE. 253 
 
 Ing equipment, and which would form valuable contribu- 
 tions to knowledge, while at the same time providing most 
 fruitful examples for instruction in the very department of 
 engineering in which future progress is to be expected, the 
 subject of the manufacture of power and its utilization to 
 the greatest advantage. 
 
INDEX 
 
 Academie des Sciences, Tournaire's, 
 
 communication to, 14-19 
 Action of heat on metals, 86 
 Actual behavior of gases in nozzles, 222 
 Adiabatic compression, 31, 45, 54, 
 
 123 ; 133 
 
 efficiency of, 118 
 expansion, 35, 37, 48, 69, 70 
 
 law of, 112 
 flow, formula for, 181 
 Advantages of high compression, 128 
 
 148 
 
 of regenerator, 147 
 Aerial motors, 163 
 Air compressor, Rateau, 240 
 compressors, 197 
 
 efficiency of, 91 
 
 Analysis of mixed turbines, 156 
 Applications of the gas turbine, 105, 
 
 218 
 
 Armengaud and Lemale, 227 
 Armengaud and Lemale turbine, 
 
 illustrated, 238 
 Armengaud, Re"ne", 26, 227 
 Atkinson, James: Discussion of Neil- 
 son Paper, 86 
 Atomizer for gas turbine, 239 
 
 Banki motor, 132 
 Barber's turbine, 11-13 
 Barbezat, Alfred, 26, 241 
 Barkow, R., 8, 27, 121, 130 
 Baumann, A., 27 
 Blades, losses in, 191 
 Blast furnace gases, 176 
 Bochet, A., 26, 221 
 Boulton's patents, 21, 22 
 Bourdon, M., 201 
 Bourne's suggestions, 20, 22 
 Bray ton engine, 31, 32 
 
 Bucholz turbine, 104 
 
 Burdin, 14, 15, 19 
 
 Burstall, F. W.: Discussion of Neil- 
 son Paper, 82 
 
 Butler, Edward: Discussion of Neil- 
 son Paper, 96 
 
 Carnot's formula, 29, 91 
 
 cycle, 29, 74, 116 
 Cassier's magazine, 26, 227 
 Centrifugal force, stresses due to, 48 
 Circulation of cooling water, 39 
 Civil engineers of France: Discus- 
 sion before, 108-221 
 Clark, Ade, 84 
 
 Classification of gas turbines, 191 
 Clerk, Dugald, 30 
 Clerk, Dugald: Discussion of Neilson 
 
 Paper, 97 
 Combes, 14 
 
 Combination cycles, 167 
 Combined gas and steam turbines, 
 
 60-70, 153 
 
 turbines and reciprocating en- 
 gines, 102 
 Combustible, influence of nature, 176 
 
 mixtures, 207 
 
 Combustion apparatus, Davey, 80 
 chamber, Boulton's, 22 
 cooling of, 152 
 details of, 208 
 dimensions of, 209 
 injection of steam into, 157 
 Lemale, 236 
 lining for, 242 
 for mixed turbine, 154 
 cycles without compression, iso- 
 
 pleric, 137 
 
 experiments with Davey appa- 
 ratus, 81 
 
 255 
 
256 
 
 INDEX. 
 
 Combustion, isobaric, 124, 144 
 
 nozzle, de Laval's, 25 
 
 temperatures, efficiencies for 
 
 various, 170 
 limitations of, 152 
 
 turbine, 230 
 
 diagram for, 231 
 
 under constant pressure, 110 ? 
 121 
 
 under constant volume, 110 
 Comparative efficiencies, 131 
 
 table of cycles, 75 
 Compound, efficiency, 92 
 Compression, adiabatic, 31, 45, 54, 
 123, 133 
 
