Seam Abrasion Resistance: Protect Stitching Where Garments Rub

An exposed stitch line can wear through while the cloth beside it still looks sound. A thread with a higher breaking force may leave that contact problem unchanged. The useful question is how the sewn joint handles rubbing in the place where the garment will be used.

Seam abrasion resistance concerns wear of the assembled seam. It depends on the thread and the seam's shape. That shape controls how much thread meets the rubbing surface. A cone test, a flat fabric test and a test on a raised seam each load a different object.

This guide shows how to compare wear in ordinary garment seams. Its worked values and trial limits are fictional. They show how to read the results. They do not certify protective clothing or predict a fixed number of days of wear.

Locate the contact before choosing a stronger thread

Begin with the garment feature. A cuff edge, a pocket opening and an inside-leg seam may touch different surfaces. One can rub against equipment, another against the wearer's other leg, and another against a carried object. Name what rubs against the seam and where it touches. “High wear” gives too little detail.

Look at the garment in use as well as laid flat. A fold can expose a row that appeared protected on the table. Movement can turn an internal chain of loops toward a rubbing surface. A loaded pocket can also change the shape and position of its edge.

Record which face receives the wear. The needle side and the looper side do not have the same thread path. A good result from the outward face does not prove that the inside face will perform as well. Where both faces matter, make them separate parts of the review.

Use a simple contact map to keep the trial relevant:

These are prompts for observation, not fixed claims about every garment. Confirm the actual contact for the style under review. A sample that removes the raised edge may remove the feature that caused the complaint.

Separate abrasion from a sharp cut, heat damage or an overload. Those events may all produce a broken stitch. A new thread alone cannot resolve every cause. Keep the returned area and its care history. Do this before cutting out a test piece or making a repair.

Read the damage at the thread and the fabric

Inspect an unused matching seam beside the worn one. Note thread fuzz, flattened areas, broken filaments, missing stitch sections and damage to the surrounding cloth. Keep the original images and identify the same locations in later observations.

The surface can suggest what to check next, but it does not show the full history. In their study of sewing-thread abrasion, Munshi, Pai and Ukidve used electron microscopy to examine three sewing threads. Their published abstract describes different forms of surface damage. That limited study supports looking closely at damage. It does not give a visual scale that suits all modern garment seams.

Record when the first thread breaks. Record separately when the seam can no longer do its job. A decorative row may break without opening the joint. A load-bearing row may lose security while another visible row remains intact. Define which row the product requirement concerns.

Fabric failure also needs its own record. If the cloth wears through around sound stitches, the garment has not passed merely because the thread survived. Likewise, a coating can lose its required surface condition before a stitch breaks.

Do not count every visible change as the same event. Fuzzing, colour change, first thread break and an open joint are different observations. A report can record all of them, but it must say which one ends the test and which one determines approval.

Inspect the starting seam too. A partly cut thread, missed stitch or damaged fabric yarn is not a clean baseline for judging wear. If a piece was made badly, follow the agreed rules for replacing it. Keep a record of the fault and the reason for replacement. Do not quietly exchange only the specimens that give inconvenient results.

Change the contact path without creating another defect

The thread's position matters as much as its label. AMANN's thread-abrasion guidance explains that sewing settings and fabric affect how exposed the thread is. It also notes a risk with thicker thread. The thread may sit farther above the surface, where it faces more rubbing. Sew and compare the proposed thickness on the actual joint.

Possible proposals include moving a row away from a rubbing edge or reducing a raised profile. A fold or another construction detail may shield part of the thread path. Check the resulting joint as a whole. Protection on one face can add bulk or create a new rubbing ridge. It can also expose the opposite face.

Keep strength, movement and comfort in the decision. Coats' Soft and Secure Seams guide treats strength, stretch, appearance, bulk and softness as linked design demands. A seam that survives rubbing but becomes stiff against the skin may still be unsuitable.

