A fabric strip can be 200 mm long while the yarn removed from it measures 218 mm. Both lengths can be correct. The yarn curves over and under the crossing threads. When straightened, it is longer than the distance it spans in the cloth.

That extra length matters when analysing a fabric or estimating yarn use. Yet a result headed “9% crimp” tells only part of the story. First ask which yarn direction was tested. Also check the fabric state, the straightening force and the basis of the calculation. These details give the number its meaning.

This guide develops those distinctions through original examples. The numerical samples are invented teaching cases. They are not mill results or recommended test settings. Use them to check the reasoning and calculations. The actual fabric still needs the complete test method agreed for it.

Start with the two lengths and the denominator

Call the distance between the yarn's endpoints while it is in the fabric Lf. Call the straightened length between those same endpoints Ly. Keep both measurements in the same unit.

The fabric crimp percentage uses the in-fabric distance:

Crimp, C = 100 × (Ly − Lf) ÷ Lf.

For Lf = 200 mm and Ly = 218 mm, the extra length is 18 mm. Crimp is therefore 18 ÷ 200 × 100 = 9%.

ISO 7211-3:1984 describes this in-fabric denominator. Its public abstract also addresses the tension used to straighten the yarn. That tension depends on the yarn's nature and linear density. ISO confirmed the standard in 2023, so its old publication year alone does not make it obsolete.

Yarn take-up uses a different denominator. The convention used here takes the extra length as a percentage of the straightened yarn length:

Take-up, T = 100 × (Ly − Lf) ÷ Ly.

The same specimen gives 18 ÷ 218 × 100 = 8.26% take-up, rounded to two decimals. Nothing about the cloth has changed between the two calculations.

Direction in this example

In-fabric length, mm

Straightened yarn, mm

Crimp

Take-up

Warp

200

218

9.00%

8.26%

Weft

200

214

7.00%

6.54%

Here, warp crimp exceeds weft crimp. That is a feature of the invented data, not a rule for woven fabrics. Either direction can have the larger value. A higher crimp does not, by itself, show better or worse yarn quality.

The word “take-up” also names the loom motion that draws fabric forward. A take-up percentage in a fabric analysis is not a machine speed, gear ratio or winding setting. Write the quantity and its denominator beside the result.

Convert percentages before using them in a plan

Use C and T as percentage numbers in these equations. The two definitions then give:

T = 100C ÷ (100 + C).

C = 100T ÷ (100 − T).

Thus, 10% crimp corresponds to 9.09% take-up. Conversely, 10% take-up corresponds to 11.11% crimp. Entering one percentage in a formula intended for the other creates a length error.

Suppose an idealised calculation starts with 1,000 m of cloth and 10% take-up. The corresponding yarn length is 1,000 ÷ 0.90 = 1,111.11 m. Multiplying by 1.10 would give only 1,100 m, short by 11.11 m.

These equations relate defined lengths. They do not add loom waste, damaged material or finishing allowances. A purchasing calculation needs those other quantities separately, with their own bases.

Choose the method before choosing the straightening force

The difficult part of a crimp test is often deciding what counts as straightened yarn. Pull too little and some waviness may remain. Pull harder and the yarn itself may extend. A smooth-looking strand does not show where one process ended and the other began.

ASTM D3883-04(2020) covers yarn crimp and take-up in woven fabric. Its public guidance warns that extraction and straightening can affect the result. It also separates two ways of obtaining a result. One estimates crimp from yarn removed from cloth. The other compares yarn entering a loom with the corresponding cloth produced.

The same ASTM guidance calls for separate results when the yarn types differ substantially. Relevant differences include count, shrinkage response and weaving tension. A fabric can contain several yarn systems, and one average may hide those differences.

Use the complete agreed method for specimen preparation, tension, length measurement and reporting. A public standards summary cannot supply all those details. The numerical examples here do not claim to reproduce a full ISO or ASTM procedure.

Why a stronger pull can give a higher reported crimp

In their 2016 crimp-measurement paper, Mertova, Neckar and Ishtiaque examined a tensile-curve approach. The work first appeared online in 2015. The public abstract reports that the results depended on the applied force range. Choosing the force is therefore part of defining the measurement. It is more than a handling detail.

