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

Подождите немного. Документ загружается.

origin, the presence of well-deﬁned subsurface

cracking was observed. This layer had the

appearance of smeare d metal and base metal

pullout. Flat cracking, suggestive of fatigue

cracking, was observed to emanate from the

ﬂaw (Fig. 68). The ﬂaws were located at

4.3 mm (0.017 in.) intervals, identical to the

feed rate of the injection drilling process.

During the injection drilling process, three

cutters are used. Coolant is forced through a

central hole to cool the cutting tools and to ﬂush

the chips. AerMet 100 tends to form long strands

of material during machining and does not want

to form chips. Hot chips can contact the freshly

machined surface. These chips or long strands

are under pressure at the cutter or follower

and can be forced onto the newly machined

surface by the follower. If the temperature and

pressures are high enough, solid-state welding of

the chips and bore surface can occur. As the

cutter boring bar moves, pullout can occur. The

Fig. 64

ASTM B7 low-alloy steel bolt grade. Fracture initiated along threads, with typical and pronounced beach marks (i.e., cyclic

fracture propagation) and transgranular fracture mode. (a) Location of bolts in pump coupling. (b) Beach marks showing

asymmetrical bending with initiation at high stress-concentration factor at bolt threads. (c) Transgranular fracture morphology

Fig. 65 Schematic of the failed arresting hook shank showing location of loads

Fig. 66

Machining marks found on the inside of the bore, at

the origin of cracking

82 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_43-86.pdf/Chap_02/ 18/8/2008 2:57PM Plate # 0 pg 82

examined ﬂaws matched the machining feeds

and speeds.

The arresting hook failed by fatigue,

initiating at ﬂaws created during the ﬁnal

machining process. The defect morphology

suggested localized solid-state welding and

pullout from chip contact with the freshly

machined surface. The surface roughness and

ﬁnish of the inner bore did not meet drawing

requirements.

Summary

In this short overview of the possible

mechanisms of failure for steels, the following

were discussed:

 Techniques for examining fractures

 Ductile and brittle failures

 Intergranular failure mechanisms

 Fatigue

The previous disc ussion has shown that it is

important to look at not only the fracture surface

but at all the factors (manufacturing history,

service conditions, and loading). All the tools

available to the metallurgist should be used—

these include photography, fractography, and

Fig. 67

SEM examination of the fracture surface. (a) Fatigue

striations emanating from the fracture origin. (b)

Machining marks found on the surface of the inner bore. (c) Well-

deﬁned layer showing fatigue emanating from the damaged

material at the surface of the inner bore

Fig. 68

Metallography of the arresting hook shank. (a) Typi-

cal quenched and tempered martensite found. This is

typical for the hardness of the arresting hook shank. (b) Pulled

material at 4.3 mm (0.17 in.) intervals along the inner bore of the

arresting hook shank. Origin is to the left. (c) Secondary cracking

observed at the location of pulled material

Overview of the Mechanisms of Failure in Heat Treated Steel Components / 83

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_43-86.pdf/Chap_02/ 18/8/2008 2:57PM Plate # 0 pg 83

metallography—to understand the sources and

root cause of failure.

REFERENCES

1. Hippocr ates, Epidemics, Book I, Section XI

2. D.S. MacKenzie, G. Chen, M. Barker,

C. Ramsay, J. Alderson, L. Dharani, and

J. Yu, Forensic Investigation of Failed Mast

Arms of Trafﬁc Signal Supported Struc-

tures, Transport. Res. Record, Des. Struc.

2002, 2002, p 9–16

3. “Loss of Control and Impact with Paciﬁc

Ocean Alaska Airlines Flight 261 McDon-

nell Douglas MD-83, N963AS about 2.7

Miles North of Anacapa Island, California,

January 31, 2000,” NTSB/AAR-02/01,

National Transportation Safety Board,

Adopted Dec 30, 2002, Notation 7263E

4. “Uncontained Engine Failure/Fire Valujet

Airlines Flight 597 Douglas DC-9-32,

N908VJ, Atlanta, Georgia, June 8, 1995,”

