Fabric Testing Before Bulk: Shrinkage, Colourfastness and Dimensional Stability

Fabric approval is usually a conversation about appearance: the colour, the hand, the weight, whether the drape is right. All of that matters, and none of it predicts how the fabric will behave when it is cut, sewn, pressed and washed.

Behaviour is measurable before production. That is the whole argument for testing — not that it eliminates risk, but that it converts a category of risk from a discovery into a number, at a point when changing something is still cheap.

A Fabric Can Be Correct and Still Fail in Production

The sentence buyers find most frustrating from a factory is that the fabric met the specification. It is frustrating because it is often true, and because it does not help with the garment that no longer fits.

Specifications describe a fabric; production subjects it to processes. Three mechanisms do most of the damage:

• Heat and moisture. Steam, pressing and washing relax tensions built into the fabric during weaving, knitting, dyeing and finishing. When those tensions release, the fabric changes dimension — usually shrinking, sometimes growing.
• Mechanical action. Agitation in a wash, abrasion against a seam, friction against a lining. These produce surface change and, on some fabrics, transfer of colour.
• Time and tension under load. A bias-cut panel hung under its own weight continues to move; elastic recovers or does not; a stretched seam recovers or stays stretched.

None of these is a defect. They are properties. The failure is not that the fabric does them; it is that nobody measured them before cutting, so the pattern was never adjusted for them and the care instruction was written from habit.

Dimensional Change: Shrinkage, Relaxation and Growth

Dimensional stability is the most commercially important test in womenswear, because a garment that changes size after the first wash produces returns, and it is the test buyers most often skip.

Shrinkage, and where it comes from

Most shrinkage comes from tension locked into the fabric during manufacture, released by heat or moisture. It behaves differently by fibre and by construction.

• Viscose and some regenerated fibres are notably prone to losing dimension in water, and often more in the first wash than the second.
• Knits can shrink in length while growing in width, or the reverse, which is worse for fit than uniform change.
• Loosely woven fabrics tend to move more than tightly woven ones of the same fibre.
• Washed and garment-dyed finishes usually reduce later shrinkage because some of the movement has already happened — which is precisely why the wash process has to be repeated consistently in production.
• Press-only relaxation is a separate effect: fabric can change dimension with steam alone, before it has ever been washed.

Relaxation, and the bias problem

Relaxation is dimension change without washing. A fabric that has been rolled under tension continues to settle after it is unrolled, and a bias-cut panel settles more than a straight one because it is hanging on the diagonal of the weave.

This is why a bias garment is a scheduling problem as well as a fabric problem. Cutting first and measuring immediately produces a garment that will not measure the same the following week. The interval matters, and the only way to know how much is to observe the behaviour rather than assume it.

Growth, and why it is easy to miss

Not all dimensional change is loss. Some fabrics extend under steam or under their own weight, and a garment that grows is often harder to correct than one that shrinks, because there is no spare material in the seam allowance to recover. On a long skirt or a maxi dress, growth shows up as a hem that drops unevenly, which reads to a customer as a defect even when the construction is correct.

Colourfastness — to Washing, Rubbing, Light and Perspiration

Colourfastness is not one property. It is several, each with its own cause and its own failure mode in the market.

Two practical observations. Saturated shades are the hardest to make fast, so an intense black, a deep red or a vivid jewel tone deserves more testing rather than less — and on a dress where the bodice and skirt are separate dye lots, transfer between them is the specific risk. And the failure that costs most is the one that happens in the customer's wash, which means the wash test is not optional on anything that will be machine washed.

Seam Slippage, Pilling and Abrasion

Beyond dimension and colour, three more behaviours cause recoverable but irritating failures. They share a characteristic: each is cheap to test and expensive to explain.

• Seam slippage. On a smooth, tightly woven fabric with low friction, yarns can pull apart along a seam under tension. It shows up as opening at a shoulder, a waistband or a side seam, and it is more likely on a fabric that was chosen for its drape.
• Pilling. Fibre ends working loose and tangling into small balls, typically under friction — under an arm, at a seat, along a hip where a bag rubs. More likely on loose-spun yarns and blends, and often worse after the first few washes than on first wear.
• Abrasion. Surface wear showing as thinning or shine on areas of friction. Particularly visible on a smooth fabric, where the first sign is a change in sheen rather than a hole.

On occasionwear, an additional behaviour deserves attention: snagging. Beading, embroidery, lace and fine tulle catch on each other, on packaging and on the customer's own jewellery. It is not usually a lab test so much as a handling test — how the garment survives being folded, bagged and unfolded — and it is worth observing before deciding how a delicate piece is packed.

One further point on embellishment: the tests that matter are often the least formal. A beaded panel that sheds a few beads when the garment is folded is telling you something about packing, not about the beading, and no laboratory will report it. Buying a sample, wearing it, washing it if the instruction allows, and folding it the way a warehouse will — that is the test, and it takes an afternoon.

