Canale L.C.F., Mesquita R.A., Totten G.E. Failure Analysis of Heat Treated Steel Components

Подождите немного. Документ загружается.

Fig. 43 Martensitic hardness as a function of carbon content in carbon and alloy steel. Source: Ref 80

Fig. 44

Dependence of the carbon gradient as a function

of case depth for three carburized steels that were

carburized under the same conditions: 925



C and 10 h. 1,

chromium-molybdenum steel (0.56% Cr, 0.16% Mo); 2, carbon

steel; 3, nickel steel (3.5% Ni)

Fig. 45

Microstructures of the carburized case structure

of two different samples of 20H steel that were

carburized in the same load. Carburizing temperature:

930



C for 7 h; hardening temperature: 860



C for 0.5 h. (a)

Surface. (b) Core. Etchant: 3% HNO

. Original magniﬁcation:

500·

214 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_177-240.pdf/Chap_06/ 18/8/2008 3:20PM Plate # 0 pg 214

established between the gaseous atmosphere and

the steel surface:

2CO

O CO+H

+2H

When the reactions proceed from left to right,

decarburization will occur. These are the reverse

of the carburization process.

Some of the more commonly reported causes

of decarburization include a malfunctioning

endogas generator, such as soot accumulation

hindering catalyst activity; excess moisture in

Fig. 47

Illustration of lath martensite in the core of carbur-

ized SAE 1524 steel; water quenched from 925



(1700



F). Etchant: nital. Scale = 10 mm. Courtesy of G. Vander

Voort, Buehler Ltd., Lake Bluff, IL

Fig. 48

Illustration of lath martensite in the core of carbur-

ized SAE 8115 steel; water quenched from 925



(1700



F). Etchant: nital. Scale = 10 mm. Courtesy of G. Vander

Voort, Buehler Ltd., Lake Bluff, IL

Fig. 46

Microstructure of SAE 8620 case of a mold taken just

below the surface. Etchant: alkaline sodium picrate

boiling (60 s), area just below the surface. Original magniﬁca-

tion: 500·. Courtesy of G. Vander Voort, Buehler Ltd., Lake

Bluff, IL

Fig. 49

Micrograph of poorly carburized SAE 8620 mold

showing decarburization at the surface (note patches

of ferrite and pearlite). Below this zone is where the grain-

boundary carbides are seen. Original magniﬁcation: 500· .

Courtesy of G. Vander Voort, Buehler Ltd., Lake Bluff, IL

Fig. 50

This is a higher magniﬁcation of the decarburized

microstructure shown in Fig. 49 of the surface of a

poorly carburized SAE 8620 mold (note patches of ferrite and

pearlite). Original magniﬁcation: 1000·. Courtesy of G. Vander

Voort, Buehler Ltd., Lake Bluff, IL

Sources of Failures in Carburized and Carbonitrided Components / 215

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_177-240.pdf/Chap_06/ 18/8/2008 3:21PM Plate # 0 pg 215

the furnace atmosphere; air contamination and

leakage (Fig. 51); heating in aged (deoxidized)

salt baths; and improper carbon potential selec-

tion (Ref 94).

If a carburized steel is at a temperature above

the Ac

(approximately 900



C) in a decarbur-

izing atmosphere, the carbon potential will be

low and the surface carbon content will also be

low, since the carbon in the steel surface and the

related gaseous reactions shown previously will

be driven to this equilibrium condition. This will

result in a decarburized layer being produced,

and the depth of the decarburized layer will

depend on residence time in the furnace under

these conditions.

If the temperature of the steel is below Ac

and above Ac

(800 to 840



C), there is a dif-

ferent decarburization condition. In this case,

the carbon content rapidly decreases from “A”

to “B,” as shown in Fig. 52 (Ref 94). Further

decreases in carbon content will result in a

material of carbon content “C” in equilibrium

with material of carbon content “B”. Therefore,

further loss of carbon by decarburization must

result in the formation of ferrite containing

carbon content “C”.

If the atmosphere carbon content is controlled

to carbon potential “D”, then ferrite cannot form.

Instead, a gradient is formed between carbon

contents “A” and “D”.

Decarburization is typically classiﬁed as total

or partial. Figure 53(a) illustrates a case of total

decarburization of 1018 steel and is charac-

terized by a ferritic surface layer (Ref 94).

