05 heat treatments to produce ferrite and perlite

103

CHAPTER 5

Heat TreatmentsTo Produce Ferr teand Pea rlite

This chapter describes heat treatments such as full anneal-ing, normalizing and spheroidizing that have been developed to pro-duce uniformity in microstructure, improve ductility, reduce residualstresses, and/or improve machinability of steels. The microstructuresthat are produced by these heat treatments consist of various distribu-tions of ferrite and cementite, and therefore are produced by relativelylong holding times at temperature and slow cooling rates . The latterconditions permit the diffusion-controlled formation of desirable fer-rite and cementite microstructures . The heat treatments are describedin terms of the Fe-C diagram and the various transformation diagramsdiscussed in earlier chapters. Relationships for the mechanical prop-erties of mixtures of ferrite and pearlite as influenced by steel composi-tion are presented in the last section of this chapter.

Full Annealing and HomogenizingThe term annealing has been used in its broadest sense to refer to

any hect treatment that has as its objective the development of anonmartensitic microstructure of low hardness and high ductility. Thisunderstanding of annealing is much too broad, however, and a numberof more specific annealing heat treatments have been developed anddefined. Full annealing is a heat treatment accomplished by heatingsteels into the single-phase austenite field and slowly cooling, usually

1

104 / PRINCIPLES OF HEAT TREATMENT OF STEEL

if.in a furnace, through the critical transforfhation ranges . When the term {annealing is used without an adjective in reference to carbon steels,full annealing is the implied heat treatment practice (Ref 5.1).

Figure 5.1 shows the temperature ranges for severa¡ heat treat-ments involving austenitizing superimposed on the Fe-C diagram. Asshown , the temperature for full annealing is a function of the carboncontent of the steel, staying just aboye the A3 temperature forhypoeutectoid steels and aboye the Al for hypereutectoid steels. Thecritical temperatures will vary somewhat with the alloy content of thesteel, but the objective of heating into the single-phase austenite fieldfor low- and medium-carbon steels and finto the austenite -cementitefield for high-carbon steels remains the same no matter what thesteel composition.

The reason for heating the hypereutectoid steels in the two-phasefield is to agglomerate or spheroidize the proeutectoid cementite. Ifsuch steels are heated aboye Arm, proeutectoid cementite would formon slow cooling at the austenite grain boundaries, as discussed in l osChapter 2. The resulting network of carbides on the austenitic grainboundaries provides an easy fracture path and renders the steel brittleto forming or service stresses . Figure 5.2(a) shows a carbide networkdeveloped in SAE 52100 steel, a high-carbon bearing steel containingnominally 1% carbon and 1.5% chromium. Figure 5.2(b) shows howfracture produced by impact loading has followed the carbide network li^along prior austenite grain boundaries in a microstructure similar tothat shown in Fig. 5.2(a). In Fig. 5.2(b), the steel has been hardenedby quenching from the austenite -cementite field and martensite coex-ists with the carbide network. A carbide network formed on slowcooling from aboye Acm in a 52100 steel is shown in Fig. 5.3(a). Pearliteinstead of martensite has formed withín the austenite grains. Theobject of full annealing high-carbon steels in the austenite -carbidefield, then, is to break up such continuous carbide networks by

agglomeration into separated, spherical carbide particles. The driv-ing force for this process is the reduction in austenite/cementite inter-face area and thus the reduction in interfacial energy that accompaniesspheroidization. Figure 5.3(b) shows the partial spheroidization of acementite network. Although the structure was formed during austen-itizing for hardening, the austenitizing temperature ranges forhardening and full annealing are identical in high-carbon steels.

Not only is the temperature range of heating an important part offull annealing , but the slowcooling rate associated with full annealingis also a vital part of the process. Figure 5.4 compares schematictemperature-time schedules for full annealing and the normalizing

1

HEAT TREATMENTS TO PRODUCE FERRITE AND PEARLITE / 105

eC °F

1600 2912

1500 2732

1400 2552

1300 2372

1200 2192Hot working

and1100 Homogenizing 2012

1000 1832

Normalizing900 1652

800 Annealing 1472

700 1292

Carbon content in weight percent

Fig. 5.1. Portion of the Fe-C diagram with temperatureranges for fui¡ annealing , normalizing, hot workingand homogenizing indicated . (Courtesy of M. D. Geib,Colorado School of Mines , Golden , Colo.)

heat treatments discussed in the next section of this chapter. Thecooling rates are superimposed on a schematic CT diagram for ahypoeutectoid steel. The slow cooling rates characteristic of furnacecooling insure that the austenite transforms first to proeutectoid ferriteand then to pearlite at temperatures approaching the equilibrium A3and Al temperatures. As a result the ferrite will be equiaxed andrelatively coarse-grained, and the pearlite will have a coarse inter-lamellar spacing. The latter microstructural characteristics lower hard-ness and strength and increase ductility, the major objectives of the fullannealing treatment. Once the austenite has fully transformed toferrite ánd pearlite, the cooling rate could be increased to reduce thetime of annealing and thereby improve productivity. A number of