NTSB/AAR-96/03, National Transporta-

tion Safety Board, Adopted July 30, 1996,

Notation 6579A

5. W.M. Garrison and N.R. Moody, J. Phys.

Chem. Solids, Vol 48, 1987 p 1035

6. J.F. Knott, Advances in Fracture Research,

Proc. Seventh International Conference on

Fracture (Oxford, U.K.) 1989, p 125

7. J.F. Knott, Fracture 1977, Vol 1, Proc.

Fourth International Conference on Frac-

ture, June 1977 (Waterloo Canada), p 61

8. H.G. Wilsforf, Mater. Sci. Eng., Vol 59,

1983, p 1

9. B. Bøgner, G. Rørvik, and L. Marken, Bolt

Failures—Case Histories from the Nor-

wegian Petroleum Industry, Microsc. Micro-

anal., Vol 11 (Suppl . 2), 2005, p 1604–1605

10. T.L. Anderson, Fracture Mechanics, CRC

Press, New Yor k, 1995

11. C. Ruggieri, Numerical Investigation of

Constraint Effects on Ductile Fracture in

Tensile Specimens, J. Braz. Soc. Mech. Sci.

Eng., Vol 26 (No. 2), 2004

12. “Sta ndard Test Methods for Determining

the Inclusion Content of Steel,” E45–05,

ASTM International

13. “Premium Aircraft-Quality Steel Clean-

liness: Magnetic Particle Inspection Proce-

dure,” AMS 2300

14. “Aircraft-Quality Steel Cleanliness:

Magnetic Particle Inspection Procedure,”

AMS 2301

15. “Aircraft-Quality Steel Cleanliness: Mart-

ensitic Corrosion-Resistant Steels Magnetic

Particle Inspection Procedure,” AMS 2303

16. “Special Aircraft- Quality Steel Cleanliness:

Magnetic Particle Inspection Procedure,”

AMS 2304

17. R. Kiessling and N. Lange, Non-Metallic

Inclusions in Steel, Part I–IV, Iron and Steel

Institute, London, 1978

18. S. Maropoulos and N. Ridley, Inclusions

and Fracture Characteristics of HSLA Steel

Forgings, Mater. Sci. Eng. A, Vol 384 (No.

1–2), Oct 25, 2004, p 64–69

19. “Railway Investigation Report, Derailment

and Collision Canadian National Train

No. U-783-21-30 and Train No. M-30 6-31-

30 Mile 50.84, Saint-Hyacinthe Subdivision

Mont-Saint-Hilaire, Quebec, December 30

1999,” Report R99H0010, Transportation

Safety Board of Canada

20. M.L . Williams, STP 158, ASTM, 1954,

p 11–44

21. M.E . Shank, STP 158, ASTM, 1954, p

45–110

22. Fractography, Vol 12, Metals Handbook,

9th ed., ASM International, 1987

23. J. McCall and P. French, Metallography in

Failure Analysis, Plenum, New York, 1978,

24. “Standard Test Methods for Notched

Bar Impact Testing of Metallic Methods,”

E23–88, American Society for Testing and

Materials

25. “Standard Test Method for Plane Strain

Fracture Toughness, K

, of Metal lic Mate-

rials,” E399–74, American Society for

Testing and Materials

26. “Standard Test Method for Conducting

Drop-Weight Test to Determine Nil-

Ductility Transition Temperature of Ferritic

Steels,” E208–95A, American Society for

Testing and Materials, 1983

27. “Standard Test Method for Dynamic Tear

Testing of Metallic Materials,” E604–83,

American Society for Testing and Materi-

als, 1983

28. J.M. Kraft, A.M. Sullivan, and R.W. Boyle,

Proc. Symp. Crack Propagation, Cranﬁeld,

U.K., 1961

29. W.F . Brown and J.E. Strawley, STP 381,

American Society for Testing and Materi-

als, 1965, p 133

30. A.S. Tetelman and A.J. McEvily, Fracture

of Structural Materials, Wiley, New York,

1967

84 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_43-86.pdf/Chap_02/ 18/8/2008 2:57PM Plate # 0 pg 84

31. J.F. Knott, Fundamentals of Fracture

Mechanics, Wiley, New York, 1973

32. A.H. Cottrell, Proc. R. Soc. A, Vol 276

(No. 1), 1963

33. A.H. Knott, Mater. Sci. Eng., Vol 7 (No. 1),

1971

34. J.S. Colton, Class notes, “Great Boston

Molasses Disaster, Jan 15, 1919,” ME6222

Manufacturing Processes and Systems,

Georgia Institute of Technology

35. G. Krauss, Steels: Heat Treatment and

Processing Principles, ASM International,

1990, p 231, 237

36. G. Thomas, Retained Austenite and Tem-

pered Martensite Embrittlement, Metall.