Usually, there is a gradient from total to partial

decarburization with increasing depth from the

surface. Partial decarburization is illustrated in

Fig. 53(b) and is often characterized by grain-

boundary ferrite at the surface (Ref 94). In this

case, nital etching will reveal a structure more

gray in color than would be achieved with a

higher-carbon martensite. The quenched surface

of partially decarburized steel is typically bai-

nitic or martensitic. The ﬁnal microstructure

produced by a speciﬁc level of decarburization is

dependent on the steel alloy and cooling rate.

Decarburization is accompanied by surface

hardness reduction. However, partial or shallow

decarburization may not necessarily be detected

by macrohardness determination. If decarbur-

ization is suspected, microhardness determina-

tions, in view of their sensitivity to the presence

of such microstructures, should be used.

Decarburization can exhibit dramatic effects

on the residual stress of a component, as illu-

strated by Fig. 54 (Ref 94). In this example,

the surface carbon content of a carburized

3.5Ni-1.5Cr steel would be approximately 1%,

and the surface residual stresses would be

compressive at 4 392 MPa. After decarburiza-

tion reduces the carbon content to 0.64%,

the surface residual stresses were found to be

nearly 0. Finally, when decarburization re-

duced the surface carbon content to 0.35%,

Fig. 52

Iron-carbon equilibrium diagram to explain dec-

arburization

Fig. 51

Effect of air ingression into the carburization atmo-

sphere (N

/4% natural gas) on the decarburization of

SAE 8620 after 2 h at 850



C. Source: Ref 94

216 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_177-240.pdf/Chap_06/ 18/8/2008 3:21PM Plate # 0 pg 216

the surface residual stresses were tensile at

226 MPa.

Various reports have shown that decarbur-

ization can result in large decreases in bending

and contact fatigue strength (Ref 94). Since

wear-resistance properties are typically depen-

dent on achieving optimal hardness, reduction of

carbon content and thus surface hardness will

have a correspondingly adverse effect.

Inﬂuence of Grain Size

Grain size is one of the most important and

characteristic features of steel. Grain size in-

ﬂuences mechanical and plastic properties,

especially impact resistance and also steel har-

denability. Grain size is characterized by the size

of the austenite grain, and it is dependent on

various factors, such as degree of cooling and the

deoxidation process during steelmaking (Ref

95). Generally, an ASTM grain size of 6 to 8 is

speciﬁed. Figure 55 illustrates grain sizes of

ASTM No. 6 to 9.

Grain growth is affected by temperature.

Typically, grain growth increases with tem-

perature and time at temperature. Aluminum

may be added to steel to provide resistance to

grain growth (grain reﬁned). Alloying elements

such as nickel and molybdenum also provide

greater resistance to grain coarsening at typical

carburizing temperatures than plain carbon

steels. Although grain coarsening is usually not a

problem for carburizing temperatures up to

925



C, carburizing at temperatures greater than

1000



C is typically accompanied by some grain

coarsening, yielding a mixed-grain structure.

Reheating at 820 to 860



C can be performed to

reﬁne the mixed-grain structure.

Fine-grained steels are less hardenable than

coarse-grained steels with the same composi-

tion, and this generalization is true for case

structure also. The slower the cooling of the steel

during steelmaking, the larger the grain, since

there are fewer nucleation sites formed. Steels

Fig. 53

Micrographs illustrating total and partial decarbur-

ization. (a) Total decarburization of 1018 steel

caused by a furnace air leak. Etchant: 1% nital. Original magni-

ﬁcation: 500·. (b) Illustration of partial decarburization. Original

magniﬁcation: 190·

Fig. 54

Effect of decarburization on the residual stresses of

carburized and hardened 3.5Ni-1.5Cr steel. The

carbon content at 0.002 mm was approximately 1% for curve 1,

0.64% for curve 2, and 0.35% for curve 3.

Sources of Failures in Carburized and Carbonitrided Components / 217

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_177-240.pdf/Chap_06/ 18/8/2008 3:21PM Plate # 0 pg 217

are typically deoxidized by silicon and manga-

nese prior to aluminum addition. This will

inhibit the undesirable formation of AlN or

particles, which provide nucleation sites

for coarse-grained carbides that lead to the

formation of coarse ferrite/carbide (pearlitic)

grain structures on cooling.

Grain size is possible by control of the steel

composition during the steelmaking process.

Subsequent to this, control is by proper heat

treatment. Heating the steel to the upper critical

temperature, Ac

, will typically produce an

average minimum grain size. Heating to higher

temperatures will increase the grain size. Also,

quenching from the Ac

temperature will pro-

duce ﬁne grain size, and quenching from a

higher temperature would yield a coarser grain

size.