Consider a separate demonstration using matched yarn specimens from 200 mm fabric gauges. The laboratory declares three conditions for the demonstration. Suppose the corresponding straightened lengths are 217, 218 and 222 mm. The reported crimp values would be 8.5%, 9.0% and 11.0%.

The calculation alone cannot identify which force was suitable. The last value could include more yarn extension. The first could still contain some waviness. Neither explanation follows from the readings alone: examine the method and the way the yarn responds.

Keep the force units explicit. A force reading differs from a hanging mass. Force divided by yarn linear density is another quantity again. Copying a number from an old table without its units or material limits can produce a different test.

Elastic, textured or highly extensible yarns need particular care. Ask the laboratory whether the method is suitable before treating the result as purely geometric crimp. Document any departure from the agreed method. A convenient measurement should not silently become a claim of standard conformity.

Keep the fabric state consistent

A result taken from loom-state cloth need not match a result from relaxed or finished cloth. Name the stage being tested. Include any washing, heat treatment or other processing that could affect the result.

ISO 139:2005, confirmed in 2025, addresses atmospheres for textile conditioning and testing. Its published 2011 amendment belongs with the applicable edition. The public scope allows an alternative atmosphere when the parties agree.

For a comparison, record the actual conditioning and test conditions. A room reading alone does not show that the specimens received the required treatment. If the fabric states differ, say so before blaming weaving for a difference in results.

Preserve the specimen's identity while removing the yarn

Before removing yarn, mark warp and weft and record the sample position. Mark the cloth face too, if it matters to the test. Record whether the specimen came from the body, a border, a selvedge or a distinct pattern area. An edge yarn may serve a different function from a body yarn.

Establish the in-fabric gauge without introducing unintended stretch or distortion. Then preserve the relationship between that gauge and the yarn endpoints being measured. A lost mark or an extra cut changes the length pair that belongs in the formula.

During removal, avoid treating a damaged specimen as an ordinary valid observation. Note breaks, slipping marks and yarns that cannot be extracted intact. Use the method's rules to replace damaged specimens. Keeping only yarns that are easy to handle could bias the sample.

The record should make it possible to match each Ly value to its own Lf value. A neat column of straightened lengths is not enough. Each reading needs its original gauge and yarn group.

An apparent negative result deserves investigation. If a 200 mm fabric gauge gives a 198 mm removed-yarn length, the stated formula gives −1%. Do not reverse the subtraction or replace the answer with zero just to make it look plausible. Check the marks and units first. Then review the fabric tension state and straightening conditions to decide whether the pair is valid.

In a purely geometric model, the straight path between two points is the shortest. A curved path cannot be shorter. Actual testing, however, compares material before and after extraction under defined conditions. A tensioned cloth gauge and a retracted yarn may not share the intended reference state. The negative result calls for a review of the measurement basis. It does not mean negative waviness. Do not hide it within the average. Retain the original reading and the reason for any correction.

Include enough of the structure to answer the question

A short gauge may cover an area with many long floats. It could describe that area well but miss much of the figured pattern. Decide what the sample must represent before choosing an averaging rule.

Kolcavová Sirková and Mertová examined pattern transitions in their 2025 jacquard study. They also compared two ways to measure crimp. Its discussion shows why the length observed in relation to the pattern matters. It also identifies subjective straightening as a weakness of the thumb method.

That research does not establish one universal gauge for every jacquard fabric. Identify which part of the structure the measurement includes. Then check whether that part represents the pattern you intend to assess.

Suppose a stripe behaves differently from the rest of the cloth during finishing. Start the investigation with separate results for the stripe and the background. To estimate total yarn use later, combine the groups using their actual quantities. Equal weight would be wrong if one group contributes much more yarn.

Make the average match the sampling question

Yarn measurements sit within a sampling structure. Several yarns can come from one specimen, and several specimens can come from one roll. A production lot can contain several rolls. More yarns from one small patch do not create more independent rolls.

Consider three cloth panels from different recorded positions. Four warp yarns are measured from each under the same agreed conditions. Every in-fabric gauge is 200 mm.

