The mechanism is thermal expansion, and the central fact is that the materials inside a connector expand at different rates. A PA9T housing, an FR4 insert, copper alloy contacts, solder, and the PCB substrate each follow their own thermal curve. Across a 125°C span, those small differences accumulate. The parts stay assembled, yet the interfaces between them change. On warm-up, a pin and the plastic holding it grow by different amounts; on cool-down, the same mismatch reverses. Each cycle is a small, repeatable displacement through the same set of interfaces. Repeated displacement is the wear path that matters because it acts on fit, normal force, and joint geometry rather than on visible outer shape. Self-heating from the power side makes the picture more concrete. A 40A supply generates heat where current crosses the contact interface, so the connector temperature is ambient plus internal rise. In a warm cabinet, a 40A load can push the contact area toward the top of the rated band while the room remains comfortably within it. On a cold start, the same parts sit near the bottom of the band. The material response follows the actual local temperature, so the cycle amplitude at the contact can be wider than the ambient swing suggests. That difference is why current load, enclosure airflow, and duty cycle belong in the same discussion as the -20°C to +105°C range. The useful design question is how often the direction of thermal movement flips at the hottest and coldest points of the assembly.