The solder joint didn't fail because it got hot. It failed because it got hot, then cold, then hot again — and the two materials bonded at that joint expanded and contracted at different rates every single time. After enough cycles, the accumulated mechanical fatigue cracked the joint. Not from temperature as a thermal stress. From temperature change as a mechanical stress. That distinction — between temperature as an environment and temperature cycling as a fatigue mechanism — is what temperature cycling testing is built around. Miss it, and the test profile you write will be optimised for the wrong failure mode.
The physics: CTE mismatch fatigue
Every material expands when heated and contracts when cooled. The rate at which it does so is its Coefficient of Thermal Expansion (CTE), measured in parts per million per degree Celsius. Copper: approximately 17 ppm/°C. FR-4 PCB laminate: 14–18 ppm/°C in-plane, 50–70 ppm/°C through the thickness. A ceramic MLCC capacitor: 7–10 ppm/°C. An aluminium enclosure: 23 ppm/°C.
When these materials are bonded together — as they are in every PCB assembly — and the temperature changes, they fight each other. The copper trace wants to expand more than the ceramic component. The solder joint between them absorbs the difference. Every thermal cycle accumulates a small increment of plastic strain in the solder. The Coffin-Manson relationship describes what happens next: cycles to failure decrease as plastic strain per cycle increases. A wider temperature range, a faster ramp rate, and a larger CTE mismatch all increase strain per cycle and reduce the number of cycles to failure. Temperature cycling testing exploits this relationship deliberately — applying enough cycles at a sufficient range to precipitate the failures that field service life would eventually produce, but in days rather than years.
What temperature cycling testing actually measures
Temperature cycling testing exposes a product to repeated transitions between a cold extreme and a hot extreme, under controlled conditions, for a defined number of cycles. The chamber ramps from cold to hot, dwells at each extreme until the product reaches thermal equilibrium, ramps back, and repeats. The test is not measuring heat damage — it is measuring fatigue accumulation. A product that survives 1,000 cycles between -40°C and +85°C has demonstrated resistance to a specific calculated amount of CTE mismatch fatigue. Whether that amount corresponds to the field service life depends on how accurately the test profile represents the product's actual thermal history in deployment.
The four parameters that define severity
Temperature range (T-low and T-high). The wider the range, the larger the thermal excursion per cycle, the more strain per cycle, and the faster the fatigue accumulates. For JEDEC JESD22-A104 Condition B: -55°C to +125°C. For automotive underhood to ISO 16750-4 Class VI: -40°C to +150°C. For consumer electronics: -40°C to +85°C. A component qualified at -40/+85 is not qualified for automotive underhood use. These ranges are not interchangeable.
Ramp rate. Faster ramps produce higher thermal gradients across the DUT during transition, which adds stress on top of the bulk CTE mismatch strain. Most standards specify 3–20°C/min. The achievable ramp rate under your DUT load is almost always lower than the spec sheet figure. How to calculate it — and which chamber decision it drives — is at Benchtop or Floor-Standing Environmental Chamber? The Decision Comes Down to One Number.
Dwell time. The time at each extreme must be sufficient for the DUT — not just the chamber air — to reach thermal equilibrium. Insufficient dwell means the product never reaches T-high or T-low, which makes the test less severe than specified. A 10-minute dwell may be adequate for a 50g PCB and completely insufficient for a 2 kg metal assembly. Verify with a thermocouple on the DUT during a trial run, not by trusting the chamber air sensor. The full dwell calculation is in Writing a Temperature Cycling Test Profile: The Parameters That Change Your Results.
Number of cycles. JEDEC JESD22-A104 requires 1,000 cycles for high-reliability qualification. AEC-Q100 Grade 1 specifies 1,000 cycles at -40°C to +125°C. The number is derived from field reliability models, not chosen arbitrarily.
Temperature cycling vs. thermal shock: the test that looks the same
They share a standard document — IEC 60068-2-14 — but they are different tests. Test Na (thermal shock) uses two pre-conditioned zones and transfers the DUT between them in under 30 seconds, targeting brittle fracture under instantaneous gradient stress. Test Nb (temperature cycling) uses a single zone that ramps at a controlled rate, targeting accumulated CTE mismatch fatigue. They are not interchangeable. A product that passes temperature cycling has not been tested for thermal shock failure modes, and vice versa. The full distinction — with the specific failure modes each test finds — is at Thermal Shock Testing: Why Slow Ramps Miss the Failures That Matter.
