The scanning electron microscope (SEM) is one of the most versatile instruments for investigating the microscopic features of most solid materials. The SEM provides the user with an unparalleled ability to observe and quantify the surface of a sample. This article discusses the development of SEM technology and operating principles of basic systems of SEM. The basic systems covered include the electron optical column, signal detection and display equipment, and the vacuum system. The processes involved in the preparation of samples for observation using an SEM are described, and the application of SEM in fractography is discussed. The article covers the failure mechanisms of ductile failure, brittle failure, mixed-mode failure, and fatigue failure. Lastly, image dependence on microscope type and operating parameters is also discussed.

Metallographic examination is one of the most important procedures used by metallurgists in failure analysis. Typically, the light microscope (LM) is used to assess the nature of the material microstructure and its influence on the failure mechanism. Microstructural examination can be performed with the scanning electron microscope (SEM) over the same magnification range as the LM, but examination with the latter is more efficient. This article describes the major operations in the preparation of metallographic specimens, namely sectioning, mounting, grinding, polishing, and etching. The influence of microstructures on the failure of a material is discussed and examples of such work are given to illustrate the value of light microscopy. In addition, information on heat-treatment-related failures, fabrication-/machining-related failures, and service failures is provided, with examples created using light microscopy.

One of the two AISI 4340 steel pitch horn bolts from the main rotor hub assembly failed while in service. Optical microscope revealed evidence of corrosion pitting in regions adjacent to the fracture. Fractographic examination utilizing a scanning electron microscope revealed multiple crack origins which assumed a “thumbnail” shape and displayed surface morphologies which resulted from intergranular decohesion. Many of the crack sites initiated from corrosion pits. Energy dispersive spectroscope performed on areas within the crack initiation site showed the presence of chlorides. The failure was attributed to stress-corrosion cracking. Short- and long-term recommendations to prevent future failures are given.

Following the crash of a Mirage III-0 aircraft (apparently caused by engine failure), a small crack was detected in a bolt hole in the wing main spar (AU4SG aluminum alloy). Because this area was considered to be critical to aircraft safety and similar cracking was found in other spars in service, the Royal Australian Air Force requested that the crack growth rate during service be determined. The loading history of the aircraft was made available in the form of flight by-flight records of the counts from the vertical accelerometer sensors fitted to the airframe and a series of “overstress” events recorded during the life of the aircraft. The bolt hole was examined by eddy current testing, visual examination, high-powered light microscope, and scanning electron microscope. Simulation tests were also conducted. The use of simulation specimens permitted actual crack growth rate data to be determined for the configuration of interest.

An investigation of a damaged crankshaft from a horizontal, six-cylinder, in-line diesel engine of a public bus was conducted after several failure cases were reported by the bus company. All crankshafts were made from forged and nitrided steel. Each crankshaft was sent for grinding, after a life of approximately 300,000 km of service, as requested by the engine manufacturer. After grinding and assembling in the engine, some crankshafts lasted barely 15,000 km before serious fractures took place. Few other crankshafts demonstrated higher lives. Several vital components were damaged as a result of crankshaft failures. It was then decided to send the crankshafts for laboratory investigation to determine the cause of failure. The depth of the nitrided layer near fracture locations in the crankshaft, particularly at the fillet region where cracks were initiated, was determined by scanning electron microscope (SEM) equipped with electron-dispersive X-ray analysis (EDAX). Microhardness gradient through the nitrided layer close to fracture, surface hardness, and macrohardness at the journals were all measured. Fractographic analysis indicated that fatigue was the dominant mechanism of failure of the crankshaft. The partial absence of the nitrided layer in the fillet region, due to over-grinding, caused a decrease in the fatigue strength which, in turn, led to crack initiation and propagation, and eventually premature fracture. Signs of crankshaft misalignment during installation were also suspected as a possible cause of failure. In order to prevent fillet fatigue failure, final grinding should be done carefully and the grinding amount must be controlled to avoid substantial removal of the nitrided layer. Crankshaft alignment during assembly and proper bearing selection should be done carefully.

The quantitative characterization of fracture surface geometry, that is, quantitative fractography, can provide useful information regarding the microstructural features and failure mechanisms that govern material fracture. This article is devoted to the fractographic techniques that are based on fracture profilometry. This is followed by a section describing the methods based on scanning electron microscope fractography. The article also addresses procedures for three-dimensional fracture surface reconstruction. In each case, sufficient methodological details, governing equations, and practical examples are provided.

