Knitted Bag Pilling Resistance Testing: Choosing Between the Pilling Box, the Random Tumble Method and Martindale Abrasion

The Decision Order: Four Questions Before Any Test

Pilling testing is usually commissioned backwards. A buyer asks for a pilling test, a laboratory runs the method it has, and a grade arrives that nobody specified an acceptance level for. Four decisions taken in order before the request goes out prevent that, and each one narrows the choice to something defensible.

The first question is whether the bag carries a durability or appearance claim. If it does, the test method and the acceptance grade are evidence a buyer may need to produce later, so they should be chosen for defensibility rather than speed. If it does not, a screening result is enough and the simpler method wins.

The second is whether the yarn is staple spun or continuous filament, because the two pill by different mechanisms. Staple yarns pill by fibre ends migrating out of the yarn and tangling. Filament yarns mostly snag rather than pill, and a pilling box result on a filament structure tells a buyer very little about how the bag will look after a month.

The third question is surface: a brushed or raised surface behaves differently from a flat one and generally needs a gentler method. The fourth is jurisdictional, whether the buyer's market and retailer follow ISO methods or ASTM methods, because a grade obtained under one is not directly comparable with a grade obtained under the other.

Method One: The Pilling Box

The pilling box under ISO 12945-1 is the traditional method: specimens are mounted on tubes and tumbled in a cork-lined box rotating at a constant speed, so the fabric rubs against itself and against the cork. It is a gentle, multi-directional action and it reproduces the kind of abrasion a bag gets in general use rather than against a single hard surface.

The end point is expressed in revolutions, and programmes commonly run to 7,200 or 14,400 revolutions depending on how severe a test is wanted. The higher figure roughly doubles the exposure and is the one to specify where a bag is sold with a durability claim. Whichever is chosen, the number must appear on the report, because a grade with no end point is not a result.

The box has one limitation worth knowing: it is slow, and it under-represents abrasion against hard edges such as a desk corner or a car door frame. A bag that grades 4 in the box can still show surface fuzzing where a customer habitually sets it down on one corner, which is a different mechanism and needs the Martindale method to see.

Method Two: Random Tumble

Random tumble under ASTM D3512 is the North American counterpart and it is faster and more aggressive. Specimens are tumbled in a cylindrical chamber with a small amount of cotton lint and a controlled air stream, and the standard exposure is 30 minutes. The lint is deliberate: it accelerates pill formation and makes the test discriminate more quickly.

Because it is more aggressive, a random tumble result is not interchangeable with a pilling box result on the same fabric. A structure that grades 4 in the box may grade 3 to 4 on random tumble, and neither number is wrong. Buyers comparing quotes across two programmes should check which method produced each grade before concluding that one yarn outperforms another.

The method suits staple-spun yarns and knitted structures with a slightly raised surface, which is where pilling actually occurs in service. On a flat filament structure it produces little and the result is uninformative, which is why the second decision in the sequence above, staple or filament, comes before the method choice rather than after it.

Method Three: Martindale Abrasion and Pilling

The Martindale method under ISO 12945-2 rubs a specimen against a standard abrading fabric in a figure-of-eight path under a defined pressure, and it is the method of choice where abrasion and pilling need to be assessed together. The end point is expressed in rubs, and a common bag programme figure is 7,000 rubs with a grade assessed afterwards.

Its advantage over the box is that it also predicts wear and, at higher rub counts, breakdown. Running the same specimens to 7,000 rubs for a pilling grade and then continuing to a defined end point for abrasion gives a buyer two numbers from one set of specimens, which is efficient and, more importantly, comparable across lots.

The pressure setting matters and should be stated. Standard Martindale work uses a pressure appropriate to the end use, and for a bag body a figure in the region used for upholstery and heavy textiles is more realistic than the apparel setting. A laboratory running an apparel pressure on a bag body will return a flattering number that the field will not reproduce.

