front usually swings around about 15 or more(Fig. 5a and c). A third difference arising from rotation is in the distribution of the initiation sites of a multiple-origin crack Moderate stress concentration c (approx) Severe stress concentration Fig. 5 Typical fatigue marks on the fracture surface of a uniformly loaded rotating shaft. Marks are produced from single and multiple origins(arrows) having moderate and severe stress concentration; shaded areas are final-fracture zones. Shaft rotation is clockWISe In a nonrotating shaft subjected to unidirectional bending, the origins are located in the region of the maximum-tension zone(Fig. 2). In a nonrotating shaft subjected to reversed bending, the origins are diametrically opposite each other(Fig 4). In rotary bending, however, every point on the shaft periphery is subjected to a tensile stress at each revolution therefore, a crack may be initiated at any point on the periphery(Fig 5b and d The crack surfaces are pressed together during the compressive component of the stress cycle, and mutual rubbing occurs A common result of final fracture is that slight movement of one side of the crack relative to the other side frequently causes severe damage to the fracture surfaces and tends to obliterate many marks. However, al though the high spots on one surface may rub the high spots on the other, the marks in the depressions are retained. Because the depressions are negative images of the damaged high spots on the opposing surface, they provide useful evidence; therefore, it is desirable to examine both parts of a cracked or fractured shaft The similarity in macroscopic appearance of fractures in shafts resulting from rotating-bending fatigue and from single overload torsional shear of a relatively ductile metal frequently results in misinterpretation. The fracture surface shown in Fig. 6(a) was the result of fatigue, as evidenced by the ratchet marks around the periphery and the pronounced beach marks. Under the low magnification of the fractograph shown in Fig. 6(b), beach marks are not visible, because they were obliterated by rubbing. The presence of ratchet marks around the periphery is also an indication of rotating-bending fatigue. However, the metal smearing apparent on the fracture surface and the twisting deformation of the shaft shown in Fig 6(c)indicate torsional shear and would preclude mistaking this fracture for a fatigue fracture Thefileisdownloadedfromwww.bzfxw.com
front usually swings around about 15° or more (Fig. 5a and c). A third difference arising from rotation is in the distribution of the initiation sites of a multiple-origin crack. Fig. 5 Typical fatigue marks on the fracture surface of a uniformly loaded rotating shaft. Marks are produced from single and multiple origins (arrows) having moderate and severe stress concentration; shaded areas are final-fracture zones. Shaft rotation is clockwise. In a nonrotating shaft subjected to unidirectional bending, the origins are located in the region of the maximum-tension zone (Fig. 2). In a nonrotating shaft subjected to reversed bending, the origins are diametrically opposite each other (Fig. 4). In rotary bending, however, every point on the shaft periphery is subjected to a tensile stress at each revolution; therefore, a crack may be initiated at any point on the periphery (Fig. 5b and d). The crack surfaces are pressed together during the compressive component of the stress cycle, and mutual rubbing occurs. A common result of final fracture is that slight movement of one side of the crack relative to the other side frequently causes severe damage to the fracture surfaces and tends to obliterate many marks. However, although the high spots on one surface may rub the high spots on the other, the marks in the depressions are retained. Because the depressions are negative images of the damaged high spots on the opposing surface, they provide useful evidence; therefore, it is desirable to examine both parts of a cracked or fractured shaft. The similarity in macroscopic appearance of fractures in shafts resulting from rotating-bending fatigue and from single overload torsional shear of a relatively ductile metal frequently results in misinterpretation. The fracture surface shown in Fig. 6(a) was the result of fatigue, as evidenced by the ratchet marks around the periphery and the pronounced beach marks. Under the low magnification of the fractograph shown in Fig. 6(b), beach marks are not visible, because they were obliterated by rubbing. The presence of ratchet marks around the periphery is also an indication of rotating-bending fatigue. However, the metal smearing apparent on the fracture surface and the twisting deformation of the shaft shown in Fig. 6(c) indicate torsional shear and would preclude mistaking this fracture for a fatigue fracture. The file is downloaded from www.bzfxw.com