Visual examination, using the unaided eye or a low-power optical magnifier, is typically one of the first steps in a failure investigation. This article presents the guidelines for selecting samples for scanning electron microscope examination and optical metallography and for cleaning fracture surfaces. It discusses damage characterization of metals, covering various factors that influence the damage, namely stress, aggressive environment, temperature, and discontinuities.

X-ray spectroscopy is generally accepted as the most useful ancillary technique that can be added to any scanning electron microscope (SEM), even to the point of being considered a necessity by most operators. While “stand-alone” x-ray detection systems are used less frequently in failure analysis than the more exact instrumentation employed in SEMs, the technology is advancing and is worthy of note due to its capability for nondestructive analysis and application in the field. This article begins with information on the basis of the x-ray signal. This is followed by information on the operating principles and applications of detectors for x-ray spectroscopy, namely energy-dispersive spectrometers, wavelength-dispersive spectrometers, and handheld x-ray fluorescence systems. The processes involved in x-ray analysis in the SEM and handheld x-ray fluorescence analysis are then covered. The article ends with a discussion on the applications of x-ray spectroscopy in failure analysis.

Some corrosion processes in the presence of chlorides, for steel embedded in concrete, are described and illustrated with the aid of scanning electron microscope EDXA data. Observations made of failure surfaces of reinforcements removed from the concrete beams after being subjected to sinusoidal load fluctuations at 6.7 Hz in air, 3% NaCl solution, and natural sea water are described. Reinforcement types studied included: hot-rolled mild steel bar, hot-rolled alloyed high strength bar, cold-worked high strength bar, galvanized bar of all these three types, nickel-clad bar and epoxy-coated bar.

Stress-corrosion cracking of low-alloy steel turbine discs has emerged as a generic concern in nuclear generating stations. An investigation that made extensive use of field metallographic techniques to examine suspected cracking in such a component is described. The crack position, and its relationship to surface topographic features, were examined and recorded by magnetic rubber and high-resolution dental rubber replicating materials. Corrosion deposits on keyway surfaces and within the crack were collected with acetate foil replicas applied and then stripped from the keyway surfaces. Microstructural details were revealed by the use of field metallographic preparation techniques and replicated by acetate foil for examination with optical and scanning electron microscopes. It was possible by these techniques to establish the cracking mechanism as stress corrosion possibly related to chloride or sulphate ion steam contaminants. Subsequent sectioning and conventional metallography confirmed both the validity of the conclusions and the replication techniques.

Corrosion in a closed-loop cooling water system constructed of austenitic stainless steel occurred during an extended lay up of the system with biologically contaminated water. The characteristics of the failure were those of microbiologically influenced corrosion (MIC). The corrosion occurred at welds and consisted of large subsurface void formations with pinhole penetrations of the surfaces. Corrosive attack initiated in the heat affected zones of the welds, usually immediately adjacent to fusion lines. Stepwise grinding, polishing, and etching through the affected areas revealed that voids generally grew in the wrought material by uniform general corrosion. Tunneling or worm-holing was also observed, whereby void extension occurred by initiating daughter voids probably at flaws or other inhomogeneities. Selective attack occurred within the fusion zone, i.e., within the cast two-phase structure of the weld filler itself. The result was a void wall which consisted of a rough and porous ferritic material, a consequence of preferential attack of the austenitic phase and slightly lower rate of corrosive attack of the ferrite phase. The three-dimensional spongy surface was studied optically and with the scanning electron microscope.

Metallography is an important component of failure analysis. In the case of a liquid metal embrittlement (LME) failure it is usually conclusive if a third phase constituent can be formed inside of the cracks after failure. In the case where it is necessary to characterize the third phase material, one can use various x-ray spectrographic techniques in conjunction with a scanning electron microscope (SEM). This study describes those metallographic and SEM analysis techniques for determining the mode of failure for a locomotive traction motor by LME.

The cause of low fatigue life measurements obtained during routine fatigue testing of IMI 550 titanium alloy compressor blades used in the first stage of the high-pressure compressor of an aeroengine was investigated. The origin of the fatigue cracks was associated with a spherical bead of metal sticking to the blade surface in each case. Scanning electron microscope revealed that the cracks initiated at the point of contact of the bead with the blade surface. Energy-dispersive X-ray analysis indicated that the bead composition was the same as that of the blade. Detailed investigation revealed that fused material from the blade had been thrown onto the cold blade surface during a grinding operation to remove the targeting bosses from the forgings, thereby causing local embrittlement. It was recommended that extreme care be taken during grinding operations to prevent the hot, fused particles from striking the blade surface.

