Fig.8 Two spiral power springs from a textile machine. Spring at left is an acceptable part, whereas spring at right took an excessive set(the inner end of the spiral is 30 out of position) because of insufficient yield strength and a decarburized surface layer. The material in the satisfactory spring had a hardness of 45 HRC, while the material in the spring that failed had a hardness of 41.5 HRC. This represents a 10% disparity in tensile and yield strengths between the two springs. The spring that failed had a 0.08 mm(0.003 in )thick surface layer of partial decarburization that further weakened the region of the cross section where maximum stresses are developed in the spring under load. The decarburized layer, which had a lower yield strength than the bulk material, yielded excessively during presetting; therefore the spring did not attain its specified shape in the free state following this operation Another material deficiency that can lead to deformation failure is variability in response to heat treatment among parts in a given production lot. Certain alloys, particularly hardenable low-alloy steels and some precipitation-hardening alloys, can vary in their response to a specified heat treatment because of slight compositional variations from lot to lot or within a given lot. This can result in some parts having too low a strength for the application even though they were properly eat treated according to specification Remedies for variability in response to heat treatment usually involve changes in the heat treating process, ranging in complexity from tailoring the heat treating conditions for each lot or sublot to making a small adjustment in the heat treatment specification. Material composition and microstructural specifications can also be tightened to provide more uniform heat treatment response. Experiments on each lot or sublot are almost al ways needed to establish parameters when heat treating conditions are tailored An adjustment in heat treating conditions was successful in avoiding variation in properties among sublots of heat treated AISI type 631(17-7 PH) stainless steel Belleville washers. Two of these washers--one of which was from an acceptable sublot and the other from a deficient sublot-were subjected to examination. The washer from the acceptable sublot had developed the required hardness upon solution heat treating at 955C(1750F), followed by refrigeration at-75C (-100 F)and aging. The other washer was soft after an identical heat treatment and yielded under load (flattened). The microstructure of the acceptable washer was a mixture of austenite and martensite, the structure of the washer that flattened consisted almost entirely of austenite Previous experience with 17-7 PH stainless steel indicated that some alloy segregation was not unusual and that relatively minor variations in composition could affect response to heat treatment, perhaps by depressing the range of martensite transformation temperatures to a variable degree. As noted in Ref 5, the solution-treating temperature has a marked effect on the martensite start(Ms)temperature in the precipitation-hardening stainless steels that are austenitic as solution annealed and martensitic as aged(17-7 PH, AM-350, AM-355 and PH15-7Mo) Consequently, although it never was clearly established whether temperature variations inside the solution-treating furnace or minor variations in composition were responsible for the observed variability in properties of the 17-7 PH Belleville washers, all sublots attained the required strength when the solution-treating temperature was lowered to 870C(1600F) Faulty Heat Treatment. Mistakes made in heat treating hardenable alloys are among the most common causes of premature failure. Temperatures that are either too high or too low can result in the development of inadequate or undesirable mechanical properties. Quenching a steel part too fast can crack it; quenching too slowly can fail to produce the required strength or toughness. If parts are shielded from a heating or cooling medium, they can respond poorly to heat treatment. as discussed in this section Two hold-down clamps, both from the same lot, are shown in Fig 9. Both clamps were bowed to the same degree after fabrication, as intended, but the clamp at the bottom flattened when it was installed. A small percentage of the clamps, all of which were made from hardened-and-tempered 1070 steel, deformed when a bolt was inserted through the hole and tightened. The clamp at top in Fig. 9 was acceptable, with a microstructure of tempered martensite and a hardness of 46 HRC, the clamp at bottom, which deformed, had a mixed structure of ferrite, coarse pearlite, and tempered martensite and a hardness of only 28 HRO Fig.9 Two hardened-and-tempered 1070 steel hold-down clamps. The clamp at top was acceptable. The clamp at bottom was slack quenched because of faulty loading practice Thefileisdownloadedfromwww.bzfxw.com
