Knitted Bag Abrasion Testing: Which Cycle Count a Knit Panel Actually Needs, and Why Martindale Alone Is Not Enough

First Question: What Is the Bag Actually Rubbed By?

Before choosing a method, name the abrasion the bag will meet in service, because the test should reproduce the dominant one. A shopper carried in the hand rubs against a coat seam and a hip. A tote set down on pavement abrades its base against concrete. A backpack rubs shoulder strap against shoulder and back panel against a shirt.

Those are three different mechanisms and they do not rank the same way. Flat surface abrasion, the mechanism Martindale reproduces, dominates for a base panel and for a body panel rubbing a coat. Flexing and edge abrasion dominate at a folded handle edge and at the corner of a base, and a flat abrasion test says little about either.

So the first output of the question is a list of surfaces, ranked. Base panel, handle edge, body panel, strap contact face. Test the top two by the method that fits them, and accept that the third and fourth are covered by inspection and field returns rather than by a laboratory figure.

Second Question: Is Martindale the Right Machine?

Martindale is the right default and the wrong universal answer. The machine rubs a specimen against a standard abradant in a Lissajous figure, which produces multidirectional flat abrasion and is therefore a good model for a bag panel rubbing fabric or a flat surface. It is specified in ISO 12947 and, in the ASTM system, in ASTM D4966.

Where it fits less well is at edges and folds, and at coatings. ISO 12947-2 explicitly excludes coated fabrics including laminated ones, and directs coated-surface abrasion to the ISO 5470 series instead. A bag with a polyurethane-coated base therefore needs a second method rather than a longer Martindale run.

The other machine a buyer will meet is the rotary platform abrader under ASTM D3884, which drives a specimen against an abradant under two rotating wheels with a vacuum removing debris. It is common in North American specifications and produces results that do not correlate with Martindale cycles, so the two should never be converted.

Third Question: What Counts as Failure?

An abrasion result is meaningless without an end point, and ISO 12947 offers three different ones across its parts. Part 2 determines specimen breakdown, meaning the point at which a specified number of threads or loops have broken. Part 3 determines mass loss. Part 4 assesses appearance change, including fuzzing, pilling and colour change.

For a knitted bag the interesting end points are usually breakdown and appearance change, and which one matters is a commercial decision rather than a technical one. Breakdown answers how long until the panel holes through. Appearance change answers how long until the panel looks tired, which for a branded bag happens long before the panel holes.

This is the single most common specification error in bag programmes. A buyer asks for 20,000 cycles, the laboratory runs to breakdown, and the report says the specimen survived. The bag then pills and greys at 6,000 cycles of real use and the buyer has a report that is technically true and commercially useless.

Fourth Question: How Many Cycles Should a Programme Specify?

Cycle counts for bag panels cluster into a few bands, and the band follows the use rather than the fashion. A lightweight promotional tote used a handful of times sits in the low thousands. A daily-use shopper sits around 15,000 to 25,000 cycles at 9 kPa. A bag whose base meets pavement regularly, or a work bag with a coated base, sits higher and is usually tested at 12 kPa.

The pressure choice matters as much as the count. ISO 12947-2 uses a nominal pressure of 9 kPa for apparel-type end uses and 12 kPa for heavier duty, and the same fabric produces a materially lower cycle count at the higher pressure. A report that states cycles but not pressure is half a specification.

Method

ISO 12947-2 for breakdown, ISO 12947-3 for mass loss, ISO 12947-4 for appearance change

Companion ASTM method

ASTM D4966 for Martindale abrasion on textile fabrics

Nominal pressure, apparel weight

9 kPa

Nominal pressure, heavier duty

12 kPa

Abradant

Standard wool abradant, changed at the intervals the method sets

Lissajous figure

Multidirectional rub, which is why the method suits flat panels

Typical body panel target

15,000 to 25,000 cycles at 9 kPa to breakdown

Typical base panel target

Tested at 12 kPa; coated bases move to the ISO 5470 series

Pilling follow-up

ISO 12945-2 modified Martindale pilling method where appearance governs

Specimens

Tested per direction where the panel has one, and on the finished construction

Two cautions on the numbers. First, a cycle count is not a service life prediction; Martindale cycles correlate loosely with wear and not at all with a specific number of months. Second, the count should be set against a reference the buyer already owns and is happy with, which is a far better calibration than any published table.