 advantages of high, 57, 128, 148, 
 221 
 
 efficiency of adiabatic, 118 
 
 for gas turbines, 83 
 
 isothermal, 68, 127, 135 
 
 work of, 113 
 Compressions, low, 70 
 Compressor, efficiency of, 115 
 
 losses, 50 
 
 Rateau, 240 
 
 reciprocating, 6 
 
 rotary, 7, 83 
 
 Strnad, 201 
 
 tests, 201 
 Compressors, air, 51, 197 
 
 efficiency of, 91 
 
 efficiency of rotary, 88 
 
 high-speed, 202 
 
 piston, 197 
 
 rotary, 203 
 
 turbine, 93, 204 
 Computations for gas turbine, 213 
 
 for mixed turbine, 155 
 Conclusions from thermodynamic 
 
 study, 169 
 
 Constant pressure, combustion under, 
 110, 121 
 
 volume, combustion under, 110 
 Construction of gas turbines, details, 
 104, 197 
 
 Cool gases, injection of, 151, 164-167 
 Cooling of combustion chamber, 152 
 
 of gas turbines, 39, 101, 105 
 
 losses, 84 
 
 of turbine wheel, 239 
 
 water circulation of, 39 
 Creusot works, 246 
 Crompton, Lt. Col. R. E. B.: Dis- 
 cussion of Neilson Paper, 87 
 Curves of isobaric cycles, 147 
 Cycle, Carnot, 116 
 
 Diesel, 118 
 
 Ericsson, 142 
 Cycles, Clerk, 30-59 
 
 combination, 167 
 
 comparative table of, 75 
 
 curves of isobaric, 147 
 
 for explosion motors, 133 
 
 gas turbines, 108, 109 
 
 involving the injection of water, 
 151 
 
 using heat regenerators, 142 
 
 using isobaric introduction of 
 heat, 121, 144 
 
 using isopleric introduction of 
 heat, 133, 150 
 
 using isothermic introduction of 
 heat, 116 
 
 table of isobaric, 146 
 
 Davey combustion chamber, 80 
 
 Davey, Henry: Discussion of Neil- 
 son Paper, 80 
 
 De Laval gas turbine, 25 
 
 Delaporte, 190 
 
 Deschamps, J., 26 
 
 Details of gas turbine construction, 
 104, 197 
 
 Development methods for gas tur- 
 bines, 169 
 
 Diagram for combustion turbine, 231 
 explosion, 229 
 of nozzle sections, 182 
 of nozzle velocities, 182 
 
 Diesel cycle, 118 
 
INDEX. 
 
 257 
 
 Diesel motor, 31, 52, 53, 83, 132 
 Difficulties with the gas turbine, 83, 
 
 98 
 Discharge from nozzles, velocity of, 
 
 103, 180 
 Discs, efficiencies of revolving, 191 
 
 friction of revolving, 192 
 Discussion on Neilson Paper, 79-107 
 Dissociation in mixed turbines, 157 
 Divergent nozzles, 84, 88, 89 
 Dowson gas, 176 
 
 Economy curves, gas turbine, 234 
 Efficiencies, comparative, 131 
 compression, 99 
 expansion, 99 
 of mixed turbines, 158 
 of revolving discs, 191 
 for various combustion temper- 
 atures, 170 
 Efficiency of adiabatic compression, 
 
 118 
 
 of air compressors, 91, 115 
 compound, 92 
 of gas turbine, mechanical, 51, 
 
 109, 112 
 
 of gas turbines, probable, 169 
 practical, 75, 220 
 of rotary compressors, 88 
 in terms of temperature ratio, 114 
 thermal, 109 
 Elastic-fluid turbines, 19 
 Elements of the gas turbine problem, 
 
 108 
 Energy conversion in nozzles, 85, 180, 
 
 216 
 
 kinetic, 69 
 Engineering Congress at Lige, 26 
 
 magazine, 26, 175 
 
 Entropy- temperature diagrams, 31, 
 38, 41, 43, 45, 53, 58, 62, 65, 67, 72 
 Ericsson cycle, 142 
 Exhaust gases for operating turbines, 
 
