Seam Strength Testing for Knitted Bags: Where the Seam Fails First, and How ISO 13935-2 and ASTM D5034 Read the Same Failure Differently

A Handle That Held in the Laboratory and Failed in the Shop

A buyer launched a knitted market tote with a strong specification: body panel tested to ISO 13934-1 at a minimum of 600 newtons, abrasion to ISO 12947 at 20,000 cycles, colour fastness to ISO 105-X12 at a minimum rating of 4. Every report came back comfortably inside limit and the launch went ahead on a twelve thousand piece first order.

Ten weeks into the season the returns desk started logging handles detaching at the top join. The fabric was undamaged in every returned bag; the join had released. Nothing in the specification had ever measured the join, so the programme had three strong reports about the panel and no report at all about the thing that failed.

The fix was not a stronger fabric. It was a seam specification: ISO 13935-2 on the handle join with a minimum force figure, a stitch density recorded on the approved sample, and a pre-bulk test on five assembled bags rather than on flat specimens. The following season's returns rate on that construction dropped to near zero without the panel specification changing at all.

Why the Seam Is Usually the Weakest Line on a Bag

A seam is a discontinuity. Stitching perforates the fabric, concentrates load at the needle holes, and introduces a second material, the sewing thread, whose strength and extension behaviour rarely match the panel. On a knitted panel the effect is amplified, because the loop structure has nothing to anchor the thread against in the way a woven interlacing does.

So a bag is a chain of parts with different strengths, and the chain breaks at its weakest link. Fabric tensile testing measures one link; seam strength testing measures the link next to it; hardware attachment testing, where the design includes a rivet or a bar tack, measures another. Testing only the fabric is measuring the strongest link and calling it the bag.

The commercial point is that a seam failure and a fabric failure look identical to a customer and completely different in a laboratory. A specification that only measures the fabric cannot distinguish them, which is why the post-mortem on a returned bag so often concludes that the reports were all in limit.

ISO 13935: The Strip and Grab Routes

ISO 13935 covers seam tensile properties in two parts. Part 1 uses the strip method, gripping the full specimen width; Part 2 uses the grab method, applying force perpendicular to the seam on a specimen gripped across part of its width. Part 2 is the one most bag programmes use, because it accommodates a made-up article and a bulkier seam assembly more easily.

The 2026 edition of Part 2 is explicit about scope: it applies to straight seams only, whether taken from previously sewn articles or prepared from fabric samples, and not to curved seams, with guidance on curved seams in an annex. The method is restricted to constant rate of extension machines, which is worth stating on the request because it excludes older constant-rate-of-traverse frames.

That scope note matters on a bag. A handle join on a shaped tote, a curved base seam and a gusset corner are all curved or angled, so strictly they sit outside Part 2. The practical answer is to test a straight representative seam for specification purposes and add a whole-bag load test for the actual geometry, which is what the next section covers.

ISO 13935-1

Seam maximum force to rupture, strip method, full width gripped

ISO 13935-2

Seam maximum force to rupture, grab method, force perpendicular to the seam

ISO 13935-2 scope

Straight seams only, from made-up articles or prepared samples, on constant rate of extension machines

ISO 13936-1

Seam slippage, determined at a fixed seam opening

ISO 13936-2

Seam slippage, determined at a fixed load

ASTM D1683

Failure in sewn seams of woven apparel fabrics, the common North American route

Specimen source

Cut from production-made bags where possible, not from laboratory-prepared swatches

Stitch record

Stitch type and stitch density per unit length recorded on the approved sample

Reported result

Maximum force to seam rupture, plus the failure mode observed

Seam Slippage Versus Seam Rupture: Two Different Complaints

Rupture is the seam coming apart. Slippage is the seam staying intact while the fabric pulls out from under the stitching, leaving a gap that looks like a failure and does not recover. On a knitted panel slippage is the more common of the two, because the loop structure offers less resistance to yarns being drawn past the thread.

