Experiment F: Failure Analysis Introduction: In this experiment, a component that has failed in service will be analyzed to determine the probable cause or causes of the failure. Each group will give a short oral presentation on their findings. Procedure For this laboratory, a broken component is required, any failed component can be used either bring something from home or work or a selection of components are available in the department. The component should have failed in service and be in a condition suitable for analysis. If in doubt about the suitability of the component ask the TA. Remember, don t fit together the fracture parts unless necessary, delicate features on the fracture surfaces can be damaged. Cleaning can remove important evidence and may contaminate or corrode surfaces. Each group must perform metallurgical analysis, and if needed a chemical analysis and SEM. Based on these results and any observations, each group will present its findings to the class in a BRIEF (5-7 minutes) oral presentation. In the presentation (and the report), the following points must be addressed: a. Function of the part in service b. Results of the chemical analysis (if appropriate) c. Description of the microstructure d. Description of the failure e. Why did the part fail (design, material, operating conditions?) f. What can be done to rectify the problem? An industrial presentation should be prepared, i.e., suppose that the finding are being presented to corporate engineers who are responsible for the design of the component. The due date will be discussed in the lab. Background: What is a failure? The fracture of a shaft in a machine is an oblivious failure. But is this example what the definition for a failure should be based on? What about a plane crash? During the course of a normal day how many situations occur where things or procedures don t work properly? A possible definition for a failure might be the inability of a component, machine or process to function properly.
In this experiment, the focus is on failures of components in service. Failure analysis will be used to determine the factors responsible for the failure of the component examined. The proper application of failure analysis techniques can provide valuable feedback to design problems and material limitations. In the real world, there may also be legal reasons for conducting a failure analysis: if a failure has caused injury or damage to expensive equipment, it will be necessary to determine where to lay the blame! Based on the results of a careful failure analysis, it is generally possible to attribute the problem to one of the following: Poor design (stress raisers, etc ) or improper selection of material Improper testing Service and operation (use, misuse, etc...) Improper fabrication or manufacturing procedures (casting, machining, improper heat treatment etc ) Improper assembly Improper maintenance Due to the very wide variety of factors which can contribute to an engineering failure, this type of analysis requires considerable knowledge in metallurgy, corrosion and machine design. In addition, there is no substitute for experience. Steps in Failure Analysis A proper failure analysis involves a systematic study of all the factors which may have contributed to the failure. The role of an engineer in a failure analysis situation is to answer one or more of the following questions: - What is the problem? - What is the main cause of the problem? - What are the potential solutions? - What is the best solution? It is important to remain objective and have clear information about the failure. In the real world, a lot of time and effort can be spent on understanding and negotiating the goals, procedures of an analysis with a client (internal or external) and the allowable costs associated with the analysis. Establishing clearly defined expectations is very important. Finding out what happened at the actual location of the failure. Investigate! Visiting the scene of the crime can often be very valuable. Workers on the shop floor may know more about what happened then the manager. Note-taking, documentation, drawings, pictures and planning are key parts of failure analysis. Do as much non-destructive analysis as possible before the destructive parts of the investigation. During the course of an investigation, it is important to identify all the possible causes of a failure and eliminate as many as possible either due to testing or the balance of probabilities. For solutions, look at all possibilities and list criteria for rejecting and accepting each solution. Select the best solution and if possible, evaluate the solution.
For the purposes of this lab, the following will service as a guide to the stages, which comprise the investigation and analysis of a failure: INITIAL OBSERVATIONS Preliminary visual study of the failed component LOTS of photographs COLLECTION OF BACKGROUND DATA (INVESTIGATION/RESEARCH) Component design (specifications, fabrications, repairs, material) How long was the part in service? Nature of stresses at time of failure? Working environment of the part (temperature, medium, corrosion) MACROSCOPIC EXAMINATION AND ANALYSIS Detailed visual study of the failed component Identification and interpretation of surface irregularities SELECTION, IDENTIFICATION, AND CLEANING OF SPECIMENS Select specimens for failure analysis Prepare metallographic sections MECHANICAL ANALYSIS Stress-strain curves, if available Hardness measurements (macro, micro) CHEMICAL ANALYSIS Verify composition of component to see whether it falls within specified limits MICROSCOPIC EXAMINATION AND ANALYSIS Metallographic analysis (correct microstructure) Scanning electron microscopy (wear, fractures, corrosion, microstructures) EDS - Elemental analysis with the SEM (inclusions, corrosion, phases) TESTING UNDER SIMULATED SERVICE CONDITIONS (if possible) May help to identity cause of failure SYNTHESIS OF FAILURE Analysis of all facts and evidence Formulation of conclusions Write report and suggest solution to problem Design? Normal wear? Material? Etc... Some Notions about Failures
