TEST STANDARD

Martindale abrasion: the end point decides your result, not the machine

August 26, 2026 · 3 min read

A buyer specifies "40,000 Martindale". A supplier tests and reports 52,000. The buyer's own lab reports 31,000. Same fabric, same instrument type, same standard number on both reports.

Nobody made a mistake with the machine. They stopped the test at different moments.

ISO 12947 is four standards, not one

The number alone does not define a test. The part number does.

PartWhat it definesWhat you get
ISO 12947-1Apparatus and general methodNo result — this is the equipment part
ISO 12947-2Determination of specimen breakdownA cycle count
ISO 12947-3Determination of mass lossMilligrams lost
ISO 12947-4Assessment of appearance changeA visual grade

A specification that says only "ISO 12947" is incomplete. Ask which part before you quote a number, because parts 2, 3 and 4 cannot be converted into each other.

The breakdown definition is not obvious

Part 2 is the one most buyers mean, and it ends when the specimen "breaks down". That word means different things depending on the fabric:

If your operator uses the woven rule on a knit, the test runs longer and the reported figure is higher. That single difference explains most disputes in the 30,000–60,000 range.

Write the breakdown criterion on the test report, not just the cycle count. It costs one line and prevents the whole argument.

Pressure: 9 kPa or 12 kPa

Two nominal pressures are used:

Running an apparel fabric at 12 kPa shortens its life considerably. The pressure is set by loading pieces onto the specimen holder, and it is easy for an operator to reach for the wrong weight when the bench holds both sets. Keep them physically separated and colour-coded.

The abradant wears out before you notice

The standard abradant is a specified worsted wool fabric, backed by felt. It is a consumable, and a tired abradant abrades more slowly — which inflates your cycle count in exactly the direction that looks good.

Replace the wool and the felt on the schedule the method requires, not when they look worn. "Looks worn" arrives long after the results have drifted.

Keep a log per station: date, batch number of the wool, cycles since change. When a result looks generous, the log is the first thing to check.

Abrasion is not pilling

The same instrument runs both tests, and the tests are routinely confused.

AbrasionPilling
StandardISO 12947-2ISO 12945-2
AbradantStandard wool fabricThe specimen itself
ResultCycles to breakdownVisual grade 1–5
Question answeredHow long until it wears throughHow ugly does it get

They use different specimen holders and different loads. A pilling result reported as an abrasion figure is not a small error — it is a different property.

Before you report

  1. Which part of ISO 12947 does the specification require?
  2. Was the breakdown criterion the right one for this fabric construction?
  3. Was the pressure 9 or 12 kPa, and does that match the end use?
  4. How many cycles has the abradant run since it was changed?
  5. Were the specimens conditioned to ISO 139 before testing?

Two laboratories that agree on these five points usually land within ten per cent of each other. Two that do not can differ by a factor of two — which is the gap between passing and failing a contract.

Cenk YILMAZ
TLI-TexLabInstruments

TexLab instruments used in this test

Need the right instrument to run this test in your lab? Let us help you choose.

Request a quote