Home / Blog / Materials & Construction

Welded vs Stitched Dry Bag Seams

The two ways a dry bag panel can be joined

Every dry bag has to close the gap between two cut edges, and there are only two families of solution. Fusion joins melt the polymer on both faces so the joint re-solidifies as one mass, whether by radio-frequency energy, hot air, a heated wedge or ultrasonics. Mechanical joins pass a thread through both panels and rely on friction, stitch density and sometimes a covering tape to resist water.

Fusion only works on thermoplastics, which is why TPU and PVC dominate this category. Once TPU has been laminated onto a woven face fabric, the fusing surfaces must still be film against film, so panels are folded, taped or edged to expose the right layer. A woven shell with a thin surface treatment has nothing to melt, and no amount of heat and pressure will turn it into a waterproof barrier.

What actually happens at a needle hole under pressure

A Tex 60 needle leaves a hole around 0.6 to 1.0 mm across, and the thread does not fill it completely. At seven stitches per inch a 60 cm seam carries about 165 rows of holes, each one a potential channel. Standing water does nothing to them; the problem begins when the bag is squeezed, sat on or submerged, because pressure pushes water towards the lowest-energy exit.

At one metre of immersion the differential across the seam is close to 9.8 kPa, which is small in engineering terms but plenty to drive water through a channel of sub-millimetre width. Polyester thread also wicks slowly along its length, so even a seam that holds shape can carry moisture into the interior over an hour. This is why a stitch line that survives a shower test can still fail an overnight ducking.

Thread choice changes the picture without solving it. Bonded polyester resists abrasion better than nylon and absorbs less water, and a bonded filament version reduces linting inside the hole, but none of those improvements closes the channel. Adding a second row of stitches does not help either: two rows mean twice the holes, and the second row usually runs within 6 mm of the first, so it sits in the same weakened zone of fabric.

Radio-frequency welding: how the weld is formed

In radio-frequency welding the two films are clamped between a shaped brass or aluminium electrode and a grounded bed, then exposed to a 27.12 MHz field. Polar molecules in the polymer oscillate and generate heat internally rather than from the outside, so both plies reach melt temperature at the same moment. Typical cycles run four to ten seconds at two to five kilowatts, with weld widths anywhere from 10 to 25 mm.

Because heating is volumetric, the weld interface reaches an even melt across its whole width, which is what produces peel strength higher than the surrounding film. The method prefers TPU and PVC, struggles with very thick or highly filled films, and needs an electrode that mirrors the part geometry. A fuller walk-through of tooling, power settings and cycle windows sits in the RF welding explainer.

Hot-air, hot-wedge and ultrasonic joins

Hot-air and hot-wedge machines heat the interface externally, typically delivering air at 350 to 500 °C through a nozzle while a roller pair presses the plies together at two to six metres per minute. Tooling is far cheaper than a shaped radio-frequency electrode, long straight seams run beautifully, and it handles polyethylene and PVC as happily as TPU. Tight curves are harder, because the nozzle needs clearance and the roller needs a consistent nip.

Ultrasonic welding at 20 to 35 kHz is the third option and rarely makes the main body seam. Its strength is localised energy for small hard points: a webbing tab, a valve flange, a strap anchor pressed onto a curved panel in under two seconds. Where a full seam needs consistent heat across 60 cm, ultrasonics excels at a 40 by 25 mm footprint that would be awkward under any platen.

Seam tape is still a stitch: reading bonded seams correctly

A taped seam is a compromise that appears on most mid-market gear: needle and thread carry the mechanical load, then a thermoplastic tape roughly 20 to 22 mm wide is heat-pressed over the stitch line at 120 to 160 °C. Done well, it produces a genuinely waterproof result on flat panels, and it lets a designer keep the strength and geometry freedom of sewing.

The weakness is the adhesive layer, not the tape film. Under heat and humidity those bonds creep; at corners the tape has to fold over itself and tends to lift within one to two seasons. Repeated creasing along the same line breaks the bond even faster, which is why a taped roll-top usually fails at the mouth fold rather than along the side seam. The tape versus weld comparison sets out where each approach earns its place.

Tear, peel and shear figures side by side

Three measurements matter and they are routinely confused. Grab or tongue tear tells you whether the fabric rips; peel tells you whether the joint opens from its edge; shear tells you how much sliding load the seam takes before it shifts. For dry bags, peel is the useful number because real failures almost always start at an edge or a corner and propagate inwards.

Test house protocols such as the coated-fabric methods maintained by ASTM International define how specimens are cut, conditioned and pulled, and the equivalent standards family published by ISO covers the same ground for European specification. Ask which method a figure came from, because a peel value is meaningless without the specimen width, pull speed and conditioning behind it.

Failure modes over three years of real use

Fresh out of packing, both systems perform. The divergence begins with exposure. Ultraviolet radiation attacks the polymer chain rather than the join, so a bag left on a deck for two summers loses film strength around the seam before the seam itself changes. Flexing does the opposite: repeated folding along one line fatigues the tape bond and the thread, leaving the welded body largely unaffected.

