Failure analysis is an investigative process that uses visual observations of features present on a failed component fracture surface combined with component and environmental conditions to determine the root cause of a failure. The primary means of recording the conditions and features observed during a failure analysis investigation is photography. Failure analysis photographic imaging is a combination of both science and art; experience and proper imaging techniques are required to produce an accurate and meaningful fracture surface photograph. This article reviews photographic principles and techniques as applied to failure analysis, both in the field and in the laboratory. The discussion covers the processes involved in field and laboratory photographic documentations, provides a description of professional digital cameras, and gives information on photographic lighting and microscopic photography. Special techniques can be employed to deal with highly reflective conditions and are also described in this article.

The US Army Research Laboratory performed a failure investigation on a broken main landing gear mount from an AH-64 Apache attack helicopter. A component had failed in flight, and initially prevented the helicopter from safely landing. In order to avoid a catastrophe, the pilot had to perform a low hover maneuver to the maintenance facility, where ground crews assembled concrete blocks at the appropriate height to allow the aircraft to safely touch down. The failed part was fabricated from maraging 300 grade steel (2,068 MPa [300 ksi] ultimate tensile strength), and was subjected to visual inspection/light optical microscopy, metallography, electron microscopy, energy dispersive spectroscopy, chemical analysis, and mechanical testing. It was observed that the vacuum cadmium coating adjacent to the fracture plane had worn off and corroded in service, thus allowing pitting corrosion to occur. The failure was hydrogen-assisted and was attributed to stress corrosion cracking (SCC) and/or corrosion fatigue (CF). Contributing to the failure was the fact that the material grain size was approximately double the required size, most likely caused from higher than nominal temperatures during thermal treatment. These large grains offered less resistance to fatigue and SCC. In addition, evidence of titanium-carbo-nitrides was detected at the grain boundaries of this material that was prohibited according to the governing specification. This phase is formed at higher thermal treatment temperatures (consistent with the large grains) and tends to embrittle the alloy. It is possible that this phase may have contributed to the intergranular attack. Recommendations were offered with respect to the use of a dry film lubricant over the cadmium coated region, and the possibility of choosing an alternative material with a lower notch sensitivity. In addition, the temperature at which this alloy is treated must be monitored to prevent coarse grain growth. As a result of this investigation and in an effort to eliminate future failures, ARL assisted in developing a cadmium brush plating procedure, and qualified two Army maintenance facilities for field repair of these components.

A commercial aircraft wheel half, machined from an aluminum alloy 2014 forging that had been heat treated to the T6 temper, was removed from service because a crack was discovered in the area of the grease-dam radius during a routine inspection. Neither the total number of landings nor the roll mileage was reported, but about 300 days had elapsed between the date of manufacture and the date the wheel was removed from service. The analysis (visual inspection, macrographs, micrographs, electron microprobe) supported the conclusions that the wheel half failed by fatigue. The fatigue crack originated at a material imperfection and progressed in more than one plane because changes in the direction of wheel rotation altered the direction of the applied stresses. Recommendations included rewriting the inspection specifications to require sound forgings.

The floors (fabricated from aluminum alloy 7178-T6 sheet, with portions of the sheet chemically milled to reduce thickness) of the fuel tanks in two aircraft failed almost identically after 1076 and 1323 h of service, respectively. Failure in both tanks occurred in the rear chemically milled section of the floor. An alkaline etch-type cleaner was used on the panels before chemical milling and before painting. Various tests and measurements indicated that the aluminum alloy used for the fuel-tank floors conformed to the specifications for 7178-T6. Low power magnification, fractographs taken with a scanning electron, and optical microscopic examination of the milled sections revealed extensive pitting on both sides of the floors. Evidence found supports the conclusions that the floors failed by fatigue cracking that initiated near the center of the fuel-tank floor and ultimately propagated as rapid ductile-overload fractures. The fatigue cracks originated in pits on the fuel-cell side of the tank floors. The pits were attributed to attack caused by the alkaline-etch cleaning process. Recommendations included monitoring of the alkaline-etch cleaning to avoid the formation of pits and careful inspection following alkaline-etch cleaning, to be scheduled before release of the floor panels for painting.

