A supplier reports 850 N. The customer's laboratory reports 690 N. Both used a calibrated tensile tester, both tested the warp direction, both wrote "ISO 13934" on the report.
The difference is not calibration. It is specimen geometry.
Two methods, one property
| Strip method | Grab method | |
|---|---|---|
| ISO | 13934-1 | 13934-2 |
| ASTM | D5035 | D5034 |
| Specimen width | Narrow strip, fully gripped | Wider specimen, partially gripped |
| Jaws | Grip the entire specimen width | Grip only a central portion |
| Typical result | Lower | Higher |
In the strip method, the jaws hold the full width of the specimen. Every yarn in the specimen carries load, and the measured force is the true breaking force of that width of fabric.
In the grab method, the specimen is wider than the jaws. The yarns outside the jaw faces are not gripped, but they are still woven into the yarns that are. They share load through yarn-to-yarn friction and crimp interchange.
That sharing is real, and it is why grab results come out higher.
Which is "correct"?
Both are. They answer different questions.
The strip method gives a fabric property you can compare across constructions and across laboratories. It is the cleaner number for material development and for specification limits.
The grab method is closer to how fabric fails in a seam or under a fastening, where load enters through a limited area and the surrounding fabric helps. It is the more realistic number for many end-use assessments.
The important rule is simple: never compare a grab result with a strip result. There is no conversion factor. The ratio depends on construction, yarn friction and crimp — it is not a constant you can apply.
Where the numbers actually drift
Beyond the method choice, four settings quietly move the result.
1. Extension rate. The standard sets the rate according to how much the fabric stretches. A high-elongation fabric tested at the low-elongation rate spends longer under load and generally breaks lower. Check which rate class your fabric falls into before you start, not after.
2. Gauge length. A shorter gauge length usually gives a slightly higher force, because a shorter specimen contains fewer weak points. If you cannot reach the specified gauge length on your instrument, that is a deviation and belongs on the report.
3. Jaw slippage. If the specimen creeps out of the jaws, the force curve shows a shoulder rather than a clean break, and the result is meaningless. Pneumatic jaws with the correct face material and pressure solve most of this. Serrated faces damage delicate fabrics; rubber-faced jaws slip on coated ones. Match the face to the material.
4. Jaw breaks. A specimen that breaks at the jaw line, rather than in the free length, has failed because of the clamping, not because of its own strength. Most methods require you to discard jaw breaks and test a replacement specimen. Reporting them inflates scatter and depresses the mean.
A useful habit: record where each specimen broke. If more than one in five breaks at the jaw, the clamping needs attention before the fabric does.
Conditioning applies here too
Cotton, viscose and other hygroscopic fibres change strength with moisture content. Testing a specimen straight off the roll, before it has reached equilibrium in the standard atmosphere, produces a number that depends on the weather in your lab.
Condition to ISO 139 — 20 ± 2 °C and 65 ± 4 % relative humidity — and record the conditions on the report.
The one-line fix
Most tensile disputes end the moment both laboratories write the full method on the report:
ISO 13934-1 (strip), 200 mm gauge length, 100 mm/min, conditioned per ISO 139
Not "ISO 13934". The part number, the geometry, the rate, the gauge length. Four extra words, and the argument does not start.