In this section, some notions about failures (and fracture in particular) will be discussed. It is a brief overview and students wishing to look at the subject more in depth are encouraged to do some research on their own in the library. Most books on introductory metallurgy will have at least one section on failure analysis. Also, there are quite a few books devoted entirely to the subject. Figures and micrographs are provided at the end help to illustrate the points raised in the following discussion. Primarily, the discussion applies to metals, but many non-metallic or composite materials will exhibit similar behaviour. When a part or component has fractured, it is often necessary to determine the nature of the fracture in order to draw any conclusion about the failure. Was the fracture ductile, brittle, or a combination of the two? Was fatigue, wear, or corrosion involved in the failure? Where did the fracture start at or below the surface? Did the fracture start at one point, or did it originate at several points? What was the loading of the component at the point of fracture? Tensile? Shear? Tearing? Figure 1. A - Tensile B - Shear C - Tearing Modes of Fracture - Ductile or Brittle? Figure 2. Ductile Cup - Cone Fracture. The mode of facture of a part may be difficult to discern if the background information of the failure is limited. Mix modes of fracture may occur within the same part. An understanding of fracture mechanics and fractography can be very helpful. Ductile fractures are high-energy fractures. They are the result of shear forces, which produce
plastic deformation (slip or twinning). Thus, they occur at stresses above the material s yield strength. They are characterized by stable crack propagation of the load causing cracks to propagate. When the load is removed, the crack stops advancing. Shear fractures caused by a single load are dull grey and fibrous, with edges that are usually deformed plastically. Examination of the fracture surface under the stereoscope will usually show dimples unless the fracture has been severely smeared at some point during fracture. During ductile fracture, small cavities or microvoids are initially formed by slip. These microvoids are normally associated with matrix/secondary particle (inclusions, precipitates etc) decohesion or cracking. When the microvoids join together or coalesce, they eventually grow to form a crack under continued loading. The microvoids appear as dimples after fracture. The crack propagates with the aid of stress concentration at its tip, generally moving perpendicular to the tensile force and eventually forming a shear lip at the surface (the plane of the shear lip will be at 45 from the tensile load axis). The classic example of a ductile fracture is the cup-and-cone geometry often observed during a standard tensile test on ductile materials. The reduction in area that occurs prior to the final ductile fracture of a component is commonly referred to as necking. If the material is very ductile, the sample may neck down to a point prior to final fracture. Brittle fractures are low-energy fractures. They occur at stresses well below the yield strength, are usually associated with flaws, are often catastrophic and usually occur rapidly without warning. They are distinguished from ductile fractures by the absence of gross plastic deformation and little or no evidence of reduction in area at the point of fracture. Figure 3. Brittle Fracture Brittle fractures often appear bright and crystalline. The two modes of brittle fracture are transgranular and intergranular. With transgranular fractures, each grain tends to fracture on a single cleavage plane. Thus a brittle fracture surface will often sparkle in the light when rotated in the hand. Brittle fracture surfaces sometimes have distinctive appearances from the origin of fracture, a characteristic chevron or herringbone pattern is formed which points to the fracture origin. The fracture surfaces are usually smooth compared with ductile failures and the fractured parts usually fit together well (please resist the temptation to do this unless Figure 4. A distinct Chevron pattern is apparent on the right-hand side of the fracture surface. absolutely necessary). Brittle failures occur in brittle materials, but can also occur in normally ductile materials under certain loading conditions (e.g. high strain rates, tri-axial stresses, or low temperatures).
In some materials such as certain tempered martensitic steels, the fracture is predominately brittle but has elements of ductile fracture. The fracture surface has small facets, dimples and ridges which sometimes result in a rosette-like appearance. This type of fracture mode is usually referred to as quasi-cleavage, a mixed mechanism of microvoid coalescence and cleavage. Figure 6. Quasi-cleavage Figure 6. Quasi cleavage Point of Origin Several features exist around fractures or on fractured surfaces, visible to the eye, that indicate the area where a fracture has initiated or its point(s) of origin. Branching cracks will usually originate from a single crack or point. A network of cracks will commonly have a larger or dominate crack. On fracture surfaces, chevron marks, radial marks or ridges can point to the region where the crack initiated. In the case of many fatigue failures, beach marks or alternating bands radiate out from the origin like waves on a beach. These marks can be said to identify the fracture origin as crack length increases, beach marks increase in size and spacing so that the location of the smallest beach mark indicates the origin of the failure. Figure 7. Radial and chevrons marks point to the region from which the crack originated. Figure 8. Locating crack origins.
