故障解析

Stress relaxation is the time-dependent loss of load when a spring is held at a constant compressed height at elevated temperatures. It is a form of creep. For a wave spring, the relaxation rate depends on the material's metallurgical stability and the initial stress level. For example, SAE 1070 carbon steel begins to relax significantly above $250^\circ F$ ($121^\circ C$), losing 5-10% of its load. In contrast, Inconel X-750 can operate at $1100^\circ F$ with minimal relaxation. The failure is typically identified when an assembly (like a mechanical seal) begins to leak because the spring is no longer providing sufficient seating force. Designers use the 'Larson-Miller Parameter' to predict the long-term relaxation and compensate by over-designing the initial load or selecting a more stable superalloy.

Fretting fatigue is a failure mechanism that occurs in nested wave springs where the parallel layers of wire rub against each other during cyclic loading. This micro-motion (fretting) removes the protective oxide layer of the metal and creates small pits that act as stress concentrators. Eventually, these pits initiate fatigue cracks that propagate through the wire thickness. This is particularly common in high-frequency applications like vibration isolators. To prevent fretting, engineers specify dry-film lubricants (like MoS2 or PTFE) or ensure the spring is submerged in oil. Failure analysis typically reveals 'red rust' (cocoa) in carbon steel or shiny worn patches in stainless steel at the contact points between turns, accompanied by crack initiation sites.

Permanent set occurs when the internal stress in the wave spring exceeds the yield strength ($\sigma_y$) of the material, causing plastic deformation. This is often diagnosed when the free height of the spring ($H_{free}$) is significantly shorter after the first compression or after a period of service. The maximum stress in a wave spring occurs at the ID or OD of the wave peaks. If the calculated stress $\sigma = \frac{3 \pi P D}{4 N^2 b t^2}$ exceeds the tensile yield of the material (e.g., $210 ksi$ for 17-7PH CH900), the spring will not return to its original height. Engineers use 'preset' operations—compressing the spring to solid during manufacturing—to induce beneficial residual stresses and minimize further set during operation. If a spring fails in the field, a decrease in $H_{free}$ is the primary indicator of over-stressing.

Galvanic corrosion occurs when two dissimilar metals are in electrical contact in the presence of an electrolyte (like salt water). Stainless steel (302/316) is more noble than carbon steel. In this 'galvanic couple,' the carbon steel housing acts as the anode and corrodes preferentially, while the stainless ring acts as the cathode. The corrosion typically happens inside the groove, where the housing material is eaten away. This 'undercuts' the ring, eventually leading to a loss of the groove wall and the ring being ejected under load. In marine environments, this is prevented by using matching materials, applying protective coatings, or using sacrificial anodes. The 'Area Rule' also applies: a small anode (housing) and a large cathode (ring) is the worst-case scenario.

In pneumatic actuators, the piston often strikes the retaining ring with high kinetic energy $E_k = 0.5 \cdot m \cdot v^2$. Unlike static loads, impact loads create dynamic stress waves that can exceed the material's yield strength for a fraction of a millisecond. This can cause the spiral ring to 'dish' or even fracture due to the high strain rate sensitivity of some steels. To analyze this, the 'dynamic load factor' $L_d$ is used, where $P_{dynamic} = P_{static} \cdot L_d$. For sudden impacts, $L_d$ can be as high as 2.0. If failure occurs, the solution is to incorporate a 'buffer' or 'shock absorber' (like a polyurethane washer) between the piston and the ring, or to increase the ring's thickness $t$ to increase its stiffness and energy absorption capacity.

Edge margin is the distance $z$ from the groove to the end of the shaft. If $z$ is too small, the material between the groove and the shaft end can fail in 'double shear' or 'shear-out.' The failure looks like a small 'plug' of material being pushed off the end of the shaft. The required margin is typically $z \ge 3 \cdot d$ (where $d$ is groove depth). If a failure occurs even with $z > 3 \cdot d$, it may be due to 'moment-induced bending' of the shaft end. This is common in brittle materials like cast iron. Engineers should perform a finite element analysis (FEA) to ensure that the combined stress (axial shear + bending) at the corner of the groove does not exceed the material's shear strength.