Cold is the other divider. Below about minus 20 °C some plasticised films stiffen and the stress concentration at a weld edge rises sharply, while thread and tape generally stay more tolerant because neither relies on a melted interface. The practical consequence is that a welded high-specification body can still be the right choice everywhere except a single high-flex zone.

Whatever the join, the failure sequence is well documented in the seam failure walkthrough and quantified against service age in the seam integrity timeline. Both matter more than the marketing adjective on the hang tag, because a seam does not usually fail all at once; it gives up a few drops per minute for a season before it becomes obvious.

Design consequences for pattern and features

Choosing welding changes what a designer is allowed to draw. Nothing can pierce the waterproof shell without becoming a leak path, so every lash point, strap anchor and handle has to arrive as a welded patch. That constraint produces the clean, tape-free silhouettes seen across modern TPU waterproof backpacks, and it also means features tend to be simple, large and rounded rather than decorative.

Sewing imposes almost no geometry limits, which is why early samples and small numbered runs are usually built on a needle first. The hidden cost sits later: each additional stitch line is another candidate failure point and often another tape operation. As volumes rise past a few thousand units, the recurring labour of taping usually exceeds the one-off cost of shaped tooling, and the economics flip decisively towards fusion.

Repair, rework and what failure costs you

A failed weld region is best repaired as a patch rather than re-fused, because once the film has contaminated or thinned, it will not melt back to specification. A clean room-temperature patch with a suitable adhesive and a rolling pressure step recovers most of the original performance within a day of cure. Field repair options are covered in the seam tear repair guide.

Sewing retains a real advantage here. A split stitch line can be unpicked and re-run by any competent repair shop, and lifted tape can sometimes be re-pressed with a household iron through a silicone sheet. That repairability counts for expedition operators who work three days from the nearest town, even where the original spec preferred a tape-free construction.

Testing a seam you are evaluating

You do not need a laboratory to separate a good join from a poor one. A backlight reveals pinholes and uneven melt immediately, a spring balance gives a usable peel estimate, and thirty minutes under a metre of water with paper towel inside will expose any continuous leak path. Do the checks in that order, because destructive testing on a sample you have not yet inspected visually is wasted information.

For immersion claims specifically, the grading comes from IEC 60529 rather than from a brand slogan, and the definitions are maintained by the IEC. IPX7 describes thirty minutes at one metre, nothing more. A seam that passes can still weep under two hours of drag or against a loaded edge pressing it against a hull.

Choosing by use case: a decision table

Most buyers do not need the highest absolute specification; they need the join that survives their specific pattern of abuse. The failures that reach our desk come from three situations: someone creasing the same fold line every day, someone dragging a bag behind a boat, and someone leaving a bag wet and rolled for a month.

Match the join to the dominant risk rather than to the price point. Daily commuting mostly demands tape quality at stress points; sea kayaking demands film integrity under pressure; a rental fleet demands a repair process that any staff member can follow in ten minutes.

Writing the seam line into a product brief

Specify the join as a performance requirement rather than a preference. Name the process, minimum weld width, peel target and test method, then require validation pieces rather than accepting a certificate alone. Terms like AQL 2.5 describe how samples are drawn for inspection; they say nothing about the design unless you also state what constitutes a defect at the seam.

Cost drivers are predictable once you know them: shaped electrodes scale with part geometry, cycle times of four to ten seconds set hourly throughput, and the rejection rate at first article usually falls sharply after the first hundred pieces. Our production has run since 2014 in an SGS-verified 4,950 m² facility with 7 lines, 149 machines and 137 staff turning out around 200,000 pieces a month; standard terms are a 500-piece minimum, sampling in 6 to 10 working days, bulk in 35 to 50 days, FOB Xiamen, T/T 30/70. We document the full process from sampling to bulk for every programme, and current line specifications sit alongside the product range.

One caution on spec writing: do not mix rating systems. A hydrostatic head figure, an ingress protection grade and a spray test describe three different things, and quoting whichever happens to look best is how products end up returned. Pick one primary test, name secondary evidence, and require the same protocol on every production batch.

Care habits that extend seam life either way

Most premature seam failure is caused by storage rather than by the water. A bag compressed into the bottom of a locker with the same three creases every week will fail at those creases whether the panels are welded or sewn. Store bodies loosely rolled, rotate the fold line every few weeks, and never leave a wet bag sealed shut for longer than a day.

Clean with fresh water at ambient temperature, because hot water accelerates hydrolysis in adhesives and plasticisers, and avoid solvents entirely on taped seams since one wipe with alcohol-based cleaner can soften the bond line permanently. Full washing detail is set out in the roll-top cleaning guide. A seam inspected monthly in thirty seconds will outlast one that is never looked at, regardless of which join you chose.