Home / Blog / Materials & Construction

RF Welding Explained for Dry Bags

What radio-frequency welding does to a TPU film

A radio-frequency welder is a press with a radio transmitter attached to it. The upper tool, usually a shaped brass electrode, carries the high-frequency energy; the lower platen is grounded. When the press closes on two layers of film, an alternating electric field passes through the material between the two electrodes and the film becomes the dielectric of a capacitor.

Because the heat is generated inside the material rather than conducted in from a hot surface, the whole thickness reaches melt temperature at close to the same moment. That is why a 0.5 mm TPU laminate can be fused in under three seconds while a hot-air welder would need several times longer and would scorch the face fabric first.

When the generator stops, the press holds the parts together while the melt cools. This cooling phase is not optional: the polymer chains need to re-entangle under pressure, and releasing the platen early produces a seam that looks perfect and peels apart in the hand. Cooling under pressure typically runs 0.8 to 2 seconds, roughly half the dwell time.

Why TPU welds and polypropylene does not

The process depends on dielectric loss. Materials whose molecules carry an electrical dipole, such as TPU, PVC, EVA and some polyamides, try to align with the alternating field, and the friction of that rotation becomes heat. Materials without a dipole, such as polyethylene and polypropylene, barely respond at all, which is why a polypropylene woven bag cannot be radio-frequency welded no matter how much power you apply.

This is the practical dividing line in waterproof bag construction. A TPU-laminated shell welds cleanly and quickly; a polyethylene-coated shell has to be joined with hot air or a hot wedge, which melts the coating from the outside and needs slower, hotter cycles that are harder to keep consistent across a long seam.

Coating weight changes the picture too. A film with a higher TPU content responds faster and produces a broader processing window, which is why production lines prefer 0.4 to 0.6 mm laminates for welded bags. The trade-offs between the two common thicknesses are set out in our film thickness comparison.

The four parameters a technician actually sets

Every weld on a production line is defined by four numbers: generator power, platen pressure, dwell time and cooling time. Voltage and frequency are fixed by the machine design, so these four are what a technician tunes when a new fabric lot arrives or the weather changes the moisture content of the material.

These ranges interact, which is why a weld recipe is transferred between machines as a set rather than as individual numbers. Raising power by 20% without shortening the dwell will over-weld; raising pressure without adding dwell will squeeze molten polymer out of the joint and leave a thin, weak band that looks narrower than it should.

Ambient conditions matter more than people expect. A humid morning puts moisture on the film surface, moisture absorbs energy, and the first few welds of a shift can run cold. Production lines handle this by running scrap parts through the first cycles and logging the first inspection results separately.

Inside a weld cycle, step by step

1. Operator loads the panels against the fixture stops on the lower platen.
2. Press closes and applies the set pressure before any energy is applied.
3. Generator delivers 27.12 MHz energy for the programmed dwell time.
4. Generator stops, pressure holds while the melt cools and re-solidifies.
5. Press opens, operator unloads and the seam goes to visual inspection.
6. Every 30 minutes, one part is pulled for a peel test on a tensile rig.

The sequence looks simple, but the order of pressure before energy is what makes it repeatable. If energy is applied while the press is still closing, the gap between the electrodes changes during heating, the field strength changes with it, and the seam strength varies from part to part.

Weld geometry: width, corners and electrode shape

Weld width is a design decision, not a machine limitation. A closure seam on a small roll-top needs 8 to 10 mm of weld; a seam carrying a handle, a D-ring or a shoulder strap anchor needs 15 to 20 mm, because the load spreads across the bonded area and peel stress concentrates at the ends of the band.

Corners are where welded bags fail. At an inside corner the electrode has to change direction, the field crowds, and the film tends to thin. Good tooling uses a radius of at least 6 mm at every direction change and avoids sharp 90-degree turns in the weld path entirely.

Electrode shape also determines how the seam looks. A flat electrode leaves a plain pressed band; a stepped electrode produces a recessed line that hides the melt edge and gives a cleaner appearance on a coloured shell. Both are equally waterproof, so the choice is cosmetic unless the step reduces weld width.

Strength numbers worth asking for

A weld is measured three ways: peel, shear and burst. Peel is the weakest mode and therefore the one that matters, because a bag in service is much more likely to peel a seam open than to slide it apart. Shear numbers look impressive and tell you very little about real life.

Ask for the peel number together with the failure mode. A weld that peels at 45 N through the film is a good weld, because the film itself failed before the bond did. A weld that peels at 45 N cleanly at the interface is a marginal one, and it will degrade faster in UV and salt.

Our line runs peel tests every 30 minutes on a calibrated tensile rig and logs the results against the machine and the operator. Finished goods are inspected to AQL 2.5 with hydrostatic and air tests on the seam path, and the wider quality picture is covered in our quality control notes.

RF welding against the alternatives

Radio-frequency welding wins on consistency at volume, because the energy goes exactly where the electrode is and the cycle is timed rather than judged. Hot wedge welding is the right answer for very long straight seams and for materials that will not respond to a dielectric field; sewing plus tape remains the fallback where a fabric simply cannot be welded.

The taped route has one structural weakness: the tape is an adhesive layer bonded to the fabric, and adhesive is the first thing to let go in heat, salt and age. Our seam tape versus welded seam comparison goes through the service-life data, and the stitched versus welded guide covers why needle holes are the underlying problem.