The austenite grain size at the onset of trans-

formation during the quenching process will

inﬂuence the martensite platelet size and thus

will affect microcracking potential, the amount

of retained austenite formed, and the frequency

and depth of internal oxidation (Ref 95).

Coarse grain structure is observed on etched

microsections, most often in the form of coarse

grain structure of martensite relative to ﬁner

retained austenite structure. Of the various fac-

tors affecting grain size, the primary factor is

furnace treatment. Although coarse-grained

steels exhibit better machinability, they gen-

erally possess lower toughness and ductility and

exhibit a greater tendency for distortion and

cracking than ﬁne-grained steels. Coarse-

grained steel also exhibits a more limited range

of thermal treatment temperatures, and they

possess better hardenability with higher as-

quenched hardness. Additionally, coarse-

grained steels typically possess lower impact

resistance and a lower yield point.

Carburized steel case structures with a

coarse-grained structure, along with signiﬁcant

amounts of overcarburizing, are characterized

by decreased mechanical properties. In this case,

the grain-boundary structure contains a con-

tinuous network carbide structure that is difﬁcult

to remove, and within the grain structure there

are typically acicular carbides. Such micro-

structures are fatigue sensitive, with cracking

throughout the brittle network carbide structure.

The presence of these defective microstructures

can be prevented by controlling the carburizing

temperature and carbon potential or by using a

steel with characteristically ﬁne-grained struc-

ture.

An example of grain size formation during

austenitization of steel 40 is shown in Fig. 56.

The irregular austenitic grain boundaries result

from the short heating time and also are related

to a prior normalizing and annealing process.

Fig. 55

Comparison of nominal ASTM No. 6 to 9 grain

sizes. Etchant: nital. Original magniﬁcation: 100· .

Source: Ref 96

218 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_177-240.pdf/Chap_06/ 18/8/2008 3:21PM Plate # 0 pg 218

This thermal history produces the variable grain

size shown.

Isogai et al. reported that grain sizes of

approximately 8 were required for carburized

transmission gear steel to achieve the required

high fatigue strength, pitting strength, and

impact strength in severe-use environments (Ref

97). To assure a ﬁne grain size, after the

carburizing process, the chromium-containing

steel, such as SCr420, was reheated to 820 to

870



C for 20 to 60 min in an atmosphere con-

taining the carburizing gas (carbon potential of

0.75%), quenched in oil until the steel was

120



C, and then tempered at 120 to 200



Treatment of the carburized steel in this way

permits substantial reductions in grain size and

corresponding improvements in fatigue strength

and impact strength. Similar improvements were

achievable with carbonitrided steel (Ref 97).

Internal Oxidation (Ref 77)

The depth of internal oxidation that originates

the carburizing atmosphere may vary from 1 to

30 mm (Ref 93, 98, 99). Internal oxidation con-

sists of a continuous oxide layer on the surface,

on the order of 0.01 mm, due to oxygen reaction

with the carburized steel surface. In this region,

the oxygen content can be 10 to 20 times that of

the core. During the gas carburizing process,

oxides will not only form on the surface but also

penetrate into the steel surface. Since internal

oxide formation is a diffusion process, the depth

and extent of oxide penetration is dependent on

the square root of the total carburizing time and

temperature, as well as the steel alloy chemistry.

The total depth of internal oxidation can be

calculated from (Ref 100):

=½(2D

 C

)=(n  C

)t=k

 t

=½k

 t

0:5

where X

is the depth of oxygen penetration, D

is the diffusion coefﬁcient of oxygen in the alloy,

is the oxygen concentration at the steel alloy

surface, C

is the concentration of the base

metal in the alloy (e.g., silicon, chromium,

manganese, titanium, vanadium), and n is the

stoichiometric factor.

Internal oxidation appears on the polished

metallographic specimen in the form of very

small inclusions concentrated in an austenite

grain or within the grain boundary. The prob-

ability of oxidation within the grain increases as

the grain size decreases (Ref 101). Figure 57

shows grain-boundary oxidation of carburized

20MnCr5 steel. Figure 58 also shows grain-

boundary oxidation, but it is accompanied by

nonmartensitic transformation products. Inter-

nal oxidation occurs in two zones: an inner zone

and an outer zone. Oxides of chromium-man-

ganese are typically formed in the outer zone,

both within the grain and in the grain boundaries.

