The information provided in this article is intended for those individuals who want to determine why a casting component failed to perform its intended purpose. It is also intended to provide insights for potential casting applications so that the likelihood of failure to perform the intended function is decreased. The article addresses factors that may cause failures in castings for each metal type, starting with gray iron and progressing to ductile iron, steel, aluminum, and copper-base alloys. It describes the general root causes of failure attributed to the casting material, production method, and/or design. The article also addresses conditions related to the casting process but not specific to any metal group, including misruns, pour shorts, broken cores, and foundry expertise. The discussion in each casting metal group includes factors concerning defects that can occur specific to the metal group and progress from melting to solidification, casting processing, and finally how the removal of the mold material can affect performance.

This article presents a failure analysis of an aluminum cylinder head on an automotive engine. During an endurance test, a crack initiated from the interior wall of a hole in the center of the cylinder head, then propagated through the entire thickness of the component. Metallurgical examination of the crack origin revealed that casting pores played a role in initiating the crack. Stress components, identified by finite element analysis, also played a role, particularly the stresses imposed by the bolt assembly leading to plastic strain. It was concluded that the failure can be prevented by eliminating the bolt hole, using a different type of bolt, or adjusting the fastening torque.

A commercial hybrid-iron golf club fractured during normal use. The club fractured through its cast aluminum alloy hosel. Optical analysis revealed casting pores through 20% of the hosel thickness. Mechanical properties were determined from characterization results, then used to construct a finite element model to analyze material performance under failure conditions. In addition, a full scale structural test was conducted to determine failure strength. It was concluded that the club failed not from ground impact but from a force reversal at the bottom of the downswing. Large moments generated during the downswing aggravated by manufacturing defects and stress concentration combined to create an overload condition.

A roadarm for a tracked vehicle failed during preproduction vehicle testing. The arm was a weldment of two cored low-alloy steel sand castings specified to ASTM A 148, grade 120–95. A maximum carbon content of 0.32% was specified. The welding procedure called for degreasing and gas metal arc welding; neither preheating nor postheating was specified. The filler metal was E70S-6 continuous consumable wire with a copper coating to protect it from atmospheric oxidation while on the reel. Analysis of the two castings revealed that the carbon content was higher than specified, ranging from 0.40 to 0.44%. The fracture occurred in the HAZ , where quenching by the surrounding metal had produced a hardness of 55 HRC. Some roadarms of similar carbon content and welded by the same procedure had not failed because they had been tempered during a hot-straightening operation. Brittle fracture of the roadarm was caused by a combination of too high a carbon equivalent in the castings and the lack of preheating and postheating during the welding procedure. A pre-heat and tempering after welding were added to the welding procedure.

During the operation of tractors with cantilevered body, the lateral wall of the hypoeutectic cast iron cylinder blocks cracked repeatedly. Three of the blocks were examined. The grain structure of the thick-walled part consisted of uniformly distributed graphite of medium flake size in a basic mass of pearlite with little ferrite. But the thin-walled part showed a structure of dendrites of precipitated primary solid solution grains with pearlitic-ferritic structure and a residual liquid phase with granular graphite in the ferritic matrix. The structure was formed by undercooling of the residual melt. In this case, it was promoted by fast cooling of the thin wall and had comparatively low strength. The fracture formation in the cylinder blocks was ascribed primarily to casting stresses. They could be alleviated by better filleting of the transition cross sections. The fracture was promoted by the formation of undercooled microstructure of low strength in the thin-walled part. Similar damage appeared in a cylinder head, in which case, the cracks were promoted by a supercooled structure.