Panel

Four straightened lengths, mm

Mean length, mm

Mean crimp

A

216, 218, 220, 218

218

9%

B

220, 222, 224, 222

222

11%

C

212, 214, 216, 214

214

7%

The overall mean of these twelve crimp values is 9%. It is also the mean of the three panel means because every panel contributes the same number of yarns.

Panels B and C differ by four percentage points, which the overall figure hides. Report the panel results when position matters. The variation between positions may be the reason for the investigation.

There are twelve yarn observations but only three sampled panels. They are not twelve independent production trials. Whether three panels are enough depends on the method and the decision. This teaching dataset does not set a sampling requirement.

Unequal gauges require an explicit averaging rule

For a separate calculation, imagine two valid length pairs. In the first, 100 mm of cloth gives 108 mm of yarn. In the second, 200 mm of cloth gives 222 mm of yarn.

Their individual crimps are 8% and 11%, so their simple mean is 9.5%. Pooling the lengths gives a different result:

Pooled crimp = 100 × (330 − 300) ÷ 300 = 10%.

The simple mean gives equal weight to each observation. The pooled calculation gives twice as much cloth-length weight to the second pair. Neither arithmetic operation changes the raw readings, but they answer different questions.

Use the averaging rule required by the method or agreed analysis. Do not mix gauges merely to obtain a preferred percentage. A pooled value may help with a separate length estimate. Label that use clearly so nobody mistakes it for the standard test result.

Check whether the apparent difference exceeds reading sensitivity

Small differences in crimp can arise from small differences in either length. Two decimal places in the answer do not prove that the measured lengths are that precise.

Return to the 200 mm and 218 mm example. For this sensitivity calculation alone, assume each measured length has a possible error of up to 0.5 mm.

Within those bounds, the lowest crimp pairs the shortest yarn with the longest fabric gauge:

100 × (217.5 − 200.5) ÷ 200.5 = 8.48%.

The highest uses the longest yarn length and shortest gauge:

100 × (218.5 − 199.5) ÷ 199.5 = 9.52%.

The nominal answer is 9%, but this limited model spans roughly 8.48–9.52%. This is a worst-case calculation using two assumed length bounds. It is neither a confidence interval nor a complete statement of uncertainty.

It leaves out force control, specimen handling, material variation and other influences. Its purpose is to show why a small reported difference needs measurement context.

Another laboratory might report 9.3% on apparently similar cloth. First compare the specimen state and yarn group, then the gauge, force and calculations. Do not infer a process change from the decimal difference alone.

A repeat measurement on the same specimen can check some handling effects. Fresh specimens can reveal variation within the sampled cloth. Those checks contribute different information, so keep their identities in the results.

Separate crimp from shrinkage after treatment

Crimp compares yarn length with the distance it spans in one stated fabric condition. Fabric shrinkage compares fabric dimensions before and after a treatment. The two quantities may change together without being equal.

ISO 5077:2007 addresses dimensional change after specified washing and drying procedures. That is a different measurement question from extracting and straightening yarn. ApparelDiary's fabric shrinkage and pattern allowance guide develops the downstream sizing calculation.

Consider this original example. A marked cloth gauge measures 500 mm before treatment and 480 mm afterwards. Its length reduction is:

100 × (500 − 480) ÷ 500 = 4%.

Now use matched specimens that represent those same fabric gauges. Suppose the straightened yarn lengths are 540 mm before treatment and 528 mm afterwards. The yarn extractions require suitable companion material because the original extraction is destructive.

Initial crimp is 100 × (540 − 500) ÷ 500 = 8%. Final crimp is 100 × (528 − 480) ÷ 480 = 10%.

Both the cloth and the corresponding straightened yarn become shorter, yet crimp rises. The reason is that the cloth gauge falls by a larger proportion than the yarn length.

The yarn-length reduction in this constructed case is 12 ÷ 540 × 100 = 2.22%. Crimp rises by 2 percentage points, while fabric length falls by 4%. These are three distinct statements.

Their relationship can be checked without adding percentages:

Final yarn length ÷ initial yarn length = (480 ÷ 500) × (1.10 ÷ 1.08) = 0.97778.

This agrees with 528 ÷ 540. The calculation is a consistency check for the stated matched measurements. It does not explain what caused the shrinkage. Nor does it prove that every yarn changed by the same amount.

