Two quotes, two numbers: what the 30,000 / 50,000 line actually tells you
You are holding two fabric quotes. One mill writes "30,000 Martindale". Another writes "50,000 cycles". A third writes "commercial grade" and no number at all. None of them states which test produced the figure, how the test was stopped, or whether the figure describes the fabric or the finished chair.
That is the whole of the problem: a bare Martindale figure cannot tell you whether it is enough, because it is incomplete. Three inputs are missing before the number means anything — the test method and standard that produced it, the endpoint criterion that ended the test, and the article the report actually covers. Until those are on paper, 50,000 is not "better" than 30,000. It may not even be the same unit.
What follows is an acceptance checklist: eight checks, each with the question to send the mill or lab and the form of answer that closes it. Two limits apply throughout. First, this page contains no YIZHONG abrasion test results, no universal thresholds and no lifespan estimates — those are documents tied to a specific fabric reference and a specific project, not blog claims. Second, YIZHONG does not claim lab accreditation, BIFMA, GREENGUARD or FSC certification in this article. The method below is how a buyer audits any supplier, including us.
Check 1 — Which test produced the number?
Martindale and Wyzenbeek are different machines with different motions, different abradants and different units. American Leather's explainer describes the Wyzenbeek machine as rubbing an abrading material — usually cotton duck fabric — back and forth across the test fabric, and presents it as the North American standard method, while the Martindale test is described as the UK/European standard (American Leather, "Upholstery Fabric Rub Tests Explained"). One method reports cycles. The other reports double rubs.
The Martindale method sits inside the ISO 12947 / BS EN ISO 12947 framework, according to a technical guide published by Kothea (Kothea, "Martindale Rub Test: How Many Rubs Does Upholstery Need?", published April 2026 per the publisher's page metadata). If a report cites an ASTM designation rather than the ISO one, ask the lab to confirm the test conditions match the method you specified.
Because the units are not the same, cross-method conversion should not be used as a specification value. Gessato's guide states that a 40,000 Martindale rating corresponds to around 30,000 Wyzenbeek double rubs (Gessato, "The Martindale Test, A Complete Guide"). Record that as a publisher's approximation for orientation only — never write a converted figure into a purchase order as the required performance.
Action: write the method and the standard designation with its edition year into the purchase order, next to the number. "Martindale, ISO 12947-2, 40,000 cycles minimum" is a clause. "40,000 Martindale" is a slogan.
Check 2 — Is the number a fabric result or a furniture claim?
A rub count can migrate from a lab report onto a swatch card, then onto a quotation, then into a tender, losing its fabric reference somewhere along the way. Ask for the report by fabric reference and colourway, because a different dye lot, backing or post-treatment can be a different tested article.
Three questions close this check:
- Is the tested article the cover fabric exactly as supplied to us, or a laminate with a backcoating or FR treatment applied afterwards?
- Does the report name the model code and the fabric on our quotation?
- If the number appears only on a website, swatch card or sales email, what is the underlying report number and date?
A valid report for a different article proves nothing about this one — the same discipline that applies to structural standards applies here.
Check 3 — Where does the figure sit against published duty bands?
Published guidance exists, and it disagrees. That disagreement is the useful part.
| Publisher | What it states about the contract/restaurant band |
|---|---|
| Revival, "Commercial Upholstery Fabric: Ratings and Specs" | Duty-band table presents under ~20,000 Martindale as light duty, ~20,000+ as general contract, ~40,000+ as heavy duty for restaurants, lobbies and waiting rooms, and ~100,000+ as extra-heavy for busy dining, bars and 24/7 healthcare |
| Kothea (April 2026) | 40,000 rubs and above described as contract grade for hotels, restaurants, bars and offices |
| Haining Leading, "Martindale Rub Test Results: 50,000+ Cycles Upholstery Fabric Guide" | 30,000 or more described as commercial grade suitable for restaurant seating, with anything beyond 50,000 offering diminishing returns for most applications |
Three publishers, three different contract lines: 40,000+, 40,000+, and 30,000+. None of the supplied excerpts states a statistical basis, a sample size or a test house behind these bands — they are mill and retailer guidance, not standards-body thresholds. Read them as orientation, not as an accept/reject rule.
In practice: 30,000 sits at or below the contract line in two of the three sources and at the heavy-duty line in one. 50,000 clears all three published contract bands. That makes 50,000 easier to defend if a fabric fails early — but it does not make 30,000 automatically wrong for a low-turnover private dining room, and it does not make 50,000 automatically right for a fabric that pills or stains.
Action: choose your own threshold deliberately, write the reason on the approval sheet, and stop copying it from a mill page.
Check 4 — What stopped the test? Abrasion-through is only one endpoint
Documented Martindale practice, as described by Kothea (April 2026): a circular 140 mm specimen is mounted face-down over a foam backing, a 38 mm worsted wool disc on the upper plate acts as the abradant (wire mesh in some configurations), the test pauses every 5,000 cycles for inspection under a standardised light source, and the test ends at either two complete yarn breaks or a noticeable change in surface appearance such as pilling, flattening or loss of pile. The cycle count at that point is the reported rub count.