Cast iron bearing caps in tractor engines fractured repeatedly after only short operating periods. The fracture originated in a cast-in groove and ran approximately radially to the shaft axis. The smallest cross section was at the point of fracture. The core structure of the caps consisted of graphite in pearlitic-ferritic matrix. Casting stresses did not play a decisive role because of the simple shape of the pieces that were without substantial cross sectional variations. Two factors exerted an unfavorable effect in addition to comparatively low strength. First, the operating stress was raised locally by the sharp-edged groove, and second, the fracture resistance of the cast iron was lowered at this critical point by the existence of a ferritic bright border. To avoid such damage in the future it was recommended to observe one or more of the following precautions: 1) Eliminate the grooves; 2) Remove the ferritic bright border; 3) Avoid undercooling in the mold and therefore the formation of granular graphite; 4) Inoculate with finely powdered ferrosilicon into the melt for the same purpose; and, 5) Anneal at lower temperature or eliminate subsequent treatment in consideration of the uncomplicated shape of the castings.

Examination of a connecting-rod shell bearing from a six-cylinder gasoline engine was done after it was returned to the factory. Copper-lead alloy SAE 485 bonded to a low-carbon steel backing was used to make the bearing and the oil used in the engine was the recommended one. Measurable material loss was visible over most of the bearing halves particularly in a wide region at the centerline. A brittle waxlike substance identified to be a mixture of copper and lead sulfides covered the visible shallow pits and the darkened region. Change of oil with greater frequency to prevent the buildup of sulfur compounds or bearing halves that have corrosion-resistant overlay materials were recommended as best solutions.

The steam tug Hercules was an ocean-going and bay tug for 55 years before being retired. It is now being restored by the National Park Service. A broken steam valve was obtained for microstructural examination. The body was gray cast iron, and the stem and seat were brass. The examination centered on corrosion of the brass components. The seat and shaft were alpha brass, with a hardness of 64 and 79 DPH, respectively. A nut held the shaft onto the seat, and was alpha-beta brass with a hardness of 197 DPH. Welded on the end of the shaft was a ring of hard (DPH 294) alpha-beta brass, which seated against the nut. The brass seat and stem show little corrosion. However, the alpha-beta brass nut and welded tip showed extensive dezincification. This process of removal of Zn and the retention of Cu began in the high Zn beta phase, but eventually both phases were attacked. The depth of penetration was consistent with dezincification rates reported in the literature for such brasses in salt water if the valve had been in service about 55 years.

A sand-cast gray iron flanged nut was used to adjust the upper roll on a 3.05 m (10 ft) pyramid-type plate-bending machine. The flange broke away from the body of the nut during service. Analysis (visual inspection and 150x micrographs of sections etched with nital) supported the conclusions that brittle fracture of the flange from the body was the result of overload caused by misalignment between the flange and the roll holder. The microstructure contained graphite flakes of excessive size and inclusions in critical areas; however, these metallurgical imperfections did not appear to have had significant effects on the fracture. Recommendations included carefully and properly aligning the flange surface with the roll holder to achieve uniform distribution of the load. Also, a more ductile metal, such as steel or ductile iron, would be more suitable for this application and would require less exact alignment.

A full lift disk, made of die cast brass, which served as a lifting aid in a safety valve, had cracked in service at a number of locations in the vicinity of the threaded hole. During microscopic examination, agglomeration of oxide inclusions were noted in the region of the cracks. Because the die cast brass was alloyed with aluminum, these inclusions consisted predominantly of aluminum oxide. The tolerable limit in pores and oxide inclusions was greatly exceeded in the lift disk under examination. Above all, the numerous oxide skins disrupted the cohesion of the microstructure and were primarily responsible for the failure of the lift disk.

