CONDITIONS LEADING TO QUENCH CRACKING OTHER THAN SEVERITY OF QUENCH

R.R. Blackwood1 L.M. Jarvis1, D.G. Hoffman2 and G.E. Totten3 1. Tenaxol Inc., Milwaukee, WI USA 2. Metallurgical Services Inc., Chicago, IL 3. Union Carbide Corporation, Tarrytown NY USA

ABSTRACT One of the greatest concerns in the heat treating process is cracking of finished and semi-finished material during quenching. Many times, the only source of cracking considered is the quenchant. However, there are various sources of steel cracking other than the quench which must be considered. A tutorial on microstructural identification of various sources of cracking including: quench severity, prior steel structures, transformation temperature range, surface condition and others will be provided here. INTRODUCTION One of the greatest phenomenon of most ferrous materials is their ability to be heated and cooled to produce higher physical properties. This procedure consists of heating the metal to a high temperature (austenitizing) for a specified time to complete transformation to austenite and diffusion of constituents and then cooling in a quenching medium that produces the desired microstructure and as-quenched hardness. This hardening treatment is most often followed by a lower temperature heating process (tempering) for stress relieving and finalizing the required microstructure to achieve the necessary physical properties. This sequence is illustrated in Figure 1. A major concern of the heat treater is cracking of finished and semi-finished material

during the quenching cycle. Besides the monetary expense of the material involved, there is a loss of production time which is in addition to the monetary loss.

Figure 1 - Diagram of hardening and tempering cycle. Unfortunately, when cracking is encountered, it is often attributed to the severity of the quenching medium without microstructural verification. Although excessive quench severity is often the cause of quench cracking, there are many other sources that must also be considered. This paper will provide a tutorial on microstructural identification of various sources of steel cracking during heat treatment including: quench severity, prior steel structures, transformation temperature range, surface condition and others.

DISCUSSION

A. Quench Cracking Related to Severity of Quench One major source of cracking is excessive cooling rates during the quench (quench severity). This is illustrated in Figure 2. Note that the crack passes straight from the surface to the core. Excessive cooling rates (high quench severity) will produce greater thermal stresses in addition to greater transformation stresses. (Steel cracking during transformation to martensite is largely due to the volumetric increase that accompanies martensite formation.) If the total residual stresses in the part are sufficiently high, quench cracking, as illustrated in Figure 2, will occur.

parts prior to the quench, incompatible fluid contamination (oil -in-water or water-in-oil), excessive foaming and air entrainment, etc. All of these phenomena are capable of producing increased thermal gradients resulting in quench cracking. The effect of non-uniform quenching was illustrated with forged AISI 1045 crankshafts. Microstructural examination showed a mixture of non-uniform cross-sectional microstructures. Areas of tempered martensite adjacent to pearlite, bainite, acicular ferrite and ferrite at prior austenite grain boundaries were observed. . Some of the microstructures are illustrated in Figures 3 and 4.

Figure 2 Micrograph of AISI 4340 quenched and tempered steel illustrating macroetched pure quench crack. B. Quench Cracking Related Uniform Heat Transfer to Non-

Figure 3 - Micrograph of AISI 1045 steel asquenched and tempered, microstructure shows bands with banded tempered martensite and some bainite. The crack profile revealed evidence of tempering oxide and secondary cracking. (Magnification: 200 X; Etchant 2% Nital.)

It is important to note that quenchantrelated problems, other than excessive quench severity, are also often major contributors to steel cracking. These include: non-uniform fluid flow around the part due to poor racking of the