In the inner zone, silicon-rich oxides are typi-

cally formed exclusively within the grain

boundaries (Ref 102). Figure 59 also illustrates

intergranular oxidation of a gas-carburized steel

(Ref 79).

Typically, the greater the case depth, the

greater the degree of oxide formation at the

Fig. 56

Prior-austenite grains formed in hardened steel 40,

which were due to abnormal growth during the

austenitizing process, Etched: S. Bechet and L. Beaujurda. Ori-

ginal magniﬁcation: 500·

Fig. 57

Illustration of grain-boundary oxidation of carbur-

ized 20MnCr5 to a depth of 30 mm. Unetched. Ori-

ginal magniﬁcation: 200 · . Courtesy of Fluidtherm Technology P.

Ltd., Ambattur, India

Sources of Failures in Carburized and Carbonitrided Components / 219

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_177-240.pdf/Chap_06/ 18/8/2008 3:21PM Plate # 0 pg 219

surface and the greater the depth of internal

oxidation. The thickness of the internal oxida-

tion zone typically is approximately 5% of the

carburized layer thickness and, on occasion,

may be as high as 10%. The critical cooling rate

is greater in the internal oxidation zone. As a

result, an otherwise normal hardening produces

a greater amount of bainite, which will lead to

lower surface hardness and poorer abrasive wear

resistance.

Typically, the oxides in the outer surface

region are globular in form, while intergranular

oxides were formed further from the surface. In

one analyis of a carburized chromium-manga-

nese steel, larger globular oxides were formed in

the region closer to the surface (1.9 mm depth)

and intergranular oxides in the region farther

from the surface (2.49 mm) when the steel was

heated for 16.6 h (diffusion at 2 h at 800



followed by 3 h at 930



C and a boost cycle at

930



C for additional heating times, in this case

11.6 h) (Ref 103). For this work, glow discharge

optical emission spectroscopy, in which a sput-

ter erosion process using ionized argon gas with

a voltage of 600 V and 25 mA was used to

quantify the degree of oxidation and elemental

distribution, was performed using energy-dis-

persive x-ray analysis. When the carburized

samples were subjected to shorter heating times,

intergranular oxides formed relatively farther

from the surface.

Transmission electron microscopy was used

by An et al. (Ref 103) to identify oxide type

and morphology. Chromium and manganese

globular oxides formed nearer the surface after a

total heat carburizing cycle of 5.8 h (5+0.8 h).

An agglomerated internal oxide of chromium

Fig. 58

Illustration of grain-boundary oxidation with non-

martensitic transformation products to a depth of

approximately 30 mm. Etchant: nital. Original magniﬁcation:

200·. Courtesy of Fluidtherm Technology P. Ltd., Ambattur,

India

Fig. 59

Intergranular oxidation of the surface along prior grain boundaries in a carburized steel. Original magniﬁcation: 1000· .

Source: Ref 78

220 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_177-240.pdf/Chap_06/ 18/8/2008 3:21PM Plate # 0 pg 220

and manganese oxides in the core surrounded by

silicon oxides formed after heating for 16.6 h

(5+11.6 h) (Ref 103).

Internal oxide formation within grain bound-

aries provides sites for crack initiation (Ref 10).

This was shown in a study by Laue et al.

(Ref 104), who evaluated the fatigue behavior of

case-hardened SAE 5115 steel with internal

oxidation of the case structure. Fatigue studies

were conducted on this steel, and it was shown

that fatigue crack initiation occurred along the

oxidized grain boundaries of the steel (Ref 104).

Internal oxides form as a result of oxygen

diffusion into the surface, with subsequent for-

mation of metal oxides at carburizing tempera-

tures. The formation of these oxides is enhanced

by the presence of metals, chromium and man-

ganese and Cr-Mn-Ti, which possess a greater

afﬁnity for oxygen than iron. The susceptibility

for internal oxidation increases with increasing

concentration of these oxide-forming elements.

Lohrmann et al., referring to earlier work by

Kozlovskii and co-workers, reported that the

depth of internal oxidation was dependent on the

total oxidation potential (TOP) of the steel alloy

(Ref 105, 106):

TOP=4:87 Si+3:7Mn+1:47 Cr 3:24 Ni

 1:82 Mo

where the elemental compositions are given in

weight percent. Figure 60 provides a correlation

of the TOP and depth of internal oxidation (Ref

106).