This paper describes an investigation on the failure of a large leaded bronze bearing that supports a nine-ton roller of a plastic calendering machine. At the end of the normal service life of a good bearing, which lasted for seven years, a new bearing was installed. However the new one failed catastrophically within a few days, generating a huge amount of metallic wear debris and causing pitting on the surface of the cast iron roller. Following the failure, samples were collected from both good and failed bearings. The samples were analyzed chemically and their microstructures examined. Both samples were subjected to accelerated wear tests in a laboratory type pin-on-disk apparatus. During the tests, the bearing materials acted as pins, which were pressed against a rotating cast iron disk. The wear behaviors of both bearing materials were studied using weight loss measurement. The worn surfaces of samples and the wear debris were examined by light optical microscope, SEM, and energy-dispersive x-ray microanalyzer. It was found that the laboratory pin-on-disk wear data correlated well with the plant experience. It is suggested that the higher lead content ~18%) of the good bearing compared with 7% lead of the failed bearing helped to establish a protective transfer layer on the worn surface. This transfer layer reduced metal-to-metal contact between the bearing and the roller and resulted in a lower wear rate. The lower lead content of the failed bearing does not allow the establishment of a well-protected transfer layer and leads to rapid wear.

A stuffing box (sand cast from ASTM A 536, grade 60-45-10, ductile iron) began leaking water after two weeks of service. The machine was operating at 326 rpm with a discharge water pressure of 21.4 MPa (3100 psi). Investigation (visual inspection, mechanical analysis, and nital etched 100x magnification) supported the conclusion that the crack initiated at the inner edge of a lubrication hole and had propagated toward both the threaded and flange ends of the casting. An appreciable residual-stress concentration must have been present and caused propagation of the crack. The residual stress might have been caused when a fitting was tightly screwed into the lubrication hole, and it might have been concentrated by notches at the inner end of the hole created when the drill broke through the sidewall to the stuffing box.

A sample tube was removed from a reformer furnace for life assessment after 69,000 h of service. Sections were cut from the tube, which was a spindle cast A297 Grade HK 40 (25 Cr, 20 Ni, 0.4 C) austenitic steel of 122.5 mm OD and 10.5 mm nominal wall thickness. They were examined metallographically on transverse sections and on longitudinal sections through the butt welds joining the separate cast segments of the tube. Creep damage was mainly concentrated within the inner one third of the wall thickness. The use of damage assessment parameters in evaluating the reformer tube remaining life showed the welds to be inadequate, and to have a strength and creep resistance below those of the base metal.

This investigation involved two AISI 304L acid storage tanks and one AISI 304L spent solvent tank from a sewage treatment facility. After installation, these tanks were hydrostatically tested using sewage effluent. No leaks were found and after a year or two, the tanks were drained and filled with nitric acid in preparation for service. Three weeks later the two acid tanks were found to be leaking from the bottom. Samples from the spent solvent tank revealed that pitting was located in a depressed area near a suction hole, beneath a black residue. It was concluded that the acid tanks failed by chloride-induced pitting initiated by microbial activity. Further, the spent solvent tank failed by a similar, but anaerobic mechanism. The use of the effluent for the hydrostatic test and the failure to remove it and clean and dry the tanks was the primary cause of failure. Localized carbide segregation in the original plate served as preferential corrosion sites. Had the tanks been hydrostatically tested in a proper manner, the pitting may not have occurred.

Microstructural examinations on transverse cross sections of a steam reformer tube, showed the presence of large macrovoids elongated in the radial direction and emanating from the internal surface of the tube. The macrovoids were located at the interdendritic regions, and were partially filled by a Mn-Fe bearing chromium oxide film. The areas adjacent to the oxide film were chemically depleted in C, Cr and Mn and rich in Fe and Ni. Associated with this depletion were a large concentration of microvoids. It was suggested that the dissolution of carbides in areas surrounding the macrovoids and the concentration of stresses at their tips, caused extensive localized plastic deformation which led to the formation of microvoids and subsequently to the spalling of the oxide film. The non-protective character of the film induced a progressive deterioration of the grain boundaries properties. Grain boundary sliding and dislocation motion were enhanced, causing a local increase in the steady state strain rate and the premature failure of the tube.