Trans. A, Vol 10, 1979, p 1643–1651

37. S. Banerji, C. McMahon, and H.

Feng, Intergranular Fracture in 4340-

Type Steels: Effects of Impurities and

Hydrogen, Metall. Trans. A, Vol 9, 1978,

p 237–247

38. D. Kalderon, Steam Turbine Failure at

Hinkley Point ‘A’, Proc. Inst. Mech. Eng.,

Vol 186, 1972, p 341–377

39. L.D. Kramer and D. Randolph, Analysis

of TVA Gallatin No. 2 Rotor Burst: Part

1—Metallurgical Considerations, Proc.

1976 ASME-MPC Symposium on Creep-

Fatigue Interaction, p 1–24

40. H.K.D .H. Bhadeshia, High Performance

Bainitic Steels, Mater. Sci. Forum, Vol 500,

2005, p 63–74

41. I. Olefjord, Temper Embrittlement, Re-

view 231, Int. Met. Rev., Vol 23, 1978, p

149–163

42. B. Woodﬁne, Temper Brittleness, A Critical

Review of the Literature, JISI, Vol 173,

1953, p 229–240

43. F. Carr, M. Golman, L. Jaffee, and D. Buf-

fum, Isothermal Temper Embr ittlement of

SAE 3140 Steel, Trans. TMS-AIME, Vol

197, 1953, p 998

44. Liquid Metal Embrittlement, Failure Ana-

lysis and Prevention, Vol 10, Metals

Handbook, 8th Ed., American Society for

Metals, 1975, p 228–229

45. P. Fernandez, R. Clegg, and D. Jones,

Failure by Liquid Metal Induced Embrit-

tlement, Eng. Fail. Anal., Vol 1 (No. 1),

1994, p 51–63

46. G. Vilga nlte, G. Troiano, and C. Moss ey,

“Liquid Metal Embrittlement of ASTM

A723 Gun Steel by Indium and Gallium,”

