UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The characteristics of the formation of austenite in eutectoid steel 1947

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
UBC_1947_A7 R5 C4.pdf
UBC_1947_A7 R5 C4.pdf [ 12.72MB ]
UBC_1947_A7 R5 C4.pdf
UBC_1947_A7 R5 C4.pdf
Metadata
JSON: 1.0106999.json
JSON-LD: 1.0106999+ld.json
RDF/XML (Pretty): 1.0106999.xml
RDF/JSON: 1.0106999+rdf.json
Turtle: 1.0106999+rdf-turtle.txt
N-Triples: 1.0106999+rdf-ntriples.txt
Citation
1.0106999.ris

Full Text

L G" *> 8 1 I 1H ] ft-) C<rf . 1 THE CHARACTERISTICS OF THE FORMATION OF AUSTENITE IN EUTECTOID STEEL BY Michael Dennis Edward Robinson A Thesis Submitted ln Part i a l Fulfilment of the Requirements for the Degree of Master of Applied Science in the Department of Metallurgical Engineering The University of British Columbia September, 19^7 ABSTRACT This paper Is a report of the investigations carried out to determine the effect of mlcrostructure on the rate and nature of the formation of austenite in some .commercial steels. Previous work is reviewed and compared with the results of current experiments. Experimental procedure Is briefly discussed. Steels used for these experiments were similar to the following S.A.E. specifications; 1020, 10^5, 1080, blkQ, and 52100. Studies of the 1080 type form the greater part of the work. Results are presented in graphs of hardness against time-at-austenitlzatlon-temperature and in a series of photographs showing various stages of transformation for different prior structures. Conclusions drawn from results are discussed. The processes of nucleation and growth of austenite in pearlite, spheroidite, balnite, martensite, and sorbite are examined. The effects of lamellar spacing, size and distribution, of spheroids, and proeutectold constituents are noted, together with general considerations such as stability of micro- structures. The effect of chromium is briefly discussed. Possibilities for future work on austenitizatlon are presented, with emphasis on induction heating and flame hardening. TABLE OF CONTENTS page Object - - - - - - - - 1 Summary - - - - - - - - 2 Review of Previous Work - - - - k Formation of Austenite from Pearlite - 6 Formation of Austenite from Spheroidite - 8 Formation of Austenite from Other Structures 9 General Considerations - - - - 10 Experimental Procedure - - - - 15 Specimens Used 15 Rate of Heating of Specimens 16 Formation of Prior Structures 16 Austenitizing Specimens 19 Metallurgical Examination of Specimens 19 Graphical Presentation of Results 21 Study of Microstruetures - - - - 30 Formation of Austenite from Pearlite 31 Formation of Austenite from Spheroidite 34 Formation of Austenite from Other Structures 38 General Considerations 39 Discussion - - . - - Conclusions Pearlite Spheroidite Other Structures General Considerations Suggestions for Further Work - - - - 50 Suggestions Arising from Current Work 5° Flame Hardening and Induction Heating 51 Acknowledgment - - - - - - 53 *3 <*3 k6 Bibliography 5^ L I S T OF TABLES AND ILLUSTRATIONS page Table 1: Analyses of Steels Used - - - 15 Table 2: Origin of Prior Structures - - - 18 Plate 'I: Curve Showing Rate-of-Heatlng of Specimens - - - 17 Plates II to VII: Curves Showing Hardness 2k to vs. Tlme-at-Austenitizatlon 29 - Temperature of Specimens Figure i ; Schematic Curves Showing Relation between Transformation and Hardness. - - - 21 facing pages Figures 1 to 82: Photomicrographs Showing 30 to Progression of Austenitizatlon in Specimens - - • - kj, (1) OBJECT The object of the following investigations was to study the mechanism of the formation of austenite from various prior metallurgical structures. These investigations were resolved Into a search for the answers to the following questions; 1. What l s the nature of the nucleation of austen- ite in pearlite? in spheroidite? in other structures? 2. What is the nature of the growth of austenite in these structures? 3. How do the size and distribution of lamellae in pearlite and of spheroids in spheroidite affect the nucleation and growth of austenite? 4. What is the effect of structure on the time required for a given degree of austenitization? 5* What is the effect of carbon content on the time required for a given degree of austeniti- - zation? 6. What effect do alloy additions have on austenitization? The procedure followed in carrying out the object was arranged as follows: (a) a review of literature on the subject of austenit- ization; (b) a series of experiments with suitable specimens; (c) metallurgical study of these specimens; and (d) a comparison of experimental results with the conclusions of previous workers. This arrangement of procedure is discussed at further length in the following summary. (2) SUMMARY v The questions presented in the statement of the object form the basis for this thesis. They are discussed in the sections entitled "Review of Previous Work", "Graphical Presentation of Results", "Study of Microstructures", and "Discussion", and to some degree in "Experimental Procedure". The order of discussion is as follows: (a) The formation of austenite from pearlite; (b) The formation of austenite from spheroidite; (c) The formation of austenite from other structures; and (d) General considerations. This order has been observed, but not followed"^ s t r i c t l y , in "Review of Previous Work". In the review, only those sources relating to the questions of the object are included. Some sources have been omitted because they are repetitious; others presented theoretical considerations that would have required too much space to discuss in the review. This study of the investigations of previous workers i n - dicated the nature of the experimental work required for the current investigations'. "Experimental Procedure" outlines the work con- ducted by the writer. It deals :with the size of specimens used, the derivation of the rate-of-heating curve for the specimens, and the methods used to obtain the different prior structures. (3) The results of the examination of the specimens are given ln the sections entitled "Graphical Presentation of Results" and "Study of Microstructures". The f i r s t section discusses a group of curves showing hardness versus time-at- austenltizing-temperature; the second presents a series of photomicrographs showing different prior structures at d i f f e r - ent times of austenitization. Conclusions are drawn in both sections. The conclusions are summarized in "Discussion" and compared with those drawn by other workers, mentioned in the review. In these conclusions w i l l be found the answers, in part or in whole, to the questions asked in the object. Other questions arising from this research are presented in "Suggestions for Further Work". Parts of this section supplement the review. CO REVIEW OF PREVIOUS WORK Austenlte Is the high-temperature phase of carbon steel; i t has a face-centered cubic lattice, with the carbon atoms s t a t i s t i c a l l y distributed among the interstices of the lat t i c e . Below the c r i t i c a l temperature (Ae^), approximately 1330°F for plain carbon steels, the lattice transforms from face-centered cubic to body-centered cubic or tetragonal; i n either case, the solubility for carbon decreases and the excess carbon combines with some of the iron to form iron- carbide. Depending on the rate of cooling from the austenitic phase, the carbide i s combined in different ways with the remaining iron, or ferrlte . ( 1 , 2 ) . If the steel is again heated to above the c r i t i c a l temperature, these aggregates of iron-carbide and ferrite are re-transformed to austenlte; In other words, when carbon steel is heated to above the c r i t i c a l temperature, the phenomenon of austenitizatlon takes place. Grossman (3) outlines austenitizatlon as follows: "When • — steel reaches the temperature at which i t begins to transform, the very f i r s t minute austenlte crystals form within a pearlite island or in the boundaries between fer r i t e grains. These small austenlte grains grow across the pearlite islands and across the, ferrite grains un t i l they meet other austenlte grains grown from other nuclei. When a l l the austenlte grains have thus met, a certain set of grain sizes w i l l have been established. 