How to Replace a Bearing Model: Decode, Match, Verify

Can you read the designation?

A legible designation is the cheapest identification route you have. It encodes the bearing type, the dimension series, the bore code and the prefix/suffix options such as seals, clearance and precision, so the type, series and bore are recoverable without touching the machine^5,11,13.

The bore code follows a fixed rule: 00, 01, 02 and 03 mean 10, 12, 15 and 17 mm, and from 04 upward you multiply the code by five^11,13 (supplier-reported figure). A 6204 therefore decodes to a 20 mm bore in the 62 series, and a 6205 to 25 mm^11,13 (supplier-reported figure).

Bearings with a bore of 500 mm or larger use the actual bore size directly in the number instead¹¹. A worn or polished-off marking changes the whole route.

Seal, cage, clearance and tolerance codes cannot be recovered by physical measurement, so once the printed designation is gone a cross-reference has to take over¹².

You can still measure the bore, outside diameter and width with a caliper, and that narrows the candidate list to a handful of designations, but it cannot complete the match on its own¹².

The same 6204 with a polished-off marking yields only d, D and B from a caliper and forces the cross-reference route¹² (supplier-reported). Before you spend time on either route, check the marking under good light and record every character, including prefixes.

If any suffix is missing or illegible, treat the bearing as unidentified and start measuring.

Will it physically enter the housing?

The dimension series is the second digit group of the designation, and it sets the outside diameter at a given bore. A series mismatch produces a bearing that will not fit the housing no matter how similar the part number looks, so this check comes before any performance comparison⁵.

At a 20 mm bore, the 60, 62 and 63 series give 42 mm, 47 mm and 52 mm outside diameters respectively⁵ (supplier-reported). A 6204 cannot be dropped into a 6304 housing, and a 6205 and 6305 differ in outside diameter for the same reason^5,14 (supplier-reported).

Do clearance, seals, cage and class match?

Substitution is safe only when bore, outside diameter, width, internal clearance, sealing or shielding, cage (under demanding conditions), precision class and all other technical specifications match; part-number similarity alone is not sufficient¹⁴.

The clearance designation (C2, CN/normal, C3, C4) and the sealing configuration (open, shielded, sealed) are separate attributes from the boundary dimensions, and both must be compared¹⁴. Where operating conditions are demanding — high temperature, corrosive environment or high speed — the cage material must match too, not just the dimensions and clearance¹⁴.

Precision class is the attribute buyers most often misread. ABEC/ISO class is a tolerance band specification, not a quality endorsement, and it does not control noise, lubrication, ball grade, radial internal clearance, material, heat treatment, cage quality, speed rating or load capacity¹⁹.

One source reports that ABEC 1 serves approximately 90% of industrial applications, so a higher class is a tighter band rather than an upgrade¹⁹. A higher ABEC number does not compensate for an uncontrolled clearance, cage or material¹⁹.

If perfect alignment cannot be guaranteed, the sources point toward self-aligning designs or more generous internal clearance². For a demanding high-temperature or high-speed position, the cage material must match, and that is the case that must go to the equipment OEM or a bearing specialist before substituting¹⁴.

If your application is high-speed, high-temperature or safety-critical, put that question to the OEM in writing before you accept any candidate.

Is the load rating adequate?

Dimensional fit says nothing about whether the candidate will survive the load. Ball bearing fatigue life follows a cubic relationship with load: doubling the equivalent dynamic load reduces calculated L10 life to about one-eighth of its original value⁹.

That is why undersizing is penalized so heavily compared with sizing up, and why a lower-rated candidate must be rejected rather than accepted on dimensional fit alone⁹.

The L10 calculation is L10 = (C / P)³ × 1,000,000 revolutions, where C is the basic dynamic load rating and P is the equivalent dynamic bearing load accounting for both radial and axial components⁹.

Load ratings vary slightly between production standards, so exact dimensions and ratings must be confirmed against the current manufacturer datasheet before ordering⁵. That means the candidate's C and C0 values have to come from the supplier's current data, not from a catalogue page you found years ago⁵.

Where failure is costly or safety-critical, target a higher reliability than the default 90% in the L10 calculation² (supplier-reported). Work the numbers before the RFQ.

If your original bearing ran at a given equivalent dynamic load and the candidate's rating is lower, the calculated life drops by the cube of the load ratio — a 20% lower rating is not a 20% shorter life (supplier-reported).

Ask the supplier for the current C and C0 values and the limiting speed, and recalculate with your actual equivalent load.

