Compression Set in Elastomers: Testing, Graphs, Recovery and Seal Design

Compression set is the unrecovered thickness deformation of an elastomer after a defined compression exposure and recovery period. ASTM D395-18(2025) and ISO 815-1:2019 specify test procedures, but their method labels, specimen rules and recovery conditions are not interchangeable. For a constant-deflection test, the percentage describes residual thickness loss relative to the imposed compression, not total original thickness. Compare rubber compounds only when the method, specimen, compression, temperature, duration and recovery match. For O-rings and gaskets, the test supports material screening; it does not measure installed sealing force or prove leak resistance.

1. What Is Compression Set?

Compression set describes the deformation that remains after an elastomer is released from a specified compressive exposure and allowed to recover. A 15% result from a constant-deflection test means 15% of the applied thickness reduction was not recovered during the specified recovery period. It does not mean a seal has lost 15% of its installed contact force. This compression set definition concerns measured residual deformation, not a direct sealing-force measurement.

In a fixed-gap seal, the rubber cross-section is squeezed between the groove and mating surface. On unloading, part of the deformation recovers. Any thickness still missing after the specified recovery interval contributes to the measured compression set.

For rubber compression set, ASTM D395-18(2025) provides constant-force and constant-deflection test methods. ISO 815-1:2019 specifies constant-deflection testing for vulcanized or thermoplastic rubber at ambient or elevated temperatures. Both standards require defined testing and recovery conditions, but their procedures are not automatically interchangeable.

A reported 15% set from one supplier cannot be ranked against another supplier’s 25% without matching the test method, specimen preparation, compression, temperature, exposure duration and recovery procedure.

The thickness recovery test also does not measure the force retained by an installed seal. That requires separate stress relaxation or assembly testing.

The selection of elastomer material properties also depends on chemical exposure, hardness and part geometry, not one compression set number.

1.1 Permanent Deformation, Recovery and Compression Deflection

The three thickness quantities in a constant-deflection test are:

• Initial thickness: The specimen thickness before compression.
• Compressed thickness: The thickness maintained during a constant-deflection test.
• Recovered thickness: The thickness measured after unloading and the specified recovery period.

The difference between initial and compressed thickness is the imposed compression deflection. The difference between initial and recovered thickness is the residual deformation.

A simple example shows the distinction. A 12.50 mm rubber specimen compressed to 9.50 mm undergoes 3.00 mm of compression deflection. After unloading and recovery, suppose its thickness measures 12.00 mm. The remaining thickness loss is 0.50 mm.

The 0.50 mm thickness loss must still be divided by the denominator required by the specified test method; the equations follow in Section 2.

Compound formulation, cure history, compression duration and temperature can change recovery. Delayed elastic recovery and changes in the polymer network both contribute, so the percentage alone cannot identify a single mechanism.

Recovery time matters too. A specimen measured immediately after unloading may have a different thickness from the same specimen measured after 30 minutes. ASTM D395 specifies a 30-minute recovery for its methods, while ISO 815-1 defines recovery procedures that depend on the selected test conditions and method.

Compression deflection is the imposed thickness reduction. Compression set is the fraction that remains after the prescribed recovery.

2. How Is Compression Set Calculated?

In ASTM D395 Methods B and C, compression set calculation divides residual thickness loss by the imposed compression. A 15% result represents 15% of the imposed deflection, not 15% of the specimen’s full original thickness. ASTM Method A uses a different denominator.

2.1 ASTM D395 Method A vs Methods B/C: Different Denominators

ASTM D395 Method A holds a specified force: CA (%) = 100 × (t0 − tr) / t0. Methods B and C hold a specified deflection: CB/C (%) = 100 × (t0 − tr) / (t0 − ts). Here t0 is original thickness, tr is recovered thickness after the specified interval, and ts is the controlled compressed gap, all in mm. ISO 815-1 uses the constant-deflection equation but prescribes its own exposure and recovery procedures.

Using t0 = 12.50 mm, tr = 11.75 mm and ts = 9.50 mm gives 6.0% with Method A’s denominator and 25.0% with the B/C denominator. This comparison demonstrates the different equations; it does not describe one specimen validly tested under both methods.

