Mechanical springs are used in mechanical components to exert force, provide flexibility, and absorb or store energy. This article provides an overview of the operating conditions of mechanical springs. Common failure mechanisms and processes involved in the examination of spring failures are also discussed. In addition, the article discusses common causes of failures and presents examples of specific spring failures, describes fatigue failures that resulted from these types of material defects, and demonstrates how improper fabrication can result in premature fatigue failure. It also covers failures of shape memory alloy springs and failures caused by corrosion and operating conditions.

An agricultural tine, which is a relatively large double torsion spring with outer legs that are used to sweep through hay or other crops and turn them over, had failed. It was made hard-drawn carbon steel. Bending fatigue was revealed by visual examination to be almost certainly the cause of failure. The fatigue fracture origin was found on the inside surface of the legs at the point where they joined the coiled body of the spring. It was established that the tines after being wound up by loading with hay, sprung back through the neutral unloaded position and into the unwind direction. This movement into the unwind direction was concluded to be happening often enough to initiate fatigue. The stress relieving temperature was recommended to be increased to reduce the residual stresses from coiling and hence improve fatigue performance.

U-shaped leaf springs, intended to serve as spacers between oil tank floats and the inner walls of the containers, broke while being fitted, or after a short time in use, in the bend of the U. The springs were made of tempered strip steel of type C 88 with 0.84 % C, bent at room temperature, and electroplated with cadmium for protection against corrosion. Each fracture showed seven or eight kidney-shaped cracks. At the origins of these cracks on the concave inner surface of the springs, crater-like depressions and beads of melted and resolidified material were found. Fracture of the springs was caused by stress cracks as a consequence of local hardening. The hardening caused by melting and resolidification, and therefore the cracks in the springs, was the result of a faulty procedure during cadmium electroplating.

A valve-seat retainer spring (made of 0.23 mm thick 17-7 PH stainless steel) from a fuel control on an aircraft engine was found to be broken after 3980 h of service. The two inner tabs were found to be broken off. The part was revealed to be in relative rotation against its contacting member by the radial wear marks on the convex surface. Beach marks indicating that fatigue fracture had been initiated at the convex surface of the washer and had propagated across to the concave surface were revealed by examination of the fractured surfaces of the washer. The cracks were revealed to have originated in the 0.38-mm radius fillet between the tab and the body of the washer. It was interpreted from the analysis of the compound fracture that it was composed of fatigue fractures caused by the formed tab being loaded so as to compress the spring along the axis of its centerline and produce torsional vibrations. It was concluded that the two inner tabs had broken in fatigue as the result of cyclic loading that compressed and torsionally vibrated the spring. The fillets were replaced with slots to minimize stress concentration at the corners as a corrective measure.

A flat spring for the main landing gear of a light aircraft failed after safe execution of a hard landing. The spring material was identified by chemical analysis to be 6150 steel. The fracture was found to have occurred near the end of the spring that was inserted through a support member about 25 mm thick and attached to the fuselage by a single bolt. Brinelling (plastic flow and indentation due to excessive localized contact pressure) was observed on the upper surface of the spring where the forward and rear edges of the spring contacted the support member. It was indicated by chevron marks that brittle fracture had started beneath the brinelled area at the forward edge of the upper surface of the spring. The origin of the brittle fracture was found to be a small fatigue crack that had been present for a considerable period of time before final fracture occurred. Fracture of the landing-gear spring was concluded to have been caused by a fatigue crack that resulted from excessive brinelling at the support point. Regular visual examinations to detect evidence of brinelling and wear at the support in aircraft with this configuration of landing-gear spring were recommended.

Spring failures were investigated in this study. A seam that extended more than 0.05 mm below the wire surface was revealed and the fatigue-fracture front progressed downward from several origins. A crack that is triangular in outline was produced by each of the fronts. This was reported to have occurred when the fracture plane changed to an angle with the wire axis in response to the torsional strain. The spring failure was concluded to have originated at the seam.

A steel valve spring meeting Steel-Iron-Test 1570 fractured during the high-stress condition of the operation of its valve. Metallographic examination of a transverse section adjacent to the fracture and a longitudinal section through the crack showed the steel was free of major defects and was of high purity, although a number of minor surface defects such as rolling laps were found. The spring was heat treated and its surface strengthened by shot-peening, but the surface was also decarburized to a depth of approximately 0.03 mm which resulted in a lowering of the surface hardness. The fracture of this valve spring is therefore primarily due to surface defects, and secondly perhaps also to weak surface decarburization. No recommendation resulted from the investigation except to note that comparatively minor effects suffice to cause fractures in highly stressed springs.

