Pertanyaan Umum

Load consistency in wave springs is highly sensitive to the installed work height $H_w$. Because the spring rate $k$ is linear in its design range, any variation in the housing or shaft axial dimensions directly affects the load $P = k(L_{free} - H_w)$. In a Crest-to-Crest assembly, cumulative tolerances of the waves and the alignment of peaks can introduce a 'shimming' effect. Engineers must specify a 'load at work height' rather than just a free height to ensure the assembly maintains the required preload, especially in bearing preloading where a deviation of $\pm 10\%$ in load can significantly alter bearing life.

17-7PH (AMS 5678) is a precipitation-hardening stainless steel that offers superior yield strength and fatigue resistance compared to 300-series alloys. In the CH900 condition (cold reduced and aged at $900^{\circ}F$), the material achieves a high tensile strength exceeding 200 ksi. This allows for higher operating stresses and lower relaxation at elevated temperatures up to $650^{\circ}F$. Unlike 302 stainless, 17-7PH maintains its elastic modulus $E \approx 29.5 \times 10^6$ psi more consistently under thermal cycling, making it ideal for critical aerospace actuators and valve seals.

The spring rate $k$ for a single-turn wave spring is primarily determined by the material properties and geometry. For a given load $P$ and deflection $f$, the relationship is expressed as $k = \frac{E b t^3 N^4}{R^3 D_m}$, where $E$ is the Modulus of Elasticity, $b$ is the radial width, $t$ is the material thickness, $N$ is the number of waves, and $D_m$ is the mean diameter. Note that for precision applications, the stress $\sigma$ must also be verified using $\sigma = \frac{3 \pi P D_m}{4 b t^2 N^2}$. In multi-turn Crest-to-Crest designs, the rate is divided by the number of active turns $Z$, assuming the waves are perfectly aligned to act in series.

Impact loading (or dynamic loading) is significantly more dangerous for retaining rings than static loading. When a mass hits the retained part, the kinetic energy $E_k = \frac{1}{2}mv^2$ must be absorbed by the ring and the groove. This creates a transient load $P_{peak}$ that can be many times higher than the static weight. The safety factor $K$ must be increased from 3 (standard static) to 5 or 10 for impact applications. Failure under impact often looks like 'groove deformation' because the peak force exceeds the yield strength of the housing for a fraction of a second, causing the ring to dish and pop out. Engineers should use 'Heavy Duty' series rings (thicker $T$ and deeper $d$) and verify the design using the impulse-momentum theorem ($F \cdot Δt = m \cdot Δv$).

Unlike stamped rings with holes for pliers, spiral rings are removed by prying the end out of the groove. A 'removal notch' is a small cutout at one end of the ring that allows a screwdriver or dental pick to get behind the wire. In some high-performance applications, 'scallops' or multiple notches are added to allow for removal in tight spaces where the primary end might be inaccessible. For assemblies that require frequent maintenance (e.g., aerospace gearboxes), the inclusion of a removal notch is a critical 'design for service' feature. Without it, the ring may be impossible to remove without damaging the shaft or the groove, leading to costly component replacements.

Spiral rings are manufactured by coiling cold-rolled flat wire. This cold-working increases the tensile strength through strain hardening, but it also leaves high levels of residual tensile stress on the outer diameter and compressive stress on the inner diameter. If the ring is used as-coiled, these stresses can lead to dimensional instability. Heat treatment (stress relieving or age hardening) allows the atoms to rearrange into a lower-energy state, 'locking' the ring into its coiled shape. For carbon steel, this is a tempering process; for 17-7PH, it is precipitation hardening. The resulting ring has a more uniform stress profile, which improves its resistance to fatigue and ensures that it maintains its 'cling' diameter over the life of the product.

For a spiral retaining ring, the moment of inertia $I$ of the cross-section is calculated as $I = \frac{b \cdot t^3}{12}$, where $b$ is the radial wall width and $t$ is the thickness. However, the radial stiffness of the ring—its ability to resist expansion or contraction—is also governed by the 'curved beam' theory. The stiffness is proportional to $E \cdot I / R^3$. This means that even a small increase in the radial wall $b$ significantly increases the force required to install the ring and its ability to remain in the groove. Designers must balance the radial wall width $b$ to ensure the ring is stiff enough to hold the load but flexible enough to be installed without exceeding the yield strength $\sigma_y$ during expansion.