(b) Fig 6 Fracture surfaces of failed shafts.(a)and(b) Failure by fatigue(c)and(d) Failure by torsional shear see text for discussion The fracture shown in Fig. 6(d)also exhibits a superficial similarity to a fatigue fracture. However, it is evident that this fracture was the result of torsional shear because the entire fracture surface has a smooth texture and no well-defined final-fracture area In splined shafts, fracture resulting from torsional shear is frequently accompanied by deformation of the splines not engaged by the mating part. However, the portion of the shaft not engaged by the mating part is sometimes unavailable for examination. When macroscopic examination affords only inconclusive evidence, use of an electron microscope may reveal fatigue striations or elongated shear dimples. Also, metallographic examination of a section through the fracture surface may reveal microdeformation from torsional shear in the rotary direction that would not be present in a fatigue fracture Torsional Fatigue. Fatigue cracks arising from torsional stresses also show beach marks and ridges. Longitudinal stress raisers are comparatively harmless under bending stresses, but are as important as circumferential stress raisers under torsional loading. This sensitivity of shafts loaded in torsion to longitudinal stress raisers is of considerable practical importance because inclusions in the shaft material are almost always parallel to the axis of rotation. It is not unusual for a
Fig. 6 Fracture surfaces of failed shafts. (a) and (b) Failure by fatigue. (c) and (d) Failure by torsional shear. See text for discussion. The fracture shown in Fig. 6(d) also exhibits a superficial similarity to a fatigue fracture. However, it is evident that this fracture was the result of torsional shear, because the entire fracture surface has a smooth texture and no well-defined final-fracture area. In splined shafts, fracture resulting from torsional shear is frequently accompanied by deformation of the splines not engaged by the mating part. However, the portion of the shaft not engaged by the mating part is sometimes unavailable for examination. When macroscopic examination affords only inconclusive evidence, use of an electron microscope may reveal fatigue striations or elongated shear dimples. Also, metallographic examination of a section through the fracture surface may reveal microdeformation from torsional shear in the rotary direction that would not be present in a fatigue fracture. Torsional Fatigue. Fatigue cracks arising from torsional stresses also show beach marks and ridges. Longitudinal stress raisers are comparatively harmless under bending stresses, but are as important as circumferential stress raisers under torsional loading. This sensitivity of shafts loaded in torsion to longitudinal stress raisers is of considerable practical importance because inclusions in the shaft material are almost always parallel to the axis of rotation. It is not unusual for a
torsional-fatigue crack to originate at a longitudinal inclusion, a surface mark, or a spline or key way corner and then to branch at about 45 When a stress raiser such as a circumferential groove is present, different states of stresses exist around the stress raiser and the tensile stress is increased to as much as four times the shear stress. Therefore, the tensile stress on the 45 plane will exceed the tensile strength of the steel before the shear stress reaches the shear strength of the steel. fracture occurs normal to the 45 tensile plane, producing a conical or star-shaped fracture surface Example 2: Torsional-Fatigue Fracture of a Large 4340 Steel Shaft That Was Subject to Cyclic Loading and Frequent Overloads. The 4340 steel shaft shown in Fig. 7(a), which was the driving member of a large rotor subject to cyclic loading and frequent overloads, broke after three weeks of operation. The shaft was also part of a gear train that reduced the rotational speed of the driven member. The driving shaft contained a shear groove at which the shaft should break if a sudden high overload occurred, thus preventing damage to an expensive gear mechanism. The rotor was subjected to severe chatter, which was an abnormal condition resulting from a series of continuous small overloads at a frequency of about three per second 43希0stel Rockwell C 3010: 30.5- 4.2am ta) Rotor shott Section Fig. 74340 steel rotor shaft that failed by torsional fatigue.