Fabric performance tests for contract upholstery

At a glance

Standard numbers as named in UK and European specifications; check the current edition of each. The standards report a result and set no pass mark: the figure you require is your own, agreed with the supplier and the partner upholsterer.

Six questions, six tests

Contract upholstery is judged on six things: how long it takes to wear through, whether it pills, whether it fades, whether the dye rubs off, whether the seams pull open and how it may be cleaned. Each has its own laboratory test on a small piece of the cloth. The standards behind them share a feature that surprises specifiers: they say how to test and how to report, and none of them sets a pass mark. The figure you require is your own decision, made with the fabric supplier and the partner upholsterer for the use, and written into the specification together with the standard, its edition and the conditions of the test.

Three things to keep in view.

• These are laboratory tests on a small sample. None predicts years of service. The one published guideline we read says that abrasion results above its own threshold are not an indicator of product lifespan.
• A result belongs to one cloth, in one colourway, from one batch. A sample card saying “contract” or “performance” is not a result: ask for the report.
• These tests are about the cloth only. Fire is a separate test of the whole composite of cover, interliner and filling: see BS 7176 hazard categories and the fire documentation matrix. The strength of the seat itself is tested to BS EN 16139.

Abrasion: what a Martindale result means

The Martindale test (BS EN ISO 12947) rubs a round specimen of the cloth against a standard abradant fabric, under a set load, along a path that traces a Lissajous figure, until the cloth breaks down. The result is the number of rubs reached. It measures how the cloth stands up to being rubbed flat: a fair guide to surface wear and a poor guide to anything else.

• Load. The 1998 text of part 2 sets two loads: a nominal 12 kPa for workwear, upholstery, bed linen and technical fabrics, and 9 kPa for apparel and household textiles other than upholstery and bed linen. A figure tested at 9 kPa is not an upholstery figure.
• End-point. Breakdown is reached when two separate threads are completely broken in a woven cloth; when a thread breaks and opens a hole in a knitted one; when the pile is worn off a pile fabric; and when the first hole reaches a set size in a nonwoven. The test is paused at inspection intervals, so the reported figure is the last inspection at which no breakdown had been seen.
• Rubs and cycles. ISO 12947-1:1998 defines a rub as one revolution of the two outer drives and a cycle as the full Lissajous figure of 16 rubs. Suppliers often quote a single number without saying which. Ask for the unit and the standard.
• Not Wyzenbeek. The US test (ASTM D4157) rubs back and forth along a line. The two cannot be converted: the Association for Contract Textiles states that there is no correlation between them.
• Not for coated fabrics. BSI's listing for BS EN ISO 12947-2:2016 says it is not applicable to coated fabrics, including laminated ones, and points to the ISO 5470 series. Vinyl and polyurethane covers are tested differently.
• Repeat tests differ. ACT reports that results of repeated abrasion tests on some fabric structures can vary by as much as 60 per cent or more. Ask for the laboratory report, not a rounded figure on a sample card.

Pilling, light, rubbing, seams and cleaning

Pilling (BS EN ISO 12945-2)

Pills are small balls of fibre that form on the surface in use. The test uses a modified Martindale machine and a visual assessment of the surface after a stated number of rubs (the assessment is set out in ISO 12945-4). The standard says the aim is to rank fabrics in the order of their propensity to pill in normal wear, not to predict an amount. Some soft yarns pill even when they pass a high abrasion test, so pilling is asked for separately. Ask for the grade and the number of rubs it was assessed at.

Light fastness (BS EN ISO 105-B02)

A specimen is exposed to a xenon-arc lamp that imitates daylight, alongside a set of blue wool references, and the cloth is graded by comparing its colour change with theirs. The references run from 1 (very low) to 8 (very high) and each is about twice as light-fast as the one before, so a grade 5 is not “a bit better” than a 4. A second set, L2 to L9, is used in America and the two sets are not interchangeable. Grades are given per colourway. Ask which scale: a grade on the 1 to 8 blue wool scale is a different statement from the “grade 4 at 40 hours” in the US guideline below, which uses AATCC methods, and ACT notes that there is no direct correlation between testing hours and hours of service.

Colour fastness to rubbing (BS EN ISO 105-X12)

A dry and a wet rubbing cloth are rubbed across the specimen and the staining on the cloth is graded against a grey scale (ISO 105-A03). Pile fabrics use a rectangular rubbing finger (19 mm by 25.4 mm at 9 N). It is a test of dye coming off onto clothes, so dark, saturated colours and pile cloths deserve a look. Ask for the dry and wet grades per colourway.

Seam slippage (BS EN ISO 13936)

Seam slippage is the movement of yarns in a woven cloth when it is pulled at a seam. The standard stresses that it is a property of the fabric and should not be confused with seam strength. Part 1 uses a fixed seam opening and is not suitable for stretch fabrics; part 2 uses a fixed load and is the part the standard says suits upholstery woven fabrics, including stretch ones. Ask which part was used, the result in warp and weft, and the stitch the seam was sewn with.

Cleaning codes

The letters W, S, WS and X on upholstery cloth say how it may be cleaned: W water-based cleaner, S solvent, WS either, and X vacuum or brush only. They are a trade convention and not a test result, and no standard we have read defines them. The supplier's written cleaning instructions for the colourway are the document to ask for, and the workshop should confirm they suit the finished piece, including any interliner or backing.

One published set of minimums, and why you still choose your own figure

The Association for Contract Textiles (ACT) is a US trade body. Its voluntary performance guidelines for contract fabrics, February 2025 edition for woven fabrics for indoor use, are the one published set of figures we found from a neutral body. They are a useful example of how a guideline is built, and not a UK requirement: they use US test methods (ASTM and AATCC) and not the ISO methods above, and ACT describes them as minimum requirements, for information only, subject to change without notice.

