A frequently asked question is “why perform Thermal Shock testing instead of / or in addition to Temperature Cycle testing?”
Temperature Cycle testing is used to stress product at temperature extremes, focusing on failure modes due to mismatch of thermal coefficients.
Thermal Shock testing is meant to accelerate failure modes due to very rapid changes of temperature (<10 seconds liquid-to-liquid). Examples of these conditions include hand or wave soldering, self-heating of power semiconductors or turn on of optical devices.
Liquid-to-liquid Thermal Shock testing uses heating/cooling by conduction, whereas Temperature Cycle testing is by convection, which is a much slower process than conduction. Other differences are the number of cycles and dwell times. Temperature Cycle tests are often performed for hundreds of cycles (Telcordia GR 468 requires 100 to 500 cycles) with minimum dwell times of 10 minutes. This is aimed at fatigue modes of failure. Thermal Shock is usually specified for 15 cycles and a minimum dwell of 2 minutes. MIL-STD-883, for example, specifies three temperature ranges for Thermal Shock: 0 to 100°C, -55 to 125°C, or –65 to 150°C. Telcordia GR 468 specifies 0 to 100°C, which can be performed in water. The two higher ranges require the use of a perfluorocarbon fluid with low freezing point and high boiling point. Silicon Cert uses Galden D02-TS, which has a useful range of –97 to 175°C, in both hot and cold tanks. The basket, holding the parts under test, switches from one tank to the other in eight seconds. It returns to the load/unload position automatically after the specified number of cycles.