ARCCB-TR-99011, Army Research,

Development and Development Center,

Benet Labora tories, Watervliet Arsenal,

June 1999

47. NTSB CHI97IA117, National Trans-

portation Safety Board, adopted June 28,

1998

48. NTSB Safety Recomme ndation, in memo to

J. Garvy, Administrator A-97-83, FAA,

Aug 29, 1997

49. S.F. Ca rter, Effect of Melting Practice on

Hydrogen, J. Met. (AIME), Vol 188,

Jan 1950, p 30–40 and J. Met. (AIME), Vol

188, Feb 1950, p 245–255

50. F.B. Foley, Flakes and Cooling Cracks

in Forgings, Met. Alloys, Vol 12, 1940,

p 442–445

51. F.J. Shortsleeve and A.R. Troiano, “A Lit-

erature Survey on the Mechanism of Flake

Formation,” Interim Technical Report 1,

Contract DA 33-019-ORD-897

52. C.A. Edwards, Pickling or the Action

of Acid Solution on Mild Steel and the

Diffusion of Hydrogen Through the Metal,

J. Iron Steel Inst., Vol 110, 1924, p 9–60

53. C.A. Zapffe and M.E. Haslem, Acid Com-

position, Concentration Temperature and

Pickling Time as Factors in the Hydrogen

Embrittlement of Mild Steel and Stainless

Steel Wire, Trans. ASM, Vol 39, 1947,

p 213–236

54. J.R. Gustafson, Some of the Effects of

Cadmium, Zinc, and Tin Plating on Springs,

Proc. ASTM, Vol 47, 1947, p 782

55. K.B. Valentine, Stress Cracking of

Electroplated Lockwashers, Trans. ASM,

Vol 38, 1947, p 488–504

56. C.A. Zapffe and M.E. Haslem, Measure-

ment of Embrittlement during Chromium

and Cadmium Electroplating and the Nature

of Recovery of Plated Articles, Trans. ASM,

Vol 39, 1947, p 241–258

57. G. Nagu and T. Namboodhiri, Effect of

Heat Treatments on the Hydrogen Embrit-

tlement Susceptibility of API X-65 Grade

Line-Pipe Steel, Bull. Mater. Sci., Vol 26

(No. 4), June 2003, p 435–439

58. A. Thompson and I. Berstein, Ed., Effect

of Hydrogen on Behavior of Materials,

TMS-AIME, 1976

59. H. Johnson, J. Morlet, and A. Troiano,

Hydrogen Crack Initiation and Delayed

Failure in Steel, Trans. TMS-AIME, Vol

212, 1958, p 528–536

60. “Release of Hazardous Materials from

Cargo Tank in Middletown, Ohio, on August

22, 2003,” Accident DCA03MZ002, NTSB

Overview of the Mechanisms of Failure in Heat Treated Steel Components / 85

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_43-86.pdf/Chap_02/ 18/8/2008 2:57PM Plate # 0 pg 85

Hazardous Materials Accident Brief,

Adopted July 22, 2004

61. A.W. Loginow, Stress-Corrosion Cracking

of Steel in Liqueﬁed Ammonia Service,

Bull., Natl. Board Boiler Pressure Vessel

Insp., Vol 45 (No. 6), Oct 1988

62. “Repor t by the Danish Accident Investiga-

tion Board into the Incident to Boeing 757,

TF-FIJ at Kastrup Airport, Denmark on

June 28, 2001,” M-04001/AIG-09, Danish

Accident Review Board

63. J.E. Dorn, Ed., Mechanical Behavior of

Materials at Elevated Temperature,

McGraw-Hill, New York, 1961

64. R.W. Guard, Prod. Eng., Vol 27 (No. 10),

1956, p 160–174

65. I. Finnie and W.R. Heller, Creep of Engi-

neering Materials, McGraw-Hill, New

York, 1959

66. E.N. da C. Andrade and B. Chalmers, Proc.

R. Soc. (London) A, Vol 138, 1932, p 348

67. J.B. Conway, Trans. Metall. Soc. AIME,

Vol 223, 1965, p 2018

68. F. Garofalo, Properties of Crystalline

Solids, STP 283, ASTM, 1965

69. W. Rosenhahn and D. Ewen, J. Inst. Met.,

Vol 10, 1913, p 119

70. W.J. Plumbridge and D.A. Ryder, Metall.

Rev., Vol 14, 1969, p 136

71. P.J. Forsyth and C.A. Stubbington, J. Inst.

Met., Vol 83, 1955, p 395

72. W.A. Wood, Some Basic Studies of Fatigue

in Metals, Wiley, New York, 1959

73. A.H. Cottrel and D. Hull, Proc. R. Soc.

(London) A, Vol 242A, 1953, p 211

74. W.A. Wood, Bull. Inst. Met., Vol 3, 1955,

75. C. Laird, “Fatigue Crack Propagation,” in

STP 415, American Society for Testing and

Materials, 1967, p 136

76. P.J. Forsyth and D.A. Ryder, Metallurgica,

Vol 63, 1961, p 117

77. G. Jacoby, Current Aeronautical Fatigue

Problems, J. Schijve, Ed., Perga mon, New

York, 1965, p 78

78. W.R .Warke and J.M. McCall, “Fracto-

graphy Using the Electron Microscope,”

ASM Technical Report We-2-65, American

Society for Metals, 1965

79. G. Jacoby, Fractographic Methods, Exp.

Mech., 1965, p 65

80. P.J. Forsyth, A Two Stage Process

of Fatigue Crack Growth, Symp. Crack

Propagation, Vol 2, Cranﬁeld, U.K., 1961,

p76

81. D. Broek and J. Schijve, “The Inﬂuence

of Sheet Thickness in the Fatigue Crack

Propagation in 2024-T3 Alclad Sheet

Material,” NLR Technical Report M2129,

Amsterdam, 1963

82. N.E. Frost, K.J. Marsh, and L.D. Pook,

Metal Fatigue, Oxford University Press,

London, 1974

83. T.T. Oberg and E.J. Wad, Technical Note

Report 56-289, Wright Air Development

Department, 1956

86 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_43-86.pdf/Chap_02/ 18/8/2008 2:57PM Plate # 0 pg 86

Mechanisms and Causes of Failures

in Heat Treated Steel Parts

Debbie Aliya, Aliya Analytical, Inc.

