ASTM D395 Compression Set Testing: A Practical Guide for Seal Engineers

ASTM D395 Compression Set Testing: A Practical Guide for Seal Engineers

Published: 2026-01-05 | Reading time: 8 minutes

Why Compression Set Is the Seal Engineer's First Concern

Among the dozen or so properties routinely measured on rubber compounds, compression set (CS) occupies a unique position. Tensile strength can be thought of as the material's "nameplate" -- it tells you what the rubber is capable of in a single-use stress situation. Compression set, by contrast, is the material's "service record" -- it tells you whether the rubber will still be doing its job six months or six years from now.

For any static sealing application (O-rings, gaskets, weatherstrips, flange seals), compression set is arguably the single most predictive property of long-term sealing performance. The reason is physical: a seal works by maintaining contact stress against the mating surface. That contact stress comes from the elastic recovery force of the compressed rubber. Compression set quantifies how much of that recovery force the material permanently loses after being compressed at temperature.

A seal that loses 80% of its original compression (CS = 80%) has very little residual contact stress left. It will leak. Conversely, a compound that retains 90% of its original deflection (CS = 10%) can maintain a reliable seal through thermal cycles, pressure fluctuations, and years of service.

This guide covers the testing standards, condition selection, acceptance criteria, and compound-design principles that every engineer specifying rubber seals needs to understand.

The Standards: ASTM D395 and ISO 815

The two internationally recognized standards for rubber compression set testing are ASTM D395 (U.S. origin) and ISO 815 (international). Both standardize the same core measurement -- permanent deformation after compression at temperature -- but differ in methodological classifications and procedural details that can affect results.

The Three Test Methods

ASTM D395 defines two distinct methods. ISO 815-1 aligns closely with ASTM D395 Method B.

Why Method B Dominates Industrial Practice

Method B (constant deflection) is the workhorse of industrial seal qualification for four practical reasons that go beyond the standard text:

1. It mirrors real seal geometry. In a production seal, the compression percentage is determined by groove dimensions (gland depth relative to cross-section), not by an external applied force. A fixed 25% compression in the lab directly corresponds to a gland designed for 25% fill.

1. It eliminates a confounding variable. In Method A (constant force), the specimen continues to creep under load, meaning the actual compression percentage drifts over the test duration. This makes it unclear whether a poor result reflects inherent material set or simply increased strain. Method B locks the strain from the start.

1. It enables apples-to-apples material comparison. Comparing a 50 Shore A compound to an 80 Shore A compound at the same applied force yields different actual strain levels, confounding the comparison. Comparing both at 25% strain produces directly comparable CS values.

1. It maps directly to seal design parameters. The engineer specifying a 25% gland fill can read a 25% Method B CS value and know exactly what to expect in service, with no conversion or modeling required.

Selecting Test Conditions: Temperature and Time

The single most common question from engineers new to compression set testing is: "What temperature and for how long?" The answer is not a single number -- it depends on the material and the intended service environment.

Standard Temperature-Time Combinations

The table below presents the industry-standard condition sets, organized by material capability and application temperature range. These combinations represent decades of industrial consensus on what is both technically meaningful and logistically practical.

The fundamental rule for temperature selection: Always test at a temperature equal to or slightly above the material's maximum continuous service temperature. Testing at a lower temperature than actual service will give falsely optimistic CS values and miss degradation mechanisms that activate only above a thermal threshold. Testing at a substantially higher temperature risks activating entirely different degradation chemistry and produces results that do not predict service behavior.

Why 70 Hours?

The 70-hour duration appears throughout the standard condition sets and is not arbitrary. It represents a pragmatic balance between:

• • Sufficient time for the material to reach a quasi-equilibrium state of stress relaxation at temperature
• • Practical laboratory turnaround (a working week)
• • Reproducibility: shorter durations (22-24 h) show greater inter-lab variability

For materials with slow relaxation kinetics, extending the duration to 168 hours (one week) provides a more discriminating assessment. This is particularly relevant for FKM and FFKM compounds at their maximum rated temperatures.