• ACT's own cautions: abrasion results above its high-traffic threshold are not an indicator of product lifespan; repeat tests can differ by 60 per cent or more on some structures; and there is no correlation between the Wyzenbeek and Martindale results.
• Coated fabrics have their own set. For vinyl and polyurethane covers ACT sets a Wyzenbeek abrasion figure and tests such as flex resistance and seam strength, which shows that a woven and a coated cloth are not judged on the same tests.
• Its flammability line is a US (California) test and is not a UK fire route.

How to set your own figure

Agree it with the fabric supplier and the partner upholsterer for the use. Seats used all day by the public, in sun and cleaned often, deserve a higher requirement than a private bedroom chair. Write it as the number, the standard and its edition, the load or conditions and the end-point, so that a laboratory report can be checked against it line by line. Where the cloth is supplied by the client (COM), the evidence has to arrive with it: see fabric and leather yardage, COM and COL.

What to ask the supplier for, and where fire and strength fit

• A test report, not a card. From a named laboratory, dated, naming the standard and its edition.
• The cloth identified. Name, colourway, batch, weight and fibre content, so the report can be tied to what is delivered.
• The conditions. For abrasion: the load, the end-point, the number of rubs and the unit. For pilling: the number of rubs. For light and rubbing: the scale and the grade, per colourway.
• The cleaning instructions for the colourway, in writing.
• The fire evidence separately. A cloth is not fire-certified by its performance tests. The BS 7176 route tests the cover with its interliner and filling as a composite: see the hazard categories and the documentation matrix.
• The seat tested separately. The strength of the finished chair is tested to BS EN 16139, which does not test the cloth.

Leather is tested differently (see our leather article), as are vinyl and polyurethane covers.

Cloth [name, colourway] to achieve, on the batch supplied: abrasion not less than [ ] rubs without breakdown, tested to BS EN ISO 12947-2 [edition] at 12 kPa; pilling not worse than grade [ ] after [ ] rubs to BS EN ISO 12945-2 [edition]; light fastness not less than [ ] on the blue wool scale to BS EN ISO 105-B02 [edition]; colour fastness to rubbing not less than [ ] dry and [ ] wet to ISO 105-X12 [edition]; seam slippage not more than [ ] mm at [ ] N, warp and weft, to BS EN ISO 13936-2 [edition].

The supplier is to provide, before the cloth is cut, a test report from [a named laboratory] for each test above, naming the standard and edition, the colourway and the batch, and the cloth's written cleaning instructions. Fire performance is specified separately [BS 7176 hazard category]; the tests above are not fire tests.

For vinyl or polyurethane coated fabric, abrasion and the other tests are to be to [the methods stated by the supplier, for example the ISO 5470 series], not Martindale to BS EN ISO 12947. For leather, tests as the leather specification [ref].

Standards and sources

• BS EN ISO 12947-1:1998 and -2:2016: Martindale abrasion. BSI: current; the 2016 listing excludes coated fabrics. 1998 previews read; 2016 text not read.
• BS EN ISO 12945-2:2020: pilling, modified Martindale. BSI: current. ISO preview read.
• BS EN ISO 105-B02 and ISO 105-X12:2016: light and rubbing fastness. Previews read; confirm the current BS EN ISO editions.
• BS EN ISO 13936-1 and -2:2004: seam slippage. Previews read.
• ACT Voluntary Performance Guidelines, Woven Fabrics, Indoor, Feb 2025: US voluntary guideline; read in full. Not a UK requirement.

Questions we are asked

What Martindale number do I need for a restaurant?

The standard sets none: ISO 12947 reports a result. One published guideline, from the US Association for Contract Textiles, uses 40,000 cycles of the ASTM Martindale method as its high-traffic minimum for woven upholstery, and itself says higher results are not an indicator of lifespan. Agree a figure with the fabric supplier and the upholsterer for the use, and write the standard, the load and the end-point beside it. This sheet gives no number of its own.

Can a Wyzenbeek result be converted to Martindale?

No. ACT states there is no correlation between the two, and repeat tests on the same fabric can vary by 60 per cent or more on some structures. Ask for the test your specification names.

Are rubs and cycles the same thing?

Not in ISO 12947-1:1998, which defines a rub as one revolution of the outer drives and a cycle as the whole Lissajous figure of 16 rubs. Suppliers quote one number without always saying which. Ask for the unit, the standard and the load, which is 12 kPa for upholstery in the 1998 text.

What does a light fastness of 5 mean?

On the blue wool scale of ISO 105-B02 it means the cloth faded like reference 5, and each step up is about twice as light-fast as the one below. The same number on another scale, or in hours of exposure, means something different, so ask for the standard and the colourway.

Is seam slippage the same as seam strength?

No. ISO 13936 says seam slippage is a fabric property and should not be confused with seam strength. It measures how far yarns slide at a seam, so a smooth, loosely woven cloth can fail it while the stitching is strong.

Do these tests replace a fire certificate?

No. They describe the cloth. Fire is tested on the whole composite of cover, interliner and filling, to the BS 7176 route for contract seating. See the BS 7176 hazard categories and fire documentation sheets.

Do the tests apply to velvet, vinyl or leather?

The Martindale apparatus covers pile up to 2 mm high, and the end-point for a pile fabric is the pile worn off. Coated fabrics such as vinyl and polyurethane are outside ISO 12947-2 on BSI's listing. Leather has its own tests: ask the supplier which.