Stage II striations, typical of some(not all) fatigue failures Stress analysis It is sometimes quite apparent that an excessively high load or stress level was the direct cause or contributed significantly to the failure. Even so, an accurate stress analysis of the magnitude and type(axial, torsion, bending) of stress is required to substantiate the role of stress. In other failure analyses, the analyst may have strong evidence that the cause of a failure is related to excessively high static stresses (or cyclic stresses in the case of fatigue. In these cases, an analysis of the stress during normal operation (or abnormal operation if identified) must be conducted. Analytical, closed-form calculations based on engineering mechanics are often used by designers to predict stress levels in the early design stages This method of using known "machine design and structural" formulas to predict the stress under a given load is also helpful to the failure analyst, especially in cases where this step may not have been used in the original design of the part It is not uncommon to find products that have no record of any stress analysis in the development of the product Even if such calculations were made, the failure analyst may not have access to them. The analyst must answer the questions Was the component sized properly by the design stress analysis? Did the material have the properties assumed in the design? Did the part fail in a manner consistent with that assumed in design, or did it fail in a way not anticipated in the original design? In cases where unusual or abnormal loading is suspected, direct calculation of stresses will fall short and predict incorrect stress levels. In these cases, experimental stress analysis is used for determining machine loads and component stresses This technique normally involves attachment of strain gages to similar parts in critical areas or typical areas where the failure has occurred. The strain gages are connected to a monitoring device either directly with wires or indirectly by radio signals for monitoring moving or rotating parts. In this way, the actual dynamic stresses can be determined For products with very complex shapes and high thermal gradients, a finite-element analysis(FEA)may be required to estimate the level of stress that most likely existed in the failed component. These analyses can stand alone or can be used to help select critical locations for strain-gage attachment. Finite-element analyses can be time consuming and expensive but they are necessary for an accurate assessment of stress levels in areas of complex geometry of some components. This type of analysis is almost essential for determining stresses caused by thermal gradients such as those found in welding Overload failures are often a result of improper design or improper operation. a design analysis is essential in determining which of these is the root cause. Sometimes improper design is a result of incorrect information passed to the designer. In these cases, a failure is the only indication that the wrong input was used for the design. This is also true for fatigue failures. For proper design of rotating or moving parts, a detailed stress analysis is essential. It is much more difficult to predict dynamic stresses than static stresses Fracture Modes Because the initial steps in failure analysis of a fracture involve visual and macroscopic observation, the first impressions should be based on obvious visual evidence. The simplest and most important observations relate to deformation: Was the metal obviously deformed? If it was deformed prior to fracture, yielding and fracture have occurred due to one or more gross overloads. It is predominantly a ductile fracture or a very high-stress, low-cycle fatigue fracture, as can be demonstrated by repeated manual bending of a paper clip or wire coat hanger. The deformation is directly related to the type of stress causing fracture: tension(stretched), bending(bent), torsion(twisted), or compression(shortened or buckled), or a combination of these stress types The absence of gross deformation of the failed part indicates that the fracture is predominantly brittle. a brittle fracture should not be confused with brittle material. The shape or geometry of the part made from a ductile metal can result in an yerload failure with little overall shape change, or a failure mechanism can operate to start and grow a crack such as a tigue crack or stress-corrosion crack. When such a crack grows to the point that the remaining cross-sectional area of the part is overloaded by the normal loads, the final overload failure has little macroscale deformation associated with it Thus, on a macroscale, the failure of a ductile metal can appear brittle. Of course, overload failures of brittle material al ways appear brittle on a macroscale. It is usually more difficult to analyze a brittle fracture because there are a large number of possible mechanisms that can cause fracture with little or no obvious deformation. For single overload fractures, these include such factors as stress concentrations, low temperatures, high rates of loading, high metal strength and hardness, SSC, hydrogen embrittlement, temper embrittlement, large section size, and others. For fatigue fractures, causative factors can include stress concentrations, tensile residual stresses, large stress amplitudes, large numbers of load applications, corrosive environments, high temperatures, low metal strength and hardness, wear, and others From this discussion, it should become clear that proper failure analysis is not simple but can become exceeding complex, requiring considerable thought, examination, questioning, and reference to other sources of information in the· Stage II