References cited in this section 17. P F. Wilson, L D. Dell, and G F. Anderson, Root Cause Analysis: A Tool for Total Quality Management, AsQ Quality Press, 1993, p 50 18. G.F. Smith, Quality Problem Solving, AsQ Quality Press, 1998, p 12 19. M. Paradise, L. Unger, and D. Busch, Tap Roote Root Cause Tree TM User's Manual, Systems Improvement, Inc 996,p914 20.R J. Latino and K.C. Latino, Root Cause Analysis: Improving Performance for Bottom Line Results, Reliability Center, Inc, 1999, p 79-89 21. C. Nelms, What You Can Learn From Things That Go Wrong, Ist ed, Failsafe Network, Richmond, VA, 1994 22. H P. Bloch and F.K. Geitner, Practical Machinery Management for Process Plants, Vol 2, Machinery failure Analysis and Troubleshooting, Gulf Publishing Co., 1983, p 5-6 Introduction to Failure Analysis and Prevention James ]. Scutti, Massachusetts Materials Research, Inc. William ]. McBrine, ALTRAN Corporation Primary Physical Root Causes of Failure Categorizing schemes for the root causes of equipment failures vary among failure analysis practitioners, quality engineers, other engineers, and managers, as well as legal and insurance professionals(Ref 13, 15, 23, 24, 25, 26, 27) Grouping physical root causes into only a few fundamental categories is advantageous and informative because it defines which aspect of a product or system requires corrective action and prevention strategies. Systematic analysis of equipment failures reveals physical root causes that fall into one of four fundamental categories(Ref 28) Design deficiencies Material defects Manufacturing/installation defects Service life anomalies An effective graphical representation of the impact of defects on the service life of a component or system is provided in the application-life diagram(Fig. 5)(Ref 29, 30). The diagram is constructed by plotting the service lives of components having specific characteristics in the design/configuration, as related to the severity of a specific service condition that is anticipated for the application. Typical characteristics include strength, corrosion resistance, heat treatment condition. flaw size, surface finish, bend radius, void content (i.e, in a casting), degree of sensitization, and so forth. Examples of ervice conditions include magnitude of stress (either cyclic or static), exposure temperature, aggressiveness of environment, radiation exposure, electrical stress, and so forth Thefileisdownloadedfromwww.bzfxw.com
References cited in this section 17. P.F. Wilson, L.D. Dell, and G.F. Anderson, Root Cause Analysis: A Tool for Total Quality Management, ASQ Quality Press, 1993, p 50 18. G.F. Smith, Quality Problem Solving, ASQ Quality Press, 1998, p 127 19. M. Paradise, L. Unger, and D. Busch, TapRoot® Root Cause Tree™ User's Manual, Systems Improvement, Inc., 1996, p 9–14 20. R.J. Latino and K.C. Latino, Root Cause Analysis: Improving Performance for Bottom Line Results, Reliability Center, Inc., 1999, p 79–89 21. C. Nelms, What You Can Learn From Things That Go Wrong, 1st ed., Failsafe Network, Richmond, VA, 1994 22. H.P. Bloch and F.K. Geitner, Practical Machinery Management for Process Plants, Vol 2, Machinery Failure Analysis and Troubleshooting, Gulf Publishing Co., 1983, p 5–6 Introduction to Failure Analysis and Prevention James J. Scutti, Massachusetts Materials Research, Inc.; William J. McBrine, ALTRAN Corporation Primary Physical Root Causes of Failure Categorizing schemes for the root causes of equipment failures vary among failure analysis practitioners, quality engineers, other engineers, and managers, as well as legal and insurance professionals (Ref 13, 15, 23, 24, 25, 26, 27). Grouping physical root causes into only a few fundamental categories is advantageous and informative because it defines which aspect of a product or system requires corrective action and prevention strategies. Systematic analysis of equipment failures reveals physical root causes that fall into one of four fundamental categories (Ref 28): · Design deficiencies · Material defects · Manufacturing/installation defects · Service life anomalies An effective graphical representation of the impact of defects on the service life of a component or system is provided in the application-life diagram (Fig. 5) (Ref 29, 30). The diagram is constructed by plotting the service lives of components having specific characteristics in the design/configuration, as related to the severity of a specific service condition that is anticipated for the application. Typical characteristics include strength, corrosion resistance, heat treatment condition, flaw size, surface finish, bend radius, void content (i.e., in a casting), degree of sensitization, and so forth. Examples of service conditions include magnitude of stress (either cyclic or static), exposure temperature, aggressiveness of environment, radiation exposure, electrical stress, and so forth. The file is downloaded from www.bzfxw.com
c88 D 3 Design 2 自 gg 2 Anticipated service condition Premature failure I Prototype Increasing service life Intended life Fig. 5 Application-life diagram comparing the severity of a service condition with the service lives of products having a variable characteristic. This diagram is utilized in specific examples in text. By varying the characteristics, a family of curves is generated, contrasting the lives of components with the various characteristics and service conditions with the intended service life. Each of the curves represents a different design/configuration characteristic, with increasing degrees of durability as the curves move up the ordinate. Failures can be prevented when the curve for a specific design/configuration lies above the severity of service line, and to the left of the intended service life line. However, if the severity of service conditions increases (either intentionally during operation or as a result of some other change in the system), the propensity for