In addition, Kozlovskii et al. reported that

most steels, when subjected to the gas carbur-

izing process, will undergo internal oxidation

with a corresponding surface formation of

troostite to a depth of 0.01 to 0.03 mm. If

troostite is formed at a depth greater than

0.014 mm, there is a substantial decrease in

fatigue strength. However, the potential for

internal oxide formation can be reduced by the

addition of 5 to 10% of ammonia to the furnace

for 10 min before the carburizing process is

completed or by using steels containing 0.5%

Mo and not more than 0.5% Cr (Ref 106).

Lohrmann showed that the form and type of

internal oxide obtained was dependent on the

alloy composition of the steel. For example,

depending on the steel alloy, spot-, liner-, or

lattice-type internal oxides could be formed

(Ref 105).

Kehr and Seese examined the effect of internal

oxidation during carburizing of investment-cast

ingot-iron test specimen steels containing var-

ious amounts of chromium (0.20, 0.45, 0.91,

1.85, and 4.74% Cr) (Ref 107). It was shown that

steels containing approximately 0.50% Cr, for

Fig. 60 Correlation of calculated total oxidation potential (TOP) and average depth of internal oxidation

Sources of Failures in Carburized and Carbonitrided Components / 221

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_177-240.pdf/Chap_06/ 18/8/2008 3:21PM Plate # 0 pg 221

example, SAE 8620, were more susceptible to

internal oxidation than either plain carbon steel

or steels with greater amounts of chromium, such

as AISI 5120 (0.70 to 0.90% Cr). Figure 61

shows that the depth of oxide formation increases

as the chromium content increases up to 0.45%

(Ref 107).

The presence of internal oxidation on car-

burized steels greatly reduces bending and

contact fatigue strength and wear resistance

(Ref 93). However, surface oxides that are

formed may be removed by grinding or shot

peening. The potential for internal oxidation can

be reduced by heating steel to the carburizing

temperature under a nitrogen/hydrogen gas

mixture. During carburizing, these gases are

replaced by a carburizing atmosphere whose

oxygen activity is less than that required for the

formation of manganese II oxide or chromium

III oxide, and, in some cases, in the presence of

ammonia (Ref 100).

Finally, internal oxidation (and decarburiza-

tion) is known to lead to variations in the surface

compressive stresses in a carburized component

(Ref 108). There is one report of the presence of

internal oxidation leading to undesirable surface

tensile stresses that then led to subsequent

cracking of a carburized idler gear when used in

a diesel engine gearbox.

Carbides and Carbide Structure

Carbides formed during carburization are

treated as undesirable products to be avoided.

There are three types of carbides to be discussed

here: globular (or massive) carbides, network

carbides, and surface-ﬁlm or ﬂake carbides.

Carbides in steel are hard and brittle ceramic-

like interstitials with a high compressive

strength but low tensile strength (approximately

35 MPa, or 5000 psi).

Carbides in steel basically form when carbon

levels exceed the solubility limits of carbon in

the iron crystal structure. The allotropic nature

of iron also has different phase structures (i.e.,

crystal) with different solubility limits for car-

bon. For example, the maximum solubility of

carbon in the body-centered cubic (bcc) struc-

ture of ferrite is approximately 0.025 wt% at

723



C on the iron-carbon phase diagram (see

Appendix 9). For the face-centered cubic phase

of austenite (c), the maximum solubility limit of

carbon in c is approximately 2.06 wt% C at

1147



C. At still higher temperatures, another

type of bcc solid phase is d-ferrite. The max-

imum solubility of carbon in d-ferrite is

0.09 wt% C at 1493



C. A peritectic also occurs

at 0.16% C at 1493



The iron-carbon system has eutectic trans-

formation at 1147



C during soldiﬁcation, with

steel carbon levels of 2.06 to 6.67 wt% C. The

eutectic carbon concentration is 4.3%. Solid-

state transformations in steel include the well-

known eutectoid transformation at 733



C, with

a carbon concentration of 0.83%. At 733



austenite transforms to pearlite. Pearlite is a

eutectoid mixture containing 0.83% C and is

characterized by a ﬁne ferrite-cementite struc-

ture that forms upon austenite decomposition

during slow cooling.

The upper critical temperature (A

) is the

temperature below which ferrite starts to form

due to ejection from austenite in the hypoeu-

tectoid alloys. The lower critical temperature

) is the temperature where the austenite-to-

pearlite eutectoid transformation occurs. Aus-

tenite does not exist below this temperature. A

is the temperature below which ferrite is ferro-

magnetic.

