IGE & METHODS O^ANALYSIS of IRON AND STEEL COPYRIGHT 1ERICAN ROLLIN3 MIUU CO. MIDDLETOWN, OHIO 1920 PRESS OF THE GIBSON 8c PERIN CO. CINCINNATI, OHIO Research and Methods of Analysis IRON AND STEEL at ARMCO Second Edition Price $4.00 THE AMERICAN ROLLING MILL COMPANY MIDDLETOWN, OHIO 1920 Preface to Second Edition T HE first edition of tKis book appeared in 1912. The supply xtfas soon exhausted, and continued de- mands hav^e made it seem desirable to issue a second edition. The methods described are particularly adapted to the analysis of "Armco" products. Where -well-known methods have been de- scribed, \tfe have omitted details -which are -well understood by the skilled chemist. Where new" methods are described, we have entered into minute details. The second edition has been entirely rewritten, many methods have been added, and the entire scope amplified by the addition of new material. ARMCO RESEARCH MIDDLETOWN, OHIO 1920 444055 I a cti bfljy ' It II si ed .2 JH O O OJ II J cu o INTRODUCTION iVER since its inception, The American Rolling Mill Company has been a leader in the adapta- tion of science to practical problems of steel making. The company was organized to manufacture special grades of sheet metal suited to the needs of exacting users. A well-equipped research laboratory was, therefore, essential to the carrying out of its manufacturing program. In the establishment of such a laboratory, a new era was begun covering the manufacture of high grade iron and steel sheets and other metal products. In 1903 the development of various grades of electrical sheets for transformers, motors, generators, etc., was begun. Most satisfactory results were secured and research work on this most important line of manufacture has been con- tinued up to the present date. In 1908 for the first time in metallurgical history and contrary to established theories as laid down in authorita- tive metallurgical text books of the day, a commercially pure iron made in a modern open hearth furnace was pro- duced. This new metal was soon found to be superior to Bes- semer and Open Hearth Steel, in the essential properties of Purity, Rust-Resistance, Welding, Conductivity, and Enameling. Its metallurgical and practical development is unquestionably largely responsible for the improvement made to date in the rust-resisting qualities of the various grades of iron and steel sheets now being manufactured. In the early days, Chemistry had not in a majority of cases been applied to practical steel making beyond the determination of the elements known as "The Big Five" (Sulphur, Phosphorus, Carbon, Manganese, and Silicon). Today through research development, it is known that INTRODUCTION INTRODUCTION 7 gases such as Nitrogen and Hydrogen are elements to be considered in rust-resistance and other qualities of iron and steel. A close study of protective coatings is another branch of metallurgical research. Work along this line has been done to show the effect of impurities in the coating. It is a recognized fact that pure galvanized coatings are much more resistant to the elements of corrosion than are impure coatings, and experiments are being constantly carried on to further improve the quality and character of these coatings. Degasification of metal was not thought of a few years ago and yet today it is considered of very great importance. A modern research laboratory can now determine the gases in the metal, and w r ith this information, it is possible to maintain control of the gas content in the manufacture of the product. The Research Department of The American Rolling Mill Company was the first to make practicable, various methods for gas determination in Iron and Steel. New uses for pure iron and other special sheet metal products require a constant expansion of research work, covering such lines as vitreous enameling, japanning, weld- ing, heat treatment, forging, and casting, all of which offer wonderful fields of usefulness for the chemist and metal- lurgist. During the last ten years many grades of high polished sheets have been produced for the use of the automobile, furniture and other products. These sheets must not only have a very high finished surface, free from defects of every kind, but they must stand all sorts of drawing and spinning operations unknown to the maker and user of sheet metal just a few years ago. The modern research laboratory has been largely re- sponsible for all of these developments and it needs no prophet to foresee that many new alloys and other products, the result of special manufacture and treatment, will be developed from time to time to meet the exacting demands of industrial progress. INTRODUCTION The Iron Pillar at Delhi, India INTRODUCTION IRON PILLAR OF DELHI, INDIA, 1600 YEARS OLD Described by Sir Robert Hadfield In the 1912 Journal of the Iron and Steel Institute. He shows It to be pure Iron of the following analysis; Silicon .046 Sulphur .006 Phosphorus .114 Carbon .080 Manganese nil Copper, Etc. .034 Iron 99.720 10 INTRODUCTION In the study of corrosion problems it has been necessary to provide extensive proving grounds. This view shows one of the proving grounds where various metals are exposed to at- mospheric conditions ANCIENT IRONS AND MODERN RESEARCH |N the British Museum and elsewhere are many interesting specimens of age old irons that have resisted the "rust of time". Some of them have been taken from the tombs in the pyramids of Egypt. Their existence spans a period of 4000 years to the modern world in which steel and iron plays so large a part. For years, scientists have been seeking the secret of rust-resistance of iron and steel. They have analyzed such ancient irons as came to light for a complete understanding of their contents, they have studied their grain structure with the microscope, they have determined the gas content, and have taken into consideration the primitive methods of manufacture as compared with those of today. Out of all this has come the deductions of modern science, that is marking the pathway of progress. The American Rolling Mill Company has been an earnest investigator of the causes of corrosion. In the Research Department at Middletown is a museum of old and interesting nails and odd bits of old iron. The history together with the physical and microscopical analysis of each is carefully investigated and recorded in the archives, while the specimens are laid away under glass cases to awe the visitor with their antiquity. In fact, no sooner does an interesting example of the old iron come to light, than someone will suggest sending it to the Research Depart- ment of The American Rolling Mill Company at Middle- town, Ohio, for analysis. The study and analysis of these old irons has brought world wide recognition of Armco research work. Among the interesting old iron curios that have been sent to Middletown for analysis is a piece of iron cut from the "Merrimac," after having been in the water for more than one-half a century. It is historically interesting be- cause of the famous "Monitor and Merrimac" fight in Hampton Roads during the Civil War. 11 MCIENViRQNS AND MODERN RESEARCH HAND FORGED NAIL 145 YEARS OLD MADE BY INDIANS AND USED IN MISSION ANALYSIS Sulphur Phosphorus Carbon Manganese Copper Silicon Oxygen Nitrogen .005 .057 .015 .015 trace .048 .109 .006 MISSION SAN JUAN CAPISTRANO, CALIFORNIA, BEFORE THE EARTHQUAKE OF 1812 THE FIRST FOUNDING OF THIS MISSION TOOK PLACE Oct. 30, 1775. ANCIENT IRONS AND MODERN RESEARCH 13 Along side this is a collection of nails from the Wayside Inn at Sudbury, Mass. Nails from the famous Fairbanks homestead at Dedham, Mass., share their interest with nails from historic old missions of California. Coffin nails buried 100 years ago and still in a state of perfect preserva- tion have their own story to tell. In the corner of the room is still another notable ex- ample of rust-resisting iron the old iron links taken from the Newburyport bridge at Newburyport, Mass. Not- withstanding the fog and dampness of the New England seacoast, when the bridge was taken down in 1910, after 100 years service, the heavy "iron links were apparently as good as the day they were installed. The collection of interesting specimens is being added to every day. Recently an old iron nail was sent to the Research Department of The American Rolling Mill Com- pany, which was picked up out of the shell-torn ruins of the home of John Calvin, the great Reformist, at Noyon, France. The house was known to be at least 400 years old and yet the huge iron nail used in its construction showed no sign of corrosion. The Research department at Armco also prizes most highly specimens in its possession which were taken from the Pillar of Delhi, India the most notable example of old iron in the world today. This wonderful relic of worship dates back sixteen hundred years, and still stands today defying the elements and the "kisses" of worshipers, with no sign of disintegration by rust or corrosion. Over these specimens of old iron, the scientist w r orks like the etymologist over his insects, searching for hidden secrets. 14 ANCIENT IRONS AND MODERN RESEARCH IRON BAND TAKEN FROM A CANNON CAPTURED FROM THE BRITISH IN THE BATTLE OF TICONDAROGA. 1775-1777 TEST NO. 1036 FILE 51 ANALYSIS Sulphur .005 Phosphorus .069 Carbon .010 Manganese .010 Copper .080 Silicon trace Oxygen .063 RESEARCH ON CORROSION jURING the past few years a number of papers have been published by various investigators both in America and Europe, presenting the results of corrosion tests made under varying conditions of exposure, both in the labora- tory and in the field. The size of the test pieces in the hands of the separate investigators have ranged from small specimens which could be weighed on a chemical balance up to full size commercial sheets exposed to the natural wet and dry conditions of the outdoor atmosphere. Total and partial immersion tests in various media, with loss of weight during progressive corrosion, have been carefully recorded and plotted in the form of curves, and on these data in some cases very sweeping conclusions have been drawn. Nearly all the investigators have interested themselves in the possible beneficial or deleterious effect of more or less minute quantities of some special impurity or element, or in some grouping or variable combination of the usual constituents of commercial metals. Thus far all the experimenters have contended themselves with a discussion of the possible effect of solid constituents, without taking into account the possible effect of impurities of a gaseous nature, which are always more or less associated with iron and steel, such as carbon monoxide, carbon dioxide, oxygen, hydrogen, and nitrogen. It is self evident that the corrosion of iron and steel depends very largely upon the amount and character of surface exposed. It is obvious that metal rolled from spongy and porous steel, some of the blowholes in which extend to the surface, will be more susceptible to the action of water and oxygen, and also the nature of such rust will be different from that produced on sound metal. As unsound metal corrodes, a loose rust is produced on account of the tendency of the occluded gases to escape. A dense, sound metal, on the other hand, will form a dense, closely adherent rust, and the rate of progressive corrosion will be very materially reduced. The fact is not generally enough understood, that gases such as nitrogen, hydrogen, oxygen, and carbon monoxide may be associated with iron and steel in three separate and distinct ways. Gas may be present as an atmosphere enclosed in open blowholes, pipes or seams, 15 16 RESEARCH ON CORROSION Section through center of Steel Ingot, showing subcutaneous and deep seated blowholes. Section through center of "Armco" Ingot Iron, ingot, shoAving comparative freedom from blowholes. RESEARCH ON CORROSION 17 which may be large enough to see with the naked eye or so small as to require a microscope to resolve them. Again, free gas may be occluded among the grains or molecules of the iron, and be present in quantity even though no microscope is capable of discovering its presence. Finally, gases may be present in chemical combination with iron or manganese, or some other constituent, to form either dissolved or segregated nitrides, hydrides, oxides, or carbonyl com- pounds. That all three forms of gas inclusions affect the resistance of all metals to corrosion, there are strong apriori reasons for believing, even if there was no evidence as to fact in support of such a view. We have sawed ingots of iron and steel in two longitudinally, so that the soundness could be observed. The photographic cross sections of these experimental ingots are given. The atmospheric exposure tests which we are conducting, con- taining more than six hundred full size corrugated, 26-gauge sheets, first began to exhibit failures after about 13 months, exposure. The Bessemer Steel sheets have all failed in periods ranging from 13 to 34 months. We have now to consider why it is that Bessemer steels as a general rule rust more quickly than open hearth steels. As far as we are aware, no explanation of this frequently observed phenomenon has heretofore been advanced. Bessemer steel is made in America exclusively in acid lined converters, and, as this lining has no de- phosphorizing action, American Bessemer Steels usually run a little higher in phosphorus than the general run of basic open hearth steels. This is not, however, always true, and an analysis of a steel for its phosphorus content is by no means a definite proof of its method of manufacture. However, supposing that the general run of Bessemer steels made in America run somewhat higher in phosphorus than the general run of open hearth, is there any evidence that slightly higher phosphorus would account for any difference in corrosion resistance of the two types? As a matter of fact, phosphorus is one element of impurity in steel that has not at some time or another been held to be a prime factor in rapid corrosion. It seems to be very well assured that the tendency towards rapid corrosion, noted in the case of most Bessemer steels, must be due to some element of difference beside the usual slightly higher phosphorus content. 18 RESEARCH ON CORROSION IRON NAILS ABOUT 281 YEARS OLD TAKEN FROM FAIR BANK'S HOUSE. DEDHAM, MASS. TEST NO. 4541 FILE 1O3 ANALYSIS Sulphur .OO9 Phosphorus .005 Carbon .015 Manganese .020 Copper .01 6 Silicon .029 Oxygen .060 Nitrogen .005 Iron 99.841 RESEARCH ON CORROSION 19 Going back to the differences in method of manufacture of Bess- emer and open hearth metal, it at once occurs to us that the former type has blown through it, while it is molten and in the process of conversion, enormous quantities of air, more or less laden with mois- ture and gases. Of the two types of metal, therefore, Bessemer should be more prone to run a high gas content in the cooled and finished product than properly made open hearth steel. If this is true, the questions that at once arise are, can gas content be a factor in the relative corrosion resistance of Bessemer and open hearth metals, and further than this, is the self same factor an important one in considering relative corrosion resistance of different types of basic open hearth metal. Fortunately, there is some recent scientific data of a most in- teresting nature, which shows in a most conclusive manner what an enormous quantity of gases associate themselves with some types of commercial steel and modify or qualify their physical characteristics. In 1914, L. Baraduc Miller published in full in the Carnegie Scholarship Memoirs of the British Iron & Steel Institute a report of progress on the study of gases occluded in liquid steel. The steel which was made by the basic Bessemer process was cast into 1000 Ib. ingots and allowed to cool in a vacuum, while other ingots from the same heat w r ere cooled under atmospheric pressure in the usual manner generally practiced in metallurgical operations. One ingot of the vacuum treated steel yields 1159.8 liters of gas measured at atmospheric temperature and pressure. This gas on analysis was found to have the chemical and volume composition as shown in the following table: Per Gross Volume Per Cent Per Ton of Steel Carbon Dioxide Litres 42.2 3.6 Litres 76.7 Oxygen 10 6 9 19 3 Carbon Monoxide 352 2 30 5 640 3 Hydrogen . 604.3 52.2 1098.7 Methane 2.4 0.2 4.3 Nitrogen ... 147.7 12.7 268.5 This table shows the astonishing fact that more than 2100 liters of gas, or about 75 cubic feet per ton of steel was liberated from the vacuum treated metal. It is also shown that the principal gases, hydrogen, carbon monoxide and nitrogen, play an important role in 20 RESEARCH ON CORROSION MiCROSTRUCTURE OF LONGITUDINAL SECTION IRON CHISEL OF ANCIENT ORIGIN ABOUT 1400 YEARS OLD Described by Sir Robert Hadfieid in the 1912 "Journal of The Iron and Steel Institute." RESEARCH ON CORROSION 21 the metallurgy of steel. In another table Baraduc Miller gives the following data to show the physical character of the vacuum treated steel compared with metal from the same ladle poured and cooled in the usual manner: Test piece of metal not treated for gases, taken from Y of the 90x90 Mms. bars. Normal cooling and no heat treatment Test piece of metal treated for gases taken from a rolled round 25 millimeters in diameter. Nor- mal cooling and no heat treatment Breaking Strain Elastic Limit 40.5 Tons 27 Tons 44. 76 Tons 35 5 Tons Elongation Per Cent 28.9 Tons 24 4 Tons Hardness 112 124 Resilience 17] 31 \ Fibrous 29] 36 \ Fibrous Average 19 | Fracture 25 J 23 34 1 Fracture 34 j 33.2 Here we see the ultimate strength, elastic limit, harndess (density), and resistance to fibrous fracture all materially improved in the self same metal by the simple elimination of occluded gases. The ques- tion as to whether resistance to corrosion is also improved is not considered. Baraduc Miller recognizes the fact that the large volume of hydrogen eliminated was furnished by the decomposition of the moisture of the air used in blowing, during the conversion. To what extent this is a factor in the manufacture of open hearth steel, is uncertain, but that it is an ever present factor in the case of Bessemer steels, there can be no doubt. Baraduc Miller includes the following interesting discussion of the hydrogen content : "In the case of hydrogen, it is seen that the 5150 cubic meters of air injected into the converter contained, owing to the 5.672 grammes of water per cubic meter, 29,210.8 grammes of water capable of yielding, by complete dissociation, 3245.6 grammes of hydrogen, corresponding, at about 15 and under a pressure of 760 millimeters, with a volume of 36,058.6 liters. Now with 1098.7 liters of hydrogen given per ton of steel, measured at the ordinary temperature, the entire cast must have contained, at a given moment, 13,860.3 liters of hydrogen. It results, therefore, that the maximum amount of hydrogen 22 RESEARCH ON CORROSION M1CROSTRUCTURE OF TRANSVERSE SECTION IRON NAILS OF ANCIENT ORIGIN ABOUT 1400 YEARS OLD Described by Sir Robert Hadfield in the 1912 "Journal of The Iron and Steel Institute/' RESEARCH ON CORROSION 23 fixed, at any rate momentarily, by the steel in some form or other, dissolved or in combination, must have been 13,860.5 x 100_ -38.5 per cent of the total volume of the available hydrogen. This throws an interesting insight into the extreme solubility of the gases, and in particular of the hydrogen, in liquid steel at a high temperature. It remains to ascertain if these gases are actually in solution or in combination, and also what is left of these gases in the steels at the moment of solidification in the ordinary conditions of manufacture." The last sentence of this interesting quotation contains really the gist of the whole matter and applies quite as much to the question of nitrogen as to hydrogen. It has been pointed out in an earlier paragraph that gases may be held in steel either in a state of chemical combination or merely dissolved or occluded. Up to the present stage of progress of our studies of the effect of gaseous impurities on the corrosion resistance of iron and steel, all the evidence so far in hand seems to point to nitrogen as a more im- portant factor in respect to corrosion problems than the other gases under consideration. A vast number of analyses of all types of iron and steel have shown very significant differences in nitrogen content when taken in connection with the known behavior of certain metals under test and service. In the discussion of the possible effect of nitrogen on the corro- sion of metal, it must be borne in mind that the method used deter- mines only the fixed or nitride nitrogen that is present, and tells us nothing about occluded or dissolved nitrogen. It must be under- stood, therefore, that such data as is here presented has to do only with the nitrogen which is held in the steel in combination as nitride. Pure iron nitride has the theoretical composition Fe2N, con- taining 11.8 per cent nitrogen. Commercial irons and steels as re- ported by a number of authorities and confirmed by numberless analyses which we have made, vary in nitride nitrogen content from traces in well made open hearth metals up to approximately .050 per cent in some Bessemer steels. Although there is an abundant literature on the effect of gases and particularly nitrogen, on steel, there is very little evidence in regard to the quantity of free gas occluded in different types of metal. 24 RESEARCH ON CORROSION MfCROSTRUGTURE OF TRANSVERSE SECTION IRON BILLHOOK OF ANCIENT ORIGIN ABOUT 1400 YEARS OLD Described by Sir Robert Hadfield in the 1912 ^Journal of The Iron and Steel Institute." RESEARCH ON CORROSION 25 Austin electrically melted a number of samples of steel in a vacuum, collected and analyzed the evolved gas. One sample of mild open hearth yielded about one cubic centimeter of gas per gram, equal to about 35 cubic feet per ton of metal. The gas in Austin's experi- ments was pumped out in three stages, each of one half hour's dura- tion. The electric current was first adjusted to raise the temperature of the test bar enclosed in a water cooled steel vacuum chamber to 900C. The temperature was next increased toabo ut 1100C, and finally in the third stage the bar was melted. The first extracts of gas from the mild open hearth steel analyzed as follows: Per Cent Carbon Dioxide 7.7 Carbon Monoxide 18. 4 Hydrogen 59. 1 Nitrogen 14. 8 Two other analyses of gas evolved during the last stage in the heating yielded : I II Per Cent Per Cent Carbon Dioxide 1.4 1.5 Carbon Monoxide 65. 7 59. 4 Hydrogen and Nitrogen.. . 32. 9 39. 5 The analysis of the open hearth steel given by Austin is: Carbon 09 Silicon 05 Phosphorus 05 Sulphur 05 Manganese .75 Another medium steel which evolved an even greater quantity of gas (1.35 cc. per gram) had the analysis: Carbon 49 Silicon 35 Phosphorus 02 Sulphur 02 Manganese 72 It is interesting to compare these results with those obtained by Baraduc Miller who, as described above, cooled basic Bessemer Steel in a vacuum. 