Life testing of cyclic loaded, miniature extension springs made of 17-7 PH stainless steel wire and AISI 302 Condition B stainless steel wire has shown end hook configuration to be a major source of weakness. To avoid cracking and subsequent fatigue failure, it was found that stress concentration depended on end hook bend sharpness. Also, interference fits are to be avoided in the end hooks of small springs. Additionally, a need for careful consideration of the stress-corrosion properties of candidate materials for spring applications has been demonstrated by stress-corrosion test results for 17-7 PH CH900 and for Custom 455 CH850 stainless steels. Laboratory testing of these two materials in the form of compression springs confirmed the superiority of the 17-7 PH over Custom 455.

During testing of compressors under start/stop conditions, several helical suspension springs failed. The ensuing failure investigation showed that the springs failed due to fatigue. The analysis showed that during start/stop testing the springs would undergo both a lateral and axial deflection, greatly increasing the torsional stresses on the spring. To understand the fatigue limits under these test conditions, a bench test was used to establish the fatigue strength of the springs. The bench tests showed that the failed springs had an unacceptable surface texture that reduced the fatigue life. Based on an understanding of the compressor motion, a Monte Carlo model was developed based on a linear damage theory to predict the fatigue life of the springs during start/stop conditions. The results of this model were compared to actual test data. The model showed that the design was marginal even for springs with acceptable surface texture. The model was then used to predict the fatigue life requirements on the bench test such that the reliability goals for the start/stop testing would be met, thus reducing the risk in qualifying the compressor.

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.

Several of the springs, made of 1.1 mm diam Inconel X-750 wire and used for tightening the interstage packing ring in a high-pressure turbine, were found broken after approximately seven years of operation. Intergranular cracks about 1.3 mm in depth and oriented at an angle of 45 deg to the axis of the wire were revealed by metallographic examination. A light-gray phase, which had the appearance of liquid-metal corrosion, was observed to have penetrated the grains on the fracture surfaces. The spring wires were found to fracture in a brittle manner characteristic of fracture from torsional loading (along a plane 45 deg to the wire axis). Liquid-metal embrittlement was expected to have been caused by metals (Sn, Zn, Pb) which melt much below maximum service temperature of the turbine. The springs were concluded to have fractured by intergranular stress-corrosion cracking promoted by the action of liquid zinc and tin in combination with static and torsional stresses on the spring wire. As a corrective measure, Na, Sn, and Zn which were present in pigmented oil used as a lubricant during spring winding was cleaned thoroughly by the spring manufacturer before shipment to remove all contaminants.

The safety valve on a steam turbogenerator was set to open when the steam pressure reaches 2400 kPa (348 psi). The pressure had not exceeded 1790 kPa (260 psi) when the safety-valve spring shattered into 12 pieces. The steam temperature in the line varied from about 330 to 400 deg C (625 to 750 deg F). Because the spring was enclosed and mounted above the valve, its temperature was probably slightly lower. The 195 mm (7 in.) OD x 305 mm (12 in.) long spring was made from a 35 mm (1 in.) diam rod of H21 hot-work tool steel. It had been in service for about four years and had been subjected to mildly fluctuating stresses. Analysis (visual inspection, 0.3x photographs, 0.7x light fractographs, and metallographic examination) supported the conclusions that the spring failed by corrosion fatigue that resulted from application of a fluctuating load in the presence of a moisture-laden atmosphere. Recommendations included replacing all safety valves in the system with new open-top valves that had shot-peened and galvanized steel springs. Alternatively, the valve springs could be made from a corrosion-resistant metal-for example, a 300 series austenitic stainless steel or a nickel-base alloy, such as Hastelloy B or C.

