Thermal shock is easily treated as a fill-in-the-blanks exercise: enter a temperature range, enter a cycle count, wait for the report. But take the same 200-cycle plan and cut the transfer time from 15 minutes to 3 and the failure count can multiply several times over. Thermal shock never tested how many times a part saw high and low temperature; it tests how abruptly the temperature changed.
1. Shock and cycling are not the same thing
Temperature cycling varies temperature linearly in a single chamber and stresses the ramp rate and dwell time; thermal shock transfers rapidly between two temperature zones and stresses instantaneous thermal stress. Tests Na and Nb in GB/T 2423.22 separate these two approaches. The methods themselves are public knowledge — the problem is choosing the wrong one.
In engineering terms the difference is direct: temperature cycling more readily exposes the cumulative process of solder joint thermal fatigue, while thermal shock more readily tears material interfaces apart in a short time — potting compound debonding from the housing, permanent deformation of seals, blistering of platings. Most of these show up first under shock.
2. Mounting position sets the temperature range
| Mounting position | Common band | Note |
|---|---|---|
| Cabin | Low -40 °C, high +85 °C | Solar radiation adds on top; cabin temperature in summer can exceed ambient |
| Engine compartment | Low -40 °C, high +125 °C | Parts near heat sources go higher again, per the agreement |
| Exposed body area | Low -40 °C, high +85 to +105 °C | Subject to both radiation and splash; often run in series with spray and salt spray |
The high-temperature end is not set by "the part that feels hottest" but by the mounting-position temperature profile the OEM supplies (the temperature zone map). The same controller under a seat and in the engine bay can differ by an entire class — which is why the "mounting position" field on the request form must never be left blank.
3. Cycle count, transfer time, dwell time: none of the three can be vague
Common cycle counts run from 100 to 300, derived from design life and usage frequency, with the technical agreement as the final authority.
The shorter the transfer time, the more severe the test. A two-chamber basket system transfers in minutes; a three-chamber system with an ambient transition zone ramps more gently. The key is not what the equipment is rated for but whether the report states sample surface temperature or chamber air temperature — the air may have reached temperature while the core of the sample has not.
The same applies to dwell time: a heavy metal housing takes far longer to reach temperature than a small module. With insufficient dwell the sample never sees the full thermal stress amplitude, and extra cycles achieve nothing. The way to check is plain: attach a thermocouple to a critical location on the sample and start timing only once the curve has stabilised.
4. What the failures look like, and how to read them
Typical thermal shock failures leave traces:
- Solder joint thermal fatigue cracks: initiate at the stress concentration in the joint and propagate along the interface, most typically at BGA corners and leadless packages;
- Interfacial debonding: between potting compound and housing, or thermal pad and heat spreader — only visible by scanning acoustic microscopy or cross-section;
- Connector contact degradation: fretting wear combined with temperature change, contact resistance drifts slowly and a functional test may not flag it;
- Seal failure: after permanent compression set of the seal, a subsequent IP test lets water in — which is why the order within a climatic package matters.
When reading the result, first check whether failures cluster at the interface with the largest CTE (coefficient of thermal expansion) mismatch. Clustered means a design and material matching problem; scattered and random means go back and check whether the test conditions match the agreement. For the equivalent reasoning on structural reliability, choosing the severity level for ISO 16750-3 vibration and shock testing gives criteria you can compare against.
5. A few honest words
Thermal shock delivers the most value during development: one short exploratory run can surface structural problems such as package selection, potting process and connector fixing early. Finding them after volume production means changing tooling and process — a completely different order of cost.
When TMC LAB schedules this kind of test, the mounting position, sample state and transfer method are checked first, then the cycle count and dwell time are confirmed; if the conditions do not line up, the request goes back for confirmation rather than being run on defaults.
For the climatic load clauses defined in GB/T 28046.4 and ISO 16750-4, start with the GB/T 28046 standard page and the ISO 16750 standard page; for combined conditions in the damp heat direction, see cyclic versus constant damp heat and plan them into the same test outline.
Turnaround time is arranged according to the actual project. You are welcome to state your target date when submitting samples; we will check it against the current schedule, confirm a feasible plan with you, and update you at each milestone.