Fifth Question: Does the End Point Match How the Bag Fails?

Ask what the returns desk actually sees. If it sees holes worn through a base, breakdown is the right end point. If it sees a greyed, pilled, tired-looking body on a bag that is structurally intact, appearance change is the right end point and the breakdown figure is answering a question nobody asked.

For knitted panels the second case is more common than buyers expect. A knit surface fuzzes under abrasion before it breaks, because the loop is free to move and the fibre ends migrate to the surface. Pilling assessment under ISO 12945-2, which uses a modified Martindale approach, is usually the more informative test on a knit body.

The practical recommendation is to run both on the first programme and then drop one. Run breakdown to establish the structural floor, run appearance change or pilling to establish the commercial floor, and whichever limit binds first becomes the figure written into future specifications. That costs one extra test once.

Sixth Question: Which Panel Gets Tested?

Test the surface that fails, on the construction that will be shipped. A report on a flat swatch knitted at a different gauge than the production panel is not evidence about the bag, and this is where most abrasion data quietly stops being useful. State the gauge, the yarn count and the finished grammage on the submission form.

Where a bag uses two constructions, test both. A body knitted at E12 in a 300D yarn and a base knitted at E7 in a 600D yarn are different fabrics with different abrasion lives, and a single report on the body tells a buyer nothing about the base that meets the pavement.

Direction matters too. A knitted panel has a course direction and a wale direction and the two do not abrade alike, so where the method allows directional testing, test both and record both. Reporting only the better number is common and is worth guarding against by asking for both on the report.

Reading a Martindale Report Without Being Misled

A usable report carries seven items: the method and edition, the nominal pressure, the abradant and its change interval, the end point used, the number of specimens, the result per specimen rather than a single mean, and the tested construction. Missing any one of them and the number cannot be compared with another laboratory's number.

The per-specimen spread is the item most often suppressed and the most diagnostic. Five specimens reaching 18,000, 19,000, 20,000, 21,000 and 22,000 cycles describe a consistent fabric. Five reaching 8,000, 12,000, 20,000, 28,000 and 34,000 describe the same mean and a very different fabric, and only the full list reveals it.

Conditioning is the last thing to check. Textiles are conditioned in the standard atmosphere before testing under ISO 139, because moisture content changes abrasion behaviour measurably. A report that does not mention conditioning may be perfectly competent, but a buyer comparing two reports should confirm both were conditioned the same way.

What Abrasion Testing Does Not Predict

Abrasion testing predicts flat surface wear and nothing else. It does not predict seam failure, handle elongation, zip or closure life, colour loss, or what happens when a bag is overloaded and the loop structure distorts. Each of those has its own method, and a strong abrasion figure on a bag with a weak handle joinery is a bag that fails anyway.

It also does not predict ultraviolet degradation, which matters for a bag left in a car or on a beach, or hydrolysis for a coated panel stored in a humid warehouse. Both are ageing mechanisms with their own tests, and both produce failures that look like abrasion failures to someone who has not seen the other two.

The useful framing is that abrasion is one row in a small matrix. For a knitted bag programme that matrix is usually abrasion to ISO 12947, tensile to ISO 13934-1 or ASTM D5034, seam strength to ISO 13935-2, colour fastness to the relevant ISO 105 parts, and dimensional change to ISO 5077. Five rows, and none of them substitutes for another.

Writing the Abrasion Requirement into a Specification

The specification line needs the method and part, the pressure, the abradant, the end point, the minimum cycle count, the number of specimens, and the construction being tested. Seven fields, and every one of them prevents an argument that would otherwise happen after the goods have shipped.

Add one more field that buyers omit: the current reference. Naming the approved sample as the comparison standard, and requiring the laboratory to test against the same construction, closes the last gap through which a good number on the wrong fabric can pass.

Frequently Asked Questions

People Also Ask