Shoe work is where high frequency welding gets unforgiving. The parts are small, the stacks are thick, the edges are visible on the finished product, and the curves are real: toe springs, heel counters, arch contours. On a flat panel a 1 mm gap under the electrode is a slight weakness; on a shoe upper it is a delamination that shows up after two weeks of flexing. This article is about how HF welding with synchronous cutting actually behaves on uppers and insoles, and where the usual setup habits fail.
What "Synchronous Weld and Cut" Means in Practice
The defining feature of this machine class — the CH-12KW-DSRD and CH-15KW-DSRD in Chenghao's range — is that welding and cutting happen in the same press stroke. The die carries two working edges: an electrode face that fuses the thermoplastic layers, and a cutting rule just outside it that shears the part free. The page describes it as "high frequency welding first, then cut off at the same time," and that ordering is the whole process window: the material must reach fusing temperature before the cutting edge lands, or you get a clean-cut part with a weak seam; if the cycle runs long, the cut edge burns and whitens.
Because there is no secondary trimming station, the cut edge is the product edge. Every defect at that edge — burrs, whitening, crushed foam — ships to the customer. This is why shoe dies demand tighter tooling discipline than packaging dies, and why the electrode design choices matter more here than on almost any other product family.
The Stack Decides Everything
A shoe upper or insole is never a single film. A typical stack is a face fabric, a foam or padding layer, sometimes a reinforcing scrim, and a weldable film (PVC or TPU) somewhere in the sandwich. Only the film is strongly dielectric-active. The fabric and foam are, for HF purposes, passengers — they get hot by contact, not by the field.
Three consequences follow, and they explain most shop-floor confusion:
- The bond exists only where film meets film or film meets a compatible layer. If the die pressure pushes the foam out of the way before the film surfaces touch, the weld line is starving no matter what the power setting says.
- Foam thickness is your thermal insulator. Doubling the foam roughly doubles the lag between platen temperature and the film interface — the same reason dwell dominates over setpoint in how material properties affect the HF process.
- Fabric matters at the cut edge. Loose-weave or stretch fabrics shear badly under a worn rule; the weld can be perfect and the edge still frays.
The Land Between Weld Line and Cutting Edge
The strip of material between the electrode face and the cutting rule — the land — is where synchronous dies succeed or fail. Too narrow and the cutting action disturbs the soft weld zone, tearing it open as the rule descends. Too wide and you add an unbonded channel between seam and edge, which on waterproof or padded styles becomes a wicking path and a place for the foam to pull away.
In practice we set the land with the material, not the drawing: run a step test on the real stack, cut the coupon open, and look at where the fusion front actually stops relative to the rule mark. On thin PVC film over 2–3 mm foam the usable land is tight; on thicker stacks there is more room, but more room also means more edge area to keep clean. When a die starts producing intermittent weak spots along an otherwise consistent seam, a worn or nicked cutting rule is more often the cause than any electrical parameter.
Power and Pressure: What the 30/50-Ton Rating Is Actually For
It is worth being blunt: the 30-ton (CH-12KW-DSRD) and 50-ton (CH-15KW-DSRD) press ratings exist for the cutting stroke, not the weld. Welding a shoe stack needs force to hold the layers in contact — usually a small fraction of the available tonnage. Cutting through fabric, foam and film along a steel rule is what eats tonnage, and that demand rises with blade length and stack thickness, not with electrical power.
The practical trap is pressure density. The working table is 400 × 600 mm, but the die occupies a small part of it. Thirty tons concentrated on a die with, say, 100 cm² of active cutting edge is an enormous local pressure — more than enough to crush foam cells and blow out the weld land if the regulator is set "by feel." Set weld pressure and cut pressure as separate concerns where the machine allows it, and record both in the recipe. The factors that decide HF weld strength still apply inside the weld line; pressure beyond that line is a cutting variable.
Curvature: Why Toe and Heel Fail First
Flat samples weld beautifully; the production part delaminates at the toe. The reason is gap geometry. Over a curve, a rigid die touches the crown of the radius first, and the flanks only make contact as the stack compresses. The crown therefore receives heat and pressure earlier and longer than the flanks — so the crown over-welds while the flanks are still warming up. On deep contours the flank contact can arrive after the cutting edge has already started to land.