 100 
 
 under reduced pressure, 139 
 17 
 
 Expansion, adiabatic, 35, 37, 48, 69, 
 
 70 
 
 of air in nozzles, 223 
 exponent of, 173-175 
 law of adiabatic, 112 
 in nozzles, free, 222 
 prolonged below atmospheric 
 
 pressure, 138 
 temperature, final, 109 
 limitations of, 163 
 Experimental investigations needed, 
 
 78 
 
 researches, 214 
 turbine, at Paris, 232 
 Experiments with trial turbine, 233 
 Explosion diagram, 229 
 motors, cycles for, 133 
 turbine, Karavodine, 247 
 applicability of, 171 
 turbines, 54, 56, 57, 171, 228, 
 
 247 
 
 Exponent of expansion, 173-175 
 variations of, 174 
 
 Fernihough, 13 
 
 Final section of nozzle, 183 
 
 Flame, propagation of, 33, 101, 107, 
 
 209 
 Flow of gas, formula for, 181 
 
 of gases through nozzles, 179 
 Fluids, working, 34 
 Formula of Saint Venant, 181 
 France, Discussion before the Society 
 of Civil Engineers, 108-221 
 
 Societe des Ingenieurs Civils de, 
 
 26 
 
 Free expansion in nozzles, 222 
 Friction of discs revolving in air, 192 
 
 losses in nozzles, 49, 188-190 
 Frictional losses, 50 
 Fuel, gaseous, 207 9 
 
 liquid, 208 
 
 solid, 207 
 Furnace gases, 176 
 Future of the gas turbine, 217 
 
258 
 
 INDEX. 
 
 Gardie producer, 207 
 Gas, Dowson, 176 
 engine, 30 
 
 flow through nozzles, 179 
 formula for flow of, 181 
 lean, 176 
 regeneration, 206 
 and steam turbines, 60-70, 153 
 turbine, applications of, 218 
 
 atomizer for, 239 
 
 Barber's, 11-13 
 
 Bucholz, 104 
 
 computations for, 213 
 
 cycles, 30, 108, 109 
 
 De Laval's, 25 
 
 economy curves, 234 
 
 future of, 217 
 
 general design of, 212 
 
 large, 244 
 
 mechanical efficiency of, 112 
 
 Patschke, 104 
 
 problem, elements of, 108 
 
 scheme of, 6, 7 
 turbines, applications of, 105 
 
 classification of, 191 
 
 cooling of, 101 
 
 details of construction, 197 
 
 losses in, 50 
 
 materials for, 104 
 
 method of development, 169 
 
 pressure limits in, 111 
 
 probable efficiency of, 169 
 
 regulation of, 194 
 
 scientific investigation into, 
 28-79 
 
 small, 76 
 
 for submarine torpedoes, 
 244 
 
 temperature limits in, 110 
 
 water circulation in, 39, 101, 
 
 105 
 
 Gaseous fuel, 207 
 mixtures, 177 
 Gases, furnace, 176 
 
 injection of cool, 151 
 
 Gases in nozzles, actual behavior of, 
 
 222 
 
 velocity of discharge, 69, 103 
 General design of gas turbine, 212 
 Governing of gas turbines, 194 
 Grashof, 187 
 
 Gross work, 41, 42, 46, 49, 51 
 Guide blades, 16 
 Gutermuth, Prof., 201 
 
 Hart, G., 26, 220 
 
 Heat balance for mixed turbine, 237 
 diagrams, 212 
 motors classified, 227 
 regenerators, cycles using, 142 
 specific, 112 
 from water jacket, utilization of, 
 
 59 
 High compression, advantages of, 57, 
 
 128, 148, 221 
 
 High-speed compressors, 202 
 Hot-air turbine, Burdin's, 15 
 Stolze's, 23, 24 
 
 Influence of nature of combustible, 
 
 176 
 
 of temperature limits, 109 
 of terminal pressure, 185 
 Inge"nieurs Civils de France, Socite 
 
 des, 26 
 
 Initial cost, 92 
 Injection of cool gases, 151 
 
 after expansion, 163 
 at high velocity, 166 
 at low velocities, 164 
 of regenerator steam, 157 
 of steam, cycles using, 151 
 
 after expansion, 163 
 of water, cycles using, 151 
 
 after expansion, 163 
 Institution of Mechanical Engineers, 
 
 28 
 
 Intercooler, 140 
 Introduction, 5 
 
 Investigations, programme for future, 
 215 
 
INDEX. 
 
