A TPE or TPV seal leaks when it no longer pushes back hard enough on the surfaces it seals. Usually it has taken a compression set: squeezed for long enough at a high enough temperature, it stays partly flattened after the load comes off. Sometimes the material family is wrong for the temperature, sometimes the groove squeezes the seal too much or too little, and sometimes a fluid has swollen the seal or pulled its oil out. Check the operating conditions and the design first, then the molding, then the material.
Quick diagnosis
| What you see | Likely cause | Check first |
|---|---|---|
| The seal is flattened and stays flat after removal. It leaked after time at temperature. | More set than the material can take at that temperature, often a styrenic TPE used above its range. | Peak temperature at the seal; compression set data at that temperature |
| It leaks from the first day. | Too little squeeze (deep groove, undersize seal, grade too hard), or flash, a parting line or a gate mark on the sealing face. | Measure squeeze at worst-case tolerances; inspect the sealing face |
| The seal is cut, cracked or pushed out into the gap. | Too much squeeze or an overfilled groove, made worse by swelling or thermal expansion. | Groove fill at the hottest condition |
| It leaks in cold weather and seals again when warm. | The seal stiffens and shrinks at low temperature. | Low-temperature data on the grade data sheet |
| The seal is swollen and soft, or shrunken and hard, after contact with a fluid. | Oils and fuels swell EPDM-based TPV and styrenic TPE. Some fluids extract oil from the compound and shrink it. | Volume and hardness change after soaking in the actual fluid |
| Every part leaks in the same spot. | A weld line, gate vestige, sink or short fill at the sealing lip. | Weld line and gate location; a section through the seal at the leak |
| A new lot or a higher regrind ratio leaks more. | A hardness shift, extra heat history, or a different grade. | COA hardness for both lots; a run on virgin resin only |
Checks in order, cheapest first
- Measure a used seal. Measure the thickness of a new seal, a used one and the compressed height in the groove. Set is roughly (original − used) ÷ (original − compressed) × 100. High set confirms the material is relaxing. Low set points to the design or the surfaces.
- Confirm the real temperature and time. Find the peak temperature at the seal, not the room it sits in, and how long it stays there. A seal next to a motor or in a hot-fill line runs far hotter than ambient.
- Calculate squeeze and groove fill at the extremes. Use worst-case tolerances on the seal and the groove. Check squeeze at the coldest condition and groove fill at the hottest, with room for swelling.
- Inspect the sealing surface. Look for flash, parting lines, gate vestige, weld lines and sink on the face that seals. Moving a gate or a parting line off the sealing face can fix a leak that no material change will. See sink marks.
- Check the molding. Make sure seals are fully packed with no voids, dry TPV before molding, and hold regrind at a level you have tested. Check hardness against the specification.
- Soak seals in the real fluid. Immerse seals in the fluid they will see, at service temperature, and measure volume and hardness before and after. Swelling or shrinkage of more than a few percent changes the squeeze.
- Compare compression set data on the same basis. Same method, time, temperature, deflection and specimen type. A room-temperature value says little about a seal that runs hot.
- Change the material last. If the design, the surfaces and the molding are right and the seal still takes a set, move to a family or grade with lower set at your temperature.
How compression set is tested
Compression set is the share of the applied squeeze that does not recover, as a percentage: 0% is full recovery, 100% is none. A number means little without its test conditions. Common conditions:
| Test condition | Common choice | What to watch |
|---|---|---|
| Method | ASTM D395 Method B (constant deflection); ISO 815-1 (ambient and elevated temperature); ISO 815-2 (low temperature) | Similar approach, but compare like with like. |
| Deflection | 25% of the specimen thickness in ASTM D395 Method B; ISO 815-1 uses 25% for most rubbers but less for very hard ones (about 80 IRHD and up) | Your seal may run at a different squeeze, and a hard grade tested at a lower strain is not directly comparable. |
| Time | 22 h or 70 h in ASTM D395; ISO 815-1 commonly uses 24 h, with much longer tests for sealing work | Set keeps rising with time, so a 22 h value flatters a seal that lasts years. |
| Temperature | 23°C / 73°F, 70°C / 158°F, 100°C / 212°F, 125°C / 257°F | Ask for data at or above your peak temperature. |
| Recovery | 30 minutes at room temperature before measuring | Measured sooner, set reads higher. |
| Specimen | A molded button, or discs cut from a plaque and stacked | Molded and stacked specimens can differ. Specimens cut from your parts include your molding orientation. |
For solid seals, common O-ring practice is a static squeeze of roughly 15–30%, with the groove no more than about 85% full at the hottest condition so swelling and expansion have room. Hollow and bulb profiles are designed differently. Treat these as starting points and confirm against your seal design guide.