failure may increase, since the characteristics curves intersect the severity of service condition line to the left,that is, at an earlier point in the service Design Deficiencies Root causes of failures that stem from design deficiencies refer to unacceptable features of a product or system that are a result of the design process. This process encompasses the original concept development, the general configuration definition, and the detail design, including selection and specification of materials and manufacturing processes. Design involves identifying and defining a need for the product or system, followed by definition of the performance requirements, anticipated service conditions in the application(s), the constraints on the design, and the criticality or risks associated with failure(Ref 31). Discussion of the design process as it relates to failure analysis and prevention is provided in the article"Design Review for Failure Analysis and Prevention"in this Section Some examples of design deficiencies include unintended stress raisers due to excessively sharp notches(Ref 32)(e.g,in keyways on shafts)or insufficient radii(e.g, on shafts at bearing journals). Other examples include unanticipated residual stresses associated with heat treating configurations designed with complex geometries, or assembly stresses from configurations that contain unwanted interference. Inappropriate surface treatments could result in failures, such as the use of cadmium plating on an A286 superalloy fastener, subjected to service temperatures above 315C(600F)(the elting temperature of cadmium is 320C, or 610F). Two metals specified for use in a wear application could sustain galling if the metals are incompatible, such as sliding wear of components made from 300 series stainless steels Selection of a material that is incapable of providing adequate mechanical properties for the application(including strength, fatigue resistance, fracture toughness, corrosion resistance, elevated temperature resistance, etc )is also a type of design deficiency. Materials can exhibit anisotropy, or variability in properties within a product, such as between the thick
Fig. 5 Application-life diagram comparing the severity of a service condition with the service lives of products having a variable characteristic. This diagram is utilized in specific examples in text. By varying the characteristics, a family of curves is generated, contrasting the lives of components with the various characteristics and service conditions with the intended service life. Each of the curves represents a different design/configuration characteristic, with increasing degrees of durability as the curves move up the ordinate. Failures can be prevented when the curve for a specific design/configuration lies above the severity of service line, and to the left of the intended service life line. However, if the severity of service conditions increases (either intentionally during operation or as a result of some other change in the system), the propensity for failure may increase, since the characteristics curves intersect the severity of service condition line “to the left,” that is, at an earlier point in the service life. Design Deficiencies Root causes of failures that stem from design deficiencies refer to unacceptable features of a product or system that are a result of the design process. This process encompasses the original concept development, the general configuration definition, and the detail design, including selection and specification of materials and manufacturing processes. Design involves identifying and defining a need for the product or system, followed by definition of the performance requirements, anticipated service conditions in the application(s), the constraints on the design, and the criticality or risks associated with failure (Ref 31). Discussion of the design process as it relates to failure analysis and prevention is provided in the article “Design Review for Failure Analysis and Prevention” in this Section. Some examples of design deficiencies include unintended stress raisers due to excessively sharp notches (Ref 32) (e.g., in keyways on shafts) or insufficient radii (e.g., on shafts at bearing journals). Other examples include unanticipated residual stresses associated with heat treating configurations designed with complex geometries, or assembly stresses from configurations that contain unwanted interference. Inappropriate surface treatments could result in failures, such as the use of cadmium plating on an A286 superalloy fastener, subjected to service temperatures above 315 °C (600 °F) (the melting temperature of cadmium is 320 °C, or 610 °F). Two metals specified for use in a wear application could sustain galling if the metals are incompatible, such as sliding wear of components made from 300 series stainless steels. Selection of a material that is incapable of providing adequate mechanical properties for the application (including strength, fatigue resistance, fracture toughness, corrosion resistance, elevated temperature resistance, etc.) is also a type of design deficiency. Materials can exhibit anisotropy, or variability in properties within a product, such as between the thick