Failure modes and what they look like

• Cold weld: the seam has a glossy surface but peels with finger pressure. Cause is low power, short dwell or moisture on the film.
• Burn-through: a brown or yellow line down the middle of the weld, with the face fabric visibly thin. Cause is excess power or long dwell.
• Squeeze-out: the weld band is narrower than the electrode and a bead of polymer has pushed to one side. Cause is excess pressure.
• Arcing: a pinhole with a black halo, often near a fold or a contaminated area. Cause is moisture, dirt or a damaged electrode surface.
• Corner thinning: the film is translucent at a direction change. Cause is a radius that is too tight for the electrode.
• Green weld: the seam looks correct but creeps open over a day. Cause is releasing pressure before the melt has cooled.

Every one of these is visible without instruments, which is why operator training on visual inspection is worth more than any single test machine. A trained operator catches arcing pinholes and squeeze-out at the press, long before the part reaches an immersion tank.

Arcing deserves particular attention because it can also damage tooling. A carbon track on the electrode surface concentrates the field and makes the next arc more likely, so electrodes are cleaned at every shift change and re-faced whenever a track appears.

Testing welds on the line and in the laboratory

Line testing is statistical and fast: visual inspection of every part, an air or hydrostatic test on a sample, and a destructive peel test on a schedule. Laboratory testing is slower and rarer: tensile rigs, accelerated UV cabinets, salt spray chambers and controlled immersion tanks.

The two must agree. If laboratory peel results run 20% below line results, the difference is usually fixturing, because a hand-held strip cut at a slight angle introduces peel stress that the rig records as a lower number. Cutting jigs and a fixed strip width remove that error.

Immersion testing deserves a note on interpretation. A bag that survives 30 minutes at one metre has demonstrated that its seams close a water path under modest pressure; it has not demonstrated resistance to abrasion, puncture or long-term UV. Read our immersion testing guide before treating an immersion pass as a durability claim. The same caution applies to abrasion: a welded seam can pass immersion and still be the first place a dragged bag wears through, which our abrasion notes cover in detail.

Welding windows by film thickness

Thicker film is more forgiving in one sense and less forgiving in another. It tolerates a wider range of dwell before burning through, but it also needs more energy to reach melt through the full thickness, so an under-powered machine simply cannot weld it at any dwell time.

Face fabric matters as much as film. A 500D polyester face pulls more heat out of the joint than a 210D face does, so the same film on a heavier shell needs a longer dwell. That is one reason weld recipes are written per fabric, not per film.

What welding makes possible in bag design

Welding changes the design vocabulary. Clear windows can be welded directly into a shell without a stitched frame, handles become load-bearing panels rather than webbing tacked to fabric, and a purge valve can be seated on a welded flange instead of a glued patch.

It also changes where a bag can be loaded. A welded D-ring panel distributes a strap load across 15 to 20 mm of bond, whereas a sewn loop concentrates it at the stitch line. On a rental bag that gets dragged across a beach every day, that difference decides whether the anchor survives the season.

The limitation is tooling cost. Every weld geometry needs an electrode, and an electrode for a complex panel costs real money, which is why welded construction suits a design that will be repeated in volume. Short runs are better served by simpler tooling with more manual operations.

Energy and throughput on a real line

Energy per weld is easy to estimate. A 6 kW generator running a 2.5 second dwell delivers 15,000 joules, which is 4.2 watt-hours; a bag with six welds therefore uses about 25 watt-hours of radio-frequency energy, before counting the press, the compressor and the lighting around it.

Throughput is governed by handling rather than by welding. A 2.5 second dwell plus 1.2 seconds of cooling plus load and unload puts a simple seam cycle at 12 to 18 seconds, so a single station produces 200 to 300 parts per shift and a multi-station cell multiplies from there.

Across our own operation this scales to 7 lines and 149 machines with a 200,000-piece monthly capacity. Energy per piece is tracked station by station, and the detail of how that is measured and reduced is covered in our energy use notes.

Spec checklist for welded construction

• Ask for the peel strength number and the failure mode: film failure is good, interface failure is not.
• Ask for weld width at load-bearing points; 15 mm or more for handles and strap anchors.
• Ask whether corners use a radius of at least 6 mm rather than a sharp turn.
• Ask how often destructive peel tests are run, and what happens when one fails.
• Ask which fabric and film combination the weld recipe was written for.
• Ask for the immersion standard behind any waterproof claim rather than an adjective.
• Compare constructions across the product line and the TPU versus PVC guide before you specify.
• Check the definitions in our dry bag glossary if a term on the sheet is unclear.

For anyone specifying a welded bag for a fleet or a retail range, these eight questions separate a genuine welded construction from a sewn bag with a welded logo tab. We have produced waterproof bags since 2014 in an SGS-verified 4,950 m2 facility with 137 staff, a 500-piece minimum, sampling in 6 to 10 working days and bulk in 35 to 50 days, FOB Xiamen.

Three standards bodies sit behind the testing language used on a welded bag specification. The IEC publishes the safety standards for industrial electroheating and radio-frequency equipment, the ISO publishes coating adhesion and seam strength methods for coated fabrics, and ASTM publishes the coated-fabric test methods most North American laboratories quote.

One last practical note for owners rather than specifiers. A welded seam can be repaired in the field with a patch and a heat source, but the repair will never match the original peel strength, so treat welded construction as a reason to avoid damage rather than a reason to risk it. Our TPU construction guide and the waterproof backpack range show where welded panels sit on a finished product.