THE TERM mechanism of failure means

different things to different people. One possible

deﬁnition refers to a particular product appli-

cation, a particular type of component, and a

particular type of industry where certain envir-

onmental conditions are common. In many

cases, these mechanisms may apply only to a

limited type of material or an alloy family. In

order to understand particular failure mechan-

isms, it is important to understand the causes of

failure and categories of damage. This chapter

reviews various ways to classify failure cate-

gories. Information on mechanisms of damage,

in particular environment-material-type pairs, is

pro-vided toward the end of this chapter.

The term damage is often preferable to the

term failure, because damage is a technical term

that is very clear and has a speciﬁc meaning,

generally related to the physical condition of the

component. The term failure, on the other hand,

has many more philosophical connotations. It is

possible to see at once, perhaps, that the part is

damaged. Only at the end of the investigation

does one have a good chance of knowing with

a high degree of engineering certainty whether

the part itself failed, rather than the design, the

design system, the employee training system,

the procurement system, or the user-certiﬁcation

system. Thus, it is important to avoid using the

term failed part until one has truly determined

that the part was the problem.

Often, one of the important goals in a failure

analysis of a heat treated part is to determine

whether the damage is the result of improper

heat treatment, that is, the heat treater’s fault.

For many people in manufacturing, especially if

an independent or “job shop” heat treater is

involved, it is easy to blame the heat treater.

After all, he was the last one to touch the part

before assembly, and most people in the general

manufacturing arena understand heat treating

poorly, if at all.

If the part does not meet the speciﬁcation for

mechanical or physical characteristics after heat

treating, it may be the heat treater’s fault.

However, there may have been something

wrong with the raw material, or prior manu-

facturing processes that allowed a part that went

through the normal heat treating and inspection

process to have substandard properties, which

would not be the heat treater’s fault after all.

This chapter gives some examples of lack of

conformance to speciﬁcation that may at ﬁrst

look like the heat treater did something wrong,

but where other contributing factors made it

difﬁcult or impossible for the heat treater to meet

the speciﬁcation.

This chapter also summarizes the basic types

of damage, with particular consideration given

to whether their likelihood can or cannot be

inﬂuenced by the heat treating process. The

classical organization for types of damage

(failures) is as follows: deformation, fracture,

wear, corrosion or other environmental damage,

and multiple or complex damage. Separately

from what the damage type is, one also should

look at the potential causes, sources, or factors

promoting the damage.

Failure analysts used to be taught to classify

failures as a result of defects or abuse. There is

still a large amount of literature that presents

failure causes in such binary terms. By limiting

the analysis to one of two possible causes,

opportunities may be missed for improving

the product. There are actually several

different ways to classify causes of failures and

damage.

This chapter also describes a process that can

be used to demonstrate likelihood that a product

was abused. Sometimes, the physical evidence

speaks clearly. If this is not the case, it may be

necessary to quantitatively prove, for example,

that the part was overloaded. This may be dif-

ﬁcult to do, because exact service conditions are

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_87-109.pdf/Chap_02b/ 21/8/2008 10:50AM Plate # 0 pg 87

Failure Analysis of Heat Treated Steel Components

L.C.F. Canale, R.A. Mesquita, and G.E. Totten, editors, p 87-109

DOI: 10.1361/faht2008p087

www.asminternational.org

frequently impossible to determine with any

degree of certainty.

Besides service or maintenance abuse, four

potential sources or origins of damage and fail-

ure relate to the product life cycle, as follows:

 Design process omissions

 Undesired raw material characteristics

 Undesired component manufacturing char-

acteristics (includes heat treating)

 Improper service or maintenance conditions

In this chapter, a modiﬁed classiﬁcation is

developed speciﬁcally for failure analysis of

heat treated parts where the heat treating is

suspected to be the cause of the failure. The ﬁrst

logical potential source of poor component

performance is raw material characteristics.

Note the use of the word characteristics rather

than defects. This is an important thing to keep in

mind when performing this type of work. The

term defect has a speciﬁc legal deﬁnition that

does not necessarily apply to all cases of phy-

sical ﬂaws or suboptimal material properties.