Is a cross-brand number really equivalent?

Part-number similarity is not interchangeability. A 6205 and a 6305 look like neighbours in the same series but have different outside diameters, so the number alone cannot be treated as an equivalence claim¹⁴.

Deep groove and angular contact bearings are not interchangeable either: they differ in load direction capability, contact angle and application suitability³.

Deep groove bearings accept combined radial and moderate axial load in both directions and suit cases where space does not allow matched sets, while angular contact bearings are used for high speed, high rigidity and precision^4,9.

Because no public cross-reference table is available in the evidence, equivalence has to be established by comparing the candidate's datasheet against the original on every controlled attribute rather than by trusting a lookalike number¹⁴.

Where the designation is unreadable, the route is measurement of d, D, B/T and visible features, then a cross-reference to narrow the candidate list¹². Ask the supplier to state, in writing, which attributes of the candidate match the original and which differ.

If the supplier cannot name the differences, you do not have an equivalence claim — you have a guess.

Which sourcing route should you take?

OEM parts guarantee the exact specification and preserve warranties, but they carry a price premium and longer lead times; authorized distributors offer a middle ground with verified provenance²³. Standard in-stock bearings ship faster than custom or non-standard items, so the route you choose sets your lead time before the RFQ goes out²⁴.

One source reports OEM components command a 20% to 40% price premium, and one supplier quotes 3-7 days for standard stock against 25-45 days for custom manufacturing^23,24.

What documentation must the supplier provide?

If certification and material traceability are not confirmed before the order, you have no way to prove the replacement was made from the specified steel or process. Counterfeit bearings made with inferior steel can reduce fatigue life by up to 80%, and the difference is not visible on arrival²³ (supplier-reported).

One source estimates counterfeit bearings cost the global industrial sector about $3 billion annually in unplanned downtime and replacement costs²³. The checks below are the ones the evidence names, and they belong in the RFQ rather than in a post-delivery argument.

• ✓Ask for ISO 9001 manufacturing certification before the quotation is issued.
• ✓Request Material Test Reports confirming steel cleanliness and alloy composition.
• ✓If the bearing is destined for an explosive atmosphere or a specialised environment, require ATEX or RoHS compliance documentation with the quote.
• ✓For critical-path motors, ask whether 100% vibration and noise testing at the factory is included and how it is documented.
• ✓Confirm the supplier can trace a batch back to raw material and production date.

Can the supplier sustain the match?

For a one-off replacement, the first delivery is the whole order. For a production buy, the first delivery is a sample, and the risk is that later batches drift in clearance, noise or grease fill while the part number stays the same.

A production buyer must be confident that every batch matches the approved sample in dimensions, noise, internal clearance and grease fill, and that the supplier keeps that consistency for years²⁸.

The questions below are the ones that separate a direct manufacturer answering from its own process records from a trader forwarding them to another factory²⁸.

• ✓Ask how noise is measured and what limits apply.
• ✓Ask for the dimensional sampling plan used on each batch.
• ✓Confirm whether internal clearance is delivered as CN or C3 on every batch.
• ✓Confirm the grease fill specification and how it is controlled.
• ✓Ask whether the supplier's quality management system is aligned with ISO 9001 and ISO/TS 16949.
• ✓Ask for the technical support contact who will answer substitution questions after the order.

Where do the sources disagree?

The three disagreements below all touch money and warranty, so you need both positions visible before you commit. One source treats ABEC as a tolerance band only, while another lists precision class alongside clearance and noise grades as a catalogue attribute buyers select^19,26.

One source says standard bearings can be modified to meet specific requirements, while another says to check standard catalogue availability before considering a custom design^2,22.

One source says OEM components guarantee exact specification matches and preserve warranties, while another says that if all substitution criteria match the bearing can be directly interchanged regardless of brand^14,23. Resolve each one with the manufacturer rather than assuming either side is correct.

What the sources do not establish

• No complete code-by-code table for cage, seal/shield and precision suffixes
• No dimensional tolerance values or inch/non-standard dimension guidance
• No mapping from application requirements to required internal clearance or precision class
• No actual load rating or limiting speed values for any specific bearing
• No cross-reference table or named equivalent part numbers across brands
• No actual prices, tooling costs or availability for a specific part
• No statement of which certifications are mandatory for a given application
• No warranty terms or sample-availability commitments for any named supplier
• No guidance on axial clearance matching specifically
• No guidance on mixing bearings from different production standards in a matched set