Two tubular AISI 1025 steel posts (improved design) in a carrier vehicle failed by cracking at the radius of the flange after five weeks of service. The posts were two of four that supported the chassis of the vehicle high above the wheels. The original design involved a flat flange of low-carbon low-alloy steel that was welded to an AISI 1025 steel tube, and the improved design included placing the welded joint of the flange farther away from the flange fillet. Investigation (visual inspection and chemical analysis) supported the conclusion that the failures in the flanges of improved design were attributed to fatigue cracks initiating at the aluminum oxide inclusions in the flange fillet. Recommendations included retaining the improved design of the flange with the weld approximately 50 mm (2 in.) from the fillet, but changing the metal to a forging of AISI 4140 steel, oil quenched and tempered to a hardness of 241 to 285 HRB. Preheating to 370 deg C (700 deg F) before and during welding with AISI 4130 steel wire was specified. It was also recommended that the weld be subjected to magnetic-particle inspection and then stress relieved at 595 deg C (1100 deg F), followed by final machining.

A return bend (made from ASTM A213, grade T11, ferritic steel) from a triolefin-unit heater ruptured after two years in service. The unit operated at 2410 kPa, with a hydrocarbon feed stream (85% propylene) entering at 260 to 290 deg C and leaving at 425 to 480 deg C. The fracture was found to terminate at the welds that joined the bend to the pipeline. A high concentration of both small and large inclusions was exhibited by the metallographic examination of the steel near the fracture. Branched cracks similar to those produced by stress corrosion of steel were observed in a section through the fireside edge of the fracture surface. Scale was observed over most of the crack path which acted as a stress raiser. The effect of the oxide was magnified during thermal cycles because of differential thermal expansion, with the steel having a greater expansion coefficient than the scale. It was recommended that material that is intended for critical applications where failure cannot be tolerated should be non-destructively examined.

There was a large incidence of surface defects on the crank pins and journals and other areas of crank shafts of a high power automotive engine. The steel used was a Cr-Mo type of nitriding steel. Metallographic observations conclusively proved that the defective areas were entrapment of foreign bodies, resulting from steel making/deoxidizing/teeming stages. The occasionally globular nature of the foreign particles suggested these were formed at the liquid condition of the steel. The ratio of Mn-Si as seen on electron probe microanalysis also suggested the globules high in Mn content might have resulted in deoxidizing stage. Particularly the absence of Fe in some areas in the inclusion was indicative of precipitation deoxidation by ferromanganese/ferrosilicon. The defects apparently did not have time to coalesce and rise up to the top.

The 1040 steel crankshaft in a reciprocating engine cracked within one year of operation. The journals of the main and crankpin bearings were inspected by the magnetic-particle method. Three to six indications of 1.5 to 9.5 mm long discontinuities were observed in at least four of the main-bearing journals. A crack along the fillet, almost entirely through the web, was observed in one of the main-bearing journals. Numerous coarse segregates, identified as sulfide inclusions, were identified by macroetching the surface during metallographic examination of a section taken through the main-bearing journal at the primary crack. Fatigue cracking with low-stress high-cycle characteristics was disclosed during macroscopic examination of the crack surface. Sulfide inclusions, which acted as stress raisers, were found to be present in the region where cracking originated. As a corrective measure, ultrasonic inspection was used in addition to magnetic-particle inspection to detect discontinuities.

A 1050 steel crankshaft with 6.4 cm (2.5 in.) diam journals that measured 87 cm (34.25 in.) in length and weighed 31 kg (69 lb) fractured in service. The shaft had been quenched and tempered to a hardness of 19 to 26 HRC, then selectively hardened on the journals to a surface hardness of 40 to 46 HRC. Visual inspection and 100x micrographs showed the fracture surface as having a complex type of fatigue failure initiated from subsurface inclusions in the transition zone between the induction-hardened surface and the softer core. The fractured shaft was examined for chemical composition and hardness, both of which were found to be within prescribed limits. This evidence supports the conclusions that the failure was caused by fatigue cracks that initiated in an area having an excessive amount of inclusions. The inclusions were located in a transition zone, which is a region of high stress. No recommendations were made.