1


Fig. 5.2 . (a) Carbide network at prior austenite grainboundaries in 52100 steel. Light micrograph. Nitaletch. Magnification , 600X; shown here at 75%.(b) Fracture along grain boundary carbides in 52100steel. Scanning electron micrograph. Magnification,415 x ; shown here at 75%. (Courtesy of T. Ando,Colorado School of Mines, Golden, Colo.)

additional rules for developing optimum full annealing practicesand properties are given in Ref 5.1.

Figure 5.1 also shows the temperature range for homogenizing, atype of annealing treatment usually performed in earlier stages of steel

v

o-

t1


Fig. 5.3 . (a) Proeutectoid cementite network in normal-ized 52100 steel. (b) Residual cementite network afteraustenitizing structure in (a) at 850 °C (1562 °F) forhardening. Very fine particles are from spheroidiza-tion of cementite in original pearlite matrix. Arrowspoint to fine austenite grains that have formed onaustenitizing . ( Ref 5.2)

processing prior to hot rolling or forging, working operations that arealso performed in the same temperature range. Homogenizing isperformed at high temperatures in the austenite phase field to speedthe diffusion-controlled reduction of segregation or chemical concen-

1

MI,

108 1 PRINCIPLES OF HEAT TREATMENT OF STEEL

HeatingCyCle

a

F + A

•1

,

CoolingCyCle

P+A

Time Time

Fig. 5.4 . Schematic time-temperature cycles for nor-malizing and fui¡ annealing. The slower cooling ofannealing results in higher temperature transforma-tion to ferrite and pearlite and coarser microstructuresthan does normalizing . ( Courtesy of M. D . Geib, Col-orado School of Mines , Golden , Colo.)

tration gradients that are produced by ingot solidification. In addition,second phases such as carbides are dissolved as fully as possible. Theresulting uniformity or homogeneity of the austenite not only improveshot workability but also contributes to uniformity in the response ofa steel to subsequent annealing or hardening operations.

4

aNormalizing

Normalizing is a heat treatment which, similar to full annealing,produces a uniform microstructure of ferrite and pearlite. There are,however, severa] important differences between normalizing andannealing. Normalizing in hypoeutectoid steels is performed attemperatures somewhat higher than those used for annealing, while inhypereutectoid steels the heating temperature range is aboye the A,.,,,(see Fig. 5.1). In normalizing, heating is followed by air cooling, incontrast to the slower furnace cooling of fuil annealing.

1

1

HEAT TREATMENTS TO PRODUCE FERRITE AND PEARUTE / 109

The somewhat higher austenitizing temperatures used for normal-

izing , as compared to those used for annealing hypoeutectoid steels,in effect produce greater uniformity in austenitic structure and com-position similar to a homogenizing treatment, although at a muchlower temperature and for shorter times than those used forhornogenizing. Another of the major objectives of normalizing is torefine the grain size that frequently becomes very coarse during hotworking at high temperatures or that is present in as-solidified steelcastings. As such hot worked or cast produces are heated through theAcl and Acá temperatures, new austenite grains are nucleated and, ifthe austenitizing temperature is limited to the range shown in Fig. 5.1,a uniform fine-grained austenitic structure is produced. Exceeding theindicated temperature range might result in excessive austenitic grainsize, as discussed in Chapter 7. Normalizing, then, produces a uni-form, fine-grained austenite grain structure that in hypoeutectoidsteels transforms to ferrite-pearlite microstructures on air cooling. Theresulting microstructure may Nave good uniformity and desirablemechanical properties for a given application or may be reaustenitizedfor final hardening by quenching to martensite.

S

In hypereutectoid steels, normalizing is performed aboye the Armnot only to refine austenitic grain size but also to dissolve carbides andcarbide networks that may have developed during prior processing.The normalized structures that result respond more readily to thespheroidizing treatments for good machinability as described belowand/or provide better response to a subsequent and final hardeningheat treatment. There is the possibility that continuous carbide net-works may develop on cooling from a normalizing temperature aboyeAcm , and that as a result a somewhat brittle normalized microstructuremight develop. On subsequent austenitizing for hardening, however,the network carbides agglomerate or spheroidize somewhat (see Fig.5.3(b), and fracture toughness is in fact improved relative to a micro-structure without the partially spheroidized network (Ref 5.2).