Do not use the example to predict a washing result from an initial crimp value. The treatment, structure, fibre response and measurement conditions still need their own evidence.

Use crimp in a yarn estimate with a clear mass basis

Once the length basis is suitable, a geometric yarn estimate is straightforward. Keep the estimate tied to the material and process state it describes.

Assume a uniform warp has 5,000 ends, each contributing to 1,000 m of cloth. Its representative warp crimp is 8%. Every end is assumed to have a dry-yarn linear density of 20 tex. Its length basis must match the straightened-yarn measurement.

Each end then requires 1,000 × 1.08 = 1,080 m within the modelled cloth. Total retained warp length is 5,000 × 1,080 = 5,400,000 m.

Tex expresses grams per 1,000 m. Therefore:

Retained warp mass = 5,400,000 × 20 ÷ 1,000,000 = 108 kg.

The straight-path calculation without crimp would give 100 kg. The extra 8 kg represents the assumed curved yarn path in the retained cloth. It is not 8 kg of discarded waste.

This estimate excludes size, finish, moisture, broken yarn, loom-end losses and rejected cloth. Account for those differences before comparing the estimate with the measured weight of a finished roll.

ISO 7211-5:2020 covers linear density of yarn removed from fabric. It applies to nominally uniform yarns. Use a suitable measured count when analysing actual material. The measured count of a removed yarn may differ from the supplier's nominal count.

ApparelDiary's yarn-count conversion guide helps convert count systems before using the mass formula. Keep the chosen moisture and non-fibrous-material basis explicit.

Keep distinct warp groups separate

Consider a different warp with the same total of 5,000 ends. It has 4,000 ends at 20 tex and 8% crimp, plus 1,000 ends at 30 tex and 12% crimp. Retain the same 1,000 m cloth length and compatible dry-mass assumptions.

The first group contributes 4,000 × 1,000 × 1.08 × 20 ÷ 1,000,000 = 86.4 kg. The second contributes 1,000 × 1,000 × 1.12 × 30 ÷ 1,000,000 = 33.6 kg. Total retained warp mass is 120 kg.

Applying 20 tex and 8% to all ends would instead give 108 kg. That error comes from erasing the second group's identity, not from the mass formula.

Return to the original uniform warp and keep its other assumptions. Changing crimp from 8% to 10% would raise retained mass from 108 kg to 110 kg. The increase is 2 kg, or 1.85% of the original retained warp mass. It is not automatically a 2% rise in total factory yarn use.

Interpret extension through the fabric structure

A crimp value describes available yarn length under a defined measurement. It does not directly tell you how far the fabric will stretch at a given force. Recovery and the ability of a garment to hold its shape need their own evidence.

Shahabi and colleagues used image analysis in their 2014 study of worsted fabrics. They examined extension and crimp interchange. The public abstract describes de-crimping in the loaded direction as the fabric deforms. Crimp increases in the crossing direction.

This helps explain why pulling a woven fabric involves more than straightening isolated yarns. The crossing system participates in the response. The paper's fitted relationship belongs to its study; it is not a universal curve for all fabrics.

For product development, pair crimp measurements with the property that matters. If the concern is extension at a specified load, test that response. If the concern is permanent growth, include the required recovery assessment. If it is yarn consumption, use compatible length and mass measurements.

Likewise, a change in weave or finishing conditions may alter several features together. A higher crimp value alone does not prove a softer handle, stronger fabric or better drape. Establish those outcomes with appropriate comparisons.

Leave a result that another laboratory can interpret

A useful report joins the number to the specimen and method. Identify the fabric state, direction, yarn group and sampled positions. Include both the in-fabric and straightened lengths. Add the force basis, conditioning details and calculation used.

Keep individual values when they reveal position or pattern effects. Name the averaging rule and distinguish yarn observations from independent cloth specimens. Record damaged or excluded specimens with their reasons.

For a production estimate, also state the assumed length, end count and yarn linear density. Keep retained yarn, added materials and discarded losses separate. A test result becomes useful when its limits remain visible in the next calculation.

Start the report by making clear which two lengths were compared and under what conditions. The percentage can then support fabric analysis or a yarn estimate. It can also help focus a process investigation, while remaining distinct from shrinkage or stretch.

References