Three examples of corrosion-fatigue cracking from the toes of substantial fillet welds applied to seal-leaking riveted seams in steam accumulators are described. In the first case, this practice resulted in a disastrous explosion; in the second, which involved two identical vessels at the same location, cracking in course of development was discovered during internal inspection. Microscope examination of several specimens cut to intersect a crack showed it to be typical of corrosion-fatigue; it was in the form of a broad fissure, contained oxide deposits, and the termination was blunt-ended. The two cases not only serve to illustrate the danger of applying fillet welds to seal the lap edges of riveted seams, but point to the inadvisability of employing riveted construction for vessels intended for service under conditions involving frequent pressure and thermal fluctuations, as it is extremely difficult to maintain the tightness of riveted seams under these conditions. Such vessels are now almost exclusively of all-welded construction

An intergranular stress-corrosion cracking failure of 304 stainless steel pipe in 2000 ppm B as H3BO3 + H2O at 100 deg C was investigated. Constant extension rate testing produced an intergranular type failure in material in air. Chemical analysis was performed on both the base metal and weld material, in addition to fractography, EPR testing and optical microscopy in discerning the mode of failure. Various effects of Cl-, O2 and MnS are discussed. Results indicated that the cause of failure was the severe sensitization coupled with probable contamination by S and possibly by Cl ions.

A straight-tube cooler type heat exchanger had been in service for about ten years serving a coal pulverizer in Georgia. Non-potable cooling water from a local lake passed through the inner surfaces of the copper tubing and was cooling the hot oil that surrounded the outer diametral surfaces. Several of the heat exchangers used in the same application at the plant had experienced a severe reduction in efficiency in the past few years. One heat exchanger reportedly experienced some form of leakage following discovery of oil contaminating the cooling water. This heat exchanger was the subject of a failure investigation to determine the cause and location of the leaks. Corrosion products primarily contained copper oxide, as would be expected from a copper tubing. The product also exhibited the presence of a significant amount of iron oxides. Metallographic cross sectioning of the tubes and microscopic analysis revealed several large and small well rounded corrosion pits present at the inner diametral surfaces. The cause of corrosion was attributed to corrosive waters that were not only corroding the copper, but were corroding steel pipes upstream from the tubing.

A metallurgical failure analysis was performed on pieces of the cracked vent header pipe from the Edwin I. Hatch Unit 2 Nuclear power plant. The analysis consisted of optical microscopy, chemical analysis, mechanical Charpy impact testing, and fractography. It was found that the material of the vent header met the mechanical and chemical properties of ASTM A516 Grade 70 carbon-manganese steel material and microstructures were consistent with this material. Fracture faces of the cracked pipe were predominantly brittle in appearance with no evidence of fatigue contribution. The NDTT (Nil ductility Transition Temperature) for this material was approximately -51 deg C (-60 deg F). The fact that the material's NDTT was significantly out of the normal operating range of the pipe suggested an impingement of low temperature nitrogen (caused by a faulty torus inerting system) induced a thermal shock in the pipe which, when cooled below its NDTT, cracked in a brittle manner.

The intermediate pressure (IP) turbine of a thermal generating station is driven by steam from the boiler's reheater. On one particular IP turbine, a thick deposit was found on the insides of the rotor blade shrouds in two instances two years apart. The source of the deposits was not known; bulk chemical analysis had simply shown that iron was a major component. Optical microscopy and electron microprobe analysis were used to identify the deposits. In the first instance, the deposit was found to be debris that was left in the reheater tubes during boiler modification and swept to the turbine by the steam. There were still some of these debris particles present when the incident two years later was investigated but generally the second deposit was found to be of two layer oxide particles which were shown to have spalled from 2-14% chromium reheater tube surfaces.

An unusual type of defect was discovered during hydraulic testing of a water-tube boiler after repairs to the superheater tubes following erosion from soot-blowers. When the pressure reached 700 psi, slight leakage was found to be taking place from one of the superheater tubes in a region where there appeared to be a split, approximately 8 in. long. What was thought to be a split was actually a pronounced fold. Microscopic examination showed that a corrosion-fatigue fissure had developed from one of the inside corners of the fold, presumably as a result of the fluctuating bending stresses to which this portion of the tube would be subjected because of the discontinuity in the tube wall. It was from this fissure that the leakage occurred. It was evident that the defect developed during the manufacture of the tube, probably in the course of a drawing or rolling operation without an internal plug. The diam of this portion of the tube was reduced by local collapse and folding of the section rather than by longitudinal extension of the tube itself.

