Key takeaways
Transfer time is not a severity dial; it is part of how the method is defined. Temperature cycling answers whether a part degrades with repeated use, thermal shock answers whether it cracks under sudden change, and neither report can stand in for the other. An outline missing any of the four columns - item name, profile type, dwell criterion and verdict items - leaves the resulting data impossible to review.
A commission that says simply "high and low temperature" often turns out to mean two different things: one party expects temperature cycling, the other thermal shock. They are not a fast and a slow version of the same test. The temperature change method, the chamber arrangement and the scope for in-situ monitoring all differ, and so do the failure mechanisms each one exposes - cycling accumulates damage through repeated expansion and contraction, while shock pulls interfaces apart under a sudden gradient. Transfer time is not a severity dial; it is part of how the method is defined. Below are three common misreadings, the profile differences, the failure mapping and the four columns every outline needs.
1. Three misreadings to remove
Misreading one: only the ramp rate differs. Rate changes how stress is applied, not how severe a setting is. A slow programmed ramp gives the interior of the sample time to follow the chamber, with heat conducting gradually from the surface inward. A rapid transfer opens a short-lived temperature difference between surface and core, and between dissimilar materials, so that the gradient itself becomes the load. With identical hot and cold endpoints, the two profiles do not produce the same set of readings.
Misreading two: shock is harsher, so it covers cycling. Failures that need cycle count and dwell time at the hot end to develop - solder joint fatigue, compression set in seals, creep in potting compounds - simply do not have time to grow in a short shock profile. Conversely, delamination that opens only under an instantaneous gradient may never appear during a slow ramp. How climate items are ordered and combined is covered in scheduling climate loads; this article deals only with the division between two temperature items.
Misreading three: the item name on the report is flexible. It has to match the relevant part and clause of GB/T 28046, the corresponding part of ISO 16750, and the number and revision of the technical agreement. Different methods mean the reports are not interchangeable, and this column is what an OEM review checks first. The adoption relationship between the two standard families is covered in a separate article.
One more frequent mix-up: surface condensation and falling insulation resistance are often blamed on the temperature item, when moisture has entered through a humidity path. That belongs with the damp heat work and the cyclic versus steady-state damp heat comparison, not folded into a temperature conclusion.
2. Where the two profiles differ
No endpoint temperatures, cycle counts or transfer times are given here - only the differences in action and in how conditions are defined. All conditions follow the standard clauses, the OEM specification and the technical agreement.
| Dimension | Temperature cycling | Thermal shock |
|---|---|---|
| --- | --- | --- |
| Temperature change | Chamber air or medium follows a programmed ramp; the sample stays in one workspace | The sample transfers between hot and cold zones, or airflow reverses rapidly inside a single chamber |
| Role of dwell | Dwell at each end lets critical locations reach the specified thermal state; damage accumulates with cycle count | A brief dwell after transfer establishes the state; most of the load arrives in the transfer window |
| In-situ monitoring | Feedthrough is straightforward; the part can be powered and read during application | Moving mechanisms and strong airflow constrain feedthrough, so verdicts are usually post-test and must be agreed in advance |
| Dominant stress | Fatigue and creep from repeated expansion and contraction | Interface mismatch and brittle fracture from an instantaneous gradient |
| Condition source | Endpoints, ramp, dwell and cycle count per standard clauses and the technical agreement | Endpoints, transfer arrangement and its time definition, and cycle count per standard clauses and the technical agreement |
The third row is routinely underestimated. Whether readings can be taken during application determines how much weight the statement "function normal after test" can carry. In cycling, in-situ monitoring can show that no intermittent open occurred while stress was applied. In shock, that evidence may not be available, so the report should state the monitoring boundary plainly rather than assume the two are equivalent. A third arrangement exists for combined stresses: three-integrated testing covers the interaction effects when temperature, humidity and vibration act together, outside the scope of this comparison.
3. Failure mapping decides which one to choose
Selection should start from the failure family being hunted, not from which profile sounds harsher. The table below is a directional summary, not a standard requirement; criteria and thresholds come from the standard text, the OEM specification and the technical agreement.
| Failure type | Profile that mainly exposes it | Mechanism and where the verdict lands |
|---|---|---|
| --- | --- | --- |
| Solder joint fatigue, intermetallic growth | Temperature cycling | Driven jointly by cycle count and hot-end dwell; read resistance drift and cracks after sectioning |
| Compression set in seals, creep in potting | Temperature cycling | Time-accumulating, often paired with an IP rating re-check |
| Delamination at dissimilar-material interfaces, brittle substrate cracking | Thermal shock | Mismatch in expansion coefficients pulls the interface open under a sudden gradient; document appearance and acoustic microscopy first |
| Cracking in plating, coatings and cover glass | Thermal shock | Large surface-to-bulk temperature difference; crack orientation reflects the loading direction |
| Connector contact resistance drift, terminal fretting | Both, with different reading points | Cycling shows long-term drift, shock shows the step change after an interface opens; test points belong in the outline |
Rows two and three are the watershed: one family needs time, the other needs a gradient. The thermal shock article covers how temperature range, cycle count and transfer time are set - that is the conditions side. This article explains only why the two cannot be merged on mechanism. Other climate items load the sample differently again; the solar radiation piece and the environmental reliability checklist are useful when trimming a programme, and there is no need to push every item into one outline.
4. Four columns every test outline needs
Column one: item name and revision. Clause number, standard revision year and technical agreement number, one line each. Writing only "high and low temperature" hands the definition of the method to whoever opens the chamber.
Column two: profile type and transfer arrangement. Programmed ramp in a single chamber, or moving two-zone, three-zone switching, or single-chamber airflow reversal; when the transfer interval starts, whether the reference is the sample surface or a critical internal location, and where the sensor sits. This column is the precondition for comparing results across laboratories. Separate families of standards exist for change-of-temperature methods; their scope, adoption relationship and current revision should be taken from the standard text rather than recalled from habit.
Column three: dwell criterion. Whether the profile advances once the sample reaches thermal stability, or runs to a fixed duration. For a sample with large thermal mass the two differ considerably. Criterion, sensor location and whether the part may be powered all need to be stated.
Column four: verdict items and monitoring boundary. What is read during the test and after it - appearance, contact resistance, insulation resistance, sealing, functional status class - how to phrase the result when in-situ monitoring is not possible, plus the evidence-preservation order after a failure and whether stress continues. Leave this column blank and no later data can be reviewed.
With all four aligned, an outline usually clears review in one pass. Miss any one and it comes back at report review, by which time the sample has normally been through a complete round of stress.
A few plain words
When the choice is unclear, answer one question first: is the concern slow degradation with use, or cracking at the moment of sudden temperature change. The first belongs to temperature cycling, the second to thermal shock. If both matter, agree the sequence, sample grouping and intermediate readings in advance rather than adding them after the report exists.
When TMC LAB takes on this kind of commission, the item name and revision, profile type and transfer arrangement, dwell criterion, and verdict items with their monitoring boundary are written into the outline first, with any gap in in-situ monitoring marked separately before testing is scheduled. Endpoints, dwell, cycle count and transfer time follow the standard clauses, the OEM specification and the technical agreement; no values are preset at the outline stage.
Bringing the previous report, the page reference of the relevant clause in the technical agreement and a note on whether the sample may be powered saves far more time than a single line saying "run high and low temperature". Scheduling is arranged per project; you can state your expected delivery date when submitting samples, and we will confirm a feasible plan against the current schedule and report progress at each milestone.