The air-cooling step of a normalizing treatment produces subtle butsignificant differences in microstructures compared to those producedby full annealing. Figure 5.4 shows schematically that air coolinglowers the temperature range over which proeutectoid ferrite and

® pearlite form compared to the transformation range in full annealing.As a result, both the ferrite grain size and the pearlite interlamellarspacing are reduced compared to those in the same steel in the fullyannealed condition. The finer microstructure of a normalized steel inturn has higher strength and hardness and slightly lower ductility thana fully annealed steel.

The actual mechanical properties of any normalized or annealed

•1


steel are determined by a number of factors, the most important beingcarbon content. The higher the carbon content, the more pearlite thatforms and the higher the strength and hardness of the steel. Quantita-tive relationships for the contributions of carbon and other parametersto the mechanical properties of ferrite-pearlite steels are discussedin a later section of this chapter.

It is also important to realize that the air cooling associated with anormalizing heat treatment produces a range of cooling rates depend-ing on the section size . Heavier sections air cool at much lower ratesthan do light sections because of the added time required for thermalconductivity to lower the temperature of central portions of the workpiece. Two important consequences follow from the effect of sectionsize on cooling rate. In very heavy sections, the surface may cool atsignificantly higher rates than the interior, thus producing residualstresses . In very light sections, especially in alloy hardenable steels, aircooling may actually be rapid enough to produce bainitic or martensiticmicrostructures instead of ferrite and pearlite. The British Steel Cor-poration atlas (Ref 5.3) which plots cooling transformation as a functionof air-cooling section size (see Chapter 4) enables this effect to beevaluated. Other aspects of normalizing carbon steels are discussedin Ref 5.1.

Spheroidizing

The most ductile, softest condition of any steel is associated with amicrostructure that consists of spherical carbide partirles uniformlydispersed in a ferrite matrix. Figure 5.5 shows a spheroidized micro-structure in an 0 .66C-lMn steel. The high ductility of such a micro-structure is directly related to the continuous ductile ferrite matrix;pearlite (see Fig. 2.2) with its fine lamellar carbides separating theferrite, more effectively hinders deformation and, therefore, increaseshardness and lowers ductility compared to a spheroidized structure.The good ductility of spheroidized microstructures is extremely impor-tant for low- and medíum-carbon steels that are cold formed, and thelow hardness of spheroidized structures is important for high-carbonsteels that undergo extensive machining prior to final hardening.

Spheroidized microstructures are the most stable microstructuresfound in steels and will form in any prior structure heated at tempera-tures high enough and times long enough to permit the diffusion-dependent development of the spherical carbide partirles. As a result,there are many different heat treatment approaches for producingspheroidized microstructures. The slowest spheroidizing is associated

1


7-^17̂jU-,J•r•

P

pes{.¡J ..2Ci ¡1 ;•^'4 ,'-r îb•c^••.°°. e v;.l^ d^

V. 2:

:•$•.Q:^9• ^rp^ ;^g^d 0 <OM•..''•6p.G^^-d^'r+`^!oóá• ° á. ^. ^fJ b*.oj

Ve,

^ -y Qp^••p•O ° pop `^L oó•^.,.+^î- ..OJ - ^Ó .c'a LY oí 15^R, C.

o ^•^ •l i4 i^C.^ ^^s^C •rwLZ^p -tiY¡,..o% Ó..• e 0 ' o•^ é• ^•I

00p.-Q• ¡^̂ ! I /°.^^•YQ^ .p i^•OO ^̂ii^ ' 1^Dî x .01 ^:^^. • ^ ••'. `.

irq

• ♦ r %. ^ • •:. ., Î ! • ,' •i. • doy' O. 4 ♦ 4. •

r • • :G. ' `, ^

! e yy

Fig. S.S. Spheroidized microstructure in an Fe -0.66C-lMn alloy formed by heating martensite at 704 °C(1300 °F) for 24 h. Picral etch. Magnification , 1000 X.(Courtesy of A. R. Marder and A . Benscoter, BethlehemSteel Corp., Bethlehem, Pa.)

with pearlitic microstructures, especially those with coarse interlamel-lar spacings . Figure 5.6 shows the percent of carbides that have sphe-roidized in fine to coarse pearlites produced by isothermally transform-ing an 0.74C-0.71Si steel between 700 and 580 °C (1292 and 1076 °F),followed by annealing at 700 °C (1292 °F) (Ref 5.4). Many hundreds ofhours are required to spheroidize the pearlitic microstructures. Sphe-roidizing is more rapid if the carbides are initially in the form of discretepartirles , as in bainite, and even more rapid if the starting structure ismartensite. Spheroidizing of martensitic microstructures is most fre-quently performed on highly alloyed tool steels that form martensiteon air cooling as shown schematically in Fig. 5.7.