The AISI 1080 steel crankshaft of a large-capacity double-action stamping press broke in service and was repair welded. Shortly after the crankshaft was returned to service, the repair weld fractured. The repair-weld fracture was examined ultrasonically which revealed many internal reflectors, indicating the presence of slag inclusions and porosity. A low-carbon steel flux-cored filler metal was used in repair welding the crankshaft, without any preweld or postweld heating. This resulted in the formation of martensite in the HAZ. The repair weld failed by brittle fracture, which was attributed to the combination of weld porosity, many slag inclusions and the formation of brittle martensite in the HAZ. A new repair weld was made using an E312 stainless steel electrode, which provides a weld deposit that contains considerable ferrite to prevent hot cracking. Before welding, the crankshaft was preheated to a temperature above which martensite would form. After completion, the weld was covered with an asbestos blanket, and heating was continued for 24 h. During the next 24 h, the temperature was slowly lowered. The result was a crack-free weld.

Examples of metallic inclusions in steels of various types are presented. The structure of an inclusion in an annealed Fe-1C-1.5Cr steel consisted of ferrite with lamellar pearlite. The carbon content of the inclusion was therefore considerably lower than that of the chromium steel and was adapted to the latter by diffusion only at the periphery of the inclusion. In another section of a hardened piece of the same chromium steel, the steel in this case had a structure of martensite with hypereutectic carbide, while the inclusions consisted of a very fine laminated eutectoid of the lower pearlite range (Troostite). In a pipe of 18-8 austenitic stainless steel a weakly magnetizable spot of limited size was found. This inclusion too was probably more alloy-deficient than the austenitic steel, similar to the ones described above. All three cases were casting defects.

An automobile manufacturer rejected several 1035 steel stub axles because of what appeared to be short longitudinal cracks in the surfaces of the pins. The cracks were found when six axles were examined for defects by magnetic-particle inspection. However, metallographic examination showed that these lines were not cracks but slag inclusions at and immediately below the surface. Analysis (visual inspection, metallographic examination, and 100x/500x magnetic-particle inspection) supported the conclusions that the inclusions consisted of pieces of fireclay from channel brick that were flushed into the ingot mold. Although no true cracks were present, rejection of the stub axles was nevertheless justified. Slag streaks could reduce the strength of the stub axles and lead to the formation of fatigue fractures during operation. No recommendations were made.

After about two years in service, a 303 stainless steel valve in contact with a carbonated soft drink in a vending machine occasionally dispensed a discolored drink with a sulfide odor. According to the laboratory at the bottling plant, the soft drink in question was strongly acidic, containing citric and phosphoric acids and having a pH of 2.4 to 2.5. Investigation (visual inspection, chemical analysis, immersion testing in the soft drink, and 100x unetched micrographs) supported the conclusion that the failure was caused by the size and distribution of sulfide stringers in the alloy used in the valve. Manganese sulfide stringers in the valve were exposed at end-grain surfaces in contact with the beverage. The stringers, which were anodic to the surrounding metal, were subject to corrosion, producing a hydrogen sulfide concentration in the immediately adjacent liquid. Recommendations included changing the valve material to type 304 stainless steel.

A hot rolled, low-carbon steel pot used to melt magnesium alloys leaked, releasing about 35 kg (80 lb) of molten magnesium onto the foundry floor and causing an extensive fire. Due to the fire, the original leakage hole could not be investigated. Samples of the failed pot were polished and etched and were found to be composed of ferrite and pearlite mixtures, as would be expected. However, the sample taken from a location about 75 mm (3 in.) from the hole contained a cluster of unusually large inclusions. By removing the beryllium window from in front of the detector, EPMA spectra were obtained from the inclusions and from the steel matrix. The inclusion spectrum contained primarily iron and oxygen, whereas the matrix spectrum contained primarily iron. X-ray maps were made to show the distribution of iron and oxygen. These results indicated that the inclusions were iron oxide. A similar inclusion at the failure site in the melting pot may have reacted violently with the molten magnesium, causing the leak.