Figure 9. Fatigue cracking is comprised of two stages. Figure 10. The first stage of fatigue cracking is referred to as Stage I. Fatigue Failure Fatigue failures are the most common types of fracture in machines. They result from the application of a repetitive stress over time, sometimes many years. There are two stages to a fatigue fracture. In Stage I, the surface or interface as the result of slip is subjected to intrusions and extrusions resulting in crack propagation usually at 45 o to the tensile stress. In Stage II, the fatigue fracture is progressive, i.e. there is a certain amount of crack growth per load application. The appearance of fatigue failures has often been described as brittle because there is very little gross plastic deformation associated with them and the fracture surfaces are fairly smooth. However, because fatigue failures are progressive, they usually leave characteristic marks, called beach marks as previously described. Note, however, that not all fatigue failures show beach marks. Therefore, their absence does not necessarily preclude fatigue. At the microscopic level, fatigue striations are created on the fracture surface by each progressive step of the crack and can be used to estimate the rate of crack growth - low or high cycle fatigue. Although beach marks appear similar regardless of the type of load application, specific features vary, as shown in the table at the end of the write-up. On fact common to all fatigue failures is that the fractures are caused by a tensile load generated by tension, rotation, or bending and the fracture path is normal to the tensile load. Bending Failure
Bending is one of the most common causes of fracture in machine and structural components. Failure can result from a single application of a load greater than the overall strength of a part of can be due to a reversing load that results in a bending fatigue fracture. Torsional Failure Torsional failure are most common in shafts, including crankshafts, torsion bars and axles. The appearance of a torsion-fatigue fracture is quite different from that caused by bending. Torsionfatigue failures occur along the planes of maximum shear or along the plane of maximum tension. Maximum shear stress occurs along the axis of the shaft and at right angles to it, while the maximum tensile stress acts at an angle of 45 to the two shear stresses. In torsion, the maximum shear stress is equal to the maximum tensile stress so that which type of fracture occurs will depend on the relative values of the shear strength and the tensile strength of the part. Wear The quality of most metal products depends on the condition of their surfaces and on surface deterioration due to use. Indeed, surface deterioration is often the major factor limiting the life and the performance of machine components. Wear may be defined as unintentional deterioration resulting from used or environment due to the displacement or detachment of metallic particles from a metallic surface. Several types of wear can be identified: (A) Adhesive or metallic wear (contact with another metal) (B) Abrasion (contact with a metallic or non-metallic abrasive) (C) Erosion (contact with moving liquids or gases) In adhesive wear (scoring, galling, seizing, scuffing), tiny projections produce friction by mechanical interference, with the relative motion of contacting surfaces increasing resistance to further motion. If the driving force is sufficient to maintain movement, the interlocked particles are deformed. If they are of brittle material, they may be torn off. Thus wear resistance can be improved by preventing metal to metal contact, increasing the hardness to resist initial indentation, increasing the toughness to resist the tearing out of metallic particles, and increasing the surface smoothness to eliminate projections. In adhesive wear, the effect of heat produced by friction between contacting surfaces can reduce war resistance in several ways. It may temper hardened structures, cause phase changes that increase hardness and brittleness, and accelerate corrosion reactions. If the pressure between two mating parts and/or their temperature is sufficiently large, the parts can be welded together they may seize and thus cause complete stoppage, or if relative motion is not prevented, pieces of the opposite face may pull out. Abrasive wear occurs when hard particles slide or roll under pressure across a surface. The hard particles tend to gouge or scratch the softer material. They may also penetrate the softer material
and cause the tearing off of metallic particles. Corrosion Corrosion may be defined as the destruction of a material by chemical, electrochemical or metallurgical interaction between the environment and the material. The basic cause of corrosion is the instability of metals in their refined forms. The metals tend to revert to their natural states through the process of corrosion. Corrosion is an electrochemical process resulting in part or all of a metal being transformed from the metallic to the ionic state. It requires the flow of electricity between certain areas of a metal surface through an electrolyte, which can be any solution containing ions. An electrochemical cell is made up of a connected anode and a cathode which are both in contact with the electrolyte. Material is removed from the anode, the metal which corrodes. Figure 11. Corrosion pits on an intergranular fracture surface.
Figure 12 The formation of dimples in a material containing spherical precipitates. during the tensile stress and different loading conditions Figure 13. Ductile Dimples.
Figure 14. Dimples from a less ductile material. Figure 16. A cleavage facet on the surface of a brittle fracture. Figure 15. Directional dimples caused by shear forces.
Figure 16. SEM Micrograph showing cleavage facets on the surface of a brittle fracture. A - a grain, B - another grain, C - a carbide particle and D and E - secondary cracks. Figure 18. An intergranular fracture.
Figure 19. Fatigue fracture showing beach marks.
Figure 20. Fatigue fracture with internal origin showing beach marks. Figure 21. Stage I of a fatigue fracture.
Figure 22. Diagram showing a typical stage II fatigue fracture with crack paths indicated. Figure 23. SEM Micrograph showing fatigue striations on the surface of a fatigue fracture.