This is the single most under-asked question in a fabric approval, because the same cycle count can mean two different things. One fabric reached 50,000 and failed structurally — a hole at the seat edge. Another reached 50,000 and only looked worn — fuzzed and flattened under continuous use. For a high-turnover dining room those are different commercial risks, and they need different responses.
Ask the lab: which endpoint was recorded for this result, was the appearance-change criterion applied, and do inspection photographs exist? If the report gives only a number and no endpoint definition, the number is not yet usable.
Note the published tension honestly: Haining Leading states that the Martindale test measures abrasion resistance only — not UV fading, staining, pilling, seam slippage or damage from hard objects and claws. Kothea's description shows appearance change, including pilling, can end a Martindale test. Both can be true: a test can be terminated by a pilling observation without the test being a pilling test, and it cannot tell you whether pilling will occur on a different fabric or after cleaning.
Check 5 — Pilling needs its own test, not a higher rub count
Pilling is surface fibre entanglement and balling. It is driven by fibre type and blend, yarn twist, pile construction and cleaning regime — not by abrasion cycles alone. Buying another 20,000 Martindale cycles does not answer a pilling complaint.
So ask a separate question: which pilling test method did the mill use, and what was the result on this exact fabric? "Not tested" is a legitimate answer — it is also a risk you should accept in writing rather than discover in month four.
For a dining room, add a physical validation step: request a swatch, upholster two or three seats in the busiest section, and run a defined trial period before releasing the full order. Write both lines into the PO — the abrasion figure with its method, and the pilling position, whether a tested result or an accepted "not tested".
Check 6 — What the rub count does not cover
Abrasion resistance is one property. It does not predict fading, staining and stain release under cleaning, seam slippage, or damage from hard objects (Haining Leading, same source). For F&B specifically, put these on the same request:
- Colourfastness to rubbing and to light.
- Cleanability against the disinfectants, coffee, wine and grease your operation actually uses — by product name, not "standard cleaning".
- Seam construction, backing and any coating.
A high Martindale figure on a fabric that will not release a red-wine stain is a worse restaurant choice than a mid figure with proven cleanability. The rub count should not be the only durability line on the approval sheet.
Check 7 — Map the figure to the seat, not the room average
Upholstery wear is not uniform. Seat fronts and cushion edges take the highest abrasion; backrests take the highest soiling; armrests take combined load; booth benches wear differently again. Ask whether the quoted fabric is specified for all upholstered surfaces or only low-contact panels, and whether a heavier grade is used on seat cushions.
Also ask how the cover is fixed and whether it is field-replaceable. On a fixed banquette, a replaceable cover changes the risk profile of a mid-grade fabric more than another 10,000 cycles does.
Check 8 — The document pack that closes the question
One test report, one datasheet, one written pilling position, one structural statement — that is the pack. Specifically:
- Abrasion test report: lab name and location, report number and issue date, standard and edition, method, abradant used, specimen size and backing, inspection interval, endpoint criterion, cycles recorded, and the fabric reference tested.
- Fabric datasheet and swatch card matching that reference, with composition, backing and any finish stated.
- Pilling position in writing, plus colourfastness and cleanability data against your own cleaning chemicals.
- A structural statement for the frame, naming the standard and edition and the report that covers the shipped model — a separate question from fabric abrasion.
- A record of who supplied each document and when. An answer without an owner and a date is not an acceptance record.
Acceptance checklist you can paste into an RFQ
| # | Field | Acceptable answer |
|---|---|---|
| 1 | Test method and standard | Martindale (ISO 12947 / BS EN ISO 12947) or Wyzenbeek (ASTM D4157), with edition year |
| 2 | Result and unit | Cycles (Martindale) or double rubs (Wyzenbeek), stated with the exact fabric reference and colourway tested |
| 3 | Endpoint definition | Yarn break, appearance change, or both — and who judged it |
| 4 | Threshold | Your stated figure, with a written reason for choosing it |
| 5 | Pilling | Tested method and result, or a written accepted "not tested" |
| 6 | Non-abrasion performance | Light and rub colourfastness, stain cleanability against your chemicals, seam construction |
| 7 | Document | Report supplied, dated, and naming the article on the purchase order |
| 8 | Validation | Wear trial on a named number of seats before full release |
If you work through these eight fields, the 30,000-versus-50,000 question usually resolves itself — either one number turns out to be measured on a different method, or one report has no endpoint definition, or the higher figure arrives with no pilling data and the lower one does.
To move from comparison to approval, request the actual abrasion test report, the fabric swatch card and a seating configuration quote for your floor plan — and ask us to state, in writing, which of the fields above we can supply for the specific fabric and model you are considering.