Twist holds a spun yarn together, sets much of its strength and hand, and is one of the few spinning parameters a customer can check on the finished yarn. When a weaver complains that yarn "feels harder than last time", twist is often the first thing measured.
There are two established ways to measure it, and they are not interchangeable. Choosing the wrong one — or applying either to a yarn neither was written for — produces a precise-looking number with no meaning.
Two principles
Direct counting clamps a defined length of yarn between a fixed jaw and a rotating jaw and untwists it until the fibres or plies lie parallel. The number of turns removed, divided by the length, is the twist. It is covered by ISO 2061 and ASTM D1423.
Untwist/retwist (indirect) untwists a single spun yarn under a defined tension until the yarn has lengthened and then shortened again as twist is inserted in the opposite direction, returning to its original length. Half the total turns counted is taken as the twist. It is covered by ISO 17202 and ASTM D1422.
ISO 17202's own introduction sums up the trade-off: the direct method has long been regarded as the most accurate, while the faster untwist/retwist method has long been used in industry.
Which yarns each method suits
| ISO 2061 (direct) | ISO 17202 (untwist/retwist) | |
|---|---|---|
| Single spun yarns | Yes | Yes — its intended use |
| Single filament yarns | Yes | No |
| Plied and cabled yarns | Yes, with separate procedures | No |
| Open-end (rotor) yarns | Not suitable | Not suitable |
| Air-jet, false-twist, self-twist | — | Not suitable |
| Interlaced multifilament | Not suitable | — |
| Monofilament | Not suitable | — |
Both standards also exclude, unless the parties agree special conditions, yarns that stretch more than 0.5 % when the tension is raised from 0.5 to 1.0 cN/tex.
The open-end trap
The most common misuse is running open-end rotor yarns on a twist tester and reporting the result as a twist value. Rotor yarns have a core with real twist wrapped by "belt" fibres that do not untwist cleanly. Untwisting never reaches a point where all fibres lie parallel, and untwist/retwist does not return to a clean starting length.
Both ISO methods state that they are not suitable for open-end yarns. If a customer needs a twist figure for a rotor yarn, agree a specific in-house procedure and label it as such — do not put "ISO 2061" on the report.
Untwist/retwist depends on tension
The untwist/retwist method assumes that the yarn returns to its original length when the same amount of twist has been re-inserted in the opposite direction. That assumption only holds at the correct pretension. Too much tension and the yarn draws out during untwisting, too little and it snarls.
The pretension is set from the yarn's linear density, so you need an accurate count first. A wrong count means a wrong tension and a biased twist. This is why the two methods disagree more on fine, low-twist yarns, where the untwisted yarn has little strength to hold the tension.
Plied yarns: two twists, two directions
A two-ply yarn has ply twist (usually S) and single-yarn twist in each component (usually Z). They are measured separately:
- Measure ply twist by direct counting on the plied yarn.
- Separate a single component and measure its twist.
Report both, with directions. Note that the singles twist measured after removal from the ply is not the same as the twist the single had before plying — plying changes it. ISO 2061 gives separate procedures for each yarn type for this reason.
Direction, units and multipliers
- Direction. Always report S or Z. A twist value without direction is incomplete, especially for plied and crepe yarns.
- Units. Twist is usually expressed in turns per metre (tpm). Turns per inch (tpi) is still common; 1 tpi = 39.37 tpm. Mixing them up is surprisingly frequent when results move between systems.
- Twist multiplier. Comparing twist between counts needs a twist coefficient: in the metric system αm = tpm / √Nm, in the cotton system αe = tpi / √Ne. A yarn with higher twist but a higher count may have the same multiplier and the same character.
Specimen length and number of tests
Twist is not uniform along a yarn: thin places tend to take more twist than thick places. A very short specimen length captures this local variation, while a long one averages it out. Use the specimen length the method gives for the yarn type and draw specimens from several packages. Report the mean and coefficient of variation.
Condition the packages to ISO 139 before testing — moisture affects how fibres slide during untwisting.
Before you report
- Is the yarn within the scope of the method chosen? (Not open-end, not air-jet.)
- Direct counting or untwist/retwist — and does the report say which?
- Pretension set from a measured count?
- For plied yarns: ply and singles twist both reported, with directions?
- Units stated: tpm or tpi?
- Specimens from several packages, with CV?
Our twist tester supports both direct counting and automatic untwist/retwist, with specimen lengths from 10 mm to 500 mm and a touch-screen display. We manufacture it in our own workshop in Istanbul.