and thin portions of a casting, or between longitudinal and transverse properties in a wrought material. Note that a material can be shown to meet the properties required or specified (i.e, a separately cast tensile bar used to certify a casting, or the longitudinal tensile properties to certify a complex aluminum extrusion), but the specific properties required for the application may rely on the strength, toughness, or stress-corrosion cracking resistance in a direction other than longitudinal Design-caused failures include inappropriate geometries(as defined on the engineering drawing), which may lead to a compromise of component or system capabilities. Examples of inappropriate geometries include improper joint preparation for welding or brazing, such as an insufficient or missing groove for a groove weld, insufficient fit-up relief in a socket weld, or inadequate joint overlap in a brazed joint. Other geometry-caused failures can result from insufficient section thickness for a failure based on gross yielding, excessive section thickness in the presence of a flaw for a material of limited fracture toughness, or a fabrication configuration with an excessively sharp forming bend, with the resulting high residual stresses causing a reduction in the fatigue life For the example of the excessively tight cold-formed bend radius described previously, an application-life diagram can be constructed as shown in Fig. 6. The service condition considered is stress, and the characteristic that is varied is the radius of the cold-formed bend. Upon examination of the relationship between the characteristic curves and the intended service life, the components having the large and moderate bend radii are found to meet the intended service life at the severity of stress that is anticipated in the specific application. However, in this illustration, the component with the small bend radius sustained a premature failure at the anticipated stress level in the application, since the curve intersects the severity of stress line prior to reaching the intended service life Cold-formed products og0o66go6 Large bend radius Anticipated severity Moderate bend radius 十 Premature failure smal‖ bend radius Increasing service life Intended life Fig 6 Application-life diagram for design deficiency Some of the aforementioned deficiencies in design as well as application-life diagram concepts are illustrated in the following two case histories Example 1: Ice Cream Drink Mixer Blade Failures. Excessive assembly stresses and inappropriate detail design caused the premature failures of ice cream drink mixer blades shortly after the mixing machines were introduced into service. A mixer blade as-manufactured is shown on the left side of Fig. 7. As assembled(right side of Fig. 7), the mixer blade is slightly deformed by the contact between the wavy washer at the bottom of the assembly and the bends at the bottom shoulders of the two mixer arms. When properly torqued, the screw that fastens the wavy washer and the mixer blade to the spindle in the center of the assembly places an upward force on the bottoms of the arms(as indicated by the pair of pward facing arrows in Fig. 7). This results in the observed inward deflection of the arms(as indicated by the right and left facing arrows). More significantly, this bending force places the inside radii of the two shoulders of the mixing blade Thefileisdownloadedfromwww.bzfxw.com
and thin portions of a casting, or between longitudinal and transverse properties in a wrought material. Note that a material can be shown to meet the properties required or specified (i.e., a separately cast tensile bar used to certify a casting, or the longitudinal tensile properties to certify a complex aluminum extrusion), but the specific properties required for the application may rely on the strength, toughness, or stress-corrosion cracking resistance in a direction other than longitudinal. Design-caused failures include inappropriate geometries (as defined on the engineering drawing), which may lead to a compromise of component or system capabilities. Examples of inappropriate geometries include improper joint preparation for welding or brazing, such as an insufficient or missing groove for a groove weld, insufficient fit-up relief in a socket weld, or inadequate joint overlap in a brazed joint. Other geometry-caused failures can result from insufficient section thickness for a failure based on gross yielding, excessive section thickness in the presence of a flaw for a material of limited fracture toughness, or a fabrication configuration with an excessively sharp forming bend, with the resulting high residual stresses causing a reduction in the fatigue life. For the example of the excessively tight cold-formed bend radius described previously, an application-life diagram can be constructed as shown in Fig. 6. The service condition considered is stress, and the characteristic that is varied is the radius of the cold-formed bend. Upon examination of the relationship between the characteristic curves and the intended service life, the components having the large and moderate bend radii are found to meet the intended service life at the severity of stress that is anticipated in the specific application. However, in this illustration, the component with the small bend radius sustained a premature failure at the anticipated stress level in the application, since the curve intersects the severity of stress line prior to reaching the intended service life. Fig. 6 Application-life diagram for design deficiency Some of the aforementioned deficiencies in design as well as application-life diagram concepts are illustrated in the following two case histories. Example 1: Ice Cream Drink Mixer Blade Failures. Excessive assembly stresses and inappropriate detail design caused the premature failures of ice cream drink mixer blades shortly after the mixing machines were introduced into service. A mixer blade as-manufactured is shown on the left side of Fig. 7. As assembled (right side of Fig. 7), the mixer blade is slightly deformed by the contact between the wavy washer at the bottom of the assembly and the bends at the bottom shoulders of the two mixer arms. When properly torqued, the screw that fastens the wavy washer and the mixer blade to the spindle in the center of the assembly places an upward force on the bottoms of the arms (as indicated by the pair of upward facing arrows in Fig. 7). This results in the observed inward deflection of the arms (as indicated by the right and left facing arrows). More significantly, this bending force places the inside radii of the two shoulders of the mixing blade The file is downloaded from www.bzfxw.com
