A socket is a spring with a hole in it. Everything that matters about its long-term behaviour follows from that: a spring under load relaxes, relaxation lowers contact pressure, lower pressure raises contact resistance, and higher resistance raises temperature. The chain is slow, silent, and only visible if you measure the right thing at the right time.
What relaxation actually is
Metals under sustained stress creep. In a socket contact, the stress comes from the plug pin pushing the contact spring outward. At room temperature the creep rate is small; at 70 °C, with the socket carrying its rated current, it is not. The spring does not break. It simply stops pushing as hard.
The measurable consequence is a drop in contact force, typically 10 to 25 % over the service life of an ordinary brass contact, and faster for thin springs running hot.
| Stage | Typical contact force | What changes |
| New, from the tool | 8 to 12 N per contact | Reference value for the type test |
| After 100 insertions | 7.5 to 11 N | Surfaces polished, force slightly lower |
| After 5,000 insertions | 6 to 9 N | Wear plus first relaxation |
| After thermal cycling at rated current | 5 to 8 N | Relaxation dominates |
| Below 4 N sustained | Under 4 N | Temperature rise accelerates |
Why the temperature rise arrives late
Contact resistance and force are inversely related but not linearly. A contact at 8 N might sit at 3 mΩ; the same contact at 4 N can sit at 12 mΩ. At 16 A that is a change from 0.77 W to 3.07 W of heat in a component with almost no surface area. The socket housing insulates it. The rise shows up as a slow climb, not a step.
This is why a socket that passed its temperature rise test when new can fail the same test two years later without anything visible having changed.
The measurement that catches it early
Measure force, not just temperature. A simple pull test on a gauge pin with a spring scale gives a number that tracks relaxation directly, and it takes under a minute per sample. Record it at incoming inspection and again after any thermal cycling. A drop of more than 20 % from the reference is the trigger for a closer look, long before the temperature rise test would notice anything.
- Use the same gauge pin every time, and the same insertion depth.
- Pull slowly, at about 10 mm/s, and take the value at the moment of release.
- Measure at 20 ± 5 °C; a warm contact reads higher.
- Record three contacts per socket, not one.
Six hundred insertions is not a magic threshold either. It is roughly where the first two mechanisms, mechanical wear and initial relaxation, have both finished and the remaining margin is thin enough that a hot summer will find it.
Standards referenced: IEC 60884-1 clause 20 (temperature rise), clause 22 (withdrawal force), clause 24 for the contact material requirements.
What the standard actually requires
IEC 60884-1 sets withdrawal force, not contact pressure. The withdrawal force limits are wide: for a 16 A socket, a plug must stay in under 30 N of pull but come out under 60 N. A socket at the bottom of that range can still be compliant and still relax faster than one at the top.
This is why the type test cannot catch long-term relaxation. The requirement is a snapshot, and relaxation is a trend.
The three mechanisms behind the drop
Mechanical wear removes material from the contact surface, about 0.5 to 2 µm per thousand insertions on a brass contact. Stress relaxation converts elastic strain into plastic strain, slowly, at a rate that roughly doubles every 15 K. Thermal cycling makes the housing expand and contract around the contact, adding a small working motion that accelerates both.
Run at 20 °C and the first two dominate. Run at 70 °C with current flowing and the third joins in, which is why a socket in a wall box behind a heater ages faster than the same socket in a bench test.
Design choices that buy margin
A thicker spring section, a shorter cantilever and a contact material with higher yield strength all buy relaxation margin. So does simply running the contact cooler: a 10 K drop in operating temperature roughly halves the relaxation rate.
Where the design is fixed, the cheapest intervention is the plug. A plug whose pins are 4.80 mm rather than 4.75 mm spreads the contact slightly more and gives the spring less travel to lose.