2.2 Worked Example With Actual Thickness and Spacer Dimensions

An EPOC CRAFTER engineering case concerns a molded Parker E0740-75 EPDM O-ring for a static flange in a circulating hot-water unit. Its nominal cross-section is 3.50 mm, and the installed axial squeeze is 20.0%. Three separately molded compound disks were subjected to ASTM D395-18(2025) Method B at 70 ± 1°C for 22 hours, then measured after 30 minutes of unloaded recovery at 23 ± 2°C. The disk compression is approximately 25%, not the same as the assembled O-ring’s 20% squeeze.

Disk	Initial t0 (mm)	Gap ts (mm)	Recovered tr (mm)	Set (%)
1	12.50	9.38	12.05	14.42
2	12.48	9.36	12.01	15.06
3	12.52	9.39	12.04	15.34

For disk 1, (12.50 − 12.05) / (12.50 − 9.38) × 100 = 14.42%. Disks 2 and 3 give 15.06% and 15.34%. The project arithmetic mean is 14.94%; the ASTM triplicate reporting statistic is the median, 15.06%. All three individual readings are retained because the project acceptance limit applies per disk.

The three specimens meet the project limit of no more than 20.0% set per disk after 22 h at 70°C under Method B. This result supports the compound screening decision, not an assembled-flange leakage pass.

3. How Does Compression Set Testing Work?

A compression set test report needs enough detail to reproduce the percentage: compound and sample form, actual specimen thickness, loading or compressed gap, exposure, unloading, recovery and result statistic. The relevant material testing standards determine whether two datasets are comparable.

3.1 ASTM D395-18(2025): Methods A, B and C

Method	Constraint	Key distinction
A	Constant force	Specified force, with set relative to initial thickness; Type 1 specimen
B	Constant deflection	Spacer-controlled compression, usually near 25%; Type 1 or Type 2
C	Hardness-adjusted deflection	Compression selection depends on IRHD S2 hardness; report the actual gap and hardness

For Method A, the specified test force is 1.8 kN. For Method B, the standard Type 1 spacer is 9.5 ± 0.02 mm, and Type 2 spacer is 4.50 ± 0.01 mm. Method C prescribes nominal compressions of 25 ± 2% at IRHD S2 no higher than 80, 15 ± 2% at 81 to 89, and 10 ± 1% at 90 or above. Method B is not suitable above 90 IRHD. Do not convert an O-ring’s Shore A value directly into Method C’s IRHD S2 classification.

The EPDM case uses Method B on separately molded disks, nominally 29 mm in diameter and 12.5 mm thick. The 70°C exposure for 22 hours characterizes recovery under that test condition; it is not a continuous-use temperature rating or service-life claim.

3.2 ISO 815-1:2019: Temperature-Dependent Recovery Methods

ISO 815-1:2019 uses constant deflection but has three elevated-temperature recovery procedures. In Method A the compressed fixture leaves the oven and the specimen is released immediately, followed by 30 ± 3 minutes of recovery at laboratory temperature. In Method B the assembly cools while still compressed for 30 to 120 minutes before release, followed by 30 ± 3 minutes of laboratory recovery. In Method C the specimen is released and recovers for 30 ± 3 minutes at the exposure temperature before measurement. An ISO 815 compression set report also needs the precise release and recovery procedure.

Those A, B and C labels describe recovery procedures within ISO 815-1. They do not represent ASTM D395’s loading methods. ISO Type A disks are nominally 29.0 mm diameter and 12.5 mm thick; ISO Type B disks are nominally 13.0 mm diameter and 6.3 mm thick, whereas ASTM D395 Type 2 specifies a 6.0 mm thickness. Similar-looking specimen names are not proof of interchangeable test results.

3.3 Specimen Geometry, Compression, Conditioning and Reporting

Before comparing two reports, check the compound and cure route, molded or cut specimen, specimen size and count, hardness scale, initial thickness, actual deflection or force, plate and lubrication condition, exposure temperature and hours, and unloaded recovery conditions. ASTM D395 allows duplicate results to be averaged or triplicate results to be summarized by their median. ISO 815-1 specifies at least three pieces and reports the median to the nearest whole percent.

The manufacturing route matters. A molded disk and a disk cut from sheet can have different cure histories and surface conditions. Compare them only with the preparation method documented and consistent.

4. How to Read a Compression Set Graph

To read a compression set graph, check whether the horizontal axis represents temperature or exposure time, and whether the vertical axis is percent set under a named method. A rising curve indicates greater residual deformation at the plotted conditions. It does not establish increasing leakage in an assembled seal. Interpret the compression set temperature trend only after checking the exposure duration; a compression set time curve requires constant temperature.