Spheroidizing at rates much faster than those shown in Fig. 5.6 isaccomplished by either complete or partial austenítizing, and thenholding just below Acl, cooling very slowly through the Acl, or cyclingaboye and below Ac, (Ref 5.1, 5.5). These temperature ranges forspheroidizing are shown in Fig. 5.8. It is important to limit the austeni-tizing temperature in order to retain a degree of heterogeneity in the


Fig. 5.6 . Progress of spheroidization at 700 °C (1292 °F)of fine , medium and coarse pearlites in a steel contain-ing 0.74% C and 0.71% Si. (Ref 5.4)

f

Time

Fig. 5.7. Schematic hect treatment cycle for spheroidiz-ing an air - hardening steel . Martensite forms first andthen is tempered ciose to the Ac, to produce a spheroid-ized structure . ( Ref 5.5)

11

1


0 0.5 1.0 1.5 2.0

Carbon content in weight percent

Fig. S.S . Portion of the Fe-C diagram with temperatureranges for procesa annealing , recrystallization anneal-ing, stress relieving and spheroidizing indicated.(Courtesy of M. D. Geib, Colorado School of Mines,Golden, Colo.)

austenite , especially since undissolved carbide particles appear topromote the transformation of the austenite to spheroidized micro-structures. As noted earlier, homogenized austenite free of undissolvedcarbides as produced by normalizing or full annealing promotes theformation of pearlitic structures rather than spheroidized structures.

Spheroidized microstructures are stable because the ferrite isgenerally strain -free and because the spherical shape of the cementiteparticles is one of minimum interfacial area per unit volume of particle.Lamellar cementite particles, as present in pearlite, have a very largeinterfacial area per unit volume of particle and therefore high interfa-cial energy. In order to reduce the interfacial energy, cementite lamel-lae or plates break up into smaller particles that eventually assumespherical shapes. Figure 5.9 shows a representation of the breakupprocess of a single plate as determined by serial sectioning of a speci-men annealed for 150 h at 700 °C (1292 °F) (Ref 5.4). Once the lamellaehave broken up, the small spherical particles dissolve at the expense ofthe larger particles, again driven by the reduction in interfacial energy.

1

1

P-


(la

Fig. 5.9 . Representation of a partially spheroidizedcementite place in coarse pearlite structure annealedfor 150 h at 700 °C (1292 °F). (Ref 5.4)

The following equation (Ref 5.4, 5.7) describes the yate of coarsening ofa spheroidized microstructure:

d r _ 2-y V Fe?C Xv D (] _ 1dr V Fe RT<< 'r rl

4F p • q 1

of spheroidization is directly related to the diffusion of carbon inferrite and decreases as the average size of particles in a spheroidizedmicrostructure increases. Alloying elements slow the rate of carbon

le-

(Eq 5.1)

where y is the interfacial energy; VFe3C and VFe are the molarvolumes of cementite and ferrite; X.. is the mole fraction of carbon inequilibrium with cementite in ferrite; D fr is the effective carbondiffusion coefficient; R is the gas constant; T is the absolute temper-ature ; ri is the radius of newly created particles; andr is the mean sizeof the alreadv e heroidized articIes E uation 5 1 shows that the rafe


diffusion in ferrite, and thétefore the spheroidization process. Alsoif present, strong carbide-forming elements would have to diffusefor alloy carbide coarsening, and thus would greatly reduce the rateof spheroidizatíon.

Process and RecrystailizationAnnealing

Process and recrystallization annealing are similar subcriticalannealing treatments usually applied to restore ductility to cold workedsteel products of a variety of shapes. Since these heat treatments areperformed in the ferrite and cementite two-phase field of the Fe-Cdiagram (see Fig. 5.8), no phase transformation accompanies the mi-crostructural changes produced by the treatments. Generally, the mi-crostructure of the low- and medium-carbon steels prior to cold work isspheroidized or largely ferritic with small amounts of pearlite, bothhighly ductile microstructures. The ferrite in these microstructures isequiaxed and strain-free. Cold working deforms or work hardens theferrite, tending to elongate the ferrite grains in the direction of work-ing and introducing a high density of crystal imperfections within thegrains. On heating, high strain energy of the deformed ferrite at firstdrives recovery, a mechanism by which some of the crystal imperfec-tions are eliminated or rearranged into new configurations (Ref 5.8),and eventually drives recrystallization, a process where new, strain-free equiaxed grains nucleate and grow in the deformed ferrite(Ref 5.8). The end result of the subcritical annealing process is a resto-ration of the ductile, spheroidized microstructure which is againcapable of undergoing significant cold deformation.

