me EFFECT or CATHODIC AND' PRIES'SURE‘DIFFUSED HYDROGEN on THE mnemmw as: some PLAIN CARBON ' mas Thesis for the Degree of Ph; D, MICHIGAN STATE UNIVERSITY Douglas .3. Harvey 1955‘ , ‘ LL; i . . . .IDI.‘.u.‘,..nhut5il;..‘ This is to certify that the thesis entitled The Effect of Cathodic and Pressure-Diffused Hydrogen on the Hardenability of Some Plain Carbon Steels presented by Douglas Jack Harvey has been accepted towards fulfillment of the requirements for Ph.D. Metallurgical Eng. degree in Date November 28, 195.5 0-169 THE EFFECT OF CATHODIC AND PRESSURE-DIFFUSED HYDROGEN ON THE HARDENABILITY OF SOME PLAIN CAR BON STEELS By Douglas Jill Harvey AN ABSTRACT Submitted to the School of Advanced Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Metallurgical Engineering\ Year 1955 f,‘ Approved A ////4 :7“. W /In (I THESIS Douglas J. Harvey ABSTRACT In this work steels are hydrogenized using two methods: cath— ode charging, and heat-treating from a hydrOgen atmosphere. The hardenabilities of specimens cathodically charged before heat-treating are compared to the hardenabilities of uncharged specimens. It is concluded that the amount of hydrogen obtained from cathode charging (estimated to be 5 to 7 milliliters per 100 grams) does not meas- urably affect the depth of hardening. Specimens are austenitized in hydrogen at 15 atmospheres, and quenched. These pressure-hydrog- enated specimens show a greater depth of hardening than identical specimens heated in 1 atmosphere of nitrOgen. Hardenability is measured by the method of symmetrical U-curves. In this work three steels are used ranging in carbon content from 0.33 percent to 0.50 percent. Some conclusions drawn are: (1) Hydrogen in amounts on the order of 15 milliliters per 100 grams has a small but definite effect on the hardenability of steel. (2) The increase in hardenability brought about by hydrogen content is negligible in commercial heat-treating practice, as the hydrogen content of steel is ordinarily very low. (3) The hardenability increase brought about by hydrogen appears not to change with carbon content (as in the ii case pro 6 Douglas J. Harvey case of boron). Also, some observations concerning hydrogen- produced cracks are discussed. iii THE EFFECT OF CATHODIC AND PRESSURE-DIFFUSED HYDROGEN ON THE HARDENABILIT-Y OF SOME PLAIN CARBON STEELS BY DOUGLAS JILHARVEY A THESIS Submitted to the School of Advanced Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Metallurgical Engineering 1955 .- - .J —— * _ fl" f — —-.v.”.- hardc 3T8 C 9 ma it. 1 to (F ABSTRACT In this work steels are hydrogenized using two methods: cath- ode charging, and heat-treating from a hydrogen atmosphere. The hardenabilities of specimens cathodically charged before heat-treating are compared to the hardenabilities of uncharged specimens. It is concluded that the amount of hydrogen obtained from cathode charging (estimated to be 5 to 7 milliliters per 100 grams) does not meas- urably affect the depth of hardening. Specimens are austenitized in hydrogen at 15 atmospheres, and quenched. These pressure-hydrog- enated specimens show a greater depth of hardening than identical specimens heated in 1 atmosphere of nitrogen. Hardenability is measured by the method of symmetrical U-curves. In this work three steels are used ranging in carbon content from 0.33 percent to 0.50 percent. Some conclusions drawn are: (1) Hydrogen in amounts on the order of 15 milliliters per 100 grams has a small but definite effect on the hardenability of steel. (2) The increase in hardenability brought about by hydrogen content is negligible in commercial heat-treating practice, as the hydrogen content of steel is ordinarily very low. (3) The hardenability increase brought about by hydrogen appears not to change with carbon content (as in the ii case of boron). Also, some observations concerning hydrogen- produced cracks are discussed. iii ACKNOWLEDGMENTS The author expresses his thanks to the members of the Metallurgical Engineering faculty for their encouragement and for their many helpful suggestions concerning this investigation. iv INTRC HYDE ‘ill TABLE OF CONTENTS Solubility .................................. Hydrogen Entry Into Liquid Steel ............ I ..... HydrOgen Entry Into Solid Steel .................. Diffusion of Hydrogen in Steel ................... Effect of Hydrogen on Ductility and Impact Strength . . . . Underbead Cracking of Welds ................... Shatter Cracks or Flakes ...................... Effect of Hydrogen on the Transformation of Austenite . . EXPERIMENTAL PROCEDURE AND RESULTS Cathode Charging ............................ High-Pressure Hydrogenation .................... Furnace Operation ........................... Effect of Pressure—Diffused Hydrogen .............. Observations on Specimen Cracking ............... Grain Size Determination ...................... SUMMARY AND CONCLUSIONS .................... LITERATUR E CITED ........................... 8 10 13 14 15 16 17 22 2.2 35 42 43 45 45 80 83 LIST OF TABLES TABLE Page 1. Hydrogen Dissociation .................... 12 II. Composition of Steel Used .................. 23 III. Hardness (136° diamond 50 kilogram load) of C-1, C-3, and C-5, Cathodically Charged; and C-2, C-4, and C-6, Uncharged ............... 33 IV. Specimen C-10 Hardness Values ............. 47 V. Specimen C-12 Hardness Values ............. 48 VI. Specimen C-13 Hardness Values ............. 49 VII. Specimen C-14 Hardness Values ............. 50 VIII. Specimen C-15 Hardness Values ............. 51 IX. Specimen C-16 Hardness Values ............. 52 X. Specimen C-17 Hardness Values ............. 53 XI. Specimen C-18 Hardness Values ............. 54 XII. Specimen C-19 Hardness Values ............. 55 XIII. Specimen C—20 Hardness Values ............. 56 XIV. Specimen D-l Hardness Values .............. 57 XV. Specimen D-2 Hardness Values .............. 58 XVI. Specimen E-2 Hardness Values .............. 59 XVII. Specimen E-3 Hardness Values .............. 60 vi TABLE XVIII. Specimen E-4 Hardness Values ............. . XIX. Specimen E-5 Hardness Values .............. XX. G rain Count FIGURE 10. 11. 12. 13. 14. LIST OF FIGURES Fe-H solubility at 1 atmosphere ............. Iron-hydrogen .......................... Relationship between cooling time and (a) transformation product, (b) hardness, for Mn-Mo steel ........................ Relationship between end-of—transformation temperature and cooling time for Mn-Mo steel ................................. Influence of hydrogen on the depth of hardening on an unalloyed steel with 0.96 percent carbon, 0.13 percent silicon, and 0.28 percent manga- nese ................................. Steel No. 1, nital etch, 100x ................ Steel No. 2, nital etch, 100>< ................ Steel No. 3, nital etch, 100x ................ End—quench curves . . . ................... Typical specimen after hardness exam— ination, 5X ............................. Spray—quench fixture ...................... Half U-curves .......................... High-pressure furnace .................... Furnace shell with coil in place before intro- duction of refractory insulation and zirconia tube . . viii 19 19 21 24 25 26 27 29 30 34 36 37 FIGURE 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. Furnace heating coil .................... Assembly consisting of a 1 inch diameter by 4 inch long specimen with thermocouple in place, specimen holder, thermocouple seal tube, and furnace door ............... View of high-pressure furnace and auxiliary equipm ent ........................... View of control equipment for high—pressure furnace ............................. Half U—curves ........................ Half U-curves ........................ (a) Specimen C-IO, 5 percent nital etchant, 2X; (b) Specimen C-ZO, 5 percent nital etchant, 2X .......................... Specimens C-19 and C-20, 2X ............. Chip from Specimen C-19, 2X ............. Specimen C-18 illustrating typical frac— turing that occurred in all hydrogen-treated specimens ........................... Chips from Specimen C-18 ............... View of microcracking in hydrogen-treated Specimen E—4, picral etchant, IOOX ......... Photomicrograph of Specimen C-lO, picral etchant, 100X ......................... Specimen C-13, picral etchant, 100x ......... ix Page 37 38 40 41 64 65 66 67 67 68 69 70 71 72 73 FIGURE 30. 31. 32. 33. 34. 35. Specimen Specimen Specimen Specimen Specimen Specimen C—17, picral etchant, 100x ........... C-20, picral etchant, lOOX ........... D-l, picral etchant, IOOX D—Z, E-3, E-4, picral etchant, picral etchant, pic ral etchant , ooooooooooo Page 74 75 76 77 76’ 79 will 0| that t ertie INTRODUCTION‘ For several years it has been known that the element iron will occlude hydrogen. It has also been recognized for some time that the absorption of hydrogen will greatly affect the physical prop- erties of a steel. Since hydrogen definitely occurs in steel, it should, of course, be considered an alloying element even though its presence in any significant amount may be temporary. Hydrogen atoms are so small that they can diffuse freely through the iron lattice even at temperatures as low as -78°C. (1). Because of this active diffusion, the hydrogen content of a steel may change in a very short period of time. There is very little information available concerning the sol- ubility of hydrogen in steel. However, there is reliable information on the solubility of hydrogen in iron. The difference between hydro- gen solubility in iron and in plain carbon steel is probably quite small. There is some experimental evidence to support this belief (2). Results of Sieverts' early work concerning the solubility of hydrogen in iron are shown in Figure 1 (3). This work agrees well with four other more recent investigations (4, 5, 6, 7). , °C 2000 .13.... - _ I600 I200 .. , _____ 800 FE- H SOLUBILITY AT AT I ATM 400 _.____L o 4 e :2 I6 20 cc H2/IOO GM FE (APTII SIEVEI'I’S I’ll) Figure 1. Fe—H solubility at 1 atmosphere. Solubility curves prepared by Zapffe using the above work are shown in Figure 2. The similarity between this and Sieverts' work can be seen by comparing the curve for 1 atmosphere in Fig- ure 2 with the curve shown in Figure l. Sieverts and his colleagues confirmed that hydrogen dissolves atomically (8), and that the solubility varies with the pressure ac- cording to the following relationship (9): [H] = K‘(PH2)l/2 where [H] is the hydrogen concentration, P is the pressure, and K' is a constant. The solubility curves for pressures above and below 1 atmos- phere shown in Figure 2 were calculated using this formula. Also, data for the curves below 400°C. were similarly estimated, since no experimental data are available for the lower temperatures (10). There is some evidence that the hydrides FeH, FeH and 2. FeH3 exist at low temperatures (below 300°F.) (11). Also, Simons and Ham (12) attempted to explain the diffusion of hydrogen on the basis of hydride formation. Another investigator (13) observed a thin white constituent in the area near cracks which had been caused by hydrogen. This material was believed to be a hydrogen-rich phase, but no definite evidence such as an identification of the Fe-H Iron-ll ytlrogen BY CARL A. ZAPFFE" Re/a/ive Volumes of Hydrogen to Iron 0.4 08 /.2 I6 2.0 2.4 l I | I I I I I I | I I TI—I] °C A tom/c Percentage Hydrogen °F Weight Percentage Hydrogen l l l I I 1 l 5 I0 I5 20 25 30 Cubic Cenfimefers of Hydrogen per I00 Grams of Iran Figure 2. Iron-hydrogen. supposed compound was offered. In fact, there is doubt that any stable hydrides exist. Under certain conditions hydrogen may be present in iron and steel in quantities that considerably exceed equilibrium amounts. This overcharging may occur when a metal is exposed to cathodic or chemically liberated monoatomic hydrogen. It may also occur when a cold-worked metal is exposed to gaseous hydrOgen (14). A metal may acquire amounts of hydrogen considerably above the solu- bility limit by rapidly changing conditions of temperature or pressure. For example, if molten iron is saturated with hydrogen and then allowed to solidify rapidly and to cool at room temperature, the resulting product will contain a quantity of hydrogen several times the equilibrium amount. The overcharging in this case results from the increased solubility of hydrogen in liquid iron over solid iron and the increase in solubility brought about by high temperatures. Similar results may be obtained by heating a sample of iron in high-pressure hydrogen. This high-pressure hydrogeneration tech- nique was used by Hobsen and Sykes to study the effect of hydrogen On the ductility of low-alloy steel (15). A hydrogen content of 7.6 milliliters per 100 grams was obtained by heating a steel specimen for two hours at 600°C. in a hydrogen pressure of 56 atmospheres. It is widely accepted that hydrogen dissolves interstitially in iron and steel. It is also possible that hydrogen finds its way into dislocations