26 RESEARCH ON CORROSION op to SS O 00 en a 3* Hi RESEARCH ON CORROSION 27 Baraduc Miller obtained 75 cubic feet of gas per ton of metal, while Austin obtained in the case of the open hearth steel 35 cubic feet per ton. In the first case, however, it is presumable that some part of the gas collected would have escaped into the air, had it been allowed to cool in a normal manner. In Austin's work, however, he started with a finished bar of cold steel, and it is difficult to escape the conclusion that all the gas evolved was actually held in some form or another in the body of the metal. That the gas was present in blowholes, is impossible, when we consider the volume relations which follow from Austin's experi- ments. Austin collected gas which, calculated to a tonnage basis, amounted, as has been shown, to about 35 cubic feet per ton of metal. Now, since a cubic foot of steel weight 480 Ibs., there must have been a volume of gas present equal to approximately 7.5 cubic feet to each cubic foot of metal. This is a most extraordinary conclusion and one that will perhaps be unwelcome information to many steel users. If the physical properties of steels of various types are affected by the nature and quantity of the occluded gases which they contained, it seems fair to inquire whether it might not be true that these same factors might exert an important influence on the much dis- cussed question of corrosion resistance in relation to the chemical constitution of iron and steel. The alleged good or bad effect of minute differences in the per- centage composition of the metal has been so much discussed and argued about by a great number of investigators, that there is little left to be said or claimed in regard to the influence of these solid constituents. The effect of gas content has, however, been curiously overlooked in the discussion of corrosion problems heretofore, and yet it is proba- ble that this one factor is the most important of all with relation to all the commercial metals, no matter whether we are considering a steel roofing sheet or a brass or bronze condenser tube. Of the atmospheric corrosion tests consisting of many hundred full sized corrugated sheets, which have been described, exactly thir- teen sheets out of about six hundred of the same age were found to have failed in periods ranging from 13 months to three years. It at once occurred to us that it would be interesting to select samples from ten of these failed sheets, and, after cleaning off the adherent rust with hydrochloric acid, have nitrogen and carbon monoxide deter- minations made, which would then be compared with similar analyses made from ten sheets taken at random, that were still in excellent 28 RESEARCH ON CORROSION PURE IRON NAILS TAKEN FROM GRAVE OF SOLDIER AFTER BEING BURIED FOR 100 YEARS IN FORT ST. CLAIR CEMETERY, EATON, OHIO. ANALYSIS OF NAILS Sulphur Phosphorus Carbon Manganese Copper Silicon Oxygen Hydrogen Nitrogen Iron .007 .100 .020 .004 trace trace .031 trace .008 99.830 RESEARCH ON CORROSION 29 condition on the test rack. Fortunately, in some instances strips of the original test sheets as they were before exposure had been pre- served so that it was possible also to determine if any change of chem- ical composition of the base metal had occurred during the three years of outdoor exposure. This work was immediately put under way, with the following results: GOOD SHEETS BAD SHEETS Sheet No. Carbon Monoxide Nitrogen Sheet No. Carbon Monoxide Nitrogen 250 .003 .007 88 .011 .014 241 .005 .006 91 .016 .015 266 .006 .004 529 .012 .032 252 .003 .003 626 .016 .016 337 .009 .005 625 .024 .021 290 .016 .005 530 .014 .017 336 .011 .003 623 .013 .011 662 .019 .004 624 .021 .006 299 .014 .003 531 .015 .019 435 .012 .003 101 .027 .006 The bad sheets which failed, referred to in the above table, had all been purchased in the open market as examples of typical Bessemer steels, but the present point of interest in connection with them is the fact that they showed by analysis from two to five times the con- tent of nitride nitrogen shown by the companion sheets which had not failed. Taken in conjunction with the great difference in lasting quality of the two sets of sheets, the results are suggestive and significant. In the course of our experience a number of cases had been brought to our attention, in which steel in various kinds of service had failed in an extraordinary manner and in comparatively brief periods of time. Previous chemical analyses that had been made for the usual solid impurities, as well as microscopic examinations had not as usual revealed the reason for the sudden corrosion of the materials. In view of the significance of the results shown in the above table, how- ever, it became a matter of interest to see what a complete analysis, including all gases, would indicate. The first case was that of a steel pipe which was perforated with holes after being buried for 14 months in an alkali soil. The second case is that of some railroad spikes that had corroded in an unusual and dangerous manner after being in use only a few years. An illustration of these failures is given. The results of the analysis of these failures is given in the following Table and for the purpose of comparison the analysis of a very pure open hearth iron is also given : 30 RESEARCH ON CORROSION FAILURE OF STEEL PIPE AFTER 14 MONTHS IN ALKALI SOIL DUE TO HIGH NITROGEN ANALYSIS Silicon trace Sulphur 049 Phosphorus 098 Carbon 090 Manganese 357 Copper 014 Oxygen 032 Hydrogen trace Nitrogen 041 Iron . . .99.319 RESEARCH ON CORROSION 31 COMPARATIVE ANALYSIS OF STEEL AND ARMCO INGOT IRON Pipe Spike Armco Failure Failure Ingot Iron Carbon 090 .100 .010 Manganese 357 .405 .025 Phosphorus 098 .108 .004 Sulphur 049 .041 .025 Silicon trace trace trace Copper 014 trace .040 Nitrogen .041 .033 .004 Oxygen 032 .082 .015 Carbon Monoxide 052 .058 .015 Carbon Dioxide trace trace trace Hydrogen trace trace trace Other instances like the above, in which rapid failure has been accompanied by high nitride nitrogen content have been found, but as the present intention is rather to inquire into the subject than to arrive at final conclusions, a slightly different phase of the subject will now be referred to. Among the examples of Bessemer steels included in the atmos- pheric exposure tests above referred to, Sheet No. 531 of Bessemer steel was perforated with numerous rust holes in 13 months, as was also sheet No. 88. Sheet No. 624 also of Bessemer steel did not fail until 33 months of exposure. The complete analyses of these three samples is given in the following table: Good Bessemer Poor Bessemer Poor Bessemer Sheet No. 624 Sheet No. 88 Sheet 531 Carbon .010 .030 .015 Manganese .480 .400 .310 Phosphorus .082 .064 .060 Sulphur .044 .031 .063 Copper trace .028 .010 Carbon Monoxide .021 .011 .015 Oxygen .055 .031 .071 Nitrogen. .006 .014 .019 In view of the great difference in lasting quality under atmos- pheric corrosion, these analyses are a'gain suggestive. The nitride nitrogen of the two poor sheets is, however, from 2 to 3 times higher in the poor sheets than in the good sheet. It has frequently been noted by a number of investigators that sheets that fail rapidly under atmospheric corrosion tests cover them- selves with a much lighter colored and more loosely adherent rust than the more resistant sheets. We heated small samples of iron and steel in an atmosphere of ammonia gas, and thereby raised the content of nitride nitrogen. 32 RESEARCH ON CORROSION FAILURE OF STEEL JRAILROAD SPIKES AFTER A FEW YEARS DUE TO HIGH NITROGEN ANALYSIS Silicon .... Sulphur. . . . Phosphorus Carbon. . . . Manganese Copper. . . . Oxygen . . . Hydrogen . Nitrogen Iron . . trace . .041 . .108 . .100 . .405 . trace . .082 trace 033 .99.231 RESEARCH OX CORROSION 33 In every case it has been found that with increased nitrogen content the attack of dilute acid on the specimen is stimulated. Two samples of pure iron were cut from the same sheet containing a normal con- tent of nitride nitrogen of .003 per cent. One of these samples was heated in ammonia gas until the nitrogen had risen to .127 per cent. The two samples w^ere then suspended side by side in 25 per cent sulphuric acid. The nitrified sample was destroyed in 6 days, while the untreated sample lasted 11 days. Unfortunately, no rational corrosion test is known, so that it has been impossible up to this time to test the corrosion resistance under atmospheric or various forms of service exposure, of test pieces that have had the nitrogen content increased by artificial means. In the attempt to find an explanation of why high fixed nitrogen and high gas content stimulate rapid corrosion, we feel justified in stating that there is a considerable body of evidence which points in the direction of combined and occluded gas as being a very important factor in the corrosion resistance of various types of metal. We have in our collections numerous examples of the old, hand forged, ancient irons that have shown very extraordinary resistance to corrosion under exposure, lasting in some cases for many centuries. It has always been a mystery, what gave these old irons their dura- bility. With the repeated reheatings and hammering that these hand forged metals underwent, it is reasonable to suppose that they were wonderfully densified and degasified. They have always shown themselves to contain very low percentages of nitrogen. 34 RESEARCH ON CORROSION LJ U U C.) ID O O O < < n CM co - ct tr o o o ^ h- i < J O 01 O [fl - !f) CM ID O CM CM - sr -- n O O O O N CD I 3 U cr on no u 3 I Z Z I ff O < CL (/) m o j o o: z 3 I < ' < (/) CL (^ Z n CD o Q < o n o r^ h rr CD a o - n h ^r U M U Z I il O < Q_ in 03 CD j o or z 3 I < < tn o 3 I Q_ U I GO' U MAGNETIC TESTING 36 MAGNETIC TESTING OF ELECTRIC SHEETS MAGNETIC TESTING OF ELECTRIC SHEETS. |LECTRICAL sheet steel products are today bought and sold on specification requiring careful tests for magnetic properties, both by the producer and consumer. If the consumer does not have the facilities for magnetic testing he must rely upon the producer, and in this case it is well for him to know in detail what methods are being used. It is for the special benefit of this class of customers that we in- clude the following paragraphs in this book. Others who are more ad- vanced in magnetic testing w T ill find the information useful in compar- ing results. CORE LOSS TEST The power consumption in electrical sheet steel, when subjected to an alternating magnetization is known as the core loss. The stand- ard core loss is the total power in w^atts consumed in each kilogram of material at a temperature of 25 degrees C when subjected to a har- monically varying induction, having a maximum of 10,000 gausses and a frequency of 60 cycles per second. It is represented by the symbol W 10/60. The method of test which we use is an outgrowth of that orig- inally used by Professor Epstein and adopted by The American Society for Testing Materials in 1911. A photograph shows the arrange- ment of coils and sample pieces. The magnetic circuit consists of 10 kg. (22 Ibs.) of the test materi- al, cut with a sharp shear into strips 50 cm. (19 11/16 ins.) long and 3 cm. (1 3-16 ins.) wide, half parallel and half at right angles to the direction of rolling, made up into four equal bundles, two containing material parallel and two containing material at right angles to the direction of rolling, and finally built into the four sides of a square with butt joints and opposite sides consisting of material cut in the same manner. No insulation other than the natural scale of the ma- terial (except in the case of scale-free material) shall be used between laminations, but the corner joints shall be separated by tough paper 0.01 cm. (0.004 in.) thick. 37 38 MAGNETIC TESTING OF ELECTRIC SHEETS TESTING COILS Epstein Core loss tests MAGNETIC TESTING OF ELECTRIC SHEETS 39 The magnetizing winding consists of four solenoids surrounding the four sides of the magnetic circuit and joined in series. A sec- ondary coil is used for energizing the voltmeter and the potential coil of the wattmeter. These solenoids are wound on a form of any non-magnetic non- conducting material of the following dimensions: Inside cross-section .................... 4 by 4 cm. Thickness of wall ................ not over 0.3 cm. Winding length .......................... 42 cm. The primary winding on each solenoid consists of 150 turns of copper wire uniformly wound over the 42 cm. length. The total re- sistance of the magnetizing winding is between 0.3 and 0.5 ohm. The secondary winding of 150 turns of copper wire on each solenoid is similarly wound beneath the primary winding. Its resistance should not exceed 1 ohm. A voltmeter and the voltage coil of a wattmeter are connected in a parallel to the terminals of the secondary winding of the apparatus. The current coil of the wattmeter is connected in series with the pri- mary winding. A sine wave electromotive force is applied to the primary wind- ing and adjusted until the voltage of the secondary circuit is given by the equation: 4fNnBM 41D108 in which f =form factor of primary E. M. F. =1.11 for sine wave N = number of secondary turns = 600 n number of cycles per second = 60 B = maximum induction =10,000 M = total mass in grams = 10,000 1 = length of strips in centimeters =50 D = specific gravity = 7. 5 for high-resistance steel 7.7 for low-resistance steel E = 106.6 volts for high-resistance steel for sine voltage. = 103.8 volts for low-resistance steel for sine voltage. A specific gravity of 7.5 is assumed for all steels having a resis- tance of over 2 ohms per meter-gram, and 7.7 for all steels having a 40 MAGNETIC TESTING OF ELECTRIC SHEETS O be MAGNETIC TESTING OF ELECTRIC SHEETS 41 resistance of less than 2 ohms per meter-gram. These steels are desig- nated as high and low resistance steels, respectively. The wattmeter gives the power consumed in the iron and the secondary circuit. The loss in the secondary circuit is given in terms of the total resistance and voltage. Subtracting this correction term from the total power gives the net power consumed in the steel as hysteresis and eddy current loss. Dividing this value by ten gives the core loss in w^atts per kilogram. Sampling From each annealing box of material we take sample sheets from top, center and bottom. Each sample repre- senting not over 10,000 Ibs. of material. If there is more than one heat of steel in the box the samples are so taken to represent each heat. All samples are recorded by box number, position of lot and heat number. Procedure 1. Cut the test material into strips 3 x 50 cm., half parallel and half at right angles to the direction of rolling. 2. Place on the balance a pile of strips weighing 2.5 kg. Add a second pile of the same kind, bringing the weight up to 5 kg. In each case the weight is taken to the nearest strip. Add in succession two piles of 2.5 kg. each, of the other kind of strips, bringing the weight up to 7.5 kg. and 10 kg. respectively. 3. Secure each bundle by string or tape (not wire) and insert in the apparatus as indicated. 4. Apply the alternating voltage to the primary coil and tap the joints together until the current has a minimum value, as shown by an ammeter in series. Then clamp the corners firmly by some suitable device. 5. Shunt the ammeter and adjust the primary current until the voltmeter indicates the proper value. This adjustment may be made by an auto-transformer, by varying the field of the alternator, or by both, but not by the insertion of resistance or inductance in the pri- mary circuit. Simultaneously the frequency must be adjusted to 60 cycles. 6. Read the wattmeter. 7. Calculations. Subtract from the wattmeter reading the in- strument losses, which will be constant for any set of instruments and voltage, and divide by 10. The result is the standard core loss. 42 MAGNETIC TESTING OF ELECTRIC SHEETS . MAGNETIC TESTING OF ELECTRIC SHEETS 43 AGING TESTS Sheet steel after annealing is peculiar in its magnetic behavior and tends to go back to the unannealed state when continuously sub- jected to temperatures somewhat above normal room temperature. This increase in core loss is known as aging and is generally expressed as an aging coefficient in percent based on the initial core loss. The standard test is based on a heating at 100 degrees C for 600 hours in a constant temperature oven such as shown in the accompany- ing photograph. Core loss tests to be made before and after the heating period. PERMEABILITY TESTS The above title is used in general to cover tests necessary to ob- tain data for normal induction curves or magnetization and satura- tion curves as they are sometimes called. From this data permea- bility values may be calculated and actual permeability curves plotted. For this test we use the Burrows compensated double yoke method (described in the standard Hand-book for Electrical Engineers and also in Technical Paper No. 117 of the Bureau of Standards.) A photograph shows the apparatus as used in our laboratory. The normal magnetic induction is the induction produced by a magnetizing force in a given piece of magnetic material which has been previously demagnetized and then subjected to many reversals of the given magnetizing force. Both the induction B and the magnetization force H is expressed in terms of the C. G. S. electromagnetic unit (gauss). The test material consist of 5 kg. of the strips cut as indicated for the standard core loss test. The magnetic circuit is a rectangle having the test material for one pair of opposite sides, and the same or different material for the other pair, which may be shorter. The joints at each corner are alternately butt and lap, or may be clamped on the edges. The magnetomotive force is applied in two sections. The main magnetizing coils consists of two equal and uniformly wound sole- noids surrounding the test material. The compensating coils con- sist of two solenoids surrounding the yoke strips. The test coil surrounds the middle portion of each bundle of test material. Four other test coils of half the number of turns are 44 MAGNETIC TESTING OF ELECTRIC SHEETS a g 'I o I MAGNETIC TESTING OF ELECTRIC SHEETS 45 placed over the test material, approximately midway between the yokes and the center. The two center test coils are joined in series and the four end test coils are joined in series. The corresponding ballistic deflections, due to these two test coils, are measures of the magnetic fluxes through the underlying portions of the magnetic cir- cuit. By connecting the two test coils so that the induced electro- motive forces oppose each other, and adjusting the current through the compensating magnetizing coils so that there is no resulting ballis- tic deflection, an approximate uniformity of flux is secured through the greater portion of the test material, and the induction may be measured ballistically in the regular manner. The magnetizing force when the flux is adjusted to uniformity is that calculated from the uniform winding of the main magnetizing solenoids. The cross-section of the magnetic circuit is determined as in the standard core loss test. For curve work we obtain magnetizing force or H values corres- ponding to induction values "B" of from 2000 to 20,000 gausses by steps of 2000. For obtaining permeability values at low and high inductions we determine "H" values for three values of "B" namely 6000, 10,000 and 16,000 gausses. For routine commercial testing we find the above values very satisfactory for checking and comparisons. METALLURGICAL CONTROL MICROSCOPICAL AND PHYSICAL TESTING 48 METALLURGICAL CONTROL METALLURGICAL CONTROL |N addition to careful and constant chemical control the most exacting metallurgical supervision is necessary in the manufacture of "Armco" products. This supervision begins with the raw materials and follows through every operation to the finished product. A constant watch is kept upon the quality of the raw materials to insure that they are suitable to enter into the manufacture of Armco quality materials. In the Open Hearth Department in addition to the vigilant chemical control, a record is kept of the pouring temperatures of all heats and the greatest care is used to maintain uniform pouring tem- peratures. These temperatures are measured by means of optical pyrometers which will be described later. In the Blooming, Bar and Sheet Mills every care and precaution is used to see that the materials are fabricated in a manner to produce the highest quality products. The conditions of manufacture are watched continuously by the Operating and Research Departments to see that they remain con- stant and do not vary from Armco standards. In the Annealing De- partment where improper heat treatment can so readily produce un- desirable properties in the sheets, a complete system of thermoelectric pyrometers is installed to record and to facilitate the control of an- nealing temperatures. In the Galvanizing Department thermoelectric pyrometers are also installed to safeguard the uniformity of manu- facturing operations and thereby to insure the maintenance of Armco quality. In the Finishing Department the quality of the product is carefully tested and proved by microscopic and ductility tests as well as by thorough visual inspection. In fact throughout the whole mill every manufacturing operation is surrounded with every safeguard which experience and science can devise to insure the greatest uniformity and the highest quality in the finished products. 49 50 METALLURGICAL CONTROL METALLURGICAL CONTROL 51 SCIENTIFIC HEAT TREATMENT It is probable that the greatest advance which has been made in the metallurgy of iron and steel in recent years is the development of heat treatment upon a scientific basis. It has not been a great improvement in the art made at a single step by a single invention or discovery, as some of the improvements in the past, such as the Bessemer converter and the Siemens regenerative furnace. But scientific heat treatment has been a gradual development, the result of much painstaking investigation and research upon the part of many metallurgists. The fuller understanding of its principles and the broader application of them has resulted in higher quality in iron and steel products than was formerly thought possible. Heat treatment to the metallurgist means the application of heat or rather the manipulation of heat for the purpose of producing desired physical or structural properties in the material treated. Furthermore heat treatment is usually applied to finished or semi- finished material. In the forging or rolling of metal, stresses and strains are set up and only by heat treatment can they be removed. This heat treatment when done at the proper time and temperature, produces physical changes in the metal giving it the desired proper- ties. In the sheet metal industry this heat treatment is called annealing. It has been found that the length of time and the uniformity of the temperature are important factors. Improper annealing can ruin a sheet of metal for the purpose intended. It can make a soft sheet extremely brittle, yet re-annealing this material under proper con- ditions will bring it back again as ductile as before. Especially in meeting the demands for a high finished or glossy sheet with deep drawing properties as demanded for automobile bodies and fenders, heat treatment has made a most important contribution to the industries of today. The presence of strains in the sheets due to the rolling is also a contributing factor towards rapid corrosion and it has been found necessary to relieve such strains before coating as in the case of galvanized sheets which are to be used under exposure to the elements. The narrow range between good and bad heat treatment can best be controlled by the use of metallurgical microscopes which reveal the grain structure for examination. Microscropic exam- ination serves as a guide in obtaining a standard and uniform heat 52 METALLURGICAL CONTROL METALLURGICAL CONTROL 53 treatment in the manufacture of the product. An example of the results of proper and improper annealing as detected by the micro- scope is shown on page 54. Both micrographs were taken from the same piece of steel; one after having been improperly annealed, and the other after having been properly annealed. The elongated grains and the strained condition of the metal is readily seen in the improperly annealed sample, and may be compared with the well- rounded equiaxed grains in the properly annealed sample. Due to careful scientific research the organization of The American Rolling Mill Company has been able to produce many specialties in the iron and steel line. Notable among these are American Ingot Iron, Enameling Sheets, Automobile Sheets, Deep Drawing Sheets and sheets for electrical machinery, such as, trans- formers, motors and generators. In addition, billets are supplied from which wire is made with high electrical conductivity for use in telegraph and telephone lines. Billets are also furnished for the manufacture of welding rods and wire for the Electric or Acetylene Gas Welding of iron and steel plates or castings. METALLURGICAL CONTROL 2, and in the filtrate the Al is again precipitated as phosphate as before. 2 It is generally impossible to decompose all the A1 2 3 by a sodium carbonate fusion alone. ARSENIC 77 DETERMINATION OF ARSENIC DISTILLATION METHOD Very good results can be obtained by the Distillation Method in determining arsenic in iron and steel. The methods used in the determination of this element require the use of either ferrous sulphate or cuprous chloride. We prefer the latter. The method used in our laboratory has been carefully worked out and has been found to give very satisfactory results. The essential details being as follows: Dissolve 20 grams of drillings in a 6 inch diameter casserole, using 300 cc. of nitric acid (1.20 Sp. Gr.). Heat slowly in order to prevent the reaction from becoming too violent. Evaporate to dryness on hot-plate and bake until no more nitrous fumes are evolved. Remove from hot plate and allow to cool, transfer the ferric oxide to distillation flask to which is attached a 50 cc. pipette which dips into a beaker con- taining 200 cc of. distilled water. Place in distillation flask 40 grams of cuprous chloride and 300 cc. of concentrated hydrochloric acid, boil until about two-thirds of the hydrochloric acid has distilled over. From the beginning of the distillation pass hydrogen sulphide gas into the distillate while same is being heated almost to the boiling point with the use of a small electric hot-plate. This will insure a rapid pre- cipitation of arsenious sulphide which settles readily and can be easily filtered. After the distillation is discontinued remove the beaker containing the distillate from the source of heat, dilute to 500 cc. with cold water and continue to pass hydrogen sulphide gas into the solution until cold. Disconnect pipette and rinse inside and outside with two or three cc. of concentrated ammonia, allowing washings to run into beaker containing arsenious sulphide. Allow the precipitate to settle and filter on asbestos using a Gooch crucible, w^ash with distilled water until free from acid. Trans- fer the asbestos felt containing the arsenious sulphide to a 250 cc. beaker. Add 10 cc. of fuming nitric acid or 10 cc. of nitric acid (1.42 Sp. Gr.) and 1 gram of potassium chlorate. Evaporate to dryness. Dissolve the arsenic acid in dilute hydrochloric acid, filter and wash with hot distilled water. Concentrate to about 20 cc., heat to boiling 78 ARSENIC and add 10 cc. of magnesia mixture, and 10 cc. of ammonia (0.95 Sp.Gr.) Continue to boil for 15 minutes, remove from the source of heat, add 20 cc. of alcohol and let stand for 5 hours. Filter and wash the pre- cipitate on a weighed Gooch crucible, dry, ignite and weigh as mag- nesium pyroarsenate which contains 48.27% arsenic. Magnesia Mixture Dissolve 110 grams of crystallized magnesium chloride in a small amount of water. Add 140 grams of ammonium chloride, make dis- tinctly ammoniacal with ammonia, and dilute to 2000 cc. with dis- tilled water. Allow the solution to stand and siphon off the clear solution for use. BORON 79 THE DETERMINATION OF BORON *} After considerable experimenting with various methods suggested for the determination of boron in metallurgical products, we have found the following plan to be both accurate and rapid: Fuse one-half gram of the powdered sample with potassium nitrate and sodium carbonate in a platinum crucible. Pour the melt upon an iron plate, transfer with the crucible to a porcelain dish and boil with just enough water to effect disintegration. Add a little sodium peroxide during the boiling to precipitate manganese. Filter into an Erlenmeyer flask and wash with hot water. Reserve the residue for manganese determination. Acidify the filtrate with hydrochloric acid and then add a moderate excess of calcium carbonate. Connect the flask with an upright re- flux condenser and boil about ten minutes to remove all carbon dioxide. Filter, cool to room temperature, add phenolphthalein and run in N/10 sodium hydrate to pink color; then add about one gram man- nite. This destroys the color, and more sodium hydrate is added until the color is permanent, even on the addition of more mannite. One cc. N/10 sodium hydrate equals .0011 gram boron. No deduction is required for blank. One-half gram steel containing 4 per cent, manganese to which .03 gram boron was added in the form of fused boric acid gave by this method .0299 gram boron. Dissolve residue from the fusion in hydrochloric acid, add 15 cc. sulphuric acid and evaporate to strong sulphuric fumes. Cool, dilute, heat till clear and make up to 250 cc. Take 50 cc. for manga- nese determination by the bismuthate method, and 50 to 100 cc for iron determination. Manganese may be determined on a separate sample if desired by the bismuthate method without separating the boron. Silica is determined in the usual way, and carbon by direct ignition in oxygen, spreading the sample over ignited asbestos in the combustion boat. 80 BORON CARBON 81 DETERMINATION OF CARBON COLORIMETRIC METHOD In determining carbon by color it is essential that the standard contain approximately the same percentage of carbon as the sample, and also that the materials have had the same heat treatment. For the analysis of pure iron we furnish free a vial of standardized American Ingot Iron. Dissolve .5 gram of the sample and .5 gram of the standard in 10 cc. of nitric acid, (1.18 Sp. Gr.), using 10 in. by 1 in. test tubes, heat over a Bunsen flame until the metal is dissolved and tubes are free from brown fumes. Cool gradually and pour into carbon comparison tubes, dilute standard with distilled water to at least 20 cc. (depending upon the carbon content) and add water to tube containing sample until colors match. The percentage of carbon present is determined by the following formula: cc. of Standard : cc. of Sample :: percentage of Carbon in Standard : X. 82 CARBON CARBON 83 DETERMINATION OF CARBON BY COMBUSTION The determination of carbon in iron and steel is made by direct combustion, using a % inch bore x 30 inch silica tube heated in an electric furnace. A temperature of 1000 to 1030 C., is maintained with the use of a rheostat. The apparatus being standardized oc- casionally with the use of a platinum, platinum-rhodium platinum- rhodium thermo couple. We have found that a gas furnace is not reliable where the gas pressure fluctuates considerably. This is due to the danger of over- heating and causing devitrification of the silica tube, rendering it porous and the results obtained unreliable. The following method has been found to be extremely accurate when dealing with traces of carbon such as exist in American Ingot Iron. The method we employ is essentially as follows: Use a platinum or nickel boat approximately G'^/^'x/^", on which place a Y%" layer of 60 or 90-mesh alundum, free from carbon and alkalies, which has been heated to 1000 C., before being used for the first time. Make a channel in the alundum and place therein 4 grams of low carbon steel or American Ingot Iron, in the form of fine drillings. Spread compactly and evenly, then cover the borings with alundum. Connect a Meyer bulb containing 75 cc. of clear barium hy- droxide solution (30 grams Ba(OH) 2 ' 8H 2 . O, in 1 liter of water) direct with the silica tube. Insert the boat containing the drillings into the central portion of the silica tube and quickly connect with the source of purified oxygen which is passed into the apparatus at the rate of not more than 100 cc. per minute, continuing the operation for 20 minutes. The flow of oxygen gas is regulated with the use of a high pressure reducing valve which will maintain a uniform rate of flow. The Meyer bulb is disconnected, the solution filtered and washed with boiled distilled water. Care should be taken to filter under conditions which will prevent contamination by any carbon dioxide which may be formed within the laboratory. The filter paper containing the barium carbonate is placed in a platinum crucible and ignited at a low temperature until all volatile matter has been driven off, and finish at a red heat until all carbon from the paper has been consumed. 