The cause of the fatigue failure in the retaining ring of the compressor region of an aero-engine turbine was found to be the presence of a high concentration of nonmetallic inclusions. The results of chemical analysis were used to estimate the phases present. The most frequently observed inclusions were spinel solid solutions of the type MO middot; N2O3, where M = Fe, Mn, or Mg and N = Cr or Al. The detrimental inclusions were corundum, calcium aluminates, cristobalite, and silicates. The most detrimental phases were traced on the surfaces of the specimens fractured using impact loading; the comparison is being made with the polished surfaces and the tensile specimen fracture surfaces. The inclusions in the failed retaining ring were compared with the ones in a similar component obtained from a used engine. In the case of the latter, a large number of fine and elongated (Mn, Cr, Fe)S inclusions were present along with spinels. The nondeformable, rigid oxide particles are considered more undesirable than the sulfides as far as fatigue life of the component is concerned. It has been reported that the presence of sulfides may eliminate the stresses due to oxides.

The master connecting rod of a reciprocating aircraft engine revealed cracks during routine inspection. The rods were forged from 4337 (AMS 6412) steel and heat treated to a specified hardness of 36 to 40 HRC. H-shaped cracks in the wall between the knuckle-pin flanges were revealed by visual examination. The cracks were originated as circumferential cracks and then propagated transversely into the bearing-bore wall. No inclusions in the master rod were detected by magnetic-particle and x-ray inspection. Three large inclusions lying approximately parallel to the grain direction and fatigue beach marks around two of the inclusions were revealed by macroscopic examination of the fracture surface. Large nonmetallic inclusions that consisted of heavy concentrations of aluminum oxide (Al2O3) were revealed by microscopic examination of a section through the fracture origin. The forging vendors were notified about the excess size of the nonmetallic inclusions in the master connecting rods and a nondestructive-testing procedure for detection of large nonmetallic inclusions was established.

An aircraft was grounded when illumination of the transmission oil-pressure light and an accompanying drop in pressure on the oil-pressure gage was reported by the pilot. No discrepancy in the bearing assemblies and related components was revealed by teardown analysis of the transmission. The center bearing of the transmission input-shaft ball-bearing stack had a broken cage and one ball was found to have been split into several pieces. Several scored balls and flaking damage in the raceways of the inner and outer rings was observed. The origin (area in rectangle) was oriented axially in the raceway and was flanked by areas of markedly different-textured flaking damage. Stringers of nonmetallic inclusions were revealed at the origin during metallographic examination of a section parallel to the axially oriented origin. Thus it was concluded that the failure was caused by contact fatigue mechanism (flaking) activated by the subsurface nonmetallic inclusions.

A number of failures involving carbon and alloy steels were analyzed to assess the effects of inclusions and their influence on mechanical properties. Inclusions, including brittle oxides and more ductile manganese sulfides (MnS), affect fatigue endurance limit, fatigue crack propagation rates, fracture toughness, notch toughness, and transverse tensile properties, and do so in an anisotropic manner with respect to rolling direction. Significant property anisotropy has been documented in the failures investigated, providing evidence that designers failed to account for it. Typical fracture morphologies observed in such cases and metallographic appearances of MnS-containing materials are illustrated.

The spontaneous breakage of tempered glass spandrel panels used to cover concrete wall panels on building facades was investigated. Between January 1988 and August 1990, 19 panel failures were recorded. The tinted panels were coated on their exterior surfaces with a reflective metal oxide and covered on the back surfaces with an adherent black polyethylene plastic. Macro fractography, SEM fractography, EDX analysis, and photo elasticimetry were conducted on four of the shattered panels. Small nickel sulfide inclusions were found at the failure origins. Failure of the panels was attributed to growth of the inclusions, coupled with high residual stresses. Fracture mechanics analysis showed that the residual stresses alone were high enough to cause fracture of the glass, with a flaw of the size observed.

The handle from a consumer product exhibited an apparent surface defect. The handle had been injection molded from a medium viscosity grade ABS resin. The anomalous appearance was objectionable to the assembler of the final product and resulted in a production lot being placed on quality-control hold. Investigation included visual inspection, 24x micrographs, and FTIR in the reflectance mode. The spectrum obtained on the included material was characteristic of polybutadiene, the rubber-modifying agent present in ABS. This supported the conclusion that the inclusion's most likely source was an undispersed gel particle formed during the production of the molding resin.