A coupling in a line-shaft vertical turbine pump installed in a dam foundation fractured after a very short time. The coupling material was ASTM A582 416 martensitic stainless steel. Visual, macrofractographic, and scanning electron microscopic examination of the coupling showed that the fracture was brittle and was initiated by an intergranular cracking mechanism. The mode of fracture outside the crack initiation zone was transgranular cleavage. No indication of fatigue was found. The failure was attributed to improper heat treatment during manufacture, which resulted in a brittle microstructure susceptible to corrosion. The crack initiated either by stress-corrosion or hydrogen cracking. It was recommended that the couplings in the system be examined for surface cracking and, if present, corrective measures be taken.

An AMS 4126 (7075-T6) aluminum alloy impeller from a radial inflow turbine fractured during commissioning. Initial examination showed that two adjacent vanes had fractured through airfoils in the vicinity of the vane leading edges, and one vane fractured through an airfoil near the hub in the vicinity of the vane trailing edge. Some remaining vanes exhibited radial and transverse cracks in similar locations. Binocular and scanning electron microscope examinations showed that the cracks had been caused by high-cycle fatigue and had progressed from multiple origins on the vane surface. Structural analysis indicated that the fatigue loading probably had been caused by forced excitation, resulting in the impeller vibrating at its resonant frequency. It was recommended that the impeller design, control systems, and material of construction be changed.

An investigation of a Stirling engine after an aborted test run revealed that the regenerator screens had suffered substantial damage. During the run, the individual screens oscillated as the helium working fluid was shuttled through the regenerator. In localized areas, the 41 mu m (1600 mu in.) diam type 304 stainless steel wire screening had been torn and pieces were missing. Scanning electron microscope revealed that the fracture had occurred at wire crossover locations by a fatigue mechanism. The problem was solved by sintering the individual screens into a single unit.

Following an accident in which a light pickup truck left the road and overturned, one of the rear axles, made of approximately 0.30C steel, was found to be fractured adjacent to the bearing lock nut. A keyway was present in the failed area, as were threads for the lock nut. Fracture surfaces of the failed axle and exemplar fractures obtained from simulation tests were studied using scanning electron microscope. The examination showed that the outer perimeter fracture in the axle was very flat and composed of cleavage and that the interior portion was composed of both cleavage and dimples. No evidence of prior cracking was found. The exemplar specimens from the simulation impact testing failed in a manner consistent with that observed in the axle. The examination confirmed that the failure was a one-time impact overload fracture and not the result of any prior crack in the material, indicating that the axle failure did not initiate the accident.

Several type 316L stainless steel wires in an electrostatic precipitator at a paper plant fractured in an unexpectedly short time. Failed wires were examined using optical and scanning electron microscope, and hardness tests were conducted. Fractography clearly established that fracture was caused by fatigue originating at corrosion pits on the surface of the wire. It was recommended that higher-molybdenum steel in the annealed condition be used to combat pitting corrosion.

Numerous flaws were detected in a steam turbine rotor during a scheduled inspection and maintenance outage. A fracture-mechanics-based analysis of the flaws showed that the rotor could not be safely returned to service. Material, samples from the bore were analyzed to evaluate the actual mechanical properties and to determine the metallurgical cause of the observed indications. Samples were examined in a scanning electron microscope and subjected to chemical analysis and several mechanical property tests, including tensile, Charpy V-notch impact, and fracture toughness. The material was found to be a typical Cr-Mo-V steel, and it met the property requirements. No evidence of temper embrittlement was found. The analyses showed that the observed flaws were present in the original forging and attributed them to lack of ingot consolidation. A series of actions, including overboring of the rotor to remove indications close to the surface and revision of starting procedures, were implemented to extend the remaining life of the rotor and ensure its fitness for continued service.

A failed crosshead of an industrial compressor was examined using optical and scanning electron microscope. The crosshead was an ASTM A148 grade 105-85 steel casting. On the basis of the observations reported and available background information, it was concluded that the failure began with the initiation of cracks at slag inclusions and sharp fillets in weld-repair areas in the casting. The weld-repair procedures were unsatisfactory. The cracks propagated in a fatigue mode. he casting quality was judged unacceptable because of the presence of excessive shrinkage porosity. It was recommended that crosshead castings be properly inspected before machining. Revision of foundry practice to reduce or eliminate porosity was also recommended.