Elmendorf tear test (ISO 13937-1): five reasons your tear results scatter

The Elmendorf test takes a few seconds per specimen. You clamp the fabric, the knife makes a slit, you release the pendulum and the display gives you a tear force. There is very little for an operator to adjust, which is exactly why labs are surprised when two Elmendorf results for the same fabric disagree.

The pendulum is rarely the problem. The disagreement usually comes from what was torn, how it tore and which part of the scale the reading fell on.

ISO 13937 has four parts, and they are not interchangeable

ISO 13937-1 is the ballistic pendulum method: it measures the force needed to propagate a single-rip tear of defined length from a cut, when the force is applied suddenly. The other three parts use a tensile testing machine and a slow, constant rate of extension.

A specification that says "tear strength, ISO 13937" without a part number cannot be tested. A trouser result and an Elmendorf result on the same fabric are different numbers produced by a different loading regime, and there is no conversion factor between them. If your customer quotes a limit, ask which part it was written against before you run anything.

1. The fabric is outside the method's scope

ISO 13937-1 states that it is mainly applicable to woven fabrics and that, in general, it is not applicable to knitted fabrics and woven elastic fabrics. In a knit, the "tear" tends to become loops pulling out or the specimen extending rather than a clean rip running from the slit. The pendulum still swings and the display still shows a number, but the number does not describe tear propagation.

If you are asked for an Elmendorf value on a knit or a stretch woven, say so on the report or agree another method with the customer. A number produced outside the scope is worse than no number, because someone will put it in a specification.

2. The reading sits at the wrong end of the scale

Elmendorf testers are supplied with interchangeable pendulums or augmenting weights so that different capacity ranges can be covered. The standard requires the reading to fall within a defined working portion of the pendulum's range; check the relevant clause for the limits that apply to your edition.

The practical rule: a reading close to the bottom of the range is imprecise, because friction and small losses become a large share of the result. A reading close to the top risks the pendulum not completing the tear. Run one or two trial specimens, then choose the capacity that puts the results comfortably inside the working range, and record which pendulum or weight was fitted. When results from two labs differ by a constant offset, a different pendulum capacity is one of the first things to check.

3. The tear did not follow the intended line

The method assumes the tear runs straight from the end of the slit across the remaining length of the specimen. Some fabrics do not cooperate: the tear turns sideways, threads slip out of the weave instead of breaking, or the specimen tears out of the clamp.

These specimens are not valid tests, and the standard sets out how deviating tears are to be treated and reported. The common mistake is to average them in with the good specimens because "the machine gave a reading". Mark each specimen as you remove it, keep the rejected ones aside, and report how many were discarded and why. If most specimens of a fabric deviate, that is itself a finding: the fabric may need a different tear method.

4. Direction is labelled ambiguously

"Warp tear" means different things to different people. A specimen cut with its long side parallel to the warp is torn across the weft threads — or along the warp, depending on how you describe it. Two labs can both write "warp" and be reporting opposite directions.

Avoid the shorthand. On the report, state which set of threads was broken by the tear (for example, "tear across weft threads") and how the specimen was cut. It costs one line and prevents a dispute that otherwise looks like a quality problem in the fabric.

5. The instrument is assumed to be fine

A pendulum tester has few moving parts, but those parts matter:

• Level. The instrument must be level for the pendulum to start from the same energy every time.
• Free swing. Check that the pendulum, released with no specimen, returns the zero reading the manufacturer specifies. Friction in the bearing reduces every result.
• Clamps. The specimen must be held firmly across its full width so it does not slip during the tear.
• Knife. A blunt or misaligned knife changes the slit length and leaves a ragged starting point.

Write these checks into a short daily or weekly routine. They take a few minutes and remove the most common hidden cause of drift.

Elmendorf under ASTM D1424

ASTM D1424 uses the same falling-pendulum principle. Its specimen geometry gives a tearing distance of 43 mm on a specimen 63 mm wide. Do not assume the ISO and ASTM specimens are cut with the same die; use the die specified by the standard you are reporting against, and do not mix specimens from different dies in one result.

Like every mechanical textile test, tear force depends on moisture content. Condition specimens in the standard atmosphere of ISO 139 before testing.

Before you report

1. Is the fabric within the scope of ISO 13937-1, or is another part more suitable?
2. Did the readings fall within the working range of the fitted pendulum?
3. Were deviating or slipping tears rejected rather than averaged in?
4. Does the report state which threads were torn, not just "warp" or "weft"?
5. Were level, free swing, clamps and knife checked?
6. Were the specimens conditioned to ISO 139?

Our Elmendorf tear tester covers capacities from 800 cN to 6400 cN with interchangeable ranges and gives statistics directly on the display. We build it, along with our tensile testers for the other parts of ISO 13937, in our own workshop in Istanbul.