 259 
 
 Isobaric combustion, 124, 144 
 cycles, curves of, 147 
 
 table of, 146 
 introduction of heat, cycles 
 
 using, 121, 144 
 Isopleric combustion cycles without 
 
 compression, 137 
 cycles, regeneration with, 150 
 introduction of heat, cycles using, 
 
 133 
 Isothermal compression, 68, 127, 135 
 
 cycles, partial, 121, 130 
 Isothermic introduction of heat, 
 cycles using, 116 
 
 Jets, velocities in steam, 48 
 Josse, Professor, 168 
 
 Karavodine turbine, 247 
 Kinetic energy, 47, 69, 85 
 
 Lame, 14, 19 
 
 Langen, Felix, 27 
 
 Laplace, 112 
 
 Large gas turbine, 244 
 
 Laval turbine, test of, 223, 224 
 
 Law of adiabatic expansion, 112 
 
 Laws of thermodynamics, 112 
 
 Lean gas, 176 
 
 Lechatelier, M., 112 
 
 Lemale, Charles, 26 
 
 Lemale combustion chamber, 236 
 
 Length of nozzle, 183 
 
 Leonardo da Vinci, 9 
 
 Letombe, L., 26, 221 
 
 Liege, engineering congress at, 26 
 
 Limitation of temperature of com- 
 bustion, 152 
 
 Limitations of expansion tempera- 
 ture, 163 
 
 Lining for combustion chamber, 242 
 
 Liquid fuels, 105, 177, 208 
 
 oxygen, turbines using, 162 
 
 London, W. J. A.: Discussion of 
 Neilson Paper, 100 
 
 Lorenz, 187 
 
 Losses in blades, 191 
 
 in gas turbines, 50 
 Low compressions, 70 
 Lucke, Charles E., 26, 28, 175, 222, 
 223 
 
 Martin, H. M.: Discussion of Neilson 
 
 Paper, 88 
 
 Materials for gas turbines, 93, 104 
 strength at high temperatures, 
 
 211 
 
 Maximum temperature, 110 
 Mechanical efficiency, 51, 109, 112 
 engineers, institution of, 28 
 features of turbines, 179 
 Mekarski, M., 197 
 Metals, action of heat on, 86 
 Mixed turbines, 60-70, 102, 235 
 analysis of, 156 
 combustion chamber for, 154 
 computations for, 155 
 dissociation in, 157 
 efficiencies of, 158 
 heat balance for, 237 
 table of, 160 
 Morin, 14, 19 
 
 Moss, Dr. Sanford A., 7, 26 
 Motor losses, 50 
 
 Multiple turbines, Tournaire's, 15 
 Murdock, 11 
 
 Nature of combustible, influence of, 
 
 176 
 Negative work, 36, 41, 42, 49, 51, 83, 
 
 85 
 
 Neilson, R. M., 26, 28-79, 93 
 Nozzle, final section of, 183 
 length of, 183 
 sections, diagram of, 182 
 
 ratio of, 184 
 
 velocities, diagram of, 182 
 velocity in neck of, 184 
 Nozzles, actual behavior of gases in, 
 222 
 
260 
 
 INDEX. 
 