Fig. 8 Two spiral power springs from a textile machine. Spring at left is an acceptable part, whereas spring at right took an excessive set (the inner end of the spiral is 30° out of position) because of insufficient yield strength and a decarburized surface layer. The material in the satisfactory spring had a hardness of 45 HRC, while the material in the spring that failed had a hardness of 41.5 HRC. This represents a 10% disparity in tensile and yield strengths between the two springs. The spring that failed had a 0.08 mm (0.003 in.) thick surface layer of partial decarburization that further weakened the region of the cross section where maximum stresses are developed in the spring under load. The decarburized layer, which had a lower yield strength than the bulk material, yielded excessively during presetting; therefore the spring did not attain its specified shape in the free state following this operation. Another material deficiency that can lead to deformation failure is variability in response to heat treatment among parts in a given production lot. Certain alloys, particularly hardenable low-alloy steels and some precipitation-hardening alloys, can vary in their response to a specified heat treatment because of slight compositional variations from lot to lot or within a given lot. This can result in some parts having too low a strength for the application even though they were properly heat treated according to specification. Remedies for variability in response to heat treatment usually involve changes in the heat treating process, ranging in complexity from tailoring the heat treating conditions for each lot or sublot to making a small adjustment in the heat treatment specification. Material composition and microstructural specifications can also be tightened to provide more uniform heat treatment response. Experiments on each lot or sublot are almost always needed to establish parameters when heat treating conditions are tailored. An adjustment in heat treating conditions was successful in avoiding variation in properties among sublots of heat treated AISI type 631 (17-7 PH) stainless steel Belleville washers. Two of these washers—one of which was from an acceptable sublot and the other from a deficient sublot—were subjected to examination. The washer from the acceptable sublot had developed the required hardness upon solution heat treating at 955 °C (1750 °F), followed by refrigeration at -75 °C (-100 °F) and aging. The other washer was soft after an identical heat treatment and yielded under load (flattened). The microstructure of the acceptable washer was a mixture of austenite and martensite; the structure of the washer that flattened consisted almost entirely of austenite. Previous experience with 17-7 PH stainless steel indicated that some alloy segregation was not unusual and that relatively minor variations in composition could affect response to heat treatment, perhaps by depressing the range of martensitetransformation temperatures to a variable degree. As noted in Ref 5, the solution-treating temperature has a marked effect on the martensite start (Ms) temperature in the precipitation-hardening stainless steels that are austenitic as solution annealed and martensitic as aged (17-7 PH, AM-350, AM-355 and PH15-7Mo). Consequently, although it never was clearly established whether temperature variations inside the solution-treating furnace or minor variations in composition were responsible for the observed variability in properties of the 17-7 PH Belleville washers, all sublots attained the required strength when the solution-treating temperature was lowered to 870 °C (1600 °F). Faulty Heat Treatment. Mistakes made in heat treating hardenable alloys are among the most common causes of premature failure. Temperatures that are either too high or too low can result in the development of inadequate or undesirable mechanical properties. Quenching a steel part too fast can crack it; quenching too slowly can fail to produce the required strength or toughness. If parts are shielded from a heating or cooling medium, they can respond poorly to heat treatment, as discussed in this section. Two hold-down clamps, both from the same lot, are shown in Fig. 9. Both clamps were bowed to the same degree after fabrication, as intended, but the clamp at the bottom flattened when it was installed. A small percentage of the clamps, all of which were made from hardened-and-tempered 1070 steel, deformed when a bolt was inserted through the hole and tightened. The clamp at top in Fig. 9 was acceptable, with a microstructure of tempered martensite and a hardness of 46 HRC; the clamp at bottom, which deformed, had a mixed structure of ferrite, coarse pearlite, and tempered martensite and a hardness of only 28 HRC. Fig. 9 Two hardened-and-tempered 1070 steel hold-down clamps. The clamp at top was acceptable. The clamp at bottom was slack quenched because of faulty loading practice The file is downloaded from www.bzfxw.com
(stacking), and it failed by distortion (flattening) because of the resultant mixed microstructure Analysis of the heat treating process revealed that the parts were stacked so that occasional groupings were slack quenched as a result of shielding; this promoted the formation of softer high-temperature transformation products. When the loading practice was changed to ensure more uniform quenching(so that transformation to 100% martensite was accomplished on all parts), the problem was solved xample 4: Bending of an Aircraft-Wing Slat Track. A curved member called a slat track(Fig. 10), which supported the extendable portion of the leading edge of the wing on a military aircraft, failed by bending at one end after very short service. It was estimated that the slat track, fabricated from heat treated 4 140 steel, had undergone only one high-load Region of ailure Fig. 10 4140 steel slat track from a military aircraft wing. The track bent because one end did not become fully austenitic during heat