A steel spring used in an automotive application suddenly began to fail in the field, although “nothing had changed” in the fabrication process. Fatigue tests using springs fabricated prior to field failures lasted 500,000 cycles to failure, whereas fatigue tests performed on springs fabricated after field failures lasted only 50,000 cycles to failure. It was discovered that the percent coverage of shot peening prior and subsequent to the increase in failure incidence was much less than 100%, with a shot peening time of 12 min. The residual-stress state of “as fabricated” springs in three conditions were evaluated using XRD: springs manufactured prior to failure incidence increase, 12 min peen; springs manufactured following failure incidence increase, 12 min peen; and 60 min peen. The conclusion was that the failure occurred because low peening time significantly decreased the compressive residual-stress levels in the springs. Recommendation was made to increase the time the spring was shot peened from 12 to 60 min.

A large number of electropolished copper parts showed evidence of discoloration (tinting) after electropolishing. Because these parts are used in a high-vacuum application, even trace amounts of organic materials would be problematic. Scanning electron microscopy of nondiscolored and discolored areas both showed trace amounts of residue in the form of adherent deposits. EDS, FTIR spectroscopy, XPS, and secondary ion mass spectroscopy (SIMS) analyses indicated that the discoloration to the copper components was due to the development of CuO at localized regions. It was recommended that process changes be made to completely remove residual processing fluids from the part surfaces before electropolishing. The use of more aggressive detergents was suggested, and it was recommended also that a filtering and recirculating system be considered for use in the cleaning and electropolishing tanks.

Helicopter rotor blade components that included the horizontal hinge pin, the associated nut, and the locking washer were examined. Visual examination of the submitted parts revealed that the hinge pin, fabricated from 4340 steel, was broken and that the fracture face showed a flat beach mark pattern indicative of a preexisting crack. The threaded area of the pin had an embedded thread that did not appear to come from the pin. A chemical analysis was conducted on the embedded thread and on an associated attachment to determine the origin of the thread. Analysis showed that the thread and nut were 4140 steel. Scanning electron fractographic examination of the fracture initiation site strongly suggested that the fracture progressed by fatigue. It was concluded that the failure of the horizontal hinge pin initiated at areas of localized corrosion pits. The pits in turn initiated fatigue cracks, resulting in a failure mode of corrosion fatigue. It was recommended that all of the horizontal hinge pins be inspected. Those pins determined to be satisfactory for further use should be stripped of cadmium, shot peened, and coated with cadmium to a minimum thickness of 0.0127 mm (0.0005 in.).

Inspections and microstructural analysis revealed intergranular corrosion of 6061-T6 aluminum alloy aircraft fuel line beneath ferrules. The cause of the corrosion was traced to the fuel line marking process, which involved electrolytic labeling. Although subsequent rinsing of the fuel lines washed off most of the electrolyte, some was trapped between the 6061-T6 tubing and the ferrule. This condition made intergranular corrosion of the fuel lines inevitable. The attack caused grains to become dislodged, giving the appearance of pitting. Corrosion penetrated approximately 0.13 mm (0.005 in.) into the tubing. Experiments indicated that the corrosion products were inactive. It was recommended that another marking process be used that does not involve corrosive materials. The prevention of electrolyte from being trapped between the tubing and ferrules by using a MIL-S-8802 sealant was recommended.

A T-bolt was part of the coupling for a bleed air duct of a jet engine on a transport plane. Specifications required that the 4.8 mm diam component be made of AISI type 431 stainless steel and heat treated to 44 HRC. The operating temperature of the duct is 425 to 540 deg C (800 to 1000 deg F), but that of the bolt is lower. The T-bolt broke after three years of service. The expected service life was equal to that of the aircraft. It was found that the bolt broke as a result of SCC. Thermal stresses were induced into the bolt by intermittent operation of the jet engine. Mechanical stresses were induced by tightening of the clamp around the duct, which in effect acted to straighten the bolt. The action of these stresses on the carbides that precipitated in the grain boundaries resulted in fracture of the bolt. Due to the operating temperatures of the duct near the bolt, the material was changed to A-286, which is less susceptible to carbide precipitation. The bolt is strengthened by shot peening and rolling the threads after heat treatment. Avoiding temperatures in the sensitizing range is desirable, but difficult to ensure because of the application.