Work rolls made of indefinite chill double-poured (ICDP) iron are commonly used in the finishing trains of hot-strip mills (HSMs). In actual service, spalling, apart from other surface degeneration modes, constitutes a major mechanism of premature roll failures. Although spalling can be a culmination of roll material quality and/or mill abuse, the microstructure of a broken roll can often unveil intrinsic inadequacies in roll material quality that possibly accentuate failure. This is particularly relevant in circumstances when rolls, despite operation under similar mill environment, exhibit variations in roll life. The paper provides an insight into the microstructural characteristics of spalled ICED HSM work rolls, which underwent failure under similar mill operating environment in an integrated steel plant under the Steel Authority of India Limited. Microstructural features influencing ICDP roll quality, viz. characteristics of graphite, carbides, martensite, etc., have been extensively studied through optical microscopy, quantitative image analysis (QIA), and electron-probe microanalysis (EPMA). These are discussed in the context of spalling propensity and roll life.

An extensive metallurgical investigation was carried out on samples of a failed roller bearing from the support and tilting system of a basic oxygen furnace converter used in the steel melting shop of an integrated steel plant. The converter bearing was fabricated from low-carbon, carburizing grade steel and had failed in service within a year of fitting to a repaired shaft. Microscopic observations of both the broken roller and inner-race samples revealed subsurface cracking and preponderance of brittle oxide and other macroinclusions. Electron probe microanalysis studies confirmed that the brittle oxides that formed stringers were alumina, and the other macroinclusions were complex silicates. Both the alumina and silicate inclusions were deleterious to contact-fatigue properties. Microstructurally, the carburized regions of the broken roller and of inner-race samples contained high-carbon tempered martensite. Microhardness measurements revealed that. Although the core hardness of the roller and the inner-race samples were similar, the surface hardness of the roller was approximately 8.5 HRC units harder than that of the inner-race. SEM observations of the roller fracture surface revealed striations indicative of fatigue, and EDS analyses corroborated a high incidence of silicate inclusions at crack sites. The study suggests that the failure of the bearing occurred because the hardness difference between the roller bearing and the inner-race surfaces resulted in wear of the inner-race. The wear led to shaft misalignment and play during service. The misalignment, coupled with the presence of inclusions, caused fatigue failure of the roller bearing.

Although a precise understanding of roll failure genesis is complex, the microstructure of a broken roll can often unravel intrinsic deficiencies in material quality responsible for its failure. This is especially relevant in circumstances when, even under a similar mill-operating environment, the failure involves a particular roll or a specific batch of rolls. This paper provides a microstructural insight into the cause of premature breakage of a second-intermediate Sendzimir mill drive roll used at a stainless steel sheet rolling plant under the Steel Authority of India Limited. Microstructural issues influencing roll quality, such as characteristics of carbides, tempered martensite, retained austenite, etc., have been extensively studied through optical and scanning electron microscopy, electron-probe microanalysis, image analysis, and x-ray diffractometry. These are discussed to elucidate specific microstructural inadequacies that accentuated the failure. The study reveals that even through retained austenite content is low (6.29 vol%) and martensite is non-acicular, the roll breakage is a consequence of intergranular cracking caused by improper carbide morphology and distribution.

Three samples from a ruptured 316 stainless steel tube were examined. The tube, 114 mm OD, wall thickness 8.00 mm, with 13 mm thick 321 stainless steel fins welded to the outer surface of the tube, was part of a heater through which sour gas, containing methane plus H2S and CO, passed at 1150 psig. The sour gas was heated to 600 deg F by burners playing on the outside of the tube burning “sweet” gas plus air. The inner and outer surfaces of all samples showed evidence of corrosive attack. Electron probe microanalysis showed the corrosion products contained sulfur with iron, together with nickel to a lesser extent. Local thinning, cavitation, and ductile deformation markings associated with the unmatched sample taken from the center of the fire showed the tube ruptured as a result of overheating. Overheating while the temperature recorder was off the chart caused severe loss of tube strength, resulting in ductile rupture. The minimum overheating temperature could be deduced at around 1200 deg F due to the presence of a eutectic observed metallographically within the surface corrosion products.