Stress relaxation is the decrease in stress (and thus load) in a spring held at a constant deflection over time. At $500^{\circ}F$, even though 17-7PH is within its operating range, the thermal energy allows for micro-plastic deformation via dislocation climb and grain boundary sliding. The amount of relaxation is a function of the initial stress level and time. For instance, a spring stressed to $100$ ksi at $500^{\circ}F$ might lose $5-10\%$ of its load within the first 100 hours. If the assembly depends on a specific preload to prevent vibration-induced wear, this loss can be catastrophic. Engineers must compensate for this by either over-designing the initial load or selecting a more creep-resistant material like Inconel X-750 for high-temperature service.

'Oil canning' or 'snap-through' is a form of elastic instability that occurs in single-turn wave springs when the ratio of the wave height $h$ to the material thickness $t$ is too high. If the spring is compressed, it may reach a point where it 'snaps' into an inverted shape or a flattened state with a sudden drop in load. This is mathematically similar to the buckling of a shallow arch. To avoid this, designers should keep the $h/t$ ratio within a range where the spring rate remains positive. If the rate $dk/df$ becomes negative, the spring is unstable. This is a critical failure mode in switch mechanisms where a positive tactile 'click' is required, but it is a failure in structural preload applications.

Wave springs are coiled from flat wire that has been cold-rolled. This rolling process aligns the grain structure of the metal in the longitudinal direction. When the wire is coiled into a spring, the bending stresses occur along the length of the wire. Because the material is stronger and more ductile along the grain than across it, the longitudinal grain orientation is ideal for the cyclic bending that wave springs undergo. If the springs were stamped from sheet metal (where grain direction varies relative to the wave geometry), they would be more susceptible to cracking and inconsistent spring rates. The coiling process ensures that every wave in the spring has the same optimal grain orientation, maximizing fatigue life and load consistency.

The fundamental difference lies in the arrangement of the waves. In a Crest-to-Crest spring, the waves are in 'series', meaning the total deflection is the sum of the deflections of each turn. The rate is $k_{series} = \frac{k_{single}}{Z}$. In a Nested spring, the turns are in 'parallel', meaning each turn experiences the same deflection and the total force is the sum of the forces. The rate is $k_{parallel} = n \cdot k_{single}$. This means that for the same space envelope, a Nested spring will be many times stiffer than a Crest-to-Crest spring. Specifically, if a spring has 3 turns, the Nested version is 9 times stiffer than the Crest-to-Crest version ($3$ in parallel vs $1/3$ in series). This makes Nested springs suitable for extremely high loads with small deflections, while Crest-to-Crest is for large deflections with lower loads.

Shear failure occurs when the axial load $P$ exceeds the shear strength of the ring's cross-section. Forensic evidence of shear failure includes a clean, polished appearance on the sheared surface of the ring, often with a distinct 'step' between the part of the ring that remained in the groove and the part that was forced out. This differs from 'dishing' failure, where the ring would show signs of bending and twisting. Measurement of the groove depth $d$ and the ring thickness $T$ after the event is crucial. If the ring sheared, it often indicates an unexpected impact load or a significant underestimate of the static thrust. One must also inspect for 'galling' on the ring faces, which suggests high-pressure sliding occurred before the final shear event.

The edge margin is the distance between the groove and the end of the shaft or bore. If the edge margin is too small, the housing material will fail in shear or 'blow out' when the ring is loaded axially. The required edge margin $z$ is typically calculated as $z = \frac{3 \cdot P}{\pi \cdot D \cdot \sigma_y}$, where $P$ is the thrust load and $\sigma_y$ is the yield strength of the housing material. A general guideline for steel housings is an edge margin of $3 \cdot d$ (three times the groove depth), while for aluminum housings, $4 \cdot d$ or $5 \cdot d$ is recommended. Failure to provide sufficient edge margin will lead to a catastrophic failure of the assembly, even if the retaining ring itself is intact.

