Bearing Supplier Evaluation: Capacity, Quality Control and Delivery Checks

Content

1 Supplier evaluation begins with a definition you can audit
2 Manufacturing capacity: verify the three stages that set bearing life
3 Quality control: ask for documents, not promises
4 Delivery reliability: three numbers over twelve months
5 A weighted scorecard you can defend in a review
6 Where the supply model changes the questions

A buyer compares three quotations for the same 6200-2RS deep groove ball bearing and finds a price spread the drawing cannot explain: identical bore, identical seal, identical load rating. The differences sit where quotations never look, in heat treatment control, raceway grinding capability, the declared tolerance class and the finished stock behind the delivery date. Bearing supplier evaluation is the documented proof of those differences, and it produces three separate answers on manufacturing capacity, quality control and delivery reliability.

Supplier evaluation begins with a definition you can audit

A bearing supplier evaluation is a documented check of whether a supplier can hold a stated tolerance class and a stated delivery promise across a defined number of production lots. It is not a judgement about whether the samples in a presentation look acceptable.

Definition

Manufacturing capacity for bearings means the volume a supplier can produce while still holding dimensional, runout and noise limits inside the declared tolerance class. A line that holds normal grade at high volume is not interchangeable with a line that holds Class 5 at the same volume, even when the catalogue lists both.

Three questions separate them: which process stages run in-house, how the tolerance class is measured at the end of the line, and what happens to rings that fail that measurement. Price, packaging and lead time come afterwards, because a low quotation built on uncontrolled heat treatment is a deferred cost rather than a saving.

Manufacturing capacity: verify the three stages that set bearing life

Bearing life is decided at three process stages, so capacity claims should be checked against those stages rather than against machine counts.

Heat treatment control

Ask for the steel grade and the furnace record. Bearing steels such as 100Cr6 (GCr15, AISI 52100) specified under ISO 683-17 are through-hardened to a defined hardness range, but retained austenite, decarburisation depth and dimensional change after quenching decide whether rings grind cleanly. A supplier without batch-level furnace charts is buying heat treatment from someone else.

Raceway grinding and superfinishing

Roundness, raceway profile and surface finish set noise, friction and rating life under ISO 281. Ask for roundness values in micrometres from the last three lots, and ask which machines cover your bore range. A photograph of a grinding line proves nothing about the size band you actually purchase.

Assembly, clearance and final measurement

Clearance group, grease fill, seal drag and vibration testing are fixed at this stage. Ask for the clearance distribution across recent lots, not for the nominal group printed on the drawing.

Tapered roller bearings sharpen the matching question, because cup and cone are ground as a set and a mismatched pair destroys the preload calculation. The way the 30200 series is organised by size band shows why the machine list matters more than the model list.

30200 Series Tapered Roller Bearings by Size BandMatched cup-and-cone sets in bore sizes from 17 to 220 mm, with load ratings and speed data for machine-list checks.View Product →

Table 1: evidence that turns a capacity claim into a verified process.
Stage	Evidence to request	An answer that fails the check
Ring and roller material	Mill certificate with heat number, steel grade (100Cr6 or GCr15) and ISO 683-17 reference	Steel grade stated on the quotation only
Heat treatment	Batch furnace charts: hardness range, decarburisation depth, dimensional change	Hardness described as meeting requirements, with no records
Raceway grinding	Roundness and surface finish in micrometres from the last three lots	Catalogue noise figures with no lot data
Assembly and test	Clearance distribution, grease specification, vibration or noise results	Inspection report covering samples only

Quality control: ask for documents, not promises

A quality system is only as strong as the records it leaves behind, so request four document sets before the first order: material certificates, lot-level tolerance measurements, capability studies for critical dimensions, and failure analysis records.

Tolerance class has to be written into the drawing. ISO 492 defines the classes above the normal grade (Class 6, 5, 4 and 2), DIN 620 uses the matching P6, P5, P4 and P2 designations, and ABMA Standard 20 ABEC-1 through ABEC-7 sit on the same ladder. When the drawing stays silent, the supplier quotes normal grade and every price comparison afterwards is unreliable. A detailed walkthrough of how manufacturers check materials, tolerances and testing is available in this review of bearing manufacturing quality checks.

Capability data is the second test. For critical dimensions, a Cpk of 1.33 or higher is the usual buyer requirement, and it must come from a stated sample size and production period; a 20-piece study proves one machine on one shift, not a process. Deep groove ball bearings in the 6200 series are the standard test case, because the same bore size appears in normal, P6 and P5 grades and the price gap between them is the clearest signal of which class the supplier intends to ship.