Figure 4 - Micrograph of AISI 1045 steel as quenched and tempered, illustrates representative under-heated microstructure observed adjacent to cracking. (Magnification: 400 X Etchant 2% Nital.) C. Prior Steel Structure The structure of the steel prior to hardening, e.g., extruded, cast, forged, coldformed, etc., may enhance the potential for cracking during the quench. Each as-formed structure requires a specific time and temperature cycle to condition the material for proper hardening. For example, cast structures must be homogeneous, cold-formed structures require normalization and annealing and forged structures must be grain-refined by normalization. Cracking may be caused by microstructures resulting non-uniform heating or cooling. In Figure 5, microstructures obtained on forged AISI 403 stainless steel valve stems that exhibited longitudinal cracking after quenching. Microstructural analysis suggested that cracking was caused by thermal stresses during forging or during heating prior to forging. Figure 5 illustrates evidence of a coarse grain condition associated with high-temperature surface oxidation. Further examination revealed evidence of high and low thermal oxidation within the crack profile. The presence of this condition suggests cracking occurred prior to, or during, forging. Figure 5 - Micrograph of AISI 403 stainless steel as-forged; microstructure is predominantly a mixture of carbide particles in a matrix of ferrite. No evidence of quenching and tempering was observed. High and low temperature oxidation can be observed on the surface of the sample and within the crack profile. (Magnification 100X; Etchant Villelas) D. Non-Uniform Heating Cracking may also be due to localized over-heating is provided in Figure 6 which is a microstructure of AISI 4140 tube end sections. Circumferential cracking at a mid-thickness location of the sample was reported on the tube end. The tubes were reportedly austenitized and then spray-quenched. Microstructural examination of the steel revealed a coarse grain condition due to overheating during austenitization prior to quenching. The cracking occurred along the coarse grain boundary as illustrated in Figure 6. However, microstructural analysis of samples from other regions of the tube indicated fine grain martensite. Taken together, these data suggest that the austenitization furnace contained hot spots which caused localized overheating and grain coarsening. The overheated locations cracked in the presence of quench stresses. Cracking occurred in the mid-thickness locations

due to the inherent weakness of the material centers carried over from the original billet or casting.

Figure 6 - Micrograph of AISI 4140 steel as quenched and tempered; microstructure is tempered coarse grain martensite with intergranular quench cracking along the prior austenite grain boundaries.. (Magnification 100X; Etchant 2% Nital) E. Excessive Heating Rate One source of cracking, which may appear similar to quench cracking, can also occur due to excessive heating rate to the austenitizing temperature. This is illustrated for AISI 4140 steel illustrated in Figure 7. In this case, surface oxidation and decarburization within the crack, which would not have been obtained if the cracking occurred during the quench was observed.

Figure 7 - Micrograph of AISI 4140 steel as quenched and tempered; microstructure is tempered martensite with evidence of decarburization and high-temperature oxidation on the surface of the crack profile. (Magnification 50X; Etchant 2% Nital) F. Transformation Temperature Range The steel transformation temperature range (MS - Mf) may exhibit substantial effects on cracking propensity. Table 1 provides a listing of the MS and temperatures of a number of selected steel alloys with varying hardenability. [1] In general, as steel hardenability increases, there is a greater depth of transformation to martensite producing greater transformational stresses. In addition, if the Mf temperature is sufficiently low, the transformation to martensite will be incomplete and resulting in increases retained austenite. There will be a corresponding increase in residual stresses as the retained austenite slowly transforms to martensite with time.

Table 1 Martensite Start (MS) and Martensite Finish (Mf) Temperatures Obtained on Selected AISI Steels A.I.S.I. Quench Ms M f - oF No. Temp oF 1065 1500 525 300 1090 1625 420 175 1335 1550 640 450 3140 1550 630 440 4130 1600 710 550 4140 1550 640 525 4340 1550 550 330 4640 1550 640 490 5140 1550 630 460 8630 1600 690 540 8695 1550 275 -9442 1575 620 410 Figure 8 - Micrograph of AISI 4140 steel as quenched and tempered: microstructure is tempered martensite with quench crack in the area of dimensional change. (Magnification: 100X; Etchant 2% Nital)

Excessively high austenitizing temperatures increase the surface-to-core temperature differentials which results in a corresponding increase in residual stress and cracking potential. G. Stress Risers from Prior Machining, Laps and Seams Surface conditions from prior machining conditions will act as stress risers which are areas of dimensional changes (Figure 8). Examples of such stress risers include: fillets (Figure 9) , thread and gear roots, machining marks (Figure 10 and 11), and rolling seams (Figures 11, 12 and 13) and forging laps (Figures 14. and 15) Forging laps are due to concentrations of oxides which are folded in during the forging process. The presence of these oxides prevents the hot steel from welding to itself during the forging process. This will lead to cracking as shown (Note: the sample is first viewed in the unetched condition to locate the crack. Then the sample is etched, if desired, for microstructural identification.)