treatment, producing a low-strength structure of ferrite and tempered martensite Investigation. Hardness measurements were taken at various points along the length of the track. The end that bent(at right, Fig. 10)had a hardness of 30 HRC, compared with a hardness of 41 hrC for the remainder of the part Metallographic examination of specimens from both ends of the slat track revealed that the microstructure of the end that bent contained a large number of ferrite islands in a matrix of tempered martensite. The microstructure of the opposite end contained no ferrite Conclusions. Bending had occurred in a portion of the slat track because service stresses had exceeded the strength of the material in a region of mixed martensite and ferrite. It was determined that the most likely cause of the mixed structure was nonuniform austenitization during heat treatment. The end that bent never became fully austenitic, because the furnace temperature was locally too low or because the soaking time was too short or both. It was decided that material or design changes were not warranted, considering the nature of the failure and the probable cause of the mixed microstructure in one end of the slat track Corrective Measures. Steps were taken to improve control of temperature of parts during austenitization Warpage and Residual Stress Effects. Warpage during heat treatment or during stress-relief anneal ing is also a common type of distortion failure. Warpage can result from changes in residual stress or from thermal or phase transformation induced stresses and strains that are introduced during heating or cooling. when relief of residual stress causes distortion the amount of distortion is proportional to the decrease of the residual stress. When distortion is caused by thermal or transformational stress, the extent of the distortion is greater for parts that have complex configuration or large differences in section thickness and for faster heating or cooling rates. Warpage can also be caused by inadequate support in the heat treating furnace, leading to sagging due to the weight of the part Warpage can be the result of plastic deformation that occurred in some region of the part at elevated temperature or during a change in temperature. Dimensional changes accompanying stress relief are the result of relief of those stresses involving both elastic and plastic strains. Distortion that occurs during other types of heat treatment involves mainly plastic strain and generally results in high levels of residual stress in the warped part. The magnitude and distribution of residual stresses are determined by the composition, shape, size, and heat treating conditions of a given part Warping is most often severe in heat treatments that involve quenching In hardenable steels, the principal cause of warping on quenching is nonuniform rates of transformation. The effect of transformational stresses may be intensified if a nonuniform composition exists; this nonuniformity may be the result of segregation, or it may be the result of processing, as in a carburized part. Such inhomogeneities may produce a variation in transformation temperature at locations that are geometrically equivalent and that cool at the same rate Nonuniform transformational stresses that result from inhomogeneity can also occur during tempering Warpage can often be minimized by modifying the heat treating conditions. For example, slow heating and cooling rates are less likely to cause warping because local variances in temperature and in rates of temperature change are minimized Preheating before austenitizing is often used as a means of minimizing warping of some tool steels and heavy sections because preheating reduces the temperature gradient between the surface and the interior of the part. Induction hardening and nitriding have been used to minimize warping when surface hardness is of primary importance to the performance of
(stacking), and it failed by distortion (flattening) because of the resultant mixed microstructure. Analysis of the heat treating process revealed that the parts were stacked so that occasional groupings were slack quenched as a result of shielding; this promoted the formation of softer high-temperature transformation products. When the loading practice was changed to ensure more uniform quenching (so that transformation to 100% martensite was accomplished on all parts), the problem was solved. Example 4: Bending of an Aircraft-Wing Slat Track. A curved member called a slat track (Fig. 10), which supported the extendable portion of the leading edge of the wing on a military aircraft, failed by bending at one end after very short service. It was estimated that the slat track, fabricated from heat treated 4140 steel, had undergone only one high-load cycle. Fig. 10 4140 steel slat track from a military aircraft wing. The track bent because one end did not become fully austenitic during heat treatment, producing a low-strength structure of ferrite and tempered martensite. Investigation. Hardness measurements were taken at various points along the length of the track. The end that bent (at right, Fig. 10) had a hardness of 30 HRC, compared with a hardness of 41 HRC for the remainder of the part. Metallographic examination of specimens from both ends of the slat track revealed that the microstructure of the end that bent contained a large number of ferrite islands in a matrix of tempered martensite. The microstructure of the opposite end