Testing standard
ASTM D395
Standard Test Methods for Rubber Property—Compression Set
Written and technically reviewed by Dak System Inc. engineeringLast reviewed
ASTM D395 measures compression set — how much of a squeeze a rubber never gives back. A disc is held compressed in a bolted jig at temperature for a set period, released, allowed to recover for 30 minutes, and re-measured. Method B holds a constant deflection and is the common one; Method A holds a constant force. The result is the percentage of the original deflection that did not recover.
At a glance
- Test type
- Compression — the specimen is squeezed
- Published by
- ASTM
- Edition
- D395-18
- Material
- Rubber, elastomers & foams
- Runs on
- Series 7200 and Series 9000
What the test does
A moulded rubber disc is placed between two ground parallel steel plates. In Method B, hardened spacer bars set a fixed deflection — most often 25 % of the original thickness — and the plates are bolted down until they bear on the spacers. In Method A, a constant force is applied instead and the deflection settles wherever the compound allows. The clamped assembly goes into an air-circulating oven for a specified period, commonly 22 hours at 70 °C. The specimen is then released, allowed to recover for exactly 30 minutes at 23 ± 2 °C, and its thickness measured again at the same central point.
What it measures, and why it matters
The result is compression set: the percentage of the imposed compression that never came back. It is the single most useful screening property for anything that seals by being squeezed. A gasket, an O-ring or a bonded mount works because it is compressed and pushes back against the joint, and a compound that keeps a large share of its deformation has stopped pushing back with the force the designer assumed. Set therefore explains a great many joints that sealed perfectly on assembly and wept months later. It is also sensitive to cure state, which makes it a sharp production control — an under-cured batch sets badly long before anything shows in hardness or tensile strength.
Specimen and jig
The specimen is a moulded disc, and its thickness is the entire measurement — every other dimension only exists to keep the disc stable under the plates.
- Type 1 disc
- 29.0 ± 0.5 mm diameter × 12.5 ± 0.5 mm thickThe standard specimen for Method B.
- Type 2 disc
- 13.0 ± 0.5 mm diameter × 6.3 ± 0.3 mm thickUsed where material is scarce, and its results do not pool with Type 1.
- Plied specimens
- Permitted, from thinner sheetPlies are not bonded. A plied specimen sets differently from a moulded one because each interface can slip, so the build is reported.
- Deflection, Method B
- 25 % of original thickness — spacer bars 9.5 ± 0.02 mm for the Type 1 discSet by the spacer bars, not by the operator. Spacer height is what the result actually depends on, so it is the dimension worth checking.
- Plate condition
- Ground flat and parallel
- Measure thickness at the centre
- Same point, before and afterDakA disc that has barrelled is thicker at its rim than its middle, and measuring a different spot afterwards invents set that is not there.
Method A and Method B are different tests with the same name. One holds a force and lets the deflection find its own level; the other holds a deflection and lets the force decay. Their numbers are not interchangeable and the method letter belongs beside every figure.
Time and temperature
- Common default
- 22 h at 70 °CWidely specified, but it is a default and not the only combination — the material specification governs.
- Other combinations
- 24 h at 100 °C, 70 h at 23 °C, and othersHigher temperature and longer time both increase set, so a figure without its conditions means nothing.
- Recovery
- 30 min at 23 ± 2 °CRubber keeps recovering for hours. Fixing recovery at 30 minutes is what makes two laboratories comparable — measuring at 20 or at 60 gives a different answer from the same specimen.
- Release from the jig promptly
- Then start the recovery clockDakSpecimens left clamped while the oven cools carry extra set that belongs to the cooling, not to the test.
Calculations
CB = (t₀ − t₂) / (t₀ − tₛ) × 100
- t₀
- original thickness, mm
- t₂
- thickness after recovery, mm
- tₛ
- spacer thickness, mm
The denominator is the deflection actually imposed, not the original thickness. A set expressed against original thickness is a different and smaller number, and mixing the two is the commonest arithmetic error in this test.
CA = (t₀ − t₂) / t₀ × 100
- t₀
- original thickness, mm
- t₂
- thickness after recovery, mm
Here the denominator IS the original thickness, because no fixed deflection was imposed.
How the test runs
- 01Mould or cut discs to the Type 1 or Type 2 dimensions and let them rest after moulding.
- 02Measure and record the original thickness at the centre of each disc.
- 03Select spacer bars giving the specified deflection — 25 % for Method B.
- 04Place the discs between the plates, clear of one another and of the spacers.
- 05Bolt the plates down until they bear on the spacers along their whole length.
- 06Put the assembly into the air-circulating oven, already at temperature.
- 07Hold for the specified period, timing from when the oven returns to temperature.
- 08Remove the assembly and release the specimens immediately.
- 09Let them recover on a poor thermal conductor for 30 min at 23 ± 2 °C.
- 10Re-measure thickness at the same central point.
- 11Compute set against the imposed deflection for Method B, or original thickness for Method A.
Never stack discs to reach a thickness. A plied build is permitted and reported as such; an ad-hoc stack of whatever is to hand is not the same specimen and will read differently.