84 CARBON The barium carbonate is weighed, multiplied by 6.08 and divided by the weight taken, which will give percentage of carbon. For the analysis of high carbon steel we employ the rapid method in which we take 1/^-grams, and absorb the carbon dioxide in soda lime contained in a Fleming bulb. A complete determination can be made in less than seven minutes. IRON YARNING TOOL FOUND UNDER WATER MAINS AT CINCINNATI, OHIO. BURIED IN THE GROUND ABOUT 51 YEARS TEST NO, 4524 FILE 1O3 Sulphur .030 Phosphorus .030 Carbon .015 Manganese .112 Copper .080 Silicon .027 Oxygen .050 Nitrogen .004 CARBON 85 CARBON IN IRON 1 By T. D. YENSEN. 2 It has long been known that carbon has a great influence upon the properties of iron and iron alloys. On account of the small quantities of carbon involved 0.1 per cent being regarded as a medium carbon steel and the presence of carbonaceous matter in almost everything that is used in making chemical analyses, the correct de- termination of carbon in iron is the most difficult problem in iron and steel analysis. Not only are the total quantities of carbon very small, but carbon may exist in iron in various forms, one or more of which may influence the properties considerably, while others may have little or no influence upon these properties. The accurate determina- tion of these small quantities and the separation of the different forms is, therefore, of great importance, and a large amount of work has been done lately to discover suitable methods to accomplish this purpose. In another paper, Mr. C. J. Rottman is giving an account of the various methods that have been and are being used for the determina- tion of carbon in iron and steel. He is also describing certain im- provements in the method of analysis in connection with the com- bustion method particularly in regard to absorbing the CO 2 improvements that greatly diminish the errors in the analysis. How- ever, even these improvements were not considered sufficient. The method described in this paper is a combustion method, but it elimi- nates some of the most important sources of error inherent in the standard method, namely those due to absorption and weighing, and makes it possible to decide accurately whether all the carbon has been burnt out of the sample. The standard method consists in heating the sample in the form of shavings or chips in a gas-tight tube to a temperature of 800 to 1000 C., passing oxygen through the tube and absorbing the resulting CO 2 in a bulb containing KOH or the equivalent. The increase in weight of the KOH bulb gives the weight of CO 2 . Pre- cautions are taken, of course, so that presumably nothing but the CO 2 from the sample is absorbed by the KOH bulb. As ordinarily practiced this method is satisfactory for carbon content of 0.1 per cent, or more, and if great care is exercised the error should be within 0.01 per cent. Mr. Rottmann states that with his improvements he is able to get an accuracy of 0.002 per cent. While this may be correct when the analysis is done with painstaking care, it is very 1 Transactions American Electro Chemical Society, April 1920. 2 Westinghouse Research Laboratory, East Pittsburgh, Pa. 86 CARBON doubtful if better than d= 0.01 can be obtained with the analysis in the hands of a regular analytical chemist. Errors in the Present Method The chief sources of error in the present method are: 1. Contamination of sample due to oil, grease, dust or dirt of any kind. 2. Dust or dirt or other carbonaceous matter in the combustion tube, in the sample holder, or in the connections between the tube and the rest of the apparatus. 3. Adsorbed CO or CO 2 in the walls of the combustion tube, or in the sample holder. 4. Admission of CO or CO 2 in opening the combustion tube. 5. Incomplete washing of the oxygen before it enters the com- bustion tube. 6. Incomplete oxydation of the carbon, resulting in some CO instead of all CO 2 . 7. Incomplete absorption of CO 2 in the KOH bulb. 8. Weighing of the KOH bulb; moisture and dust collects on the bulb in uncertain amounts, and the weighing itself can at best be done with an accuracy of 0.1 mg. 9. Carbon left in sample. Some of these sources of error can be eliminated in the present method by careful manipulation, thus (1) can largely be taken care of by careful sampling and boiling the sample in ether prior to placing it in the combustion boat. (2) and (3) can be minimized by burning out the system, in- cluding the sample holder, with oxygen, prior to introducing the sam- ple, while (5) can be easily eliminated by means of active KOH and soda- lime in the train, according to standard practice. This leaves (4), (6), (7), (8) and (9) as sources of error that can not readily be eliminated when the present method is used. The resulting errors vary to such an extent that it is difficult to get satisfac- tory blanks, and it is, therefore, necessaryto make radical modifications. CARBON 87 Mr. Ryder's Results Mr. H. M. Ryder has done a great deal of work on the elimination -of gases from metals, including iron and iron-silicon alloys, by heating the samples in vacuo and analyzing the gases given off. 3 It was found that large quantities of CO and CO 2 were given off below 600 C., whereas additional CO was given off in varying amounts, depending upon the alloy, at and above the A 2 transformation point. Based on these results it was concluded that the CO and CO 2 given off below 600 exist in the metal as adsorbed gases while the CO given off at and above A 2 is due to chemical reaction between the combined or graphitic carbon in the metal and the iron oxide present. It is quite probable that these different forms of carbon, i. e., that existing as adsorbed gases and that existing as combined or graphitic carbon, have different effects upon the physical properties of the metal and it is, therefore, of great importance to differentiate between them. This differentiation is taken care of in the new method as described below. The New Method A sketch of the apparatus is shown on Page 88, and is self explana- tory. The following is a condensed statement of the procedure: (a) Preparation of the Sample) . The sample is carefully collected to keep out foreign matter and the weighed portion then ** cleaned with ether in the apparatus shown on Page 87. The ether is evaporated in an Erlen- meyer flask and the vapor passed through the sample held in the Gooch crucible. The con- densed vapor again passes through the sample on its way from the condenser to the bottom of the flask and is reheated by the rising vapors. The sample is thus exposed to a constant stream of hot, clean ether, carrying oily and greasy matter down into the bottom of the flask. This procedure should minimize source of error No. 1. Apparatus for cleaning Sample. (b) Preparation of Combustion Boat and Tube. The sample is then placed on a layer of specially prepared alundum in an alundum combustion boat that has previously been heated in the combustion tube to 1000 in a stream of oxygen. The sample is also covered with a layer of alundum. This precaution should eliminate 3 H. M. Ryder: "A Precision Method for the Determination of Gases in Metals. "Trans. Am. Electrochem. Soc. (1918), 33, 197. "Analysis of Small Quantities of Gases." Jour. Am. Chem. Soc. (1918), 40, No. 11, 1656. 2 is obtained. This can be done without introducing additional errors, which is not the case when the ordinary combustion method is used. Source of error No. 9 is thus eliminated. Details of Apparatus (a) Furnaces. The 1000 furnace is a platinum-wound furnace with a porce- lain tube 24 in. (60 cm.) long and 1/4 in. (4.4 cm.) bore, insulated with fire clay and sil-o-cel. It can be maintained at 1000 with an input of 1 kw., and at 1400 with 3 kw. The 600 furnace is 12 in. (30 cm.) long, \% in. (4.4 cm.) bore, consists of a silica tube wound with nichrome ribbon and insulated by means of magnesia pipe covering. (b) Quartz Tube. The quartz tube is 1% in. (3.5 cm.) inside by liMi in. (4.1 cm.) outside diameter by 6 ft. (180 cm.) long. One end is permanently sealed with a glass cap and cement, while the other end is provided with a ground glass joint. (c) The CuO Tube. The CuO tube is a Pyrex glass tube % in. (1.6 cm.) diameter, wound with asbestos and nichrome ribbon, without any heat insulation on the outside. It is filled with fine copper wire and heated to 400. (d) Connections. All connections between the different parts of the apparatus are made with sealed hard glass tubing, eliminating sources of error due to leaks. 90 CARBON (e) Analyzing Apparatus. The analyzing apparatus was constructed in accordance with a design originated by Mr. Ryder. A sketch of same is shown on Page 90. The apparatus is made from hard glass throughout. The sketch shows the mercury in the cutoffs in position ready for analyzing the CO 2 frozen out in the liquid air trap. The volume then includes the trap, the bulb of the McLeod gage and the connecting tubing. Up to the zero points of the mercury column the volume is 259 cc. As the pressure increases, the mercury falls in the various cut-offs, thus in- creasing the volume of the system. The volume, being a function of the pressure, was consequently determined for various pressures by means of the auxiliary bulb. The relation between the pressure and the carbon is shown graphically on Page 91. Expressed in an equation this relationship is: C = (0.168 + 0.000095 P) P, where C = mg. of carbon and P = pressure in mm. Hg. Pressures up to 2 mm. are measured on the McLeod gage, while higher pressures are measured by means of the difference in level of the mercury columns of the mercury pump cut-off. ftu,. Bu/t fir C*Mr An*/. Sys/en. Analyzing Apparatus for Carbon Determination. CARBON 91 Pressures can be measured on the McLeod gage with an accuracy of 0.01 mm. Hg, and from the above equation it will be seen that this corresponds to a carbon content of 0.00168 mg. or 0.0000168 percent carbon if a 10 g. sample is used. This is far beyond the required accuracy, so that no difficulty will be had in eliminating sources of error No. 7 and No. 8. (/) Diffusion Pump. The pump was constructed in accordance with the design de- veloped by Mr. J. E. Shrader, 4 and is of the type that can be operated at fairly high backing pressures. It is capable of evacuating the system down to 0.0001 mm. Hg in 5 minutes. Preliminary Tests Numerous preliminary tests were made to weed out the weak points of the apparatus and to determine the limits of accuracy. %*>' \ A OfS /I Ce ^ ^^ x ^ r /? r'' ^ I/1L+I ^ ' N, '//fl it / ','jk ^ ^* t'/t 7 /;; w / t- Combustion Tubes. 4. Originally the house vacuum was used to evacuate the furnace down to a pressure of about 10 mm. Hg. The house vacuum was then cut off and the final evacuation done with a diffusion pump. However, large variations in the blanks (from 0.0005 mg. to 0.05 mg.) led to investigating the house vacuum system as a source of CO 2 . The house vacuum as the name implies is used by the whole laboratory, and consequently there is a large variation in the pressure, ranging from a few mm. to 2 or 3 cm. Hg when someone is evacuating a large volume. It is conceivable that a sudden increase in the line pressure may cause a reverse current into the analyzing system and if there is any CO 2 in the line this will be frozen out in the liquid air trap. It was consequently decided to do all evacuation with the backing pump for the diffusion pump, and this resulted in a decided improvement. The apparatus in its final shape is therefore as shown on Pages 88 and 93, using an individual pump for the evacuation. * With the furnaces cold and with no sample in the boat, but otherwise running the apparatus as usual, the amount of C obtained is only 0.002 mg. (with a variation of from 0.001 to 0.003 mg.). 94 METALLURGICAL CONTROL With the furnaces heated as usual (to 600 and to 1000 C. re- spectively) and going through the regular procedure of analyzing a sample except that no sample is used, the following results are typical. Table II Blanks of Apparatus Procedure same as usual, except no Sample in Boat. Test No. Tube Evac. Boat Cold Boat heated in Vacuoto600 J Boat heated to and at 1000 Tube filled with O 3 , then evacuated. 104 108 124 Mean o Variatio min. Final Pres. sure mm. min. Final Pres. sure mm. Carb. Obt'd mg. 1st Period 2nd Period min. Final Pres- sure mm. Carb. Obt'd mg. min. Final Pres- sure mm. Carb. Obt'd mg. 23 14 51 F 3Te n fron 0.010 0.015 0.002 sts i Mean 11 20 20 0.000 0.003 0.001 0.013 0.027 0.017 0.019 0.007 27 26 36 0.010 0.010 0.005 0.029 0.025 0.038 0.031 0.007 25 40 o'.oio 0.005 0'.030 0.044 0.037 0.007 The maximum variation in the blank is therefore less than d=0.010 mg. C, which for a 10 g. sample amounts to 0.0001 per cent. For samples of different carbon contents the probable errors in the analysis are therefore as follows: Table III Probable Errors in Samples of Various Carbon Contents. 10 g. Samples. Carbon Contents Errors in Analysis Percent Mg. Mg. () Percent (. ) 0.1000 10 0.01 - 0.1 0.0100 1 0.01 1.0 0.0010 0.1 0.01 10.0 0.0001 0.01 0.01 100.0 Results (a) Electrolytic Iron. A sample of doubly refined electrolytic iron serving as our standard of low carbon iron was analyzed ac- cording to the above method, giving the following results: CARBON 95 Table IV Carbon Content of Electrolytically Refined Iron. Standard Sample. 5 g. Samples. g Sample Heated in Sample Heated to and at 1000. Tube Filled with O 2 and Evac. Net Carbon Obtained U Vacuo to 600* I 8 1st Period 2d Period 1st and Below At w 2d 600 1000 y, V Carbon Obtained Carbon Obtained Carbon Obtained Car. Ad- Com- Total Is B Obt'd sorbed bined H 1 Gross Blank Net Gross Blk. Net Gross Blk. Net. Net. Min. Min. Mg. Mg. Mg. Min Mg. Mg. Mg. Min Mg. Mg. Mg. Mg. Percent Percent Percent 110 18 20 0.286 0.019 0.267 31 0.219 0.031 0.188 29 0.093 0.037 0.056 0.244 0.0053 0.0049 0.0102 111 15 20 0.320 0.019 0.301 28 0.220 0.031 0.199 25 0.067 0.037 0.030 0,229 0.0060 0.0046 0.0106 112 15 20 0.295 0.019 0.276 28 0.235 0.031 0.204 29 0.076 0.037 0.039 0.243 0.0055 0.0049 0.0104 Mean of Three Same IPS 0.0056 0.0048 0.0104 Variation from Mean () 0.0004 0.0002 0.0002 * Carbon eliminated during 2d Period at 600 in Vacuo is < 0.01 mg. These results show that the total carbon content of this material is 0.0104 (0.0002) percent. The important fact to be noted from these results is that of the total carbon content, more than half (0.0056 (0.0004) percent) is in the form of adsorbed gases, such as CO or CO 2 , given off at or below 600, and that the combined carbon content is only 0.0048 (0.0002) percent. The above standard sample of electrolytic iron was carefully analyzed by Mr. C. J. Rottmann, according to the "Old" method, the average of 11 separate analyses being 0.0177 percent with a maxi- mum variation from the mean of 0.0080 percent. This amount includes all carbonaceous matter introduced into the combustion tube, i. e., CO2 admitted in opening and closing the tube, adsorbed gases and combined carbon. In order to determine the amount of CO 2 admitted to the combus- tion tube while removing the old sample and introducing the new one, the gas was analyzed during the preliminary evacuation with the sample cold. Six tests gave values of carbon introduced in this way varying from 0.15 mg. to 0.67 mg. Based on a 5 g. sample this corresponds to 0.0030 0.0134 percent. The results are tabulated in Table V for the sake of comparing the "Old" and the "New" methods. 96 CARBON Table V Comparison of "Old" and "New" Methods of Analysis. Electrolytically Refined Iron. Standard Sample. "New" Method Mean of 3 Tests Percent C "Old" Method Mean of 11 Tests Percent C 1. Carbon introduced while opening and closing combustion tube .... 0.0080d=0 0050 2 Carbon as adsorbed Gases 0.0056 0.0004 3 Carbon in Combined Form 0.0048 0.0002 Total, 0.01840.0056 0.0177 0.0080 The results are in remarkably close agreement, both as to total carbon and as to variation. It will be noted that the variation is largely due to item 1, namely carbonaceous gases introduced into the combustion tube while opening and closing the tube, gases that have no connection with the sample at all, but amounting to 30-130 percent of the total carbon content (items 2 and 3). This fact explains the large variation usually obtained in the carbon content of low-carbon iron. (&) Vacuum Fused Electrolytic Iron. Table VI shows some re- sults obtained with electrolitic iron after being melted in a vacuum furnace and forged into rods. 2-202 and 2-210 were melted in an Arsem type furnace and 2-251 in a Tungsten wound furnace. (C) Bureau of Standards Samples. As a check of the constants of the apparatus two of the Bureau of Standards standard steel samples were analyzed, the results being given in Table VII. Judging from these results the agreement is as close as can be expected, and the conclusion is justified that the constants of the apparatus are correct' 5 . Modifications of the Method. The apparatus as described in this paper is necessarily elaborate, especially the analyzing part of it, in order to eliminate all possible errors. However, great simplifications can be made to meet the requirement of different users. In another apparatus used in this laboratory the anlyzing system consists of two good stopcocks, 5 It should be noted that for such large carbon contents, the accuracy of the apparatus is not as great as for smaller carbon contents, because in the former case the pressures are too large to be read on the McLeod gage and must be read on the barometer directly. In the above cases the pressures were in the neighborhood of 20 mm., and the readings can be read with an accuracy of only 0.5 mm., resulting in probable errors of 2.5 percent. CARBON 97 Table VI Vacuum Fused Electrolytic Iron. 5 g. Samples. Test No. 121 119 120 Sample No. Description of Sample Net Carbon Obtained 600 Percent 1000 Percent Total Percent 2-202 2-202 2-202 As forged. Outer half of Rod, except surface Ether cleaned . . . 0.0104 0.0000 0.0000 0.0159 0.0038 0.0045 0.0263 0.0038 0.0045 Forged and Annealed in Vacuo. All of Rod except surface. Not cleaned . . Forged and Annealed in Vacuo. All of Rod except surface. Ether cleaned. 115 116 114 117 2-210 2-210 2-210 2-210 As forged. Outer half of Rod, except surface. Not cleaned 0.0027 0.0015 0.0004 0.0009 0.0042 0.0039 0.0030 0.0015 0.0069 0.0054 0.0034 0.0024 As forged. Outer half of Rod, except surface. Ether cleaned As forged. Center of Rod. Ether cleaned Forged and Annealed in Vacuo. All of Rod except surface. Not cleaned . . 122 2-251 As forged. Center of Rod. Ether cleaned 0.0006 0.0025 0.0031 Previous analyses by the "Old" method have given 0.02-0.04 percent [Carbon for the above samples, *. e., up to 10 times what the "New" method gives. a liquid air trap, and a long mercury column similar to the arrange- ment shown on Page 88, and it gives very satisfactory results. Further- more, split furnaces can be used, arranged on swinging arms, thus doing away with the long combustion tube. This modification would probably result in lower blanks, because the carbon obtained in the blanks is probably, in part at least, due to diffusion of CO 2 from the atmosphere through the heated portions of the tube. Finally the mercury diffusion pump may be eliminated, depending entirely on the oil pump for evacuation. This is done in the apparatus re- ferred to above. In short, the apparatus can be made as simple or as elaborate as the requirements call for. Summary and Conclusions. 1. It has been shown in this paper that by the method described great accuracy in determining the products of the combustion tube is possible ; 0.001 mg. carbon can readily be measured by the analyzing system. 2. The blanks of the apparatus are small and consistent, the maximum variations amounting to only 0.007 mg. carbon, which is the probable error of the apparatus. This, for a 10 g. sample CARBON containing 0.01 percent carbon would result in an error of 0.00007 percent carbon or 0.7 percent of the total. This is 100 times better than is possible by any of the present methods. Table VII Analysis of Bureau of Standards Standard Samples. 2 g. Samples. No. 20a Acid Open Hearth Steel, 0.4 percent Carbon. the Bureau of Standards, 0.393 percent Carbon. No. 15a Basic Open Hearth Steel, 0.1 percent Carbon. the Bureau of Standards, 0.109 percent Carbon. Containing, according to Containing, according to Test No. Sample Net Carbon Obtained 600 Percent 1000 Percent Total Percent 102 103 105 106 107 No. 20a . 0.015 0.010 0.367 0.365 0.381 0.375 No. 20a Mean 0.0125 0.0025 0.366 0.001 0.378 0.003 Variation from Mean (->-) No. 15a .... 0.005 0.009 0.006 0.112 0.094 0.101 0.117 0.103 0.107 No. 15a No. 15a. Mean 0.007 0.002 0.102 0.010 0.109 0.007 Variation from Mean (db) 3. By carrying out the analysis in three stages, namely (a) with the sample evacuated cold, (b) heated to 600 in vacuo, and (c) heated to 1000 in oxygen, it is shown that the total amount of carbon obtained from electrolytic iron is divided about equally between (a), (b) and (c) ; in other words, that about 0.005 percent is due to gases admitted to the tube in introducing the sample, that about 0.005 percent is present in the iron as adsorbed or occluded gas, and that the remainder 0.005 percent is the combined carbon. 4. A great deal of variation exists in the amount of carbon obtained during the first stage of analysis, (3a), and it has been shown that the great variations in the carbon contents of low carbon iron usually obtained may be attributed to this source. 5. The new method can be modified to suit any given condition; but for carbon contents of 0.10 and above the established methods, when carefully carried out, are sufficiently reliable to make more re- fined methods unnecessary. It is only for very low carbon contents CARBON that the errors in the established methods are so large as to conceal the true condition. The author wishes to express his appreciation to Mr. A. L. Shields for his conscientious work in connection with the analyses and to Mr. C. J. Rottmann for his valuable advice based on his large experience in all matters pertaining to analytical chemistry. Westinghouse Research laboratory, January 15, 1920. 100: : CARBON Drilling pig iron for chemical analysis. All pig iron is carefully analyzed, five samples being taken from each car-load received CARBON CARBON AN ELECTROLYTIC RESISTANCE METHOD FOR DETERMINING CARBON IN STEEL 1 By J. R. Cain and I. C. Maxwell 2 INTRODUCTION The purpose of this study was to investigate the accuracy, speed, practicability of a method for determining carbon in steel, dependent in principle on passing the carbon dioxide produced by direct com- bustion of the metal into a solution of barium hydroxide of known electrical resistance; after complete absorption of this gas the re- sistance is again determined and from the increase in this (due to precipitation of barium ions) the percentage of carbon is deduced. This method is new in principle and it is believed that the principle can be applied generally in many cases where the substance being determined precipitates another substance from solution with re- sultant change in resistance. The assembly of apparatus for de- termining resistance is also new, 3 and offers many advantages for technical work over the methods hitherto in general use for measure- ment of electrolytic resistances, which require the use of induction coils or high frequency generators, tuned telephones, balanced in- ductances and capacities, etc. Other new features are the application of the nomograph 4 for the graphical representation of resistance data and the use of special conductivity cells with adjustable electrodes to facilitate the manufacture of any number of such cells with the same cell constant. Much work has been done by others on electro-chemical analytical methods. In general, these fall into three groups in which an end- 1 Complete equipment for determination of carbon by this method may be obtained from Arthur H. Thomas Co., Philadelphia. 