6200 Series Deep Groove Ball Bearings in Multiple GradesSingle-row bearings from 6200 to 6244 with P0/P6/P5 precision, open, ZZ or 2RS options, and load-speed tables for grade comparison.View Product →

What the quotation states

Higher precision grade available on request
Quality management system certified
Seven-day lead time for stock items
Zero defect policy

What the audit verifies

Lot measurements at that grade from three production runs
Certificate scope covering the bearing line in the current cycle
Confirmed dates and OTD rate over the last 12 months
Failures grouped under ISO 15243 with corrective actions

Delivery reliability: three numbers over twelve months

Delivery reliability is measured on three numbers, each calculated over at least 12 months and at least 20 order lines.

On-time delivery rate: order lines shipped on or before the confirmed date, divided by all lines in the period.
Lead time spread: the gap between the quoted lead time and the 90th percentile of actual lead times, which is the figure production planning experiences.
Order fill rate: the share of ordered quantity shipped from stock without a clearance or grade substitution.

Ask for the raw shipping log rather than the percentage, and reconcile it against your own goods-receipt dates. A supplier reporting a high OTD rate on lines that were re-confirmed twice is not reliable; it is merely persistent.

Illustrative example, not a supplier record

Twenty order lines in one quarter. Eighteen shipped on the confirmed date. Two shipped late, by 4 days and 11 days, both from the same size band.

OTD = 18 / 20 = 90%

The concentration matters more than the average: both late lines came from one size band, which points to scheduling rather than transport.

Stocked items are easier to verify, because a physical count during a video walkthrough settles the availability claim in minutes. Mounted units such as the UCP pillow block range are the usual starting point for that check, since they are bought as replacement parts rather than as project quantities.

UCP Series Pillow Block Ball Bearings for Shaft SupportCast-iron or aluminum housed units from UCP203 to UCP320, with shaft diameters 17–100 mm and bolt sizes M10–M30 for replacement checks.View Product →

A weighted scorecard you can defend in a review

Score the supplier on a 100-point scale with fixed weights and a written threshold per line, so engineering, purchasing and management review the same evidence.

Table 2: a 100-point evaluation model for bearing suppliers, with the evidence behind each weight.
Criterion	Weight	Evidence	Threshold
Manufacturing capacity	25	In-house process list, machine list for the size band, heat treatment charts	All three stages in-house or under a documented subcontract
Quality control	30	Material certificates, lot measurements, Cpk study for critical dimensions	Cpk of 1.33 or higher with the tolerance class on the drawing
Delivery reliability	25	12-month shipping log, fill rate, lead time spread	OTD of 95% or higher across 20 or more lines
Documentation and traceability	10	Certificate scope, lot marking, ISO 15243 failure analysis	Heat number traceable to the finished lot on every sampled lot
Response and support	10	Engineering replies, replacement policy, sample lead time	Written replies within two working days

Apply two hard rules before reading the totals: a supplier scoring below 70% on quality control is disqualified regardless of its total, and any criterion without documentary evidence scores zero rather than an average. Both rules remove the temptation to reward a fast quotation.

Where the supply model changes the questions

A factory and a distributor answer different questions, so identify which role you are buying from before the scorecard is applied.

Suppliers that pair catalogued ranges with export and distribution, such as Shanghai Yinin Bearing & Transmission Company, are assessed on traceability rather than furnace charts: which factory produced the lot, whether the mill certificate travels with the shipment, and who owns the claim when a bearing fails inside its rating life. The same three headings apply, but the evidence shifts to lot marking, stock accuracy and claim handling.

Application changes the specification as well. Where one bearing reference must match a vehicle platform instead of a shaft diameter, the candidate range narrows before scoring begins, as the automotive bearing applications pages show.

How many samples should an evaluation order include?

Order 30 to 50 pieces drawn from at least three production lots, and ask the supplier to print the lot numbers on the packing list. One lot shows what a single machine did on a single day; three lots show whether the process holds. Keep the parts for a witness test in your own metrology lab and record roundness, clearance and noise results against the declared tolerance class.

Which tolerance grade should the RFQ specify?

Specify the grade from ISO 492 or the matching DIN 620 P-class and state it on the drawing rather than in the email. Normal grade covers general machinery, P6 covers electric motors and pumps where running accuracy affects noise, and P5 is where grinding and measurement time push cost up sharply. Mixing ABEC and ISO designations in one RFQ is the most common cause of a misleading price comparison.

What on-time delivery threshold should trigger a corrective action?

Below 95% OTD measured over 12 months and at least 20 order lines, request a written corrective action with the affected size bands named. Below 90%, treat the supplier as a planning risk and hold safety stock equal to the lead time spread. Track fill rate alongside OTD, because a supplier that ships on time but substitutes a different clearance group has not delivered the order.

Can a distributor pass an evaluation that a bearing factory would fail?

Yes, because the criteria differ. A distributor is audited on traceability, stock accuracy and claim handling, while a factory is audited on heat treatment records, grinding capability and lot-level measurement. Grade both roles under the same three headings, but for a distributor weight the manufacturer certificate and lot marking higher than the distributor inspection room.