During a routine shear-pin check, the end lug on the barrel of the forward canopy actuator on a naval aircraft was found to have fractured. The lug was forged from aluminum alloy 2014-T6. Investigation (visual inspection, 2x views, and 140X micrographs etched with Keller's reagent) supported the conclusion that the cause of failure was SCC resulting from exposure to a marine environment. The fracture occurred in normal operation at a point where damage from pitting and intergranular corrosion acted as a stress raiser, not because of overload. The pitting and intergranular attack on the lug were evidence that the surface protection of the part had been inadequate as manufactured or had been damaged in service and not properly repaired in routine maintenance. Recommendations included anodizing the lug and barrel in sulfuric acid and giving them a dichromate sealing treatment, followed by application of a coat of paint primer. During routine maintenance checks, a careful examination was suggested to look for damage to the protective coating, and any necessary repairs should be made by cleaning, priming, and painting. Severely corroded parts should be removed from service.

New aircraft wing panels extruded from 7075-T6 aluminum exhibited an unusual pattern of circular black interrupted lines, which could not be removed by scouring or light sanding. The panels, subsequent to profiling and machining, were required to be penetrated inspected, shot peened, H2SO4 anodized, and coated with MIL-C-27725 integral fuel tank coating on the rib side. Scanning electron microscopy and microprobe analysis (both conventional energy-dispersive and Auger analyzers) showed that the anodic coating was applied over an improperly cleaned and contaminated surface. The expanding corrosion product had cracked and, in some places, had flaked away the anodized coating. The corrodent had penetrated the base aluminum in the form of subsurface intergranular attack to a depth of 0.035 mm (0.0014 in.). It was recommended that a vapor degreaser be used during cleaning prior to anodizing. A hot inhibited alkaline cleaner was also recommended during cleaning prior to anodizing. The panels should be dichromate sealed after anodizing. The use of deionized water was also recommended during the dichromate sealing operation. In addition, the use of an epoxy primer prior to shipment of the panels was endorsed. Most importantly, surveillance of the anodizing process itself was emphasized.

This paper summarizes the results of a failure analysis investigation of a fractured main support bridge made of 7075 aluminum alloy from an army helicopter. The part, manufactured by “Contractor IT,” failed component fatigue testing while those of the original equipment manufacturer (OEM) passed. Metallurgical data collected during this investigation indicated that the difference in fatigue life between the components fabricated by IT and by OEM may be attributable to a difference in dimensions at the web where fatigue crack initiation occurred. The webs of the two OEM parts examined had cross-sectional thicknesses significantly larger than the web cross-sectional thicknesses of the IT components. Recommendations included changing the web reference dimension of 0.38 in. to include a tolerance range based upon a fracture mechanics model. Also, the shot peening process should be controlled especially at the critical areas of the web, to assure complete coverage and proper compressive residual stresses.

An articulated rod (made from 4337 steel (AMS 6412) forging, quenched and tempered to 36 to 40 HRC) used in an overhauled aircraft engine was fractured after being in operation for 138 h. Visual examination revealed that the rod was broken into two pieces 6.4 cm from the center of the piston-pin-bushing bore. The fracture was nucleated at an electroetched numeral 5 on one of the flange surfaces. A notch, caused by arc erosion during electroetching, was revealed by metallographic examination of a polished-and-etched section through the fracture origin. A remelted zone and a layer of untempered martensite constituted the microstructure of the metal at the origin. Small cracks, caused by the high temperatures developed during electro-etching, were observed in the remelted area. It was concluded that fatigue fracture of the rod was caused by the notch resulting from electroetching and thus electroetched marking of the articulated rods was discontinued as a corrective measure.

The spindle of a helicopter-rotor blade fractured after 7383 h of flight service. At every overhaul (the spindle that failed was overhauled six times and reworked twice), any spindle that showed wear was reworked by grinding the shank to 0.1 mm (0.004 in.) under the finished diam. The spindle was then shot peened with S170 shot to an Almen intensity of 0.010 to 0.012 A. Following shot peening, the shank was nickel sulfamate plated to 0.05 mm (0.002 in.) over the finished diam, ground to finished size, and cadmium plated. Visual and stereomicroscopic exam showed faint grinding marks and circumferential grooves on the surface near the fillet at the junction of the shank and fork, which should have been peened over and covered with peening dimples. Evidence found supports the conclusions that the spindle failed in fatigue that originated near the junction of the shank and fork. The nonuniformity of the shot-peened effect on the shank and fillet portions of the spindle resulted from incomplete peeing. The fracture was of the low-stress high-cycle type, initiated by stresses well below the gross yield strength and propagated by thousands of load cycles. No recommendations were made.