2 Journal of Industrial and Engineering Chemistry, Vol. 11, No. 9, page 852. September, 1919. 3 The elements of this were described by Weibel and Thuras, This Journal, 10 (1918), 626. 4 The mathematical work in constructing the nomograph shown on page 112 was done by Mr. H. M. Roeser of the Bureau of Standards at the request of the senior author, who suggested its appli- cation to electrolytic resistance data. A paper on this subject is in preparation by Mr. Roeser. References on the nomograph are: "Traite de Nomographie," by M. d'Ocagne, Gauthiers-Villars, Paris; "Graphical Methods," by Carl Runge, Columbia University Press, New York; "Graphical Interpolation of Tabulated Data," by H. G. Deming, J. Am. Chem. Soc., 39 (1917), 2388; "The Nomon, a Calculating Device for Chemists," by H. G. Deming, Ibid., 39 (1917), 2137. \18fc.%r :'/:*?; S3 S /. ' "; CARBON point is shown electrochemically by the following methods: (1) The unknown concentration is obtained from curves expressing a relation between cubic centimeters of titrating solution and conductivity (or a related quantity) of the solution titrated j 1 (2) the unknown is obtained from curves giving the relation between cubic centimeters of titrating solution added and the corresponding electromotive forces of a cell composed of a normal electrode and an electrode not acted upon by the solution being titrated, the latter being the electrolyte; 2 (3) special application of Method 2 used for determining hydrogen ion in acidimetry sjrfd alkalimetry and in precipitations from neutralized solutions. 3 Such methods suffer by comparison with the present for the following reasons: (1) A curve has to be plotted for every determina- tion, which consumes much time; (2) the apparatus required to de- termine carbon with an accuracy of 0.01 per cent carbon would be too delicate and inconvenient of manipulation for every-day use; (3) the difficulty in some cases of fixing with sufficient definiteness the inflection or break in the curve denoting the end-point of the titration. Upon further comparing these methods with the present, it is seen that the latter dispenses with one operation common to all the others, namely, the addition of successive portions of a titrating solution and the determination of the resistance at each addition, resulting in ad- ditional time-saving. From an inspection of the chemical equation for the reaction underlying the present method, Ba(OH) 2 + CO 2 = BaCO 3 + H 2 O, it is evident (when any given conductivity cell is used) that the only factors which act to change the conductivity of the barium hydroxide used for absorption are (1) the amount of carbon dioxide absorbed, which determines the disappearance from solution of the barium ion, and (2) the temperature. Since carbon dioxide precipitates barium without leaving reaction products in the solution to increase the con- ductivity (such as would remain if, for instance, sodium sulfate were the precipitating agent for the barium) it can be seen that the present 1 Harned, J. Am. Chem. Soc., 39 (1916), 252; Findlay, "Practical Physical Chemistry," Ostwald- Luther, "Physikalisch-Chemische Messungen." 2 Loomis and Acree, Am. Chem. J., 46 (1911), 585, 621 (a bibliography is also given); Hildebrand, J. Am., Chem. Soc., 35 (1913), 869; Kelly, Ibid., 38 (1916.) 341. 3 Hildebrand, Loc. cit. See also Weibel and Thuras, Journal of Industrial and Engineering Chemistry, 10 (1918), 626, for another electrolytic method. CARBON 103 method should give the maximum possible change of resistance for a given amount of barium removed a condition tending to secure a high degree of sensitiveness. 1 However, the temperature coefficients of resistance of barium hydroxide solutions in the range of concentra- tions herein employed average nearly 1.7 per cent per degree, hence it is evident that the accuracy of the method will be largely affected by temperature if due correction is not made. In developing this method it was deemed necessary: (1) To con- struct the curve showing resistance as a function of concentration of barium hydroxide solutions ranging from very concentrated to very dilute and to select the portion of this curve showing the maximum change of resistance for a given change of concentration ; (2) to devise an apparatus which, when the selected barium hydroxide solutions were used in it, would completely absorb the carbon dioxide at the highest rates of passage of the gas current. The same absorption apparatus, in order that the method might meet the requirements of convenience and rapidity, should permit resistance determinations to be made without transfer of the solution to another vessel; it should also be easy to fill and empty ; (3) to determine the temperature coefficients of the barium hydroxide solutions in the selected range of concentration; 2 (4) to prepare a chart enabling the operator to read directly therefrom the percentage of carbon, all temperature cor- rections being incorporated ; (5) to design an apparatus for the elec- trical measurements possessing the necessary robustness, reliability, simplicity, and protection from corrosion by the laboratory fumes. The Resistance of Barium Hydroxide Solutions A curve was prepared showing the relation between electrical resistance and barium hydroxide concentration when the latter was varied from practically saturation to nearly zero. The data in the literature being insufficient for this purpose, the de- terminations were made by Mr. Louis Jordan of this Bureau and are represented on Page 104. The solutions for constructing the curve were prepared from J. T. Baker's analyzed barium hydroxide by diluting a stock solution of this with carbon dioxide-free water and determining their strength by titration against standard hydrochloric acid using methyl orange as indicator. The resistance measurements were all made in the same conductivity cell and at practically the same 1 Compare the conditions in Harned's work with Ba(OH) 2 solutions. Loc. cit. 2 There is some possibility of placing the absorption vessel in a constant temperature bath to- gether with a compensating cell in the other arm of the bridge. This would remove the necessity for temperature correction. The method described herein, however, is believed to be simpler. 104 CARBON temperature (27 C.) 1 . No great accuracy is claimed for these results, which are used only for establishing the form of the curve and 60. 50. 40. 30. 20. 10 , . ,, , : ." l%2%3%4*5% 10% GAMS BA(OH) 2 PERIOD cc. V s.0 6.0 r.o 3,0 ao 15% 2i>% ds% 3fe%* %CAR30N selecting a portion of it for more exact redetermination. The portion of the curve selected for use in this method is that between A and B. This region gives the maximum change of resistance per unit change of concentration consistent with the use of solutions sufficiently con- centrated to effect complete absorptions of carbon dioxide under the conditions imposed. Comparing the resistance changes with changes of concentration of barium hydroxide corresponding to 1 per cent carbon on different parts of the curve, it is seen, for example, that these changes in resistance are approximately six times as great on the portion 1 They were not corrected by any temperature coefficient since, as an inspection of Table 1 shows, they were made at near enough to one temperature to give roughly the form of the curve, which was all that was desired. CARBON 105 AB as on the portion CD. The use of the most dilute solutions possible is, of course, also desirable from the saving in barium hydroxide effected. Solutions to the left of A, even when used in very efficient absorbing vessels will not retain all the carbon dioxide except at rather slow rates of aspiration. TABLE I Data for Resistance-ConcentrationCurve for Ba(OH) Solutions in the Region 2 Per Cent to 30 Per Cent Carbon Equivalent Strength Cell Constant = 0.715 Ba(OH) 2 per 200 cc. Soln. Equivalent t Temperature Resistance Grams Per cent Carbon Deg C. Ohms 0.612 1.12 1.68 2.20 2.67 3.17 3.64 4.05 4.43 4.94 5.31 5.88 6.45 6.93 7.34 7.84 8.45 2.14 3.92 5.87 7.70 9.35 11.1 12.7 14.2 13.5 17.9 18.6 20.6 22.5 24.2 25.6 27.4 29.5 27.9 27.3 27.3 27.3 27.3 27.2 27.0 26.9 27.0 27.2 27.0 26.9 26.0 26.8 21.2 21.2 27.3 171.6 97.9 68.5 53.5 44.9 38.7 34.6 31.2 29.2 26.2 24.7 22.7 21.8 20.1 20.8 19.7 16.4 1 The method used in Tables I and II and elsewhere throughout this paper for expressing the strength of the barium solution in terms of "equivalent per cent carbon" was chosen for convenience in using the nomograph described subsequently. By "equivalent per cent of carbon" is meant the amount of carbon expressed as percentage on the basis of 2-g. samples being used, necessary to pre- cipitate completely all the barium ions from the solution concerned. This amount of carbon is calculated from the equation; Ba(OH) 2 + CO 2 = BaCO 3 -f- H 2 O. For instance, in the first horizontal column of this table it is seen that 2.14 "equivalent per cent carbon" corresponds to 0.612 g. Ba(OH) 2 ; the former figure was obtained by solution of the proportion: Mol. Wt. Ba(OH) 2 : At. Wt. Carbon :: Wt. of Ba(OH) 2 in 200 cc. Sample 171.38 : 12.00 :: 0^612 : X whence X 0.0428 g. carbon or the "equivalent per cent carbon" = (*/2)100 2.14. Soln. : Wt. of Carbon in The portion of the curve selected for use by this method is again shown on Page 107. The procedure used in determining the points on it was the same as for constructing the curve shown on Page 104, except that more care was taken to secure accurate readings and the resultswere corrected by the coefficients given under the heading "Temperature Coefficients." Table II gives the data used in constructing this curve. 106 CARBON TABLE II Data for Resistance-Concentration Curve of Ba(OH) 8 Solutions in the Region 4 Per cent to 5 Per cent Carbon Equivalent Strength Cell Constant = 0.715 Ba(OH) 2 per Observed Corrected 200 cc. Soln. Equivalent Temperature Resistance Resistance Grams Per cent Carbon Deg. C. Ohms Ohms 1.164 4.075 23.6 100.9 98.49 .198 4.194 24.2 97.4 96.08 .239 4.337 24.2 94.6 93.26 .242 4.348 23.2 95.8 92.81 .280 4.482 24.4 91.4 90.42 .286 4.500 26.6 87.6 90.01 1.301 4.553 . 27.5 85.3 88.92 1.303 4.559 26.4 86.8 88.79 1.346 4.712 25.1 86.0 86.06 1.385 4.848 25.8 82.7 83.90 1.403 4.912 24.4 84.2 83.94 1.460 5.110 23.1 82.8 80.11 The Absorption Apparatus The essential features desirable in the absorption apparatus are: (1) It must retain all carbon dioxide when gas passes at the rate of 300 to 400 cc. per min.; (2) it must permit resistance measurements "to be made without transfer of the solution to another cell; (3) tem- peratures of solutions should be easily read to 0.1 ; (4) the cell should be easy to clean and fill with fresh solution ; (5) all cells should be built with the same cell constant, 1 or the latter should be capable of ad- justment to one value for all, so that the chart or set of tables may be used for all cells. Fleming 2 and others have shown that the combustion of a sample of steel and the sweeping out of the products of combustion from the apparatus where combustion tubes of the usual length and bore are used, can be accomplished in 5 to 6 min. ; this requires passage of the oxygen at the rate of about 300 to 400 cc. per min. Soda lime has been found to absorb all the carbon dioxide at this and higher rates. However, barium hydroxide solutions of all concentrations are much less efficient absorbents than soda lime. The absorbing efficiency increases slightly with the concentration, but even when the most concentrated solutions were used it was impossible to absorb com- pletely all the carbon dioxide in the usual types of gas absorption vessels, nearly all of which were tried. The method of testing these forms of apparatus was to burn a sample of steel, having a gas meter before the furnace to control the rate of passage of the oxygen, and to 1 The "cell constant" is not a constant at all, as Washbourne and others have shown, but for want of a better term this expression has been used throughout this paper. 2 Iron Age, 93 (1913), 64. CARBON 107 attach after the tube being tested a second absorption vessel con- taining a little clear barium hydroxide solution; a test was not con- sidered satisfactory if the second tube showed any cloudiness. Satisfactory absorption was secured in a vessel similar to that described by Weaver and Edwards, 1 with suitable modifications, the details of which were developed by Mr. S. M. Hull, formerly of the Chemical Division of the Bureau of Standards. For use in the present work, electrodes were sealed into this absorption tube in the middle reservoir. In order always to secure the same cell constant, the area and distance apart of the electrodes were originally adjusted before sealing into the reservoir. This was found to be an extremely un- certain and difficult operation, and this difficulty as well as the general fragility of the apparatusled to its abandonment and the search for some- thing simpler. After the completion of investigationsdescribedin another paper, 2 it was found that by burning the sample as therein described (namely, by placing it on a preheated boat, allowing boat and sample to further preheat for a minute in the furnace and then admitting oxygen at 300 to 400 cc. per min.), it was possible to completely burn a 2-g. sample in 1^ to 2 min. Since the first few hundred cubic centimeters of oxygen combine with the iron, there is produced a 1 Journal of Industrial and Engineering Chemistry, 7 (1915), 534. 2 Cain and Maxwell, Journal of Industrial and Engineering Chemistry, 10 (1918), 520. 108 CARBON CARBON 109 much better partial pressure of carbon dioxide than is secured where a sample burns slowly and this makes it possible to use a very simple absorption tube, such as that shown on Page 108, for the carbon dioxide. It was also found simpler to build a cell whose electrodes could be adjusted to secure a given cell constant than to attempt to obtain the same result with the electrodes sealed in a fixed position in the cell. This fact and the use of the special method of combustion described led to the design of the apparatus shown on Page 108. This is not fragile and the glass parts can be built by any glass blower of ordinary skill ; it meets all the listed requirements of the absorption apparatus. The adjustment of the cell constant is made by moving the electrodes up and down after loosening the stuffing box; a marked change takes place as they approach the meniscus. Initial approxi- mate similarity can be attained by the glass blower with comparative ease. Once the electrodes are set in the proper position, this is main- tained by cementing with DeKhotinsky cement. The electrodes are platinized and the cell constant is determined as described under "Oper- ating Suggestions." A certain amount of adjustment of the constant of the cell can be secured by removing or adding platinum black to one or the other of the electrodes during the platinizing operation by the use of an auxiliary electrode. Some cells in long use at the Bureau have differed in cell constant by 0.04 without affecting the accuracy of results. Actually, it is easily possible to adjust cell constants within 0.005, and this practice is to be recommended. Temperature Coefficients The solutions for determination of these coefficients were prepared and standardized as already described. The conductivity cells used for the work were kept in a thermostatically controlled chamber where the temperatures were maintained within 0.01 C. Resistances of barium hydroxide solutions equivalent to ap- proximately 4.0, 4.25, 4. 50, 4. 75, and 5.0 per cent carbon were determined at 20, 25, and 30. The experimental values and the corresponding calculated values for a and B are shown in Table III. These values for a and B were calculated by substituting the values for temperature and resistance from Table III for t in the equation 1 1 + a[t 25] + B[t 25] 2 , Rt R 25 and solving for a and B. 110 CARBON Methods for Direct Reading of Carbon Percentages Several methods were tried for graphically representing the rela- tion between carbon percentages and the corresponding observed temperature and resistance measurements. At first tables were constructed in which temperature corrections were calculated and applied for every tenth of a degree and every tenth of an ohm, but these were found to be too cumbersome. Other tables were then constructed giving the temperature corrections only, with the idea of adding or subtracting these each time a reading was made. This condensed the tables very considerably but increased the amount of calculation necessary. A series of curves was then constructed showing the relation between resistance and per cent carbon, a curve being constructed for every tenth of a degree. Such curves are con- venient to use only if they are plotted on an inconveniently large scale ; even then it is very fatiguing to use them for many readings at a time. TABLE III Data for Temperature Coefficients of Resistance Concn. of Ba(OH) 2 Soln. 1 Per cent C. Temperature Resistance Approx. Deg. C. Ohms a B X 10- 4 5.0 20 135.25 0.01674 0.2687 25 124.02 30 114.37 4.75 20 142.18 25 130.36 0.01680 0.3505 30 120.16 4.50 20 149.07 25 136.62 0.01686 0.3085 30 125.91 4.25 20 156.18 25 143.07 0.01687 0.1652 30 131 .89 4.00 20 165.18 25 151.28 0.01687 0.0898 30 139.48 1 200 cc. of the solution contained barium hydroxide approximately equivalent to the carbon in a 2-g. sample with 5 per cent C, or 100 cc. of the solution contained barium hydroxide approximately equivalent to the carbon in a 1-g. sample with 5 per cent C; or approximately 0.7125 g. Ba(OH).,. The other solutions contained barium hydroxide approximately equivalent to 1-g. samples with 4.75 per cent, 4.50 per cent, and 4.25 per cent, 4.00 per cent, respectively. After a little study it was found that the B term could be omitted from the equation for correcting temperatures to 25, provided this equation is applied only between the temperatures 15 and 35 ; at higher or lower temperatures the correction for the B term is ap- preciable. If the laboratory temperature is not within these limits, CARBON 111 the stock solution in carboy F on Page 114 must be brought within this range by placing it in a bath of cold water. The equation (1) R 25 = R/[l + a(t 25) + B(t 25) 2 - then reduces to R/ R 25 = R/[l + a(t 25)], in which a = 0.01681, this value being taken from Table III. After trials of other forms of curves, it was found that the parabolic form answered the requirements of fitting the observations sufficiently well and of permitting the construction of a nomograph; the equation to the curve shown on Page 107 was calculated on the assumption that it was parabolic, using the methods of least squares and the observations shown in Table II. This equation was found to be C 2 13.589 C +. 63.191 0.2478 R 25 = or, C 2 - - 13.589 C + 63.191 -- 0.2478 [l + a(t 25)] = whence, (2) C = 6.79 }Wo.9912R(0.5798+.01681/ 67.11), C being the equivalent per cent of carbon in the sense already explained. From Equation 2 the nomograph shown on Page 112 was prepared. This nomograph can be used for all cells having cell constants not too dif- ferent from 0.715, which was the cell constant of the cell used when data for the curve shown on Page 107 were obtained. As has been shown, the form of cell used allows the electrodes to be adjusted always to secure this result. (Page 108.) The use of the nomograph is very simple. A straight edge is made to correct the observed values for temperature and resistance and the intersection with the third (middle) ordinate gives the 'con- centration of the solution in terms of carbon percentage; this may be termed the first concentration. After the combustion is ended a similar set of readings is taken and subtraction of the second con- centration reading from the first gives directly the per cent of carbon if a 2-g. sample has been taken; or, if a 1-g. sample has been used, the result is multiplied by two. The scales can be read to 0.005 per cent C, 0.05 C., and 0.05 ohm. It was found by comparison of chart readings in a number of cases with the resistances computed by Equation 1 that the error of the chart is less than 0.005 per cent carbon. 112 CARBON .98 4 ^-J3 4.00 + 10 +ZO + 60 470 4-90 SOO ti\ ^* a fer* 7 78 77 --20 CARBON 113 Apparatus for Determining Electrical Resistance With the co-operation of the Leeds and Northrup Company, different forms of apparatus for measuring electrolytic resistances were built and tried. Two forms of high frequency generator in connection with tuned telephones were used ; also one or two forms of induction coil as sources of alternating current, and an interrupter such as is used in wireless investigations; the latter was quite un- satisfactory, as might be expected, due to polarization. Good results were obtained with the high frequency generators and tuned tele- phones, provided that inductances and capacities were used as directed by recent investigators. Such refinements are inconvenient in a method such as the present, intended for works use, and there are two other important objections from the same standpoint: (1) the diffi- culty of detecting a minimum in the telephone when working in a noisy place such as the usual works laboratory and the fatigue of the operator who would have to make possibly 80 or 90 determinations in an 8-hr, day, and (2) the liability to deterioration of a high frequency generator when operating continuously 24 hrs. per day as might happen in technical use. These considerations led to the development of a method of measuring electrolytic resistances by the use of commercial 60- or 25-cycle current. It was found after trial of the Weibel 1 galvanometer, a vibration galvanometer and of a direct current galvan- ometer operating off a crystal rectifier placed in the secondary of a transformer (the cell being in the primary), that the Weibel galvan- ometer offered a very satisfactory zero instrument for such measure- ments of electrolytic resistance as had to be made in this work. DIAGRAM OF BRIDGE USED FOR RESISTANCE MEASUREMENTS The bridge shown on Page 113 was accordingly constructed by Leeds and Northrup. The resistance coils and galvanometer are enclosed in a hermetically sealed box so that they are efficiently protected from cor- 1 E. E. Weibel, Bureau of Standards, Scientific Paper, 297 (1917). 114 CARBON JN-J rosion. There are three dials tens, units, and tenth ohms, respec- tively and a fourth which adds 100 to the readings of the others, when this is desired. An accuracy of 0.1 ohm in the readings is all that is necessary, although much better than this can be done. This instrument has been in use several months and has been very satis- factory. A perfect minimum is always registered by the galvanome- ter, provided the electrodes are well platinized. CARBON 115 Procedure for Determining Carbon A nichrome-wound furnace is used for heating. Porcelain or glazed quartz tubes may be employed; they should be fitted with a sheet nickel sleeve to protect internally from fused iron oxide which is thrown off from the burning steel. The two absorption vessels are rilled to the 200 cc. mark with barium hydroxide solution from the stock bottle F (Page 114; see also Operating Suggestion 1). Oxygen or air freed from carbon dioxide is passed for a few seconds to mix the solutions and their temperatures and resistances are then recorded. In the meantime the combustion boat filled with alundum sand has been preheating in the furnace, which for this work should be main- tained continuously at 1050 to 1100, perferably the latter tempera- ture. 1 This is an extremely important point, for if the temperature is too low or the oxygen is not pure or is not admitted at 300 to 400 cc. per min. after the combustion starts, the rapid combustion es- sential to successful absorption cannot be secured. The boat is re- moved from the furnace and when at a low red heat the sample is distributed on the alundum 2 and the boat replaced in the furnace and left to preheat, without oxygen passing, while the next sample is being weighed. Oxygen is now passed at the rate of 300 to 400 cc. per min. for the next two minutes; then the stopcock, is turned to the position, which admits carbon dioxide-free air; this should pass at the same rate as the oxygen. During this combustion period, if directions have been followed, all carbon dioxide will have been removed from the furnace, but some still remains in the large bulb of the absorption apparatus. The air removes this. The advantage of the use of air at this stage is obvious : a saving of oxygen is effected and the furnace is immediately available for burning the next sample. While air is passing through the first tube (requiring \ 1 A to 2 min.) the combustion of the next sample proceeds as already directed, using the second absorption tube. The second reading of resistance and temperature on the first tube then follows, and if the solution is not too dilute it can be used for absorbing the carbon dioxide from other samples; other- wise, a little is allowed to flow into the reservoir G, and the tube is filled to the mark again with fresh solution. Of course it is an economy in time for the operator, wherever possible, to choose conditions (weight of sample, carbon content of same, etc.) so as to get the maxi- mum number of determinations for a single filling, since in this way the 1 The melting point of gold is a convenient temperature check. If this metal is not melted the temperature is too low. See paper cited, by Cain and Maxwell. 