Retaining Ring Groove Calculator

Groove diameter, width, depth and the ring’s radial cling for DIN 471 external and DIN 472 internal rings, with BOTH thrust limits — the ring’s own shear and the groove material’s yield — computed at either maker’s safety factors, the governing one identified, and the corner radius on the retained part derating it. For DIN 471 geometry the groove almost always governs, which is not what most people expect.

Two limits, and the smaller one wins

Two limits, two makers, and which one actually governs

DIN 471 external rings — every groove dimension

The corner radius on the retained part — how these actually fail

DIN 472 internal rings — the same dimensions, the other way round

What this page refuses to compute, and why

Two limits, and the groove is the one that gives way

A retaining ring has two thrust limits and the smaller one wins — and for DIN 471 geometry the smaller one is almost always the groove. Rotor Clip and Smalley publish the same pair of formulas: the ring’s own shear capacity, π·D·T·S_s divided by a safety factor, and the groove material’s yield, π·D·depth·σ_y divided by another. Both scale with the diameter, so which governs depends only on the ring thickness against the groove depth and on the two safety factors. Set them equal and the crossover needs σ_y/S_s = (T/depth)·(F_groove/F_ring). With Rotor Clip’s factors of 4 and 2 that is (T/depth)/2, and T/depth for DIN 471 runs from 4.0 at Ø3 down to 1.33 at Ø60. So below about Ø30 the crossover would require the SHAFT to yield at a higher stress than the ring shears at, which does not happen — the groove governs whatever you make the shaft of. Above about Ø35 with a hardened shaft, the ring can be the weaker part.

That conclusion reframes the failure mode, and it is worth stating because it is not what people expect. A retaining ring failure is usually the groove wall deforming and letting the ring climb out, not the ring cutting through. Which means a harder shaft buys capacity where a better ring does not — and it means the geometry at the groove’s mouth matters more than the ring’s own strength. The two makers also differ on the safety factor for exactly the term that is least often binding: Rotor Clip use 4 on ring shear, Smalley 3. Both are on the selector and neither is averaged.

The corner radius on the retained part is the largest single effect on this page and the one most often left off a drawing. A square-cornered part turns the thrust into a pure axial load on the ring’s flat face. Put a radius or a chamfer on it and part of that thrust becomes a radial component pushing the ring up the radius — out of a groove that is only two to three per cent of the shaft diameter deep. Rotor Clip publish the derating as a straight proportion, P″ = P′·R_max/R, and their own worked example puts the scale of it beyond argument: a Series HO-100 ring “which abuts a square-cornered part … has a static thrust capacity of 5,950 lbs. The same ring, seated next to a part having the maximum allowable corner radius or chamfer, has an allowable load of 1,650 lbs.” A 72% loss, for a corner nobody dimensioned. Exceed the maximum and you are outside the data entirely, with a ring that can roll out under load.

What holds it in is the radial cling, and this page computes it. The ring’s free inside diameter sits below the groove diameter by (d2 − d3)/2 — 0.35 mm on a Ø30 external ring — and that is the grip resisting anything that tries to lift the ring out. It is also the quantity centrifugal force works against on a rotating assembly, which is why it is printed. There is a second small dimension worth reading: the axial slack, groove width minus ring thickness, which DIN sets at 0.10 mm for rings up to 1.75 mm thick and 0.15 mm above. That is how far the retained part can move before the ring takes load, and on a reversing assembly it turns a steady thrust into an impact every cycle. Bowed and dished rings exist to take up exactly that slack, acting as a light spring at the cost of thrust capacity.

And one refusal, stated rather than glossed. This page does not compute a maximum RPM. The mechanism is real and published — Smalley: “failure happens when these centrifugal forces are great enough to expand and lift the retaining ring from the groove” — but Smalley decline to publish a general formula and refer critical cases to their engineers, and the closed form that IS published elsewhere carries a “factor for number of turns (1 turn = 1.909, 2 turns = 3.407, 3 turns = 4.958)”, which makes it a spiral-wound ring formula rather than a stamped circlip one. Its own worked example also did not reproduce from the equation as transcribed, by a factor of about eighty. Two independent reasons not to print it. DIN 471’s own F_N and F_R load columns are left out for a related reason: no fetch returned them together with the material assumptions and safety factor that make them mean anything, and a load figure with an unknown basis is worse than none on a page whose whole point is which limit governs. For the axial location this ring is an alternative to, the set screw dimensions calculator covers the collar-and-screw route; for a shoulder machined into the shaft instead, the corner radius discussion above is the same geometry; and for locating rather than retaining, the dowel pin and hole calculator.

Frequently asked questions

Related calculators

References

1. DIN 471, Retaining rings (circlips) for shafts, and DIN 472 for bores. Cited by number. The groove columns used here are from Westfield Fasteners’ DIN 471 and DIN 472 specification sheets, with rows 3–6 mm independently confirmed against Aspen Fasteners and rows 10 mm and 50 mm against American Ring. Every one of the 25 rows satisfies groove depth = (d₁ − d₂)/2 exactly, which is how the column mapping was established after a first fetch returned the headers offset by one. Material and hardness from American Metric’s catalogue: C60 or C75 spring steel at 47–54 HRC. The standard’s own F_N and F_R load columns are NOT reproduced — no fetch returned them with the definitions and material assumptions that make them meaningful, and a load figure whose basis is unknown is worse than none.
2. Rotor Clip Company. Formulas: Retaining Ring Load Capacity. The two limits used here, verbatim: ring shear P_r = (G_f × D_s × T × π × S_s) / F_s with “F_s = Safety factor (typically 4)”, and groove yield P_g = (G_f × D_s × d × π × σ_y) / F_s with a safety factor “typically 2”. Also the corner-radius derating, P″_r = (P′_r × R_max) / R, and the worked example that makes the case: “A Series HO-100 ring which abuts a square-cornered part … has a static thrust capacity of 5,950 lbs. The same ring, seated next to a part having the maximum allowable corner radius or chamfer, has an allowable load of 1,650 lbs.” — a loss of 72%.
3. Smalley Steel Ring Company. Load Capacity. The same two limits with different safety factors: P_R = (π × D × T × S_S) / K with “K = Safety factor (3 recommended)” and P_G = (π × D × d × S_Y) / K with 2 recommended, and the instruction to take “the lesser of the two”. Also that the formulas “do not take into account any dynamic or eccentric loading”, and maximum groove-bottom radii of 0.005 in up to 1 in diameter and 0.010 in above. Two named makers, two different safety factors on the same formula: both are printed here rather than averaged.
4. On ring speed: Smalley’s RPM Capacity states the mechanism — “failure happens when these centrifugal forces are great enough to expand and lift the retaining ring from the groove” — and declines to publish a general formula, asking instead that critical applications be referred to their engineers. Design World and Motion Control Tips publish a closed form, but it carries a “factor for number of turns (1 turn = 1.909, 2 turns = 3.407, 3 turns = 4.958)”, which makes it a SPIRAL-WOUND ring formula and not a stamped-circlip one; the worked example printed alongside it did not reproduce from the equation as transcribed, by a factor of about eighty. This page therefore states the mechanism and the quantity that resists it — the ring’s radial cling, (d_groove − d_ring inside)/2 — and refuses to compute a speed limit. See the note beside the results.
5. Aspen Fasteners. Metric DIN 9021 Flat Fender Washers, Metric DIN 433 Flat Washers, Metric DIN 127 Type B Helical Spring Split Lock Washers and Metric DIN 471 External Retaining Rings for Shafts. Distributor specification sheets, used for the printed dimensions those standards are cited for.