0 (5) Bain (4) accompanying his discussion with several photomicrographs "by V i l l e l a , says that above the c r i t i c a l temperature, « the reaction between f e r r i t e and carbide to form austenlte is inaugurated. The reaction necessarily begins at the carbide-ferrite inter- face. — • The carbon originally In the carbide particles near the interface has moved out some 0.0003 inch in as short a time as 5 seconds and in 60 seconds sufficient carbon to form a com- plete austenitic matrix has diffused outward from the particles to a distance at least equal to half the spacing between carbide p a r t i c l e s . — This steep concentration gradient i s propitious for rapid carbon diffusion." Davenport and Bain (5), in discussing the laws applying to austenitic grain establishment in higher carbon steels, state that the f i r s t nuclei form at carbide-ferrite interfaces (a) where manganese, or a similar element, is at higher concentrations, (b) where carbide, i f lamellar, i s in the thinnest plates, or (c) where carbide, If spheroidal, is in the finest particles. In the i n i t i a l stages these nuclei are very numerous but as. time progresses, those more favorably located, meanwhile absorbing the larger areas of ferrite and carbide. These premises had already been set forth by Grossman (6), who discussed them with particular reference to low carbon steels. Davenport and Bain (5) also indicate that the laws of grain growth are closely related to and dependent upon the phenomena of the formation of austenlte. (6) The foregoing ideas have also been discussed in part or in whole by Carpenter and Robertson (7), Hultgren (8), and others. Roberts and Mehl (9) present a very thorough b i b l i - ography of work done prior to 194-2. These fundamentals of austenitizatlon have been developed more thoroughly by Roberts • and Mehl (9, 10), Digges and Rosenberg (11,12,13,14), and Baeyertz (15). The Formation of Austenlte from Pearlite Baeyertz (15) states, in substance, that: on heating aggregates of ferrite and pearlite, the pe a r l l t l c areas become austenitic f i r s t , followed by free f e r r i t e . Similarly, in aggregates of cementite and pearlite, the trans- formation of the pearlite is followed by that of the cemetite, which Is complete only when the A c m is exceeded. Austenlte forming at the A C 1 must be of eutectoid content and for this reason would naturally be expected to nucleate at phase boundaries between fer r i t e and cementite. At f i r s t , "numerous small austenlte grains replace small portions of the cementite lamellae and the ferrite between two cementite lamellae." As the extremely small grains meet each other, the larger probably grow at the expense of the others, this effect increasing as the grains become established. On increasing the temperature, these grains apparently grow in two ways, the f i r s t by transformation of contiguous ferrite and (7) cementlte and the second by atomic rearrangement of small particles of austenlte enveloped by the larger grains, Baeyertz (15) also mentions two inhibiting effects, (a) undissolved particles of oxides, carbides, etc., which produce considerable discrepancy in the size of the austenlte grains and, in general, tend to reduce the grain size; and, (b) , the fact that austenlte does not readily grow across ferrite grain boundaries between pearlite colonies. These effects insure that the resulting austenlte has a grain size equal to or less than the size of the pealite colonies from which i t was formed, provided the austenlte so formed is not then subjected to a coarsening cycle. Davenport and Bain (5) state that the f i r s t nuclei form at the carbide-ferrite interfaces where the lamellae are the thinnest. Baeyertz (15) says that the nuclei expand more rapidly along the pearlite lamellae than across them, thus producing grains roughly rectangular in section, with the advancing boundaries being commonly perpendicular to the lamellae. Digges and Rosenberg (Ik), in their study of a high- purity iron-carbon alloy (0»50 per cent carbon), determined that, in a mixture of fine pearlite and ferrite, austenlte nucleates preferentially at the ferrite-carblde interfaces near the pearlite-ferrite boundaries and at pearlite colony boundaries and secondarily between the pearlite lamellae. (8) The predominant growth occurs In the directions of the lamellae* Roberts and Mehl (9) concluded that nuclei formed at pearlite colony boundaries were relatively few compared to those formed at ferrite-cementite interfaces and that a minor , source was at ferrite grain boundaries at contact with carbides. In discussing rates of nucleation and growth of pearlite, Mehl (10) showed that the rate of austenltization i s the faster the finer the spacing of the pearlite. Roberts and Mehl (9) stated that as the spacing of the pearlite decreases, the rates of nucleation and growth increase and that the over- a l l increase is roughly proportional to the carbide-ferrite interfacial area. Fine pearlite produces a slightly finer grain than coarse pearlite on austenitizatlon, since the rate of nucleation increases as the interlamellar spacing decreases. The Formation of Austenlte from Spheroldlte Davenport and Bain (5) state that the f i r s t nucleation in spheroldlte occurs where the carbide i s in the finest particles. Baeyertz (15) states that the transformation begins in the grain boundary regions of the prior austenlte. The nuclei form at ferrite-cementite phase boundaries around the cementite particles which l i e in the ferrite grain boundaries. (9) Dlgges and Rosenberg (14) determined that austenite nucleates at the ferrite-carbide interfaces, primarily at the carbide network and occasionally within the network. At f i r s t growth is preferentially along the network, Roberts and Mehl (9) state, in substance, that: the austenitizatlon of spheroldlte is essentially similar to that of pearlite and approaches general nucleation. In low carbon steels austenlte forms around each particle and grows into ferrite; in high carbon steels austenite, forms around some particles and growth includes the rest. The Formation of Austenite from Other Structures Baeyertz (15) states that the transformation of martensite i s essentially the same as that of spheroldlte, commencing in the grain boundary regions of the prior austen- i t e . The nuclei form around cementite particles which l i e in the ferrite grain boundaries. Martensite may be considered as extremely fine spheroldlte, since i t is necessarily temper- ed before the transformation begins. Roberts and Mehl (9) claim that the carbides in the martensite precipitate on heating and that austenitizatlon proceeds similarly to that in spheroldlte. Differences i n rate are functions of the degree of spheroidizatlon. No literature was available regarding the austenit- izatlon of structures other than those already mentioned. (10) General Considerations In discussing austenitic grain establishment in higher carbon steels, Davenport and Bain (5) atate that the f i r s t nuclei form at carbide-ferrite interfaces where manganese, or a similar element, is at higher concentration. Dlgges and Rosenberg, (14) studying a 5°-carbon steel, concluded that the i n i t i a l austenite i s fine-grained, regardless of the rate of heating through the Ac-_. Rapid heating through the Ac1-Ac^ range produces fine-grained austenlte. They further con- cluded that the rate of growth, rather than the rate of nucleation, was the important factor in establishing the f i n a l austenitic grain size of this steel, Baeyertz (15) concluded from her studies that prior structure determines only grain size at the time of formation, except where undissolved particles in the austenite inhibit the action. Mehl (10) shows that the rate of austenitizatlon i s the greater the coarser the grain size of the steel, Roberts and Mehl (9) have made what appears to be the most academic approach to the problem. They assume the pro- cess to be established as one of nucleation and growth, the nucleation being, a structure-sensitive property. Above the equilibrium temperature, atomic mobility is increased and the energy required for the formation of austenite decreases as the temperature i s raised. (11) The actual location of the nuclei is a matter of s t a t i s t i c a l probability, as the number of possible positions i s a function of the interfacial area of the ferrite-cementite interfaces. Even though the energy required for formation i s less at grain boundaries, the effect of ferrite grain boundaries i s small, , as nucleation would have"to begin at the point where a ferrite grain boundary meets a cementite particle. Nuclei formed at pearlite colony boundaries are relatively few compared to those formed at ferrite-cementite interfaces. According to Roberts and Mehl (9),.the austenite nuclei form around cementite particles lying in ferrite grain boundaries and often form f i r s t at prior austenite grain boundaries. In the case of spheroidite and martenslte, a greater number of cementite particles l i e in ferrite grain boundaries than in the case of pe a r l i t i c structures. Roberta and Mehl logically assume from these and similar factors that the nucleation of austenite Is structure-sensitive, dependent mainly on the size and shape of the carbide particles and to a much smaller degree upon the presence of ferrite grain boundaries. In discussing the growth of austenite nodules, Roberts and Mehl reason as follows: cementite and ferrite react at the advancing interface to form austenite, the rate of interaction depending on the rate of solution of carbide and also on the rate of migration of carbon atoms in the austenite. (12) Therefore the rate of growth increases with an increase of temperature, as the diffusion factor and the effect of carbon concentration gradients increase with temperature. Hence, the rate of growth is also a structure-sensitive property and i s presumably dependent on the rate of diffusion and perhaps on the rate of the reaction at the interfaces. Interference with normal growth is due primarily to impingement, second- a r i l y to grain coalescence (which would speed the reaction) and to pro-eutectoid constituents in hypfo- and hyper- eutectold steels. Roberts and Mehl (9) state that areas with high manganese and (or) low phosphorus and silicon w i l l nucleate f i r s t , longitudinally along the rows. Johnson and Mehl (16), in discussing the determin- ation of austenitic grain size, give the following mathemat- i c a l relations? Where N i s the total