contained no ferrite. Conclusions. Bending had occurred in a portion of the slat track because service stresses had exceeded the strength of the material in a region of mixed martensite and ferrite. It was determined that the most likely cause of the mixed structure was nonuniform austenitization during heat treatment. The end that bent never became fully austenitic, because the furnace temperature was locally too low or because the soaking time was too short or both. It was decided that material or design changes were not warranted, considering the nature of the failure and the probable cause of the mixed microstructure in one end of the slat track. Corrective Measures. Steps were taken to improve control of temperature of parts during austenitization. Warpage and Residual Stress Effects. Warpage during heat treatment or during stress-relief annealing is also a common type of distortion failure. Warpage can result from changes in residual stress or from thermal or phase transformation induced stresses and strains that are introduced during heating or cooling. When relief of residual stress causes distortion, the amount of distortion is proportional to the decrease of the residual stress. When distortion is caused by thermal or transformational stress, the extent of the distortion is greater for parts that have complex configuration or large differences in section thickness and for faster heating or cooling rates. Warpage can also be caused by inadequate support in the heat treating furnace, leading to sagging due to the weight of the part. Warpage can be the result of plastic deformation that occurred in some region of the part at elevated temperature or during a change in temperature. Dimensional changes accompanying stress relief are the result of relief of those stresses involving both elastic and plastic strains. Distortion that occurs during other types of heat treatment involves mainly plastic strain and generally results in high levels of residual stress in the warped part. The magnitude and distribution of residual stresses are determined by the composition, shape, size, and heat treating conditions of a given part. Warping is most often severe in heat treatments that involve quenching. In hardenable steels, the principal cause of warping on quenching is nonuniform rates of transformation. The effect of transformational stresses may be intensified if a nonuniform composition exists; this nonuniformity may be the result of segregation, or it may be the result of processing, as in a carburized part. Such inhomogeneities may produce a variation in transformation temperature at locations that are geometrically equivalent and that cool at the same rate. Nonuniform transformational stresses that result from inhomogeneity can also occur during tempering. Warpage can often be minimized by modifying the heat treating conditions. For example, slow heating and cooling rates are less likely to cause warping because local variances in temperature and in rates of temperature change are minimized. Preheating before austenitizing is often used as a means of minimizing warping of some tool steels and heavy sections because preheating reduces the temperature gradient between the surface and the interior of the part. Induction hardening and nitriding have been used to minimize warping when surface hardness is of primary importance to the performance of
a part Warping is of particular concern with respect to tool steels, and preventative measures have been published(e.g ee " Control of Distortion in Tool Steels"in Heat Treating, Volume 4 of ASM Handbook) In heat treatments requiring rapid cooling or quenching, excessive warping can usually be reduced by changing the quenching conditions. In many instances, the orientation of a part as it enters the quenchant will influence the amount of distortion that occurs. Quenching in special fixtures(quench presses) is widely used in certain industries to minimize distortion by providing different, controlled cooling rates at different locations in a given part. Martempering has also been used to minimize distortion, because transformation rates upon cooling to the Ms temperature are equalized While heat treatment operations are a common cause of distortion and residual stress, it should be noted that other manufacturing operations also cause deformation directly or through introduction of residual stresses. Residual stresse are created by strain gradients in a component. The strain gradients may be created mechanically or thermally. Very fev manufacturing operations create uniform strain gradients, and uniform strain from mechanical working almost never occurs. Parts that are cold worked, including both cold-formed parts and stampings, will have a residual stress pattern characteristic of the process. Machining and other finishing processes introduce surface residual stresses that are often tensile and may be significant, especially in parts with thin sections. Grinding processes can be controlled to minimize residual stresses, but less-controlled grinding processes can produce tensile surface stresses and even cracks in severe cases Castings, at least until they are stress relieved, may have residual stresses arising from the sequence in which the solidification occurred. The liquid material contracts as it solidifies. Regions that are the last to solidify are constrained by the surrounding material. Similarly, welding will produce residual stresses and possible