The onset of leakage adjacent to two butt welds in a 2 in. bore pipe was traced to the development of fine cracks. The pipe carried 40% sodium hydroxide solution. The actual temperature was not known, but the pipeline was steam traced at a pressure of 30 psi, equivalent to a temperature of 130 deg C (266 deg F). Magnetic crack detection revealed circumferential crack-like indications situated a short distance from the butt weld. Cracking originated on the bore surfaces of the tube and was of an intergranular nature reminiscent of caustic cracking in steam boilers. The strength of the solution of caustic soda and possibly the temperature also were in the range known to produce stress-corrosion cracking of mild steels in the presence of stresses of sufficient magnitude. In this instance the location of the cracking suggested that residual stresses from welding, which approach yield point magnitude, were responsible. As all other welds were suspect, the remedy was to remove the joints and to reweld followed by local stress relief.

A welded elbow assembly (AISI type 321 stainless steel, with components joined with ER347 stainless steel filler metal by gas tungsten arc welding) was part of a hydraulic-pump pressure line for a jet aircraft. The other end of the tube was attached to a flexible metal hose, which provided no support and offered no resistance to vibration. The line was leaking hydraulic fluid at the nut end of the elbow. Investigation supported the conclusion that failure was by fatigue cracking initiated from a notch at the root of the weld and was propagated by cyclic loading of the tubing as the result of vibration and inadequate support of the hose assembly. Recommendations included changing the joint design from a cylindrical lap joint to a square-groove butt joint. Also, an additional support was recommended for the hose assembly to minimize vibration at the elbow.

A weld in a fuel-line tube broke after 159 h of engine testing. The 6.4-mm (0.25-in.) OD x 0.7-mm (0.028-in.) wall thickness tube and the end adapters were all of type 347 stainless steel. The butt joints between tube and end adapters were made by automated gas tungsten arc (orbital arc) welding. It was found that the tube had failed in the HAZ. Examination of a plastic replica of the fracture surface in a transmission electron microscope established that the crack origin was at the outer surface of the tube. The crack growth was by fatigue; closely spaced fatigue striations were found near the origin, and more widely spaced striations near the inner surface. The quality of the weld and the chemical composition of the tube both conformed to the specifications. However, the fuel-line assembly had vibrated excessively in service. The fuel-line fracture was caused by fatigue induced by severe vibration in service. Additional tube clamps were provided to damp the critical vibrational stresses. No further fuel-line fractures were encountered.

A system of carbon steel headers, handling superheated water of 188 deg C (370 deg F) at 2 MPa (300 psi) for automobile-tire curing presses, developed a number of leaks within about four months after two to three years of leak-free service. All the leaks were in shielded metal arc butt welds joining 200 mm (8 in.) diam 90 deg elbows and pipe to 200 mm (8 in.) diam welding-neck flanges. A flange-elbow-flange assembly and a flange-pipe assembly that had leaked were removed for examination. Investigation (visual inspection, hardness testing, chemical analysis, magnetic-particle testing, radiographic inspection, and 2% nital etched 1.7x views) showed varying IDs on the assemblies and supported the conclusions that the failures of the butt welds were the result of fatigue cracks caused by cyclic thermal stresses that initiated at stress-concentrating notches at the toes of the interior fillet welds on the surfaces of the flanges. Recommendations included using ultrasonic testing to identify the appropriate joints and then replacing them. Special attention to accuracy of fit-up in the replacement joints was also recommended to achieve smooth, notch-free contours on the interior surfaces.

This article briefly reviews the general causes of weldment failures, which may arise from rejection after inspection or failure to pass mechanical testing as well as loss of function in service. It focuses on the general discontinuities observed in welds, and shows how some imperfections may be tolerable and how the other may be root-cause defects in service failures. The article explains the effects of joint design on weldment integrity. It outlines the origins of failure associated with the inherent discontinuity of welds and the imperfections that might be introduced from arc welding processes. The article also describes failure origins in other welding processes, such as electroslag welds, electrogas welds, flash welds, upset butt welds, flash welds, electron and laser beam weld, and high-frequency induction welds.