Why seals take a set
- Elastomers relax under constant squeeze. Under load the polymer chains slowly rearrange, and the seal pushes back less over time. Heat speeds this up sharply.
- In TPV, crosslinked rubber does the springing back. TPV is fine particles of crosslinked EPDM in polypropylene. The rubber recovers; the PP around it can flow a little under load, more as it gets hot. That is why TPV generally holds a set much better than styrenic TPE at temperature, while thermoset EPDM, crosslinked all the way through, is generally still better at high temperature. See TPV vs. EPDM.
- In styrenic TPE, the crosslinks soften with heat. The polystyrene domains that hold SEBS together are physical, not chemical, and they soften as the temperature rises. Add the PP and oil in the compound, and set climbs quickly at elevated temperature.
- Hardness changes the balance. Within one TPV family, harder grades usually contain more PP and, on most data sheets, take more set at the same test conditions. Softer grades need more squeeze for the same sealing force.
- Fluids change the volume. Oils and fuels swell EPDM-based TPV and styrenic TPE, which can overfill the groove. Fluids that pull oil out of the compound shrink the seal and lower its sealing force.
- Molding leaves orientation behind. Soft TPV shrinks about 4.1% along the flow at 49A and about 0.7% across it. That frozen-in orientation can relax in service and change the seal’s shape unevenly.
Material changes, and what each one costs you
| Change | Why it helps | What it costs you |
|---|---|---|
| TPV in place of styrenic TPE | Lower compression set at elevated temperature and better oil resistance. | Usually a higher price, drying before molding (80°C / 176°F, 3 h), fewer very soft grades, and good bonding mainly to PP when overmolded. |
| A TPV sealing grade with lower set at your temperature | Better recovery in the same process and tool. | A higher price, a narrower hardness range, and a new grade for your customer to approve. |
| Thermoset EPDM rubber in place of TPV | Generally the lowest set at high temperature among these options. | Not a pellet you run on your press: it needs a rubber molding process with a cure step, cured scrap cannot be remelted and molded again, and cycles are usually longer. |
| A softer grade with a redesigned groove | Lower set in many TPV families and a lower closing force. | More squeeze needed, more risk of extrusion into gaps, and steel changes to the groove. |
| An oil-resistant elastomer (NBR/PP TPV, TPC or TPU) where oils are present | Less swelling in oils and fuels. | A higher price. TPU and TPC are usually harder, and their set at temperature differs from TPV, so ask for data at your conditions. See TPV vs. other elastomers. |
Ask for compression set at your conditions. A room-temperature value says little about a seal that runs at 100°C / 212°F.
Often confused with
- Stress relaxation. The seal keeps its shape but presses less. Compression set measures shape; compression stress relaxation (ISO 3384) measures the drop in counterforce directly and is often the better test for critical seals.
- Swelling in a fluid. A swollen seal can leak or extrude for a different reason. See chemical-resistant plastics.
- Sink or short fill on the sealing lip. A molding defect that leaks from day one. See sink marks and short shots.
- Softening under heat. A seal that deforms while hot rather than slowly over time is above its heat limit. See parts soften under heat.
Quick answers
What is compression set?
It is the part of a squeeze that a rubbery material does not recover after the load is removed, as a percentage. 0% means it springs fully back and 100% means it stays flattened. It is measured under set conditions, usually 25% squeeze for a fixed time at a fixed temperature.
Is TPV as good as EPDM rubber for seals?
At room temperature and moderate heat, TPV comes close and molds like a plastic. At high temperature, thermoset EPDM generally holds a lower set. Compare data at your temperature and time. See TPV vs. EPDM.
What compression set should I specify?
Specify it with its conditions: method, time, temperature and deflection, for example 22 or 70 hours at your peak service temperature at 25% squeeze. Base the limit on parts that seal in the field, then add a margin.
Why does my gasket leak after a few weeks?
Usually it has relaxed: the squeeze, held at temperature, became partly permanent and the seal pushes back less. Check the peak temperature at the seal and the groove fill, then compare the material’s compression set at that temperature.
Send us a photo
A polymer chemist will look at it and reply by email. There is no charge and no obligation.