Figure 5.10 shows the effects of cold work and recrystallizationannealing on a very low-carbon (0.003%) sheet steel (Ref 5.9). Becauseof the low carbon content no carbides are visible and the only second-phase particles present are oxide and sulfide inclusions. Figure 5.10(a)shows the deformed microstructure produced by a 60% reduction insheet thickness by cold rolling. The elongated grain structure andetching effects associated with the deformation within the grains areclearly visible. Figure 5. 10(b) shows the same steel after about 80%recrystallization has been produced by annealing at 538 °C (1000 °F) for2 h. The recrystallized ferrite grains are quite fine and equiaxed, andetching shows no evidence of deformation.

A third stage of the annealing process, grain growth or coarseninghas not yet developed in the example shown in Fig. 5.10(b). Prolongedheating beyond the point of recrystallization would cause such grain

1

116 / PRINCIPLES OF HEAT TREATMENT OF STEEI

Fig. 5.10. (a) Microstructure of an Fe-0.003C alloy coldrolled 60%. (b) Microstructure of an Fe -0.003C alloyafter annealing at 538 °C (1000°F) for 2 h. About 80%of structure has recrystallized to fine equiaxed grains.Light micrographs. Nital etch. Magnification, 100X;shown here at 75%. (Courtesy of D. A. Witmer, Beth-Iehem Steel Corp., Bethlehem, Pa.)

growth to occur, a process driven by the reduction in grain boundary orinterfacial energy made possible by an increase in grain size. Refer-ence 5.8 describes the mechanisms of recovery and recrystallizationstages of annealing in detail.

r

HEAT TREATMENTS TO PRODUCE FERRITE AND PEARUTE / 117

Stress Relieving

A number of thermal and mechanical processes produce residualstresses that might be detrimental to the performance of fabricatedsteel parts or assemblies . The residual stresses may cause distortion,cracking during heat treatment or processing , or failure below designstresses in service. One source of residual stresses is the cooling ofheavy sectíons after austenitizing. Even during air cooling , the surfaceof a heavy section may transform to ferrite and cementite well beforethe center. When the center eventually transforms, the volume expan-sion associated with the ferrite formation is restrained by the cooler,alreadv transformed surface. As a result the center is compressed and

the surface put in tension . Quenching to form martensite produces asimilar but even more severe residual stress problem, even in smallersections , and is one reason why hardenable steeis are alloyed to permitmartensite formation at lower cooling rates. Martensitic steels, asdiscussed in Chapter 8, are invariably tempered , a process that re-duces residual stresses and increases ductility and toughness . Machin-ing and cold work also may introduce residual stresses in steel, due todifferences in the amount of deformation between the surface andinterior regions of a part . Welding is another process which producesresidual tensile stresses . As the weld metal solidifies and contracts, it isrestrained by the adjacent base metal . As a result, stress-relievingtreatments are frequently specified for welded assemblies.

Residual stresses are reduced or eliminated by subcritical heattreatments performed at temperatures either somewhat below oroverlapping those used for process or recrystallization annealing(see Fig. 5.8). Heating to and cooling from the stress-relief tempera-ture must be done slowly, especially in heavy sections or large weldedassemblies , in order to avoid introducing new thermal stresses andpossible cracking during the stress-relief treatment itself (Ref 5.10).

The objective of stress relieving is not to produce major changes ofmechanical properties by recrystallization as do the subcritical anneal-ing treatments discussed in the previous section . Rather the stressrelief is accomplished by recovery mechanisms that precede recrystal-lization, a situation that is aided by the difference in the kineticsbetween the two mechanisms (Ref 5.8). Recovery starts almost im-mediately on heating and reaching temperature . The rate of recovery isvery high initially and decreases with increasing time at temperature.Recrystallization , on the other hand, reguires an incubation period andstarts very slowly. Therefore it is possible to relieve residual stresseswithout changing mechanical properties significantly. For example, a

1


recent investigation (Ref 5.11) of stress xelief in cold extruded mildsteel bars shows that residual stresses are almost completely relievedwithout any hardness decrease after heating at 500 °C (932 °F) for 1 h.The latter result shows that the strengthening produced by cold work-ing can be used without the harmful effects of residual surface tensilestresses present in the as-deformed condition. Other aspects of stressrelieving are discussed in Ref 5.12.

Mechanical Properties ofFerrite-Pearlite Microstructures

As discussed earlier, the objective of normalizing and full-annealingheat treatments of carbon steels is to produce microstructures consist-ing of ferrite and pearlite. Some relationships that show the effect ofvarious parameters on the mechanical properties of steels containingferrite and pearlite are described here.

Figure 5.11 shows a set of mechanical properties for ferrite-pearlitemicrostructures as a function of steel carbon content. Yield and tensilestrengths increase and reduction of area, a measure of ductility, de-creases with increasing carbon content because of the increase inpearlite content. The microstructures vary from essentially 100% fer-rite to 100% pearlite as carbon is increased to 0.8%, the eutectoidcarbon content. The divergence of the yield and ultimate strengthcurves with increasing carbon content indicates that pearlite increasesthe work hardening rate.