or other lattice defects. Evidence for this is furnished by the ability of cold-worked metals to absorb a greater amount of gas than similar annealed metals. According to Hagg (16), to form an interstitial alloy, the ratio of the atomic diameter of the inter- stitial solute atom to the solvent atom must be less than 0.59, and the solvent must be one of the transition metals. It is generally believed that there are only four elements with small enough atoms to dissolve interstitially in iron (which, of course, is a transition metal) (17). The four interstitial elements are hydrogen, boron, carbon, and nitrogen. For several years many investigators have carefully studied the qualitative and quantitative effects of the various alloying ele- ments on the hardenability of steel. At the present time the harden- ability effects of most elements found in steel can be expressed in terms of a multiplying factor (18). This factor depends on the behavior and the percentage of the element. Hydrogen remains as One of the recognized important elements found in steel for which little or no hardenability information is available. Among the interstitial elements the influence of carbon is well documented. Boron has been found to have a very pronounced effect on the hardenability of steel even though the addition is very small. According to Grange and Garney (19), a boron content as low as 0.001 percent will give the maximum hardenability effect. The multiplying factor for this quantity of boron is approximately 1.5 (20). This means that 0.001 percent boron is equivalent to 0.1 per- cent molybdenum or 0.3 percent chromium. The effect of boron reaches a maximum, as does the influence of carbon in the presence of boron (21). It has been established that nitrogen will aid in the stabilization of austenite in stainless steels. For instance, the nickel content of 18-8 may be partly replaced by nitrogen (22). It might be expected then that nitrogen would arrest the austenite transforma- tion in ordinary carbon steels and thus have an effect on hardenability. However, this has not been proven because of certain difficulties such as low solubility and compound formation (23). In the light of the influence of the other interstitial elements it seems entirely possible that the element hydrogen might have an effect on the sta- bility of austenite or on the hardenability of steel. There is a limited amount of information concerning the in- fluence of hydrogen on the transformation of austenite (24, 25, 26). This information will be discussed in detail in the second part of this report. A533 HYDROGEN AND STEEL Solubility As stated previously, the equilibrium solubility of hydrogen in iron varies as the square root of the hydrogen pressure and in— creases with temperature. M. A.rmbruster (27) made a study of the solubility ‘of hydrogen in iron, nickel, and certain steels. This work indicated that the solubility-pressure relation of S ___ p1/2 applies to certain steels as well as to iron. In addition, it was found that the relation between temperature and solubility for solid iron can be expressed as follows: l2) 2 -(l454/T) + 1.946 Log (S/p1 where S is solubility in micromoles of hydrogen per 100 grams of iron, T is the absolute temperature, and p is pressure in millimeters of mercury. With a change of constants, this formula may be applied to steel. However, it was found that the solubility of hydrogen in low-alloy steels differs little from that of pure iron. The hydrogen content of a steel may be expressed by atomic percentage, weight percentage, micromoles per 100 grams, relative volumes of hydrogen to iron, or cubic centimeters of hydrogen per 8 . .n__. 100 grams of steel. The methods of relative volumes and cubic centimeters per 100 grams seem to be the most pOpular. Hydrogen may enter liquid steel and then become trapped when the steel solidifies, or may enter solid steel. It is generally believed that hydrogen enters solid steel as a charged atom. The charged atoms may be the result of cathodic deposition, a chemical reaction, or ordinary dissociation of the hydrogen molecule Hydrogen Entry Into Liquid Steel One of the major sources of hydrogen in liquid steel is water. This may be water vapor or water of hydration of components in the furnace charge. Hydrogen may also be introduced into liquid steel by ferro-alloys, and slag-making constituents. Furnace fuels may be a source of hydrogen, since they introduce hydrogen and hydrocarbon into the furnace atmosphere. Barraclough (28) found the hydrogen content of various steels sampled at tapping to vary from 4.6 to 10.3 milliliters per 100 grams. Ba rraclough‘s work indicated that the hydrogen content was more de- pendent upon the compOsition of the steel than on the kind of furnace or process used. However, he presented some evidence that indicated acid heats will have a slightly lower hydrogen content than basic heats of the same chemical composition. (,L. 10 Carney, Chipman, and Grant (29) reported in their work that the rate of solution of hydrogen into molten iron was very rapid. Only a few minutes were required for solution after hydrogen was introduced into the furnace atmosphere. The evolution of hydrogen from the molten metal was also rapid after removal of the hydrogen source from the furnace atmosphere. Mallett (30) found, in his in— vestigation concerning the introduction of hydrogen into molten metal during arc welding, that the hydrogen content of the resulting weld metal can be closely estimated from an analysis of gases in the arc atmosphere. His work indicates that hydrogen in the arc atmosphere results from the water gas reaction: CO+HO‘——_—COZ+H 2 2 He also found that certain arc welding atmospheres contain as much as 40 percent hydrogen, which accounts for the high hydrogen con- tent of some are welds. Hydrogen Entry Into Solid Steel In pickling Operations, the hydrogen atoms may be furnished by the following reaction (31): > ‘80 + 2H Fe + H2504 Fe 4 During the electrodeposition of a metal on steel there is an evolution of hydrogen ions on the steel surface. These hydrogen ions III I II II‘ ‘5' l I|. ll readily enter the iron lattice. Deliberate cathodic charging will cause the solution of a quantity of hydrogen several times the equi- librium amount at room temperature. Cathodic charging has been used by several investigators in their study of the effect of hydro- gen on the pr0perties of metals. It is believed by Smith (32) that steel will absorb hydrogen during heat-treatment from the steam reaction: >FeO+2H +H Fe ZO The reaction: Fe + ZHZO > Fe(OH)2 + 2H during rusting may cause the introduction of hydrogen into a steel. Ordinary dissociation of the hydrogen molecule H 2H 2 will produce atomic hydrogen for entry into steel placed in an at- mosphere of hydrogen. The amount of dissociation, and consequently the amount of hydrogen absorbed, will depend upon the temperature and pressure. Since this reaction causes an increase in volume, an increase in pressure will lower the degree of dissociation. How— ever, the dissociation is strongly endothermic and will be increased with high temperature. Giangue (33), using spectrographic data, calculated the dissociation values for hydrogen at various tempera- tures. Some of these values are given in Table I. It is evident l I 7 ‘ l I { |vt .l I ll 7.]