Grading: What a Grade of 4 Actually Looks Like

Pilling is graded by comparing the tested specimen against standard reference photographs under controlled lighting, on a scale from 5 down to 1. Grade 5 is no change, grade 4 is slight surface fuzzing or a few partially formed pills, grade 3 is moderate pilling with clearly formed pills of varying size, and grades 2 and 1 describe dense pilling covering much of the surface.

The realistic acceptance level for a bag body is grade 4 or better after the agreed end point, and grade 3 to 4 is often accepted on a textured or recycled yarn where some surface character is inherent to the material. Grade 3 on a smooth filament body is a failure, because there is no yarn reason for it and it indicates a process problem rather than a material character.

Grading is subjective by nature and that is worth managing rather than denying. Three assessors grading independently and taking the median is the standard control, and using standard lighting rather than a bench lamp is the second. Where a result sits on a half-grade boundary, commission a second assessment before treating it as a pass or a fail.

Specimens per test

4 minimum, two in each direction, cut away from the selvedge and from any seam

Conditioning

4 hours minimum at 20 degrees Celsius and 65 percent relative humidity per ISO 139

Pilling box end point

7200 revolutions for screening, 14400 where a durability claim is made

Random tumble end point

30 minutes with cotton lint, per ASTM D3512

Martindale pilling end point

7000 rubs, grade assessed immediately afterwards

Acceptance grade, smooth body

4 or better; grade 3 indicates a process problem rather than a material character

Acceptance grade, recycled or textured yarn

3 to 4 acceptable, with the character agreed on the approved reference sample

Assessment control

Three independent assessors, median grade taken, under standard lighting

Re-test trigger

Any yarn lot change, yarn count change, or finish change

The approved reference sample is the quiet fix for the subjectivity problem. If the buyer has signed off a physical reference and the reference is photographed alongside the tested specimen, the grade becomes a comparison against something both sides have held rather than against a memory of what the swatch looked like at approval.

Which Fibres Pill and Why the Yarn Comes First

Pilling is mostly decided at yarn selection, not at testing. A test tells a buyer what the yarn does; it does not change what the yarn does. The variables that matter are fibre type, staple length, yarn twist and whether the yarn has been treated to resist pilling, and all four are set before the first panel is knitted.

The practical consequence for a programme is that pilling testing is best run at yarn selection rather than at bulk. Testing three candidate yarns at sampling and choosing on the result costs one laboratory booking, whereas testing the chosen yarn at bulk produces a number and leaves the buyer with the same yarn and a shorter window in which to change it.

Setting an End Point and an Acceptance Grade

Two numbers make a pilling requirement: the end point and the grade. Grade 4 alone is meaningless, because grade 4 after 1,000 Martindale rubs and grade 4 after 7,000 are very different products. Write both, in the order the test runs them, and the requirement becomes checkable by anyone holding a report.

Setting the end point to match the product is the part buyers under-specify. A lightweight promotional tote that will be used for a season does not need 14,400 revolutions in a pilling box. A premium everyday bag sold with a two-year expectation does, and specifying the lower figure for it is how an appearance complaint arrives in month four.

One more field belongs in the requirement: what happens at a borderline result. Agreeing in advance that a half-grade below the acceptance level triggers a second assessment by three assessors, rather than an immediate rejection, resolves the most common dispute in pilling testing before it starts and costs nothing to write down.

Where Pilling Testing Fits in the Programme

Pilling testing sits at yarn selection and again at bulk confirmation, and nowhere else. Two bookings per programme is the normal arrangement: one during sampling at 7 working days, when a yarn change is still a decision rather than a disruption, and one on the first bulk lot to confirm the production yarn performs as the approved yarn did.

It does not sit at goods-in. Pilling is a laboratory result on specimens, not an attribute counted on a sample, and no amount of looking at a bag on an inspection table will reveal it. A buyer who wants pilling controlled specifies the yarn, tests it, and then checks at bulk that the yarn is the same yarn, which is a documentation check rather than a test.

Frequently Asked Questions

People Also Ask