In addition to the mechanical properties that characterize thestrength and ductility of steels, toughness or the energy absorbedduring fracture is also of considerable engineering importance. Fer-ritic steels are unique in that they show a transition from ductíle tobrittle fracture when broken at successively lower temperatures.Generally the transition for a given steel and microstructure is deter-mined by breaking a series of V-notched bars by impact loading attemperatures aboye and below room temperature. The specimens andmachine for the testing are standardized in what is known as theCharpy impact test (Ref 5.15). The ductile fracture typical of highertemperatures proceeds by the growth of microvoids around carbidesand/or inclusion particles, a fracture process that requires largeamounts of shear or plastic deformation and, therefore, absorbs con-siderable energy. In contrast, the low-temperature brittle fracture offerrite proceeds by cleavage between {100} planes of the ferrite grains.Little plastic deformation accompanies cleavage, and therefore littleenergy is absorbed during this type of fracture.

1


200

160

120w

80

nT

40

0

100

80

60

40

Notched Impact Tests

Smooth TenslleTests20

00̀

x

Reduction in area

1 1 1 1 1 1_ 10.1 0.2 0.3 0 . 4 0.5 0.6

Weight Percent Carbon0.7 0.8

160

120

80

40

0

- 40

0.9

Fig. 5.11. Mechanical properties of ferrite-pearlitemicrostructures as a function of carbon content.(Ref 5.13)

Figure 5.12(a) shows a mixture of cleavage and ductile fracture onthe overload fracture surface of a mild steel. Particles are associatedwith some of the large conical boles in the fracture surface, and clustersof fine dimples characteristic of ductile fracture are visible in the upperportions of the micrograph. Figure 5.12(b) shows primarily cleavagefracture and is characterized by very fíat fracture facets. Cleavage steps(arrows) from one set of cleavage planes to another in a given grain arealso a characteristic feature of cleavage fracture.

When energy absorbed during impact fracture is plotted as afunction of testing temperature, a transition curve results. Figure 5.13shows a family of transition curves for steels containing from 0.11% to0.80% carbon. The energy absorbed by ductile fracture is known as the


Fig. 5.12 . (a) Mixture of cleavage and ductile fracture.Note fine dimples characteristic of ductile overloadfracture in upper part of micrograph. Magnification,S00 x ; shown here at 75%. (b) Cleavage fracture.Arrows point to steps between cleavage planes ofdifferent elevation in a given grain. Magnification,1000 X ; shown here at 75%. Scanning electron micro-graphs . (Courtesy of D. Yaney, Colorado School ofMines , Golden, Colo.) 1

19

11


711%Carbon0200 .

p_i150 20% Carbon0

fl 100Ú

.

0.31 % Carbon^qG 0.41 % CarbonE 0.49% Carbon

C 50 60% Carbon0 .0.69% Carbon0.80% Carbon

0PA -150 -100 -50 0 50 100 150 200.Test temperature, °C

P̂ Fig. 5 . 13. Change in impact transition curves with- increasing pearlite content in normalized carbon

p

steels . ( Ref 5.18)

n

"shelf energy" because it reaches a plateau or is essentially constant asa function of temperature. Figure 5.13 shows that increasing carboncontent lowers shelf energy, and that, therefore, increasing amounts of

pearlite adversely affect the ductile fracture toughness. The transitiontemperature marking the transition between ductile and brittle frac-ture is also adversely affected by increasing carbon content. Figure

5.13 shows that high-carbon steels with large amounts of pearlite havehigh transition temperatures and therefore will fail in a brittle mannereven well aboye room temperature. Low-carbon steels, on the otherhand, have subzero transition temperatures and are quite tough atroom temperature. Changes in shelf energy and transition tempera-ture (taken as the testing temperature where fracture is 50% cleavage

and 50% shear) as a function of carbon content are summarized in theupper part of Fig. 5.11.

At any given cárbon content, the mechanical properties and tough-ness of a steel may be significantly affected not only by pearlite contentbut also by ferrite grain size and chemical composition. The refinementof ferritic grain size, for example, increases both strength and tough-ness. Good steel mili and heat treatment practice is therefore directedto producing as fine a ferrite grain size as possible for critical applica-tions. Alioying, low finishing temperatures for hot rolling, and low

flC1

1

1

122 / PRINCIPIES OF HEAT TREATMENT OF STEEL

austenitizing temperatures for normal'izing are all techniques used tokeep grain size small. Recently, a new class of ferritic steels has beendeveloped to take advantage of the high strength and toughness of veryfine-grained steels. The steels are referred to as high-strength low-alloy (HSLA) steels (Ref 5.16), and usually contain less than 0.2%carbon. Grain size control is achieved by microalloying with smallamounts of vanadium or niobium that produce very fine carbides. Thecarbides limit austenite recrystallization and/or grain growth duringhot rolling at low finishing temperatures (see Chapter 7) and as a resultthe ferrite that forms from that austenite on cooling is remarkably fine.