! III. I ‘II I '51 ill 1 .II 1 i I III. .1111 Ill 1! ..II 11 till 411 ll 1 II III! 12 TABLE I HYDROGEN DISSOCIATION (from spectrographic data by Giangue) Degrees Percent Dissociation Fahrenheit at 1 Atmosphere - 4 77 (1.8 a 6.6) x 10 3 -19 435 (4.4 :i: 6.7) X 10 -7 1335 (1.3 :1: 0.8) X 10 -4 2235 (9.5 :I: 3.4) X 10 3135 0.086 :I: 0.011 4035 1.31 :i: 0.13 4935 8.1 :I: 0.65 5830 29.7 :1: 1.1 6740 63.3 :I: 2.2 7630 95.7 :i: 0.1 13 from this table that very little atomic hydrogen is available from dissociation for occlusion at room temperature. However, at higher temperatures there is a considerable amount available. Diffusion of Hydrogen in Steel Barrer (34) states that a hydrogen atom may diffuse inter- stitially as a proton through the metal lattice. There seems to be little doubt that hydrogen atoms, which dissolve interstitially, diffuse through the interstices of the iron lattice (35). However, there is some doubt as to whether hydrogen atoms diffuse as protons (posi- tively charged hydrogen atoms). X-ray data indicate a measurable amount of distortion of the ferrite lattice caused by the solution of hydrogen. If the hydrogen atoms dissolved as protons, there would be no such distortion of the iron lattice. Smith (36) made the following remarks concerning the diffu- sion of hydrogen in iron: 1. Diffusion of hydrogen, at least in iron, occurs at the same rate, through single-crystals and polycrystalline mass. 2. Diffusion is not facilitated by grain boundaries but is hin- dered when they are very numerous. 3. Diffusion is at least approximately prOportional to the square root of the impelling pressure difference. Like the similar relation for solubility, this probably fails for extreme conditions. (37] It has also been observed that diffusion occurs at an accelerated rate through stres sed metal. l4 Geller and Sun (38) calculated diffusion constants for iron and certain alloy steels. In general they found that alloy additions such as silicon, chromium, and nickel decrease diffusivity and that the diffusivity of hydrogen is much lower in gamma iron than in alpha iron at the same temperature. Another and more detailed investigation (39) gives further support to these principles. Effect of Hydrogen on Ductility and Impact Strength The reduction of ductility is the most significant effect of the solution of hydrOgen in iron and steel. Usually this reduction of ductility is proportional to the amount of hydrogen dissolved in the steel. However, at least one investigation (40) has shown that after a certain minimum ductility value is reached additional hydrogen has no further effect. The partial or almost complete effusion of hydro- gen from the metal will be accompanied by a complete return of ductility. The loss of ductility caused by hydrogen occlusion is often commercially eradicated by annealing the steel for a period of time sufficient to remove a major portion of the hydrogen (41). Sims and his colleagues (42) found that a 100 hour aging treatment at 400°F. was sufficient to return the ductility to a normal value and lower the hydrogen content from 0.28 to 0.04 relative vol- ume in a cast carbon steel. The exact amount of hydrogen necessary 15 to cause embrittlement depends upon a number of factors such as composition, cleanliness, thermal history, and degree of segregation of hydrogen within the specimen (43). In an investigation concerning the effect of hydrogen on the tensile properties of steel, Hobson and Hewitt (44) found that the significant factors are: "(1) Hydrogen content, (2) alloy type, (3) heat treatment, microstructure, and tensile strength, (4) previous history . . ., (5) rate and type of testing, (6) temperature (of testing), and (7) direction of stress." (45) They also found that with amounts of hydrogen usually found in finished steel (1 to 4 milliliters per 100 grams) the effect on ductility at room temperature should not be severe unless the steel is hardened and very lightly tempered or extremely spheriodized. Underbead Cracking of Welds The cracks that appear in the base metal adjacent to metal deposited by the metallic arc process are believed to result from hydrogen (46). There is fairly conclusive evidence that hydrogen is dissolved in the liquid metal during welding and then diffuses into the base metal (47). Most of this diffusion takes place when the zone near the weld is in the austenitic condition. Apparently by ill I 'II '1‘ ll I'll 7.1.1 I I." 1,- l I. 1 16 some little-understood mechanism, this hydrogen dissolved in the base metal causes "underbead cracks." Some of the characteristics of underbead cracks are as fol- lows: (1) They increase with increasing hardenability. (2) They form at room temperature. (3) They require a period of time to form. (4) Martensite must be present. (5) The arc atmosphere must contain hydrogen (which is to say hydrogen must be present in the base metal) (48). It appears that in some way, in a manner not completely explained, the hydrogen embrittles the untempered martensite which is then susceptible to cracking. Shatter Cracks or Flakes The occurrence of an abnormal fracture appearance in a steel is often considered as evidence of the presence of hydrogen. These areas of abnormal fracture have been called by such names as flakes, snow flakes, or fish eyes. They are referred to as shatter cracks when the failure occurs during rolling or forging. Segregation of hydrogen is believed to be an important factor in the formation of this type of fracture. Derge and Duncan (49) concluded that "thermal segregation" due to temperature gradients during cooling is a greater factor in the case of