2 Experience has shown that no loss of carbon occurs unless the sample contains extremely fine particles; with most steels these can first be removed without causing an error in the carbon deter- mination. This point should always be tested, however, in burning new steels. 116 CARBON second resistance and temperature readings serve as initial values for the next combustion and so on. The solution should not be used when it is more dilute than corresponds to 4 per cent carbon (i. e., 99.5 ohms at 25 ; see nomograph, on Page 1 12, for the limiting resistances cor- responding to other temperatures), since its absorptive power at rapid rates of passage of the oxygen is then less, and some carbon dioxide may be lost. The data relating to combustions should be recorded as obtained. It is convenient to use a tabular form for record showing : (1) Designation of sample; (2) weight taken; (3) cell used; (4) initial temperature; (5) final temperature; (6) initial resistance; (7) final resistance; (8) initial concentration; (9) final concentration; (10) carbon percentage in sample; (11) remarks. There is ample time for entering this information while other operations are going on. A very good way is to enter "final temperature" below "initial tempera- ture," and "final resistance" after "initial resistance" for each sample, since this relates these quantities in an easy manner for reading the nomograph. The speed of the method naturally depends on the skill of the operator and the ingenuity displayed in arranging a cycle of operations which secures the best speed under his working conditions. Operators at the Bureau of Standards when working on a series of Bureau of Standards' analyzed samples averaged one determination per 4J^ to 5 min. The accuracy of results is shown in Table IV. TABLE IV Results by Electrolytic Resistance Method B. S. Standard Sample 1 Used lla. . B. S. Value for Carbon (Av. by Direct A^t. Used Combustion) Grams Per cent 2.0 223 Value by Electrolytic Resistance Method 0.21 Analyst Maxwell lla 2.0 0.21 126 Ua 16a 2.0 0.409 2.0 1.0 2.0 0.813 1.0 2.0 0.5 990 0.41 0.41 0.42 0.81 0.82 0.80 1 00 Swindells Swindells Maxwell 2la Sugar 0.5 1.0 2.0 2.0 0.617 2.0 Gram 0.00421 0.00421 0.00421 1.00 1.00 0.98 0.62 0.62 Gram 0.0046 0.0042 0.0040 Cain Cain Maxwell Maxwell Maxwell Maxwell CARBON 117 Operating Suggestions l^Stock barium hydroxide solutions are conveniently made in one to two carboy lots by adding solid barium hydroxide to the carboy nearly filled with water (agitating with air) until the equivalent strength approaches 5 per cent carbon ; subsequent additions can then be made by adding a saturated barium hydroxide solution. Of course, it is not necessary to make up exactly to the equivalent of 5 per cent carbon ; an approximation to this is all that is desired. The strength of the solution is determined from time to time during the standardization by running a portion of the solution into the cell and measuring its resistance. If a set-up like that shown on Page 114 is used in measuring the resistance, it is not necessary to remove the carboy from the shelf or to break any of the connections during these opera- tions. If it is desired to economize in the use of barium salt, the waste solution in reservoir G can be brought up to strength as described, after first decanting it off from the barium carbonate that has settled out. A still further economy can be effected by drying and calcining to oxide the precipitated barium carbonate; this oxide can then, of course, be used again. 2 The cell constants should be checked from time to time. This may be done (1) by turning standard samples, or (2) by determin- ing the resistance of a N/ 10 solution of pure potassium chloride. This solution should be prepared on the day it is used, since it has been found that stock solutions change during the course of this work. Table V shows the resistivities at various temperatures of N/10 potassium chloride solutions. The cell constant is computed from the formula R = NS, or N = R/S, where R is the observed resistance, N the cell constant, and S the resistivity, taken from Table V, for the same temperature. If a change in cell constant has taken place it is most probable that the electrodes need replatinizing. Directions for this are given below. If Method 1 is used, any marked deviation from the correct carbon value of the standard steel may be due to a change in the cell constant, and this should then be checked by Method 2, unless the error in the carbon determination is suspected to be due to some other cause. In the present work no deviation of cell con- stants has been observed until after several months' use, and then the change is sudden and erratic. When the cell constant changes it should be brought back to the original value by the methods already described under the heading "The Absorption Apparatus." 3 If the absorption vessels are not to be used for some time they should be cleaned with hydrochloric acid (not over 2 to 3 per cent 118 CARBON HC1) followed by distilled water. Extreme care should be taken that none of the hydrochloric acid or chlorides enters the reservoir for waste barium hydroxide solution, if this is to be used again. 4 To platinize the electrodes, the cap carrying them is removed from the cell and they are first cleaned with sulfuric acid and chromic acid mixture followed by distilled water. Then they are immersed in a vertical position in a solution made of 100 g. water, 3 g. chloroplatinic acid, and 0.02 to 0.03 g. lead acetate. Current is passed through the solution by connecting the electrodes to three dry batteries in series; the current is passed for 5 min., reversing TABLE V Specific Resistivity of N/IQ KC1 Solution at Various Temperatures (from Landolt-Bornstein Physickalish- Chemische Tabellen, 4th Ed., Page 1117). Temperature Resistance Deg. C. Ohms 15.0 95.42 16.0 93.28 17.0 91.32 18.0 89.37 19.0 87.49 20.0 85.69 21.0 83.96 22.0 82.30 23.0 80.71 24.0 79.11 25.0 77.64 26.0 76.16 27.0 74.79 28.0 73.42 29.0 72.10 30.0 70.82 31.0 69.59 32.0 68.40 33.0 67.20 34.0 66.09 35.0 64.98 very half minute. Finally, an auxiliary platinum electrode is in- troduced and current passed with this as anode for another 2 min., after which the electrodes are washed thoroughly with distilled water and are then ready for use. Summary ^ I This paper describes the fundamental principles of a method for determining carbon dioxide by absorbing it in barium hydroxide solution and measuring the resistance change of the solution in relation to its concentration. II A suitable absorption vessel with electrolytic resistance cell incorporated is illustrated and described. CARBON 119 III Resistance measurements of barium hydroxide solutions varying in connection from 3.08 g. Ba(OH) 2 per 1. to 42.25 g. Ba(OH) 2 per 1. were determined approximately, and determinations accurate within 0.01 ohm were made of the resistances of twelve different solutions varying from 5.820 g. Ba(OH) 2 per 1. to 7.300 g. Ba(OH) 2 (OH) 2 perl. IV Temperature coefficients of resistance of these twelve solutions were determined in the range 20.00 to 30.00 C. V Based on the measurements of resistance of the barium hydroxide solutions, solutions with concentrations varying between 5.820 g. Ba(OH) 2 per 1. and 7.300 g. Ba(OH) 2 per 1. were chosen as most suitable for this method. VI Special resistance-measuring apparatus was developed which simplified these measurements by dispensing with tuned telephones, high frequency generators, and balanced inductances and capacities. VII A convenient nomographic method of applying necessary temperature corrections to resistance measurements was developed. VIII The method permits of accurate determinations of carbon in steel (i. e., within 0.01 per cent carbon), by the direct combustion method in 4J/2 to 5 min. A cknowledgment The authors desire to acknowledge the work of Mr. H. E. Cleaves, former member of the Chemistry Division of the Bureau of Standards, who carried out some preliminary measurements, and the assistance of Messrs. Silsbee, Agnew, and Isler of the Electrical Division of the Bureau, who co-operated effectively in the selection of a suitable alternating current galvanometer. The Leeds and Northrup Com- pany also co-operated fully by constructing .and loaning electrical measuring apparatus for experimental w T ork and by finally building in practical form the apparatus that was developed. Bureau of Standards Washington, D. C. 120 CARBON CHROMIUM 121 CHROMIUM METHOD FOR CHROMIUM AND VANADIUM IN CHROME-VANADIUM STEELS BUREAU OF STANDARDS Dissolve the sample (2.00 g.) contained in a 300 cc. covered Erlen- meyer flask in 10 cc. of 30% sulphuric acid and 20 cc. of water. Di- lute to 200 cc. with boiling water and add from a burette sodium bicarbonate solution (80 g. per 1.) until a permanent precipitate is formed, and then 4 cc. more. Boil one minute, allow to settle and filter rapidly. Wash 4 to 5 times with boiling water. (The nitrate may be used for the determination of the manganese.) Ignite the residue on the filter paper in an iron crucible and fuse with 10 to 12 times its volume of sodium peroxide. Dissolve the melt by immersing in 100 cc. of water, remove the crucible and destroy any remaining- peroxide. (This may be accomplished by heating for one-half hour on the steam bath.) Filter into a 200 cc. graduated flask and fill up to the mark water. Acidify 100 cc. of this solution with sulphuric acid and titrate with ferrous sulphate and permanganate. The chromium may be calculated from the number of cc. of permanganate equivalent to the chromium reduction. Neutralize the rerriaining 100 cc. of solution with sulphuric acid and then add 3 cc. more of sulphuric acid (sp. gr. 1.84). Heat the solution to boiling and reduce in a Jones reductor allowing the reduced solution to flow into a ferric alum solution contained in the receiving flask. Remove the receiving flask, decolorize the ferric iron by adding a few cc. of phosphoric acid anol titrate hot with potassium permanganate. Subtract from the permanganate used one-third the number of permanganate equivalent to the chromium reduction in the first 100 cc. aliquot. The remainder represents the amount of permanganate required to oxidize the vanadium from V 2 O 2 to V 2 O 5 . In passing through the Jones reductor the chromium is reduced to CrO and the vanadium to V 2 O 2 . These react with the ferric alum more or less completely reducing some of it to the ferrous con- dition. The permanganate oxidizes the chromium to the Cr 2 O 3 condition, the vanadium to V 2 O 5 and the iron to the ferric condition. 122 CHROMIUM DETERMINATION OF CHROMIUM BY BISMUTHATE METHOD This method is based on the fact that chromium and manganese are oxidized by sodium bismuthate in either nitric acid, sulphuric acid, or a mixture of nitric and sulphuric acids. Nitric acid alone is generally used, and only in rare instances will it be necessary to add sulphuric acid to facilitate the solution of the metal. Some manganese is oxidized to permanganic acid, which is decomposed by boiling, forming nitrate of manganese and manganese dioxide. The manganese dioxide is removed by nitration through asbestos, washing the asbestos well with a 3% solution of nitric acid. If any chro- mium is present it will be indicated by a yellow color in the nitrate. Dissolve 3 grams of the sample in a mixture of 70 cc. of water and 30 cc. of nitric acid, (1.42 Sp. Gr.). Boil until metal is in solution. Cool slightly and add 2 grams of sodium bismuthate, taking care to wash all bismuthate from the neck of the flask. Boil for 15 minutes, or until permanganic acid is decomposed. Filter with suction on asbestos supported by a tuft of glass wool in a 3" glass funnel. Wash with 3% nitric acid. Cool to tap water temperature and dilute to 500 cc. with distilled water. Add a measured excess of ferrous ammonium sulphate solution until free from yellow tints. Titrate the excess with standard permanganate to faint pink color that persists for 30 seconds. Titrate with permanganate 50 cc. of ferrous ammonium sulphate containing the same amount of acid and water as the test. The amount of ferrous ammonium sulphate oxidized by the chromic acid and measured in terms of permanganate is multiplied by the iron factor x .31, or 1 cc. of tenth normal potassium permanga- nate equals .00173 gram of chromium. The ferrous ammonium sulphate is prepared by dissolving 50 grams of the salt in 2 liters of 10% by volume of sulphuric acid. When this strength solution is used, 1 cc. will equal about J/2 cc. of tenth normal permanganate. The following table indicates the accuracy of the method, showing in some instances a very small percentage of chromium in the presence of a very high percentage of manganese: Chromium Added Chromium Found Manganese 1.44 1.42 .67 1.44 1.48 .01 .37 .37 .38 .33 .35 .55 .29 .31 .75 .24 .26 .95 .12 .11 1.30 .04 .04 1.90 CHROMIUM AND VANADIUM 123 DEMOREST 1 METHOD FOR THE DETERMINATION OF CHROMIUM AND VANADIUM Dissolve 2 grams of the sample in a 400 cc. flask by a mixture of 12 cc. of strong sulphuric acid and 50 cc. of water, heat the flask until solution is complete, set it off the hot plate and add, very cautiously, 25 cc. of nitric acid, (1.42 Sp. Gr.) The iron is immediately oxidized to the ferric state with the evolution of much nitrous fumes. Heat the solution to boiling until the brown fumes are all driven off, then set the flask off the hot plate and add sodium bismuthate until every- thing in solution is oxidized and the manganese appears as perman- ganic acid and does not disappear on shaking. Dilute the solution to 200 cc., add a little more sodium bismuthate, and boil the solution for 20 minutes to decompose the permanganic acid to manganese dioxide. Cool the solution by adding 50 cc. of water and filter through asbestos. Then make the volume up to 300 cc. and cool to tap-water temperature, when it is ready for the chromium titration after the addition of 5 cc. of syrupy phosphoric acid to decolorize the iron. To titrate, run in TV/100 ferrous sulphate solution until all chromium and vanadium are reduced. This can be discerned by testing a drop on a white plate with a drop of ferricyanide. If a blue color is obtained enough ferrous sulphate has been added. Add TV/100 permanganate until a pink color appears which persists on shaking, then add a few more cubic centimeters of TV/100 permanganate and stir the solution for a minute. Then add TV/100 ferrous sulphate until the pink just disappears. The total ferrous sulphate minus the permanganate multiplied by 0.0001733 equals the chromium present. Add to the solution enough ferrous sulphate to reduce the vana- dium and to have a considerable excess, mix the solution and add about 1 gram of 20-mesh ignited natural manganese dioxide. Shake the solution until a test with ferricyanide on a white plate shows that all ferrous iron has been oxidized. Then filter through asbestos (using suction) and titrate. Add TV/100 permanganate until a persistent pink color is obtained, then add several more cubic centimeters and shake the solution for a minute. Now add N / 100 ferrous sulphate until the pink color just disappears. The permanganate used minus the ferrous sulphate used multiplied by 0.00051 equals the vanadium present. NOTES ON THE PROCESS. The bismuthate oxidizes the chromium, vanadium and manganese to chromic acid, vanadic acid and permanganic acid. Boiling destroys the permanganic acid and 1 The Journal of Industrial and Engineering Chemistry, December, 1912. 124 CHROMIUM AND VANADIUM manganese dioxide precipitates and is filtered off. When ferrous sulphate is added, chromic and vanadic acids are reduced to trivalent chromium and quadrivalent vanadium; then when permanganate is added vanadium only is oxidized back and the ferrous sulphate minus the permanganate measures the chromium. Vanadium is again reduced by ferrous sulphate (not measured), the excess of ferrous sulphate oxidized by manganese dioxide, leaving the vanadium ready to be titrated. The following are some results obtained by the above method: PERCENTAGES CHROMIUM Present Found 0.078 0.086 0.155 0.160 0.155 0.157 0.233 0.236 0.078 0.078 0.233 0.225 3.100 3.095 A blank must be run on the chromium determination, as a small amount of MnO 2 persists in solution and runs uniformly the same. COPPER 125 DETERMINATION OF COPPER BY COLORIMETRIC METHOD Dissolve 10 grams of drillings in a mixture of 25 cc., 1.84 sp. gr., sulphuric acid and 250 cc. of distilled water, using a 500 cc. flask. Heat carefully until the borings have dissolved, dilute to 400 cc. with distilled water and add 0.5 gram zinc sulphide*, cork flask for a few minutes, filter on an 11 cm. paper, wash the residue with hydrogen sulphide water, open paper against side of funnel, add 20 cc. of hot nitric acid, 1.18 sp. gr., to the residue on the paper, allowing the solu- tion to run into the flask in which the borings had been dissolved. Wash the paper with 2% nitric acid solution, evaporate the filtrate to about 15 cc., remove from hot plate and add ammonia water (1:3) just sufficient to precipitate the ferric hydroxide. Filter into a 100 cc. Nessler tube and w r ash with hot water. The presence of copper will be indicated by the blue color of the filtrate from the ferric hydroxide. To another Nessler tube add about 50 cc. of distilled water and 5 cc. of (1 :3) ammonia water. Then add from a burette a' standard copper solution until the colors match when diluted to the same volume. Modification. It is sometimes found convenient to modify the method for determining copper. After determining the sulphur by the evolution method the hydrochloric acid solution can be used for the determination of this element as follows: The solution from the determination of sulphur is neutralized with ammonia until there is a slight precipitate of ferrous hydroxide. Acidify with 5 cc. of hydrochloric acid, heat to boiling point, add 0.5 gram zinc sulphide, when dissolved dilute to 400 cc. with distilled water, cork flask for a few minutes, and filter rapidly on an 11 cm. paper. Wash with hydrogen sulphide water and finish the determina- tion as previously described. Caution The success of the colorimetric method for determining copper depends upon carefully following each detail. Some- times a green color will be obtained instead of a blue. This is usually due to the use of too much ammonia and can be corrected by acidifying the green solution with dilute sulphuric acid (1:1) and then making the solution faintly ammoniacal. * The use of zinc sulphide was suggested to us by W. F. Clark, Dunston on Tyne, England. 126 COPPER Standard Copper Solution. The standard copper solution can be prepared by dissolving 7.856 grams of crystallized copper sul- phate in about 200 cc. of distilled water and 10 cc. of nitric acid, 1.42 sp. gr., and diluting to 2 liters. This solution can also be prepared by dissolving 2 grams of copper in 20 cc. of dilute nitric acid, and diluting to 2 liters. Each cubic centimeter represents 0.01 per cent copper when using 10 grams for analysis. COPPER 127 DETERMINATION OF COPPER BY IODIDE METHOD The colorimetric method for determining copper is not sufficiently accurate when this element is in excess of .15 per cent. When such is the case proceed as outlined for the determination of this element by the colorimetric method to the point where dilute ammonia is added to precipitate the ferric hydroxide. Instead of filtering into a 100 cc. Nessler tube use a 250 cc. beaker for this purpose. After washing a few times with hot water dissolve the ferric hydroxide with hot dilute hydrochloric acid, allowing the solution to run into the flask in which the original precipitation was made, wash a few times with hot water, add ammonia water 1:3 just sufficient to precipitate the ferric hydroxide and leave a very slight excess of ammonia present, heat to boiling, filter, and allow the filtrate to flow into the 250 cc. beaker containing the major portion of the copper. Add 5 cc. of. sulphuric acid, 1.84 sp. gr., evaporate on hot plate to dense fumes, cool, add 20 cc. of water, make slightly alkajine with ammonia, boil off excess ammonia then neutralize with glacial acetic acid adding 5 cc. in excess, cool and add 5 grams of potassium iodide crystals, then a few cubic centimeters of starch solution, and titrate with TV/20 sodium thiosulphate solution to the disappearance of the blue color. (12.4 grams per liter, 1 cc. thiosulphate equals .00318 gram copper). The thiosulphate can be conveniently standardized by using 25 cc. of the standard copper sulphate solution described in the Colorimet- ric Method for determining copper, adding the same reagents, and titrating under the same conditions as described in the regular method. If preferred the thiosulphate can be standardized by dissolving 5 grams of potassium iodide in 500 cc. of water, adding 25 cc. concen- trated hydrochloric acid, then exactly 25 cc. of standardized N/ 10 potassium bichromate solution. Add starch solution and titrate with the thiosulphate solution to the disappearance of the blue color. Each cc. of N/ 10 potassium bichromate solution equals .00636 gram copper. Standard Starch Solution. The starch solution used in this method is prepared as described under the determination of sulphur in iron and steel, Page 191. 128 COPPER Limestone is used in Open Hearth furnaces for fluxing the impurities in the manu- facture of ARMCO Ingot Iron products. This shows method of sampling each carload received HYDROGEN 129 DETERMINATION OF HYDROGEN The method for determining hydrogen by heating the metal in a partial vacuum and measuring the liberated hydrogen is very la- borious. The hydrogen existing as ammonia would not be estimated by the above method. Our method is based on the fact that hydrogen is liberated by heating the metal to a red heat in an atmosphere of oxygen. The hydrogen is oxidized to water, which is absorbed in phosphoric anhy- dride. The apparatus used consists of a 12" gas blast furnace for burning the sample, and a 12" electric furnace for purifying the oxygen gas. The oxygen gas is passed through a M" silica tube (T-2) at the rate of 25 cc. per minute. This furnace is heated to about 750 C., which is sufficient to purify the oxygen gas. The impurities are ab- sorbed by passing the gas first into a solution of caustic potash (K-2) and then through a bottle containing caustic soda (K), and finally through a tube containing phosphoric anhydride opened up in glass wool (P-l). The purified oxygen gas then enters the rw' silica tube (T-l) where it combines with the hydrogen which is liberated from the metal, thus forming water which is absorbed in a 4" glass stop- cock U tube containing phosphoric anhydride opened up with glass wool (P-2). This weighed 4" U tube is connected with the 7 A" silica tube (T-l), after the sample has been placed in the combustion tube. The weighed U tube is then connected with bottle containing sulphuric acid (Cone.). The combustion tube is heated to a temperature not to exceed 800 C., as a higher temperature and a higher rate of oxygen than prescribed may generate enough heat to damage the apparatus. The oxygen passes through the entire apparatus at the rate of 25 cc. per minute for 30 minutes, and from 10 to 40 grams of drillings are used for a determination. After the test is complete there should be some metal w r hich has not been oxidized, as it has been found unnecessary to oxidize all of the metal in obtaining accurate results. . After having ignited the sample for 30 minutes, the weighed U tube (P-2) is disconnected, and connected with an aspirator for the purpose of replacing the oxygen in the tube with dry air. 130 HYDROGEN HYDROGEN 131 A suitable aspirator for this purpose consists of a one gallon aspirator bottle filled with water connected with absorbing tubes as shown in photograph of apparatus for the determination of oxygen and carbon monoxide in iron and steel on page 158. The increased weight of the U tube which is due to the water which has been absorbed is multiplied by .11190 then by 100, and divided by the weight of sample taken. This gives the percentage of hydrogen in the metal. 