number of nodules per unit volume of totally reacted sample, Nt i s the total number of grains per unit area of polished surface, 0 i s the rate of growth in mm. per s e c , and N v i s the rate of nucleation, On the effect of banding and dendritic patterns, then N_ = 0.896(N )^ (1) (2) (13) Roberts and Mehl (9), assuming that N Y i s constant and that the reaction proceeds by general nucleation, show that N Y can be calculated from the rate of growth and the time of half reaction, t ^ Q , as follows: k /N VG 3 . t 5 0 - o .9 (3) and y ^ G 3 . t 2 Q s 0.66 - - - (4) Apparently equation (4), using the time for 20 per cent reaction, gives better agreement with actual data than does equation (3) . They found that N V Increases with (a) decrease in the spacing of pearlite, (b) increase in temperature, (c) decrease in the grain size of the steel, and (d) time at a continuous rate, up to 15 or 20 per cent re- action. Roberts and Mehl also outlined the three general cases covering nucleation and growth of austenite: (1) K increases at a faster rate than G with increase in reaction temperature and the grain size w i l l therefore decrease with increasing heating rate. (2) G increases at a faster rate than N with Increase in reaction temperature and the grain size w i l l therefore increase with increasing heat- ing rate. (3) Both N and G increase at the same rate and there w i l l be no effect of heating rate on the grain size. In conclusion, Roberts and Mehl (9) summarize their findings as follows: (1) Austenite forms by nucleation and growth from ferrite-cementite aggregates; both phenomena are structure-sensitive. (2) Austenlte formed in eutectoid steels i s not homogeneous when the fer r i t e disappears as a structural constituent, for undissolved carbide remains, the solution rate of which is dependent on both time and temperature. Carbon concentration gradients exist after the carbide is no longer v i s i b l e . (3) 'The rate of formation of austenite increases continually with temperature above the Ae^ > in terms of nucleation and growth. (4) As pearlite spacing increases, the rates of nucleation and growth increase. The over- a l l increase i s roughly proportional to the carbide-ferrite interfacial area. (5) The rate of growth is less for aluminum - k i l l e d steels while the rate of nucleation remains about the same, accounting for the finer grain size in the former. Where the rate of growth is about the same, the rate of nucleation i s higher, thus preserving a high Ny/G ratio. (6) Fine pearlite produces a slightly finer grain than coarse pearlite, since the rate of nucleation increases with finer spacing. Some literature on induction heating has been re- viewed in "Suggestions for Further Work". Although some phases of austenitizatlon are discussed in these references, they were not necessary to the undertaking of the current experimental work, outlined in "Experimental Procedure". (15) EXPERIMENTAL PROCEDURE Specimens Used Five Steels were used for the experimental work: Steel 1 Steel 2 Steel 3 Steel 4 Steel (5 S.A.E. 1020 S.A.E. 1045 S.A.E. 4l40 Atlas "KK", similar to S.A.E. 52100 Atlas "Maple Leaf Special", similar to S.A.E. 1080. , / ' / The larger part of the work was done with Steel , 7; Steels 1 and 2 were used to ill u s t r a t e transformation ln low carbon steels; Steels 3 and 5 were used to examine the effect of additions of chromium. The analyses for these samples is given in Table I. Steel C Mn Cr Mo Si 1 0.16 0.72 — — 0.025 2 0.43 0.69 ~ — 0,21 3 0.48 0.89 0.53 0.23 0,28 5 1.03 0.40 1.25 .05 0.2? ^ 0.82 0.12 — — 0,17 Table I : Analyses of Steels Used Bars of these steels were turned to 7/16" diameter (tolerance 0.005") and discs 1/8" thick (tolerance 0.002") were cut from the round bars. A small hole was d r i l l e d through each specimen close to the circumference, thus pro- viding a means of suspending the specimens while heat treating them. Several specimens were d r i l l e d radially to the center and the hot Junctions of chromgl-alumel thermo- couples made with 22 gauge wire were wedged into the holes. 1 (16) Apparatus consisted of two electric salt-bath furnaces (a one-pound and a five-pound capacity), temperature control instruments, chromel-alumel thermocouples, a potentio- meter, a stop-watch, a quench tank, and a supply of fine chromel wire for suspending the specimens in the salt bath. Rate of Heating of Specimens In order to determine the actual time of the speci- mens at austenitizing temperature, a rate-of-heating curve had to be constructed. This was done by attaching the thermo- couples wedged in the specimens to the potentiometer and then plunging the specimens into the salt bath. The time required for the specimen to rise to a temperature indicated by a predetermined setting on the potentiometer was measured with a stop-watch. By repeating this procedure for several . different settings of the potentiometer, sufficient data was obtained to construct the curve shown.in Plate I, page 17. Repetition of the whole procedure several times proved the curve to be accurate enough for the purpose. Formation of Prior Structures Each set of specimens was divided Into groups of ten; each group of ten specimens was heat treated to form a pre-determined prior structure. Full use of '^-Curves'* and other heat-treating data was made at this point (17,18). A l i s t of these structures i s given in Table II, page 18. 1300 £1100 I N SOO f340°F tSa/rBath Temp i ei-ature, I4SS°F 1 ts° \ i i j IS ZS 3S "Time,, Seconds 45 SS PLATE I Graph showing Rate of Heating of Specimen i n S a l t Bath at 1455°E 10°. The specimen, 1/8" t h i c k "by 7/16" diameter, was suspended i n the s a l t hath by a chromel-alumel thermocouple, wedged i n a r a d i a l l y d r i l l e d hole so that the hot j u n c t i o n was at the center of the specimen. The l a g i n the curve at "A" i n d - i c a t e s decalescence on passing through the c r i t - i c a l temperature and i s associated w i t h the abs- o r p t i o n of heat as f e r r i t e transforms to austen- ite« The time necessary t o reach a u s t e n i t i z i n g temperature, 20 sejconds, has been subtracted from from the t i b e s of immersion] of specimens i n the s a l t bath; hence, P l a t e s I I t o V I I and Figures 1 t o 82 are based on the true times of a u s t e n i t ~ i z a t i o n f o r the specimens. (24) (18) Steel Preliminary Succeeding Resulting Specimen Used Austenitizatlon Treatment Structure Ident'n 1080 i hr., 1450 3h hr., 1300 C Pearlite 7b n M 1 hr., 1275 M " 7c n n { hr., 1250 H " 7d n tt { hr., 1225 R 0 7e tt i t t nr., 750 Upper Bainite 7m n i i 2 hr., 550 Lower " 7n tt t t Brine quenched Martensite 7o u 0 Quenched drawn at 800 % hr. Sorbite 7P JJ a 8^ hr., 1290- 1340 Spheroldlte 7q it H 10 hr., 0 11 7r tt a 12 hr., " n 7s t t » 14 hr., « St 7t ri H 1 hr., 1250 M Pearlite 7t it n 20 hr.,1280- 1330 PSC " 7g 1020 t t 1 hr., 1250 U " l f n a 20 hr.,1280- 1330 Spheroldlte Ig 1045' i t 1 hr.,1250 M Pearlite 2f u i t 20 nr.,1280- 1330 C Pearlite . 2g 52100 tt 1 hr.,1250 Spheroldlte 5t H i t 20 nr.,1280- 1330 i t f g 1045 2 hr., 1550 2 hr.,1250 M Pearlite 2a 4140 tr 2 hr.,1250 u 3a 52100 tt 2 hr.,1250 11 5a (i) Temperatures indicated are in degrees Fahrenheit, plus or minus 10 degrees. ( i i ) "Succeeding Treatment" followed "Preliminary Austeniti- zatlon without delay. ( i l l ) Spheroidlzlng cycles resembled a sine curve; each complete cycle within the temperature ranges shown required approximately 15 minutes. (iv) A l l heat treating :cycles used were determined by refering to "S-Curves". (v) C - Coarse, M - Medium, F - Fine, L- Lower, PS - Partly Spheroidized, U - Unresolved at X730. Lamellar spacing varied from 2 microns for Coarse Pearlite to less than •5 microns to less than 1 micron in spheroldlte. (vi) Original structures of steels: 1080, annealed; 1020, 1045, and 4l40, normalized; 52100, spheroidized. Table II: Origin of Prior Structures Used in Final Austenitizatlon Experiments. (19) Austenitization of Specimens After the prior structures were formed, each specimen was subjected to the fi n a l heat treatment. One specimen was reserved from each group of ten for examination in the "as l s " condition. Each of the remaining nine was immersed in a salt bath, temperature 1455°F plus or minus 10 degrees, for a measured time, 10 seconds to 10 minutes, and quenched in brine. A nine per cent salt solution was used after the recommend- ations of G i l l (19). This treatment produced a group of ten specimens for each prior structure; each specimen in each group has been austenitized for a different period of time. Commercial heat treating salts were used throughout with one exception. A "home made" salt (20) was used extensively and proved satisfactory for intermittent use at temperatures between 1200 and 1500°F; the composition of this salt was as follows - barium chloride 48 . 