distortion. Electrodischarge machining also involves localized melting and resolidification, and associated residual stresses, although on a smaller The residual stresses existing in a component will always be balanced so that the sum of the forces is 0. Thus,a compressive residual stress cannot exist without tensile residual stresses somewhere else in the component. If a machining process removes material that is under a residual stress, the stress distribution in the remaining material will change and the component may warp unacceptably during the machining process Shot peening is another common process in the manufacture of some types of parts and is intended to introduce a surface compressive residual stress to improve fatigue resistance. However, shot peening may also cause warping, again more likely in parts with thin sections. Furthermore, if the peening is excessive, the high surface compressive stresses will be balanced by high subsurface tensile stresses, which, in turn, can cause further problems. Carburizing usually results in compressive residual stress at the surface Regardless of the origin of the residual stresses, they will add to the service-induced stresses and, if a critical point exceeded, local yielding and distortion will be observed at a nominal stress at which yielding would not be anticipated Residual stress effects can be difficult to verify in a failure analysis because the residual stresses existing in the deformed or fractured part are different than they were prior to the failure and the presence or absence of residual stress is not obvious in any case. Methods for measurement and study of residual stress include hole drilling, x-ray diffraction, and neutron diffraction techniques (e.g, see the articles"Residual Stress Measurements"in Mechanical Testing and Evaluation, Volume 8 of ASM Handbook, and"X-Ray Diffraction Residual Stress Measurements in Failure Analysis this Volume ) More detailed discussion of residual stress is also available in Ref 6 For the purposes of this article, without going into the details on the many causes and effects of residual stress, the main oint is that residual stress is closely tied to distortion. The failure analyst must be aware of the potential contribution of residual stress to failures involving both distortion and fracture. Various types of residual stresses can be introduced in complex ways by manufacturing operations. For example, it is commonly understood that quenching operations create surface longitudinal and hoop stresses. This is true for lower-hardenability materials, but not for high-hardenability steels (Ref 7). Another example of the complexity of the effect of manufacturing on residual stresses is shot peening. If peening intensity is too high, large tensile stresses can occur below the surface, which can cause problems in bending and torsion Faulty Case Hardening. Carburizing, which both increases the surface hardness of a part and provides resistance to wear and indentation, can, if improperly controlled, produce a case that has too low or too high a carbon content. With too low a carbon content, the surface may not be hard enough to withstand normal service loads. This condition may be ccompanied by shallow case depth, which aggravates the problem. With an excessively high carbon content, which generally is the result of a high carbon potential or improper diffusion during the carburizing cycle, large amounts of retained austenite may be present in the carburized zone after heat treatment, depending on the composition of the steel Retained austenite softens the surface and, under certain conditions, may transform to martensite in service. When transformation in service occurs, the resulting untempered martensite may crack and thus promote early failure by surface fatigue, or a distortion failure in a close-fitting assembly may occur because of the volume change that accompanies transformation In addition to the carbon level and surface hardness, a carburized case is also characterized by its thickness or depth below the surface. Improper case depth can lead to some of the same types of problems associated previously with improper carbon content. The case depth is also important when maximum stresses occur at the surface of a part, as they Thefileisdownloadedfromwww.bzfxw.com
a part. Warping is of particular concern with respect to tool steels, and preventative measures have been published (e.g., see “Control of Distortion in Tool Steels” in Heat Treating, Volume 4 of ASM Handbook). In heat treatments requiring rapid cooling or quenching, excessive warping can usually be reduced by changing the quenching conditions. In many instances, the orientation of a part as it enters the quenchant will influence the amount of distortion that occurs. Quenching in special fixtures (quench presses) is widely used in certain industries to minimize distortion by providing different, controlled cooling rates at different locations in a given part. Martempering has also been used to minimize distortion, because transformation rates upon cooling to the Ms temperature are equalized throughout the part. While heat treatment operations are a common cause of distortion and residual stress, it should be noted that other manufacturing operations also cause deformation directly or through introduction of residual stresses. Residual stresses are created by strain gradients in a component. The strain gradients may be created mechanically or thermally. Very