132 HYDROGEN Chemist making a calorimeter test on coal to determine its heating value IRON 133 GRAVIMETRIC DETERMINATION OF IRON Dissolve about 1 gram 1 of the sample, accurately weighed, in a 600 cc. Jena glass beaker 2 , on the water bath, using 25 cc. of a 10% solution of hydrochloric acid. When solution is complete, add 200 cc. of distilled water, heat on water bath to about 80 C., and pass a moderate current of hydrogen sulphide gas through the solution for 20 minutes. Remove from water bath, add 200 cc. more of cold distilled water, and continue the stream of hydrogen sulphide for another 20 minutes, or until the solution is cold. Filter from the precipitate and wash thoroughly with hydrogen sulphide water containing a small amount of hydrochloric acid 3 , collecting the nitrate in an 800 cc. Jena glass beaker 2 . Test the residue for iron 4 . Evaporate the nitrate on the water bath until the volume is reduced to about 200 cc. and all the hydrogen sulphide is driven off. Then add 5 cc. of concentrated nitric acid and 10 cc. of concentrated hydrochloric acid , and heat to about 90 C., on the water bath and add a slight excess of warm dilute ammonia 6 . Allow the precipitate to settle, decant through a 15 ctm. ashless filter, transfer the precipitate to the filter 7 and wash with boiling water until 5 cc. of the washings give no opalescence with silver nitrate 8 . (Collect the filtrate and the first washings in a clean 800 cc. beaker and reserve for determination of manganese). . Dry the ferric hydroxide precipitate at 95-100 C 9 ., and then separate as perfectly as possible from the filter paper, in a room free from draught, placing the dry ferric hydroxide in a small porcelain dish on a white glazed paper. Cover the porcelain dish with a watch glass and then ignite the filter paper in a weighed porcelain crucible. Transfer the precipitate from the porcelain dish to the crucible; cover with a platinum cover and ignite for 10 minutes over a Bunsen burner; remove the cover, incline the crucible slightly and ignite for another 10 minutes. Place in desiccator, cool and weigh. Repeat the ignition until the weight remains constant, taking care not to heat more than a few minutes at a time and not at too high temperature 10 . In the meantime evaporate the filtrate for the determination of manganese, to about 200 cc. Add ammonia, heat to boiling and pre- cipitate the manganese with a saturated solution of bromine. Boil 134 IRON for a few minutes, filter on a small ashless filter, ignite and weigh as Mn 3 O 4 . This weight subtracted from the weight of Mn 3 O 4 in the sample, calculated from the total manganese, gives the amount of Mn 3 O 4 in the ferric oxide. To determine silica in the ignited precipitate, transfer this to a small platinum dish and digest with concentrated hydrochloric acid on the water bath and evaporate to dry ness 11 . Redissolve, dilute with hot water and filter from the silica on a small ashless filter. Wash with hot dilute hydrochloric acid and hot water until the silica is free from iron. Ignite in a platinum crucible, cool and weigh, and then evaporate with 2 drops of sulphuric and 1 cc. of hydrofluoric acid; ignite, cool and weigh. The difference between the two weighings then repre- sents the amount of silica in the ferric oxide 12 . The total amount of silica, manganese oxide, chromic oxide, phosphoric acid and alumina (calculated from analysis), subtracted from the weight of impure ferric oxide, gives the weight of pure Fe 2 O 3 , which contains 69.94% iron (Fe = 55.84). NOTES ON GRAVIMETRIC DETERMINATION OF IRON (1). By employing a power-driven centrifuge of large capacity the washing of precipitate can be perfectly and quickly performed largely by decantation, thereby enabling the operator to use a sample as large as 5 grams. (2). Beakers, funnels, watch glasses and glass rods must be thoroughly cleaned with warm concentrated hydrochloric acid before use, in order to prevent any foreign iron from entering the solutions. (3). The presence of acid is necessary to secure a perfect re- moval of iron from the residue and the filter paper. The hydrogen sulphide removes the following elements: Silver, lead, mercury, gold, platinum, tin, antimony, .arsenic, copper, cadmium, bismuth, molybdenum, tellurium, selenium, germanium, iridium, osmium, palladium, rhodium, ruthenium, tungsten, vanadium. (4). If great accuracy is desired, the analysis should be rejected whenever more than traces of iron are detected in the sulphide residue. (5). The nitric acid is added to oxidize the iron, and the hy- drochloric acid to secure sufficient ammonium chloride to keep zinc, IRON 135 cobalt, nickel and part of the manganese in solution. In nickel steels the precipitation should be repeated several times. (6). A large excess of ammonia is objectionable, as it causes some of the iron to become colloidal. (7). Ferric hydroxide adheres to the beaker and the glass rod. In order to remove this quantitatively, a few drops of concentrated nitric acid are introduced into the beaker and by means of the glass rod all ferric hydroxide is easily brought in solution. After dilution with water the iron is reprecipitated with ammonia and transferred to the filter. Nitric acid is used in order to prevent introduction of chlorides. (See 8). (8). When iron is precipitated with ammonia, small amounts of basic iron salts are always thrown down with the hydroxide. The amount and the composition of the basic salts vary according to the conditions. Thus in a solution of sulphate of iron larger amounts of basic salts are formed than from solutions of ferric nitrate or chloride. In a cold solution more basic salts are formed than when the solution is nearly boiling, 'and in addition the ferric hydroxide tends to assume a colloidal state, especially in the presence of a large excess of ammonia. Basic chloride of iron is volatile on ignition, hence the necessity of eliminating chlorides. Warm water decomposes chloride of iron, leaving the hydroxide free from chlorine. (9). The filter paper will become brittle if heated at a tempera- ture above 100 C. Dry at least 10 hours. Heat gradually in crucible. (10). When ferric oxide is ignited at too high a temperature, some magnetic oxide of iron (Fe 3 O 4 ) is always formed, causing low results. The formation of magnetic oxide takes place much more readily when the ignition is performed in a platinum crucible. A convenient arrangement consists of placing a small porcelain crucible within a covered platinum crucible, whereby the contact with plati- num is avoided and the danger of overheating greatly reduced; at the same time the disadvantage of using a porcelain crucible alone is overcome. (11) If the oxide has been heated to a high temperature it is difficult to dissolve in concentrated hydrochloric acid. To secure solution in a reasonable length of time in cases when the oxide has been overheated, it is advisable to grind the oxide carefully in an 136 IRON agate mortar and then ignite it for a minute, reweigh and determine the silica on this portion and from the result calculate the silica in the original amount of oxide. (12). In all exact gravimetric determinations of iron, allowance must be made for silica in the ferric oxide. Most steel and iron contain silicon, and considerable amounts are always dissolved from the glassware during the chemical operations. Glass is readily attacked by warm ammonia. Part of the silica is undoubtedly derived from the ammonia which ordinarily has been in contact with common glass. Most of the phosphorus is evolved during the solution in hydrochloric acid. If, however, the sample contains more than a few thousandths of 1% of phosphorus, the ferric oxide should be analyzed for phos- phorus. This can be done in the filtrate from the determination of silica, and the amount found, figured to phosphoric anhydride, added to the other impurities. MANGANESE 137 MANGANESE PERSULPHATE METHOD FOR THE DETERMINATION OF MANGANESE This method is of more value in the determination of manganese in steel than in pure American Ingot Iron. However, with proper care accurate results can be obtained on American Ingot Iron. The method is as follows: Dissolve .5 gram of the sample in a 250 cc. Erlenmeyer flask using 25 cc. (1.20 Sp. Gr.) nitric acid, heat on water bath until brown fumes are gone. Remove flask from the water bath and add 40 to 50 cc. N/ 100 silver nitrate, return flask to water bath and heat to 50 to 60 degrees C. Add about 2 grams of crystallized ammonium per- sulphate and maintain the solution at 50 to 60 degrees C., for a few minutes. Cool, dilute with 100 cc. distilled water and titrate with sodium arsenite to pale green color. It is necessary to keep the temperature between 50 and 60 degrees C. during the few minutes required for the perfect oxidation of the manganese, otherwise results will be erratic, especially when analyzing American Ingot Iron. 138 MANGANESE AMERICAN INGOT IRON Sulphur Phosphorus Carbon Manganese Silicon Copper Oxygen Nitrogen Iron .032 .008 .011 .017 trace .030 .025 .003 99.874 Microstructure and Analysis MANGANESE 139 DETERMINATION OF MANGANESE BISMUTHATE METHOD This method was perfected by Professor D. J. Demorest, of the Ohio State University and is substantially as follows: Dissolve 1 gram of the sample in 30 cc. of Nitric Acid (Sp. Gr. 1.13) and boil until brown fumes disappear. After cooling somewhat one-half gram of sodium bismuthate is added, a little at a time, until the resulting permanganic acid or manganese dioxide persists after a few minutes boiling. Now add 3 cc. of a 5% solution of potassium nitrite to reduce the manganese compounds, and boil the solution a few minutes to expel the nitrous fumes. Cool to tap water temperature and when cold add sodium bis- muthate, a little at a time, while the solution is shaken, until about }/ NEWBURYPORT LINK Sulphur Phosphorus Carbon Manganese Silicon Copper Oxygen Nitrogen Iron .014 .023 .040 .008 .028 trace .027 .003 99.867 After 100 years service Microstructure and Analysis MANGANESE 143 TEST TO INDICATE WHETHER METAL IS IRON OR STEEL In making this test without the use of a balance the following table can be used w r hen metal is in sheet form and the gauge is known. These weights represent the number of grams in 1 square inch: Grams Grams Grams Gauge Sq. Inch Gauge Sq. Inch Gauge Sq. Inch 12 14.00 17 7.10 22 4.00 13 12.00 18 6.40 23 3.61 14 10.00 19 5.66 24 3.20 15 9.00 20 4.82 25 2.80 16 8.00 21 4.41 26 2.41 As an illustration, suppose we have 16-gauge material. A square inch weighs 8 grams, hence a strip /^"xl" weighs approximately 1 gram, or /4 r/ x}x>" would also equal 1 gram. If the sheet is gal- vanized the coating need not be removed before making the test. Take equal portions of American Ingot Iron and sample to be tested, equal to }/ gram, and place in separate 10"xl" test tubes. Add to each tube 15 cc. 1 of dilute nitric acid, 1.18 specific gravity 2 . Place a test tube in holder, using care to incline the tube away from spectators while being heated with an alcohol lamp. The metal will disappear in a few minutes, but continue heating until no more brown fumes are given off. Allow solution to cool and heat the other tube in same manner and cool. The solutions can be compared at this point, the darker one containing the highest percentage of carbon. To each test add J/2 gram sodium bismuthate 3 and agitate for a few minutes. Then add sufficient water to half fill the tubes and mix thoroughly. Allow the tubes to rest for several minutes until the undissolved sodium bismuthate settles. By comparing the clear solutions, American Ingot Iron will show a light pink color, while steel will yield a purple color due to manganese present. 1 If no graduate is available, the volume of acid can be estimated by noting the depth of 1" diameter tube; each inch is equal to about 10 cc. 2 Nitric Acid of 1.18 specific gravity can be prepared by adding 1 part 1.42 specific gravity nitric acid to 2 parts water. 3 The amount of sodium bismuthate that can be placed on a dime represents about H gram. 144 MANGANESE -"- -j*. -v BESSEMER STEEL Sulphur Phosphorus Carbon Manganese Silicon Copper Oxygen Nitrogen Iron .050 .100 .120 .480 trace .010 .025 .010 99.2O5 Microstructure and Analysis MOLYBDENUM 145 MOLYBDENUM THE DETERMINATION OF MOLYBDENUM (AND COPPER) IN STEEL BUREAU OF STANDARDS Dissolve enough steel to give .06 to .1 g. molybdenum, in 1:3 nitric acid. When solution is complete, add 10 cc. of sulphuric acid and evaporate to fumes. Cool and dilute. If the residue is light colored and evidently silica no filtration is required. In case the residue is dark colored or indicates tungstic acid, filter it off, ignite it carefully in platinum, fuse the residue with sodium carbonate, thoroughly extract the melt with water, add two grams tartaric acid to the filtered water extract, acidify to two per cent by volume sulphuric acid and saturate with hydrogen sulphide. Digest one hour at approximately 50 C., filter off any dark sulphide and wash with a one per cent by volume sulphuric acid saturated with hydrogen sulphide and containing a little tartaric acid. Dissolve the precipitate in aqua regia and add the resultant solution to the solution of the sulphides which is obtained as described below. Dilute the main solution to 200 cc., add o grams tartaric acid and adjust the acidity to Yf/o sulphuric acid (by volume). Pass in hydrogen sulphide until the iron is reduced, the molybdenum (and copper) is precipitated, and the solution is saturated with gas. Digest at 50-60 for one hour or longer. Filter, preferably on a Gooch, and wash with the wash water specified above. Dissolve the pre- cipitate in aqua regia, unite with any recovery obtained as above and reserve the solution. Occasionally the filtrate from the precipitated sulphides contains some molybdenum which was not precipitated on account of reduction of the molybdenum. It is desirable to test this filtrate as follows: boil out most of the hydrogen sulphide, oxidize the iron and molyb- denum by means of bromine water, boil out excess bromine and again proceed with the hydrogen sulphide precipitation. In case molyb- denum sulphide is indicated it is to be recovered as in the regular procedure and added to the reserved solution. 146 MOLYBDENUM The reserved solution will contain the molybdenum and copper originally present in the steel and may contain a little iron. Treat this solution with 5 cc. of sulphuric acid and evaporate to fumes. Dilute and in case the solution is colored by reduced molybdenum, oxidize with a little permanganate solution. Then add a 10% solution of sodium hydroxide until in slight excess. Boil, filter, and wash with hot 1% sodium hydroxide solution. The insoluble con- tains the copper together with a little iron, and the copper may be determined electrolytically. The nitrate contains the molybdenum and this is most con- veniently determined as follows: acidify the solution, add sulphuric acid until it contains 3% by volume, warm the solution and reduce in a Jones' reductor containing a solution of ferric alum and phos- phoric acid in the receiver. The molybdenum is thus reduced to the trivalent condition in the reductor and is partially oxidized by the ferric sulphate in the receiver with the formation of an equivalent amount of ferrous sulphate. Titrate the resultant solution with tenth normal permanganate. The molybdenum and ferrous sulphate are oxidized to their higher valencies by the permanganate and the calculations are based on complete reduction to Mo 2 O 3 and subse- quent oxidation to MoO 3 . NICKEL 147 BRUNCK'S 1 GRAVIMETRIC DETERMINATION OF NICKEL Dissolve 1 gram of steel or iron in a 150 cc. Erlenmeyer flask with the use of 30 cc. of dilute nitric acid, (1.20 Sp. Gr.). Boil until brown fumes are expelled. Place 5 grams of pow^dered citric acid in an 800 cc. beaker. Add the solution in the Erlenmeyer flask to the dry citric acid, wash the flask thoroughly with water, and make the volume in the beaker up with water to 300-500 cc., depending upon the amount of nickel present. The higher the nickel the more water necessary. Make faintly alkaline with ammonia, then acid with acetic acid. Note: Acetic acid is preferred to hydrochloric on account of a more perfect separation of nickel from manganese when manganese is present. The faintly acetic acid solution is heated to near the boiling point, the beaker is removed from the source of heat, and from 15 to 25 cc. (depending upon the amount of nickel present) of a 1% alco- holic solution of dimethylglyoxime added. The solution is then made faintly alkaline with dilute ammonia, the nickel being precipitated as scarlet nickel glyoxime. The solution is kept hot for about 1 hour and filtered on a weighed Gooch crucible, washed thoroughly with hot water, and dried at 110-120 C. for 45 minutes. This scarlet pre- cipitate contains 20.31% nickel. If the percentage of nickel is low we use 3 grams of the sample which we dissolve in 50 cc. of nitric acid (Sp. Gr. 1.20). We also use 15 grams of citric acid. The presence of chromium or cobalt does not interfere with the precipi- tation. 1 Stahl u. Eisen; 28 (1) p. 331; 1908. 148 NICKEL View of milling machine showing milling of sheet bars for the determination of oxygen and carbon monoxide NICKEL 149 DETERMINATION OF NICKEL IN STEEL TITRATION METHOD The determination of nickel in steel can be finished in less than 25 minutes with the use of the method devised by the Buckeye Steel Casting Company. Dissolve 1 gram of steel or iron in a 150 cc. Erlenmeyer flask with the use of 30 cc. of dilute nitric acid, (1.20 Sp. Gr.). Boil until brown fumes are expelled. Wash the solution into an 800 cc. beaker, add 20 cc. of Citric Acid solution. Make faintly alkaline with ammonia, and acidify slightly with acetic acid. (Acetic acid is preferred to hydrochloric on account of a more perfect separation of nickel from manganese when manganese is present.) The faintly acetic acid solution is heated to near the boiling point, the beaker is removed from the source of heat, and from 15 to 25 cc. (depending upon the amount of nickel present) of dimethylglyoxime solution added. The nickel being precipitated as scarlet nickel glyoxime. Bring to boil and filter immediately washing the precipitate from the filter with the use of hot water into a 250 cc. beaker. Dissolve the nickel glyoxime in 20 cc. of aqua regia, bring to boil in order to decompose the glyoxime. Dilute with about 50 cc. of cold water, make faintly ammoniacal and cool in ice water. Make solution up to 150 cc. with cold water, add 10 cc. of potassium iodide solution, then 1 cc. of silver-nitrate solution and titrate with potassium cyanide solution (be sure to do this under the hood on account of the poisonous nature of the hydrocyanic acid) until the turbidity disappears and the solution clears. The potassium solution cyanide is standardized with the use of Bureau of Standards nickel steel, or with the use of .2 to .3 grams of nickel ammonium sulphate which has been added to one gram of iron or steel free from nickel and analyzed according to the above method. The following solutions are used in the determination of nickel by this method : Potassium Iodide Dissolve 8 grams of potassium iodide in one liter of distilled water. Silver Nitrate Solution Dissolve 5 grams of crystallized silver nitrate in one liter of dis- tilled water. 150 NICKEL Potassium Cyanide Solution Dissolve 13 to 14 grams of potassium cyanide in water and when in solution add 5 grams of potassium hydroxide which has also been dissolved in water and dilute the solution to one liter. Potassium cyanide containing sulfide cannot be used; as it forms a precipitate of silver sulfide which is not dissolved by potassium cyanide. Dimethylglyoxime Solution Dissolve 20 grams of dimethylglyoxime in 1300 cc. of concentrated ammonia, make the solution up to 2 liters with the use of 700 cc. of distilled water. Aqua Regia Solution As this acid decomposes upon standing it is desirable to make up fresh solutions each day; mixing 80% concentrated nitric acid and 20% concentrated hydrochloric acid. Citric Acid Solution Dissolve 600 grams of citric acid in one liter of distilled water. In standardizing the Potassium Cyanide Solution the blank produced from 1 cc. of silver nitrate solution should be deducted before determining the strength of the potassium cyanide, and this blank amounting usually to .3 cc. should be deducted from each de- termination. It will be satisfactory to determine the blank using the same amount of water (made ammoniacal) as the volume of a determination (200 cc.) using about 10 cc. of potassium iodide and 1 cc. of silver nitrate, titrating with potassium cyanide until the solution clears. NITROGEN :5l NITROGEN DETERMINATION OF NITROGEN We use the Allen method perfected by Professor J. W. Langley. This method is based on the reaction by which the combined nitrogen in iron or steel is estimated as ammonia by the solution of the metal in hydrochloric acid. The reagents required are: Hydrochloric Acid of 1.1 specific gravity, free from ammonia, which may be prepared by distilling pure hydrochloric acid gas into distilled water free from ammonia. To do this, take a large flask fitted with a rubber stopper carrying a separatory funnel tube and an evolution tube. Place in the flask strong hydrochloric acid, connect the evolution tube with a wash bottle connected with a bottle con- taining the distilled water. Admit strong sulphuric acid free from nitrous acid to the flask through the funnel tube, apply heat as re- quired, and distil the gas into the prepared water. Test the acid by admitting some of it into the distilling apparatus, described further on, and distilling it from an excess of pure caustic soda, or determine the amount of ammonia in a portion of hydrochloric acid of 1.1 specific gravity, and use the amount found as a correction. Solution of Caustic Soda, made by dissolving 300 grams of fused caustic soda in 500 cc. of water, and digesting it for 24 hours at 50 C., on a copper-zinc couple prepared by rolling together about 6 square inches each of zinc and copper foil. Nessler Reagent. Dissolve 35 grams of potassium iodide in a small quantity of distilled water, and add a strong solution of mercuric chloride little by little, shaking after each addition, until the red precipitate formed dissolves. Finally the precipitate formed will fail to dissolve; then stop the addition of the mercury salt and filter. Add to the filtrate 120 grams of caustic soda dissolved in a small amount of water, and dilute until the entire solution measures 1 liter. Add to this 5 cc. of a saturated aqueous solution of mercuric chloride, mix thoroughly, allow the precipitate formed to settle, and decant or siphon off the clear liquid into a glass-stoppered bottle. Standard Ammonia Solution. Dissolve .0382 gram of ammonium chloride in 1 liter of water. One cc. of this solution will equal .01 milligram of nitrogen. 152 NITROGEN Distilled water free from ammonia. If the ordinary distilled water contains ammonia, redistil it, reject the first portions coming over, and use the subsequent portions, which will be found free from ammonia. Several glass cylinders of colorless glass of about 160 cc. capacity are required. The best form of distilling apparatus consists of an Erlenmeyer flask of about 1500 cc. capacity, with a rubber stopper carrying a separatory funnel tube and an evolution tube, the latter connected with a condensing tube, around which passes a constant stream of cold water. The inside tube where it issues from the condenser should be sufficiently high to dip into one of the glass cylinders placed on the working table. The determination of nitrogen is made as follows: Place 40 cc. of the caustic soda, which has been treated with the copper-zinc couple, in the Erlenmeyer flask, add 500 cc. of water and about 2 grams, 20-mesh zinc to prevent bumping, and distil until the dis- tillate gives no reaction with the Nessler reagent. While this part of the operation is in progress, dissolve 3 grams of the carefully washed drillings in 30 cc. of the prepared hydrochloric acid, using heat if necessary. Transfer the solution to the bulb of the separatory funnel tube, and when the soda solution is free from ammonia, very slowly drop the ferrous chloride solution into the boiling solution in the flask. When about 50 cc. of water has been collected in the cylinder, remove it and substitute another cylinder. Place 1J/2 cc. of the Nessler reagent in a cylinder, dilute the distillate to 100 cc. with the special distilled water and pour it into the cylinder, containing the Nessler reagent. Take another cylinder, place therein 1^ cc. of the Nessler reagent and 100 cc. of the special distilled water to which 1 cc. of the ammonium chloride solution has been added, and compare the colors of the solutions in the two cylinders. If the solution in the cylinder containing the ammonium chloride solution is lighter in color than that in the cylinder containing the distallate, place lj/2 cc. of the Nessler reagent in another cylinder, pour into it 100 cc. of water containing 2 or more cc. of the ammonium chloride solution, and repeat this operation until the colors of the solu- tions in the two cylinders correspond after standing about 10 minutes. When about 100 cc. have distilled into the second cylinder, replace it and test as before. Continue the distillation until the water comes over free from ammonia, then add together the number of cc. of ammonia solution used, divide the sum by 3, and each .01 milligram will be equal to .001% of nitrogen in the steel. OXYGEN 153 OXYGEN CARBON MONOXIDE THE DETERMINATION OF OXYGEN AND CARBON MONOXIDE IN IRON AND STEEL Ledebur 1 about 40 years ago, proposed heating the borings in an atmosphere of hydrogen in order to determine the oxygen con- tent. The method as proposed was too laborious, consequently was not universally adopted. He recommended a preliminary heating of the borings in nitrogen. Up until the time Cushman 2 published his paper on the determination of oxygen and showed that an analysis could be made in less than an hour, very little use had been made of the method as proposed by Ledebur. However, since Cushman's paper appeared, a considerable amount of work has been done on this subject by many chemists. The Ledebur method for determining oxygen is recognized as having its limitations, but where manganese and silicon are low, such as in pure iron, we have found the method of great help in main- taining a uniform product. We have modified the Ledebur method so that we determine the oxygen and carbon monoxide in one opera- tion. For mill practice where samples can be taken from bars they should first be cleaned from all mill scale or surface oxide with the use of an emery wheel. The sample should then be placed in a milling machine which should be run at very slow speed in order to avoid oxidizing the millings. A light transverse cut should be taken en- tirely across the bar and the millings discarded in order to remove any oxidized cavities which were not removed by the emery wheel. The sample should be the average of the entire cross section if possible, as there is some difference in gas content between the interior and exterior portions of bars. The sample must be free from all dirt, and samples should not be ground in the vicinity where a sample is being milled, on account of the danger of contamination from finely divided particles of oxide of iron. The millings should be removed from the sample by the use of a magnet, and placed in a dry glass stoppered bottle: It is of the utmost importance that millings of uniform size be used for analysis, 1 Leitfaden fur Eisenhutten Laboratories, Eighth Edition, 1908, page 139. 2 Determination of Oxygen in Iron and Steel, by Allerton S. Cushman, Journal of Industrial and Engineering Chemistry, June, 1911. 154 OXYGEN OXYGEN 155 those passing a twenty mesh and remaining upon a forty mesh sieve being selected. The millings should remain in the unstoppered bottle for half an hour in a desiccator containing concentrated sul- phuric acid. A 30-gram sample is placed in a J/2"x^"x6" platinum or pure iron boat, (For cast iron use a porcelain boat), which is placed in the %"x30" silica tube, "T". In most descriptions for determining oxygen by the Ledebur method, hydrogen is generated by the action of some acid upon zinc contained in a Kipp's generator. It has been found that hydrogen so prepared contains considerable carbon monoxide and carbon dioxide, whereas hydrogen produced by the electrolytic process and stored in tanks is practically free from these two gases, is much easier to handle, and is cheaper. APPARATUS FOR THE DETERMINATION OF OXYGEN AND CARBON MONOXIDE IN IRON AND STEEL . 