Choose the compound from the medium, not from the datasheet headline: NBR handles mineral oil and grease to about +100 °C, FKM takes heat and aggressive chemicals to about +200 °C, EPDM owns water and steam but not oil, and PTFE covers the extreme edge where no elastomer survives. Confirm every figure against the compound data sheet.
The four questions that decide the material
Material selection follows the application, in a fixed order. Working through the four questions in sequence eliminates most candidate compounds before you reach a catalogue.
What is the medium. Mineral oil, water, steam, brake fluid, a solvent and a diluted acid each point at a different family, and the choice made here is usually final. NBR and EPDM are close to opposites: NBR resists oil and swells in water-based glycol fluids, EPDM resists water, steam and glycol and swells in mineral oil. Getting this question wrong cannot be fixed by changing hardness later.
What is the temperature. Temperature sets the ceiling on the compound and also decides which grades within a family are usable. A nitrile that is fine at +80 °C continuous loses its elasticity far faster at +100 °C, and the failure is not an immediate leak but a compression set that appears after some months in service.
What is the pressure. Pressure decides hardness and whether the seal needs mechanical support. Above a certain pressure and clearance combination, a soft compound is extruded into the gap and the seal is destroyed regardless of how chemically compatible it is.
Static or dynamic. A static face seal and a reciprocating rod seal are different problems. Dynamic service needs lower friction, tighter squeeze control and more attention to surface finish, and it rules out compounds that would be perfectly adequate in a static groove.
Two physical facts are worth adding to the list even though they are not questions. Surface finish and installation damage cause more premature O-ring failures in the field than chemical attack does, and a scratched groove or a seal nicked during fitting will leak with any material you choose.
The four materials compared
The table below gives the ranges that are broadly agreed in engineering practice. They are starting points for a shortlist, not a specification. Every compound has variants: a peroxide-cured EPDM, a low-temperature FKM or a high-hardness nitrile will sit outside the general figures, so the data sheet for the exact compound is the only authoritative source.
Two entries in that table deserve emphasis because they are the ones buyers get wrong most often. FKM is not a universal upgrade over NBR, since it is poor in hot water, steam, ammonia and brake fluid, which are exactly the media where EPDM is the better answer. PTFE is chemically the most resistant material in the list and the worst at being an O-ring, because it has no elastic memory of its own.
Hardness is a separate decision
Compound and hardness are chosen independently, and the catalogue reflects that: this site lists O-ring stock as O-Ring NBR 70, NBR 80, NBR 90, EPDM 70, FKM 75 and VMQ 70, where the number is the nominal Shore A hardness.
Shore A 70 is the default for general static sealing. Harder compounds in the 80 to 90 range resist extrusion better at higher pressure but need more squeeze and are less forgiving of a rough groove. Softer compounds around 60 suit low pressure, large tolerances and situations where the assembly force is limited. Hardness does not change chemical resistance, so raising it never fixes an incompatible medium.
Squeeze, groove and the four failure modes
An O-ring seals because it is deformed. The groove depth is designed to compress the cord by a percentage known as squeeze, and the whole seal depends on that deformation remaining available as force. Typical practice is roughly 20 to 30 percent for a static face seal, 15 to 25 percent for a static radial seal and 10 to 20 percent for dynamic service, with the groove width set slightly wider than the cord so the material has somewhere to move.
Four failure modes account for most seal failures in industrial service.
Extrusion. Pressure pushes the seal into the clearance gap between the metal parts. It shows as a chewed or flattened edge on the low-pressure side. Remedies are a harder compound, a smaller gap, a lower pressure, or a backup ring.
Compression set. The rubber takes a permanent deformation and stops pushing back. This is the classic slow leak that appears after a year at temperature, and it is the reason the temperature question matters more than the peak rating suggests.
Ozone and weather cracking. Fine surface cracks appear at right angles to the stretch direction. NBR is vulnerable outdoors and EPDM is essentially immune, which is why EPDM is the usual choice for exposed outdoor seals.