The effects of the various microstructural and composition param-eters on the mechanical properties of steels with ferrite-pearlitemicrostructures have been statistically analyzed by multiple linearregression analysis (Ref 5.17). The resulting empirical equations arelimited to steels containing less than 0.25% carbon and therefore tomicrostructures that are largely ferritic. The following selected equa-tions for yield strength and impact transition temperature illustrate theeffects of the various parameters:

Yield strength (MPa) (±31 MPa) =

K + 37(%Mn) + 83(% Si) + 2918 (Nf) + 15.1(d")(Eq 5.2)

Impact transition temperature (°C) (-t30 °C) =

-19 + 44(%Si) + 700N f - 11.5(d" ) + 2.2(% pearlite)(Eq 5.3)

In these equations, K is 88 M Pa for air-cooled steel and 62 M Pa forfurnace-cooled steel (Ref 5.20); N f is the free nitrogen dissolved in theferrite lattice (i.e., not combined as a stable nitride); and d is the meanlinear intercept in polygonal ferrite (mm). The beneficial effect of theferrite grain size on both strength and toughness is apparent froni theequations. The effect of manganese and silicon is to increase both yieldand tensile strength by solid solution strengthening of the ferrite. Themanganese and silicon replace ¡ron on the bce lattice of ferrite, and aresaid to dissolve substitutionally. Nitrogen, however, dissolves intersti-tially and the equations show that it not only is a very potent strength-ener of mild steels but also significantly promotes brittle cleavagefracture. The equations also show that the colonies of pearlite inlow-carbon steels have no statistically important effect on yieldstrength. However, other equations show that pearlite does increase

e

e

r

C HEAT TREATMENTS TO PRODUCE FERRITE AND PEARLITE / 123

tensile strength , an effect attrilltited to the decrease in yield extensionproduced by larger pearlite content (Ref 5.17). Figure 5.11 shows asimilar greater effect of pearlite on ultimate strength than on yieldstrength, especially beyond the 0.25% carbon for which Eq 5.2 andEq 5.3 hold.

In addition to the largely ferritic microstructures of mild steels with

carbon contents less than 0.25%, ferrite-pearlite microstructures inmedium -carbon steels (Ref 5.21) and fully pearlitic microstructures ofeutectoid steels (Ref 5.22) have also been analyzed to correlate variousmicrostructural parameters with mechanical properties. Whereas low-carbon steels are used for structural applications where relatively lowstrength but good ductility and toughness are required for formabilityand service conditions, the eutectoid steels are used where high hard-ness and wear resistance are of major concern. Railroad rails are anexample of an important application where the latter properties areimportant, and the need for improved rail steels has led to the develop-ment of the following equations for yield strength and toughness ofpearlitic steels (Ref 5.22):

o Yield strength (MPa) --

9

C

2.18(S-") - 0.40(P-W)- 2.88(d-t)+ 52.30 (Eq 5.4)

Transition temperature (°C)

- 0.83(P-w) - 2.98(d-") + 217.84 (Eq 5.5)

where S is the pearlite interlamellar spacing, P is the pearlite colonysize, and d is the austenite grain size.

Equations 5.5 and 5.6 show that the most important parameteraffecting yield strength is the interlamellar spacing while the transitiontemperature is affected most strongly by the austenite grain size fromwhich the pearlite develops. The transition temperature for fullypearlitic steels is invariably aboye room temperature, and thereforecleavage on the {100} planes of the ferrite in pearlite is the characteris-tic fracture mode. The size of the cleavage facets is a strong function ofbut always smaller than the austenite grain size and appears to berelated to common ferrite orientations in several adjoining pearlitecolonies (Ref 5.22). Air cooling or normalizing of eutectoid rail steels

containing only manganese and silicon produces fine interlamellarpearlite spacings, generally on the order of 2000 A (200 nm) (Ref 5.24).Alloying of rail steels with vanadium, chromium and molybdenum

produces even finer pearlite interlamellar spacings (Ref 5.25 - 5.27).