Compression Percentage Selection

The standard compression percentages are:

• • 25% -- The default for most seal materials. Used universally for quality control.
• • 15% -- For high-hardness materials (Shore A > 80) or ebonite, where 25% compression approaches or exceeds the material's compressive yield point.
• • 50% (ASTM D395 specific) -- For cellular rubber, sponge, and soft foam materials.

A practical rule of thumb from seal engineering: if a compound passes at 25% compression (CS < 20%), it will perform even better at the lower compression percentages (15-20%) typically used in actual gland designs. The 25% test is conservative relative to most real-world installations.

Material-by-Material CS Performance Data

The following table provides representative compression set values for common rubber materials tested per Method B at 25% compression, using the standard temperature-time condition for that material class. These are industry-typical values, not specification limits for any single supplier.

The data highlights a pattern that every seal engineer should internalize: cure chemistry dramatically affects CS. Moving from sulfur to peroxide cure in NBR can cut CS from 30-50% down to 15-25% -- a 40-point improvement that makes the difference between a failing seal and a passing one, with no change to the base polymer.

Pass/Fail Criteria: Application Dictates Acceptance

There is no single universal pass/fail number for compression set. The acceptable CS value depends on the criticality of the seal, the operating environment, and the consequences of leakage. Applying an aerospace standard to a construction weatherstrip is wasteful; applying a weatherstrip standard to an O-ring in a hydraulic actuator is dangerous.

Acceptance Criteria by Application Class

Principles for Setting CS Criteria

1. Do not blindly pursue the lowest possible CS. Pushing CS below 10% may require trade-offs in tensile strength, tear resistance, or compression modulus that compromise other aspects of seal performance. A compound optimized for CS alone is rarely a good production compound.

1. Scale expectations to temperature. CS values naturally increase with test temperature. A CS of 30% at 200°C (for FKM) is not intrinsically worse than 15% at 100°C (for NBR) -- the materials are operating on different parts of the thermal spectrum.

1. Compare within material classes. Using FKM-grade CS expectations to evaluate NBR is a category error. Each elastomer family has its own CS baseline driven by polymer backbone chemistry.

1. Watch the delta, not just the absolute. The change in CS after aging (&#916;CS) is often a more sensitive indicator of material degradation than the aged CS value alone. A compound that starts at 10% CS and ages to 25% (+15 points) may be more concerning than one that starts at 20% and ages to 28% (+8 points), even though the absolute aged value is lower in the first case.

Compound Design: Four Knobs for Lower Compression Set

Achieving low compression set is primarily a compound formulation challenge, not a polymer selection challenge. The same base polymer can deliver CS values ranging from excellent to unacceptable depending on how the compound is designed.

1. Cure System: The Dominant Lever

The crosslink chemistry is the single most powerful factor controlling compression set. The underlying mechanism: CS arises from irreversible rearrangement of the crosslink network under sustained stress at elevated temperature. Crosslink bonds that are thermally labile will break, allow chain segments to slip, and re-form in new positions -- yielding permanent set. Crosslink bonds with high thermal stability resist this sequence.

The rule is straightforward: higher bond dissociation energy means better CS. The C-C crosslink (350 kJ/mol) from peroxide cure is more than twice as stable as the polysulfidic crosslink (150 kJ/mol) from conventional sulfur cure. This is not a marginal difference -- it is a fundamental chemical advantage that plays out in every test.

2. Filler System: Building a Resilient Network

Fillers serve a mechanical as well as a reinforcing function. High-structure reinforcing fillers create a secondary "filler network" within the rubber matrix that physically constrains molecular chain movement and reduces the irreversible slippage that contributes to CS.