hydrogen distribution 17 than dendritic segregation during freezing. For example, samples from ingots air-cooled after pouring showed a greater degree of hydrogen segregation than similar ingots that were water-cooled. This higher hydrogen content of the center of the ingots is believed to be the cause of cracks during rolling or forging. Sims (50) attributes the occurrence of fish eyes to the dif- fusion and segregation of hydrogen into voids or discontinuities in the steel. He explains that the hydrogen will diffuse into the voids and form hydrogen molecules (H2). Carney, Chipman, and Grant (51) calculated that pressure as high as 218,500 pounds per square inch could be developed by virtue of hydrogen building up in these rifts or discontinuities in the steel. "It is postulated, therefore, that the molecular hydrogen present under high pressure in the cavities rushes into slip planes, as soon as slip starts, springs them apart and renders that part of the steel incapable of further plastic deformation." (52) Effect of Hydrogen on the Transformation of Austenite In his work on hard-zone cracking of welds, Cottrell (53) noted that presence of this type of failure could be related to the temperature for completion of the austenite transformation during cooling. Since it had also been established that the presence of 18 hydrogen was necessary for the formation of this type of crack, Cottrell decided to study .the effects of hydrogen on the transforma- tion of austenite. Small specimens measuring 0.25 inch long by 0.225 inch outside diameter by 0.150 inch inside diameter were used. These small, thin specimens were charged cathodically, induction- heated, and then cooled by a blast of nitrogen. Simultaneous tem- perature and dilatation measurements were recorded. Figure 3, taken from Cottrell's work, illustrates his findings on the relation- ship between cooling time and transformation product, and cooling time and hardness. This diagram shows that, for a given cooling time, the hydrogen treatment had little effect on the final hardness, but the presence of hydrogen seems to increase the amount of mar- tensite formed. In Cottrell's work the amount of martensite was estimated from dilatation. The work reported above also indicated that hydrogen has no effect on the MS temperature, but does have an effect on the end of transformation temperature. This effect is illustrated in Figure 4, taken from Cottrell's work. This work also indicated that, "when the steel is supersaturated with hydrogen im- mediately before the dilation test, there are more pauses in the transformation to martensite, and the temperature for completion of transformation is lowered considerably for a given cooling rate." (54) MARTENSITE,°/o LI' kg LOAD). D PN & O o I I I: L I i : 4 O O Untmoted l Hydrogen—treated I 0 Weld heat-affected zona‘ I N O O I I I I I I O;I 1 02K 03. 0-4 X“ IO AVERAGE HARDNE SS 1 '00 400 so 20 S.T|ME TO COOL FROM s7o°vo 300°C.,l¢C. A Figure 3. Relationship between cooling time and (a) transformation product, (b) hardness, for Mn—Mo steel (Cottrell). RATE OF COOLING AT END-OF—IRANSFORMATION TEMPERATURE,°C. 5. TIME TO COOL FROM to Figure 4. Relationship between end-of-transformation temperature and cooling time for Mn-Mo steel (Cottrell). I illllll lilltj 5 .Stundon but 1001)" aogliiht, flIlSClllIl‘Bt'nll in Wassrl' :IIILN‘I. 10 stunden bel 1000° gegluht, anschlIeBI-nd in \Vfln‘fll’ nbuel. ‘/l nat. Gr. geglflht in: ansnebrannter Hollkohle trockenem Wauerstoft Figure 5. Influence of hydrogen on the depth of hardening on an unalloyed steel with 0.96 percent carbon, 0.13 percent silicon, and 0.28 percent manganese (by Houdremont and Heller). 21 20 Houdremont and Heller (55) reported that hydrogen had an effect on the hardenability of a steel containing 0.96 percent carbon. In this work specimens were heated for various periods of time in burned-out carburizer or wet hydrogen at 1 atmosphere. The re- sults are shown in Figure 5. From these photographs taken from the work of Houdremont and Heller, it appears that the specimens heated in wet hydrogen hardened to a greater depth than the speci- mens heated in the burned carburizer. The specimens were all quenched in water. The photographs of the fractured specimens also indicate that ten hours in wet hydrogen give a greater depth of hard- ening than five hours. Harness U-curves presented in the work re- veal that the specimens were seriously decarburized on the surface from the long periods of heating. Houdremont has also reported a portion of this work in a book (56). EXPERIMENTAL PROCEDURE AND RESULTS Cathode Charging For this portion of the work, steel number 1 was used. The composition of this steel is given in Table II, and the microstruc— ture is shown in Figure 6. In the first phase of this work, harden- ability measurements were attempted using the Jominy end-quench method. Specimen A-l was cathodically charged for twenty-five hours in a 20 percent solution of H2504. After charging, the speci- men was rapidly heated in an agitated molten salt bath to 1550°F. It was determined experimentally that the temperature of the center of the 1 inch round bar would be within 10° of the temperature of the salt bath in three minutes. After heating, the specimen was transferred to the Jominy fixture and end-quenched. Specimen A-2 was heated in the same manner as Specimen A-l, but was uncharged. The resulting curves are shown in Figure 9. These first results were encouraging, but inconclusive. Fur- ther work with this procedure indicated that the curves were not reproducible. This was attributed to the insensitiveness of the end- Cluench method. Apparently, if cathodic hydrogen had an effect on hardenability, it was too small to show distinctly by this method. 22 23 TABLE II COMPOSITION OF STEEL USED (percentages) Steel Number Element 1 2 3 Carbon ...................... 0.50 0.42 0.33 Manganese .................... 0.77 0.76 0.78 Phosphorus ................... 0.015 0.046 0.010 Sulphur ...................... 0.038 0.034 0.026 Silicon ....................... 0.20 0 27 0.21 Nickel ....................... 0.04 0.07 0.05 Chromium .................... 0.04 0.20 0.09 Molybdenum ................... 0.05 0.06 0.05 Figure 6. Steel No. l, nital etch, lOOX. 24 25 lOOX. nital etch, Steel No. 2 Figure 7 . «a Figure 8. Steel No. 3, nital etch, lOOX. 26 27 I6 20 24 28 32 12 DISTANCE FROM QUENCHED END I/IS THS End—quench curves. Figure 9. 