arms(at the bottom of the blade)in tension. When the mixer is running, the rotational forces further add to the tensile loads on the inside radii of the shoulders Fig 7 Ice cream mixer blade as manufactured (left) and assembled to spindle(right) Analysis of the failed mixer blades revealed multiple fatigue crack origins on the inside radii of the bends at the bottom shoulders(Fig. 8). Metallographic examination of the arm materials revealed additional problems with the configuration the shoulders on the arms were cold bent. introducing tensile residual stresses on the inside radii of the shoulders and creating a localized area of fatigue susceptibility due to the inherent notch sensitivity of cold-formed 300 series stainless steel B Fig. 8 Fracture surface of failed ice cream mixer blade. arrows indicate fatigue crack origins.13× Clearly, the physical root cause is the design of the mixer blade, which defined two bend that contained tensile residual stresses, tensile assembly stresses, and a notch-sensitive microstructure that added to the normal operating rotational and vibratory stresses. The net effect was a reduction in the life of the blade causing loss of function Corrective-action recommendations included the addition of a stand-off washer between the wavy washer and the bottom shoulders of the blade, or modification of the shape of the wavy washer to prevent contact with the blade shoulders as assembled
arms (at the bottom of the blade) in tension. When the mixer is running, the rotational forces further add to the tensile loads on the inside radii of the shoulders. Fig. 7 Ice cream mixer blade as manufactured (left) and assembled to spindle (right) Analysis of the failed mixer blades revealed multiple fatigue crack origins on the inside radii of the bends at the bottom shoulders (Fig. 8). Metallographic examination of the arm materials revealed additional problems with the configuration: the shoulders on the arms were cold bent, introducing tensile residual stresses on the inside radii of the shoulders and creating a localized area of fatigue susceptibility due to the inherent notch sensitivity of cold-formed 300 series stainless steel. Fig. 8 Fracture surface of failed ice cream mixer blade. Arrows indicate fatigue crack origins. 13× Clearly, the physical root cause is the design of the mixer blade, which defined two bend areas that contained tensile residual stresses, tensile assembly stresses, and a notch-sensitive microstructure that added to the normal operating rotational and vibratory stresses. The net effect was a reduction in the life of the blade causing loss of function. Corrective-action recommendations included the addition of a stand-off washer between the wavy washer and the bottom shoulders of the blade, or modification of the shape of the wavy washer to prevent contact with the blade shoulders as assembled
Example 2: Sprocket Locking Device Failure.(Ref 33). A design deficiency involving improper materials selection was revealed through the analysis of a failed tapered-ring sprocket locking device. The device is used to attach a chain sprocket to a shaft without the use of a locking key, enabling the shaft to either drive or be driven anywhere on the shaft (see Fig 9). The configuration consists of an assembly of four tapered rings(Fig. 10) that are retained by a series of cap screws. As shown in Fig. 11, when the screws are tightened, the middle wedge-shaped rings are pulled closer, forcing the split inner ring to clamp tightly onto the shaft, and the split outer ring to force tightly against the inside diameter of the sprocket. When properly assembled and torqued, the sprocket is fixed to the shaft ocking device Fig 9 Sketch of tapered-ring locking device application Occ Fig. 10 Four tapered rings of locking device. Arrow indicates crack in one of the middle Ings. Cap Outer ring screw ⑥ ring O6⑩ Fig. ll Plan view (left) and cross section (right) through tapered-ring locking device assembly During initial assembly of a new locking device by the manufacturer during a bench test, one of the wedge-shaped middle rings fractured prior to having been fully torqued, preventing the sprocket from being locked to the shaft. The failed assembly was investigated for root cause. One of the middle rings had cracked(Fig. 10, 12a). Examination of the fracture Thefileisdownloadedfromwww.bzfxw.com
Example 2: Sprocket Locking Device Failure. (Ref 33). A design deficiency involving improper materials selection was revealed through the analysis of a failed tapered-ring sprocket locking device. The device is used to attach a chain sprocket to a shaft without the use of a locking key, enabling the shaft to either drive or be driven anywhere on the shaft (see Fig. 9). The configuration consists of an assembly of four tapered rings (Fig. 10) that are retained by a series of cap screws. As shown in Fig. 11, when the screws are tightened, the middle wedge-shaped rings are pulled closer, forcing the split inner ring to clamp tightly onto the shaft, and the split outer ring to force tightly against the inside diameter of the sprocket. When properly assembled and torqued, the sprocket is fixed to the shaft. Fig. 9 Sketch of tapered-ring locking device application Fig. 10 Four tapered rings of locking device. Arrow indicates crack in one of the middle rings. Fig. 11 Plan view (left) and cross section (right) through tapered-ring locking device assembly. During initial assembly of a new locking device by the manufacturer during a bench test, one of the wedge-shaped middle rings fractured prior to having been fully torqued, preventing the sprocket from being locked to the shaft. The failed assembly was investigated for root cause. One of the middle rings had cracked (Fig. 10, 12a). Examination of the fracture The file is downloaded from www.bzfxw.com