Chain link, a part of a mechanism for transferring hot or cold steel blooms into and out of a reheating furnace, broke after approximately four months of service. The link was cast from 2% Cr austenitic manganese steel and was subjected to repeated heating to temperatures of 455 to 595 deg C (850 to 1100 deg F). Examination included visual inspection, macrograph of a nital-etched specimen from an as-received chain link 1.85x, micrographs of a nital-etched specimen from an as-received chain link 100x/600x, normal microstructure of as-cast standard austenitic manganese steel 100x, micrograph of a nital-etched specimen that had been austenitized 20 min at 1095 deg C (2000 deg F) and air cooled 315x, and micrograph of the same specimen after annealing 68 h at 480 deg C (900 deg F) 1000x). Investigation supported the conclusions that the chain link failed in a brittle manner, because the austenitic manganese steel from which it was cast became embrittled after being reheated in the temperature range of 455 to 595 deg C (850 to 1100 deg F) for prolonged periods of time. The alloy was not suitable for this application, because of its metallurgical instability under service conditions.

A ductile iron T-hook hook was reported to have fractured in service. It was further reported that the hook had been subjected to a load that did not exceed 5900 kg (13,000 lb) at the time of fracture. No information was provided regarding the type of metal used to manufacture the hook. A failure analysis was requested to determine the cause of fracture. Two hooks were submitted for examination. Analysis (visual inspection, 2.7x light fractography, chemical analysis, 110x SEM fractography, 27x/110x/215x nital-etched micrographs) supported the conclusions that this component fractured in service as a consequence of ductile tensile overload. Evidence indicates that the fractured region was subjected to a load exceeding the capacity of the material. Because the information available from the service application indicated that the component had not been subjected to a stress that exceeded 5900 kg (13,000 lb), the observations made in this investigation suggested that either the load was underestimated or that the indicated load was applied at a more rapid rate (perhaps with a jerk), which would tend to increase the effective force of the load.

A 58.4 cm (23 in.) diam heavy-duty brake drum component of a cable-wound winch broke into two pieces during a shutdown period. Average service life of these drums was two weeks; none had failed by wear. The drums were sand cast from ductile iron. During haul-out, the cable on the cable drum drove the brake drum, and resistance was provided by brake bands applied to the outside surface of the brake drum. Friction during heavy service was sufficient to heat the brake drum, clutch mount, and disk to a red color. Examination of the assembly indicated that the brake drum would cool faster than its mounts and would contract onto them. Brittle fracture of the brake drum occurred as a result of thermal contraction of the drum web against the clutch mount and the disk. The ID of the drum web was enlarged sufficiently to allow for clearance between the web and the clutch mount and disk at a temperature differential of up to 555 deg C (1000 deg F). With the adoption of this procedure, brake drums failed by wear only.

The hot gas casing of a gas turbine used for peak load power production had developed extensive cracking during operation. The operating time was 18,000 h, and it had been subjected to 1,600 operating cycles. The gas temperature on the hot side was 985 deg C, on the cold side 204 deg C, the material being AISI 321 stainless steel. The purpose of the present study was to determine optimum repair welding procedures on the premise that the material was basically sound and undamaged by creep. The cracking was the result of thermal fatigue, and such cracks can propagate at elevated temperature, with damage ahead of the crack tip occurring by means of very local processes of creep. Metallographic examination disclosed heavy surface layers of carbides, such that the material was extremely brittle when subjected to bending. Accordingly, although it was demonstrated that the casing could be welded successfully, it was suggested that the remaining useful life was effectively exhausted and that it should be replaced. Thermal stresses produced during operation would rapidly result in additional cracks.

Lakes in zinc die castings are areas encompassed by irregular lines or waves on flat or slightly contoured surfaces which are intended to look uniform. The laked areas have to be removed by polishing before the castings can be plated. This adds considerably to the overall cost of production. Castings examined were of an automobile name-plate holder with two flat sides of approximately 113 sq cm. All castings produced during a trial showed laking defects, the number and position varying from casting to casting. It was found that formation of metal waves and lakes depended primarily on the design of the gate and runner system and operating conditions. High flow efficiencies, with adequate feeding to all sections of the die, and short cavity fill times are desirable.