Installasjon og montering

When installing an internal retaining ring deep within a housing, the bore must have a lead-in chamfer (typically $15^∘$ to $20^∘$ and at least 1.5 times the ring thickness in length). This chamfer acts as a natural compressor for the ring. Without a proper chamfer, the sharp edge of the ring will catch on the bore entrance, causing the ring to 'cock' and potentially scoring the precision-honed surface of the bore. In hydraulic systems, this scoring can create leak paths for high-pressure fluid. The chamfer should be smooth and free of burrs to ensure the ring slides effortlessly to its destination groove.

Self-locking rings feature a small 'tab' on one turn and a 'slot' on the other. During installation, the ring is expanded or compressed like a standard ring, but once it seats in the groove, the installer must ensure the tab 'clicks' into the slot. This often requires a final axial 'set' with a tool. Once locked, the ring cannot expand radially, making it ideal for high-RPM applications. However, this means the ring is much harder to remove; usually, the locking tab must be manually disengaged with a specialized tool. In automated assembly, sensors must verify the 'lock' state by measuring the final $OD/ID$ of the ring.

'Snap-Back' occurs when an internal ring, which is being compressed for insertion into a bore, is released too quickly or is not fully constrained, causing it to spring back violently to its free diameter. This can cause injury or damage to the bore's lead-in chamfer. To avoid this, technicians should use a 'tapered sleeve' (the inverse of a mandrel). The ring is compressed into the large end of the sleeve and then pushed through the sleeve into the bore. This provides continuous radial constraint. For manual installation, ensuring that the ring's gap is correctly oriented and using a controlled-compression tool is essential for safety and part integrity.

Unlike stamped circlips, which have 'ears' with holes for pliers, spiral retaining rings are low-profile and have no protruding lugs. To facilitate removal, a small 'removal notch' (or slot) is provided on one end of the ring. This allows a technician to insert a screwdriver or dental-style pick under the end of the ring and pry it out of the groove. In aerospace applications, where 'FOD' (Foreign Object Debris) is a concern, the notch design must ensure that the ring end does not snap off during removal. Furthermore, the notch must be positioned so it doesn't create a stress riser in a high-load area of the ring.

A tapered mandrel is used to expand the ring gradually as it is pushed onto the shaft. The leading diameter of the mandrel should be slightly larger than the shaft diameter, and the taper angle should be between $3^∘$ and $5^∘$. A steeper angle increases the risk of over-stressing the ring (permanent set), while a shallower angle makes the tool excessively long. The mandrel surface must be hardened (HRC 58-62) and polished to a mirror finish ($R_a < 0.2 μm$) to minimize friction and prevent galling. For multi-turn rings, the mandrel must also account for the 'winding' motion as the ring expands, often requiring a helical lead-in.

In mechanical seals, the wave spring must maintain a precise face pressure. The preload $P$ is sensitive to the installed height $L_i$. Due to the stack-up of tolerances in the seal housing, gland, and carbon face, $L_i$ can vary significantly ($ΔL$). Since $P = k \cdot (L_0 - L_i)$, any variation in $L_i$ results in a variation in $P$. If $k$ is high, even a small $ΔL$ can cause the seal to leak (low $P$) or wear prematurely (high $P$). Designers often select Crest-to-Crest springs with a 'flatter' spring rate (lower $k$) over a longer travel to minimize the sensitivity of the preload to these unavoidable manufacturing tolerances.

Installing nested wave springs into deep blind holes requires a dedicated insertion tool to prevent the turns from tangling or 'shingling' (where layers overlap incorrectly). The tool should be a plunger with a diameter slightly smaller than the spring $ID$. The spring is pre-compressed onto the tool, then inserted. It is vital to ensure that the individual layers are seated flush against the bottom of the hole. For high-volume automotive assembly, automated pick-and-place systems use a vacuum-assisted mandrel that holds the spring by its $ID$, ensuring that the multiple turns remain perfectly concentric during the high-speed insertion into the transmission housing.