Do not turn up tension until the stitches appear buried. Excess tension can change balance, pucker the cloth or limit seam extension. Inspect the formed stitch on both faces and retain the approved balance. The useful setting depends on the actual thread, material, stitch and machine.

There is no fixed needle-to-looper thread ratio that solves every wear problem. Nor does one fibre name establish a ranking for every seam. Check the thread's construction, size and finish for the intended use. Read supplier comparisons with their test conditions. Check which materials were used before applying the result elsewhere.

Keep the variables clear. If the proposal changes thread, stitch type and seam position together, it tests a complete new assembly. It does not prove which single change caused the result. The comparison can still help the team choose a joint. The record must state every change between the two versions.

First correct rough guides, damaged needles and poor stitch formation. A suitable thread can be weakened on its way into the seam. A sample sewn slowly by hand may escape that damage. Testing it would not show whether the normal sewing route harms the thread.

Ask what the reported abrasion number actually measures

The method name, specimen and end point belong beside the result. A large number without those details is difficult to use.

ISO 12947-2:2016 addresses fabric breakdown under the Martindale method, with inspection at set intervals. It remains current following confirmation in 2023. Its scope leaves out coated fabrics, including laminated fabrics. A result for flat cloth does not, by itself, describe a raised joint sewn from that cloth. The shape and contact have changed.

Seam-specific procedures also exist. AMANN's seam-abrasion guidance describes its own procedure based on the Martindale principle. This shows why the laboratory must name any change made to test a seam. The same machine can be used in different ways. A changed setup may no longer follow the standard method.

SATRA TM95:2020 addresses abrasion and snagging of a material or thread in a stitched seam. Its public scope covers finished leather and coated fabric. Similar materials may also fit, and it covers all thread types. Ask the laboratory whether the actual garment joint fits that scope. The method's name alone does not establish suitability for every woven or knitted seam.

ASTM D4966-22(2026) is a fabric Martindale method. Its public guidance identifies several sources of variation, including the rubbing material, specimen tension and contact pressure. It also warns against predicting wear life from the test alone. Evidence must link the test result to the intended use.

Agree the complete procedure with a competent laboratory. Public scopes help choose the question; they do not supply the full instructions. Record any change from the named method. Do not call a development screen proof of certified compliance.

Keep the counted unit and end point unchanged

A machine cycle, a rubbing stroke and a drum revolution need not mean the same thing. Preserve the unit used by the procedure. Do not rename all of them “rubs” and compare their totals as if they were identical.

Ask when the specimen is inspected and what counts as failure. The first detected broken stitch differs from a fabric hole or a visible change in colour. Two suppliers can report the same count while measuring different events.

The contact arrangement also needs to match. A flat holder can press a thick joint down. The exposed shape may then differ from the shape on the garment. The laboratory should confirm how it mounts the feature and whether its thickness fits the equipment.

Build a comparison that another laboratory can understand

Give enough detail for someone else to make the same sample. Name the fabric lot, material state, face, direction, thread in each position, needle, stitch type, density and seam dimensions. Keep the approved pattern and sample together.

Prepare the baseline and proposal from comparable material. Take pieces from across the relevant fabric area and label each one. A proposal cut from one unusually strong area should not be compared with a baseline cut from damaged cloth.

Ask the laboratory to manage machine position, run order and rubbing-material changes. Run the baseline and proposal in a way that avoids giving one group only a fresh rubbing surface. Keep the procedure's rules for replacement and cleaning.

The following record makes the comparison easier to audit:

Include a scale and the piece's identity in each photograph. Use the same view at each check. Use light and image size that allow a fair comparison. Can the relevant row be seen without taking the seam apart? If it cannot, plan another way to observe it. Unpicking it at every check would change the test object.

Set the rule before seeing the results. State whether every valid test piece must pass a count. Set any separate appearance limit. Also state what happens if the fabric fails first. Keep a missing result visible in the record. Resolve it through the agreed checks.