Figure 9 - Micrograph of AISI 4142 steel as quenched and tempered; microstructure is tempered martensite with quench crack at fillet radius. (Magnigication 100X; Etchant 3% Nital.)

Figure 10 - Micrograph of AISI 4140 steel as quenched and tempered; microstructure is tempered martensite with quench crack initiating from machine groove.(Magnification: 100X; Etchant: 2% Nital.)

Figure 12 - Micrograph of AISI 8630 steel as quenched; microstructure is martensite where cracking initiated from rolling seam.

Figure 11 - Micrograph of AISI 4118 carburized steel as quenched and tempered; microstructure is tempered martensite (unetched) with quench crack propagating from machine burr (Magnification 200X.)

Figure 13 - Micrograph of AISI Type 403 stainless steel as quenched and tempered; microstructure is tempered; microstructure is predominantly tempered martensite with cracking promoted by the seam. (Magnification 100X; Etchant Vilelias.)

Elemental analysis shored that the carbon content of the steel was higher than the specification value (3.56% C versus 3.10-3.45 % C). The surface hardness was 52HRC versus a specification value of 45HRC. The hardness was 94HRB. Microstructural evidence suggested quenching from an excessively high austenitizing temperature which contributed to the formation of excessive amounts of retained austenite at the Core lobe surface and within the induction hardened case, as shown in Figure 16.

Figure 14 - Micrograph of AISI 1030 steel as direct forge-quenched and tempered; microstructure is tempered martensite (unetched) with forged in scale adjacet to cracking. (Magnification 100X.) H. Grinding Cracks An AISI G-3500 gray iron camshaft was induction hardened to a depth of 0.0600.150 inches and a lobe hardness of HRC=45. Cracking was observed by magnetic particle inspection of the cam lobe after subsequent machining.

Figure 16 Micrograph of induction hardened AISI G-3500 gray iron illustrating retained austenite at the lobe tip. (Magnification 200X, 3% Nital) Grinding stresses, apparently caused during machining of the cam lobe surface apparently caused localized martensitic transformation of the retained austenite. Cracking occurred due to the martensitic conversion, as shown in Figure 17.

Figure 15 Micrograph of AISI 1045 as-forged steel illustrating a forging lap. (Magnification 30x, 2% Nital)

Figure 17. Micrograph of induction hardened AISI G-3500 gray iron illustrating crack propagation into the induction hardened case. (Magnification 100X, 3% Nital) I. Alloy Inclusion Defects

Another common source of steel cracking problems is alloy inclusions due to poor homogenization of the steel composition. The most common alloy inclusions are sulfides, silicates, oxides and scale. Examples are provided in Figures 18-22. In addition to microstructural analysis, elemental analysis may also provide an invaluable insight into potential for inclusion formation. For example, sulfide inclusions may be obtained if insufficient manganese is present. Generally, the manganese content should be approximately five times the sulfur content to convert sulfur to manganese sulfide.

Figure 19 - Micrograph of AISI 4140 steel as quenched and tempered; microstructure is tempered martensite (unetched) with quench cracking promoted by non-metallic inclusions. (Magnification 100X.)

Figure 18 - Micrograph of AISI 4140 steel as quenched and tempered; micristructure is tempered martensite (unetched) with cracking at inclusions. (Magification 100X).

Figure 20 - Micrograph of AISI 1144 steel as quenched and tempered; microstructure is tempered martensite where cracking is promoted by inclusion defects. (Magnification 200X; Etchant 2% Nital.)

bainite appeared to originate from carbon and manganese segragation. Within the material. Which would lead to increased internal stresses. Both surface and subsurface cracking was observed. Figure 24 shows subsurface cracking which appreared to be intergranular.

Figure 21 - Micrograph of AISI 1144 steel as quenched and tempered; microstructure is martensite (unetched) where cracking is promoted by inclusion defects. (Magnification: 100X) Figure 23 - Micrograph of AISI 4140 steel as quenched and tempered; microstructure shows bands of tempered martensite and tempered martensite/bainite. (Magnification: 50 X, Etchant 2% Nital.)