29 Tank of Electrolytic Hydrogen. E Electric Preheating Furnace 850 C. T Silicia Tube Mx30". K -Bottle containing sodium hydroxide sticks. vS Bottle containing concentrated sulphuric acid. P Bottle containing phosphoric anhydride on glass wool. T! Silica tube 7 /8"x3Q" in which is placed boat containing sample. G Gas Furnace Run at 1000 C. PI Absorption Tube containing phosphoric anhydride opened up with glass wool. LP U Tube containing phosphoric anhydride used as a trap. I Glass tube containing iodine pentoxide. O Furnace heated by Bunsen Burner to 150 C. B Meyer Bulb containing barium hydroxide solution. The hydrogen is passed through a /4"x30" silica tube "T" con- tained in an electric furnace "E" heated to 850 C. It is then passed through a bottle "K" containing sticks of sodium or potassium hy- droxide, which removes water and carbon dioxide, then through a wash bottle "S" containing concentrated sulphuric acid, and then through a bottle "P" containing phosphoric anhydride opened up with glass wool. It then passes at the rate of 100 cc. per minute into the K"x30" silica tube "T". 156 OXYGEN Place the boat containing the millings in the silica tube "T", and insert stopper which is connected with a weighed U tube "P", con- taining phosphoric anhydride opened up with glass wool. This "U" tube is conneceted with a short length of rubber tubing to the glass tube "I", which passes through the small furnace "O" which is main- tained at a temperature of 150 C., with the use of a Bunsen burner. This tube contains iodine pentoxide which oxidizes the carbon monoxide to carbon dioxide, the latter being absorbed in a .2 "N" solution of barium hydroxide contained in a Meyer tube "B". The iodine which is formed by the reaction is absorbed by the barium hydroxide, but does not interfere with the precipitation of barium carbonate. The iodine acts upon the rubber tubing making it brittle, this action can be lessened by passing a glass rod greased with vaseline through the new rubber tubing. After heating for 30 minutes at 1000 C., the gas is turned off and the air blast allowed to cool the silica tube "T", for ten minutes. The absorption tube is detached from the apparatus and is con- nected with the aspirator shown in engraving. About 500 cc. of air purified by passing through stick potash, "K", sulphuric acid "S", and phosphoric anhydride "P" respectively, is passed through the weighed U tube "X". This is done for the purpose of displacing the hydrogen gas and prevents errors which may arise should the tubes be weighed filled with hydrogen, some of which may be displaced by air should the stopper become dislodged. Another advantage is that the U tubes are quickly cooled by aspirating air through them, so that the errors from weighing tubes at different temperatures are eliminated. The increased weight of the tube "P", due to the water which was absorbed is multiplied by .8888, divided by the weight taken and multiplied by 100, which gives the per cent of oxygen. The barium carbonate is filtered on an eleven cm. filter paper, the bulb and paper being washed with boiled distilled water free from carbon dioxide. The filtering should be done at a location where there is no fuel being burned, otherwise some carbonic acid gas would be absorbed. The filter paper containing the barium carbonate is ignited first at low temperature and finally at a red heat, and the white barium carbonate weighed. This figure is multiplied by .1418, divided by the weight taken and multiplied by 100 which gives the per cent of carbon monoxide present. OXYGEN 157 A blank determination is run on the apparatus each day, the apparatus being adjusted until the final blank on the U tube amounts to less than .003 grams. The final figure is subtracted from the results obtained from each determination. With the use of electrolytic hydrogen there will be no blank to be subtracted from the barium carbonate. The following results have been obtained on various samples of iron and steel: Material Oxygen Carbon Monoxide Bureau of Standards, No. 8a 064 .065 Bureau of Standards, No. 30 024 '.029 American Ingot Iron 027 .013 Basic Open Hearth Steel 024 .073 Puddled Iron 572 .076 Iron Link Newburyport Bridge 027 .020 -. J. R. Cain, Bureau of Standards Technologic Paper No. 118, in studying the Ledebur Method for determining Oxygen in iron and steel has developed an electrolytic method for producing pure hy- drogen. The following ig a description of the method: Electrolytic Hydrogen Generator and Reservoir Hydrogen gas was generated by the electrolysis of a saturated solution of barium hydroxide mixed with a 25 per cent solution of sodium hydroxide in a large pyrex glass U tube, using a platinum anode and a nickel cathode. Platinum and nickel were used as the electrode materials because they have a low oxygen and a low hy- drogen overvoltage, respectively. The generator (Page 160) consists of a U tube containing the elec- trolyte and submerged in a jar through which flows cold water. The bottom of the U is filled with sea sand to hinder the passage of dis- solved gas from one limb of the U to the other, as suggested by Lewis, Brighton, and Sebastian.* The current passing through the genera- tor is regulated by means of a rheostat in the circuit. During the course of the investigation the current used by the operator was 3.3 amperes, which liberates about 1.5 liters of hydrogen gas per hour. The hydrogen gas reservoir and pressure-maintaining bottle is connected to the cathode side of the generator, and to the anode side a small U tube of 10 mm. inside diameter is attached which con- tains mercury to balance the pressure in the receiving system. With this arrangement the generator operates automatically. As hydrogen is generated and delivered to the gas reservoir water is displaced from * J. A. C. S. 39, 1917, 2248 158 OXYGEN Determination of Oxygen in Iron and Steel APPARATUS FOR DISPLACING HYDROGEN IN ABSORPTION TUBE WITH DRY AIR K Tubes containing calcium chloride. S Tubes containing concentrated sulphuric acid. P Tubes containing phosphoric anhydride on glass wool. X Absorption U Tube, used for weighing the moisture obtained from oxygen in the sample. OXYGEN 159 the gas reservoir and forced into the water bottle above. The in- creased pressure thus produced forces down the liquid in the cathode side of the generator an amount approximately equal to the height to which the water level in the water bottle is raised. After a certain volume of hydrogen gas has been generated and stored in the reser- voir, the level of the electrolyte on the hydrogen side of the generator will be forced down out of contact with the cathode, thus automat- ically breaking the circuit. The pressure of the hydrogen that is de- sired for the experiment is regulated by adjusting the height of the water bottle, which then determines the amount of mercury that must be added to the U tube on the oxygen side of the generator. The hydrogen gas as it is drawn off from the reservoir for use is passed through the catalyzer and purifying train. 160 OXYGEN H,0 Wtt Electrolytic hydrogen venerator Electrolytic hydrogen generator PHOSPHORUS 161 PHOSPHORUS DETERMINATION OF PHOSPHORUS ALKALI TITRATION METHOD Dissolve 2 grams of the sample in 40 cc. of nitric acid, 1.18 specific gravity, using a 300 cc. Erlenmeyer flask. Heat on hot plate until metal is in solution and add 5 cc. of saturated solution permanganate of potash. Boil until brown precipitate is formed. Now add four cc. of hydrochloric acid, 1.20 specific gravity or a sufficient amount to clear the solution by boiling a few minutes. Avoid an excess of hy- drochloric acid as it interferes, with the precipitation of phosphorus when extremely low. Remove from hot plate, cool somewhat, and cautiously add ammonia, .90 specific gravity, shaking flask occasionally, until a heavy precipitate of ferric hydroxide is formed. Then add nitric acid, 1.42 specific gravity, shaking occasionally, until precipitate dissolves and a clear amber-colored solution is obtained. It is very essential that an excess of nitric acid should be avoided, as it interferes with the precipitation of phosphorus when this element exists in traces. Heat or cool solution to 85 C., and add 50 cc. of ammonium molybdate solution. Shake well and allow to stand at least J/ hour or until precipitate settles. Filter and wash with 2% nitric acid solution until free from iron, and finally with distilled water containing about 1 gram of potassium nitrate to liter until free from acid. Transfer filter and contents to tumbler containing 50 cc. of boiled distilled water. Disintegrate paper w r ith two stirring rods and add sufficient standard sodium hydroxide to dissolve the yellow precipitate and render the solution pink when phenolphtalein indicator is added. Now run in standard nitric acid until pink color disappears, then finish the titration with standard alkali, the end point being a faint pink color. TITRATIOX EXAMPLE Standard Alkali Standard Acid Last Reading 27 .5 cc. Last Reading 14.7 cc. First Reading 17 .3 First Reading 6.7 10.2cc. S.Occ. 2.2 cc. x .01 = .022% Phos. Ib2 PHOSPHORUS Standard Solutions The standard nitric acid and alkali used for titrating are about .15 normal. One cc. being equal to .01% Phosphorus when a 2-gram sample is used for analysis. A stock solution of sodium hydroxide is prepared by dissolving 192 grams of sodium hydroxide in water and adding enough barium hydroxide solution to precipitate all carbonates, then diluting to two liters. Use 50 cc. of stock solution diluted to two liters for the standard alkali solution. A stock solution of nitric acid for titrating can be prepared by mixing 367 cc. of concentrated nitric acid, specific gravity 1.42, with enough boiled distilled water free from carbon dioxide to make two liters. Use 50 cc. stock solution diluted to two liters for the standard acid solution. Nitric Acid for dissolving the sample can be prepared by adding 1 part of nitric acid, 1.42 specific gravity, to 2 parts of water. The specific gravity of this mixture will be very close to 1.18. Ammonium Molybdate Solution Preparation Place in 5-pint bottle: 500 cc. Cone. HNO 3 , Sp. Gr. 1.42. 1700 cc. Distilled Water. Place in 400 cc. beaker: 90 g. Molybdic Acid. 100 cc. Distilled Water. 100 cc. Cone. Ammonia. Add ammonium molybdate slowly to acid in bottle while stirring. Mix thoroughly then add 2 drops only saturated ammonium phosphate solution. Agitate and let settle. Use 40 cc. to 50 cc. of clear solution. PHOSPHORUS 163 MODIFICATION FOR DETERMINING PHOSPHORUS IN CHROME-VANADIUM STEEL On account of the vanadium interfering with the determination of phosphorus we use the method of Hagmaier, described in Metal- lurgical and Chemical Engineering, Vol. XI, p. 28. This method is about as follows: Dissolve 2 g. of the steel in aqua regia in a 4-in. casserole, evaporate to dryness and bake. Cool, dissolve in 35 cc. of concentrated hydrochloric acid, dilute with water and filter from silica. Reduce the filtrate with sulphurous acid. When entirely reduced add 5 cc. of 90% acetic acid and 10 cc. of a saturated solution of cerium chloride. Add dilute ammonia with constant stirring until the solution becomes turbid. Then heat the solution to boiling, allow to settle, and filter. The cerium phosphate will filter rapidly. Wash the precipitate several times with hot w r ater and then dissolve off the paper with hot 1 :1 nitric acid. Precipitate the phosphorus from this solution in the regular manner with ammonium molybdate and titrate with alkali as previ- ously described. Add ammonia very slowly, as it is impossible to obtain proper conditions if an excess is added and an attempt is made to neutralize with acid. Should an excess inadvertently be added it is best to start another determination instead of attempting to neu- tralize the excess of ammonia. 164 PHOSPHORUS i! O QJ I O CJ ' CQ PIN HOLE TEST 165 PIN HOLE TEST LEAD COATED, TIN AND TERNE PLATE Dr. Allerton S. Cushman has devised a very simple test to de- termine the number of pin holes per square foot. The test consists of exposing a full sized sheet to the action of distilled water. The pin holes appear as rust spots. The four sides of the sheet are bent so as to make a pan 1" deep. The pan is thoroughly cleaned with several applications of gasoline and then flooded to a depth of Y^' with distilled water. After one week's exposure the water is removed and the pin holes are counted. 166 PIN HOLE TEST -o a oj - i tn .19 S3 g.s /. o bfl C I SULPHUR 191 SULPHUR DETERMINATION OF SULPHUR / BY EVOLUTION Dissolve 5 grams of the sample in 100 cc. of hydrochloric acid, 1.10 specific gravity, contained in a 500 cc. flask fitted with rubber stopper containing thistle tube and educt tube, passing through a reflux condenser to prevent acid distilling over. The educt tube dips almost to the bottom of a 10"xl" test tube containing 50 cc. of cadmium chloride solution. A low flame is applied and flask heated until all metal has dissolved and all gas has been driven out of flask, as is evidenced by the steam condensing in cadmium chloride tube. The contents of test tube are washed into an 800 cc. beaker and sufficient water is added to bring the volume to 500 cc. Add 2 cc. of starch solution and 50 cc. of hydrochloric acid, 1.20 specific gravity. The solution is then titrated with standard iodine solution to blue color. The standard iodine solution is prepared by dissolving 8.4 grams of iodine and 20 grams of potassium iodide in 50 cc. of distilled water. When iodine is in solution dilute to 2 liters and standardize with steel of known sulphur content. One cc. should equal .01% of sulphur when using 5 grams. A solution of potassium iodate (KIO 3 ) can be used instead of the iodine and potassium iodide just described. The potassium iodate is more stable. The Starch Solution can be prepared along any of the following lines: (1) One gram arrowroot mixed in 10 cc. of cold water which is poured into 100 cc. of boiling water and immediately removed from the source of heat. (2) Dissolve 20 grams of soluble starch in 100 cc. of distilled water to which can be added 40 grams of potassium iodide, free from iodine. This mixture is then poured into 900 cc. of distilled water. Potas- sium iodide makes the starch more sensitive and it should not require 192 SULPHUR more than .2 of a cc. of iodine to give a permanent blue color in water containing a small amount of hydrochloric acid. (3) If soluble starch is not available, corn starch can be used as follows : To a mixture of 10 grams of corn starch and 50 cc. of distilled water, slowly add a solution containing 5 grams of caustic potash and 50 cc. of distilled water, until the starch changes to a clear paste. Dilute to 500 cc. with distilled water and add 10 grams of potassium iodide crystals, free from iodine. (4) Prepare 500 cc. of a saturated solution of sodium chloride, and also a solution containing 100 cc. of 80% acetic acid in which 5 grams of starch has been dissolved. Pour into the sodium chloride solution and boil until clear. Make up to 600 cc. with distilled water, using 2 cc. for each determination of sulphur. Chemist pulverizing sample of coal for chemical analysis SULPHUR 193 DETERMINATION OF SULPHUR GRAVIMETRIC METHOD We prefer the Gravimetric Method for the determination of sulphur where great accuracy is desired. We use the Bureau of Standards' method which is essentially as follows: Dissolve the sample (4.57 grams) in 250 cc. of copper-potassium chloride solution (300 g. KCL-CuCl 2 and 100 cc. HC1 per liter) and filter the residue on asbestos. Wash 2 or 3 times with 5% hy- drochloric acid and then return residue and asbestos pad to the beaker and treat with 20 cc. of nitric acid (Sp. Gr. 1.42). Heat and add KC1O 3 until all carbonaceous matter is destroyed. Add a little (5 cc.) hydrochloric acid to dissolve the precipitated manganese dioxide and filter through asbestos again. Evaporate the solution to dry ness, take up in 10 cc. of hydrochloric acid and evaporate to dryness again. Take up in 5 cc. of 2% hydrochloric acid and 20 cc. of water and filter through paper. Precipitate the sulphuric acid in the boiling filtrate with 2 cc. of hot 10% barium chloride solution. Digest a short time on the hot plate and filter. Wash the barium sulphate with water until free from chlorides, ignite slowly, and weigh. The weight of barium sulphate in grams multi- plied by 3 is equal to the percentage of sulphur. 194 SULPHUR 100,000 Ib. Riehle Testing Machine, showing arrangements for carrying out tensile tests at high temperature SULPHUR 195 DETERMINATION OF SULPHUR OXIDATION METHOD In a 400 cc. beaker dissolve 5 grams of the steel in a mixture of 40 cc. of nitiic acid, (Sp. Gr. 1.42) and 5 cc. of hydrochloric acid, (Sp. Gr. 1.20) add 0.5 grams of sodium carbonate and evaporate the solution to dryness. Add 40 cc. of hydrochloric acid, (Sp. Gr. 1.20) evaporate to dryness and bake at a moderate heat. After solution of the residue in 30,cc. of hydrochloric acid, (Sp. Gr. 1.20) and evapora- tion to syrupy consistency, add 2 to 4 cc. of hydrochloric acid (Sp. Gr. 1.20), and then 30 to 40 cc. of hot water. Filter and wash with cold water, the final volume not exceeding 100 cc. To the cold filtrate add 10 cc. of the barium chloride solution. Let stand at least 24 hours, filter on a 9-cm. paper, wash the precipitate first with a hot solution containing 10 cc. of hydrochloric acid, (Sp. Gr. 1.20), and 1 gram barium chloride to the liter, until free from iron; and then with hot w^ater till free from chloride. Ignite and weigh as barium sulphate. Keep the washings separate from the main filtrate and evaporate them to recover any dissolved barium sulphate. NOTE: A blank determination on all reagents used should be made and the results corrected accordingly. Barium Chloride Dissolve 100 grams of barium chloride in 1000 of distilled water. SULPHUR PhysicaljTesting'Machine capacity 100,000 Ibs., illustrating method used for checking the accuracy of the machine TIN AND TERNE PLATE 197 TIN PLATE METHOD FOR SAMPLING AND ANALYSIS OF TIN, TERNE AND LEAD-COATED SHEETS Four 2 by 4-in. pieces are cut, one from each end and each side of the sheet, parallel with the sides and equidistant from the ends, as shown on Page 198. One sheet from each grade or shipment is taken for analysis. These samples, before weighing, should be thoroughly cleaned with chloroform, carbon tetrachloride or gasoline. Each piece is then cut in half, marking one half "A" and the other half "B". The four pieces comprising lot A are then accurately weighed together, cut into small pieces about Y% in. square, thoroughly mixed, and used for the determination of tin and lead. The four pieces comprising lot B are reserved for the analysis of base metal and the direct de- termination of coating as a check on the analysis of lot A. A templet should be provided, made preferably from steel Y% in. thick and exactly 2 by 4 in. A scribe is used to accurately mark the sections to be cut. The templet is then used to subdivide the 2 by 4 in. specimens into two pieces, 2 by 2 in. The sections for analysis are then cut with tinner's shears. METHOD OF ANALYSIS DETERMINATION OF TIN Three 5-gram portions of the finely cut sample of lot A are placed into three 300 cc. Erlenmeyer flasks, each fitted with a one-hole rubber stopper containing a glass tube bent twice at right angles, one end of which projects through the rubber stopper for a short distance, the other end being long enough to reach almost to the bottom of a beaker, placed on a level with the flask, containing about 300 cc. of dilute sodium-bicarbonate solution. Add 75 cc. of concentrated hydrochloric acid, connect the flask with the stopper containing the glass tube, and place the flask on a hot-plate. Heat gradually at first until most of the metal is in solution. The long end of the glass tube, in the meantime, is submerged in the beaker. 198 TIN AND TERNE PLATE -r *T-# 4\ ! % _v <- > >l TIN AND TERNE PLATE 199 The hydrochloric acid solution is finally brought to boiling and when all the metal is dissolved the beaker containing dilute sodium- bicarbonate solution is replaced by one containing a saturated solution of the same. Remove the beaker and flask to a cool place. This will cause a small amunt of the sodium-bicarbonate to enter the flask and exclude the air. The solution is finally brought to a low temperature, preferably with ice water. This solution is then diluted to about 200 cc. with oxygen-free water which contains several cubic centimeters of starch solution, and titrated with N/20 iodine solution. We have found this strength of iodine solution to be the most satisfactory for this method. The distilled water free from oxygen is obtained in any of three ways: (1) By passing carbon dioxide through the cold distilled water; (2) By boiling vigorously and cooling; or (3) by adding a few cubic centimeters of concentrated hydrochloric acid to the water and then about 2 g. of sodium bicarbonate stirring vigorously. By running this determination in triplicate, the first titration serves as a control to indicate the number of cubic centimeters of iodine required, whence the two succeeding titrations may be made very rapidly and should check very closely. Standardizing the Iodine Solution: About 0.1 g. of pure tin and 4 g. of iron filings are dissolved in 75 cc. of concentrated hydrochloric acid, etc., as under the determina- tion of tin. One cubic centimeter of TV/20 iodine = 0.002975 g. of tin. Calculation Weight of tin' Wt. of tin on 5 g. x Wt. (g.) of 16 sq. in. - x 8.6421 = number of 5 pounds per case of 112 sheets, 20 by 28 in. DETERMINATION OF LEAD Dissolve 10 g. of the finely cut sample of lot A in 150 cc. of nitric acid (1:1). Heat until free from brown fumes and dilute to 1 liter and mix thoroughly. Take 100 cc. of this solution, add 10 cc. of concentrated nitric acid, electrolyze at a temperature of 50 to 60 C., using 1 to 2 amperes and 2.3 to 2.5 volts. The weight of PbO 2 is multiplied by 0.866. 200 TIN AND TERNE PLATE If the base metal contains an appreciable amount of manganese the lead should be determined as sulfate. Calculation Weight of lead: PbO 2 found (g.) x 0.866 x 10 = Pb; Pb x Wt. (g.) of 16 sq. in. x 8.6421 = number of pounds per case of 112 sheets, 20 by 28 in. DIRECT DETERMINATION OF THE WEIGHT OF COATING The remaining four pieces representing lot B are used for the analysis of the base metal and incidentally can be used for the direct determination of the weight of coating. The four 2 by 2-in. pieces are carefully weighed together and each piece is wrapped with a stiff platinum or nickel wire in such a manner that it may be placed in the acid in a horizontal position. Heat 60 cc. of concentrated sulphuric acid contained in a 400-cc Jena glass beaker to at least 250 C., immerse each piece separately in the hot acid for exactly 1 minute, and remove to a 600-cc. Jena beaker con- taining 50 cc. of distilled water. Immerse momentarily and rub the surface while washing with about 50 cc. more of distilled water, using a wash bottle for this purpose. The four samples are thoroughly dried, reweighed, and used for the analysis of base metal. The loss .in weight represents the coating and some iron. The sulphuric acid contained in the 400 cc. beaker is cooled and combined with the washings in the 600 cc. beaker. Two hundred cubic centi- meters of concentrated hydrochloric acid are added and the solution boiled for a few minutes. The solution is cooled, poured into a gradu- ated 500 cc. flask and filled to the mark with distilled water. DETERMINATION OF IRON Place lOOcc. of this solution in a 300cc. Erlenmeyer flask, add 1 cc. of a standard solution of potassium permanganate to oxidize the iron and tin, heat to boiling and reduce with a few drops of stannous chloride. Cool, pour into a liter beaker containing 400 cc. of dis- tilled water, add 25 cc. of mercuric chloride, followed by 10 cc. of phosphoric acid and manganese-sulphate solution, and titrate with TV/10 potassium permanganate. TIN AND TERNE PLATE 201 Calculation Four pieces 2 by 2 in. weigh 28.5686 g. Same after stripped in acid 24.1620 g. Loss, coating plus iron 4.4066 g. Iron as found by titration 0.4887 g. Weight of coating 3.9179 g. 3.9179 x 8.6421 = number of pounds per case of 112 sheets, 20 by 28 in. Tin in 100 cc. x 5 x 100 = percentage of tin. Weight of Coating Percentage of lead is obtained by difference. In the analysis of tin plate, the weight of coating is expressed in pounds per box, which is a half case, or 112 sheets 14 by 20 in.; hence to obtain the weight of coating per box on tin plate, the number of pounds as obtained above is divided by two. CHECK DETERMINATION OF TIN The remainder of the solution which has been used for the de- termination of iron can be used for the determination of tin as follows : Place three portions of 100 cc. each in three 300 cc. Erlenmeyer flasks. If any of the lead sulphate should or should not be removed in any of these portions, the accuracy of the tin determination is not affected. Add 1 g. of powdered antimony, connect with rubber stopper and glass tube described in the method of determination of tin in the sample of lot A, place on a hot-plate, using dilute sodium biarbonate solution as a trap, and boil until the solution becomes decolorized. Replace the dilute sodium-bicarbonate solution with a saturated solution of the same, remove from the hot-plate, cool, dilute and complete the determination as described under the first method. CONCLUSIONS We claim for this method that the sample shows a true average of the coating on the plate, since we have checked the coating very closely by this method and by sampling from the center of the sheet, even with such large samples as 10 by 10 in. When 5 g. of the sample are taken for the determination of tin, an area of about 2.5 sq. in. is re- presented in the case of 40-lb., 1C plate, and of about 3 sq. in. in the case of 25-lb. plate; while, of course, it is double this in the determina- tion of lead. Furthermore, the amount of sample taken here for 202 TIN AND TERNE PLATE analysis is a representative quantity from 16 sq. in. and not merely from one particular section of 2.5 of 3 sq. in.; and is as large as many laboratories are using and larger than most are using. In addition, we believe that this method is more truly average than any method we have investigated; and moreover, the sheet is not destroyed so far as usefulness is concerned, but may be sheared down to a smaller size. While it is not necessary to determine the weight of coating directly by the sulphuric acid method, in addition to the determination of the lead and tin (on lot A), it will, however, serve