Six Fabric Tests Before a Roll Is Cut - and Why Each One Exists

GSM, D65 colour, crockmeter, coating adhesion, hydrostatic head, Martindale. What each test catches, what it costs when skipped, and the pass marks we use on every roll for bag manufacturing.

6 min readquality controltestinginspection

D65 colour matching light cabinet in the fabric inspection room

Every test on our inspection frame exists because we once skipped it and paid for the container. This is the list, in the order a roll meets it, with the failure each one prevents.

1. GSM weight check: the truth serum

Tool: circular GSM cutter and balance. Pass: within +/- 5% of spec. Catches: mills substituting lighter yarn or fewer picks. "600D" from two mills can differ by 40 gsm; the lighter one wears through at the base corners first.

2. D65 light-cabinet colour matching

Tool: light cabinet at D65 (daylight), also checked under TL84 (store) and A (incandescent). Pass: within shade band A/B of the sealed lab dip. Catches: shade drift between dye lots, and metamerism - two fabrics that match under shop lights and clash outdoors. Every roll, not a sample per lot, because dye lots split.

3. Crockmeter rubbing fastness (ISO 105-X12)

Tool: crockmeter, dry and wet. Pass: grade 4+ dry, 3-4+ wet (ISO 105-X12). Catches: dye that transfers to clothing. Dark linings and PU leather trims are the usual offenders; indigo denim is tested with fixing.

4. Coating adhesion and defects

Tool: inspection frame at 100 m per roll, tape peel on coating. Pass: coating does not lift; defects logged per 100 m under the 4-point system, max 20 points. Catches: PU or PVC that peels after a season, plus weaving faults, oil marks and holes that would otherwise be cut into panels.

5. Hydrostatic head (ISO 811) for waterproof claims

Tool: hydrostatic head tester. Pass: as specified - 1,000 mm standard PU, 2,000 mm outdoor, > 5,000 mm TPU. Catches: coatings applied too thin. Run on pre-production samples and on every TPU lot.

6. Martindale abrasion, tensile and tear strength

Tool: Martindale abrasion tester; tensile/tear tester. Pass: Martindale 15,000+ for 600D polyester, 50,000+ for 1000D nylon; tensile and tear to spec. Catches: base fabrics that will not survive the bag's life. Run on pre-production samples and any new mill.

What we add by market

  • GCC / hot climates: light-fastness (ISO 105-B02) grade 5+, perspiration fastness, 50 C conditioning.
  • EU retail: REACH screening, phthalate test on PVC, cold-flex at -10 C.
  • US: CPSIA lead and phthalate, Prop 65.
  • Recycled: GRS transaction certificate reconciled to roll numbers.

What each failure costs when the test is skipped

The tests are cheap; skipping them is not. Real cases from the last two years, anonymised:

Skipped testWhat shippedCost of the miss
GSM check6,000 m of "600D" at 242 gsm instead of 300Bags failed drop test; full re-cut, 18 days late
D65 colourTwo dye lots split under daylightRetailer rejected 40% of a container at the DC
Rubbing fastnessNavy body marked white piping on 3,000 gym bagsFull rework with solvent wipe; brand pulled the reorder
Coating adhesionPU lifted at stress seams after 6 weeksWarranty claims ran 11% on the line

Each test costs minutes on an inspection frame. Each miss above cost more than a year of testing every roll we ship.

After the fabric

Fabric tests stop fabric problems. Seam, zipper, handle and load failures are caught on the finished bag - the six-gate, fourteen-test bag QC process at our sister factory covers those, from cutting-room checks to a 100% final inspection before packing.

Every roll we ship carries its test record. Ask for it with your first order; you should never have to take a fabric supplier's word for a number that a machine can print.

Building your own incoming-QC bench for under $600

Importers who buy fabric direct can replicate 80% of our frame checks with four purchases:

ToolCostCatches
GSM cutter + 0.01 g scale~$180Underweight cloth, coating drift
Portable D65 lamp box~$250Lot-to-lot shade splits
Manual crockmeter~$120Rubbing fastness failures
30x thread counter loupe~$25T-count shortfalls

The routine: cut one disc per colour per lot, weigh it, count the threads, rub-test, and view against the retained swatch under D65. Fifteen minutes per delivery, and the four most expensive fabric failures are caught while the truck is still at the dock. Hydrostatic head and Martindale stay with the lab - but those two are exactly what a supplier's per-roll test report should already be giving you in writing.