Thus, the use of the word defect may have

undesirable cascading consequences, especially

when personal injury or large ﬁnancial losses

were a result of the component malfunction.

The second potential source of poor perfor-

mance is undesirable component characteristics.

For example, castings, forgings, and machined

or molded components may have discontinuities

or microstructural features present that make it

appear to the casual observer that there was a

heat treating problem. These discontinuities can

contribute to a poor heat treating outcome, a

poor service outcome, or both. The third poten-

tial trouble spot is design characteristics, many

of which may fall under the subclassiﬁcation of

inadequate attention to detail. The fourth cate-

gory of causes of failure in this scheme is true

heat treating process problems. Note that in this

classiﬁcation, these are, so far, all things that can

go wrong in the engineering and manufactur-

ing of the component. Finally, service and/or

maintenance abuse can be considered.

Types of Damage and Failure

The four basic types of damage (deformation,

fracture, wear, and corrosion) are brieﬂy re-

viewed in this section, and discussions address

whether each of the four types of damage can or

cannot be created or facilitated as a result of a

problem in the heat treating process.

Deformation

Figure 1 shows the result of a major defor-

mation event. Note the image of a cylindrical

structure. There is a dark sign-wave-shaped

band where the material is crumpled. The weight

of the structure above that area created the force

that caused the cylinder to suddenly deform.

There were also some tears or fractures as well,

but the main visible damage type is deformation.

There are different ways to categorize defor-

mation. One way is to compare gradual to sud-

den deformation. Gradual deformation can

occur when something is loaded and the load is

sustained. Due to the sustained load, the struc-

ture can stretch or bend. This type of deforma-

tion can happen during heat treating. Imagine a

part that has a protrusion that is not supported by

an appropriate ﬁxture. The part is heated to red

heat. Depending on the particular conﬁguration

and the presence of residual stresses in the

component, the protruding feature may droop

due to gravity, or it may change its shape in some

other way to relieve residual stresses from ear-

lier parts of the processing. It is important to

consider this type of deformation in a compo-

nent that has a critical dimensional envelope. If

the part goes outside of the speciﬁed or required

envelope and is placed into service, the stresses

that are experienced by the dimensionally non-

conforming component may be quite different

from the original design intent.

Gradual onset deformation may also happen

in parts that are loaded near the elastic limit.

It is important to note that the published elas-

tic limit for different materials is a numerical

quantity that is determined by using a speciﬁc set

of test parameters, usually of relatively short

duration. Long times at a stress level that may be

Fig. 1

Example of a sudden deformation event due to

buckling

88 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_87-109.pdf/Chap_02b/ 21/8/2008 10:51AM Plate # 0 pg 88

considered to be below the yield strength

may actually cause unacceptable permanent

deformation. Since the yield strength or elastic

limit can be affected by heat treating, it is poss-

ible, although probably not a frequent cause,

that this type of service-related gradual onset

deformation is also related to a heat treating

problem. Gradual onset deformation could also

be a result of loading the part beyond its design

intent or actual physical limitations.

The alternative to gradual onset of deforma-

tion is sudden onset. Examples of sudden onset

deformation are buckling instability in com-

pression of columns or torsion of tubes. A roof

truss that collapses under a snow load is such an

example. Roof trusses are not usually heat

treated for strength, at least not separately from

the thermal element of the hot rolling. However,

it is possible that annealed steel may be used in

some structural components. Most sudden onset

damage is primarily related to the basic geo-

metry and modulus of elasticity, which is not a

strong function of any heat treating process.

Thus, heat treating problems are generally a

minor or insigniﬁcant factor in most sudden

onset damage events.

Another way to classify deformation is by

level or degree: elastic (in other words, if the

load is removed, deformation is relieved) or

plastic (which is permanent deformation). A

possible example of an elastic deformation

failure is a spring that does not have the correct

spring constant. Imagine a coil spring that is

supposed to stretch out 0.1 mm when subjected

to 12 N of force. What if the spring stretches out

0.05 or 0.2 mm? Can that type of failure be due

to heat treating? What are the causes or the

factors that allow these two types and two levels

of deformation to happen?

As with sudden onset, the two main factors

that control elastic deformation are geometry

and modulus of elasticity. These are not factors

that are greatly inﬂuenced by the heat treater.

