Common Heat Staking Failures and How to Fix Them

Most heat staking failures fall into a small number of named categories: under-formed heads, over-formed heads, cracked or split bosses, sticking and stringing on retract, witness marks on cosmetic surfaces, loose joints after cool-down, and inconsistent cycle behaviour from one part to the next. This guide names each failure, identifies the root cause, and lists the design or process fix. Most of them are fixable before the part hits the line if the boss geometry, the resin, and the cycle parameters are reviewed together.

Failures usually come from one of four sources: wrong boss geometry, wrong resin choice, wrong tip profile or temperature, or wrong cycle parameters. The right fix depends on which source it is — changing the dwell time when the boss is undersized just damages more parts faster, and changing the resin when the cycle is mistuned just moves the failure around.

Under-formed Heads

Head does not fully fill the tip cavity or does not spread wide enough to clamp the captured part. Root causes: stud volume too small for the head profile chosen; tip temperature too low (hot probe) or cycle energy too low (impulse); insufficient cycle dwell. Fix: recalculate stud volume against the head profile, increase impulse energy setting incrementally, or change to a head profile that needs less plastic (flush instead of dome, for example).

Over-formed Heads

Head spreads beyond the clearance hole and creates a cosmetic flash or bulges around the captured part. Root causes: too much stud volume, tip temperature too high, dwell too long. With impulse Weld by Energy control this is essentially eliminated because the press delivers a metered energy and stops; with hot-probe staking it is the most common over-cycle failure.

Cracked or Split Bosses

Boss splits longitudinally during or after staking. Root causes: boss wall too thin relative to stud diameter; stud volume too high for the boss diameter; sharp corners at the boss base creating stress concentration; resin too brittle (acetal, PPS in some grades) for the geometry chosen. Fix: thicken the boss wall to at least the boss wall thickness rule, add a generous radius at the base, reduce stud volume, or switch to a more forgiving resin if cosmetic surface allows.

Sticking and Stringing on Retract

Molten plastic clings to the tip on retract, leaving a stringer across the part or pulling the head off-shape. Root cause: tip retract while plastic is still molten — standard hot-probe behaviour. Fix: impulse heat staking with cold-tip retract eliminates this by design (the tip cools under compressed air before retract); for hot-probe systems, add a coated or surface-treated tip and tune retract timing.

Witness Marks on Cosmetic Surfaces

Visible drag, shine, or imprint on the A-surface of the part. Root causes: hot tip dragging molten plastic across the surface on retract; tip pressing into the boss past the design head height. Fix: impulse cold-tip retract eliminates the drag; encoder-controlled head positioning eliminates the over-press; for hot-probe systems, tighten cycle parameters and accept that some risk remains.

Loose Joints After Cool-Down

Head appears formed but the captured part rattles or pulls free. Root causes: head does not clamp because clearance hole is too large or head spread is too small; plastic shrinks more than expected during cool-down; captured part has spring-back that the head cannot resist. Fix: tighten clearance hole tolerance, choose a head profile with more clamp area (dome instead of flush), or add a secondary feature (snap, rib) to constrain the captured part.

Inconsistent Cycle Behaviour

The same recipe produces different head quality from part to part. Root causes: resin batch variation, ambient temperature drift, operator variation, tip temperature drift on hot-probe systems. Fix: impulse Weld by Energy control eliminates most of this by holding delivered energy constant; for time-based hot-probe systems, the realistic answer is to upgrade to impulse control rather than chase the drift continuously.

Get a Process Review

Send failed parts and the cycle parameters as part of any quote and we will identify the root cause and recommend the fix — sometimes a tooling change, sometimes a geometry change at the next moulding revision. Contact us for a process review.