Guide
What causes fretting corrosion at a press fit or bearing seat, and what should the drawing change?
Fretting corrosion is surface damage that forms where two loaded metal parts that are meant to stay still relative to each other slip back and forth by a tiny amount, usually under vibration or a rotating load. The slip wipes off the protective oxide, the fresh metal oxidizes, and the oxide debris then grinds the joint further. On a press fit, bearing seat, spline or bolted flange it shows up as reddish-brown or black powder, dull polished patches and a fit that has gone loose.
By Saola Manufacturing engineering team · Published · Updated
Fretting is a joint problem, not a material defect. The same bearing, shaft and housing can run clean in one machine and fret in another because the load, vibration, fit, surface texture and clamp force differ. That is good news for the engineer who has to fix it: nearly every lever that controls fretting can be written on the drawing or the RFQ. This guide shows how to recognize the damage on a returned part, how to rank the likely causes, which fix fits which cause, and what wording stops the problem from coming back on the next lot.
On this page
What does fretting corrosion look like, and how is it different from ordinary rust?
Fretting corrosion sits only on the contact surface of a joint that should not move, and it comes with wear: fine oxide powder, shallow pits or grooves and glazed patches. Ordinary rust forms on exposed surfaces and does not change the size of a fit.
AMPP, the corrosion society, places the damage at the high points where loaded surfaces touch and move against each other again and again, names vibration as the most common driver, and says pits or grooves plus oxide debris are its signature. Its short list of where to look is machinery, bolted assemblies and ball or roller bearings (AMPP, fretting corrosion). It also flags surfaces vibrated during transportation as at risk, which explains damage found on parts that have never run.
The signs that point to fretting rather than general corrosion are:
- Location. The damage is confined to the mating area of a bore, journal, spline, key, flange face or bolt shank, often heaviest at the edges of the contact where slip is largest.
- Debris color. On steel the powder is reddish-brown, often called cocoa, although ordinary steel corrosion product is gray; on aluminum it is black, although ordinary aluminum corrosion product is white (Fitch, Machinery Lubrication). Grease near the joint often turns brown or black with the debris.
- Surface texture. Matte or polished bands, shallow pitting and a “frosted” look in the seat, with the original machining marks partly erased.
- Loss of fit. A bearing ring that went in with a press now slides out by hand, a pulley wobbles, or a bolted joint has lost clamp.
Do not confuse it with false brinelling, which is fretting on the raceways of a bearing that vibrates while standing still: the marks are on the raceways inside the bearing, typically a row of dents spaced like the rolling elements, not on its bore or outside diameter. Both share the root cause, small oscillating motion under load, but the fix for false brinelling is in shipping and standstill conditions rather than in the seat.
What causes fretting corrosion at a press fit or bearing seat?
The root cause is always micro-slip under load at a joint designed to be static. In practice, the most common reasons are a fit that is too loose for the load, a seat that is out of round or too rough, a clamp or preload that is too low, and vibration or thermal cycling that the design did not allow for.
Ranked by how often they explain a fretted seat on a returned part:
- Fit too loose for the load direction. A bearing ring that rotates relative to the load needs an interference fit; with a clearance or light transition fit, elastic flexing of the ring under each pass of the load works the interface and the ring creeps and frets. The guide to clearance, transition and interference fits explains the classes.
- Interference lost to surface texture. A coarse turned seat has peaks that flatten on assembly, so the effective interference is smaller than the measured one, and the remaining grip may not stop slip.
- Poor seat geometry. Lobing, taper or an oval bore means the ring is only held at a few points. The unsupported arcs flex and fret even when the average diameter is in tolerance.
- Low clamp or preload. Bolted flanges, couplings and clamped hubs fret when the transverse load exceeds the friction that the bolt preload can carry. Under-torqued bolts, and soft gaskets or paint in the joint that relax and shed preload, all lower the friction the joint can carry.
- Thermal mismatch. A steel bearing in an aluminum housing loses interference as the housing warms because aluminum expands more than steel; a fit chosen at room temperature can be loose at operating temperature.
- Vibration with no running load. Transport, standby equipment and machines near vibrating plant can fret joints that would be fine in service.
A NASA study of fretting in titanium, Monel-400 and cobalt found three mechanisms acting in sequence: adhesion and plastic deformation first, spall-like fatigue pits after a few hundred cycles, and then oxidation and abrasion by debris particles as the most significant factor; it also found that the hardness of a metal’s oxide relative to the metal predicted how prone it was to progressive damage (Bill, NASA TN D-6660). In a press fit, the practical consequence is that once hard oxide powder is in the joint it keeps abrading the seat, which can loosen the fit further and allow more slip. Fretting also matters beyond appearance. Fatigue cracks can start in the fretted zone, typically near the edge of a press fit on a rotating shaft, and ASTM E2789, the guide for fretting fatigue testing, states that fretting fatigue should be considered a system response rather than a material response, because life depends on both materials, the contact geometry and how load and displacement are applied.