Resolve conflicting reports before combining them

Two laboratories may report different counts for apparently similar samples. First check the source records. Were the fabric, finish and sewn construction actually the same? Was the same face rubbed, and did both reports use the same end point?

Check the inspection schedule as well. One laboratory may examine the row more often and detect a break sooner. That difference alone does not show that one material wears worse. Ask for the last intact observation as well as the first failed one.

The ASTM guidance also identifies poor between-laboratory precision for its fabric method. It does not recommend the method for deciding whether to accept a commercial shipment. Do not turn two unrelated reports into a universal supplier ranking. Ask the laboratories and buyer to resolve the basis of comparison.

If another comparison is needed, agree a new plan. Include both seam constructions and use comparable test pieces. Keep a known reference sample where the procedure calls for it. That sample may help explain a change in test response. It cannot make an unsuitable holder or mounting method valid.

Preserve any invalid or stopped result with its reason. A sample that slipped in its holder is different from one whose stitch broke. Repeating the former under agreed rules does not justify deleting the latter. The final table should account for every prepared piece. Show its valid result or explain why that result is missing.

Worked comparison: protected stitching and an inspection cutoff

Consider two fictional versions of a workwear pocket edge. Version A has a more exposed load-bearing row. Version B places the same thread and stitch type beside a folded edge. The row is less exposed but remains visible for inspection.

The team wants to compare complete constructions. The fold changes the contact profile, so this is not a test of the thread's intrinsic abrasion resistance.

A laboratory uses one agreed development procedure for both versions. It confirms that the fixture loads the intended feature. Five independent test pieces, called specimens, are prepared for each version. They come from the same fabric lot and state. All are sound before testing.

For this teaching case, inspections occur after each 1,000 machine-defined cycles. Testing ends when a broken stitch is first found in the named row. If the row stays intact, testing stops at 12,000 cycles. Fabric breakdown would also stop the test and be reported separately. None occurs in the results below.

The fictional first approval rule requires all five rows to stay intact through 8,000 cycles. It also requires no fabric breakdown by that point. These counts and rules are project assumptions, not standard requirements.

At the 8,000-cycle check, only one A specimen remains intact. Its observed intact fraction is 1 ÷ 5 × 100 = 20%. All five B specimens remain intact, giving 5 ÷ 5 × 100 = 100%.

That is an 80 percentage-point difference in these small groups. A fails the stated all-specimen rule. B passes this initial screen, with its other product checks still outstanding.

This result supports developing B further under the tested setup. It does not prove a failure rate for the whole population. Nor does it show five times the service life or prove that B suits every pocket shape. Repeat the comparison where differences in material or sewing conditions could change the outcome.

The inspection count is not an exact failure time

A piece first found broken at 6,000 cycles was sound at the earlier 5,000-cycle check. The break occurred after 5,000 and by 6,000. The report can give 6,000 as the first detection count. It cannot say the break occurred at that exact cycle.

The observation interval is 1,000 cycles. That interval describes how often the specimen was checked. It is not a statement about machine accuracy.

For B, two specimens remain unbroken when testing stops. The NIST reliability handbook distinguishes such cutoff observations from known failures. It also explains data where failure is known only within an interval. Both distinctions matter here.

It may be tempting to enter 12,000 as a failure for the two survivors:

(10,000 + 12,000 + 11,000 + 12,000 + 9,000) ÷ 5 = 10,800 cycles.

That arithmetic is correct, but “mean seam life of 10,800 cycles” would be a false label. Two values are stopping counts, not observed breaks. The other three are first-detection counts, not exact failure times.

Report the individual records and state that 2 of 5, or 40%, remain intact at the cutoff. The test has not established when those two will fail. To estimate a failure distribution, get a suitable analysis and sampling plan. Do not replace unknown failure times with the count at which testing stopped.