1$, potassium chloride 3 Q»7# sodium chloride 21.2$. Metallurgical Examination of Specimens Each specimen was spli t across the diameter; one half was mounted for metallographic examination (Figures 1 to 82, "Study of Microstructures"): the other half was- tested for hardness with a Rockwell hardness tester. Results of the hard- ness tests are given in Plates II to VII, "Graphical Present- ation of Results". (20) The hardness readings were converted to the Brinell scale (3000 kg. load, 10 mm. ball), the only suitable scale covering the complete range of hardnesses. Rockwell "A", "B", and "C" scales were used in the actual determinations. A four per cent solution of pic r i c acid in ethyl alcohol proved to be the best general etch for metaliographic examination (21). Its use brought out excellent detail in most structures encountered, especially in pearlite and spheroldlte; i t was particularly valuable for distinguishing martensite (quenched austenlte) from untransformed structures by coloration. True austenitizatlon times (time in salt bath minus 20 seconds, from Plate I) are used in "Graphical Presentation of Results u and "Study of Microstructures". A l - though the term "austenite" is used in discussing these results, the product actually studied was martensite, formed by quenching the specimen to "freeze" the structure at the desired time of austenitizatlon. 1 (21) GRAPHICAL PRESENTATION OF RESULTS The graphs shown in Plates II to VII inclusive are derived from the hardnesses found for each specimen. This i s a reliable method for determining relative rates of austenitization. This i s not to say that percentage trans- formation can be predicted from hardness or vice versa, although a certain degree of familiarity with the curves permits prediction within rough limits. According to Roberts and Mehl (9), increase in hardness lags behind per cent transformation up to 10 or 15 per cent trans- formation. The curves then converge and cross and the hardness ap- u proaches a maximum at Jj 85 to 90 per cent transformation, i.e., the per cent trans- formation lags behind the Increase in Time of Austenitizing Temperature Figure i : Schematic Curves Showing Relation Between Transformation and hardness in the later Hardness. stages of the reaction. The latter lag i s apparent in the current studies. The studies do not, however, show the I n i t i a l lag in increase of hardness. (22) Examination of the graphs i n Plates II to VII, pages 2,4 to ZSy indicates the following conclusions? Plate II; Steels 7b,7c,7d,7e, cut from the same bar of Atlas "Maple Leaf Special"(S.A.E. 1080) are successively finer grades of pearlite formed at 1300, 1275, 1250, and 1225 F respectively. The curves show that the finer the pearlite the sooner the hardness increases. Approximately f u l l hardening is reached at 80 seconds. (2) Plate III: Steels 7q, 7r, 7s 7t, cut from the bar above and spheroidized 8-f, 10, 12, 14 hours of spheroids (range of average sized particles, O.75 to 1.5 microns) i s not particularly pro- nounced, the curves definitely show that the finer the spheroldlte the sooner does the Increase in hardness occur. Complete hardening is not attained at 280 seconds, as this time is not sufficient to austenitize the spheroids completely; maximum hardnesses are approached at 40 seconds austenitizatlon of these specimens. (3) Plate IV; Steels 7m, 7n, 7o, 7p, cut from the bar above and having prior structures of upper bainite, lower bainite, martensite, and sorbite (tempered martensite) respectively. The graphs show that these structures transform far more rapidly than pearlite or spheroldlte. Full hardening Is reached in 15 seconds for these finer structures; there is no significant d i f - ference in the time necessary to completely harden bainite, martensite, or sorbite. (4) Plate V: Steels 7f, 7g, 5f, 5g, the f i r s t two cut from the bar above, the second from the same bar of Atlas "KK" (52100). Specimens (f) were annealed 1 hour at 1250°F and specimens (g) heat treated 20 hours at 1280-1330°F after austenltlzing \ hour at 1450 F. The curves i n - dicate that structures taking longer periods to develop, i.e., the more stable structures, take longer periods to austenitize. (5) Plate VI: Steels 3f, 3g, 2f 2g, the f i r s t two cut from the same bar of 4l4o, the others from the same bar of 1045• Specimens (f) and (g) treated as in (4). In addition to confirming the conclusion in (4), these curves also (23) Indicate that the addition of chromium retarded the transformation to austenite^ This i s to he expected, since chromium is a strong carbide former and i s known to retard the transformation either to or from austenite(22,23). Comparison between steels 2 and 7 (see Plates V I , V) indicates that decreasing carbon below the eutectoid point increases the re- action time* (6) Plate V I I ; Steels 2a, 3a, 5a, cut from the same bars as 2, 3, 5 (Plates V, V I ) , respectively. A l l specimens austenitized 2 hours at 1550°F followed by annealing 2 hours at 1250°F, prior to f i n a l austenitizing treatments at 1455° F» No conclusions are drawn from information i n Plate V I I . These specimens were austenitized one at a time and probable fluctuations of temperature between specimens prohibit accurate comparisons. Specimens 7b, c, d, e were austenitized simultaneously, however, as were specimens 7<1, *\ s, t and 7m, n, o, p; a l l - specimens (f) and (g) were also austenitized together. Curves for specimens l f and l g were not drawn, as the range of hardnesses for this steel was so small that such curves were useless for comparison. These latt e r speci- mens, however, furnish some of the best photomicrographs depicting the formation of austenite in low carbon steel (see Figures 1 to 4, "Study of Micro structures"). The following "Study of Microstructures" is intended as a companion piece to the graphs discussed in the present section; the same specimens have been used for each study. S ^ S & s & BrineII Hardness (3000 Hg., 10 mm.) |LATE I I N Graphs showing Hardness v s . Time at A u s t e n i t i z i n g Temperature. S a l t Bath Temperature, 1455°E± 10°. SAE 1080 - 7b - Coarse P e a r l i t e 7c - Medium P e a r l i t e 7d - Medium P e a r l i t e 7e - Fine P e a r l i t e ~_T??—- I £ S S s Brine// Hardness (3000 Kg., 10 mm.) 25 1 S 3 PLATE I I I Gra.phs showing Hardness vs. Time at A u s t e n i t i z i n g Temperature .Salt Bath Temperature, 1 4 5 5 6 F ± 1 0 O . SAE 1080 - 7q - Fin e s t Spheroidite 7r - 7s - 7t - Coarsest Spher'ite'' PLATE IV Graphs shewing Hardness vs. Time at Austenit i z i n g Temperature .Salt Bath Temperature, 1455°E±10°. SAE 1080 - 7m - Upper B a i n i t e 7n - Lower Bainite 7o - Mart en s i t e 7p = S o r b i t e " / t 1 PLATE V . \ Graphs showing Hardaess vs. Time at A u s t e n i t i z i n g Temperature .Salt Bath Temperature, 1455°E± 10° . SAE 1080 - 7f - Coarse P e a r l i t e 7g - Spheroidite SAE 52100- 5f - Spheroidite / 5g - " (More homogeneoj^ ^ Graphs showing Hardness v s . Time at i A u s t e n i t i z i n g Temperature .Salt Bath \ Temperature, 1455°E±10o. ;! SAE 4140 - 3f - P e a r l i t e j 3g - Coarse P e a r l i t e SAE 1045 - 2f - P e a r l i t e 2g - Coarse P e a r l i t e PLATE VII Graphs showing Hardness v s . Time at A u s t e n i t i z i n g Temperature. S a l t Bath Temperature, 1455°P±10°P. SAE 1045 - 2a - P e a r l i t e 'SAE 4140 - 3a - P e a r l i t e SAE 52100- 5a - P e a r l i t e ? i g . l : l f k X730 Annealed 1020 A u s t e n i t i z e d 15 sec. F i g . 2 : l f C X7 30 A u s t e n i t i z e d 5 sec \ u s t e n i t i z e d 280 sec. (30) 5TUDY OF MICROSTRUCTURES Most of the photomicrographs are arranged so that each set represents the progression of transformation In a particular steel. Sets have been grouped, where possible, to show the -differance in reactions for steels having a relatively slight difference in prior structure. A l l specimens were austenitized in a salt bath at 1455°F, plus or minus 10 degrees. Reference to Plate I, page 1? shows that austenitizatlon temperature, approximately 1340°F, was reached at 20 seconds In the salt bath; specimens required a further 35 seconds to rise to the temperature of the bath. Times mentioned for the Figures shown do not i n - clude the i n i t i a l 20 seconds required to raise the temperature of the specimens to the point where austenitizatlon commenced* Figures 1 to k, facing page, are presented as an introduction and show progressive austenitizatlon in low carbon steel (S.A.E. 1020). The original structure, (Figure 1) consists of grains of very fine pearlite (unresolved at X 730) distributed in bands in a ferrite matrix* As shown in Figure 2, austenitizatlon. of the pearlite i s effected within 5 seconds; the austenite has already begun to grow in the surrounding f e r r i t e . Figures 3 and k show the advance of austenite (darker areas), preferentially along the ferrite grain boundaries. L i t t l e transformation has occurred between 15 and 280 seconds. According to the iron, iron-carbide equilibrium Fig.7 : 7bF X 715 A u s t e n i t i z e d 20 sec. F i g . 6 : 7"b̂  X 715 A u s t e n i t i z e d 15 sec. F i g . 