28 It was decided to use "symmetrical U-curves" (57), produced from fully quenched specimens, as a measure of hardenability. In the case of symmetrical U-curves, the two halves are mirror im- ages. Each half is produced by averaging hardness values taken along several radii. A. typical specimen after hardness examination is shown in Figure 10. All hardness tests were made with a Wilson "Tukon" ma- chine using a 136° diamond with a 50 kilogram load. This machine was fitted with a stage and specially built indexing fixture. The positions for the hardness measurements were accurately located using this equipment. Spray-quenching was used in order to obtain a reproducible quenching rate. The spray—quench fixture is shown in Figure 11. It consisted of a 4 inch pipe jacketed along a portion of its length with a 6 inch pipe. Water is forced at the specimen through 168 one-eighth inch holes arranged in eight rows. The holes are one- fourth inch apart in each row. When the 1 inch round by 4 inch long specimen is dropped down the 1 inch pipe, it is stopped and held in place by the microswitch actuating mechanism. The micro- switch starts the 3 horsepower centrifugal pump, which sprays the Specimen with 200 gallons of water per minute. The spray-quench equipment is shown, along with pressure furnace, in Figure 17. Figure 10. Typical specimen after hardness examination, 5X. 29 30 I II N ._ _ __ m , n _T-|-II.IU=I-lNI m w \ //////////////////// ' \\\\\ 4'2——--I llllllllllllllllllll \ V I ‘5‘ I r 4— \.\ v “\R I //////////////////X/ ///./A I. . .. ._._..._..._,.._...=_._.._.._..._.._..._,.._,.._..._.._.7////// _ ////// //// / ///////// \\\ .__ spscmsu I_\ WATER MANIFOLD Spray-quench fixture . Figure 11. 31 The water was recirculated from the 100 gallon tank shown in the photograph. A. 2 inch gate valve was used to control the water en- tering the quenching fixture. The half U-curves prepared from specimens quenched with this equipment indicated that the quenching rate was very uniform from specimen to specimen. A series of six specimens was quenched using this spray equipment. Specimens C-1, C-3, and C—5 were cathodically charged for twenty-four hours in 20 percent H250 using a current of 3 4 amperes. Because of acid attack, the Specimens were machined to the final 1 inch diameter after charging. All specimens were heated for 3.5 minutes in an agitated salt bath and then spray-quenched with water. The total time required for machining, heating, and quenching was less than fifteen minutes. This speed was thought necessary to prevent the loss of hydrogen. Specimens C-2, C-4, and C-6 were heated and quenched in a similar manner. The specimens were sectioned and hardness values were taken on twelve equally spaced radii. The hardness tests were taken every millimeter starting 0.5 millimeter from the surface of the specimen. A typical specimen after hardness examination is shown in Figure 10. Averages of the twelve sets of hardness values were used to produce half U-curves. The tabulated results of these six III 1 i] 32 specimens are shown in Table 111. Only the average values are given in this table. The two half U-curves shown in Figure 12 represent an average of the three charged specimens (C-1, C-3, C-5) and an average of the three uncharged specimens (C-2, C—4, C-6). Even though the charged specimens hardened to a greater depth, the dif- ference is so small as to be almost negligible. At this point it was concluded that the effect of cathodic hydrogen, if any, was very small. It was estimated, using data from another investigator (58), that the hydrogen content of the charged specimens was between 5 and 7 milliliters per 100 grams immediately after charging. Since the solubility of hydrogen in steel at 1550°F. is approximately 5 milliliters per 100 grams, and the heating time for the Specimens was so short, it was thought that there was adequate hydrogen available to nearly saturate the austenite. As a hydrogen content nearly equivalent to saturation at l at— mosphere had a very small or no effect, it was decided to use high- pressure hydrogenation. Some of the advantages; of this technique are: (1) increase in the hydrogen solubility, (2) longer austenization time, and (3) better temperature control. TABLE III 33 HARDNESS (136° DIAMOND 50 KILOGRAM LOAD) OF C-1, C—3, AND C-5, CATHODICALLY CHARGED; AND C-2, C-4, AND C-6, UNCHARGED Depth Specimen Numbe r Specim en Numbe r (nun) Avg. Avg. C-1 C-3 C-5 C-2 C-4 C-6 0.5 769 773 770 770 766 770 780 '772 1.5 738 745 742 742 739 737 745 740 2.5 694 691 686 690 689 689 691 690 135 598 598 587 594 587 598 585 590 4.5 440 453 436 443 430 446 426 434 5.5 347 355 348 350 343 350 341 345 6.5 318 324 318 320 317 318 323 319 7.5 308 313 307 309 309 301 304 304 8.5 307 308 303 306 306 299 302 302 9.5 302 305 301 302 301 299 299 302 10.5 294 298 298 297 298 294 295 296 iii-l I' will '1 ll! 34 DPH 800 ;, 700 5f eoo £ ;; 400 :31 fit; 1' :3 H 300 f' o 1| 2 3'4 51611-77118 9 I0 II DISTANCE FROM SUR ACE (MM) Figure 12. Half U-curves. 35 High-Pres sure Hydrogenation A. sketch of the pressure—tight furnace is shown in Figure 13. The furnace shell was made from a low carbon steel forging 8-1/2 inches in diameter and 11 inches long. The low carbon steel cover flange was welded on. The furnace cavity was bored to a size of 8-1/2 inches deep by 5 inches in diameter, leaving a wall thickness of 1-3/4 inches. The cover was held on by fifteen 1 inch studs and nuts. A. gasket was made from l/16 inch thick c0pper. Five lands 1/8 inch wide were turned into the furnace gasket seat as shown in Figure 14. A. 2 inch thick by 11-1/2 inch diameter mild steel cover was provided. A heating coil made from eighteen turns of 0.201 inch cromel wire was used. Nickel leads were welded on as shown in Figure 15. This coil had a cold resistance of 0.118 ohm. The coil leads were made gas—tight through the furnace wall by a pile Of mica washers held in place by jam-nuts. Shorting was prevented on either side of the mica pile by wrapping the coil leads with mica sheet. Insulation for the furnace was cut from silica insulation bricks. A 1-5/16 inch diameter zirconia refractory tube having a wall thick- ness of 1/8 inch was placed inside the coil to prevent the specimens from shorting the heating coil. pro}! I \\\\ E: 36 me fin monk SZOOEN E 52: II_ was; 3.2.59... E \\\NI \ox\v0\. /////%//////// zu!_owum .7; n ..N\H .DHMOmv oodfiudm endomoumIflmmm mH ousmfih \\\; 44mm x» 114.. «J bio: on \nbxu ltflaqonowv 1' 0': «52:225...- amlfiPJ. PR...” .onloouh .o\:I,4.I.m,H . ,0! . a.