For a bore-piloted wave spring, the outside diameter ($OD$) of the spring is the primary datum. The housing bore must be sized to accommodate the $OD$ at its maximum radial expansion (at solid height). If the bore is too tight ($D_{bore} < OD_{max}$), the spring will 'hoop' and seize, causing the spring rate to spike. Conversely, if the bore is too loose, the spring can shift off-center, leading to an uneven load $P$ across the wave peaks. In precision assemblies, a tolerance of H7/h6 is often targeted for the bore, and the spring $OD$ is specified with a tolerance that accounts for the manufacturing variation of the flat wire width $b$.

Springs with a high $L_0/D$ ratio (typically $> 1.5$) are prone to buckling during installation and operation. When compressed, the spring acts like a slender column. If not properly constrained by a bore or guided by a pilot (shaft), the spring will bow laterally, leading to non-axial loading and catastrophic failure of the waves. The critical buckling load is defined by $P_{crit} = π^2 \cdot E \cdot I / (K \cdot L)^2$. To prevent this, a housing bore should be used with a clearance of approximately 0.05mm to 0.15mm per side. Additionally, if the spring is used in a dynamic system, the guiding surface must be hardened to prevent the 'snaking' spring from wearing into the housing.

In high-speed bearing applications (e.g., turbochargers), the interface between the wave spring and the bearing race is critical. 'Plain ends' terminate at a wave peak, which can create localized 'point' loading, potentially leading to race distortion or uneven wear. 'Shim ends' (or flat ends) are integrated into the final 360 degrees of the spring, providing a flat parallel surface. This ensures that the spring force is distributed evenly over the entire circumference of the bearing race ($360^∘$ contact). This uniform preload is essential for preventing ball skidding and maintaining the stiffness of the spindle assembly at high RPMs, though it does slightly increase the spring's solid height.

Installing a retaining ring in a blind hole (where the bottom of the bore is not accessible) requires an internal spiral ring and a deep-reach installation tool. The tool must compress the ring sufficiently to clear the bore diameter while guiding it to the groove depth. A 'plunger and sleeve' setup is typically used. One challenge in blind holes is ensuring that no debris is trapped in the groove, which would prevent the ring from seating. Furthermore, removal from a blind hole is significantly more difficult; a removal notch is mandatory, and special 'puller' tools may be required to hook the notch and unwind the ring from deep within the assembly.

A spiral retaining ring is designed to work against a sharp-cornered mating part. If the retained component has a large radius or chamfer at its base, it will contact the ring at a point further from the groove wall. This increases the 'lever arm' of the thrust load, creating a large bending moment that encourages the ring to 'dish' and fail. If a radius is unavoidable (e.g., due to stress concentrations on a shaft), a hardened 'back-up washer' with a sharp corner should be placed between the radiused part and the retaining ring. This ensures the load is transferred to the ring as close to the groove wall as possible, maximizing the shear capacity.

Spiral retaining rings often sit flush within the groove, making them difficult to remove without damage to the housing. To facilitate maintenance, rings are designed with a 'Removal Notch' (a small cutout on one end) or a 'Scalloped End'. A technician can insert a screwdriver or dental pick into the notch and pry the end of the ring out of the groove, after which it can be unwound. In aerospace applications, where disassembly must be non-destructive, the choice of end configuration is vital. Without a notch, the ring may need to be destroyed to be removed, risking damage to the precision-machined groove walls.

Ensuring a spiral ring is fully seated is critical, as a partially seated ring will fail at a fraction of its rated load. Automated verification can be achieved through: 1) Laser profiling, which checks the axial position and 'flatness' of the ring; 2) Vision systems that look for the characteristic 'gap' or the overlap of the multi-turn ends; or 3) A 'push-back' test where a calibrated axial force is applied to the component being retained to ensure it doesn't move. In some cases, acoustic emission sensors can detect the 'click' of the ring snapping into the groove. For safety-critical parts, the groove diameter itself must be inspected via air-gauging prior to ring installation.