few manufacturing operations create uniform strain gradients, and uniform strain from mechanical working almost never occurs. Parts that are cold worked, including both cold-formed parts and stampings, will have a residual stress pattern characteristic of the process. Machining and other finishing processes introduce surface residual stresses that are often tensile and may be significant, especially in parts with thin sections. Grinding processes can be controlled to minimize residual stresses, but less-controlled grinding processes can produce tensile surface stresses and even cracks in severe cases. Castings, at least until they are stress relieved, may have residual stresses arising from the sequence in which the solidification occurred. The liquid material contracts as it solidifies. Regions that are the last to solidify are constrained by the surrounding material. Similarly, welding will produce residual stresses and possible distortion. Electrodischarge machining also involves localized melting and resolidification, and associated residual stresses, although on a smaller scale. The residual stresses existing in a component will always be balanced so that the sum of the forces is 0. Thus, a compressive residual stress cannot exist without tensile residual stresses somewhere else in the component. If a machining process removes material that is under a residual stress, the stress distribution in the remaining material will change and the component may warp unacceptably during the machining process. Shot peening is another common process in the manufacture of some types of parts and is intended to introduce a surface compressive residual stress to improve fatigue resistance. However, shot peening may also cause warping, again more likely in parts with thin sections. Furthermore, if the peening is excessive, the high surface compressive stresses will be balanced by high subsurface tensile stresses, which, in turn, can cause further problems. Carburizing usually results in compressive residual stress at the surface. Regardless of the origin of the residual stresses, they will add to the service-induced stresses and, if a critical point is exceeded, local yielding and distortion will be observed at a nominal stress at which yielding would not be anticipated. Residual stress effects can be difficult to verify in a failure analysis because the residual stresses existing in the deformed or fractured part are different than they were prior to the failure and the presence or absence of residual stress is not obvious in any case. Methods for measurement and study of residual stress include hole drilling, x-ray diffraction, and neutron diffraction techniques (e.g., see the articles “Residual Stress Measurements” in Mechanical Testing and Evaluation, Volume 8 of ASM Handbook, and “X-Ray Diffraction Residual Stress Measurements in Failure Analysis” in this Volume). More detailed discussion of residual stress is also available in Ref 6. For the purposes of this article, without going into the details on the many causes and effects of residual stress, the main point is that residual stress is closely tied to distortion. The failure analyst must be aware of the potential contribution of residual stress to failures involving both distortion and fracture. Various types of residual stresses can be introduced in complex ways by manufacturing operations. For example, it is commonly understood that quenching operations create surface longitudinal and hoop stresses. This is true for lower-hardenability materials, but not for high-hardenability steels (Ref 7). Another example of the complexity of the effect of manufacturing on residual stresses is shot peening. If peening intensity is too high, large tensile stresses can occur below the surface, which can cause problems in bending and torsion. Faulty Case Hardening. Carburizing, which both increases the surface hardness of a part and provides resistance to wear and indentation, can, if improperly controlled, produce a case that has too low or too high a carbon content. With too low a carbon content, the surface may not be hard enough to withstand normal service loads. This condition may be accompanied by shallow case depth, which aggravates the problem. With an excessively high carbon content, which generally is the result of a high carbon potential or improper diffusion during the carburizing cycle, large amounts of retained austenite may be present in the carburized zone after heat treatment, depending on the composition of the steel. Retained austenite softens the surface and, under certain conditions, may transform to martensite in service. When transformation in service occurs, the resulting untempered martensite may crack and thus promote early failure by surface fatigue, or a distortion failure in a close-fitting assembly may occur because of the volume change that accompanies transformation. In addition to the carbon level and surface hardness, a carburized case is also characterized by its thickness or depth below the surface. Improper case depth can lead to some of the same types of problems associated previously with improper carbon content. The case depth is also important when maximum stresses occur at the surface of a part, as they The file is downloaded from www.bzfxw.com