A 9310 steel gear was found to be defective after a period of engine service. A linear crack approximately was discovered by routine magnetic-particle inspection of an electron beam welded joint that attached a hollow stub shaft to the web of the gear. The welding procedure had a cosmetic weld pass on top of the initial full-penetration weld. There were no other known service failures of gears were welded by this method. One zone of the welded joint showed incomplete fusion, surrounded by two zones containing fatigue beach marks This indicated that the incomplete-fusion zone was the site at which primary fracture originated. The possible causes of incomplete-fusion include localized magnetic deflection of the electron beam, a momentary arc-out of the electron beam, and eccentricity in the small weld diam. The failure was attributed to fatigue originating at the local unfused interface of the electron beam weld, which had been the result of a deviation in the welding procedure. Examination of the possible causes of failure gave no evidence that a recurrence of the defect had ever occurred. Thus, there was no basis on which to recommend a change in design, material, or welding procedure.

Following three similar failures of load chains on manually operated geared pulley-blocks of 1-ton capacity, a portion of one of the chains was obtained for examination. The chain was made of mild steel and the links had been electrically butt-welded at one side. In the case of the sample obtained, the failure in service had resulted from fracture of one of the links in the plane of the weld. Six of the other links in the vicinity showed cracks in the welds in various stages of development. Microscope examination showed a crack in an early stage of development and also from an apparently sound link, the prepared surfaces lying in the planes of the links. This examination revealed that the welds were initially defective. Discontinuities were present in both cases adjacent to the insides of the links, of a type indicative of either inadequate fusion or incomplete expulsion of oxide, etc., at the time of the upset, i.e. the pressing together of the ends of the links to complete the welding. It was evident from the examination that the service failures were due to the use of chain that was initially defective.

Extra high strength zinc-coated 1080 steel welded wire was wound into seven-wire cable strands for use in aerial cables and guy wires. The wires and cable strands failed tensile, elongation, and wrap tests, with wires fracturing near welds at 2.5 to 3.5% elongation and through the welded joints in wrap tests. The welded wire was annealed by resistance heating. The wire ends had a chisel shape, produced by the use of sidecutters. Tests of the heat treatment temperatures showed that the wire near the weld area exceeded 775 deg C (1425 deg F). Metallographic examination revealed martensite present in the weld area after the heat treatment. The test failures of the AISI 1080 steel wire butt-welded joints were due to martensite produced in cooling from the welding operation that was not tempered adequately in postweld heat treatment, and to poor wire-end preparation for welding that produced poorly formed weld burrs. The postweld heat treatment was standardized on the 760 deg C (1400 deg F) transformation treatment. The chisel shape of the wire ends was abandoned in favor of flat filed ends. The wrap test was improved by adopting a hand-cranked device. Under these conditions, the welded joints withstood the tensile and wrap tests.

Airfoil-shape impingement cooling tubes were fabricated of 0.25 mm (0.010 in.) thick Hastelloy X sheet stock, then pulse-laser-beam butt welded to cast Hastelloy X base plugs. Each weldment was then inserted through the base of a hollow cast turbine blade for a jet engine. The weldments were finally secured to the bases of the turbine blades by a brazing operation. One of the laser beam attachment welds broke after a 28-h engine test run. Exposure of the fracture surface for study under the electron microscope revealed the joint had broken in stress rupture. Failure was caused by tensile overload from stress concentration at the root of the laser beam weld, which was caused by the sharp notch created by the lack of full weld penetration. Radiographic inspection of all cooling-tube weldments was made mandatory, with rejection stipulated for joints containing subsurface weld-root notches. In addition, all turbine blades containing cooling-tube weldments were reprocessed by back-brazing. Back brazed turbine blades were reinstalled in the engine and withstood the full 150-h model test run without incident.

A large diameter steel pipe reinforced by stiffening rings with saddle supports was subjected to thermal cycling as the system was started up, operated, and shut down. The pipe functioned as an emission control exhaust duct from a furnace and was designed originally using lengths of rolled and welded COR-TEN steel plate butt welded together on site. The pipe sustained local buckling and cracking, then fractured during the first five months of operation. Failure was due to low cycle fatigue and fast fracture caused by differential thermal expansion stresses. Thermal lag between the stiffening rings welded to the outside of the pipe and the pipe wall itself resulted in large radial and axial thermal stresses at the welds. Redundant tied down saddle supports in each segment of pipe between expansion joints restrained pipe arching due to circumferential temperature variations, producing large axial thermal bending stresses. Thermal cycling of the system initiated fatigue cracks at the stiffener rings. When the critical crack size was reached, fast fracture occurred. The system was redesigned by eliminating the redundant restraints and by modifying the stiffener rings to permit free radial thermal breathing of the pipe.