How do you diagnose fretting on a returned part?
Map where the damage is, measure both mating diameters and their form against the drawing, and check how the joint was assembled and loaded. The pattern of damage usually tells you whether the fit, the geometry, the clamp or the environment is to blame.
Work through this checklist before deciding on a fix. It is written so a quality engineer can run it on the bench with ordinary gauges.
- Photograph before cleaning. Record debris color and where it sits. Wiping first destroys the most useful evidence.
- Map the damage. Damage all around a bore points to a ring that creeps under a rotating load. Damage in one arc points to a stationary load zone with too little support, or to an out-of-round seat. Bands at the ends of a press fit point to edge slip from shaft bending.
- Measure both parts. Check bore and shaft diameter at two or more heights and two or more angles, and check roundness and cylindricity on a form tester or CMM where you have one; a two-point micrometer can miss odd-lobed out-of-roundness. Compare against the drawing, then compute the actual interference, not the nominal one.
- Check the surface. Compare the undamaged part of the seat with the drawing’s surface texture requirement. A seat rougher than specified loses grip on assembly.
- Check the assembly record. Was the part heated or pressed, was a lubricant or compound used, and what torque did the bolts get? A dry press with oil on one part is a different joint from the one designed.
- Check the service. Ask about vibration, start-stop cycling, operating temperature and how the machine was shipped.
- Check the lubricant. If an anti-fretting grease or paste was used and the joint still fretted, the slip amplitude is probably too large for a lubricant to manage; lowering friction can even let a marginal joint move more.
- Check for cracks. On shafts, inspect the edges of the fretted band with dye penetrant or magnetic particle before returning the part to service.
If the measured parts are inside the drawing tolerance and the joint still frets, the drawing is the problem, not the supplier. If one part is outside, the fix starts with the supplier’s process and inspection, and the surface finish guide helps when texture is the gap.
How do you fix or prevent fretting corrosion?
Stop the slip first: a tighter fit, a better seat, more clamp or a retaining compound. Where some motion cannot be removed, separate or harden the surfaces with a coating or an anti-fretting lubricant, and never raise interference without checking stress in the mating part.
Use the table to match the fix to the cause found in the diagnosis. Fit classes are written in the ISO 286 hole-and-shaft system; for rolling bearings, take the recommended class from the bearing maker’s catalog for your load direction and magnitude, and see the press fit tolerance guide for the interference each class gives.
| Cause found | First fix | Watch out for |
|---|---|---|
| Ring rotates relative to load, fit loose | Move to the interference class the bearing maker recommends for a rotating load | Excess interference raises hoop stress and cuts bearing life; check internal clearance |
| Seat rough, grip lost on assembly | Specify a finer surface texture on the seat (fine turned or ground) | Apply only to the fit length, not the whole part |
| Bore out of round or tapered | Add roundness or cylindricity tolerance to the seat | Thin-wall housings distort when clamped; machine and inspect clamped if needed |
| Bolted joint slips under side load | Raise bolt preload, add dowels or a shear feature | Torque must be stated for the lubricated or dry condition actually used |
| Small gaps or uneven contact in a cylindrical fit | Anaerobic retaining compound at assembly | Name grade and cure; plan for removal on serviced parts |
| Steel parts with unavoidable micro-motion | Oiled manganese phosphate on one part, verified by test | Coating adds thickness; dimension the fit after coating. MIL-DTL-16232 ties it to wear and break-in, not to fretting by name |
| Aluminum housing or shaft with micro-motion | Hard anodize (Type III) on the aluminum part, or a steel sleeve; verify by test | Anodize grows the part; mask or size the bore for the build-up |
| Fretting only after shipping | Lock or support rotors in transport; anti-fretting grease | Same mechanism as false brinelling; the seat may be fine |
Fit. Interference is the strongest lever, but it is not free. A NASA study of inner-ring interference fits found that the hoop stress they add reduces calculated ball bearing fatigue life, with life factors from 1.00 down to 0.38 in the worst of more than 1,150 cases analyzed (Oswald, Zaretsky and Poplawski, NASA). Go to the class the load needs, not one class tighter for safety, and check that a thin housing or a cast boss can take the stress.
Retaining compounds. Single-component anaerobic adhesives used to lock, seal and retain cylindrical assemblies are classified by ASTM D5363. They fill the gaps between asperities, so the load is shared over the whole seat and the joint resists slip. Name the grade on the drawing or assembly instruction, and remember that a high-strength grade may need heat to remove a bearing during service.
Coatings. Heavy manganese or zinc phosphate for ferrous parts is covered by MIL-DTL-16232, and the specification associates the oil-holding manganese type with wear prevention and break-in of bearing surfaces. It does not name fretting, so treat a phosphate seat as an option to prove on test parts rather than a guaranteed cure. Hard anodize on aluminum is Type III under MIL-A-8625, now issued as MIL-PRF-8625F with the same types (see the MIL-A-8625 types guide). Both change size, so the anodizing dimensional change guide applies to any coated fit.