8 : 7"bG X 715 A u s t e n i t i z e d 30 sec. (3D diagram (20), the austenlte:ferrite ratio in a completely- reacted, 0*20 per cent carbon alloy is about 8:7 at 1450°F. Hence, transformation would not be expected to continue a great deal beyond that shown in Figures 3 and 4. Comparison of these figures indicates that, after the pearlite has trans- formed, the austenlte advances along the ferrite grain boundaries and expands laterally u n t i l the limiting carbon concentration is reached* Formation of Austenlte from Pearlite Figures 5 to 8, facing page, and Figures 9 and 10 are studies of transformation occurring in coarse pearlite. Figure 5 r e P r e s e r r t a ^ e original structure, formed by cooling from 1450 to 1300°F and holding at the latter temperature for 3̂- hours. Average spacing i s about 1.5 microns per lamella. The f i r s t noticeable transformation to austenite has occurred at 15 seconds, and is manifested by "fusion" of some of the lamellae (see Figure 6). Figure 7 indicates that nucleation has commenced preferentially at the boundaries of the pearl- ite colonies and that growth has begun to include these colonies before further significant nucleation has occurred. Figure 8 (30 seconds austenitizatlon) shows points of random nucleation and growth after the primary reaction is well consolidated. Figures 9 and 10, taken at higher magnification, are perhaps the most interesting of the series. Interpretation Fig*9 •  7bV X 1120 A u s t e n i t i z e d 20 seconds. (32) of the evidence shown at lower right and top center of Figure 9 and at upper le f t of Figure 10 indicates the following conclusions regarding the austenitizatlon of pearlite, 1. Nucleation starts preferentially at the ferrite-cementite interface near a pearl- ite colony boundary. 2. I n i t i a l growth tends to "hug" the carbide lamella on which nucleation has occurred and proceeds along the side of the lamella for some distance before dissolving the carbide and before crossing the intervening expanse of ferrite to the next lamella. 3. The preferred growth of austenlte i s along and between the lamellae at the fe r r l t e - carbide interface. Growth is retarded by pearlite grain boundaries, carbide lamellae, and areas of ferrite (lamellar or not) in that order when they exist at right angles to the direction of growth. k. Small traces of the carbide lamellae are " l e f t behind" by the advancing austenlte; these are dissolved later as the trans- formation continues. Figures 11 to 15 (facing page 33) picture the trans formation in pearlite formed by quenching S.A.E. 1080 from 1450 to 1275°F and holding one-half hour at the latter temper ature. The f i r s t signs of austenitizatlon occur at 15 seconds (see Figure 12) and are indicated by a slight thicken ing and "welding" of the carbide lamellae. No significant difference exists between the character of transformations in this and the preceding structure, 7b. An instance of the fusion of the lamellae at the grain boundary, in the direction of the boundary, before ? i g . l l . 7cC X 715 ? i g . 13: 7cF A u s t e n i t i z e d X 715 20 s e c . ? i g . ! 4 : 7cG X 715 A u s t e n i t i z e d 30 s e c . •'ig.15: 7cH A u s t e n i t i z e d X 715 40 sec. F i g . 16: 7dF X 715 A u s t e n i t i z e d 20 sec. ' i g . 17: 7dG X 715 A u s t e n t i i z e d 30 sec. Fig.18: 7dH X 715 A u s t e n i t i z e d 40 sec, I f f . 1 9 ; 7eC X 715 Aus t e r i i t i z e d 5 sec. Fig.20;7e"B X715, 1 Plg.21:7eG X 715 Fig.?.2:7eH X 715 Aust. 15 sec. Aust. 30 sec. tost. 40 sec. (33) transformation proceeds along the lamellae is shown in the upper right corner of Figure 14 (austenitized 30 seconds). From this and similar instances, i t may he inferred that the growth of austenite tends to parallel pearlite grain bound- aries. This, however, i s a secondary preference and the growing areas of austenite are roughly rectangular, with the length parallel to the lamellae and the breadth parallel to the grain boundaries, or across the lamellae. Figure 15 shows the transformation to be about 75 per cent complete at 40 seconds within the austenitizing range, or above the Ac^. Figures 16, 17 and 18 show transformation in pearlite formed at 1250°F; Figures 19 to 22 show transformation in pearlite formed at 1225°F. These specimens exhibit 7b and 7c, except that the transformation is the faster, the finer i s the pearlite. Average lamellar spacing i s approximately 1.5 microns per lamella for steel 7b, 1 micron for steel 7c, 0.8 for 7d, and 0.7 for 7e. Two more conclusions are added to those already formed regarding transformation in pearlite: 5« The secondary preference for the growth of austenite, nucleating as in 1., l s perpendicular to the pearl- ite lamellae next to the pearlite grain boundaries. This and the preference indicated in 3» result ln roughly rectangular austenite grain. 6. The finer the spacing of the pearlite lamellae, the more rapid is the transformation of pearlite to austenite, other variables being equal. 7qA X 840 P i g . 2 4 : 7qS X 840 Spheroidized 8-g- h r . A u s t e n i t i z e d 15 sec. SAE 1080 (34) Formation of Austenlte from Spheroldlte Figures 23 to 42 show a group of spheroidized structures of S.A.E. 1080 (Atlas "Maple Leaf Special") at various stages of transformation. Steel 7q, austenitized one-half hour at 1450°F and cooled to and held at 1290-1340°F for 8£ hours, i s shown In Figures 23 to 26; steel 7r, 10 hours; steel 7s, 12 hours; and steel 7t, 14 hours, are shown in Figures 27 to 32, 33 to 36, and 37 to 42, respectively. The maximum diameters of the spheroids in steels 7q, 7r, and 7t are approximately 1,5, 2, 3, and 4 microns, respectively average diameters are roughly ,75, 1, 1, and 1,5 microns, respectively. The structures are progressively more stable from 7q to 7t, that i s , more of the spheroids approach maximum size in the longer-treated steels and the spheroids are more evenly distributed. There are two hints of impending transformation at 10 seconds at austenitizing temperature for these specimens. On comparing specimens 7rD, 7sD, and 7tD (Figures 28, 34, 38) with specimens 'A1 for the same steels (Figures 27, 33, 37), a "thinning out" of the finer spheroids can be seen in specimens 'D', indicating that some of them have commenced transformation within 10 seconds. A slight difference in re l i e f is also noticeable;, in specimens 'A1, the r e l i e f Is a l l within the boundaries of the spheroids but in specimens gift.27; 7rA X 840 F i g . 2 8 : 7rD' X 840 F i g . 2 9 : 7rE X 840 Spher'zd 10 h r . A u s t e n i t i z e d A u s t e n i t i z e d SAE 1080 10 seconds 15 seconds Fig.30:7rF X 840 Fig.31:7rG X 840 Fig.32:7rK X 840 A u s t e n i t i z e d A u s t e n i t i z e d A u s t e n i t i z e d 20 seconds 30 seconds 280 seconds (35) fD f, there i s not only r e l i e f w i t h i n the perimeters of the spheroids but a l s o a f a i n t o u t l i n e around the outside of them. This l a t t e r phenomenon of r e l i e f i n m l c r o s t r u c t u r e i s the only evidence In any of the photomicrographs of the a c t u a l n u c l e a t i o n of a u s t e n i t e i n s p h e r o i d i t e . I t i n d i c a t e s that the spheroid i s the nucleus and that the transformation begins on the whole surface of the spheroid r a t h e r than at a p o i n t . I t i s perhaps more l o g i c a l to assume that a nucleus forms at a p o i n t on the surface of the spheroid and t h a t t h i s p o i n t expands r a p i d l y to form an enveloping f i l m of a u s t e n i t e about the spheroid. The d i f f e r e n c e between specimens ,D' and 'E* i s q u i t e pronounced. The t r a n s f o r m a t i o n has progressed from l e s s than one per cent to about 80 per cent i n specimen 7qe, over 5° Pe** cent l n 7rE, about 40 per cent l n 7sE, andabout 25 per cent i n 7tE. These increases have been e f f e c t e d by h o l d i n g the specimens f o r an a d d i t i o n a l 5 seconds above the Ac^ (see Figures 24, 29, 35, 39)» I n these specimens, the transformation has apparently o r i g i n a t e d - i n the spheroid network at the g r a i n boundaries of the f e r r i t e matrix and subsequent growth has taken place along the boundaries. These f e r r i t e g r a i n boundaries are not p a r t i c u l a r l y evident i n the Figures shown. Specimens had to be etched deeply to b r i n g out the boundaries and the n e c e s s a r i l y long exposures of the p l a t e s e l i m i n a t e d the d e t a i l t o such an extent that photomicrographs so taken proved u s e l e s s . Eig.33: 7sA X 840 Pig.34; 7sD X 840 Spheroidized 12 h r . A u s t e n i t i z e d 10 sec. SAE 1080 ?ig.35;7sE X 840 A u s t e n i t i z e d 15 sec. Fig.36: 7sF X 840 A u s t e n i t i z e d 20 sec. gig.37; 7tA X 840 Spheroidized 14 h r . gig.38: 7tD X 840 Austenitized 10 sec. gig.39: 7tE X 840 A u s t e n i t i z e d 15 sec. f i g . 4 0 ; 7tF X 840 A u s t e n i t i z e d 20 s e c . I ' - . ov P OO " fl b • .• e v * r . * V!, « - V - , ?ig.41; 7fr X 840 A u s t e n i t i z e d 30 sec. (36) Figures 26, 32, and kZ show the transformation at 280 seconds. There i s l i t t l e difference between these speci- mens and those austenitized for 3° seconds (specimens 'G-' ), the transformations being virtually complete in both cases. Fewer fine spheroids remain at 280 seconds; the coarser spheroids are slower to dissolve. This i s no doubt a matter of stability and concentration gradients. In the f i r s t , instance, the spheroid i s the most stable of the constituents in normal steels at temperatures below the Ae^; i t would naturally be expected to be the most sluggish to react to austenitization. In the second place, as the concentration of carbon in the austenite increases, the diffusion of carbon from the combined form to the solid solution decreases. The same considerations may be applied to explain the differences in rates of formation of austenite in pearl- i t e . The f i r s t austenite to form, Just above the Ac^ must be of eutectoid composition; i f the carbide lamellae are thin, a small growth of austenite includes the carbide lamellae and the intervening layers of ferrite; i f the lamellae are relatively thick, the same growth might not totally consume the lamellae or cross the intervening f e r r i t e . In the latter case, the carbon concentration gradient has been increased by interposing an intermediate concentration (austenite) between the carbide and the f e r r i t e . It i s also reasonable to assume that thicker plates of carbide are more (37) stable than thinner ones and, hence, retard the transformation to austenite. Austenitized 2 8 0 seconds. These, then, are the conclusions regarding the austenitization of spheroidite: 1. Nucleation takes place on the surface of each spheroid as a whole rather than at a point on a carbide-ferrlte Interface; l f the very f i r s t origin of a nuclei i_s a point, then that point expands almost instantaneously to form an "envelope of nucleation" covering the spheroid. 