Tie bars of a dragline excavator each consisted of a rectangular section steel bar to which eye-pieces, to facilitate anchorage, were attached by butt-welds. Failure of one weld in each bar after seven years of service allowed the boom to fall and become extensively damaged. The appearance of the fracture faces of the two welds showed partial-penetration joints. Failure in each bar had taken place through the weld metal. The presence of built-in cracks introduced zones of stress concentration and the fluctuating loads to which the ties were subjected in service served to initiate fatigue cracks. While the partial-penetration type of weld may be tolerated in a component subjected to bending stresses it is undesirable in one that is required to withstand fluctuating tensile stresses.

Caustic cracking is the term used to describe one of the forms in which stress-corrosion cracking manifests itself in carbon steels. In the present study, persistent leakage occurred after ten weeks of service from tube expansions in the steam and mud drum of a two-drum D type boiler, which failed to respond to repeated expansion. The leakage was traced to circumferential cracking in the portion of Fe-0.11C-0.46Mn-0.018S-0.011P tubes within the expanded region. Microscopic examination indicated that all cracks started from the outer surface of the tubes in the expanded portion. The form of cracking which was mostly intergranular. Examination at higher magnification disclosed that a selective attack had taken place on the carbide constituents of the pearlite grains. An alkaline deposit on the fireside surface of the tube resulted from the evaporation of boiler water which had found its way past the tube expansions. This indicated that this operation had not resulted in a satisfactorily tight joint.

During a refit of a twenty-year-old Naval destroyer, two cracks were found on the inside of the killed carbon-manganese steel hull plate at the forward end of the boiler room. The cracks coincided with the location of the top and bottom plates of the bilge keel. Metallurgical examination of sections cut from the cracked area identified lamellar tearing as the principle cause of the cracking. This was surprising in 6 mm thick hull plates. Corrosion fatigue and general corrosion also contributed to hull plate perforation. Although it is probable that more lamellar tears exist near the bilge keel in other ships and may be a nuisance in the future, the hull integrity of the ships is not threatened and major repairs are not needed.

Three links of a chain showing unusually strong wear were examined. Corresponding to the stress, the wear was strongest in the bends of the links, but it was especially pronounced in the bend in which the butt weld seam was located. Investigation showed the links were manufactured from an unkilled carbon-deficient steel, and were case hardened to a depth of 0.8 to 0.9 mm. The peripheral structure at the places not showing wear consisted of coarse acicular martensite with a high percentage of retained austenite. The links therefore were strongly overheated, probably directly heated during case hardening. The butt weld seams were not tight and were covered with oxide inclusions. Given that wear occurred preferentially at the welds it may be concluded that this weld defect contributed to the substantial wear. This leaves unanswered whether these chains could have withstood the high operating stress if they had been welded satisfactorily and hardened correctly, and whether it made any sense to case harden highly stressed chains of this type.

Welded low-carbon steel bomb fins were rejected because of poor weld practice. Visual and metallographic examination revealed that the resistance plug welds that attach the outer skin to the inner spar displayed inadequate weld penetration. Recommended changes to the resistance welding parameters resulted in acceptable welds.

This case study describes the failure analysis of a steel nozzle in which cracking was observed after a circumferential welding process. The nozzle assembly was made from low-carbon CrMoV alloy steel that was subsequently single-pass butt welded using gas tungsten arc welding. Although no cracks were found when the welds were visually inspected, X-ray radiography showed small discontinuous surface cracks adjacent to the weld bead in the heat affected zone. Further investigation, including optical microscopy, microhardness testing, and residual stress measurements, revealed that the cracks were caused primarily by the presence of coarse untempered martensite in the heat affected zone due to localized heating. The localized heating was caused by high welding heat input or low welding speed and resulted in high transformation stresses. These transformation stresses, working in combination with thermal stresses and constraint conditions, resulted in intergranular brittle fracture.

This article describes some of the welding discontinuities and flaws characterized by nondestructive examinations. It focuses on nondestructive inspection methods used in the welding industry. The sources of weld discontinuities and defects as they relate to service failures or rejection in new construction inspection are also discussed. The article discusses the types of base metal cracks and metallurgical weld cracking. The article discusses the processes involved in the analysis of in-service weld failures. It briefly reviews the general types of process-related discontinuities of arc welds. Mechanical and environmental failure origins related to other types of welding processes are also described. The article explains the cause and effects of process-related discontinuities including weld porosity, inclusions, incomplete fusion, and incomplete penetration. Different fitness-for-service assessment methodologies for calculating allowable or critical flaw sizes are also discussed.