An electron probe microanalyzer was applied to the study of service failures (due to severe heating) of aluminum wire connections in residential electrical circuits. Perturbed regions in which the composition underwent a change during the failure were revealed by optical and scanning electron microscopy of the contacts. A sequence of iron-aluminum compositions that shift from the pure aluminum of the wire to the nearly pure iron of the screw was revealed by analyses of two distinct layers formed on the aluminum/iron region. The compositions were found to correspond to specific intermetallic compounds found in the aluminum-iron phase diagram. Similar compositional variations were noted at the aluminum/brass interface. It was concluded that the failure of the electrical junction due to extreme heating was related to the formation of intermetallic compounds at the current carrying interfaces. These intermetallics were established to have a high resistance causing significant resistive heating.

A nickel alloy cylinder plated with chromium along its inner liner, installed in a commercial ice cream freezer, showed gray discoloration along its OD surface. The discolored parts exhibited significantly reduced cooling efficiency as compared with new cylinders. During operation, the OD of the cylinder was exposed to liquid ammonia refrigerant containing lubricant from the compressor. The lubricant (mineral oil) was intended to separate from the ammonia and be recirculated through the compressor. Nondestructive portable optical microscopy, XRF, EDS, and XPS analyses showed that the discoloration on the cylinder was associated with metal oxidation products coated with a thin oil film. One of the recommendations was to plate the OD of the cylinder with hard chromium to increase its resistance to erosion. Another recommendation was to reduce the amounts of water contamination in the refrigerant.

A failed laser mirror and another complete mirror of the same construction were analyzed. The laser mirror consisted of three layers of material brazed together to form channels through which the cooling water flows. Samples were analyzed with light optical and scanning electron microscopy. The corrosion product contained molybdenum and copper with a trace of gold. The base material was analyzed as molybdenum with negligible alloying additions. The primary mode of corrosion attack on the base material appeared to be intergranular, although uniform corrosion was evident also. It was concluded that corrosion attack sufficiently weakened the base material and the brazed joints, allowing catastrophic failure of the mirror due to the pressure of the cooling water. It was recommended that the mirrors be cleaned of all corrosion products present as a result of past service conditions and proof tested. It was recommended that the water system consisting of deionized water and formaldehyde be replaced with water having a low oxygen content and a cathodic inhibitor (oxygen scavenger).

A 1-in. diam pump spindle fractured within the length covered by the boss of the impeller which was attached to the spindle by means of an axial screw. The pump had been in use in a chemical plant handling mixtures of organic liquids and dilute sulfuric acid having a pH value of 2 to 4 at temperatures of 80 to 90 deg C (176 to 194 deg F). The fracture was unusual in that it was of a fibrous nature, the fibers-which were orientated radially-were readily detachable. The surface of the spindle adjacent to the fracture had an etched appearance and the mode of cracking in this region suggested that failure resulted from an intergranular attack. Subsequent microscope examination confirmed the generally intergranular mode of failure. A macro-etched section near the fracture revealed a radial arrangement of columnar crystals, indicating that the spindle was a cast and not a wrought product as had been presumed. Spectroscope examination showed this particular composition (Fe-23Cr-18Ni-1.8Mo-1.2Si) did not conform to a standard specification and is apparently a proprietary alloy. It was evident that the particular mode of failure was related to the inherent structure of the material.

A number of machined end frame steel forgings made of Cr-Si-Mn alloy showed tiny cracks during magnetic particle inspection after heat treatment. The cracks were mostly confined to base edges and fillet radius. No significant abnormality was observed in chemical composition and microstructure. SEM, optical microscopy, and gas analysis revealed that the subsurface discontinuous cracks at the bore edges and in the fillet radius of the heat-treated end frame component had occurred due to hydrogen embrittlement, and not because of faulty heat treatment. This conclusion was supported by the presence of cracklike indications in machined bore surface of the annealed part.

While a chain sling was being used to lift a casting one of the links ruptured. The sling, reputed to be of the electrically-welded steel type, was at least eight years old and had been overhauled several times during its working life. Examination showed the links were scarf-welded. Furthermore, the welds were at the ends and not at the sides as is usual in the case of electrically-welded chains. A transverse section from one side of a link was examined microscopically. This showed the material to be wrought-iron of satisfactory quality. It was concluded this chain sling had been made from wrought-iron, forge welded in the usual manner, and that it was not electrically-welded steel as had been supposed. Failure was attributed to embrittlement in service of the surface material of the links. If it had been realized that the sling was made from wrought-iron then it would doubtless have been subjected to periodical annealing in accordance with Statutory requirements, which would have restored the ductility of the surface material.