striations, typical of some (not all) fatigue failures Stress Analysis It is sometimes quite apparent that an excessively high load or stress level was the direct cause or contributed significantly to the failure. Even so, an accurate stress analysis of the magnitude and type (axial, torsion, bending) of stress is required to substantiate the role of stress. In other failure analyses, the analyst may have strong evidence that the cause of a failure is related to excessively high static stresses (or cyclic stresses in the case of fatigue). In these cases, an analysis of the stress during normal operation (or abnormal operation if identified) must be conducted. Analytical, closed-form calculations based on engineering mechanics are often used by designers to predict stress levels in the early design stages. This method of using known “machine design and structural” formulas to predict the stress under a given load is also helpful to the failure analyst, especially in cases where this step may not have been used in the original design of the part. It is not uncommon to find products that have no record of any stress analysis in the development of the product. Even if such calculations were made, the failure analyst may not have access to them. The analyst must answer the questions “Was the component sized properly by the design stress analysis? Did the material have the properties assumed in the design? Did the part fail in a manner consistent with that assumed in design, or did it fail in a way not anticipated in the original design?” In cases where unusual or abnormal loading is suspected, direct calculation of stresses will fall short and predict incorrect stress levels. In these cases, experimental stress analysis is used for determining machine loads and component stresses. This technique normally involves attachment of strain gages to similar parts in critical areas or typical areas where the failure has occurred. The strain gages are connected to a monitoring device either directly with wires or indirectly by radio signals for monitoring moving or rotating parts. In this way, the actual dynamic stresses can be determined. For products with very complex shapes and high thermal gradients, a finite-element analysis (FEA) may be required to estimate the level of stress that most likely existed in the failed component. These analyses can stand alone or can be used to help select critical locations for strain-gage attachment. Finite-element analyses can be time consuming and expensive, but they are necessary for an accurate assessment of stress levels in areas of complex geometry of some components. This type of analysis is almost essential for determining stresses caused by thermal gradients such as those found in welding. Overload failures are often a result of improper design or improper operation. A design analysis is essential in determining which of these is the root cause. Sometimes improper design is a result of incorrect information passed to the designer. In these cases, a failure is the only indication that the wrong input was used for the design. This is also true for fatigue failures. For proper design of rotating or moving parts, a detailed stress analysis is essential. It is much more difficult to predict dynamic stresses than static stresses. Fracture Modes Because the initial steps in failure analysis of a fracture involve visual and macroscopic observation, the first impressions should be based on obvious visual evidence. The simplest and most important observations relate to deformation: Was the metal obviously deformed? If it was deformed prior to fracture, yielding and fracture have occurred due to one or more gross overloads. It is predominantly a ductile fracture or a very high-stress, low-cycle fatigue fracture, as can be demonstrated by repeated manual bending of a paper clip or wire coat hanger. The deformation is directly related to the type of stress causing fracture: tension (stretched), bending (bent), torsion (twisted), or compression (shortened or buckled), or a combination of these stress types. The absence of gross deformation of the failed part indicates that the fracture is predominantly brittle. A brittle fracture should not be confused with brittle material. The shape or geometry of the part made from a ductile metal can result in an overload failure with little overall shape change, or a failure mechanism can operate to start and grow a crack such as a fatigue crack or stress-corrosion crack. When such a crack grows to the point that the remaining cross-sectional area of the part is overloaded by the normal loads, the final overload failure has little macroscale deformation associated with it. Thus, on a macroscale, the failure of a ductile metal can appear brittle. Of course, overload failures of brittle material always appear brittle on a macroscale. It is usually more difficult to analyze a brittle fracture because there are a large number of possible mechanisms that can cause fracture with little or no obvious deformation. For single overload fractures, these include such factors as stress concentrations, low temperatures, high rates of loading, high metal strength and hardness, SSC, hydrogen embrittlement, temper embrittlement, large section size, and others. For fatigue fractures, causative factors can include stress concentrations, tensile residual stresses, large stress amplitudes, large numbers of load applications, corrosive environments, high temperatures, low metal strength and hardness, wear, and others. From this discussion, it should become clear that proper failure analysis is not simple but can become exceedingly complex, requiring considerable thought, examination, questioning, and reference to other sources of information in the