Brittle intergranular fracture, typical of a hydrogen-induced delayed failure, caused the failure of an AISI 4340 Cr-Mo-Ni landing gear beam. Corrosion resulting from protective coating damage released nascent hydrogen, which diffused into the steel under the influence of sustained tensile stresses. A second factor was a cluster of non-metallic inclusions which had ‘tributary’ cracks starting from them. Also, eyebolts broke when used to lift a light aircraft (about 7000 lb.). The bolt failure was a brittle intergranular fracture, very likely due to a hydrogen-induced delayed failure mechanism. As for the factors involved, cadmium plating, acid pickling, and steelmaking processes introduce hydrogen on part surfaces. As a second contributing factor, both bolts were 10 Rc points higher in hardness than specified (25 Rc), lessening ductility and notch toughness. A third factor was inadequate procedure, which resulted in bending moments being applied to the bolt threads.

Two freshwater tanks (0.81 mm (0.032 in) thick, type 321 stainless steel) were removed from aircraft service because of leakage due to pitting and rusting on the bottoms of the tanks. One tank had been in service for 321 h, the other for 10 h. There had been departures from the specified procedure for chemical cleaning of the tanks in preparation for potable water storage. The sodium hypochlorite sterilizing solution used was three times the prescribed strength, and the process exposed the bottom of the tanks to hypochlorite solution that had collected near the outlet. Investigation (visual inspection, 95x unetched images, chemical testing with a 5% salt spray, chemical testing with sodium hypochlorite at three strength levels, samples were also pickled in an aqueous solution containing 15 vol% concentrated nitric acid (HNO3) and 3 vol% concentrated hydrofluoric acid (HF) and were then immersed in the three sodium hypochlorite solutions for several days) supported the conclusion that failure of the stainless steel tanks by chloride-induced pitting resulted from using an overly strong hypochlorite solution for sterilization and neglecting to rinse the tanks promptly afterward. Recommendations included revising directions for sterilization and rinsing.

A power plant using two steam generators (vertical U-tube and shell heat exchangers, approximately 21 m (68 ft) high with a steam drum diameter of 6 m (20 ft)) experienced a steam generator tube rupture. Each steam generator contained 11,012 Inconel alloy 600 (nickel-base alloy) tubes measuring 19 mm OD, nominal wall thickness of 1.0 mm (0.042 in.), and average length of 18 m (57.75 ft). The original operating temperature of the reactor coolant was 328 deg C (621 deg F). A tube removal effort was conducted following the tube rupture event. Investigation (visual inspection, SEM fractographs, and micrographs) showed evidence of IGSCC initiating at the OD and IGA under ridgelike deposits that were analyzed and found to be slightly alkaline to very alkaline (caustic) in nature. Crack oxide analysis indicated sulfate levels in excess of expected values. The analysis supported the conclusion that that the deposits formed at locations that experienced steam blanketing or dryout at the higher levels of the steam generators. Recommendations included steam generator water-chemistry controls, chemical cleaning, and reduction of the primary reactor coolant system temperature.

The accident at Three Mile Island Unit No. 2 on 28 March 1979 was the worst nuclear accident in US history. By Jan 1990, it was possible to electrochemically machine coupons from the lower head using a specially designed tool. The specimens contained the ER308L stainless steel cladding and the A533 Grade B plate material to a depth of about mid-wall. The microstructures of these specimens were compared to that of specimens cut from the Midland, Michigan reactor vessel, made from the same grade and thickness but never placed in service. These specimens were subjected to known thermal treatments between 800 and 1100 deg C for periods of 1 to 100 min. Microstructural parameters in the control specimens and in those from TMI-2 were quantified. Selective etchants were used to better discriminate desired microstructural features, particularly in the cladding. This report is a progress report on the quantification of changes in both the degree of carbide precipitation and delta ferrite content and shape in the cladding as a function of temperature and time to refine the estimates of the maximum temperatures experienced.