(a) Shear groove designed to protect gear mechanism from sudden overload Dimensions are in inches. (b)Star-shaped pattern on a fracture surface of the shaft. (c) longitudinal and transverse shear cracks on the surface of the shear groove, which resulted from high peak loads caused by chatter Investigation. Examination disclosed that the shaft had broken at the shear groove and that the fracture surface contained a star-shaped pattern( Fig. 7b). Figure 7(c) shows the fracture surface with the pieces fitted back in place. The pieces were all nearly the same size and shape, and there were indications of fatigue, cleavage, and shear failure in approximately the same location on each piece. The cracks were oriented at approximately 45 to the axis of the shaft, which indicated that inal fracture was caused by a tensile stress normal to the 45 plane and not by the longitudinal or transverse shear stress that had been expected to cause an overload failure Examination of the surfaces of one of the pieces of the broken shaft revealed small longitudinal and transverse shear cracks at the smallest diameter of the shear groove. Also, slight plastic flow had occurred in the metal adjacent to these cracks. Cracking occurred at many points in the groove in the shaft before several of the cracks grew to a critical size No surface irregularities were present in the shear groove at any of the shear cracks. The structure of the metal was normal, with a uniform hardness of 30 to 30.5 HRC across the section, indicating a strength in the expected range for quenched-and-tempered 4340 steel shafts. A hot-acid etch showed the steel to be free from pipe, segregation, or other irregularities Conclusion. The basic failure mechanism was fracture by torsional fatigue, which started at numerous surface shear racks, both longitudinal and transverse, that developed in the periphery of the root of the shear groove. These shear cracks resulted from high peak loads caused by chatter. Stress concentrations developed in the regions of maximum shear, and fatigue cracks propagated in a direction perpendicular to the maximum tensile stress, thus forming the star pattern at 45 to the longitudinal axis of the shaft. The shear groove in the shaft, designed to prevent damage to the gear train, had performed its function, but at a lower overload level than intended Corrective Measures. The fatigue strength of the shaft was increased by shot peening the shear groove, and chatter in the machine was minimized The relative extent of development of two torsional-fatigue cracks mutually at right angles can indicate the magnitude of the torque reversals that have been applied If the cracks are of approximately the same length, the indications are that the torque reversals have been of equal magnitude, but only if the cracks are in a comparatively early stage of development. Thefileisdownloadedfromwww.bzfxw.com
torsional-fatigue crack to originate at a longitudinal inclusion, a surface mark, or a spline or keyway corner and then to branch at about 45°. When a stress raiser such as a circumferential groove is present, different states of stresses exist around the stress raiser, and the tensile stress is increased to as much as four times the shear stress. Therefore, the tensile stress on the 45° plane will exceed the tensile strength of the steel before the shear stress reaches the shear strength of the steel. Fracture occurs normal to the 45° tensile plane, producing a conical or star-shaped fracture surface. Example 2: Torsional-Fatigue Fracture of a Large 4340 Steel Shaft That Was Subject to Cyclic Loading and Frequent Overloads. The 4340 steel shaft shown in Fig. 7(a), which was the driving member of a large rotor subject to cyclic loading and frequent overloads, broke after three weeks of operation. The shaft was also part of a gear train that reduced the rotational speed of the driven member. The driving shaft contained a shear groove at which the shaft should break if a sudden high overload occurred, thus preventing damage to an expensive gear mechanism. The rotor was subjected to severe chatter, which was an abnormal condition resulting from a series of continuous small overloads at a frequency of about three per second. Fig. 7 4340 steel rotor shaft that failed by torsional fatigue. (a) Shear groove designed to protect gear mechanism from sudden overload. Dimensions are in inches. (b) Star-shaped pattern on a fracture surface of the shaft. (c) Longitudinal and transverse shear cracks on the surface of the shear groove, which resulted from high peak loads caused by chatter Investigation. Examination disclosed that the shaft had broken at the shear groove and that the fracture surface contained a star-shaped pattern (Fig. 7b). Figure 7(c) shows the fracture surface with the pieces fitted back in place. The pieces were all nearly the same size and shape, and there were indications of fatigue, cleavage, and shear failure in approximately the same location on each piece. The cracks were oriented at approximately 45° to the axis of the shaft, which indicated that final fracture was caused by a tensile stress normal to the 45° plane and not by the longitudinal or transverse shear stress that had been expected to cause an overload failure. Examination of the surfaces of one of the pieces of the broken shaft revealed small longitudinal and transverse shear cracks at the smallest diameter of the shear groove. Also, slight plastic flow had occurred in the metal adjacent to these