Diagnosis of centrifugal dislodgement involves inspecting the ring and the groove for specific indicators. If the ring is found outside the groove and appears 'stretched' (its free diameter is now larger than its original manufactured diameter), but there are no shear marks on the groove edge, it likely expanded due to RPM. If the ring has 'burnished' or 'shiny' marks on its inner diameter, it indicates that it was spinning relative to the shaft before it dislodged. The forensic engineer would calculate the lift-off speed using $N = \sqrt{(K \cdot E \cdot I) / (w \cdot R^3)}$ and compare it to the motor's peak RPM. If the motor exceeded this speed, the design requires a self-locking spiral ring or a higher-interference fit.

Groove deformation occurs when the axial load $P$ exceeds the compressive yield strength of the groove material, even if the retaining ring itself remains intact. This is common when high-strength steel rings are used in soft aluminum or plastic housings. The ring 'sinks' into the groove wall, creating a 'ramped' surface. As the load increases, this ramped surface creates a radial outward force that eventually pushes the ring out of the groove. The failure is analyzed using the formula for bearing pressure: $\sigma_b = P / (\pi \cdot D \cdot d)$. If $\sigma_b$ exceeds the housing yield strength, failure is imminent. The fix is to increase the groove depth $d$, use a harder housing material, or use a wider radial wall $b$ on the ring to spread the load.

A snapping sound, often called 'oil canning' or 'buckling,' occurs when a wave spring turn flips or shifts suddenly during compression. This is usually due to a lack of radial guidance or an uneven wave height distribution. When the spring is compressed, the mean diameter expands; if the spring is constrained unevenly, the stress builds up until it is released by a sudden movement. This is detrimental to reliability because the sudden release of energy causes high-frequency stress waves (shocks) that can accelerate fatigue. Furthermore, it results in a non-linear load-deflection curve with significant hysteresis. The solution involves improving the concentricity of the housing and ensuring the spring's waves are uniform within $\pm 2$ degrees of phase.

Fretting corrosion occurs at the contact points between the wave spring and the mating surfaces (or between turns of a multi-turn spring) when subjected to low-amplitude, high-frequency vibration. The small relative motion removes the protective oxide layer (e.g., the Cr2O3 layer on stainless steel), leading to rapid oxidation and the formation of abrasive debris (red rust on carbon steel, black powder on stainless). This debris accelerates wear and can lead to premature failure. In subsea couplings, this is mitigated by using Inconel X-750 and applying a solid-film lubricant like Molybdenum Disulfide (MoS2) or a silver plating. The lubricant reduces the coefficient of friction and prevents the metal-to-metal contact that drives fretting.

Fatigue failure typically manifests as a clean, brittle-appearing fracture originating from a wave crest or trough, where the bending stress is maximum. Scanning Electron Microscopy (SEM) will reveal 'striations'—microscopic ridges each representing one load cycle. In the case of a Crest-to-Crest spring, the crack usually starts at the inner diameter (ID) of the crest because that is where the highest tensile stress occurs during compression. If the fracture surface shows 'beach marks' (macroscopic ridges), it indicates periods of varying load intensity. To prevent recurrence, the stress range $Δ\sigma$ must be reduced, or the surface finish must be improved (e.g., by vibratory finishing) to remove micro-scratches that act as stress risers.

Load loss in the absence of cracks usually points to stress relaxation or 'creep.' In a recent aerospace valve failure, 17-7PH springs lost 15 percent of their preload after 500 hours at 300 C. Detailed metallurgical analysis showed that while 300 C is within the 'rated' range, the combination of high mean stress (70 percent of yield) and temperature accelerated the dislocation movement within the martensitic laths. The solution was to switch to Inconel X-750, which has a higher creep resistance. Another possibility is 'wear' at the contact points (crests); if the crests wear down by just 0.05 mm in a spring with a low spring rate, the load drop can be substantial due to the $P = k \cdot Δx$ relationship.

Permanent set is identified when the free height $H_0$ of the spring after loading is significantly less than the original $H_0$ as manufactured. This is a result of the material being stressed beyond its proportional limit. The most common cause is 'over-compression'—taking the spring too close to its solid height. The stress at solid height $\sigma_s$ should never exceed the yield strength $\sigma_y$. If a design requires the spring to be compressed near solid, a 'preset' operation (compressing the spring to solid during manufacturing) can be performed to 'remove' the initial set and induce beneficial residual stresses. Another cause is 'thermal relaxation' where the operating temperature exceeds the material's limits, causing the microstructure to rearrange and lose its elastic energy.