Reading a test report: passes that hide failures

A report full of green ticks can still be hiding problems. Three reading habits catch them:

  • Check the method year. "ISO 105-X12" alone is not enough - the rubbing cloth and pressure changed between editions. A pass to a superseded edition is a soft pass; ask for the current one.
  • Check the specimen source. "Sample submitted by client" means the mill chose the best metre of the roll. Frame tests on randomly drawn metres (we pull at 20%, 50%, 80% of roll length) describe the roll you will actually cut.
  • Check units against the claim. Hydrostatic head quoted in mbar looks bigger than mm H2O (1 mbar = 10.2 mm); a "2,000" rating in the wrong unit is a fifth of what the brief asked for. The same trick appears with N vs kgf on tear numbers.

None of this requires a lab background - only the habit of reading the header and footer of the report, not just the verdict column.

Sampling plans: how many rolls to test at each order size

Testing every roll is our default on programme fabric; buyers running their own QC need a plan that scales:

Order sizeMinimum plan
Under 1,000 m (1-3 rolls)Test every roll - the lot is too small to average
1,000-5,000 mEvery roll for GSM + shade; 1 in 3 for full six tests
5,000-20,000 mAQL-style: sqrt(n) rolls full-tested, all rolls shade-checked
20,000 m+Per dye lot: first, middle, last roll full-tested per lot

Two principles behind the table. Shade and GSM drift within a production run, so those two never skip a roll. Destructive tests (tear, abrasion) stabilise per dye lot, so sampling them is safe once the lot has three clean data points. And any single failure resets the plan to 100% inspection for that lot - a rule worth writing into the purchase contract verbatim.

The two tests we added after real failures

Six tests are the frame minimum; two more entered the routine because a customer paid for their absence. Seam slippage (ISO 13936) joined after a lightweight ripstop lining opened along the seam under load despite passing tear. Pilling (ISO 12945) joined after a brushed-face polyester fuzzed white on a black bag within a month of retail. Neither is in the standard six because most bag fabrics pass them trivially - both are now automatic on any construction under 250 gsm or with a brushed or peached face.

Where the six tests sit in the order timeline

Testing is not one event. Pre-production: full six on the first roll of each colour, before the cutting room touches anything - this is the gate that stops a bad lot. In-line: GSM and shade on every roll as it arrives; a 30-second check per roll. Post-production: the finished-bag QC inherits the fabric data, so a seam failure can be traced to a roll number rather than argued about. The order of operations matters because a fabric problem found at finished-bag QC has already consumed the cutting, sewing and packing labour of every bag it touched.

Buyer FAQ

What tests should bag fabric pass before cutting?

Six on every roll: GSM within +/-5%, D65 colour match to the sealed dip, rubbing fastness 4+ dry / 3+ wet, coating adhesion with no lift, hydrostatic head per spec, and Martindale abrasion of 20,000+ cycles for body fabrics.

What is a D65 light cabinet?

A viewing booth with standardised daylight (6500 K). Colour approved under warehouse fluorescents can split visibly in daylight (metamerism); D65 is the light source retail buyers audit under.

What tests should fabric pass before bag production?

At minimum: GSM weight, D65 colour match, rubbing fastness (ISO 105-X12), coating adhesion, hydrostatic head and Martindale abrasion. Add seam slippage and pilling for lightweight or brushed fabrics.

How much does fabric testing cost?

On an inspection frame, minutes per roll and no external cost. Third-party lab confirmation of the six tests runs $150-300 per fabric per season. Either is a rounding error against one rejected container.

What is hydrostatic head?

The water column height, in mm, a coated fabric resists before leaking (ISO 811). 1,000 mm resists light rain; 3,000 mm is a real waterproof claim; TPU laminates exceed 10,000 mm.

Free colour card and swatches

Get the colour card and three fabric proposals for your bag

Tell us the bag type and target price. We reply with two or three fabric options, swatches by courier and, if you need the bags made, a sample plan from our sister factory.