4.1 Temperature, Exposure Time and Recovery Conditions

Read the test method first, then the specimen type, deflection, temperature, time and recovery protocol. For a time plot, check whether every curve was exposed at the same temperature. For a temperature plot, check whether exposure duration and post-release measurement timing were held constant. A comparison of 22 hours at 70°C with 72 hours at 100°C combines two changing inputs and cannot isolate either effect.

Higher temperatures and longer exposure can increase residual deformation, but the magnitude and trend depend on formulation and test history. Do not assign an invented generic curve to EPDM, silicone, FKM or TPU.

4.2 Comparing Curves Without Claiming Seal-Life Prediction

A compression set curve can screen formulations only when specimen size, deflection, loading method and recovery are aligned. It cannot supply retained seal contact force or a safe replacement interval. Media aging, thermal cycles and gland geometry require separate evaluation.

5. Compression Set vs Creep and Stress Relaxation

Property	What is observed	What it does not establish
Compression set	Residual thickness loss after prescribed unloading and recovery	Contact force or leakage rate
Creep	Deformation developing with time under maintained load or stress	Specified unloaded recovery percentage
Stress relaxation	Decline in force or stress while deformation is held	Compression set from thickness measurements

The distinction changes what an engineer should request. For a clamped gasket whose installed gap stays nearly fixed, stress relaxation testing may be needed to estimate how its clamping reaction changes. For a part under a sustained force, creep may control dimensional movement. For an elastomer tested after unloading, compression set measures how much deformation remains. These properties are related through material behavior, but one result cannot substitute for the others.

Plastic creep behavior concerns time-dependent deformation during continued loading, rather than thickness recovery after unloading.

6. What Does Compression Set Mean for O-Rings and Gaskets?

A standardized rubber disk and an installed O-ring have different contact boundaries. Disk compression set informs compound selection; the finished gland still requires verification at the specified temperature, water pressure and assembly geometry. An O-ring compression set or gasket compression set value therefore informs material screening, while installed seal compression set behavior depends on joint geometry.

In the EPDM case, a 3.50 mm circular cross-section compressed into a 2.80 mm gland depth gives 0.70 mm of axial squeeze, or 20.0%. The reported groove width is 4.60 mm, corresponding to an approximate cross-sectional fill of 74.7% from the circular area divided by the rectangular gland area. These dimensions describe the assembled geometry, not the standardized disk’s 25% test compression.

6.1 Seal Squeeze, Preload Loss and Leak Testing

The flange uses metal-to-metal limiting contact and bolts to maintain its installed geometry. Water pressure is specified not to exceed 0.30 MPa, over a stated operating temperature range of 20 to 80°C. The three tested disks meeting the internal 20.0% compression set criterion support the compound screening decision at 70°C for 22 hours. They do not quantify residual contact pressure or demonstrate zero leakage over the operating range.

Assembly validation must specify pressure, hold time and allowable leakage. The seal can also be affected by groove tolerances, flange deflection, installation damage and thermal cycling. Consumer electronics enclosures illustrate how unequal compression around a joint can defeat an otherwise suitable elastomer.

7. Choosing Low Compression Set Rubber and Elastomers

Low compression set rubber is useful when the test condition resembles the application. Check fluid compatibility and temperature first, then hardness, installed squeeze, production route and matched recovery test data. No universal percentage automatically approves every O-ring or gasket.

7.1 EPDM, Silicone, FKM and NBR: Compare Only Matched Test Conditions

EPDM is often evaluated for water service; silicone for applications requiring temperature flexibility; FKM for selected oils and chemicals; and NBR for many oil-contact parts. These are compatibility starting points, not a compression set ranking. Compare the actual compound, cure, hardness and test conditions. A silicone rubber compression set result does not establish performance for another silicone grade. EPDM compression set must be measured for the actual compound rather than inferred from the polymer family.

Rapid injection molding of TPU or TPE components introduces processing history that must be represented in the test pieces. Suppliers should identify the exact formulation, production route and test method rather than substitute a family-level material claim.

7.2 TPU, TPE and Additive-Manufactured Parts: Method Boundaries

For TPU compression set or TPE compression set, specimen preparation, process temperature and recovery method matter. Printed elastomers add build orientation, infill and internal voids; a cellular printed part is not equivalent to a solid molded rubber disk.

Additive manufacturing processes need tests on representative build architecture when the printed part is used as a seal. Urethane vacuum casting has its own cure and geometry variables. Neither manufacturing process implies a universal compression set percentage.