The five failure modes temperature cycling finds
Solder joint fatigue is the primary target: cracking at the interface between leadless ceramic components — MLCCs, BGAs, LGAs — and the PCB pad. PTH barrel cracking occurs in multi-layer PCBs where copper barrels connecting layers through the z-axis fatigue from the dramatic CTE mismatch between copper and FR-4 in the z-direction (17 ppm/°C vs. 50–70 ppm/°C). Wire bond heel cracking accumulates at the point where bond wire exits the bond pad. Package delamination occurs when CTE mismatch between layers in plastic-encapsulated components drives shear stress beyond adhesive strength. Conformal coating microcracking degrades moisture protection precisely where humidity damage is most likely to follow — connecting directly to the failure modes in Humidity Testing in Electronics.
The standards
JEDEC JESD22-A104 defines conditions A through J for semiconductor packages, from 0/+70°C (commercial plastic) to -55/+125°C (high reliability). AEC-Q100 references JESD22-A104 for automotive IC qualification. MIL-STD-883 Method 1010 covers military microelectronics. IEC 60068-2-14 Test Nb covers the international baseline. ISO 16750-4 Test Z/AD references IEC 60068-2-38 for automotive combined cycling. The broader standards map — how these fit together — is at IEC, MIL-STD, ASTM, ISO: The Environmental Testing Standards Map Every Engineer Needs.
The step most programmes skip
Every temperature cycling test produces a pass or fail. What most programmes don't do: destructive physical analysis on failed units — and on units that survived. Cross-sectioning solder joints, examining PTH barrels under SEM, and mapping crack initiation sites converts a compliance test into an engineering tool. A product that passes 1,000 cycles with no electrical failures but shows 40% crack propagation in the solder joints is different from one that passes with no crack initiation at all. The qualification report won't tell you which one you have unless you look.
Acceleration factors and field life equivalence
The value of a temperature cycling test depends on how accurately the test conditions represent the field conditions — and on the acceleration factor that relates cycles in the chamber to cycles in the field. The Norris-Landzberg model is the most widely used for solder joint fatigue: it relates the acceleration factor to the temperature range ratio, the maximum temperature ratio, and the cycling frequency ratio between test and field. A product cycling between -40°C and +85°C at 5°C/min in the chamber may experience 3–5 field thermal cycles per day in a vehicle in a continental climate. One thousand test cycles represents three to five years of field thermal history — but only if the model assumptions hold for the specific solder alloy, PCB substrate, and component geometry in the assembly.
Most qualification programmes use the acceleration factor to justify the test severity and cycle count. Fewer use it in reverse — to estimate what field life the test result predicts. A product that barely passes 1,000 cycles at -40/+85 has a different field life margin than one that passes with no detectable crack initiation after 1,000 cycles. Both pass. Both get the same certificate. The difference is only visible in the destructive physical analysis that most programmes don't run.
Post-test analysis
A temperature cycling test that ends at a pass result has told you that the product survived the specified number of cycles. A temperature cycling test with thorough post-test analysis has told you something more useful: where the fatigue accumulated, which joints are closest to failure, and what the margin between test completion and failure onset actually is. Cross-sectioning solder joints under optical microscopy reveals crack initiation and propagation. Scanning acoustic microscopy maps delamination in BGA packages without destruction. X-ray inspection identifies barrel cracking in PTH vias. These analyses are not part of most standard qualification programmes — they are engineering tools that answer the question the pass/fail result doesn't ask. The chamber types that run these tests are covered in Not All Environmental Test Chambers Are Equal — Here's How to Tell the Difference.
Connecting temperature cycling to the broader test programme
Temperature cycling does not stand alone in a complete qualification programme. It targets CTE mismatch fatigue — a specific failure mechanism. Thermal shock targets a different mechanism: brittle fracture under instantaneous gradient stress. Humidity testing targets moisture-driven degradation. Vibration testing targets mechanical fatigue from dynamic loads. HALT targets multiple mechanisms simultaneously at extreme stress levels to find design margin. Each test reveals failures that the others miss. A product that passes temperature cycling and humidity testing separately has not been tested for failures that require both stresses simultaneously — that requires combined environment testing, covered at Combined Environment Testing: The Only Way to Find Failures That Need Two Stresses to Appear. The complete test programme structure — which standards require which combination of tests — is at IEC, MIL-STD, ASTM, ISO: The Environmental Testing Standards Map Every Engineer Needs. The chamber selection decision that follows from knowing which tests the programme requires is at Not All Environmental Test Chambers Are Equal — Here's How to Tell the Difference.