Elgiloy (a cobalt-chromium-nickel alloy, UNS R30003) offers the highest level of corrosion resistance for spiral retaining rings. It is virtually immune to hydrogen embrittlement and has exceptional resistance to stress corrosion cracking in sulfide and chloride environments. Mechanically, Elgiloy provides a very high modulus of elasticity ($28.5 \times 10^6$ psi) and can be heat-treated to a hardness of 45-55 HRC. It maintains these properties from cryogenic temperatures up to $850^{\circ}F$. This makes it ideal for downhole oil tools and chemical pumps where 17-7PH or 316SS would fail due to chemical attack. The primary disadvantage is the high material cost, which usually limits its use to critical safety-of-flight or mission-critical subsea components.

The radial expansion $\Delta r$ of a retaining ring due to rotation is given by $\Delta r = \frac{\rho \cdot \omega^2 \cdot R^3}{E}$, where $\rho$ is the density, $\omega$ is the angular velocity, and $R$ is the mean radius. The ring will fail when $\Delta r$ exceeds the groove depth $d$. More specifically, the maximum speed $N$ (in RPM) is calculated as $N = \sqrt{\frac{440 \times 10^6 \cdot E \cdot I}{w \cdot \rho \cdot R^4}}$, assuming the ring is made of steel. If the application requires a speed higher than this limit, a 'Self-Locking' ring or a heavier cross-section ring must be used. Additionally, the fit must be tight; a loose ring will start to expand at a lower RPM than a ring with high initial 'cling' (interference).

'Permanent set' is an immediate plastic deformation that occurs the first time a spring is compressed to its work height or solid height if the internal stress exceeds the material's elastic limit. It results in a permanent loss of free height. 'Relaxation', on the other hand, is a time-dependent loss of height and load that occurs under constant stress, especially at elevated temperatures (creep). Both result in a lower-than-intended preload. To minimize permanent set, manufacturers often 'preset' the springs by compressing them to solid height during production. This induces beneficial residual stresses (autofrettage) that allow the spring to operate at higher loads without further deformation in the field.

Measuring the work height $H_w$ requires a precision load tester. The spring is compressed to the specified work height $H_w$, and the resulting force $P$ is recorded. It is incorrect to measure the free height $L_f$ and then subtract the deflection, as the spring rate is often slightly non-linear in the first $20\%$ and last $20\%$ of the stroke. The load should be measured during the 'compression' stroke to avoid the influence of hysteresis. For high-precision springs, the 'shimmed' ends must be perfectly flat against the tester's plates. Any deviation in the parallelism of the testing plates will result in an inaccurate load reading due to uneven wave compression.

Beryllium Copper (typically Alloy 25, UNS C17200) is used for wave springs in electronics due to its excellent electrical conductivity and non-magnetic properties. Beyond conductivity, it can be heat-treated to reach tensile strengths up to 200 ksi, comparable to some steels. This allows the spring to function as both a high-force mechanical contact and an EMI/RFI shield. In medical or scientific equipment like MRI machines, its non-magnetic nature is essential. The material is typically aged at $600^{\circ}F$ to achieve its maximum spring properties. However, designers must be cautious of its cost and the environmental regulations regarding beryllium handling during the manufacturing phase.

The number of waves $N$ has a dramatic effect on the spring's performance. The spring rate $k$ is proportional to $N^4$ ($k \propto \frac{E \cdot b \cdot t^3 \cdot N^4}{D_m^3}$). Adding waves significantly increases the stiffness. However, as $N$ increases, the allowable deflection $f$ decreases because the physical distance between wave crests becomes smaller, leading to steeper wave angles and higher bending stresses for the same amount of deflection. The stress $S$ is proportional to $1/N^2$. Therefore, a spring with more waves can carry a higher load $P$ but at the cost of reduced travel. Engineers must balance $N$ to achieve the required force $P$ within the available axial space without exceeding the material's yield strength.

'Ring walking' is a phenomenon where the retaining ring slowly rotates and eventually moves out of its groove. This is usually caused by 'precession' due to a combination of vibration and a loose fit between the ring and the groove. If the ring's free diameter is not small enough to provide sufficient 'cling' on the shaft, or if the groove is oversized, the ring can vibrate. Under cyclical axial or radial loads, the ring may rotate slightly each cycle. Over time, this can wear the groove edges or the ring itself. Troubleshooting involves checking the 'cling' (interference fit) of the ring and ensuring the groove width $W$ is within the specified tolerance (typically $0.003$ to $0.005$ inches wider than the ring thickness $T$).