Using a sharp-edged tool like a chisel or a hardened screwdriver to remove a spiral ring can create a 'nick' or 'gouge' in the groove wall or on the ring itself. If the ring is reused (which is generally discouraged), this nick acts as a massive stress riser $(K_t)$. Under cyclic axial loading, a fatigue crack will rapidly grow from this site. Furthermore, damage to the groove wall can prevent the next ring from seating properly, leading to the 'pop-out' failure described in ID 651. Proper training for field service personnel should emphasize the use of the removal notch and a rounded-edge 'pick' tool to preserve the integrity of the precision-machined groove.

Stress Corrosion Cracking (SCC) is the sudden failure of a ductile material when subjected to a tensile stress in a corrosive environment (like seawater). For 302/304 stainless steel spiral rings, chloride ions penetrate the passive layer, and the residual stresses from the coiling process provide the energy for crack propagation. The ring may literally 'snap' while sitting idle in the groove. Troubleshooting involves looking for 'branching' cracks under a microscope. To prevent SCC, engineers should specify 316 Stainless Steel (for its Molybdenum content) or, for even higher risk environments, nickel-based alloys like Inconel 625, and ensure the rings are properly passivated per AMS 2700.

Fretting fatigue occurs when there is minute, oscillatory relative motion between the retaining ring and the groove wall, typical in splined shafts subjected to torsional vibration. This motion breaks down the protective oxide layer of the ring, leading to 'pitting' and the formation of iron-oxide debris (often appearing as reddish 'cocoa' powder on carbon steel). These pits act as stress concentrators, leading to crack initiation. If a spiral ring fails in this manner, the fracture will show multiple initiation sites. Solutions include increasing the axial preload to 'clamp' the ring in place or using a material with higher galling resistance like Nitronic 60.

'Fly-out' occurs when a ring expands due to centrifugal force and leaves the groove. Forensic signs include the ring being found 'loose' on the shaft, often with significant wear or heat-discoloration on its inner diameter from spinning against the shaft outside of the groove. Unlike a thrust failure, the groove itself might remain undamaged. To confirm this, the engineer should calculate the $N_{max}$ using the formula from ID 622. If the operating RPM was within $20\%$ of $N_{max}$, fly-out is highly likely. The remedy is to specify a 'Self-Locking' spiral ring or to increase the interference fit by reducing the ring's free diameter $D_i$.

When a spiral ring fails below its rated thrust capacity, the root cause is often 'groove deformation' or 'ring dishing.' If the groove material (e.g., aluminum) is soft, the groove wall yields under the axial load, creating a ramp (chamfer). The ring then 'dishes' (cones) and follows this ramp, expanding radially until it clears the groove. This is a common failure mode in lightweight aerospace gearboxes. Analysis involves measuring the groove wall angle after failure. If the angle exceeds $10^{\circ}$, the groove is the culprit. The fix is to either harden the groove (e.g., anodizing or heat-treating) or use a thicker ring to distribute the bearing load over a larger groove area.

Acoustic noise in wave spring assemblies is usually caused by 'stick-slip' friction between the spring and the housing (bore) or between the turns of a nested spring. As the spring deflects, its diameter changes, and if the interface is dry or the surface finish is too rough (e.g., $> 32$ RMS), the spring 'stutters' rather than sliding smoothly. This is common in steering column assemblies. Troubleshooting involves checking for wear marks on the bore wall. The solution is often the application of a dry-film lubricant (like PTFE or $\text{MoS}_2$) to the spring or improving the housing surface finish. In some cases, switching to a different material like Beryllium Copper can reduce noise due to its different friction coefficient.

In applications like automotive valvetrains or high-frequency vibration isolators, if the excitation frequency matches the natural frequency ($f_n$) of the wave spring, resonance occurs. The natural frequency is $f_n = \frac{1}{2 \pi} \sqrt{\frac{k}{m}}$, where $m$ is the effective mass. Resonance causes the waves to oscillate with amplitudes much higher than the design deflection, leading to rapid fatigue failure. Symptoms include 'dancing' of the spring or localized wear on the ID/OD. To fix this, the spring rate $k$ or mass $m$ must be changed to shift $f_n$ out of the operating range, or damping must be introduced, often by using a nested wave spring where inter-turn friction dissipates the vibrational energy.