do in bending or torsion. To avoid failures due to distortion, as well as other failures, a designer must define the desired case characteristics and then work with the heat treater to ensure that the process is capable and controlled to produce that case. This communication must also include the necessary quality control. As should be clear from this discussion, surface hardness testing cannot completely characterize a case Example 5: Seizing of a Spool-Type Hydraulic Valve. Occasional failures were experienced in spool-type valves used in a hydraulic system. When a valve would fail, the close-fitting rotary valve would seize, causing loss of flow control of the hydraulic oil. The rotating spool in the valve was made of 8620 steel and was gas carburized. The cylinder in which the spool fitted was made of 1117 steel, also gas carburized Investigation. Low-magnification visual examination of the spool and cylinder from a failed valve revealed some burnishing, apparently the result of contact between the spool and the inside cylinder wall. Measurement of the surface profile of the spool showed no evidence of wear or galling. When this profile was compared with that of a spool from a valve that operated satisfactorily, no significant difference was found Metallographic sections were made of the cylinder from the failed valve and from the one that operated satisfactorily. The microstructure of the carburized case on the cylinder from the satisfactory valve was well-defined martensite interspersed th some austenite(light-etching )areas, Fig. 1la. On the other hand, the microstructure of the case from the failed valve contained a much greater amount of retained austenite, especially near the surface(Fig. 11b). In addition, some patches of untempered martensite were found close to the surface of the failed valve cylinder with use of a more aggressive etching procedure Fig. 1l Cross sections through the carburized 1117 steel cylinders from two spool-type hydraulic valves. The cylinder of the valve that operated satisfactorily (a) had little retained austenite in the case, whereas the cylinder of the seized valve(b) had much retained austenite that transformed to martensite in service resulting in size distortion ( growth).500× Microhardness traverses of the two sections revealed that, in general, the hardness of the case from the failed part was about 100 Knoop points lower than the case hardness of the unfailed part. At the surface, the case hardness of the failed part was 300 Knoop points lower Conclusions Momentary sliding contact between the spool and the cylinder wall (probably during valve opening)cau unstable retained austenite in the failed cylinder to transform to martensite. The increase in volume resulted in suffic
do in bending or torsion. To avoid failures due to distortion, as well as other failures, a designer must define the desired case characteristics and then work with the heat treater to ensure that the process is capable and controlled to produce that case. This communication must also include the necessary quality control. As should be clear from this discussion, surface hardness testing cannot completely characterize a case. Example 5: Seizing of a Spool-Type Hydraulic Valve. Occasional failures were experienced in spool-type valves used in a hydraulic system. When a valve would fail, the close-fitting rotary valve would seize, causing loss of flow control of the hydraulic oil. The rotating spool in the valve was made of 8620 steel and was gas carburized. The cylinder in which the spool fitted was made of 1117 steel, also gas carburized. Investigation. Low-magnification visual examination of the spool and cylinder from a failed valve revealed some burnishing, apparently the result of contact between the spool and the inside cylinder wall. Measurement of the surface profile of the spool showed no evidence of wear or galling. When this profile was compared with that of a spool from a valve that operated satisfactorily, no significant difference was found. Metallographic sections were made of the cylinder from the failed valve and from the one that operated satisfactorily. The microstructure of the carburized case on the cylinder from the satisfactory valve was well-defined martensite interspersed with some austenite (light-etching) areas, Fig. 11a. On the other hand, the microstructure of the case from the failed valve contained a much greater amount of retained austenite, especially near the surface (Fig. 11b). In addition, some patches of untempered martensite were found close to the surface of the failed valve cylinder with use of a more aggressive etching procedure. Fig. 11 Cross sections through the carburized 1117 steel cylinders from two spool-type hydraulic valves. The cylinder of the valve that operated satisfactorily (a) had little retained austenite in the case, whereas the cylinder of the seized valve (b) had much retained austenite that transformed to martensite in service, resulting in size distortion (growth). 500× Microhardness traverses of the two sections revealed that, in general, the hardness of the case from the failed part was about 100 Knoop points lower than the case hardness of the unfailed part. At the surface, the case hardness of the failed part was 300 Knoop points lower. Conclusions. Momentary sliding contact between the spool and the cylinder wall (probably during valve opening) caused unstable retained austenite in the failed cylinder to transform to martensite. The increase in volume resulted in sufficient