A cadmium-plated music-wire return spring that operated in a pneumatic cylinder designed for infinite life at a maximum stress level of 620 MPa failed after 240,000 cycles. An extremely hard and small kernel, which looked like a weld deposit, was observed at the center of the fractured surface. The kernel was assumed to have resulted from extreme localized overheating. These springs were reported to have been barrel electroplated after fabrication. The intermittent contact with the dangler (suspended cathode contact) as the barrel rotated allowed high local currents when the last contact was broken was revealed to have resulted in an arc that caused local melting of the metal being plated. The molten metal was interpreted to have been quenched instantly by the plating solution and by the mass of the cold metal of the spring. The hard spot caused by arcing during plating was concluded to be the reason of the fatigue failure. Rack plating or barrels with fixed button contacts at many points instead of dangler-type contacts were recommended to avoid hard spots.

Grease-wiper springs for cams formed from stampings of 0.25-mm thick carbon spring steel (0.65 to 0.80% C) fractured at the 0.025 mm radius on the stamped 135 deg corner at a 90 deg bend after 5,000,000 cycles. Tool marks 2 to 2.3 mm from the center of the stamped bend were disclosed by visual examination. Fatigue striations originating from cracks at the 0.025 mm radius inside corner at the bend were revealed by SEM of the fractured surface. The maximum stress at the bend, in stock of maximum thickness and as a function of the radius of the 135 deg corner, was indicated by stress calculations to be very close to the maximum allowable fluctuating stress for the material. The wiper springs were concluded to be fractured in fatigue and the cyclic loading resulted from cam rotation. The maximum applied stress approached the allowable limit due to high stress-concentration factor in the spring (caused by the very small inside radius). The corner radius was increased to 0.76 mm and the tools were re-polished to avoid tool marks.

A medium-carbon helical spring was installed in a machine assembly that was welded into its final location. Weld spatter was not prevented from landing on the wire surface by any shield. An elongated drop and two tiny droplets of metal were observed a short distance from the fracture. No droplets were revealed at the origin of the fracture, but it was assumed that a drop of molten metal landed at the origin. Adherence of the spatter drop was expected to have been affected by the opening and closing of the fatigue crack. Weld spatter bead was concluded to have caused the fatigue fracture.

A 6150 flat spring was found to be failed. The face of the spring was revealed to be under tensile stress. The failure was concluded to have begun at the dark spot on the edge where roughness resulted from shearing during the blanking operation.

Two outer valve springs made from air-melted 6150 pretempered steel wire broke during production engine testing. The springs were 50 mm in OD and 64 mm in free length, had five coils and squared-and-ground ends, and were made of 5.5 mm diam wire. It was revealed that fracture was nucleated by an apparent longitudinal subsurface defect. The defect was revealed by microscopic examination to be a large pocket of nonmetallic inclusions (alumina and silicate particles) at the origin of the fracture. Partial decarburization of the steel was observed at the periphery of the pocket of inclusions. Torsional fracture was indicated by the presence of beach marks at a 45 deg angle to the wire axis. It was established that the spring fractured by fatigue nucleated at the subsurface defect.

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.

A copper alloy C51000 (phosphor bronze, 5%A) failed prematurely during life testing of several such springs. The wire used for the springs was 0.46 mm (0.018 in.) in diam and was in the spring-temper condition. The springs were revealed to be subjected to cyclic loading, in the horizontal and vertical planes during the testing. The fracture was revealed to have occurred in bend 2. An indentation, presumably caused by the bending tool during forming, at the inner surface of the bend where fracture occurred was revealed by microscopic examination. Spiral marks produced on springs during rotary straightening were observed. A crack that had originated at the surface at the inside bend and had propagated toward the outside of the bend was revealed by microscopy of a longitudinal section taken through bend 2. The small bend radius was interpreted to contribute to spring fatigue as a result of result in straining at the bend zone. The spring was concluded to have failed in fatigue. It was recommended that the springs should be made of wire free from straightener marks and the bending tool should be redesigned so as not to indent the wire.

The power-type counterbalance spring, formed from hardened-and-tempered carbon steel strip and subsequently subjected to phosphating treatment, fractured at the two locations during fatigue testing. A rust colored dark band at the inside edge of the fracture surface was disclosed during investigation. Etch pits were revealed by the cleaned surface which were never observed on properly phosphated coating. It was interpreted that the spring had been subjected to an abnormal acid attack in pickling or phosphating which had resulted in considerable absorption of hydrogen by the metal and hence embrittlement. The part was concluded to have cracked during phosphating or excessive acid pickling before phosphating.