Figure 22 - Micrograph of AISI 4150 steel as quenched and tempered; cracking initiates from silicate and sulfide inclusions. (Magnification: 100X; Etchant 2% Nital.) J. Chemical Segregation - Banding As indicated above, steel cracking may result from material chemical (segration which was evident from banding as shown in Figure 23. Microstructural analysis showed that bands of tempered martensite were associated with bands of tempered martensite and bainite. The

Figure 24 - Micrograph of AISI 4140 steel as quenched and tempered, microstructure shows bands with banded tempered martensite and tempered martensite/bainite microstructure illustrated in Figure 20. Here a representative view of subsurface cracking that was obtained is illustrated. (Magnification: 100 X, Etchant 2% Nital.)

In a similar case, AISI 1144, a resulfurized steel, pins were with subsequent cracking at the pin tips accompanied with soft spotting. (The pins were through-hardened prior to induction hardening of the pin tip.) Microstructural analysis showed that the cracking and soft spotting condition was due to stringer inclusions with bands of gross chemical segregation, significantly greater than normally observed with this grade of steel. The stringers act as stress concentration sites for crack initiation in the presence of quenching stresses. (See Figure 25) Figure 26 - Micrograph of AISI 4140 steel as quenched and tempered; microstructure is tempered martensite where cracking is promoted by alloy depletion. (Magnification: 100X; Etchant: 2% Nital.) L. Porosity Steel porosity, variation which is due to voids steel castings due to trapped gases, is another source of potential steel cracking. This is illustrated in Figure 27.

Figure 25 - Micrograph of AISI 1144 steel as quenched; microstructure shows shows ferrite bands and inclusions exiting the pin tip. (Magnification: 100 X; Etchant 3% Nital.) K.. Alloy Depletion Another source of non-homogeneous steel composition is alloy depletion. As with non-metallic inclusions, alloy depletion leads to greater stresses and cracking as illustrated in Figure 26. Examples of alloys particularly susceptible to alloy depletion are AISI 4100, 4300 and 8600 series. Figure 27 - Micrograph of AISI 8630 cast steel as quenched and tempered; microstructure is tempered martensite, pearlite and ferrite showing a potential cracking condition. (Magnigication 100X; Etchant: 2% Nital) M. Improper Steel Chemistry Steel hardenability is determined by its chemistry. The quench conditions required to obtain the desired properties are a funtion of the hardenability. Therefore, if the steel chemistry is incorrect, the selected quench process conditions may, if too severe, lead to cracking. Unfortunately, this problem is not uncommon.

An example illustrating this problem was afforded by a quench cracking problem obtained with AISI 1070 steel bearing raceways. Metalographic analysis confirmed that quench cracking had occurred. However, the steel chemistry, see Table 2, was incorrect for a plain AISI 1070 steel. The higher carbon, higher hardenability stee with a high manganese content, would be more susceptible to quench cracking using the same quenchant that the 1070 heat treating process demands.

is recommended that metallurgical analyses be conducted to determine the root cause of steel cracking during heat treat processing.

REFERENCES 1. Practical Data for Metallurgists, Book available from The Timken Company, Canton, OH. R.M. Brick and A. Phillips, Structure and Properties of Alloys, 2nd Ed., 1949, McGraw-Hill Inc., New York, NY,. p. 52.

Table 2 2. Comparison of Steel Obtained and Specification Range of Steel Chemistry of AISI 1070 Steel Used for Bearing Raceways Element Carbon Manganese Phosphorous Sulfur Silicon Nickel Chromium Molybdenum Copper Specification Range for AISI 1070 (%) 0.65-0.75 0.60-0.90 0.11 0.026 0.10-0.20 Obtained (%) 0.74 0.97 0.04 0.05 0.23 0.07 0.11 0.22 0.10

CONCLUSIONS Although quench cracking of steel may arise from insufficiently low quench severity, there are numerous other potential contributors to this problem. They include: non-uniform quenching due to poor system design, racking procedures which inhibit uniform quenchant flow around the part during the quench or contaminated quenchants. However, other potential sources of cracking are due to mechanical or material flaws which include: non-metallic inclusions, laps or seams, stress risers from prior machining, alloy non-uniformity and porosity. These problems are not readily detectable without microstructural characterization. Therefore, it