Keep a fabric failure from disappearing behind a thread result

Suppose a later candidate leaves the load-bearing row intact but wears through the folded fabric at the 7,000-cycle check. That is a different failure mode. It fails the example's fabric requirement before the 8,000-cycle target.

Do not write “thread passed 12,000 cycles” if testing stopped at 7,000. The thread was observed intact only to the stopping point. Nor should a fabric hole be recorded as a broken-thread event.

The complete assembly may need a wider bearing area, another fold or a different material. Those are proposals for further checking. They are not automatic remedies, and they may affect fit, bulk or sewing access.

A seam can also retain unbroken threads while losing some load-bearing capacity. If the product needs a strength check after abrasion, set that test out separately. Use suitable matched pieces and the agreed sequence. A seam that looks intact still needs a test to show how much strength remains.

Preserving these distinctions prevents a narrow thread improvement from hiding a garment failure. It also makes the next trial more useful: the team knows which part wore out and what remained sound.

Test the required care state with separate specimens

Washing, drying and production finishing can change the joint that reaches the wearer. The question is which state the requirement covers. An unwashed development seam and a finished garment after care are different test objects.

ISO 6330:2021 provides domestic washing and drying procedures for textile tests. Its public scope describes a complete test as washing plus drying. It notes that the chosen settings can affect results. Record the actual route and number of repetitions. “Washed” alone is not enough.

Keep industrial laundering and garment finishing distinct from that domestic route. Ask for the procedure appropriate to the product. Do not claim that one route represents every other form of care.

For an original planning example, compare two constructions in two states, with five abrasion specimens in each group:

2 constructions × 2 treatment states × 5 specimens = 20 abrasion specimens.

If the team also retains two unabraded companions per construction and state, that adds 2 × 2 × 2 = 8 companions. Total preparation becomes 20 + 8 = 28 pieces. The companions are additional pieces, not specimens that have already been worn in the machine.

This allocation is illustrative. Use the laboratory's method and the project's evidence needs to set the actual count. Extra pieces may be needed for setup, invalid tests or another property.

Do not merge results from the two care states to improve an average. A strong result before washing cannot cancel a failure in the state the buyer requires. Record both, along with the treatment identity and any change in seam shape.

Confirm use and preserve the approved construction

A controlled test helps answer one wear question. A garment trial then helps check whether the sample represented the real contact. Use a defined activity, recorded care and agreed inspection points. Do not turn a few informal wear comments into a lifetime guarantee.

Examine starts, ends, crossings and thickness changes. A long straight test row may leave these weak points unexamined. Check both the protected section and the place where protection ends.

A revised seam may require different material use. ApparelDiary's sewing-thread consumption guide helps translate the approved stitch construction into thread needs. When reducing material use, keep the feature that produced the accepted wear result.

Release the exact construction that was assessed. Record the seam dimensions, materials, thread positions, balance and treatment route. Keep the specimen identities, method version and individual results with the approval.

ApparelDiary's quality assurance guide explains how documented controls support consistent output. For this task, review any change in the exposed thread path before it enters production.

When a later complaint appears, compare the returned seam with that record. Determine whether the same construction was made and whether its actual use matched the trial. The next improvement should follow that evidence, with the worn material and thread telling the same story as the test report.

References

• AMANN: Sewing Thread Abrasion Resistance.

• AMANN: Seam Abrasion Resistance.

• Coats: Soft and Secure Seams.

• Munshi, Pai and Ukidve: Studies on Abrasion of Sewing Threads with Scanning Electron Microscopy.

• ISO 12947-2:2016: Fabric Abrasion by Martindale, Specimen Breakdown.

• ASTM D4966-22(2026): Abrasion Resistance of Textile Fabrics.

• SATRA TM95:2020: Abrasion and Snagging Resistance, Drum Method.

• NIST Engineering Statistics Handbook: Censoring.

• ISO 6330:2021: Domestic Washing and Drying Procedures for Textile Testing.