Stretch, Recovery and Why Both Matter

Stretch fabric is judged on two numbers, not one. How far it extends is the visible property; whether it returns is the one that determines whether the garment still fits after an evening.

• Extension decides how the garment has to be pattern-cut relative to a rigid fabric — a bodysuit or a bodycon dress relies on extension for fit rather than on ease.
• Recovery decides whether the garment returns to its intended dimensions after being stretched. Poor recovery shows up as bagging at the knees, sagging at the waist or a neckline that no longer sits where it did.
• Direction matters. Stretch is rarely the same along the fabric and across it, and a pattern cut without knowing which is which will fit differently on the body than on the table.
• Recovery degrades. A fabric that recovers well when new may recover less after repeated wear and washing, which is a fit problem that appears weeks after delivery rather than at the point of sale.

For a fitted stretch garment, both properties belong in the specification alongside the composition, because two fabrics described by the same fibre blend can differ enough in recovery to produce two different garments.

Which Tests Apply to Which Fabric

How to Specify a Test So the Result Is Comparable

A test result is only useful if it describes the same thing twice. Most disagreements about whether a fabric passed are really disagreements about what was tested.

• Name the method and the conditions. Temperature, cycle, duration and whether the sample was laundered before testing. A wash test at one temperature and a wash test at another are different tests.
• Say what is being measured and how. Dimensional change measured between marked points before and after, on a conditioned sample, is repeatable. "It shrank" is not.
• Test the garment, not only the fabric, where the risk is in the combination. Colour transfer between a lining and a shell, or shrinkage that only appears once a seam constrains the panel, will not show on a flat cloth sample.
• Test after the same finish production will use. A garment-dyed style needs the garment's own process reproduced, not the mill's standard one.
• Agree the acceptance criteria in advance and write them into the specification. A result has to be compared against something the buyer set, not against a general expectation.
• Keep the sample. The tested piece is the reference. Without it, a later dispute has no physical baseline.

What to Do With a Result That Is Marginally Out

Testing exists to produce decisions, and the interesting decisions are the marginal ones. A result clearly inside the limit needs nothing; a result clearly outside needs a different fabric. The ones in between have options.

1. Adjust the pattern for expected movement. If the fabric will shrink predictably, cutting to allow for it is legitimate and normal — provided the allowance is based on a measured number rather than an assumption.
2. Change the care instruction to match the fabric. A garment that cannot be machine washed without changing dimension can be labelled accordingly, as long as the instruction is honest and the market accepts it. This is usually cheaper than changing the fabric and worse for the customer relationship.
3. Change the finish. A different wash or a different dye process sometimes moves the number materially, and it is worth asking before abandoning the fabric.
4. Accept and monitor. Where the risk is confined to a specific construction — a bias panel, a long skirt — a targeted inline check during production may be a more proportionate response than rejecting the fabric.
5. Change the fabric. Sometimes the right answer, and it is far better reached at sampling than after a production run.

Whichever is chosen, the decision belongs in the specification rather than in a conversation. A care instruction written to accommodate a known shrinkage is a decision. The same care instruction written because nobody tested is an accident that happens to be defensible.

Tests, Cost and Time: What Is Worth Doing on a Small Order

Buyers on small orders reasonably ask whether testing is worth it. The honest answer is that some of it is nearly free and some of it depends on the order.

• Always worth it: a dimensional change test in the way the garment will actually be washed, on anything that will be machine washed. This is where the returns come from.
• Always worth it on saturated shades: colourfastness, especially where two dye lots meet in one garment.
• Worth it on structured or fitted pieces: recovery, on any fabric that carries the fit.
• Usually worth it on embellished pieces: a handling test, because snagging is a packaging and presentation issue that shows up as a customer complaint.
• Possible to defer on a small first order: exhaustive endurance testing on a style you intend to validate before scaling. Cheap validation, then test properly when the style becomes real.

The cost calculus is not the test fee; it is the comparison against the value of the goods at risk and the cost of a return. On a small order the test fee can look disproportionate to the invoice, and it remains small next to a season of returns, a rework, or a customer who no longer trusts the fit of the brand.

Where Testing Sits in the Development Sequence

Testing is not a single event; it is two moments in a programme, and buyers who collapse them into one usually test too late to act and too early to be accurate.

1. Before sampling, test for selection. When two candidate fabrics are close on appearance, behaviour is the tie-breaker. A dimensional test and a recovery test at this stage are cheap and they change the decision rather than confirming it.
2. Before cutting, test for production. Once the fabric, the construction and the finish are settled, test what production will actually do — including the wash the garment will be given, on a sample produced that way. This is the test that protects the order.
3. At first piece, observe. Some behaviours only appear once a garment is made: seam slippage under real seam tension, growth in a long bias panel, snagging on a beaded surface. Watching the first pieces is not a substitute for a lab test, and it catches the things a lab test cannot see.
4. During production, monitor the lot. If the fabric arrives in more than one lot, the tested result describes the tested lot. Dimensional behaviour on the lot being cut is worth a check, particularly where the risk is visible — a bias skirt, a long drop.