Cementite (Fe

C) is the principal carbide of

iron and carbon, with an orthrhombic crystal

structure. Cementite is harder and more brittle

than ferrite. Steel hardness increases with in-

creasing cementite content. Increasing carbon

content increases the amount of cementite but

does not affect the amount of ferrite present,

since ferrite is saturated with 0.22% C. If there is

less than 0.83% C, iron and carbon will combine

to form Fe

C until no carbon remains. The

Fig. 61

Effect of chromium content of steel on the depth of

oxidation

222 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_177-240.pdf/Chap_06/ 18/8/2008 3:21PM Plate # 0 pg 222

cementite formed in this manner will combine

with the required amount of ferrite to form

pearlite, and any remaining ferrite will be in the

structure as free ferrite (proeutectoid ferrite).

Pearlite will form if the carbon content in the

austenite is 40.83%, and excess carbon will

form cementite. Excess cementite (proeutectoid

cementite) will deposit in the grain boundaries.

Because cementite (Fe

C) contains a speciﬁc

amount of carbon and iron, pearlite also contains

a speciﬁc amount of cementite and ferrite.

As a phase in steel, the chemical composition

of cementite will contain carbides of other car-

bide-forming elements, such as chromium and

manganese Alloying elements of chromium,

manganese, nickel, and other elements are, of

course, commonly used in alloy steels for

property improvement. They also impact the

properties of ferrite and cementite, because they

partition differently in the phases. For example,

chromium and manganese partition in cementite

instead of ferrite. However, nickel and silicon

tend to favor partitioning in ferrite. Chromium,

manganese, molybdenum, and titanium are thus

cementite stabilizers in steels, while nickel and

silicon are ferrite stabilizers. Interestingly, while

chromium, manganese, molybdenum, and van-

adium show no negative effect on cementite

formation, titanium, nickel, and silicon exhibit a

negative effect on cementite formation (Ref

109).

Cementite develops different morphologies

and distributions depending on the process of

cementite formation. Figure 62 illustrates three

microstructural forms of cementite: lamellar,

mixed, and granular (Ref 110). Cementite may

also be classiﬁed as reticular, acicular, or gran-

ular (Ref 110). Reticular cementite, also known

as shell-type cementite, possesses a crack-

sensitive network or platelet structure. Acicular

cementite, or needlelike structure, refers to a

lamellar structure of cementite in ferrite, shown

in Fig. 62(b). Finally, cementite exhibits a

granular or grainy appearance, as shown in

Fig. 62(c).

Cementite forms during soldiﬁcation from a

liquid or during solid-state transformations.

When cementite originates by crystallization

from a liquid melt, it is referred to as pri-

mary cementite (Fe

CI). Secondary cementite

(Fe

CII) is formed from austenite by hyper-

eutectoid alloys (carbon40.8%). Tertiary cem-

entite (Fe

CIII) is formed at temperatures below

723



C by precipitation in the grain boundaries

(which become richer in carbon with the

decreasing carbon content in a-iron).

Globular Carbides. Slowly heating a steel

to the carburizing temperature in the presence of

a carburizing atmosphere through the ferrite-to-

austenite temperature transformation region will

lead to unconnected globular carbide formation

either within the ferrite grains or at the former

ferrite grain boundaries, as shown in Fig. 63.

This process is favored by high carbon potentials

and also by reduction of normal heating rates

typically involved during carburizing, by

excessive furnace loading, or by a furnace mal-

function. Globular carbide formation may also

be enhanced by austenitic nuclei or by localized

concentrations of carbide-forming elements.

The problem of globular carbides may also

coexist with other problems, such as retaining

austenite or quench cracking.

When carburizing steels (0.15 to 0.25% C),

which are commonly ferritic with localized

areas of spheroidal carbides due to prior nor-

malizing and subcritical annealing, are heated

through the Ac

temperature, the high-carbon

regions begin to transform to austenite, resulting

in the formation of localized regions of carbon

and carbide-forming elements in addition to

undissolved carbides in the presence of the

Fig. 62 Cementite structures of CT60 steel with (a) lamellar, (b) mixed, and (c) granular cementite. Original magniﬁcation: 500·

Sources of Failures in Carburized and Carbonitrided Components / 223

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_177-240.pdf/Chap_06/ 18/8/2008 3:21PM Plate # 0 pg 223