What the report has to contain
- Reference to ASTM D395 and the edition
- Method letter — A or B
- Specimen type, and whether moulded or plied
- Original thickness of each specimen
- Spacer thickness, and the percentage deflection it produced
- Temperature and duration under compression
- Recovery time and temperature
- Thickness after recovery
- Compression set per specimen, and the median
- Any specimen that showed splitting or surface degradation
What the machine must be capable of
Less than most rubber tests, and more than it appears. The compression is imposed by the spacers and held by the bolts, so a testing frame is not part of the measurement once the jig is clamped. What matters instead is the jig: plates ground flat and parallel, spacer bars of accurate and equal height, and bolts that can be pulled down evenly so every disc in the assembly sees the same deflection. The oven must hold its set point across the whole chamber, because a jig in a cold corner is running a different test. The thickness gauge has to resolve small differences reliably, since the result is a ratio of two small numbers.
What goes wrong in practice
Nearly all of the scatter in this test comes from three places. Plates that are not parallel, or bolts not pulled evenly onto the spacers, give different discs different deflections while the arithmetic assumes they all saw the spacer height. Measuring thickness at a different point after the test than before invents set that is not there, because a recovered disc is barrelled rather than flat. And the recovery period is treated as approximate when it is not — rubber recovers quickly at first and then slowly, so twenty minutes overstates the set and an hour understates it, from the same specimen. A fourth error is arithmetic: Method B divides by the imposed deflection, Method A by the original thickness, and using the wrong denominator produces a plausible number that is simply wrong.
ASTM D395 or ISO 815-1
| ASTM D395 | ISO 815-1 | |
|---|---|---|
| Standard specimen | Type 1: 29.0 × 12.5 mm disc | Type A: 13.0 × 6.3 mm; Type B: 29.0 × 12.5 mm |
| Common deflection | 25 % | 25 % |
| Recovery | 30 min | 30 min |
| Set expressed against | Imposed deflection (Method B) | Imposed deflection |
The two are close enough that results are often assumed to transfer, and close enough that the assumption usually survives. It is still an assumption: specimen type, temperature and duration all have to match before two figures can be compared.
Questions we are asked about this test
What is ASTM D395?
It is the ASTM test method for compression set in rubber — the proportion of an imposed compression that a compound fails to recover. A disc is held squeezed in a bolted jig at temperature for a fixed period, released, allowed 30 minutes to recover, and re-measured. Method B holds a constant deflection; Method A holds a constant force.
What is the difference between Method A and Method B?
Method B compresses the specimen by a fixed amount, usually 25 %, and lets the force decay however it will. Method A applies a fixed force and lets the deflection settle where it will. Method B is far more commonly specified. The two produce different numbers from the same compound, and they are also divided by different denominators, so the method letter has to travel with the result.
Why is 22 hours at 70 °C so common?
It is the default combination in the widely used material specifications rather than a property of the method. Both time and temperature increase set, so the pair is really part of the requirement. A compound quoted at 22 h/70 °C cannot be compared with one quoted at 24 h/100 °C.
Why does the recovery period have to be exactly 30 minutes?
Because rubber goes on recovering for many hours after release, quickly at first and then slowly. Any recovery period gives a defensible number; only a fixed one gives a comparable one. Measuring at 20 minutes overstates the set and at 60 minutes understates it, using the very same specimen.
Does compression set need a testing machine?
Not for the compression itself. The deflection is imposed by spacer bars and held by bolts, so once the jig is clamped the frame plays no further part. What the test does need is a jig with truly flat parallel plates, accurate spacers, an oven that holds temperature, and a thickness gauge good enough to resolve small differences.
What does a high compression set actually mean for a seal?
That the seal is losing the interference it was designed with. A gasket or O-ring seals because it is squeezed and pushes back; a compound that keeps a large part of its deformation is no longer pushing back as hard. High set is the usual explanation for a joint that sealed on assembly and weeps months later.
Why do my results scatter so much?
Most often the spacers or the plates. If the plates are not parallel, or the bolts are not pulled down evenly onto the spacers, different discs in the same jig see different deflections — and the calculation assumes they all saw the spacer value. Measuring the same point on the disc before and after removes another large share of the scatter.
Can compression set predict service life?
It ranks compounds, it does not predict a date. The test compresses at one temperature for one period, while a real seal sees a temperature history, a fluid, and often a changing squeeze. Set is a good screening and lot-release property and a poor substitute for testing under the conditions the part will actually meet.
Running ASTM D395 on the Series 7200 and Series 9000
Dak verifies against whichever standard the method names, and where a class applies our frames sit a class tighter than it asks.
| The method asks for | Dak supplies | |
|---|---|---|
| Capacity | Not a force measurement. Method B holds a fixed deflection in a bolted jig; Method A holds a constant force, commonly 1.80 kN on the Type 1 disc | Load cells from 1 kg to 60 ton on the Series 7200, and 0.5 to 100 kN on the Series 9000 |
| Force accuracy | ASTM E4 where a frame applies the initial compression | Verified to ASTM E4, and to ISO 7500-1 Class 0.5 |
| Gripping | Compression device: two ground parallel steel plates with hardened spacer bars, bolted to hold the deflection; an air-circulating oven holds the temperature | Our compression anvils, built to the specimen |
| Environment | Elevated temperature for the specified period — 70 °C for 22 h is the common default; recovery 30 min at 23 ± 2 °C before measuring | 3009 series chambers, −150 °C to +400 °C — temperature only |
This page describes the method as practised. The governing text is the current edition from the issuing body. Tell us what you are testing and we will answer with the machine, the fixture and a quotation.