cracks. Cracking occurred at many points in the groove in the shaft before several of the cracks grew to a critical size. No surface irregularities were present in the shear groove at any of the shear cracks. The structure of the metal was normal, with a uniform hardness of 30 to 30.5 HRC across the section, indicating a strength in the expected range for quenched-and-tempered 4340 steel shafts. A hot-acid etch showed the steel to be free from pipe, segregation, or other irregularities. Conclusion. The basic failure mechanism was fracture by torsional fatigue, which started at numerous surface shear cracks, both longitudinal and transverse, that developed in the periphery of the root of the shear groove. These shear cracks resulted from high peak loads caused by chatter. Stress concentrations developed in the regions of maximum shear, and fatigue cracks propagated in a direction perpendicular to the maximum tensile stress, thus forming the star pattern at 45° to the longitudinal axis of the shaft. The shear groove in the shaft, designed to prevent damage to the gear train, had performed its function, but at a lower overload level than intended. Corrective Measures. The fatigue strength of the shaft was increased by shot peening the shear groove, and chatter in the machine was minimized. The relative extent of development of two torsional-fatigue cracks mutually at right angles can indicate the magnitude of the torque reversals that have been applied. If the cracks are of approximately the same length, the indications are that the torque reversals have been of equal magnitude, but only if the cracks are in a comparatively early stage of development. The file is downloaded from www.bzfxw.com
Beyond this stage, one crack usually takes the lead, and such inferences are no longer justified. If the shaft transmits a unidirectional torque but two cracks develop mutually at right angles, it can be presumed that the torque was of a reversing character. If a bending stress is applied to a shaft that is transmitting torque, the angle at which any fatigue crack develops will be modified. Therefore, if the angle differs significantly from 45 to the shaft axis, the presence of a bending stress is indicated Failures of shafts Revised by Donald ]. Wulpi, Metallurgical Consultant Contact Fatigue Contact fatigue occurs when components roll, or roll and slide, against each other under high contact pressure and cyclic loading. Pitting occurs after many repetitions of loading and is the result of metal fatigue from the imposed cyclic contact stresses. Factors that govern contact fatigue are contact stress, relative rolling/sliding material properties, and metallurgical, physical, and chemical characteristics of the contacting surfaces including the oil film that lubricates the surfaces The significant stress in rolling-contact fatigue is the maximum alternating shear stress that undergoes a reversal in direction during rolling. In pure rolling, as in antifriction bearings, this stress occurs slightly below the surface and can lead to the initiation of subsurface fatigue cracks. As these cracks propagate under the repeated loads, they reach the surface and produce cavities, or pits When sliding is superimposed on rolling, as in gear teeth, the tangential forces and thermal gradient caused by friction alter the magnitude and distribution of stresses in and below the contact area. The alternating shear stress is increased in magnitude and is moved nearer to the surface by friction resulting from the sliding actior Additional information is provided in the articles"Fatigue Failures"and"Failures of Rolling-Element Bearings" in this volume Failures of shafts Revised by Donald ]. Wulpi, Metallurgical Consultant ear Wear of metal parts is commonly classified into either of two categories: abrasive wear or adhesive wear Abrasive wear, the undesired removal of material by a cutting mechanism, can reduce the size and destroy the proper shape of a shaft. The shaft may then fail by another means, such as by fracture, or may cease to perform its designed function. Foreign particles, such as sand, dirt and other debris, in the lubricant can cause wear of a shaft The a ble 3: Wear Failure of a Fuel-Pump Drive Shaft Caused by the Presence of Sand, Metallic Particles, and Vibration fuel pump in a turbine-powered aircraft failed, resulting in damage to the aircraft. The pump is shown in Fig. 8(a)and