2 . Nucleation occurs preferentially at those spheroids lying in the ferrite grain-boundary network. 3. Austenite grows preferentially along the ferrite grain boundaries while the remainder of the growth is in the form of a lateral expansion to include the ferrite grains. 4. The more homogeneous the spheroidite, i.e., the more evenly distributed in size and position are the spheroids, the longer is the time required for austenitization. Ei£ii6:7nA X 715 ?ig.47;7nD X 715 Fig.48:7nS X 715 Lower B a i n i t e A u s t e n i t i z e d A u s t e n i t i z e d SAE 1080 10 seconds. 15 seconds. M a r t e n s i t e . SAE 1080 Fig.50:7oC X 840 Austenit i zed 5 seconds. F i g . 54:7pK X 840 Austenit i z e d 280 seconds. Fig.52:7sA X 840 S o r b i t e ( M a r t e n s i t e drawn at 800°F) SAE 1080 Fig.53:7nE X 840 A u s t e n i t i z e d 15 seconds. (38) 5» The larger the spheroids, the longer i s the time required for austenitization* Formation of Austenite form Other Structures A survey of the structures containing the carbon i n finer dispersion than in spheroidite or pearlite i s given i n Figures 43 to Figures 43 to 45 show the transformation of upper bainite; 46 to 48, lower bainite; 49 to $1, marten- site j and 52 to 5^, sorbite formed by tempering martensite one-half hour at 800°F, L i t t l e can be said regarding the nucleation of austenite in any of these structures, except that It appears to be general, or occurring at random© Prior to nucleation, an agglomeration of carbides takes place ln the martensite and sorbite, reducing them to what l s , In.effect, extremely fine spheroidite. Transform- ation has apparently started before 10 seconds (see Figure 47) and is almost complete in 15 seconds (see Figures 45, 48, 51, 53) at auetenitlzlng temperature* Specimens *E* also show the two balnites to be more completely transformed at this stage than either the martensite or the sorbite* If we consider the balnites as extremely fine pearlite, and the martensite and sorbite as extremely fine spheroidite, the reason becomes apparent; nucleation and transformation are effected more easily i n the lamellar structures. It ls lik e l y , though, that the balnites undergo some form of tempering, or agglomeration of carbides, before (38) 5* The larger the spheroids, the longer i s the time required f o r aust e n i t i z a t i o n * Formation of Austenite form Other Structures A survey of the structures containing the carbon i n f i n e r dispersion than l n spheroidite or p e a r l i t e i s given i n Figures 43 to 5̂ « Figures 43 to 45 show the transformation of upper b a i n i t e ; 46 to 48, lower b a i n i t e ; 49 to 51, marten- s i t e ; and 52 to 5^, sorbite formed by tempering martensite one-half hour at 800°F. L i t t l e can be said regarding the nucleation of austenite i n any of these structures, except that i t appears to be general, or occurring at random* P r i o r to nucleation,. an agglomeration of carbides takes place i n the martensite and sorbite, reducing them to what i s , i n . e f f e c t , extremely fine spheroidite. Transform- ation has apparently started before 10 seconds (see Figure 47) and i s almost complete i n 15 seconds (see Figures 45, 48, 51, 53) at au s t e n i t i z i n g temperature* Specimens 'E* also show the two balnites to be more completely transformed at i t h i s stage than either the martensite or the sorbite. I f we consider the balnites as extremely f i n e p e a r l i t e , and the martensite and sorbite as extremely fine spheroidite, the reason becomes apparent; nucleation and transformation are effected more e a s i l y i n the lamellar structures. I t i s l i k e l y , though, that the balnites undergo some form of tempering, or agglomeration of carbides, before i F i g . 55; 7f A X 840 a r l i t e . S A E 1080 Fig.56 ; 7fC X °40 .Austenitized 5 sec. Fig.57;7fD X 840 A u s t e n i t i z e d 10 seconds. Fig.58:7f0 X 840 Fig.59;7fK X 840 ,u s t e n i t i z e d \ u s t e n i t i z e d •50 seconds. 280 seconds. F.ifl.eo: 7gA X 730 F i g . 61: 7gC X 730 P a r t l y spheroidized A u s t e n i t i z e d 5 seconds. P e a r l i t e . SAE 1080 Fig.64: 7gK X 730 A u s t e n i t i z e d 280 seconds. (39) transformation occurs; the graphs in Plate IV, "Graphical Presentation of Results", indicates this to be so. The following conclusions regarding these finer structures appear reasonable: 1. Austenite forms in bainite by random nucleation and growth; the formation resembles that in pearlite but proceeds more rapidly. 2. Austenite forms from martensite and tempered martensite by random nucleation and growth; the structures react as though they were extremely fine, homogeneous spheroidite. 3. Bainite, martensite, and tempered martensite can be austenitized more rapidly than normal pearlite or spheroidized structures. General Considerations Figures 55 to 64 present a comparison of rates of austenitization in similar structures having different degrees of stability: Figures 55 to 59 show the transformation in a structure formed by annealing steel 7 for one hour at 1250°F directly after austenitizing one-half hour at 1450°F; Figures 60 to 64 show the transformation in a structure formed by cooling to and holding at the range 1280-1330°F, directly after the same austenitization cycle. The second prior structure appears similar to the f i r s t , being only slightly coarser. In both steels, the transformation has commenced at five seconds of subsequent austenitization (specimens 'C, Figures 56 and 6 l ) . These specimens are approximately five ?ig.67:5fG X 7 3 0 Fig.68:5fH X 7 3 0 Fig.69:r3fK X 7 3 0 A u s t e n i t i z e d A u s t e n i t i z e d A u s t e n i t i z e d 30 seconds. 40 seconds. 2 8 0 seconds. (40) per cent transformed. Specimens 'D* show a slight difference in that more transformation has occurred in the finer struct- ure; specimen 7fB i s over 90 per cent transformed, 7gD i s about 85 per cent transformed, at 10 seconds at austenitizing temperature (Figures 57, 62). At 280 seconds, specimen ?fK has transformed more completely than has specimen 7gK (Figures 59, 62). The following conclusion is indicated: 1. The more stable the structure, that i s , the more fu l l y i t has developed under similar conditions, the greater Is the time required for the austenitizatlon of that structure. A l l the conclusions so far presented in this section, with the exception of those obtained from the study of Figures 1 to 4, are founded on the examination of specimens cut from the same bar of steel. This steel i s Atlas "Maple Leaf Special", a commercial grade of eutectoid tool steel similar to S.A.E. 1080. The remaining photomicrographs, Figures 65 to 82, .show general studies of transformations occurring in a few other steels. Figures 65 to 69 show the progress of trans- formation in Atlas "KK? similar,to S.A.E. 52100. The structure i s well spheroidized and appears to transform in. the normal manner, with a notable exception; some ferrite remains at the end of 40 seconds at austenitizing temperature (Figure 68, specimen 5fH)» The explanation i s that chromium is retarding the transformation; this Indicates the following conclusion: Pig,70; 3fA ' X 730 P e a r l i t i c 4140 F i g . 7 1 : 3fD X 730 A u s t e n i t i z e d 10 seconds. f i g . 72: 3fG X 730 A u s t e n i t i z e d 30 seconds. Pig.73t 3fK X 730 A u s t e n i t i z e d 280 seconds. (41) l o The presence of chromium retards the formation of austenite* Figures 70 to 73, of a S.A.E. 4140 steel, support this conclusion. Unlike the transformation in ordinary pearl- i t e , the formation of austenlte in this case begins in the pearlite. bordering on the free fe r r i t e and some of the ferrite is dissolved before the transformation consumes the larger part of the pearlite (see Figures 71, 72). Figure 73, specimen 3fK, shows undissolved pearlite remaining even after 280 seconds of austeniti zation* It i s probable that most of the •pearl 31a i s concentrated in the interior of the pearlite and that the i n i t i a l transformation occurs i n the comparatively chromium-poor pearlite and fe r r i t e . Figures 74 and 75 show normal transformation in a S.A.E. 1045 pearlite. The transformation has dissolved the pearlite before growing Into the free fer r i t e . Inspection of the inclusions In Figure 75 .does not reveal that they have in any. way interfered with the reaction. Reference to Figures 1 to 5 establishes the same situation; inclusions in the S.A.E. 1020 apparently do not interfere with the growth of the austenite. The following conclusion i s presented: 1. Visible inclusions did not interfere with the nucleation or growth of austenite in any of the specimens examined during the experiments relating to 'this report. Figures 76 to 80 also show transformation in S.A.E. 1045; the pearlite i s a l i t t l e finer and more evenly d i s t r i - buted than in the specimens shown in Figures 74 and 75* P e a r l i t i c 1045, Aust .lOsec. A u s t e n i t i z e d 15 seconds. Fig.78:2a? X 1800 Pig.79;2aF X 600 Fig.80:2aG X 600 A u s t e n i t i z e d A u s t e n i t i z e d A u s t e n i t i z e d 20 seconds. 