Thus, the heat treater is usually not at fault in the

case of sudden onset buckling of a column,

sudden onset buckling of a tube in torsion, or for

elastic deformation failures. In general, sudden

onset and elastic failures are a result of the

combination of the design or actual geometry

and the elastic modulus being insufﬁcient to

sustain the loading conditions. Sudden onset and

elastic deformation damage is only secondarily

related to the yield strength. The modulus of

elasticity is, in general, a constant. However,

modulus is not totally a constant in wrought

materials, where differences can occur due to

preferred crystal orientation. This is less likely

in heat treated steel than in steel used in the as-

worked condition. In general, in most heat

treated steels, the modulus of elasticity will vary

little. In attempting to determine the predomi-

nant factor or predominant cause in a sudden

onset deformation, one should look ﬁrst at the

geometry and the applied loading. Heat treating

issues would be considered in puzzling cases

that did not lend themselves to ready analysis.

For gradual onset deformation, the geometry

and the elastic modulus are still important, but

the heat treating can have a much more sig-

niﬁcant effect. If the yield strength is not as high

as it is supposed to be, unacceptable levels of

gradual onset deformation may occur. This

concludes the introduction on how to think about

deformation failures and whether the heat treater

even needs to be involved.

Figure 1 shows what was reported to be a

sudden onset event. This is actually an example

of a complex failure or damage mode, because

this structure was standing for quite a few

years, and then one day, the wind blew and

the damage happened. Investigators found evi-

dence of long-term corrosion on the inside,

associated with signiﬁcant wall thinning in some

areas.

As another example, if a hollow tube for a

truck drive shaft is not heat treated properly,

could it buckle more readily? If the damage is

buckling deformation, the primary factor would

be the wall thickness and the modulus of elas-

ticity. However, there may also be some strength

issues. When people talk about buckling due to

instability (or Euler’s buckling, or what is

referred to as sudden onset), the main factors are

the geometry and the modulus of elasticity.

However, with geometrical conﬁgurations that

are not exactly like the extreme examples that

Euler used to develop his theory, one can

appreciate that the equations become very com-

plicated and actually do have factors based on

strength values. However, the predominant

factors, in general, are the geometry and the

modulus. In failures of heat treated structural

steel, the modulus will be 30 million psi, with

some variations. For stainless steel, the modulus

may be a little different. If the component

material is a heavily cold-worked steel with

oriented, heavily textured microstructure, then

the modulus of elasticity may be different

in the different directions. However, most of

this orientation may be eliminated by any

Mechanisms and Causes of Failures in Heat Treated Steel Parts / 89

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_87-109.pdf/Chap_02b/ 21/8/2008 10:51AM Plate # 0 pg 89

subsequent heat treating that included an auste-

nitization phase. Note again that this discussion

is strictly speaking about deformation in the

absence of a primary fracture event.

Before leaving the subject of deformation,

consider a small spring clip that was designed to

be used at the top of the linear portion of the

stress-strain curve. Why would anyone design a

spring to be used at a stress almost at the nominal

yield strength? No experienced metallurgical

engineer would expect each one of a quarter of a

million such spring clips to sustain multiple load

cycles to the theoretical elastic limit and not

have any permanent shape change. These clips

could not even sustain the minimum 15 load-

cycle requirement without excessive permanent

deformation. This was a case of a nominally

elastic spring application where deformation

failure was caused by the heat treatment, speci-

ﬁcally by normal variations of the heat treating

process. It would be difﬁcult to blame the heat

treater in the absence of very strict quality-

control speciﬁcations. In this case, the major

cause of the failure was the design engineer’s

unrealistic expectations of uniformity of heat

treated components or the design process in the

company in which the designer was working.

Fracture

How can fractures of heat treated steel parts

be examined to determine the existence of any

factors related to the steel itself or its heat

treating? Basically, there are two types of

questions to ask. One of the things that people

like to ask is, “Why did this one break?” That is a

useful question in the case of a part that has been

used successfully for many years. Maybe there is

one part that failed out of a half-million parts that

are in service. Proper examination and evalu-

ation of the physical evidence can reveal much

to answer this question. Before evaluating the

effects of the heat treatment itself, one must ﬁrst

examine the physical condition of the damaged

property. It must be understood how the loads

interacted with the component to create new

surface area where none used to exist. The visual

appearance in three dimensions can reveal a

large amount of information on issues related to

how the part was really loaded. The colorations

and surface texture of the newly formed (unde-

sired) crack surfaces can indicate how the crack

happened and how long it took for the crack

to grow.