• • High-structure carbon blacks (N330, N220): Provide a rigid framework that resists permanent deformation. The higher the structure, the more effective the filler network at reducing CS.
• • Precipitated silica + silane coupling agent: In EPDM compounds with peroxide cure, this combination delivers some of the lowest CS values achievable.
• • Avoid low-reinforcement fillers: Calcium carbonate, clay, and talc lack the surface activity and structure to form an effective filler network. Compounds relying on these fillers will naturally show higher CS.

3. Post-Curing: Finishing the Crosslink Network

Post-curing (also called two-stage curing or oven post-treatment) is standard practice for FKM and VMQ and increasingly common for peroxide-cured HNBR and NBR. The process involves heating the molded part in an air-circulating oven after demolding, typically at a temperature 10-20°C above the intended service temperature.

The mechanism is twofold: first, post-curing drives the crosslinking reaction further toward completion, reducing the number of "dangling" chain ends that contribute to CS. Second, it volatilizes low-molecular-weight species (unreacted curatives, decomposition byproducts) that would otherwise act as internal plasticizers and increase set.

For FKM in particular, skipping the post-cure is not an option for any seal application. The improvement in CS -- often a 50% reduction -- is too large to leave on the table.

4. Supporting Factors

Several additional formulation variables influence CS, though their effects are smaller than the cure system:

• • Antidegradant package: A robust antioxidant/antiozonant system preserves the crosslink network during service, indirectly limiting the long-term increase in CS.
• • Plasticizer selection: High-molecular-weight polyester plasticizers are less mobile and less prone to migration than low-MW phthalates (DOP/DBP), contributing to lower CS over time.
• • State of cure: Extending cure time 10-20% beyond T90 (the time to 90% of maximum torque) produces a more fully developed network with measurably lower CS. The extra few minutes per cycle are a low-cost improvement.

Common Testing Pitfalls

Even with a well-designed compound and correctly selected test conditions, procedural details in the lab can significantly influence CS results. The following are the most common sources of inter-laboratory variation and erroneous data.

Cooling Time Consistency

ISO 815 specifies that after removal from the compression fixture, specimens must be cooled in standard laboratory atmosphere for 30 +/- 3 minutes before thickness measurement. The material continues to recover during this period. A specimen measured after 15 minutes will show a higher CS (less recovery) than the same specimen measured after 60 minutes (more recovery). In inter-laboratory round-robin testing, inconsistent cooling time is consistently identified as the largest single source of variability. All technicians in a given lab must use the same timer for this step.

Fixture Parallelism

Compression set fixtures consist of two flat plates between which the specimen is compressed. If the plates are not parallel, the compression percentage varies across the specimen -- the thicker side is compressed more, the thinner side less. The standard requires parallelism within 0.01 mm across the plate surface. Fixtures should be checked periodically with a micrometer, especially after any impact or drop.

Lubricant Standardization

At elevated temperatures, certain compounds (notably FKM and VMQ) have a tendency to adhere to the metal compression plates, which alters the effective test conditions. A thin film of silicone oil or PTFE spray is permitted to prevent adhesion. However, the type and amount of lubricant must be standardized within a lab, because lubricants can interact with the rubber surface and affect the measured result. The best practice is to specify the lubricant in the test procedure and never change it without a documented equivalency study.

Specimen Placement in the Fixture

When multiple specimens share a single compression fixture, the specimens at the center of the fixture experience a slightly different thermal environment than those at the edges, particularly during heat-up and cool-down. For the highest-precision work, each specimen should have its own individual spacer ring, or specimens should be arranged symmetrically in multi-position fixtures. This is especially important for materials with high thermal expansion (silicone, fluorosilicone).

Inquiry & Technical Support

Nanjing Yuhang Rubber Co., Ltd. performs compression set testing per ASTM D395 Method B and ISO 815 on all seal-grade compounds as part of standard quality control. Our laboratory maintains calibrated compression fixtures, validated temperature uniformity, and documented cooling-time procedures to ensure reproducible and defensible CS data. For technical inquiries regarding compression set specifications, material recommendations for specific seal applications, or custom compound development: Products | Contact