 Nozzles, combination, 89 
 construction of, 210 
 diverging, 84, 89 
 energy delivered from, 180 
 
 of gases in, 216 
 expansion of air in, 223 
 experiments with, 85 
 flow of gases through, 179 
 free expansion in, 222 
 friction losses in, 188-190 
 oscillations in, 90, 95, 186-188 
 rotating, 49 
 shock in, 187 
 
 Stodola's experiments with, 88 
 temperature drop in, 222 
 
 measurements in, 215 
 velocities in, 85 
 velocity of discharge from, 180 
 
 Oscillations in nozzles, 90, 95 
 Otto cycle, 30 
 
 Oxidation of turbine blades, 211 
 Oxygen, turbines using liquid, 162 
 
 Parallel flow turbine, 56 
 
 Parsons's patent, 24 
 
 Parsons turbine, 33 
 
 Partial isothermal cycles, 121 
 
 Patschke turbine, 104 
 
 Piston compressors, 197 
 
 Poisson, 112 
 
 Poisson's law, 173 
 
 Poncelet, 14, 19 
 
 Power regulation of turbines, 77 
 
 Practical efficiency, 75 
 
 Prandtl, 187 
 
 Pressure limits in gas turbines, 111 
 
 volume diagrams, 31, 38, 40, 42, 
 44, 52, 58, 61, 63, 66, 71 
 
 waves in nozzles, 90, 186, 188 
 Probable efficiency of gas turbines, 169 
 Proell, 187 
 
 Programme for future investigations, 
 215 
 
 Prolonged expansion, 138 
 Propagation of flame, velocity of, 
 101, 107 
 
 Radiation losses, 50 
 Rateau air compressor, 240 
 Ratio of nozzle sections, 184 
 Rayleigh, 187 
 Reeve, Sidney A., 26 
 Reduced pressure exhaust, 139 
 Regeneration from gas to gas, 206 
 
 with isopleric cycles, 150 
 
 by steam, 206 
 
 of waste heat, 145 
 Regenerator, 70-73 
 
 action, table of, 149 
 
 advantages of, 147 
 
 cycles using heat, 142 
 
 design, 161 
 
 steam, injection of, 157 
 
 thermal, 205 
 
 Regulation of gas turbines, 77, 194 
 Revolving discs, efficiencies of, 191 
 
 friction of, 192 
 
 wheels, construction of, 211 
 Rey, Jean, 26, 219 
 Rich fuel with water injection, 162 
 Rotary compressors, 83, 203 
 Rotating nozzles, 49 
 
 Saint Venant, formula of, 181 
 Sautter, Harle and Co., 219 
 Scheme of gas turbine, 6, 7 
 Schweizerische Bauzeitung, 26 
 Scott, E. Kilburn: Discussion of 
 
 Neilson Paper, 103 
 Section of nozzle, final, 183 
 Seguier, 14 
 Sekutowicz, L., 26 
 Sekutowicz, L., Paper by, 108-218 
 Shock in nozzles, 187 
 Small gas turbines, 76 
 Smith, Robert H.: Discussion of 
 
 Neilson Paper, 90 
 Smoke jack, 9-11 
 
INDEX. 
 
 261 
 
 Societe des Ingeriieurs Civils de 
 
 France, 26 
 
 Societe des Turbomoteurs, 224, 245 
 Specific heat, 34, 112 
 Steam, cycles using injection of, 151 
 and gas turbine, 60-70, 153 
 injection after expansion, 163 
 
 of regenerator, 157 
 jets, velocities in, 48 
 regeneration by, 206 
 Stodola, A., 88, 164, 187, 189, 190 
 Stodola, experiments with divergent 
 
 nozzles, 88 
 
 Stodola's experiments with jets, 94 
 Stolze, hot air turbine, 23, 24 
 Strength of materials at high tem- 
 peratures, 211 
 Strnad compressor, 201 
 Structural difficulties with turbines, 
 
 220 
 Submarine motors, 163 
 
 torpedoes, gas turbines for, 24 
 Suction gas for turbines, 77 
 Sulphur dioxide turbine, 140 
 