The other question that is frequently asked is,

“Why did this one break at this speciﬁc time?”

This is a different question, and to answer it,

fracture mechanics type of explanations and

theories must be explored. In such a case, one

assumes that every structure has some small

discontinuity and a related crack growth rate,

which is a function of the stress intensity (mea-

sured in terms of the mathematical product of the

nominal component stress and the crack size)

and the fracture toughness. These are functions

of service condition and material parameters.

Particularly for parts that appear to have broken

in fatigue, (consider a two-year service life when

no similar component had previously cracked in

under ﬁve years), it may be informative to look

at the microstructure and how that may have

impacted the fracture toughness. Microstructure

and fracture toughness could deﬁnitely be rela-

ted to heat treating issues. Other chapters in

this book give more information on fracture

mechanics.

For a more conceptual, lower-math-content

methodology to understand why something

cracked or why it broke, start by reviewing the

stress and strength variations that are at work in

the component. Any place where the local stress

exceeds the local strength can initiate a crack.

There can even be single grains that are low

strength for some reason. If the local strength is

lower than the local stress, then it is possible to

initiate a crack. Once a crack exists, it may or

may not progress to complete component frag-

mentation. Sometimes, cracks do not propagate.

However, cracks generally do not heal them-

selves, and they often do propagate. It is actually

reasonably straightforward to learn to look at

a component, if one has an understanding of

the loading geometry and the heat treating,

and determine if there was something wrong

with the component in question. Someone

inspecting a component with a complex shape

or complicated loading history may beneﬁt

from a good ﬁnite element analysis. However,

many components do lend themselves to ready

evaluation regarding the presence or absence of

an extraordinary factor promoting premature

fracture.

To use the stress/strength distribution concept

to analyze fractures, one should be familiar with

the six basic loading geometries, including

tension, compression, bending, torsion, contact

stress, and shear. A review of some of the

loading geometries is presented in Ref 1; how-

ever, the three-dimensional characteristics of the

90 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_87-109.pdf/Chap_02b/ 21/8/2008 10:51AM Plate # 0 pg 90

fragments are not systematically presented.

Armed with information on these basic fun-

damental loading geometries, one can learn to

predict macroscale fracture appearance in duc-

tile and brittle materials given different loading

conditions, along with the more well-known

fracture surface features, such as beach marks,

ratchet marks, and chevrons.

This is a much more powerful method of

visual examination than simply interpreting

surface texture features. With practice, one can

look at a broken component and obtain a realistic

idea of what loading conditions actually caused

the fracture. Sometimes, it can be shown that the

real loading was very different from the design

Before leaving this subject, a review is needed

on details that are often poorly understood

relating to how to distinguish ductile and brittle

cracks on the macroscale. This is clearly very

important for a heat treater to know to defend

against incorrect accusations of embrittlement.

There are some types of macroscale ductile

cracks that can easily be misinterpreted as brittle

cracks. In fact, much of the published literature

is unclear on this issue. To understand this more

clearly, see Fig. 2. The image is of a broken

tensile bar. Near the fracture, the material is

necked down. Because of the visible shape

change, this is an obvious ductile fracture. The

image in Fig. 3 is of a threaded fastener. Based

on the ridge patterns in evidence, the crack

started at the root of one of the threads and went

back into the page. There is very little indication

of any deformation visible at the macroscale.

This is correctly called a macroscale brittle

fracture.

Figure 4 has more challenging fragments for

fracture analysis. The image shows a chain link,

which did not originally have an open shape. The

two protruding ends were touching; there was no

gap. The lower portion of the link did not have a

curved arc shape as depicted. A signiﬁcant

amount of deformation was clearly associated

with this fracture event. Does that make this a

Fig. 2

Example of a macroscale ductile fracture in tensile

Fig. 3

Example of a macroscale brittle fracture in tensile

Mechanisms and Causes of Failures in Heat Treated Steel Parts / 91

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_87-109.pdf/Chap_02b/ 21/8/2008 10:51AM Plate # 0 pg 91