A 321 stainless steel radar coolant-system assembly fabricated by torch brazing with AWS type 3A flux, failed at the brazed joint when subjected to mild handling before installation, after being stored for about two years. It was revealed by visual examination of the failed braze that the filler metal had not covered all mating surfaces. Lack of a metallurgical bond between the brazing alloy and stainless steel and instead mechanical bonding of the filler metal to an oxide layer on the stainless steel surface was revealed by examination of the broken joint at the cup. It was indicated by the thickness of the oxide layer that the steel surface was not protected from oxidation by the flux during torch heating. It was concluded that the failure was caused by lack of a metallurgical bond between the brazing alloy and the stainless steel. Components made of 347 stainless steel (better brazeability) brazed with a larger torch tip (wider heat distribution) and AWS type 3B flux (better filler-metal flow) were recommended for radar coolant-system assembly.

Nodular cast iron crankshafts and their main-bearing inserts were causing premature failures in engines within the first 1600 km (1000 mi) of operation. The failures were indicated by internal noise, operation at low pressure, and total seizing. Concurrent with the incidence of engine field failures was a manufacturing problem: the inability to maintain a similar microfinish on the cope and drag sides of a cast main-bearing journal. Investigation supported the conclusion that the root cause of the failure was carbon flotation due to the crankshafts involved in the failures showing a higher-than-normal carbon content and/or carbon equivalent. Larger and more numerous cope side graphite nodules broke open, causing ferrite caps or burrs. They then became the mechanism of failure by breaking down the oil film and eroding the beating material. A byproduct was heat, which assisted the failure. Recommendations included establishing closer control of chemical composition and foundry casting practices to alleviate the carbon-flotation form of segregation. Additionally, some nonmetallurgical practices in journal-finishing techniques were suggested to ensure optimal surface finish.

Randomly selected dictating-machine drive mechanisms, which contained small ball bearings, were found to exhibit unacceptable fluctuations in drive output during the early stages of production. It was indicated that the bearing raceways were being true brinelled before or during installation of the bearings. The preinstallation practices and the procedures for installing the bearings were carefully studied. It was revealed that during one preinstallation step, the lubricant applied by the bearing manufacturer was removed and the bearing was relubricated with another type of lubricant prior to which the bearings were ultrasonically cleaned in trichloroethylene to ensure extreme cleanness. Equally spaced indentations resembling true brinelling were revealed by careful examination of the bearing raceways. It was concluded that the ultrasonic energy transmitted to the balls brinelled the raceways enough to cause fluctuations in machine output. Solvent-vapor cleaning was employed as a corrective technique for removing bearing lubricant.

Several diesel-engine rocker levers (malleable iron similar to ASTM A 602, grade M7002) failed at low hours in overspeed, over-fuel, highly loaded developmental engine tests. Identical rocker levers had performed acceptably in normal engine tests. The rocker levers were failing through the radius of an adjusting screw arm. The typical fracture face exhibited two distinct modes of crack propagation: the upper portion indicated overload at final fracture, whereas the majority of the fracture suggested a fatigue fracture. Investigation (visual inspection, 1.5x/30x/60x magnification, and nital etched 300x magnification) supported the conclusion that the rocker levers failed in fatigue, with casting defects, or spiking, acting as stress raisers to initiate failures in highly loaded engine tests. Recommendations included shot peening of the levers as an interim measure to reduce the possibility of failure and redesign to increase the cross-sectional area of the levers.

Flow forming technology has emerged as a promising, economical metal forming technology due to its ability to provide high strength, high precision, thin walled tubes with excellent surface finish. This paper presents experimental observations of defects developed during flow forming of high strength SAE 4130 steel tubes. The major defects observed are fish scaling, premature burst, diametral growth, microcracks, and macrocracks. This paper analyzes the defects and arrives at the causative factors contributing to the various failure modes.

Presence of transverse marks which were remnant of grinding was indicated in a failed valve spring made from ground rod. The shot-peening pattern was light at this location. A transverse crack was found to grow from one such mark under the influence of local stress fields until it was reoriented to the plane normal to the major tensile axis by sufficient loading. The shot-peening procedure was altered to create adequate surface compression at all stressed points on the springs.