The countermeasures are tooling-side, which is why shoe dies are milled to match the part rather than cut flat: a contoured lower die, or a resilient facing (silicone or hard rubber sheet) that lets pressure distribute into the flanks. If delamination clusters at curvature changes, the fix is in the die or the cushion — not in more power, and not in a longer cycle that will further overcook the crown.
Defects We See Most on Uppers and Insoles
| Symptom | Most likely cause | Change first |
|---|---|---|
| Clean edge, seam peels by hand | Cut landing before fusion completes | Shorten cut delay / extend weld phase |
| Whitened, shiny cut edge | Over-cycle or excess cut pressure | Shorten dwell; reduce cut pressure |
| Delamination at toe/heel curves | Rigid die gap over radius | Contour the die; check facing cushion |
| Burr or frayed edge, weld fine | Worn or nicked cutting rule | Inspect/sharpen rule; check fabric fit |
| Foam crushed flat at seam | Pressure density too high for die area | Lower regulator; verify with coupon |
| Repeated spark trips | Moist stack, metal contamination, edge flash | Dry material; inspect die edge condition |
| Size drift through the shift | Material shrinkage after hot cut | Cool under pressure; allow recovery time |
The spark-protection system (NL-5557 on both models) will save the die when a spark occurs, but treat frequent trips as process information rather than nuisance — they almost always point at moisture in the stack, metallic contamination, or a flashing edge from an over-hot weld. The systematic version of this discipline is the same one we apply to reducing defects and downtime in HF production.
Testing That Matches the Product
A peel test on a flat coupon is necessary but not sufficient. Shoe parts flex in service, so:
- Peel in three zones — toe, arch, heel — on parts from the actual die, not coupons. Curvature defects hide in the toe and heel; the arch is nearly flat and will pass while the others fail.
- Flex the seam 50–100 cycles by hand or jig before judging. An under-fused bond that passes a slow peel can open after repeated bending.
- Check the cut edge under magnification at start-up and at shift end. Edge whitening and micro-burrs are the earliest visible signs of rule wear and of cycle drift.
Choosing Between the Two Models — and Between Layouts
Both models share the 400 × 600 mm table and 27.12 MHz frequency; the differences are power (12 vs 15 kW), input (15 vs 25 kVA) and cutting force (30 vs 50 tons). For most upper and insole stacks the 12 kW machine is sufficient; the 15 kW earns its cost on thicker stacks, larger-area dies, or when cutting through dense materials where the 50-ton rating actually gets used. Buying tonnage you never cut with only costs floor space and electrical capacity.
Layout matters as much as power. The front-and-back folding table suits one operator running one machine — common in sample rooms and short runs. The left-right sliding table lets two operators alternate stations on one generator, which doubles effective throughput where the die cycle, not the loading, is the bottleneck — the same logic as the two-station setups used on the glove welding machine. If you are weighing this machine against the wider HF machine range or browsing the shoe-material machine category, start from stack thickness and die area, not from the brochure's headline tonnage.
FAQ
Can it weld and cut leather?
Genuine leather is not dielectric-active, so HF fuses the film layers in the stack, not the hide itself. Leather uppers with a PVC or TPU film layer weld well; pure leather cutting belongs on a cutting press, not an HF machine.
Why does my seam pass at the arch but fail at the toe?
Almost always die gap over the curvature. The crown of the radius welds first; the flanks make contact late. Contour the die or add a compliant facing, and peel-test at three zones.
Is the 50-ton model better for insoles?
Only if your cutting force actually demands it — long rule paths, dense stacks. Welding itself uses a small fraction of the available tonnage on both models.
What causes the cut edge to whiten?
Over-cycle or excess cut pressure. Shorten the weld phase and reduce pressure before blaming the material — whitening is heat damage at the edge, not a resin defect.
How often should the cutting rule be serviced?
Inspect at every die change and sharpen on evidence (burrs, fraying, whitening), not on a fixed calendar. Rule wear is the top cause of "weld fine, edge ugly" complaints.
Why do sparks keep tripping the machine?
Frequent trips point at moisture in the stack, metallic contamination, or flash at an over-hot edge. The NL-5557 protection prevents die damage; the trips themselves are diagnostic data.
Building Shoe Dies or Planning a Shoe Line?
Send us your stack structure, die area and target output. We help you pick the model, design the die for synchronous weld-and-cut, and set the process recipe before the first production part.
© Chenghao Machinery — High frequency welding machines and forming equipment for industrial manufacturing.