One of the more counterintuitive facts in electrical fire investigation is that a connection can glow hot enough to ignite its surroundings while the circuit draws entirely normal current. Overcurrent protection never operates, because there is no overcurrent. The breaker is doing exactly what it was designed to do, and the fire starts anyway.

The runaway that has no alarm

Heat generated at a connection is proportional to its resistance and to the square of the current through it. A sound connection has very low resistance and generates negligible heat. But the true contact area at any separable interface is a small fraction of the apparent area — current crosses at a limited number of microscopic asperities. Anything that disturbs those contact points raises resistance.

What follows is self-reinforcing. Higher resistance produces more heat at the joint. Heat accelerates oxidation of the contact surfaces and drives stress relaxation in the spring member that maintains contact force. Both effects raise resistance further. The joint degrades itself, and the process compounds until the terminal reaches temperatures capable of igniting adjacent polymeric housings and insulation. Because the current never exceeds rating, nothing upstream registers a fault.

How connections get there

Fretting corrosion is a leading path. Vibration or thermal cycling causes mated contacts to move against each other by microscopic amounts. That motion abrades the plating and generates oxide debris which, being insulating, becomes trapped in the interface. Resistance climbs in steps, which is why the early symptom is usually an intermittent fault rather than heat.

The mechanism has been studied extensively in the contact-physics literature, including vibration-induced fretting in connector systems (IEEE HOLM 2016), lifetime modelling for connectors subject to fretting (IEEE HLM 2025), and the degradation and recovery behaviour of tin-lead plated contacts (IEEE HOLM 1997).

Other routes reach the same place. Crimp defects — under- or over-compression, the wrong die, insulation trapped in the conductor barrel, damaged strands — create a high-resistance joint from the first day of service. Loss of contact force through creep and stress relaxation reduces normal force below what a stable interface requires. Dissimilar-metal joints, notably aluminium to copper, add galvanic corrosion and differential expansion. Moisture, salt, and chemical exposure attack base metal through plating pores.

What survives, and what to preserve

Connector fires destroy a great deal, but the terminal itself usually still carries a readable record. Arc and heat damage at the contact zone, oxide and debris morphology, plating wear patterns, annealing of the spring member, and the geometry of the crimp all persist. Comparing the failed termination against others from the same harness or production run is frequently what separates a manufacturing defect from service degradation.

The preservation instruction is specific and routinely violated: do not unplug it. Separating a suspect connector destroys the as-found contact evidence — the exact interface that records the failure. The assembly should be photographed in place and preserved mated, together with both halves, sufficient wire length for crimp sectioning, and the surrounding mounting hardware so the thermal relationship between the terminal and the fire damage can be established.

Why it matters

Connector and terminal failures appear in incidents far larger than the part: structure and equipment fires, product recalls, automotive and EV charging matters, and insurance subrogation. Origin-and-cause work under NFPA 921 frequently turns on whether a given connection was the heat source or merely a casualty of a fire that started elsewhere — a question answerable from the physical evidence, but only if that evidence survives.

For the full mechanism set and examination protocol, see our specialization area on connector and terminal failure analysis.