Chemical swell and hardening. The medium is absorbed, the seal grows, hardness drops and the groove fills. A small swell is sometimes designed for, but a large one means the compound is wrong.
What PTFE is not
PTFE appears in O-ring discussions and it is worth being precise about why. Polytetrafluoroethylene is a fluoropolymer with near-universal chemical resistance and the widest temperature range of anything in this table, but it does not behave elastically. It creeps under load and does not recover.
In practice PTFE is used as a backup ring behind an elastomer seal, as an anti-extrusion element at high pressure, or as the body of a spring-energized seal where a metal spring supplies the force that the material cannot. Asking for a PTFE O-ring as a drop-in replacement for a nitrile O-ring is a common mistake, and the result is normally a leak once the joint relaxes. Where a genuine elastomer has failed on chemical grounds, PTFE is the right direction, but it is a change of seal design rather than a change of material.
A working sequence
Establish the medium and its concentration first, then the continuous operating temperature rather than the peak, then the pressure and the clearance gap, then whether the seal is static or moving. Shortlist one or two compounds, confirm the specific grade against the data sheet, choose hardness from pressure and clearance, then check that the groove dimensions match the cord you intend to fit. If the application sits near a limit in more than one column, treat it as a design review rather than a purchasing decision.
Frequently asked questions
How do I choose between NBR and FKM?
Choose NBR when the medium is mineral oil, hydraulic oil or grease and the temperature stays below roughly +100 °C. Choose FKM when the temperature is higher, when the medium includes fuels, acids or aggressive solvents, or when ozone and weathering are present. Do not choose FKM simply because it costs more: it is poor in hot water, steam, ammonia and glycol brake fluid, where NBR or EPDM is the more reliable answer.
Can EPDM be used with oils?
No. EPDM has very poor resistance to mineral oils, hydrocarbon fuels and greases and will swell severely in them. Its strengths are water, steam, glycol-based brake fluid, phosphate esters, ozone and weathering, which is why it is the standard choice for outdoor seals, hot water service and automotive brake systems. If the medium contains both oil and water, neither NBR nor EPDM covers the whole duty and the specification has to be reconsidered rather than compromised.
What is the difference between PTFE and a rubber O-ring?
A rubber O-ring seals by elastic recovery: it is compressed and pushes back. PTFE does not recover, creeps under load and has to be energised by a spring or confined so that system pressure supplies the force. PTFE wins on chemical resistance and temperature range and loses on elasticity and creep, so it is normally a backup ring, an anti-extrusion element or a spring-energized seal body rather than a direct substitute.
Does a higher hardness mean a better seal?
Not on its own. Harder compounds resist extrusion at high pressure but need more squeeze and tolerate less groove roughness, so an unnecessarily hard seal can leak where a 70 Shore A seal would hold. Hardness also has no effect on chemical compatibility, so raising it never rescues a seal whose compound is wrong for the medium. Choose the compound first, then hardness from pressure and clearance.
What causes an O-ring to fail after a few months of good service?
Compression set is the usual answer, often accelerated by a temperature higher than the compound was chosen for. The rubber takes a permanent deformation, loses contact force and begins to weep rather than fail suddenly. Extrusion from excessive clearance and chemical swell from an incompatible medium produce similar slow failures, and all three are visible on the removed seal, which is why the failed part should be kept and examined rather than discarded.
Matching a compound to your application
If you are replacing a seal and are unsure which compound the original was, send the medium, the continuous temperature, the pressure and a photograph of the failed seal. We supply sealing technology from Angst+Pfister and can match the material family and hardness to the duty, then confirm the specific grade against the manufacturer's data sheet before quoting. Start from the Angst+Pfister brand page, browse the sealing technology range, or look through the full product catalogue and the brand index. Contact our team with the application details, and see the insights index for more selection guides, including how to verify that a replacement part is genuine.
Need technical documentation or a quotation for this model or solution?
Tell us the brand and model — our engineers will provide selection advice, technical documentation and a quotation.