size distortion(growth) to cause interference between the cylinder and the spool, seizing, and loss of flow control. The failed parts had been carburized in a process in which the carbon potential was too high, which resulted in a microstructure having excessive retained austenite after heat treatment Corrective Measure. The composition of the carburizing atmosphere was modified to yield carburized parts that did not retain significant amounts of austenite when they were heat treated. After this change was made, occasional valve failure by seizing ceased Faulty Repairs. Products are often repaired to correct deficiencies that are found in new parts during quality-control inspections or in used parts after they have deteriorated in service. Repair welding and brazing are generally recognized as potential sources of unwanted alterations of the properties of heat treatable alloys. Parts can be made softer or more brittle by careless repair, depending on the alloy and the conditions under which the repair was made Substitution of a part, particularly a fastener, whose properties do not match the properties of the part it replaces can lead to failure of the substitute part, failure of another part, or both, as in the following instance A carrying handle of complex configuration was secured to a heavy portable device by means of a ring clamp at one end of the handle and a special -20 hardened bolt through a small flange at the other end. For some unknown reason, the special slotted hexagonal-head bolt was replaced by a standard commercial hexagonal-head cap screw whose hardness was 93 HRB instead of the specified hardness of 28 HRC The commercial cap screw was distorted in service(Fig. 12a), causing the handle to become loose at one end. This in turn caused eccentric and excessive loading on the small flange of the handle with the result that the flange bent and then broke. Figure 12(b) shows the broken flange; the bright area next to the hole is where the loose flange chafed against the distorted bolt and is an indication of the eccentric load distribution on the flange. a fastener of the correct hardness and length is shown in Fig. 12(c)for comparison with the distorted commercial cap screw Fig. 12 Part substitution that resulted in a distortion failure .(a) distorted commercial cap screw that was used as a replacement for a hardened bolt.(b) Carry-handle flange that broke because the cap screw bent. (c)Correct replacement part References cited in this section 5. A.J. Lena, Precipitation Reactions in Iron-Base Alloys, in Precipitation from Solid Solution, American Society for Metals, 1959. p 224-327 6. G. Totten, M. Howes, and T. Inoue, Ed, Handbook of Residual Stress and Deformation of Steel, ASM International. 2002 Thefileisdownloadedfromwww.bzfxw.com
size distortion (growth) to cause interference between the cylinder and the spool, seizing, and loss of flow control. The failed parts had been carburized in a process in which the carbon potential was too high, which resulted in a microstructure having excessive retained austenite after heat treatment. Corrective Measure. The composition of the carburizing atmosphere was modified to yield carburized parts that did not retain significant amounts of austenite when they were heat treated. After this change was made, occasional valve failure by seizing ceased. Faulty Repairs. Products are often repaired to correct deficiencies that are found in new parts during quality-control inspections or in used parts after they have deteriorated in service. Repair welding and brazing are generally recognized as potential sources of unwanted alterations of the properties of heat treatable alloys. Parts can be made softer or more brittle by careless repair, depending on the alloy and the conditions under which the repair was made. Substitution of a part, particularly a fastener, whose properties do not match the properties of the part it replaces can lead to failure of the substitute part, failure of another part, or both, as in the following instance. A carrying handle of complex configuration was secured to a heavy portable device by means of a ring clamp at one end of the handle and a special 1 4 -20 hardened bolt through a small flange at the other end. For some unknown reason, the special slotted hexagonal-head bolt was replaced by a standard commercial hexagonal-head cap screw whose hardness was 93 HRB instead of the specified hardness of 28 HRC. The commercial cap screw was distorted in service (Fig. 12a), causing the handle to become loose at one end. This in turn caused eccentric and excessive loading on the small flange of the handle with the result that the flange bent and then broke. Figure 12(b) shows the broken flange; the bright area next to the hole is where the loose flange chafed against the distorted bolt and is an indication of the eccentric load distribution on the flange. A fastener of the correct hardness and length is shown in Fig. 12(c) for comparison with the distorted commercial cap screw. Fig. 12 Part substitution that resulted in a distortion failure. (a) Distorted commercial cap screw that was used as a replacement for a hardened bolt. (b) Carry-handle flange that broke because the cap screw bent. (c) Correct replacement part References cited in this section 5. A.J. Lena, Precipitation Reactions in Iron-Base Alloys, in Precipitation from Solid Solution, American Society for Metals, 1959. p 224–327 6. G. Totten, M. Howes, and T. Inoue, Ed., Handbook of Residual Stress and Deformation of Steel, ASM International, 2002 The file is downloaded from www.bzfxw.com