Beyond this stage, one crack usually takes the lead, and such inferences are no longer justified. If the shaft transmits a unidirectional torque but two cracks develop mutually at right angles, it can be presumed that the torque was of a reversing character. If a bending stress is applied to a shaft that is transmitting torque, the angle at which any fatigue crack develops will be modified. Therefore, if the angle differs significantly from 45° to the shaft axis, the presence of a bending stress is indicated. Failures of Shafts Revised by Donald J. Wulpi, Metallurgical Consultant Contact Fatigue Contact fatigue occurs when components roll, or roll and slide, against each other under high contact pressure and cyclic loading. Pitting occurs after many repetitions of loading and is the result of metal fatigue from the imposed cyclic contact stresses. Factors that govern contact fatigue are contact stress, relative rolling/sliding, material properties, and metallurgical, physical, and chemical characteristics of the contacting surfaces, including the oil film that lubricates the surfaces. The significant stress in rolling-contact fatigue is the maximum alternating shear stress that undergoes a reversal in direction during rolling. In pure rolling, as in antifriction bearings, this stress occurs slightly below the surface and can lead to the initiation of subsurface fatigue cracks. As these cracks propagate under the repeated loads, they reach the surface and produce cavities, or pits. When sliding is superimposed on rolling, as in gear teeth, the tangential forces and thermal gradient caused by friction alter the magnitude and distribution of stresses in and below the contact area. The alternating shear stress is increased in magnitude and is moved nearer to the surface by friction resulting from the sliding action. Additional information is provided in the articles “Fatigue Failures” and “Failures of Rolling-Element Bearings” in this Volume. Failures of Shafts Revised by Donald J. Wulpi, Metallurgical Consultant Wear Wear of metal parts is commonly classified into either of two categories: abrasive wear or adhesive wear. Abrasive wear, the undesired removal of material by a cutting mechanism, can reduce the size and destroy the proper shape of a shaft. The shaft may then fail by another means, such as by fracture, or may cease to perform its designed function. Foreign particles, such as sand, dirt and other debris, in the lubricant can cause wear of a shaft. Example 3: Wear Failure of a Fuel-Pump Drive Shaft Caused by the Presence of Sand, Metallic Particles, and Vibration. The fuel pump in a turbine-powered aircraft failed, resulting in damage to the aircraft. The pump is shown in Fig. 8(a) and (b)
.28oD SHt-sEre 25 dom (b) Drive shaft and impeller (al Fuel pump Fig 8 Fuel pump that failed by vibration and abrasion.(a)Configuration and dimensions (given in inches).(b) Splines on the drive shaft and in the impeller were worn away by vibration in the presence of sand and metallic particles. Detail A: Enlarged view of failure area showing worn splines This particular model of fuel pump had a history of wear failures and had been redesigned to incorporate a shaft of case hardened steel(composition not reported). Vibration was common during operation, but generally was not excessive for the aircraft Investigation. The pump and the filter chamber were found to be dry and free of any debris or contamination, except for some accumulated deposit on the filter cartridge, when the pump was disassembled in the laboratory. The drive-shaft splines that engaged the impeller were almost completely destroyed down to the roots(Detail A, Fig. 8). Extensive damage to the splines was apparent because on subsequent reassembly the shaft could be rotated without rotating the impeller The pressure side of each spline tooth in the impeller also exhibited some damage. Relatively smooth cavities and undercutting of the flank on the pressure side of the spline teeth indicated that the damage had not been caused by wear from metallic contact between the splines but by an erosion or abrasion mechanism Hardness readings taken at several axial locations on the drive shaft showed a reasonably uniform hardness of approximately 570 HV. The impeller and the retaining ring each had a hardness of approximately 780 HV, and the parts surrounding the impeller(including the vanes )exhibited a hardness of approximately 630 HV. The microstructure of the metal in the impeller exhibited scattered porosity and carbide particles and appeared to be a sintered powder metallurgy compact Metallographic examination of a section through the damaged splines and of a section through the adjacent undamaged part of the same splines disclosed no material defects. The microstructure indicated that the shaft had been satisfactorily heat treated by quenching and tempering, but there was no evidence of case hardening on the spline surfaces, a treatment commonly given to shafts of this type The worn surfaces of the splines showed evidence of cold work at the edges. Also, there was a relatively smooth worn area at the center of each tooth that appeared to be free of cold work and that appeared to have been caused by an abrasive action. The damaged side of each spline appeared as an undulating outline with some undercutting rather than a jagged or deformed shap or plastic, fibers from the cartridge, brass, and steel could be identified. Application