The two are measured by different standards. Rupture runs to ISO 13935. Slippage runs to ISO 13936, in two parts: Part 1 determines slippage at a fixed seam opening and Part 2 at a fixed load. Choosing between them is usually a matter of which the buyer's existing laboratory runs, and either is far better than neither.

The distinction matters commercially because a customer cannot tell them apart and a returns log will not either. A bag returned with a gaping seam may have a seam well inside its rupture limit and still be unsaleable, which is why a specification covering only rupture can be fully met by bags the customer rejects.

What Changes the Number: Stitch, Thread, Seam Allowance

Three variables dominate a seam result, and all three are straightforward to get right. Stitch type sets the structural behaviour: a lockstitch classified as type 301 under ISO 4915 behaves differently from a chainstitch, and a chainstitch can unravel from an end where a lockstitch will not. Record the stitch type on the approved sample.

Stitch density is the second. More stitches per unit length generally raises seam strength up to a point, after which the needle damage starts to dominate and the curve turns over. Recording density in stitches per unit length on the approved sample is the single most economical control available, and it is what a laboratory needs in order to reproduce the result.

Seam allowance is the third and the most often overlooked. A narrow allowance lets the fabric pull out rather than break, so widening it raises the measured force materially. Where a buyer sees two laboratories disagree on the same seam design, the allowance is the first thing to check and usually explains the gap.

Testing a Whole Bag Rather Than a Seam Specimen

There is a strong argument for testing the join as it exists in the bag rather than as a flat specimen. A cyclic load test on an assembled bag, loading the handle to a defined force for a defined number of cycles and then inspecting, reproduces the failure the returns desk sees and does not depend on the seam being straight.

A workable protocol is simple: load the assembled bag to the intended carry weight multiplied by a safety factor, hold or cycle it a defined number of times, then inspect the join for slippage, elongation and rupture. Five bags per colourway is enough to catch a systematic problem, and the test needs no instrument beyond a load frame.

This is not a substitute for ISO 13935-2, because it produces no comparable number. It is a complement that catches curved seams, angled joins and hardware attachment, all of which sit outside the scope of the standard and all of which appear on real bags.

Where a Knitted Seam Behaves Unlike a Woven One

A woven seam fails when the yarns break or pull out past the thread. A knitted seam adds a third mechanism: the loop unravels. Where the stitching runs across the course direction, a broken thread can initiate a run, and a run propagates far further than the original damage, turning a small join failure into a large visible one.

That is why stitch choice matters more on a knit. A chainstitch that unravels from an end is a poor choice on a knitted bag join even where the measured strength is adequate, because the failure mode after damage is worse than the measurement suggested. A lockstitch, or an overlocked edge, behaves more predictably.

The second knit-specific issue is recovery. A knitted seam can hold the test load and then leave the join permanently stretched, so the bag passes the force test and looks wrong on the shelf. Inspecting the specimen after the test, not just reading the force at rupture, is what catches this.

Setting the Acceptance Figure for a Bag Seam

Derive the seam figure from the load rather than borrowing one. Estimate the heaviest realistic carry weight, apply a safety factor, and require the join to hold that with margin, then confirm by cycling five assembled bags at the same load. A handle join on a market tote and a shoulder strap join on a travel bag justify very different numbers.

A useful rule of thumb is to require the seam to reach a defined proportion of the fabric's own tensile result, so the join is never the weakest link by design. Where the fabric reports 600 newtons and the join reports 180, the fabric is over-specified and the join is under-specified, and both numbers were expensive to obtain.

Record the failure mode alongside the force. A seam that ruptures at 420 newtons and a seam that slips at 420 newtons are not the same seam, and only the recorded mode tells a buyer which one was tested. Ask for it on the report; it costs the laboratory one line.

Writing the Seam Clause

The clause needs the method and part, the stitch type and stitch density recorded on the approved sample, the seam allowance, the minimum force, the required failure mode, and the specimen source. Add a whole-bag cyclic load test with a stated load and cycle count, and the clause covers both the flat geometry and the real one.

Frequently Asked Questions

People Also Ask