Thermal cycling and humidity testing

What the two tests do

Thermal cycling moves a part between hot and cold extremes. The materials expand and contract at different rates, so joints, bonds, seals and plated interfaces see strain. Over enough cycles, a weak solder joint can crack, a press fit can loosen and a potting compound can separate.

Humidity testing holds a part in moist air, often at a raised temperature. Moisture enters through seals, pores and material interfaces. It can lower insulation resistance, corrode a plated surface or swell a moulded housing.

The tests are often run together, but they answer different questions. A part can pass a thousand thermal cycles and fail a week of damp heat because its seal is porous. A part can pass damp heat and fail thermal cycling because two materials have very different expansion rates.

The thermal cycle profile

A thermal cycle has four parts: a cold extreme, a hot extreme, a dwell at each and a ramp between them. Each one needs a value. The extremes should reflect the service environment, with margin agreed in the specification. The dwell must be long enough for the part to reach the chamber air temperature, which depends on mass and fixture.

The ramp rate is the part most often left out. A fast ramp creates a larger temperature difference across the part, so it drives more strain. A slow ramp is gentler. If the specification does not state a rate, the result is hard to compare with another laboratory.

For illustration, a profile might run from minus 40 to plus 125 degrees Celsius, with a 30 minute dwell and a ramp of 5 degrees per minute for 100 cycles. Those numbers are illustrative only. The correct values come from the product specification, the standard or the customer requirement.

Dwell time is often set from a thermocouple on the part, not from the chamber air. A heavy casting and a thin contact need very different dwells to reach the same internal temperature. If the specification gives a fixed dwell, check that the part actually stabilises within it.

Steady and cyclic humidity

Steady-state damp heat holds a constant temperature and relative humidity, often 85 degrees Celsius and 85 per cent relative humidity, for a defined number of hours. It is a soak test. It is good at finding moisture ingress, insulation weakness and material degradation.

Cyclic damp heat alternates between a humid hot phase and a cooler phase, often with a condensation period. The cycle drives moisture in and out, which is harsher on seals and coatings than a steady soak. It is closer to a real diurnal cycle or a repeated washdown.

The two tests are not interchangeable. A steady soak is easier to run and easier to compare. A cyclic test is better at exposing a breathing seal. State which one you mean, because a supplier may quote the easier one by default.

Choosing the test

The table maps the risk to the test that is most likely to reveal it. It is a starting point for a test plan, not a substitute for the product specification.

What fails, and where

Thermal cycling failures usually start at an interface. A solder joint between materials with different expansion rates accumulates strain. A bonded insert can debond. A crimped joint can relax. A plated through-hole can crack at the barrel.

Humidity failures usually start at a seal or a surface. Moisture crosses a gasket, a potting boundary or a porous coating. Corrosion appears at a cut edge or a scratch. Insulation resistance falls before any visible damage appears.

That is why the inspection must look at the interfaces, not just the outside. A visual check of a sealed unit can pass while the inside is wet. A continuity check can pass while a joint is almost cracked. The test plan should name the checks that reveal the expected failure mode.

Damage often starts before the final check can see it. A seal may weep slightly for many cycles before insulation resistance falls. A joint may develop a hairline crack before resistance changes. Interval checks during the test can show the progression, which is more useful than a single pass or fail at the end.

Acceptance criteria and post-test checks

Acceptance should be written before the test starts. It might be no visible corrosion, insulation resistance above a stated value, contact resistance within a stated change, or no crack in a cross-section. A pass or fail decided after the parts come out is not a test.

Post-test inspection should include a functional check at the specified load or voltage, a dimensional check where thermal expansion matters, and a visual check for cracks, discolouration, seal damage and corrosion. For critical joints, add a cross-section or a dye penetrant check.

Record the chamber identity, the calibration date, the profile actually run and any interruption. A power cut or a door opening can change the result. The report should show the measured profile, not just the set point.

Samples, fixtures and sequencing

The sample size should reflect the risk and the cost. A small pre-production batch may be enough to find a design fault. A qualification programme needs enough units to cover the production spread and the critical interfaces.

The fixture matters. A part clamped to a heavy block heats and cools more slowly than a free part. A part that is powered during the test adds its own heat. State how the part is mounted, whether it is powered and what load it carries.

Sequence the tests to reflect the real life. A product may see thermal cycling, then humidity, then vibration. Running the sequence in the wrong order can hide an interaction or create one that never happens in service.

Keep a control sample at room conditions. If both the tested and the control units fail, the cause may be the build rather than the environment. If only the tested units fail, the profile is doing its job. The control is cheap and often forgotten.

Reading the result

A pass is not proof that the design is immune to the environment. It is evidence that the tested units survived the stated profile. The report should say what was tested, how many units, under which profile and with which checks.

A fail is more useful than a pass if the failure is analysed. Look at where the crack started, which interface absorbed the strain and whether the failure is consistent across samples. That tells you whether the fix belongs in the design, the process or the material.

Keep the failed parts. A cross-section of a failed joint is often the most valuable record in the whole programme, because it shows the mechanism rather than the symptom.

Compare the result with the design intent. If the part was expected to survive a certain number of cycles and failed earlier, the gap is the useful finding. Record it, and use it to set the next test rather than treating the report as a pass or fail only.