of a nope, particles of sand, paint. The residue on the filter cartridge was brown, and when viewed under a low-power micros magnet to the sample showed that it contained a large amount of iron Chemical analysis indicated that the deposit contained about 20% sand, 30%iron, and 30% organic material(paint, plastic, and filter fibers). The reddish-brown color of the deposit suggested that some of the iron present was an oxide (rust), but this was not confirmed Discussion. Vibration in the fuel pump could be expected to initiate damage, particularly when combined with an abrasive action. Under these conditions, fretting or abrasive wear can be expected on sliding-contact surfaces that are not sufficiently abrasion resistant. The residue from the filter contained significant quantities of sand and iron; the iron probably originated from the damaged shaft The examination indicated that the splines on the drive shaft had been damaged by abrasion, which could have been caused by the combined effect of vibration and abrasives, such as sand and the metal particles removed from the splines Thefileisdownloadedfromwww.bzfxw.com
Fig. 8 Fuel pump that failed by vibration and abrasion. (a) Configuration and dimensions (given in inches). (b) Splines on the drive shaft and in the impeller were worn away by vibration in the presence of sand and metallic particles. Detail A: Enlarged view of failure area showing worn splines This particular model of fuel pump had a history of wear failures and had been redesigned to incorporate a shaft of casehardened steel (composition not reported). Vibration was common during operation, but generally was not excessive for the aircraft. Investigation. The pump and the filter chamber were found to be dry and free of any debris or contamination, except for some accumulated deposit on the filter cartridge, when the pump was disassembled in the laboratory. The drive-shaft splines that engaged the impeller were almost completely destroyed down to the roots (Detail A, Fig. 8). Extensive damage to the splines was apparent because on subsequent reassembly the shaft could be rotated without rotating the impeller. The pressure side of each spline tooth in the impeller also exhibited some damage. Relatively smooth cavities and undercutting of the flank on the pressure side of the spline teeth indicated that the damage had not been caused by wear from metallic contact between the splines but by an erosion or abrasion mechanism. Hardness readings taken at several axial locations on the drive shaft showed a reasonably uniform hardness of approximately 570 HV. The impeller and the retaining ring each had a hardness of approximately 780 HV, and the parts surrounding the impeller (including the vanes) exhibited a hardness of approximately 630 HV. The microstructure of the metal in the impeller exhibited scattered porosity and carbide particles and appeared to be a sintered powder metallurgy compact. Metallographic examination of a section through the damaged splines and of a section through the adjacent undamaged part of the same splines disclosed no material defects. The microstructure indicated that the shaft had been satisfactorily heat treated by quenching and tempering, but there was no evidence of case hardening on the spline surfaces, a treatment commonly given to shafts of this type. The worn surfaces of the splines showed evidence of cold work at the edges. Also, there was a relatively smooth worn area at the center of each tooth that appeared to be free of cold work and that appeared to have been caused by an abrasive action. The damaged side of each spline appeared as an undulating outline with some undercutting rather than a jagged or deformed shape. The residue on the filter cartridge was brown, and when viewed under a low-power microscope, particles of sand, paint, or plastic, fibers from the cartridge, brass, and steel could be identified. Application of a magnet to the sample showed that it contained a large amount of iron. Chemical analysis indicated that the deposit contained about 20% sand, 30% iron, and 30% organic material (paint, plastic, and filter fibers). The reddish-brown color of the deposit suggested that some of the iron present was an oxide (rust), but this was not confirmed. Discussion. Vibration in the fuel pump could be expected to initiate damage, particularly when combined with an abrasive action. Under these conditions, fretting or abrasive wear can be expected on sliding-contact surfaces that are not sufficiently abrasion resistant. The residue from the filter contained significant quantities of sand and iron; the iron probably originated from the damaged shaft. The examination indicated that the splines on the drive shaft had been damaged by abrasion, which could have been caused by the combined effect of vibration and abrasives, such as sand and the metal particles removed from the splines. The file is downloaded from www.bzfxw.com