 Table of mixed turbines, 160 
 
 of regeneration with isopleric 
 
 cycles, 151 
 
 Temperature of combustion, limita- 
 tions of, 152 
 
 drop in nozzles, 222 
 
 of expansion, 109 
 
 limitations of expansion, 163 
 
 limits, influence of, 109 
 
 maximum, 110 
 
 measurements in nozzles, 215 
 
 ratio in terms of efficiency, 114 
 Temperatures, efficiencies for various 
 combustion, 170 
 
 in gas turbines, 86 
 
 practicable, 33, 37, 39 
 Terminal pressure, influence of, 185 
 Tests of compressors, 201 
 
 of explosion turbine, 248 
 Thermal efficiency, 109 
 
 Thermal regenerators, 205 
 Thermodynamic study, 169 
 Thermodynamics, laws of, 112 
 Torpedoes, gas turbines for sub- 
 marine, 244 
 Tournaire, 14, 19 
 
 Trial turbine, experiments with, 233 
 Turbine, applications of the gas, 218 
 atomizer for gas, 239 
 combustion, 230 
 combustion chamber for mixed, 
 
 236 
 
 compressor at Bethume, 219 
 compressors, 93, 204, 219 
 computations for gas, 213 
 construction, material for, 93 
 economy curves of gas, 234 
 experiments with trial, 233 
 future of the gas, 217 
 general design of gas, 212 
 heat balance for mixed, 237 
 Karavodine explosion, 248 
 parallel flow, 56 
 at Paris, experimental, 232 
 practical efficiency qf, 220 
 waste heat, 167 
 wheels, construction of, 211 
 Turbines, computation for mixed, 
 
 155-158 
 
 efficiencies of mixed, 158 
 elastic fluid, 19 
 explosion, 54, 56, 57, 228 
 using gas and steam, 60-70 
 liquid fuel for, 106 
 using liquid oxygen, 162 
 mechanical features of, 179 
 mixed, 235 
 
 operated with exhaust gases, 100 
 for submarine torpedoes, 244 
 table of mixed, 160 
 uncooled, 55 
 Turbomoteurs, Socie"te* des, 26, 224, 245 
 
 Uncooled turbines, 33,55 
 Utilization of heat from water jacket, 
 59 
 
262 
 
 INDEX. 
 
 Velocities in nozzles, 85 
 
 in steam jets, 48 
 
 Velocity of discharge from nozzles, 
 103, 180 
 
 of gases, 69 
 
 in neck of nozzle, 184 
 
 of propagation of flame, 101, 
 
 107 
 
 Vermand, M., 173 
 Vinci, Leonardo da, 9 
 
 Waste heat recovery, 140 
 
 turbine, 167 
 Water circulation in gas turbines, 39, 
 
 101, 105 
 
 injection, cycles using, 151 
 after expansion, 163 
 with rich fuel, 162 
 
 Water jacket, utilization of heat from 
 
 59 
 siphons for gas turbines, 97 
 
 Waves in nozzles, pressure, 90, 186, 188 
 
 Weber, 187 
 
 Wegener, Richard, 27 
 
 Wheel, water cooling for, 239 
 
 Wheels, construction of, 211 
 
 Wilkins, Bishop, 9, 10 
 
 Windmill as a gas turbine, 9 
 
 Witz, M. A., 112 
 
 Work of compression, 113 
 gross, 41, 42, 46, 49, 51 
 negative, 36, 41, 42, 49, 51 
 
 Working fluids, 34 
 
 Zeitschrift fiir des Gesamte turbin- 
 enwesen, 27 
 
UNIVERSITY OF CALIFORNIA LIBRARY 
 BERKELEY 
 
 Return to desk from which borrowed. 
 This book is DUE on the last date stamped below 
 
 LD 21-100m-9,'48(B399sl6)476 
 
YC 33383 
 
 
 793334 
 
 TOY 
 
 Engineering 
 Library 
 
 UNIVERSITY OF CALIFORNIA LIBRARY