The reason the sequence matters is that each stage can still change something. Testing after cutting produces a report. Testing before cutting produces a decision — a pattern adjusted, a wash process changed, a care instruction written honestly, or a fabric replaced while replacement is still cheap.

What the Factory Needs From You to Test Properly

A test result is only as good as the brief behind it. This is the short list a supplier needs in order to test something that describes your product rather than a generic swatch.

• The finished construction, not just the fabric name. Whether the style is lined, structured, gathered, bias-cut or heavily embellished determines which behaviours are actually at risk.
• How the garment will be washed. The intended care instruction is the test condition. If you do not know it yet, that is itself the question the test should answer.
• The wash process production will use, where the style is garment-dyed or given a special finish. Testing the mill's standard process when production will run a different one answers the wrong question.
• Where on the garment the risk sits. A bodice and skirt in separate dye lots create a transfer risk; a bias panel creates a dimensional one; a beaded surface creates a handling one.
• The acceptance criteria, before the test. A result judged against criteria set afterwards is not a result, it is a negotiation.
• The markets and any buyer standard that applies. Requirements differ by destination, and a test designed for one market may not satisfy another.

Suppliers rarely receive all six, which is why so much testing produces a document that satisfies a file and changes nothing. The list takes ten minutes to assemble and it is the difference between a test and a formality, and it also shortens the testing cycle: a brief that answers these questions can be run straight away, while a vague one usually produces a week of clarifying emails. The list takes ten minutes to assemble and it is the difference between a test and a formality.

There is a sequencing point that sits alongside the cost question, and it concerns attention rather than money. Testing takes days, and those days sit between the fabric decision and the cut. A programme that does not schedule them explicitly will find that they happen anyway — as a delay while somebody waits for a report — or that they do not happen at all, which is worse. Building the test window into the calendar at the point the fabric is confirmed is what turns testing from an interruption into a step — see the womenswear production calendar for how that window sits against the rest of the schedule.

FAQ

When should fabric testing happen?

After the fabric is chosen and before cutting. Testing after cutting tells you what went wrong; testing before cutting lets you change something. Where a garment-dyed or washed finish is involved, the test comes after the finish, on a sample produced the way production will produce it.

Who pays for fabric testing?

It varies, and it is worth settling at quotation rather than discovering later. Where the buyer's market has a testing requirement, the cost usually sits with the buyer as part of the compliance scope. Where the test is the factory's own quality control, it sits with the factory. Both arrangements are workable; the unworkable one is nobody having decided.

Is a lab dip the same as a fabric test?

No. A lab dip is a colour approval — it establishes that the shade is acceptable. A fabric test establishes how the material behaves. A fabric can have a perfectly approved shade and fail a wash test, and the two activities answer different questions.

How much shrinkage is acceptable?

There is no universal figure, and any supplier who offers one as a rule of thumb is flattening a decision that belongs to you. The acceptable range depends on the market, the product and the buyer's own standard. What matters is that the criteria are written down before the test is run, so the result can be judged rather than discussed.

The fabric passed testing but the garment still shrank. Why?

Usually because the test was on flat cloth and the garment introduces new constraints: seams that restrict recovery, a lining that behaves differently, a panel cut on the bias. Where the risk is in the combination rather than the material, the test has to include the combination.

Can I write the care label from the test result?

That is exactly what the test is for. A care instruction written from a tested wash behaviour is defensible; one written from habit is a guess that happens to have been printed. Where the fabric cannot tolerate the treatment the design implies, that is a conversation to have at sampling rather than after the labels are ordered — see apparel labelling requirements for the US and EU.

Bulk fabric arrived from a different mill lot. Does it need retesting?

If the lot differs from what was tested, the result no longer describes the material in the building. Whether that requires a full retest depends on what changed — a different shade is a colourfastness question, a different weight is a dimensional one. At minimum, the dimensional behaviour should be checked on the lot that will be cut.

Working With Luxudress

Luxudress is the factory-direct front end for womenswear production across production facilities in Guangzhou and Dongguan, covering development, sampling, material sourcing, manufacturing, inspection and export packing.

We will tell you which behaviours to test for your fabric and construction, report the method alongside the result so it can be compared against your own criteria, and flag a number that sits close to a limit rather than quietly inside it. Where a fabric is likely to move, we would rather adjust the pattern or the care instruction at sampling than explain a returned dress later.

If you are choosing between fabrics and want the behaviour measured before you commit, send the composition, the construction and how the garment will be washed. You can also browse the style library to see constructions already developed on tested fabrics, or order from ready stock if the window is too short for a new test cycle.

Request a quote and we will return a costed plan including the testing scope it assumes.