\ t, o0 o .fi 1. 19’90\ 55:55.. 9 (.0 Tor [ .ofi r3 ,MI "I. My . Zea/423120732527.4? 1,. Pop” rs 3.0 I 2:55-11:59! ............................................... I... 3 75777777. Lay/[5772. w. . . .l to .11 .9/1.“ ..f\hola. nI :.\\.a I I 0.099.71 .U" 0|! I 9;! au/’ :. I\ (V .5“, otfifi/fl) (\ifl I.~I A..qv/;.\0oI//s\(\ 32.53.53. anJ\Ion/1:”._h./o\9.\)\o.J:nsrjscslta 9.59 n D \01109)\\. .nI‘IIx...0..(\I\~ez~/9ro..c 0’)¢Il)Jfllh'~ EEEEEE. a. /. a m a < u.— mqaaoootcurh mo- '«r ‘. ‘ l‘ 4‘ if. k" l y y- — l- I \ . n ’ r - J: ». "I .L \ ._ ‘ f ,‘r |’ ..~ \ ‘ I r ’1' ~ 5_. ‘- k I ‘ - I .\.- " p I .J 0 . ‘ r ‘ ,v - . . ‘1 - t - J ' ’\ _ . ‘ . , “ - .3!“ ‘1 u I 'x * ~ I ' ,_ . / v- - 1 ‘ . . x. r ‘ )3 'Y ..-4 .-.¢:-- I .“I \~ . .' - ’ 0"; ' ' (“‘- ' .. . " 0 ‘ ‘ . ‘Ic, . ’ , " . \ 'X \ q I) I J " 'q. 0 ’- C ' ‘ ' I \ . ‘. .4 ‘ " ' 5 - l U . . ‘ . -- 1 k'\ 4;, e‘; ' I. y ': - ’Q- ‘ {‘ ., --? u; b“ .r . , 1 u .- . f-f‘ ' . ~ . . r-. '3 - ' . . , 1.; , g. " ['1 1 "p ‘, :a' 3.. '1‘ l— ‘ n. r . ‘1 .' If? . r n. o z ‘ . O) .K I a , (- 7’ ’v 1 p. n- “ I. ~ ' (“ ‘ h ’ .‘ ' \ ' \ 0v 1; »‘ ' ‘ .v; — , . .’ . - . ' J r ,0 Ir 1'1 l ‘ I .' v";“’ " j r . ' .l r, I 4' ‘I' ’j (' ' . a} --* ‘ -. .n .. ...._ , ' - ‘.- fl. w? . ' ' .. \ " 0" “‘3“ ~.’ I ‘ ‘ ' I ‘ - ‘ . J " . ' V' " ’ ‘ . 5' ’v' ' - 5"! . p ‘3 .‘ ‘ f' , v . ' <‘ ' . ' ' ' ' Ari", o- 1' 1 1".- . " - . ’2 l‘ - \‘1 \ ’. I 'r I P. ‘ ‘ . '.r ‘ h /< ,1 . . I: r / ' ' d“ . ' ’(I /¢ 1' ‘ .~ . ' ’ _4' . . (.1, ‘ / “Au. MA” .. o'..'~ '£.'. .' ‘ m. "'3. .. .. *‘ ' ’ ‘ . ‘ /—‘.=' \ 7. a r ' f! u u . ." I § '. ’,-'\ I‘_ r “ I - ‘ ' I I - v I ‘ ’ C/ s ‘, l b . . r c I a V .’ ' .51. "1 1 . '7' ‘v I ’6‘ t - rr' \ r . L- 'f " o . l ' \ A ‘ § ’ . ' ' ' 4 ‘ " ' .' -\ ‘ ° 9 9 ‘ ' '- v .,'4:‘. - [r . ‘ 4 t | ‘-‘. J ’. . ‘ r . ' ‘ '1 J .. . vO- ’ _ I ‘ ,.-\ “\ A"\ , I ' ‘ I .- . ' , 7‘: .4- l ‘ ‘ ‘9 f -’ ’5 -'A: "' iv; ‘ 4 '(l\"" ‘ :1 , ‘b I ‘ ' . . .. ’. I ‘ ., r g f ’ (I: u ‘ v | I V ‘ 0:? ’ . I . ‘ . . . Q o . \ - . . I :— I I ____ hf~ g ’-_-V #“9 _L 7"?" _’ _‘;‘ .7 . , 7 . a!!! . _ _.__ _ u , 1, #42 Figure 32. Specimen D-l, picral etchant, lOOX. 77 Specimen D-Z, picral etchant, lOOX. Figure 33. 78 1 etchant, lOOX. O p1cra Specimen E—3 , Figure 34. 79 . 1], ... . . . “mt. mm“ A? .. .. .‘Mwfl. W1 wing; was mm» — , . .. ‘x‘f "’Ij‘f ‘5’ J Specimen E-4, picral etchant, 100x. Figure 35. SUMMARY AND CONCLUSIONS It can be clearly seen from the half U-curves shown in Fig- ures 19, 20, and 21 that hydrogen has an effect on hardenability. However, the effect illustrated here appears to be much less than that found by Houdremont and Heller (59). In this earlier work by Houdremont and Heller, nothing was said about Specimen cracking. This indicates that the hydrogen con- tent of the specimens was low. The specimens were heated in wet hydrogen gas held at 1 atmosPhere. Thus, the greatest possible hydrogen content of the specimens would be on the order of 5 mil- liliters per 100 grams of steel. The results of similar hydrogen contents are shown in Figure 12 of this work. The difference is so small as to be inconclusive. Thus, the effect of a small amount of hydrogen found in this work does not agree with that found by Houdremont. Houdremont illustrated the effect of hydrogen by fracturing the quenched specimens. His results are shown in Figure 5 of this report. It was difficult to fracture the specimens examined in this work on account of cracking. However, the two specimens shown 80 81 in Figure 22 were polished and etched with 5 percent nital to show the extent of hardening. Even though these specimens exhibit a definite difference from hardness examinations (see Tables IV and XIII), this difference is not obvious from the etched Specimens. Upon examination of Figures 29 and 30, illustrating Speci- mens 13 (control) and C-17 (hydrogen—treated), it can be seen that it was necessary to go deeper into specimen C—17 to find sufficient pearlite to determine the grain size. This illustrates that the effect of hydrOgen was to produce more martensite, not mere higher hard- ness. The conclusions of this study are as follows: 1. Hydrogen in amounts on the order of 15 milliliters per '100 grams has a small but definite effect on the hardenability of steel. 2. The increase in hardenability, brought about by hydrogen content, is negligible in commercial heat-treating practice, as the hydrogen content of steel is ordinarily very low. 3. Cathodic charging will not furnish enough hydrogen to measurably affect hardenability. 4. The hardenability increase brought about by hydrogen appears not to change with carbon content (as in the case of boron). 82. 5. Specimens containing hydrogen and martensite continue to crack for several hours after quenching. 6. Flakes, large enough to distinguish as such, require a period of time to form. 7. It appears that hydrogen has no effect on the austenitic grain size. At least this was the case in this work. 10. LITERATURE CITED K. C. Barraclough, ”The Significance of Hydrogen in Steel Man- ufacture," Murex Limited Review, vol. I, p. 320 (1954). Marion H. Armbruster, ”The Solubility of Hydrogen at Low Pressure in Iron, Nickel and Certain Steels at 400 to 600 °C.," Journal, American Chemical Society, vol. 65, p. 1043 (1943). A. Sieverts, "Die Loslichkeit von Wasserstoff in Kupfer, Eisen und Nickel" (in German), Zietschrift Fuer Physikalische Chemie, vol. 77, pp. 598-606, Leipzig (1911). E. Martin, “The Occlusion of Hydrogen and Nitrogen by Pure Iron and Some Other Metals," Metals and Alloys, vol. I, pp. 831-835, New York (1930). G. A. Moore and D. P. Smith, "Occlusion and Evolution of HydrOgen by Pure Iron," Transactions, American Institute of Mining and Metallurgical Engineers, vol. 135, pp. 225-295, New York (1939). K. Iwase' and M. Fukusima, ”Absorption of Hydrogen by Metals and Some Iron Alloys,” Nippon Kinzoku Gakkai-Si, vol. I, p. 151 (1937). K. Iwase, "Occlusions of Gases by Metals in Solid and Liquid States," Science Reports, Tohoku Imperial University, vol. 15, p. 531 (1936). A. Sieverts, "Zur Kenntnis der Okklusion und Diffusion von Gasen durch Metalle," Zeitschrift Fur Physikalische Chemie (in German), vol. 60, pp. 129-201, Leipzig (1907). A. Sieverts und Hagenacker, "Uber die Absorption dis Was- serstuffs durch metallischer Nickel," Berichte Derv Deutschen Chemischen Gesellschaft, vol. 42, pp. 338-347, Berlin (1909). Carl A. Zapffe, "Iron Hydrogen,“ American Society for Metals Handbook, p. 1208, Cleveland, Ohio (19418). 