What should the drawing and RFQ say so fretting does not recur?
Put the fit class on both mating parts, add form and surface texture to the seat, name any coating with where it applies and whether the fit is measured after finishing, and state the assembly method, compound and torque basis. A bare diameter with a general tolerance leaves every one of those choices to someone else.
- Fit. Write the class on each part, such as a bearing seat on the shaft in the class the bearing maker recommends and the housing bore in its own class, not only “press fit.”
- Form. Add a roundness or cylindricity tolerance to the seat, tighter than the size tolerance, and a perpendicular or runout control on the bearing shoulder.
- Texture. Give an Ra requirement on the seat only, with the surface texture symbol, so the supplier does not grind the whole part.
- Coating. “Manganese phosphate per MIL-DTL-16232, seat dimensions apply after coating,” or “Hard anodize per MIL-A-8625 Type III, mask bore,” whichever matches the design.
- Assembly. Name the press or shrink method, the retaining compound grade if any, and the bolt torque for the actual lubrication condition, or control by clamp load.
- Inspection. Ask for the seat diameter at two heights and two angles, and form results where specified, on the first article and per lot.
On the RFQ, say that the change is a fretting fix and send a photo of the damaged part. A supplier who knows why the seat changed can flag a conflict, such as a coating that eats the new interference, before tooling rather than after the next return.
Who does what on a Saola order for a fretting fix?
Shafts, bores, bearing seats and flanges are turned or milled at our lead CNC plant, T.K Viet Nam Precision Company Limited in Vietnam. Any coating, such as anodizing, is applied at a partner plant we manage, and the quote names the production site and country of origin for each step.
Seats are machined at the lead plant, where the inspection equipment includes a 2D vision measuring machine, height gauge and hardness tester. State on the RFQ which seat diameters and form controls need measured results reported and how form is to be verified, so the quote can confirm how each one will be checked before the order is placed. See CNC turning and CNC milling for the part families that run there. The surface finishing menu at partner plants includes anodizing on aluminum; manganese phosphate is not on that menu today, so if your fix depends on it, say so on the RFQ and we will tell you before quoting whether it can be arranged. If your own line does the assembly and applies any retaining compound, the drawing only needs to state seat sizes after any coating. ISO 9001 certification is in progress, not held. Send the revised drawing, the fretted part photos and the bearing or mating-part data through the quote form.
Frequently asked
Is fretting corrosion the same as fretting wear?
They are the same process seen from two sides. Fretting wear is the material loss from small-amplitude slip; fretting corrosion is the name used when oxidation of the debris is a large part of the damage, which is the usual case in air on steel and aluminum.
Can grease alone stop fretting at a bearing seat?
Rarely on its own. An anti-fretting paste slows the damage where some motion is unavoidable, but lubricants reduce fretting rather than stop it, and lower friction can let a marginal joint slip more. If the fit is too loose for a rotating load the ring will keep creeping. Fix the fit first and use the paste as a second line.
Can a fretted shaft or housing be reused?
Only after measurement and crack inspection. Light staining on a seat that is still in size can sometimes be cleaned and reused with a retaining compound, but a seat out of size or with pits at the fit edges should be repaired, sleeved or replaced.
Why does fretting show up on parts that have never run?
Vibration in transport is enough. AMPP lists equipment vibrated during transportation as at risk, and bearings shipped in assembled machines can show false brinelling. Lock rotors and support heavy parts in the crate.
Does a harder or coated seat always cure fretting?
No. Hardening and coatings slow wear, but if the joint still slips the damage moves to the softer part or the coating wears through. Treat coatings as protection for motion you cannot remove, not as a substitute for a correct fit and clamp.
Sources
- Fretting corrosion, AMPP (Association for Materials Protection and Performance), accessed October 2026
- Fretting wear in titanium, Monel-400, and cobalt 25-percent-molybdenum using scanning electron microscopy, NASA TN D-6660, R. C. Bill, NASA Lewis Research Center, 1972
- Interference-Fit Life Factors for Ball Bearings, NASA/TM-2010-216913, F. B. Oswald, E. V. Zaretsky and J. V. Poplawski, NASA Glenn Research Center, 2010
- ASTM E2789-10(2021), Standard Guide for Fretting Fatigue Testing (scope), ASTM International, 2021
- ASTM D5363-16, Standard Specification for Anaerobic Single-Component Adhesives (scope), ASTM International, 2016 (reapproved 2024)
- MIL-DTL-16232H, Phosphate Coating, Heavy, Manganese or Zinc Base, US Department of Defense, 2020
- MIL-A-8625F, Anodic Coatings for Aluminum and Aluminum Alloys, US Department of Defense, 1993 (superseded by MIL-PRF-8625F, 2020)
- Fretting Wear in Lubricated Systems, E. C. Fitch, Machinery Lubrication, accessed October 2026
- ISO 286-2:2010, Tables of standard tolerance classes and limit deviations for holes and shafts, ISO, 2010