Spiral retaining rings are unique because they can be installed by 'winding' them into the groove. Unlike stamped rings that require circlip pliers and expand/contract significantly, a spiral ring can be started by hand. One end of the ring is placed in the groove, and the remainder is wound in by pressing it axially. For larger rings or high-volume production, a tapered mandrel (for external rings) or a tapered sleeve (for internal rings) should be used. The taper allows the ring to gradually expand/contract as it is pushed into position by a plunger. This method is superior because it ensures the ring is never over-stressed beyond its yield point, which is a common risk when using pliers on traditional snap rings.

At high rotational speeds, centrifugal forces can cause a wave spring to expand radially and shift off-center, leading to dynamic imbalance and vibration. To prevent this, the assembly should incorporate a 'pilot'—either a step in the shaft or a recess in the housing—that captures the spring's $ID$ or $OD$ respectively. In Crest-to-Crest springs, the centrifugal expansion is more pronounced at the middle turns. For speeds exceeding $3600$ RPM, engineers should calculate the 'lift-off' speed where the centrifugal force $F_c = m r \omega^2$ exceeds the radial stiffness of the spring. If lift-off occurs, the spring may rub against the bore, causing heat and potentially failing the assembly. Specialized 'Nested' springs are often more stable at high speeds due to their tighter radial footprint.

In slip clutches, wave springs provide the normal force $F_n$ required to generate friction torque $T = \mu F_n R_e n$, where $\mu$ is the friction coefficient, $R_e$ is the effective radius, and $n$ is the number of friction surfaces. The installation must ensure that the spring applies a perfectly axial load. Any non-parallelism in the mating plates will cause uneven wear and 'chatter'. Because slip clutches generate significant heat, the wave spring must be isolated from direct contact with the friction material if possible, or manufactured from a high-temperature alloy like Inconel X-750. Furthermore, the spring rate should be chosen to be relatively flat (low $k$) to maintain consistent torque even as the friction linings wear down and the spring's operating height increases.

When stacking single-turn wave springs in series to increase deflection, they must be oriented such that the waves are 'in-phase' (peak-to-peak) or using a Crest-to-Crest multi-turn design. If single-turn springs are randomly oriented, the wave peaks may slide into the valleys of the adjacent spring (nesting), which would dramatically increase the spring rate and reduce the total deflection. For high-vibration environments, using a multi-turn Crest-to-Crest spring is preferred over a stack of single-turn springs because the integral construction eliminates the risk of component misalignment and frictional wear between the individual spring interfaces.

Load relaxation, or 'set', occurs when the internal stresses in the spring exceed the material's elastic limit during the first few cycles of compression. This results in a permanent reduction in the spring's free height. To ensure stable performance in the field, manufacturers often 'preset' or 'remove set' from the springs by compressing them to their solid height or the maximum operating deflection during production. This induces beneficial residual compressive stresses on the outer fibers. After presetting, the spring will maintain a constant load-deflection curve during subsequent cycles, provided the operating stress does not exceed the newly established elastic limit. This is critical for precision bearing preload applications where a constant force is required.

As a wave spring is compressed, its mean diameter $D_m$ expands. The theoretical expansion $\Delta D$ can be estimated as $\Delta D = \frac{0.05 f^2}{D_m Z^2}$, where $f$ is the deflection. If the clearance between the spring's Outer Diameter ($OD$) and the housing bore is insufficient, the spring will bind, leading to an erratic spring rate and potential localized buckling. For a standard internal application, the bore diameter should be at least $102-105\%$ of the spring's free $OD$. Similarly, for external applications over a shaft, the shaft diameter should be at most $95-98\%$ of the spring's free $ID$. These clearances also provide the necessary volume for lubricants in high-speed rotating assemblies like clutch packs.