83 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 84 R. C. Ray and R. B. N. Sahar, ”Hydrides of Iron," Journal, Indian Chemical Society, vol. 23, p. 67 (1946). J. H. Simons and W. R. Ham, ”The Diffusion of Gases Through Metals from a Chemical Point of View," Journal of Chemical Physics, vol. 7, p. 899 (1939). J. H. Andrew, A. K. Base, H. Lee, and A. G. Quanel, Journal of the Iron 8: Steel Institute, vol. 146, p. 181 (1942). Donald P. Smith, Hydrogen in Metals, The University of Chicago Press, Chicago, Illinois, p. 32 (1948). J. D. Hobsen and C. Sykes, "Effect of Hydrogen on the Proper- ties of Low Alloy Steels," Journiof the Iron and Steel Institute, vol. 169. PP. 209-220 (1951). G. Hagg, "Eigenschaften der Phasen von Ubergangselementen in binaren Systemen mit Bor, Kohlenstoff und Stickstoff," Zeit- schrift fur Physikalische Chemie (in German), vol. 6B, pp. 221—232, Leipzig (1929). F. Seitz, The Physics of Metals, p. 38, McGraw—Hill Book Co., New York (1943). M. A. Grossman, Elements of Hardenability, American Society for Metals, Cleveland, Ohio (1952). R. A. Grange and T. M Garvey, "Factors Affecting the Hard- enability of Boron Treated Steels," Transactions, American Society for Metals, vol. 37, p. 136 (1946). M. A. Grossman, Elements of Hardenabilijy, American Society for Metals, Cleveland, Ohio, p. 138 (1952). Ibid., p. 138. W. Tofaute and H. Schottky, "Etsatz von Nickel in Austenitischen Chrom-Nickel-Stahlen durch Stickstoff," Archive fur das Eisen- huettenwesen, vol. 14, p. 71 (1940). M. A. Grossman, Elements of Hardenability, American Society for Metals, Cleveland, Ohio, p. 150 (1952). 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 85 E. Houdremont and P. A. Heller, "Wasserstuff als Legierungs- element bei Stahl und Gusseisen," Stahl und Eisen, vol. 61, p. 756 (1941). E. Houdremont, Handbuch der Sonderstahlkunde, Springer, Berlin (1943). C. L. M. Cottrell, ”Effect of Hydrogen on the Continuous—Cooling Transformation Diagram for a Manganese-Molybdenum Steel," Journal of the Iron and Steel Institute, vol. 176, pp. 273-282, London (1954). M. H. Armbruster, "The Solubility of HydrOgen at Low Pres- sure in Iron, Nickel and Certain Steels at 400 to 600°C.," Journal, American Chemical Society, vol. 65, p. 1043 (1943). K. C. Barraclough, “The Significance of Hydrogen in Steel Manufacture,” Murex Limited Review, vol. I, no. 13 (1954). D. J. Carney, J. Chipman, and Grant, ”An Introduction to Gases in Steel," Electric Furnace Steel Proceedings, American Inst. of Mining 8: Met. Engrs., vol. 6, p. 34 (1948). M. W. Mallett, ”The Water-Gas Reaction Applied to Welding- Arc Atmospheres," Welding Journal, Research Supplement, vol. 25, PP. 3965-3995 (1946). C. E. Sims, ”Behavior of Gases in Solid Iron and Steel," Gases in Metals, American Society for Metals, p. 148, Cleve- land, Ohio (1953). D. P. Smith, "Fundamental Metallurgical and Thermodynamic Principles of Gas-Metal Behavior," Gases in Metals, American Society for Metals, Cleveland, Ohio, p. 1 (1953). W. F. Giangue, ”The Entropy of Hydrogen and the Third Law of Thermodynamics: The Free Energy and Dissociation of Hydrogen," Journal, American Chemical Society, vol. 52, p. 4816 (1930). R. M. Barrer, Diffusion in and Through Solids, Cambridge University Press, p. 224 (1951). 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 86 F. Seitz, The Physics of Metals, McGraw-Hill Book Co., Inc.. New York, p. 200 (1943). Donald P. Smith, HydrOgen in Metals, The University of Chicago Press, Chicago, 111., pp. 57-58 (1948). Ibhflw p. 58. W. Geller and T. Sun, "Einfluss von Legierungszusatzen auf die Wasserstoffdiffusion im Eisen und Beitragzum System Eisen Wasserstoff," Archive fur das Eisenhuttenwesen, vol. 21, p. 423 (1950). 32' P. L. Chang, and D. G. Bennett, ”Diffusion of Hydrogen in Iron and Iron Alloys at Elevated Temperatures," Journal of the Iron 8: Steel Institute, vol. 170, pp. 205—213, London (1951). J. B. Seabrook, N. J. Grant, and D. Carney, "HydrOgen Em— brittlement of S.A.E. 1020 Steel," Transactions, American Institute of Mining 8: Met. Engrs., vol. 188, p. 1317. W. A. Bell, ”The Embrittlement of Steel by Hydrogen," Product Engineering, March, 1955, p. 192. C. E. Sims, G. A. Moore, and D. W. Williams, l'The Effect of Hydrogen on the Ductility of Cast Steels," Transactions, Amer- ican Institute of Mining & Metallurgical Engineers, vol. 176, p. 306 (1948). Ibid., p. 307. J. D. Hobson and J. Hewitt, ”The Effect of Hydrogen on the Tensile PrOperties of Steel,” Journal of the Iron and Steel Institute, vol. 173, pp. 131-140, London (1953). Ibid.,p. 139. S. A. Henes, "Arc Welding of Alloy Steels," The Welding Journal, vol. 23, p. 43, Research Supplement (1944). C. B. Voldrich, ”Cold Cracking in the Heat Affected Zone,” The Welding Journal, vol. 26, Research Supplement (1947), 5%. 153-169. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 87 M. W. Mallett and Rieppel, "Arc Atmospheres and Underbead Cracking," The Welding Jgurnal, Vol. 25, 1946, Research Sup- plement, p. 748. C. Derge and E. E. Duncan, "Thermal Segregation: A Mech— anism for the Segregation of Hydrogen in Steel," Journal of Metals, American Institute of Mining & Metallurgical Engineers, June, 1950, p. 884. C. E. Sims, "Behavior of Gases in Solid Iron and Steels," Gases in Metals, American Society for Metals, Cleveland, Ohio (1953), p. 168. D. J. Carney, J. Chipman, and N. J. Grant, ”An Introduction to Gases in Steel," Electric Furnace Steel Proceedings, Amer— ican Institute of Mining & Metallurgical Engineers, vol. 6, p. 34 (1948). Sims, op. cit., p. 171. L. M. Cottrell, "Effect of Hydrogen on the Continuous-Cooling Transformation Diagram for a Manganese-Molybdenum Steel," Journal of the Iron and Steel Institute, vol. 176, pp. 273-282 (1954). Ibid., p. 282. E. Houdremont and P. A. Heller, ”Wasserstuff als Legierungs— element bei Stahl und Gusseisen," Stahl und Eisen, vol. 61, p. 756 (1941). E. Houdremont, Handbuch der Sondenstahlkunde, Springer, Berlin (1943). M. A. Grossmann, Elements of Hardenability, American Society for Metals, Cleveland, Ohio, pp. 7—8 (1952). J. D. Hobson and c. Sykes, "The Effect of Hydrog‘en'xon the Properties of Low Alloy Steel," Journal of the Iron ,8: Steel Institute, vol. 169, p. 214 (1951). - \— \ E. Houdremont and P. A. Heller, "Wasserstuf£-als Leurungs- element bei Stahl und Eisen,“ Stahl und Eisenflx‘ol. 61, p. 756 (1941). \. 3' 8". (.553 ‘ “1““ v . H ‘9 Lid“ 1'31}. 1 Date Due {'33 ' 9 ‘....- r' .5 l '1 Demco-293 91111111191111” I I) “a ”0 m3 ['0 “113 mg “2 H ”I m