A chair that cracks in the warehouse costs a replacement. A chair that cracks in an office three thousand kilometres away costs a container, a customer and a certification review — because by then the fault is in every unit made that month.
That asymmetry is the argument for seat and back fatigue testing. This article sets the two load types against each other, and explains why a single station bench is usually the right answer for design work.
Key takeaways
- One-off load and repeated load are different physical events with different pass criteria.
- Static strength is not a proxy for durability — the two rank chairs differently.
- Fatigue failure starts where stress concentrates: welds, screw threads, bracket corners.
- A single-station bench suits development; multi-station suits production sampling.
Two load types, two different physics
The standards pair them because products meet both in service: a heavy person sitting down hard, and the same person sitting down two thousand times a year.
A static load applied once asks whether the structure has enough strength. Failure is immediate: a crack, a permanent bend, a joint pulled out. The measurement is residual deformation and the judgement is whether the part still functions.
A cyclic loading programme asks something else entirely. Each individual cycle is well below the failure load — that is the point. Damage accumulates at stress raisers until a crack initiates, then propagates until the section separates.
Why a chair can pass one and fail the other
Fatigue is governed by local stress, while static strength is governed by the section as a whole. A bracket can be generously sized and still crack, because a weld toe or a thread root concentrates stress by a factor the overall geometry never sees.
This is why stress concentration is the practical obsession of fatigue testing. The failure you find is rarely where the designer was looking; it is at the junction nobody modelled.
| Aspect | One-off (static) load | Repeated (fatigue) load |
|---|---|---|
| What it asks | Is the structure strong enough | Will it survive a service life |
| Typical failure | Cracking, permanent bend, pull-out | Crack initiation then growth at a stress raiser |
| Pass criterion | No loss of function, residual deformation within limits | Survival to the specified cycle count |
| Time on the bench | Minutes | Hours to days |
| Typical weak point | Thin sections, poor welds | Weld toes, screw threads, bracket corners |
Load values, pad dimensions, cycle counts and acceptance criteria are set by GB/T 10357.3 and EN 1728, and differ between them. Confirm against the current edition you declare.
How one bench delivers both
The hardware difference between a static test and a fatigue test is smaller than it looks: an actuator, a loading pad at a defined position, and a way to hold or repeat the load. What changes is the profile and the duration.
For the seat, the pad is applied to the seat surface and loaded downward; for the backrest, the pad is applied to the back and loaded rearward at the specified height. Both directions are covered by reorienting the actuator and repositioning the pad, which is why one frame serves seat and back without a second machine.
The loading pad deserves more attention than it usually gets. Its size and stiffness define the contact area, and therefore the local pressure under it. Substituting a smaller or harder pad concentrates load and produces failures the standard never asked about.
Why single station is a design choice, not a compromise
During development the questions are diagnostic: which part failed, at what cycle, under which orientation. One station answered carefully tells you more than four stations running in parallel, because you can watch the failure happen.
Multi-station benches belong to a different job — sampling a production batch, where the question is statistical and throughput matters more than observation.
A single station also keeps the fixture simple and the datum stable. With one mounting, the back pad and seat pad are positioned relative to the same reference, so a failure in one does not force a re-setup before the other.
Defining failure before you start
Fatigue tests fail in messy ways, so the failure criterion has to be written down before the run: complete separation, a crack of visible length, loosening of a fastener, permanent set beyond a limit, or an audible change.
Two practical requirements follow. The bench must stop on loss of load or stroke, otherwise the counter keeps running against a part that has already given way. And the report must state the cycle count at failure, not merely that the sample survived or did not.
Three setup errors worth avoiding
- Cycle rate pushed for speed. Foam and polymer components heat internally at high frequency, softening the very structure being judged.
- Pad position estimated. Loading point height on the backrest changes the bending moment; the standard specifies it for exactly that reason.
- Dwell omitted. A saw-tooth profile with no hold time is not the same load history as one with a defined dwell, even at identical peak force.
Lituo has manufactured furniture testing instruments since 2008. Its seat and back fatigue testing machine is a single station design covering one-off and repeated loading of seat and backrest to GB/T 10357.3 and EN 1728 programmes, with adjustable loading pads, programmable cycle count and loss-of-load stop. View the furniture testing range →
Frequently asked questions
Is EN 1728 the same as GB/T 10357.3?
They cover comparable ground — strength and durability of seating — but the load values, pad sizes and cycle counts differ. Results are not directly interchangeable.
Can the machine test both seat and backrest in one run?
Separately, not simultaneously. Seat and back are loaded in different directions at different points, so the fixture is repositioned between programmes.
How long does a full durability programme take?
Usually days, because cycle counts are high and rate is limited by material behaviour. Plan capacity around the slowest programme, not the fastest.
Why buy single station instead of multi-station?
For development and failure analysis, observation matters more than throughput. Multi-station pays off when the question is batch consistency rather than root cause.
Post time: Nov-02-2026