1

124 1 PRINCIPIES OF HEAT TREATMENT OF STEEL

Referentes5.1 Heat Treating of Carbon and Low-Allov Steels, in Metals Handbook, Vol

2, 8th Ed., American Society for Metals, Oh.,1964, p 1-105.2 K. Nakazawa and G. Krauss, Martensite and Fracture in 52100 Steel,

Met Trans A, Vol 9A, 1978, p 681-6895.3 M. Atkins, Atlas of Continuous Cooling Transformation Diagrams for

Engineering Steels , British Steel Corp., Sheflield, 19775.4 S. Chattopadhyay and C. M. Sellars, Quantitative Measurements of

Pearlite Spheroidization, Metallography, Vol 10, 1977, p 89-1055.5 E. C. Bollason , Fundamental Aspects of Molybdenum on Trans-

formation of Steel, Climax Molybdenum Co., London5.6 P. Payson, W. L. Hodapp and J. Leeder, The Spheroidizing of Steel by

Isothermal Transformation, Trans AS11I, Vol 28, 1940, p 306-3325.7 R. L. Fullman , Measurement of Particle Sizes in Opaque Bodies, Trans

AIME, Vol 197, 1953, p 447-4525.8 P. C. Shewmon, Transformations in Metals, McGraw-Hill, New

York, 19695.9 D. Witmer and G. Krauss , Effect of Thermal History on the Recrystal-

lization Behavior of Low Carbon 0.305 Mn Steels Containing Oxygenand Sulfur, Trans ASM, Vol 62, 1969, p 447-456

5.10 A. Vinckier and A. Dhooge, Reheat Cracking in Welded StructuresDuring Stress Relief Heat Treatments, J Heat Treating, Vol 1, 1979,p 72-80

5.11 M. B. Adeyemi, R. A. Stark and C. F. Modlen, Isothermal Stress Reliefof Cold Extruded Mild Steel Rods, Proceedings of the Metals Society,Heat Treatment '79, Birmingham, England, 22-24 May 1979

5.12 K. E. Thelning, Steel and Its Heat Treatment, Bofors Handbook, 1075,Butterworths, London, 1975

5.13 A. R. Rosenfield, G. T. Hahn and J. D. Embury, Fracture of SteelsContaining Pearlite, Met Trans, Vol 3, 1972, p 2797-2804

5.14 K. W. Burns and F. B. Pickering, Deformation and Fracture of Ferrite-Pearlite Structures, J Iron Steel Inst , Vol 202, 1964, p 899-906

5.15 G. E. Dieter, Jr., Mechanical Metallurgy, McGraw-Hill, New York,1961, p 371-375

5.16 Microalloying 75, Union Carbide Corp., 1977, distributed by AmericanSociety for Metals, Metals Park, Oh.

5.17 F. B. Pickering, The Effect of Composition and M icrostructure on Duc-tility and Toughness, in Toward Improved Ductility and Toughness,Climax Molybdenum Development Co., Japan, 1971, p 9-31

5.18 F. B. Pickering, The Optimization of Microstructures in Steel and TheirRelationship to Mechanical Properties, in Hardenabilit y Concepts withApplications to Steel, D. V. Doane and J. S. Kirkaldy (Eds.), AIME,Warrendale, Pa., 1978, p 179-228

'c

5.19 R. W. K. Honeycombe and F. B. Pickering, Ferrite and Bainite in AlloySteels, Met Trans, Vol 3, 1972, p 1099-1112

15.20 T. Gladman, D. Dulieu and I. D. Mclvor, Structure-Property Rela--tionships in High-Strength Microalloyed Steels, in Microalloying 75,

b dUnion Car e Corp., 1977, distributed by American Society for Met-ials, Metals Park, Oh.; p 32-54 4


5.21 T. Gladman, I. D. Melvor •and F. B. Pickering , Some Aspects of theStructure -Property Relationships in High-Carbon Ferrite-PearliteSteels , JISI, Vol 210, 1972, p 916-930

5.22 J. M. Hyzak and 1. M. Bernstein , The Role of Microstructure on theStrength and Toughness of Fully Pearlitic Steels , Afet Trans, Vol 74,1976, p 1217-1224

5.23 Y-J. Park and 1. M . Bernstein , Mechanism ofCleavage Fracture in FullyPearlitic 1080 Rail Steel, in Rail Steels - Developments , Processingand Use, STP 644, ASTM, 1978, p 287-302

P 5.24 D. E. Sonon , J. V. Pellegrino and J. M . IVandrisco, A MetallurgicalExamination of Control-Cooled, Carbon -Steel Rails with Service-Developed Defects , in Rail Steels - Developments, Processing andUse, STP 644, ASTM , 1978, p 99-117

5.25 G. K. Bouse , 1. M. Bernstein and D. H. Stone ' in Rail Steels -Developments , Processing and Use, STP-644, ASTM, 1978, p 145-166

5.26 S. Marich and P. Curcio, in Rail Steels - Developments , Processingand Use, STP 644, ASTM, 1978, p 167-210

5.27 Y. E. Smith and F. B. Fletcher, in Rail Steels - Developnwnts , Process-ing and Use, STP 644, ASTM, 1978, p 212-232

C:

n

1