as a check, and should ^.gree very closely with it. Furthermore, this is an excellent method for stripping the coating preliminary to the analysis of the base metal. By running the determination of tin in triplicate, as described, the method is very rapid and accurate, whereas the method as now used by many laboratories in which the plate is dissolved in an atmos- phere of carbon dioxide in a graduated flask, cooled, diluted to volume and titrated in aliquots, involves many details and is not so rapid. In this method also, no antimony is needed for the reduction of tin, since the iron in the base metal accomplishes this; moreover, in the presence of the quantity of tin here involved the antimony would have a tendency to deposit back on the plate, retarding the solution of the tin and thus giving low results. With the use of a rotating anode the proposed method is very rapid and the entire determination can be finished in a reasonable length of time. TITANIUM 203 TITANIUM BUREAU OF STANDARDS' METHOD FOR THE DETERMINATION OF TITANIUM Titanium is determined by treating 5 grams of iron with 40 cc. of hydrochloric acid (1:1) and heating until all iron has gone into solu- tion. Dissolving in this manner, all but a negligible quantity of titanium remains in the insoluble residue. The filtrate is tested for titanium by extracting the iron with ether after oxidation with a small amount of nitric acid, using the method of Rothe (Stahl und Eisen, 12, 1052 (1892), and 13, 333 (1893),) and adding hydrogen peroxide to the extracted solution, after expelling the ether and oxidizing with nitric acid. In all cases only a faint coloration is obtained. The insoluble residue is filtered off and washed with hot water, and the filter paper and carbonaceous matter are burned. The residue in the crucible is treated with hydrofluoric acid and a little sulphuric acid, and all silicon volatilized. The residue is fused with sodium carbonate, treated with water, and acidified with sulphuric acid. A sufficient amount of ferric alum is added to the standard titanium solution to give the same tint as the sample when they are at the same time dilution, for it is found that the residue from the silica always contains a little iron along with the titanium. Hydrogen peroxide is added to the solution and standard and the comparison made in a Wolff colorimeter. Reagents Peroxide Solution: Dissolve 4 grams of sodium peroxide in 125 cc. dilute sulphuric acid (1 of acid to 3 of water), and dilute to 500 cc. Concentrated Standard Titanium Solution: One-fourth gram of a standard 20% carbonless ferro-titanium is dis- solved in 30 cc. of dilute sulphuric acid (1 acid to 3 water). When so- lution is complete it is oxidized by the least possible quantity of con- centrated nitric acid, boiled for a few minutes, cooled and diluted to 204 TITANIUM such a volume that 1 cc. will contain 0.0005 gram of titanium. When using a five-gram sample 1 cc. is therefore equal to 0.01% titanium. Dilute Standard Titanium Solution: This solution is made, just before making the determination, by diluting one volume of the concentrated standard titanium solution to ten volumes. One cc. of this solution contains 0.00005 gram of titanium and is equal to 0.001% of titanium when using a 5-gram sample. VANADIUM 205 VANADIUM DOUGHERTY'S 1 METHOD FOR THE DETERMINATION OF VANADIUM IN STEEL In the application of Johnson's 2 or similar methods for the determination of vanadium in steel, considerable difficulty is often experienced in producing a colorless or "old rose" shade with ferrous sulphate in the solution containing an excess of permanganate after the preliminary oxidation of the vanadium. To obviate this diffi- culty the following method has been developed, in which this oxida- tion of the vanadium is effected by a sufficient quantity of nitric acid alone or with ammonium persulphate. Method Treat 2 to 4 grams of the drillings in a 500 cc. Erlen- meyer flask, with 60 cc. of water and 10 cc. of concentrated sulphuric acid. After heating the solution nearly to boiling, until the reaction is complete, add 40 cc. of nitric acid (Sp. Gr. 1.20), and boil thoroughly for 10 minutes to oxidize the iron and vanadium and to expel the last traces of nitrous fumes. Cool the solution, add 60 cc. of cold sulphuric acid (1:2) and dilute in a 600 cc. beaker to 450 cc. Add 3 cc. of a freshly prepared 1% solution of potassium ferricyanide, and titrate rather rapidly, with constant stirring, with N/20 ferrous ammonium sulphate, to the appearance of the first dark blue color. The end point can best be observed by looking through the side of the beaker toward the bottom of the beaker placed directly before a window. Deduction of a blank of 0.4 cc. of the ferrous solution has been found necessary, and is independent of the weight of the sample, the presence of chromium, and of the carbon content up to 0.5 per cent. For steels with over 0.50 per cent carbon, the blanks are higher; and, moreover, with 4-gram samples of such steels, the end point is rendered indistinct by a turbidity which appears toward the end of the titration. This difficulty may be avoided by adding to the solu- tion immediately after the boiling with nitric acid as above, 60 cc. of 1:2 sulphuric acid and 5 to 8 grams of ammonium persulphate (which in the absence of silver nitrate will not oxidide the Cr. and Mn.), and continuing to boil for 15 minutes, so that all nitrous oxides and hydrogen peroxide are expelled. (Before this second boiling, George T. Dougherty, The Journal of Industrial and Engineering Chemistry, May, 1915. 2 C. M. Johnson, "Analysis of Special Steels." 206 VANADIUM wash down with hot water the persulphate which sticks to the glass.) Cool, dilute and titrate as above. After such treatment the blank is .35 cc. (instead of .4 cc.) for steels with under .5 per cent carbon, and .5 cc. for .60 to .70 carbon, and .6 cc. for .90 to 1.25 carbon steels. The blanks are the same with or without the persulphate treat- ment for steels of over .50 per cent carbon. The ferrous ammonium sulphate solution may be standardized against N/W permanganate, the strength of which has been de- termined with sodium oxalate. The iron value of the permanganate multiplied by .917 gives the vanadium value. If chromium is desired it should be determined on a separate portion, using the sodium bismuthate oxidation method. USEFUL DATA 207 USEFUL DATA To find circumference of a circle multiply diameter by 3.1416. To find diameter of a circle multiply circumference by .31831. To find area of a circle multiply square of diameter by .7854. To find area of a triangle multiply base by J/2 perpendicular height. To find surface of a sphere multiply square of diameter by 3.1416. To find solidity of a sphere multiply cube of diameter by .5236. Doubling the diameter of a pipe increases its capacity four times. A gallon of water (U. S. Standard) weighs 8 Ibs. /^ oz., and con- tains 231 cubic inches. A cubic foot of water contains 1728 cubic inches, iy% gallons and weighs 621/2 pounds. A standard horse power : The evaporation of 30 pounds of water per hour from a feed water temperature of 100 deg. F., into steam at 70 pounds gauge pressure. To find capacity of tanks any size; given dimensions of a cylinder in inches, to find its capacity in U. S. gallons: Square the diameter, multiply by the length and by .0034. 1 meter = 39.37 inches. 2.54 cm. = 1 inch 28316 cc. = 1 cubic foot. 29.573 cc = 1 fluid oz. 1000 cc. = 1.05668 quarts. 3785.43 cc. = 1 U. S. Gallon (231 cu. in.) 1 gram = 15.4324 grains = .035274 oz. avoirdupois. 1 kilo = 2.2046 pounds (avoirdupois). 28.35 grams = 1 oz. 453.59 grams = 1 pound. 208 USEFUL DATA CONVERSION TABLES OF FAHRENHEIT AND CENTIGRADE SCALES Cent. Fahr. Cent. Fahr. Cent. Fahr. Cent. Fahr. Cent. Fahr. 32 200 392 400 752 600 1112 800 1472 5 41 205 401 405 761 605 1121 805 1481 10 50 210 410 410 770 610 1130 810 1490 15 59 215 419 415 779 615 1139 815 1499 20 68 220 428 420 788 620 1148 820 1508 25 77 225 437 425 797 625 1157 825 1517 30 86 230 446 430 806 630 1166 830 1526 35 95 '235 455 435 815 635 1175 835 1535 40 104 240 464 440 824 640 1184 840 1544 45 113 245 473 445 833 645 1193 845 1553 50 122 250 482 450 842 650 1202 850 1562 55 131 255 491 455 851 655 1211 855 1571 60 140 260 500 460 860 660 1220 860 1580 65 149 265 509 465 869 665 1229 865 1589 70 158 270 518 470 878 670 1238 870 1598 75 167 275 527 475 887 675 1247 875 1607 80 176 280 536 480 896 680 1256 880 1616 85 185 285 545 485 905 685 1265 885 1625 90 194 290 554 490 914 690 1274 890 1634 95 203 295 563 495 923 695 1283 895 1643 100 212 300 572 500 932 700 1292 900 1652 105 221 305 581 505 941 705 1301 905 1661 110 230 310 590 510 950 710 1310 910 1670 115 239 315 599 515 959 715 1319 915 1679 120 248 320 608 520 968 720 1328 920 1688 125 257 325 617 525 977 725 1337 925 1697 130 266 330 626 530 986 730 1346 930 1706 135 275 335 635 535 995 735 1355 935 1715 140 284 340 644 540 1004 740 1364 940 1724 145 293 345 653 545 1013 745 1373 945 1733 150 302 350 662 550 1022 750 1382 950 1742 155 311 355 671 555 1031 755 1391 955 1751 160 320 360 680 560 1040 760 1400 960 1760 165 329 365 689 565 1049 765 1409 965 1769 170 338 370 698 570 1058 770 1418 970 1778 175 347 375 707 575 1067 775 1427 975 1787 180 356 380 716 580 1076 780 1436 980 1796 185 365 385 725 585 1085 785 1445 985 1805 190 374 390 734 590 1094 790 1454 990 1814 195 383 395 743 595 1103 795 1463 995 1823 USEFUL DATA 209 CONVERSION TABLES OF FAHRENHEIT AND CENTIGRADE SCALES Cent. Fahr. Cent. Fahr. Cent. Fahr. Cent. Fahr. 1000 1832 1190 2174 1380 2516 1570 2858 1005 1841 1195 2183 1385 2525 1575 2867 1010 1850 1200 2192 1390 2534 1580 2876 1015 1859 1205 2201 1395 2543 1585 2885 1020 1868 1210 2210 1400 2552 1590 2894 1025 1877 1215 2219 1405 2561 1595 2903 1030 1886 1220 2228 1410 2570 1600 2912 1035 1895 1225 2237 1415 2579 1040 1904 1230 2246 1420 2588 1045 1913 1235 2255 1425 2597 1050 1922 1240 2264 1430 2606 .... 1055 1931 1245 2273 1435 2615 1060 1940 1250 2282 1440 2624 1065 1949 1255 2291 1445 2633 1070 1958 1260 2300 1450 2642 lOJS 1967 1265 v 2309 1455 2651 1080 1976 1270 2318 1460 2660 1085 1985 1275 2327 1465 2669 1090 1994 1280 2336 1470 2678 1095 2003 1285 2345 1475 2687 1100 2012 1290 2354 1480 2696 1105 2021 1295 2363 1485 2705 .... 1110 2030 1300 2372 1490 2714 .... .... 1115 2039 1305 2381 1495 2723 1120 2048 1310 2390 1500 2732 1125 2057 1315 2399 1505 2741 1130 2066 1320 2408 1510 2750 1135 2075 1325 2417 1515 2759 1140 2084 1330 2426 1520 2768 1145 2093 1335 2435 1525 2777 1150 2102 1340 2444 1530 2786 1155 2111 1345 2453 1535 2795 1160 2120 1350 2462 1540 2804 .... 1165 2129 1355 2471 1545 2813 .... 1170 2138 1360 2480 1550 2822 1175 2147 1365 2489 1555 2831 1180 2156 1370 2498 1560 2840 1185 2165 1375 2507 1565 2849 .... CENTIGRADE AND FAHRENHEIT CONVERSIONS To change Centigrade Temperatures to Fahrenheit Temperatures Add 40' multiply by 1.8, then subtract 40. To change Fahrenheit Temperatures to Centigrade Temperatures Add 40' multiply by .5555 (5/9), then subtract 40. 210 USEFUL DATA MELTING POINTS OF THE CHEMICAL ELEMENTS Element C F Element C F Element C F Helium > 271 259 253? 223 218 210 188 169 140 101.5 - 38.87 7.3 + 26 30 38 44 62.3 97.5 113.5 S, 112.8 S 1X 119.2 Smioe.s 155 186 217-220 231.9 271 302 > 456 434 423 369 360 346 306 272 220 150.7 37.97 + 18.9 79 86 100 111 144.1 207.5 236.3 235.0 246.6 224.2 311 367 423-428 449.4 520 576 CADMIUM. . LEAD ZINC 320.9 327.4 419.4 452 630.0 640 651 658.7 700 810 810? >Ca1700 / < Mo. 1720 1755 ? isoo <1850 1950 2200-2500? 2350? 2450? 2550 2700? 2900 3400 >3600 2696 2714 2786 2822 2939 3090 3090 >3009 6500 Barium Praseodymium Germanium. . . SILVER GOLD Rhodium Boron Iridium Ruthenium . . . Molybdenum Osmium Tantalum .... TUNGSTEN . Carbon...' Sulphur Indium Lithium Selenium TIN COPPER. . . . Manganese . . . Beryllium (Glucinum) Samarium .... Scandium .... Silicon NICKEL Bismuth Thallium OTHER STANDARD TEMPERATURES Substance Phenomenon C F Variation with pressure (pressure in mm. of Hg.) OXYGEN 183 297 4 C 183 0+0 01258(p 760) CARBON DIOXIDE 78 5 109 3 0.0000079(p 760) 2 C 78 5 | 01595 (p 760) SODIUM SULPHATE Na 2 SO 4 +10H 2 O WATER Transformation in- to anhydrous salt Boiling 32.384 100 90.29 212 0. 00001 ll(p 760) 2 C 100 HO 03670(p 760) NAPHTHALENE . . ... do 217 96 424 33 0.00002046 (p 760) 2 C 217 96+0 058 (p 760) BENZOPHENONE SULPHUR ....do do 305.9 444 6 582.6 832 3 C=305 ! 9 +0 . 063 (p 760) C 411 6 f 0908 (p 760) 71.9 per cent Ag 28 . 1 per cent Cu SODIUM CHLORIDE Eutectic freezing . . Freezing 779 801 1434 1474 6.000047 (p 760) 2 USEFUL DATA 211 C/2 Q OK X OH CO LD CO 1T3 CD n cd CM LO CM a CD 2 CO I ' 212 USEFUL DATA AN ACT ESTABLISHING A STANDARD GAGE FOR SHEET AND PLATE IRON AND STEEL Be it enacted by the Senate and House of Representatives of the United States of America in Congress assembled, That for the purpose of securing uniformity, the following is established as the only standard gage for sheet and plate iron and steel in the United States of America, namely: Number of gage Approximate thickness in fractions of an inch Approximate thickness in decimal part? of an inch Approx- imate thick- ness in millimeters Vv eight per square foot in ounces avoirdu- pois Vv eight per square foot in pounds avoirdu pois Weight per square foot in kilograms Weight per s iuare meter in kilograms Weight per square meter in pounds avoirdu- pois 0000000 1-2 .5 12.7 320 20.0 9.0 97. 215. 000000 15-32 .46 11.9 300 18.7 8.5 91. 201. 00000 7-16 .43 11.1 280 17.5 7.9 85. 188. 0000 13-32 .40 10.3 260 16.2 7.3 79. 174. 000 3-8 .375 9.5 240 15 6.8 73. 161. 00 11-321 .343 8.7 220 13.7 6.2 67. 148. 5-16 .312 7.9 200 12.5 5.6 61. 134. 1 9-32 .281 7.1 180 11.2 5.1 54. 121. 2 17-64 .265 6.7 170 10.6 4.8 51. 114. 3 1-4 .25 6.3' 160 10 4.5 48. 107. 4 15-64 .234 5.9 150 9.3 4.2 45. 100. 5 7-32 .218 5.5 140 8.7 3.96 42. 94. 6 13-64 .203 5.1 130 8.1 3.68 39.6 87. 7 3-16 .187 4.7 120 7.5 3.40 36.6 80. 8 11-64 .171 4.3 110 6.8 3.11 33.5 74. 9 5-32 .156 3.96 100 6.2 . 2.83 30.5 67. 10 9-64 .140 3.57 90 5.6 2.55 27.4 60. 11 1-8 .125 3.17 80 5 2.26 24.4 53. 12 7-64 .109 2.77 70 4.3 .98 21.3 47. 13 3-32 .093 2.38 60 3.75 1.70 18.3 40. 14 5-64 .078 1.98 50 3.12 .41 15.2 33.6 15 9-128 .070 1.78 45 2.81 .27 13.7 30.2 16 1-16 .062 1.58 40 2.5 .13 12.2 26.9 17 9-160 .056 1.42 36 2.25 .02 10.9 24.2 18 1-20 .05 1.27 32 2 .to 9.7 21.5 19 7-160 .043 1.11 28 1.75 .79 8.5 18.8 20 3-80 .0375 .95 24 1.50 .68 7.3 16.1 21 11-320 .0343 .87 22 1.37 .62 6.7 14.8 22 1-32 .0312 .79 20 1.25 .56 6.1 13.4 23 9-320 .0281 .71 18 1.12 .51 5.4 12.1 24 1-40 .025 .63 16 1 .45 4.8 10.7 25 7-320 .0218 .55 14 .87 .396 4.2 9.4 26 3-160 .0187 .47 12 .75 .340 3.66 8.0 27 11-640 .0171 .43 11 .68 .311 3.35 7.4 28 1-64 .0156 .396 10 .62 .283 3.05 6.7 29 9-640 .0140 .357 9 .56 .255 2.74 6.0 30 1-80 .0125 .317 8 .5 .223 2.44 5.3 31 7-640 .0109 .277 7 .43 .198 2.13 4.7 32 13-1280 .0101 .257 6H .40 .184 1.98 4.3 33 3-320 .0093 .238 6 .375 .170 1.83 4.0 34 11-1280 .0085 .218 5H .343 .155 1.67 3.70 35 5-640 .0078 .198 5 .312 .141 1.52 3.36 36 9-1280 .0070 .178 4M .281 .12/ 1.37 3.03 37 17-2560 .0066 .168 4M .265 .120 1.29 2.87 38 1-160 .0062 .158 4 .25 .1 3 1 22 2.69 And on and after July first, eighteen hundred and ninety-three, the same and no other shall be used in determining duties and taxes levied by the United States of America on sheet and plate iron and steel. But this act shall not be construed to increase duties upon any articles which may be imported. Sec. 2. That the Secretary of the Treasury is authorized and required to prepare suitable standards in accordance herewith. Sec. 3. That in the practical use and application of the standard gage hereby established a variation of two and one-half per cent either way may be allowed. Approved, March 3, 1893. USEFUL DATA 213 ELECTROCHEMICAL SERIES POSITION IN ELECTROCHEMICAL SERIES OF VARIOUS SUBSTANCES, IN THE ORDER OF THE MOST POSITIVE FIRST 1 Caesium 17 Nickel 33 Rhodium 2 Rubidium v 18 Thallium 34 Platinum 3 Potassium 19 Indium 35 Osmium 4lSodium 20 Lead 36 Silicon S^Lithium 21 Cadmium 37 Carbon 6 Barium 22 Tin 38 Boron 7 Strontium 23 Bismuth 39 Nitrogen 8 Calcium 24 Copper 40 Arsenic 9 Magnesium 25 Hydrogen 41 Selenium 10 Aluminum 26 Mercury 42 Phosphorus 11 Chromium 27 Silver 43 Sulphur 12 Manganese 28 Antimony 44 Iodine 13 Zinc 29 Tellurium 45 Bromine 14*Gallium 30 Palladium 46 Chlorine 15 Iron 31 Gold 47 Oxygen 16 Cobalt 32 Iridium 48 Fluorine All elements preceding iron are electro-positive to iron. All following iron are electro-negative. 214 USEFUL DATA ATOMIC WEIGHTS INTERNATIONAL ATOMIC WEIGHTS, 1918 Symbol Atomic Weight Aluminium Al 27 . 1 Antimony Sb 120 .2 Argon A 39 .88 Arsenic As 74 .96 Barium : Ba 137 .37 Bismuth Bi 208 .0 Boron B 11.0 Bromine Br 79 .92 Cadmium Cd 112 .40 Caesium Cs 132.81 Calcium Ca 40 .07 Carbon C 12 .05 Cerium Ce 140 .25 Chlorine Cl 35.46 Chromium Cr 52 .0 Cobalt Co 58 .97 Columbium Cb 93 .1 Copper Cu 63 .57 Dysprosium Dy 162 .5 Erbium Er 167.7 Europium Eu 152 .0 Fluorine F 19 .0 Gadolinium Gd 157 .3 Gallium Ga 69 .9 Germanium Ge 72 .5 Glucinum Gl 9.1 Gold.... Au 197.2 Helium He 4.0 Holmium Ho 163 .5 Hydrogen H 1 .008 Indium In 114 .8 Iodine I 126.92 Iridium '. . Ir 193 . 1 Iron Fe 55.84 Krypton Kr 82 .92 Lanthanum La 139 .0 Lead Pb 207.20 Lithium Li 6 .94 Lutecium Lu 175 .0 Magnesium Mg 24 .32 Manganese Mn 54 .93 Mercury Hg 200.6 Molybdenum Mo 96 .0 Neodymium Nd 144 .3 USEFUL DATA 215 Symbol Atomic Weight Neon !: Ne 20.2 Nickel Ni 58.68 Niton (radium emanation) Nt 222 .4 Nitrogen N 14.01 Osmium Os 190 .9 Oxygen O 16 .00 Palladium Pd 106.7 Phosphorus P 31 .04 Platinum Pt 195 .2 Potassium K 39 . 10 Praseodymium Pr 140 .9 Radium Ra 226 .0 Rhodium Rh 102 .9 Rubidium Rb 85 .45 Ruthenium Ru 101 .7 Samarium Sa 150 .4 Scandium Sc 44.1 Selenium , Se 79 .2 Silicon Si 28.3 Silver Ag 107.88 Sodium : Na 23.00 Strontium Sr 87 .63 Sulfur S 32 .06 Tantalum Ta 181 .5 Tellurium Te 127 .5 Terbium Tb 159 .2 Thallium Tl 204.0 Thorium Th 232 .4 Thulium Tm 168.5 Tin Sn 118.7 Titanium Ti 48 . 1 Tungsten W 184.0 Uranium , U 238 .2 Vanadium V 51 .0 Xenon Xe 130.2 Ytterbium (Neoytterbium) Yb 173 .5 Yttrium Yt 88 .7 Zinc Zn 65.37 Zirconium.. Zr 90.6 216 INDEX INDEX Page Aging Oven, photograph of 40 Tests 43 Alkali Titration Method for the determination of Phosphorus 161 Allen Method, determination of Nitrogen in Iron and Steel 151 Alternating Stress Tests Landgraf-Turner 69 Testing Machine, photograph of 68 Aluminum in Iron and Steel, determination of: By Bureau of Standards Method 169 By Kichline Method '. 75 Armco Ingot Iron, micrograph and analysis of 138 Ancient Irons and Modern Research 11 Arsenic in Iron and Steel, determination of by Distillation Method 77 Atomic Weights, Table of 214 Bessemer Steel, micrograph and typical analysis of 144 Billhook of Ancient Origin, photograph and micrograph of 24 Bismuthate Method for the determination of Manganese 139 For the determination of Chromium 122 Boron in Iron and Steel, determination of 79 Brinell Hardness Test 67 Testing Machine, photograph of 62 Brunck's Method for the determination of Nickel 147 Bureau of Standards Method for Aluminum in Iron and Steel 169 Chromium in Iron and Steel 121 Melting Points of the Chemical Elements 210 Molybdenum in Iron and Steel 145 Silicon in Steel 169 Sulphur in Iron and Steel (Gravimetric Method) 193 Titanium in Steel 169 Vanadium in Iron and Steel 121 Zirconium in Iron and Steel 169 Burrows' Permeability Apparatus, photograph of 42 Cain and Maxwell Method for the determination of: Carbon in Iron and Steel 101 Cain Electrolytic Hydrogen Generator and Reservoir 157 Calibrating Thermocouple, photograph of 52 Cannon, Iron Band from analysis of 14 Carbon in Iron and Steel, determination of: By Cain and Maxwell 101 By Colorimetric Method 81 By Combustion Method 83 By Liquid Air Method 85 Carbon Monoxide in Iron and Steel, determination of 153 INDEX 217 Page Chemical Analysis 73 Elements, melting points of 210 Chisel of Ancient Origin Photograph and Micrograph of 20 Chrome-Vanadium Steel, determination of phosphorus by Hagmaier Method. . . 163 Chromium and Vanadium, determination of, by Bureau of Standards Method 121 Chromium in Iron and Steel, determination of: By Bismuthate Method 122 By Demorest Method 123 Conductivity and Permeability Tests, photograph of 44 Copper in Iron and Steel, determination of: Bureau of Standards Method 145 Colorimetric Method 125 Iodide Method 127 Core Loss Test 37 Corrosion, Research on 15 Cushman Method for the determination of Spelter Coating 179 Data Useful 207 Delhi, India, Iron Pillar of Photograph of 8 Delhi, India, Iron Pillar Analysis of 9 Demorest's Method, determination of Chromium and Vanadium in Iron and Steel 123 Dougherty's Method, determination of Vanadium in Iron and Steel 205 Ductility Tests Erichsen Method 69 Electric furnace, Experimental, photograph of 70 Electrical Steels Magnetic Testing of 37 Electrochemical Series, Table of 213 Electrolytic Resistance Method for determining Carbon 101 Elements, Chemical, melting points of 210 Periodic Classification of 211 Epstein Testing Coils 38 Erichsen Method, for Ductility Tests 69 Etching Methods 59 Evolution Method for the determination of Sulphur in Iron and Steel 191 Experimental Furnace Room 71 Heat Treatment 71 Fairbank's House, Iron Nails from Analysis of 18 Furnaces, Electric, Experimental, photograph of 70 Gauge, U. S. Standard for sheet, plate iron and steel 212 Gravimetric Method (Brunck's) for the determination of Nickel in Iron and Steel 147 Gravimetric Method for the determination of Sulphur in Iron and Steel 193 Gravimetric Method for the determination of Iron in Iron and Steel 133 Hagmaier Method for the determination of phosphorus in Chrome-Vanadium Steel 163 Hardness Tests 61 Heat Treatment, Experimental 71 Scientific 51 Hydrochloric Acid Method for the determination of Spelter Coating 183 218 INDEX Page Hydrogen in Iron and Steel, determination of 129 Hydrogen Electrolytic Generator and Reservoir J. R. Cain's 157 Ingots, Split 16 International Atomic Weights 214 Introduction 5 Iron, determination of in Iron by Gravimetric Method 133 Iron Nails taken from Grave 28 Iron Specimens of Old Iron: Band from British Cannon 14 Billhook 24 Chisel 20 From Merrimac Gunboat 26 Nails, of Ancient Origin 22 Nails, hand forged used in Mission 12 Nails, from Fairbank's House 18 Nails, from Bakersfield Weir 34 Newburyport Link 142 Pillar of Delhi, India 8 Yarning Tool 84 Kichline Method, determination of Aluminum in Iron and Steel 75 Landgraf-Turner Alternating Stress Tests, photograph of 68 Alternating Stress Tests 69 Lead-Coated Sheets, sampling and analysis of: Weight of Coating 197 Pin Hole Test 165 Magnetic Testing 37 Aging Tests 43 Core Loss Tests 37 Permeability Tests 43 Manganese in Iron and Steel, determination of: By Bismuthate Method 139 By Color 141 By Persulphate Method 137 Test to determine whether Metal is Iron or Steel. 143 Maxwell and Cain, Bureau of Standards Method for the determination of Carbon in Iron and Steel 101 Melting Points of the Chemical Elements 210 Merrimac Gunboat, iron from 26 Metallurgical Control 47 Micrographs: American Ingot Iron 138 Bessemer Steel 144 Billhook 24 Chisel 20 Nails 22 Newburyport Link 142 Showing the effect of Annealing on hard drawn Wire 64 Of Steel Properly and Improperly Annealed 54 Microscopic Tests 57 INDEX 219 Page Microscope, view of Apparatus 56 Molybdenum in Steel, determination of, Bureau of Standards Method 145 Nails, analysis of: Iron Nails of Ancient Origin 22 From Bakersfield Weir 34 From Fairbank's House 18 Hand Forged Nail used in Mission 12 Newburyport Link: Analysis of 142 Micrograph of 142 Nickel in Iron and Steel, determination of: Brunck's Gravimetric Method 147 Titration Method 149 Nitrogen in Iron and Steel, determination of 151 Optical Pyrometers 55 Oxidation Method for the determination of Sulphur in Iron and Steel 195 Oxygen in Iron and Steel, determination of 153 Permeability Apparatus, Burrows photograph of 42 Permeability and Conductivity Tests photograph of 44 Permeability Tests Electrical Steels 43 Periodic Classification of the Elements 211 Persulphate Method for the determination of Manganese in Iron and Steel. . . . 137 Phosphorus in Iron and Steel, determination of: Alkali Titration Method 161 Determination of in Chrome- Vanadium Steel 163 Physical Tests 67 Physical Testing Section, photograph of 60 Pin Hole Test Lead Coated Tin and Terne Plate 165 Polishing Equipment, photograph of 58 Polishing Methods 59 Pyrometry Optical and Thermoelectric '. 55 Recording Potentiometer, photograph of 50 Research on Corrosion 15 Scientific Heat Treatment 51 Section through Split Ingots 16 Sheet Gauge U. S. Standard for Sheets, Plate Iron and Steel 212 Silicon in Iron and Steel, determination of 167 Silicon in Steel, determination of, Bureau of Standards 169 Spelter Coating, determination of on sheets and wire 183 Cushman Method 179 Lead Acetate Method 189 Spikes, Failure of Steel, photograph and analysis of 32 Steel Pipe Failure, photograph and analysis of 30 Sulphur in Iron and Steel, determination of 191 Bureau of Standards Method 193 Evolution Method 191 Gravimetric Method 193 Oxidation Method.. . 195 220 INDEX Page Table of Electrochemical Series 213 Atomic Weights 214 Conversion of Fahrenheit and Centigrade Temperatures 208 Melting Points of the Chemical Elements 210 U. S. Standard Gauge for Sheet Metal 212 Tensile Tests 67 Terne Plate Analysis of Weight of Coating 197 Pin Hole Test 165 Testing Coils, Epstein Core Loss Tests, photograph of 38 Thermocouples, Calibrating photograph of 52 Thermoelectric Pyrometers 55 Tin Plate Analysis of Weight of Coating 197 Pin Hole Test 165 Titanium in Steel, determination of: Bureau of Standards Method 169-203 Titration Method for the determination of Nickel 149 Useful Data . . 207 Vanadium in Iron and Steel, determination of: Bureau of Standards Method 121 Demorest's Method 123 Dougherty's Method 205 Weights Atomic, Table of 214 Wire, Hard Drawn and Annealed, micrographs of 64 Yarning Tool photograph and analysis of 84 Yensen, T. D., determination of Carbon in Iron 85 Zirconium determination of Bureau of Standards Method . 169 - THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW AN INITIAL FINE OF 25 CENTS WILL BE ASSESSED FOR FAILURE TO RETURN THIS BOOK ON THE DATE DUE. THE PENALTY WILL INCREASE TO SO CENTS ON THE FOURTH DAY AND TO $I.OO ON THE SEVENTH DAY OVERDUE. 3 25 1933 APR 261933 NOV 161933 REC-D LD MAY 1 8 1951 3Fflb'58HJ| REC'D LD JAN 19 LD 21-50m-l,'33 UNIVERSITY OF CALIFORNIA LIBRARY