20 seconds. 30 seconds. (42) Figure 76 (specimen 2aD) shows l i t t l e or no transformation to have occurred at 10 sdconds austenitization. Figure 77, at 15 seconds, pictures a "weld" in the lamellae at the pearlite grain boundary; the growth i s evidently about to expand i n - ward along the lamellae. Other areas of nucleation are shown in the same Figure as random thickening of the lamellae. Figure 78 shows what has happened to a pearlite grain during 20 seconds at austenltizing temperature. The untrans- formed pearlite i s outlined by the "ghost" structure in the rest of the grain, now austenitized pearlite. Fig.81: 5aF X 1800 52100, Fig.82: 5aH X 1800 austenitized 20 seconds. 52100, austenitized 30 seconds. Figures 81 and 82 show stages in the transformation of Atlas "KK", similar to S.A.E. 52100. The "ghosts" of pearlite are typical of chromium steels and are further evidence that chromium retards transformation. The conclusions presented in this study, together with those from "Graphical Presentation of Results" are summarized and discussed ln the following section of this report. (43) DISCUSSION It is the plan of this discussion to re-state the conclusions as indicated in the sections on "Graphical Presentation of Results" and "Study of Mlcrostructures", and to compare these conclusions with those of the workers cited in "Review of Previous Work". Most of the following conclusions are not original but have been checked by the writer. Summary of Conclusions Pearlite 1. Austenite nucleates preferentially at the carbide-ferrite interfaces of the lamellae near the pearlite grain boundaries; second- ary and minor points of nucleation are at the interfaces of the lamellae within the pearlite and at positions of random d i s t r i - bution. C3,i4). 2. The growth of austenite is primarily along the lamellae at the carbide-ferrite inter- face in finger-like growths; secondarily, across the inter-lamellar fe r r i t e , expand- ing at the base of the "fingers" to include the neighboring lamellae; thirdly, across the lamellae paral l e l and close to the inner "wall" of the pearlite colony bound- aries. 0 5 ) . 3. Austenitic growth is retarded by pearlite colony boundaries, carbide lamellae, and areas of fe r r i t e , in that order, when in the path of the advancing austenite. (\5). (44) In eutectoid steel, the growth preferences result in the I n i t i a l austenite grains having a roughly rectangular section. Trans- formation i s generally complete before a l l traces of the lamellae are dissolved, 05>9)- In low carbon steel, the transformation of the pearlite Is complete, or nearly so, before growth advances in a finger-like manner, expanding laterally at the base, (»5). The transformation is the faster, the finer the interlamellar spacing of the pearlite, other variables being equal. C9). A l l observations contained in the conclusions regarding the austenitization of pearlite have been mentioned, at least in substance, by previous workers. Baeyertz (15) supports conclusions 2 to 5, inclusive, on the austenitization of pearlite; Roberts and Mehl (9) mention observations identical, in substance, with those of conclusions 1, 4, and 6; Digges and Rosenberg (14) are in accord with conclusion 1, Spheroidite 1. Nucleation may occur at a point on the surface of a spheroid. However, the point, i f such i t i s , expands so rapidly to cover the surface of the spheroid that nucleation may be con- sidered to commence as an "envelope" covering the surface of the spheroid. (4). 2. Nucleation occurs preferentially about those spheroids in or near the fe r r i t e grain bound- aries and where the spheroids are in the finest particles. (i5,5). 3. Austenite grows preferentially along the ferrite grain boundary network, expanding laterally to include the fe r r i t e grains. In the early stages the growth is roughly spherical, with finger-like projections proceeding along the grain boundaries. (i4) . 4. 5. 6. (45) 4. Growth does not readily cross ferrite grain boundaries but tends to f i r s t include the fe r r i t e grain in which the transformation commenced. The more homogeneous is the spheroldlte, the more random is the nucleation and the longer time i s required for trans- formation, 6. The larger the spheroids, the longer i s the time required for complete trans-formation. Conclusions 1, 2, and 3, relating to the austeniti- zatlon of spheroldlte, have been previously Indicated. The photomicrographs of V l l l e l a , reproduced by Bain (4), support conclusion 1; i t i s unfortunate that Bain gives no references to V i l l e l a ' s work and no analyses of the steels used. Con- clusion 2 is supported by Baeyertz (15) and Davenport and Bain (5). Conclusion 3 i s confirmed in the work of Dlgges and Rosenberg (14). Other Structures 1. Austenite forms from bainite by random nucleation and growth; rate and character of transformation are what would be expected in extremely fine pearlite, 2. Austenlte forms from martensite and sorbite by random nucleation and growth; the necessary tempering of these struct- ures reduces them to what resembles extremely fine, homogeneous spheroldlte,15*,9). 3. Bainite, martensite, and tempered martensite transform more rapidly than do normal p e a r l l t i c or spheroidized structures. o (46) The essence of conclusion 2, regarding martensite and sorbite has been given by Baeyertz (15) and by Roberts and Mehl ( 9 ) . General Considerations 1. The more stable the structure, that i s , the more f u l l y It has developed under similar conditions, the more time i s required for the austenitization of that structure. 2. The presence of pro-eutectold constituents retards the transformation to austenite; i n hypoeutectoid steels there are less total points of nucleation; in hypereutec- toid steels, the massive carbides require additional time to transform after areas of eutectoid composition have transformed. In either case, areas of eutectoid com- position transform f i r s t . (i5\3). 3. The presence of chromium retards the formation of austenite and tends to pro- duce "ghosts" i n the newly transformed structures. C22,23) 4. Complete transformation i s not necessary to the attainment of maximum hardness in pearlite, bainite, martensite, or tempered martensite. CS)- 5» Maximum hardness may be obtained in pearl- ite eutectoid steel by austeniting for 80 seconds; in bainite, martensite, and sorbite at 15 seconds: maximum hardness may be approached in spheroidite by austenltizing for 40 seconds and are attained in the coarser spheroidites only beyond 280 seconds. These statements apply only to steel 7 under the conditions of the current experiments. 6. Visible inclusions did not interfere with the nucleation or growth of austenite in any of the specimens examined during the current experimenta. (47) Conclusion 2 has been previously stated by Baeyertz (15) and by Roberts and Mehl (9); conclusion 3 i s generally mentioned in texts on physical metallurgy, including Heyer (22) and G i l l and co-workers (23). Conclusion 4 has been Indicated by Roberts and Mehl (9) (see "Graphical Presentation of Results"). 0 Conclusions 4, 5, and 6, under "Spheroldlte" and concerning growth retarding influences and relative rates of growth of austenlte in spheroldlte, are not specifically mentioned In any of the references included in the review. They have, however, been Inferred in such discussions and in studies relating to grain growth and other phenomena. No claim can be made as to the originality of these conclusions; they have perhaps been stated previously in sources not i n - vestigated by the writer. None,of the literature investigated contained any significant remarks pertaining to the austenitizatlon of bainite; nothing was found that confirmed conclusion 1, under "Other Structures". Except for the bainite, the substance of conclusion 3, under the same section, and regarding relative rate of transformation, has been inferred previously by many sources. Conclusion 1, under "General Considerations", regard- ing the effect of stability of structures, applies to the austenitizatlon of most structures. It can be expanded to explain why martensite, pearlite, and spheroldlte are (48) progressively slower to react to austenitization and why finer structures usually austenitize more rapidly than others* Conclusion 5, under the same section and presenting some quantitative results regarding rates of hardening, i s indicated by the current experiments; this conclusion may he applied with caution to transformations in steels similar to those used in these experiments* Conclusion 6, stating that visible Inclusions did not interfere with austenitization, is the only one actually at variance with the work of others; Baeyertz (15) states, i n substance, that undissolved particles of oxides, carbides, etc*, produce considerable discrepancy in the size of the austenite grains and, in general, tend to reduce the grain size* While admittedly the inclusions observed in the S.A.E* 1020 and 1045 during the current experiments were not identi- fied with respect to composition, they were not found to Interfere with the austenitization of these steels in any manner (see Figures 1 to 4, 74, and 75, "Study of Micro- structures" )• Some attempts were made to check the equations of Roberts and Mehl (9) (Equations 3 and 4, "Review of Previous Work"). Not enough data was available to give other than rough estimates of the variables Involved and calculated results were unsatisfactory, although the order of the, con- stants checked (calculated constants were between 0*2 and 0*5, as compared to 0*66 for the time of one-fifth reaction as given by Roberts and Mehl). (49) Considerably more research i s necessary i f the mechanism of the nucleation and growth of austenite is to be thoroughly examined. Some of the possible directions of such investig- ations are indicated in the following section, "Suggestions for Further Work", (50) SUGGESTIONS FOR FURTHER WORK Austenitizatlon is the necessary beginning of practically a l l the time-temperature cycles used for the heat treating of steel. It follows that a thorough knowledge of the mechanism of austenitization should be a prerequisite to such heat treating. There i s , however, much research to be performed before the complete f i e l d of austenitizatlon can be thoroughly examined. Some of ,the possible avenues of research on the formation of austenite are presented in this section. Suggestions Arising from Current Work Among the phases of austenitization not dealt with in this thesis are the effects of prior austenitic grain size, alloy content (nickel, chromium, e t c ) , and rate of heating on the nucleation and formation of austenlte. Considerably more quantitative work could be performed, dealing with rates of austenitization versus pearlite spacing, spheroid d i s t r i - bution and other variables. More work of the type advanced by Roberts and Mehl (9, 10) and by Johnson and Mehl (16) on the prediction of quantitative rates of nucleation and growth would be valuable* It should be possible to compile tables or graphs showing the times required at various rates of heating for (a) the attainment of completely transformed, homogeneous austenite (5D and (b) the degree of austenitization necessary for maximum , hardening versus composition and microstructure for many commonly used steels; the form of such graphs i s indicated in Plates II to VII, "Graphical Presentation of Results". Flame Hardening and Induction on Heating Actual rates of austenitization are of v i t a l interest to heat-treaters hardening steels by flame hardening or i n - duction heating. Since both processes are necessarily rapid for surface hardening, i t i s important to know how long a particular, steel must be austenitized before i t can be hard- ened. Austenitization by flame hardening would presumably require research such as conducted in the current experiments; Investigations Into the effect of large masses of cold metal adjoining the heated part, that i s , the effect of heat flow, on austenitization are indicated. Induction heating presents another phase of the problem, namely, the effect of the heating characteristics of the pro- cess on austenitization A review of the literature on i n - duction heating proved slightly confusing ln that some conflicting opinions were presented. Tran and Osborne (25,26) claim extremely fast rates of austenitization for the process and also thorough austenitization, including solution of free ferr i t e , of a low carbon steel below the Ae^. Also claimed Is a "superhardenlng" of steels by induction heating. Martin (52) and Wiley (27) and E l l i s (28) discuss these and similar phases of induction heating. Martin and Van Note (29) review l i t e r - ature relating to induction heating and austenitization with a discussion of the effect of the pre-austenitlzing micro- structure. Poynter (30) differs the most sharply of any with the claims of Tran and Osborne (25,26). He sayss "No evidence is found to indicate that induction heating results in more rapid solution and transformation rates - - -. While the evidence obtained regarding the so-called "superhardness1' i s not conclusive, i t seems l i k e l y that the i n - creased hardness i s related to the presence of Internal stresses and not to any difference in the metallurgy" The po s s i b i l i t i e s of research on the subject of austeni- tization are far more numerous than have been mentioned here. Time or space do not permit further discussion of the subject* In any case, these po s s i b i l i t i e s w i l l be evident to those familiar with the subject of austenitizatlon* (53) ACKNOWLEDGMENT The writer i s indebted to several members of the staff of the Department of Mining and Metallurgy at the University of British Columbia* Their assistance in the work connected with the preparation of this thesis has been of considerable value* The guidance of Professor F. A. Forward, Head of the Department, during the experimental work and his suggest- ions concerning the presentation of the thesis material were exjtremely helpful* W* M* Armstrong, Professor of Physical Metallurgy and under whose direction most of the work was conducted, donated much of his time and effort to unraveling many of the problems encountered; his assistance during the photomicrography was particularly valuable* Dr. C* S* Samis, Professor of Metallurgy, made several timely suggestions. The assistance of Messrs. R.C* Hammersley, D. A. Scott, and G. Bell in determining the analyses of the specimens used i s gratefully acknowledged* BIBLIOGRAPHY (1) Sachs, G., and Van Horn, K.R«, Practical Metallurgy. 1940, Am*Soc.Met. (2) Grosvenor, A.W., Seitz, F., and others Metal - Inside Out. 1941, Am.Soc.Met. (3) Grossman, M.A», "Grain Size in Metals, with Special Reference to Grain Growth in Austenite", Grain Size Symposium. 1934, Am.Soc.Met. (4) Bain, E.C., Functions of the Alloying Elements in Steel 1959, Am.Soc.Met. (5) Davenport, E.S., and Bain, E.C., "General Relations between Grain Size and Hardenabillty and the Normality of Steels", Grain Size Symposium 1934, Am.Soc.Met. (6) Grossman, M.A», "On Grain Size and Grain Growth" Trans. A.S.S.T.. Vol. XXI, No.12, 1933. (7) Carpenter, H.C.H., and Robertson, J.M., "Structural Changes in Hypereutectoid Steels on Heating", Jour. I.S.I.. Vol. 127, No. 1, 1933. (8) Hultgren, A., "The Ac-j_ Range of Carbon Steel and Related Phenomena", Trans. A.S.S.T.. Vol. 16, 1929. (9) Roberts, G.A», and Mehl, R.F., "The Mechanism and the Rate of Formation of Austenlte from Ferrite-Cementite Aggregates", Trans. A.S.M.. Vol. 31, 1943. (10) Mehl, R.F., "The Structure and Rate of Formation of Pearlite" Trans. A.S.M.. Vol. 29, 1941. (11) Rosenberg, S.J., and Digges,' T.G., "Effect of Rate of Heating through the Transformation Range on Austenitic Grain Size", Trans. A.S.M., Vol .29, 1941. - f (12) Dlgges, T.G., and Rosenberg, S.J., . "Influence of I n i t i a l Structure and Rate of Heating on the Austenltlc Grain Size of 0,5 per cent Carbon Steels and Iron-Carbon Alloys", J ourRes.. Nat.Bur.Stds., Vol. 29, 1942. (13) Dlgges, T.G., "Influence of Austenltlc Grain Size on the C r i t i c a l Cooling Rate of High Purity Iron Carbon Alloys", Trans. A.S.M.. Vol. 31, 1943. (14) Digges, T.G., and Rosenberg, S.J., "A Metallograph!c Study of the Formation of Austenite from Aggregates of Ferrite and Cementite in an Iron Carbon Alloy of 0.5 per cent Carbon", Trans. A.S.M.. Vol. 31, 1943. (15) Baeyertz, M., "Effects of I n i t i a l Structure on Austenite Grain Formation and Coarsening", Trans. A.S.M.. Vol. 30, 1942. (16) Johnson, W.A., and Mehl, R.F., Trans. A.I.M.M.E.. Iron and Steel Division, — — — v o l . 135, P . 4 l6, 1939. (17) Atlas of Isothermal Transformation Diagrams. 1943, United States Steel Corporation, Pittsburgh. d 8 ) gutting the Heat Treatment to toe Job, 1946 United States Steel Corporation, Pittsburgh. (19) G i l l , J.P., Tool Steels. 1944, Am. Soc. Met. (20) Metals.Handbook, 1939, Am. Soc. Met. (21) V i l l e l a , J.R., Metallographlc Technique for Steel. 1938, Am. Soc. Met. (22) G i l l , J.P., "Alloy Steels", Modern. Steels. 1939, Am. Soc. Met. (23) Heyer, R.H., Engineering Physical Metallurgy. 1941, Van Nostrand. (24) G i l l , J.P., Rose, R.S., Roberts,. G.A., Johnston, H.G., and George, R.B.. Tool Steels, 1944, Am. Soc. Met. (25) Trail, MoA*, ana Osborne, H.B., Jr., "Inherent Characteristics of Induction Hardening", Surface Treatment of Metals. 1941, Am. Soc. Met. (26) Osborne, H.B., Jr., "Surface Hardening by Induction", Trans. Electrochem. Soc.. Vol. 79, 1941. (27) Martin, D.L., and Wiley, P.E., "Induction Hardening of Plain Carbon Steels" Trans. A.S.M.. Vol. 3^ , 19^5. (28) E l l i s , O.W., "Pseudomorphs of Pearlite ln Quenched Steel" Trans,. A.S.M.. Vol. 32, 1944. . . (29) Martin, D.L., and Van Note, W.G., "Induction Hardening and Austenltizing Characteristics of Several Medium Carbon Steels", Trans. A.S.M.. Vol. 34, 1946. . (30) Poynter, J.W., "Metallurgical Characteristics of Induction Hardened Steel", Trans. A.S.M.. Vol. 3.4, 19^6.

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
United States 8 0
Japan 5 0
India 2 0
Germany 2 0
Canada 1 0
City Views Downloads
Unknown 10 1
Boardman 2 0
Moscow 2 0
Vancouver 1 0
Deltona 1 0
Tokyo 1 0
Ashburn 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}

Share

Share to:

Comment

Related Items