Silicone Oil Resistance: Where It Works and Where It Swells

Why silicone swells in oil

Silicone rubber has a very low cohesive energy density, which is what gives it its wide temperature range and its flexibility. The same property means that non polar fluids such as mineral oil are chemically similar to it, and the polymer absorbs them until the chemical potentials balance.

The absorption is physical, not chemical. The oil enters the network and pushes the chains apart, so the part swells and its strength falls. It does not degrade the polymer, and if the part is removed from the oil it will shrink back, but in service the swelling is permanent because the oil is always there.

The consequences for a seal are direct. A swollen O ring no longer fits its groove, and it extrudes out of the gap under pressure. A swollen gasket grows in thickness and its compression increases, which raises the compression set and shortens its life. Neither failure looks like a material failure; both look like a design or assembly fault.

The fluid decides everything. Silicone in silicone oil swells by a few percent and is perfectly serviceable. Silicone in mineral oil swells by 30 to 40 percent and is not. The same compound, the same part, two completely different answers.

Volume swell against time and fluid

The chart on the left plots volume swell against immersion time at 100 C for three fluids. In mineral oil the swell reaches 30 percent at 168 hours and 42 at 672. In a synthetic ester it reaches 20 and 28. In silicone oil it stays between 3 and 7 percent.

The shape of the curves is a diffusion process. The initial rate is high because the concentration gradient is steep, and it falls as the material approaches equilibrium. Most of the swell happens in the first 168 hours, which is why a 72 hour test is a reasonable screening test even though it is not the equilibrium value.

The 10 percent line on the chart is the usual limit for a seal. A compound that swells more than that will not stay in its groove, and above about 15 percent the swelling is enough to change the sealing force and the assembly dimensions. Silicone in mineral oil crosses that line within a week.

Temperature accelerates the process without changing the equilibrium. A test at 100 C reaches equilibrium faster than one at 23 C but arrives at the same swell, which is why oil resistance is specified at a temperature as well as a time.

Which fluids, and how to tell

Mineral oils, diesel, kerosene and most hydraulic oils are the worst case for silicone. They are non polar hydrocarbons and they are chemically similar to the polymer, so the absorption is high and fast.

Synthetic esters, phosphate esters and some glycols are intermediate. Silicone swells less, but the esters can also extract the additives from the compound, which changes its properties even when the swell is modest.

Silicone oils, fluorinated fluids and most aqueous fluids are benign. Silicone in silicone oil is the classic compatible pair, and it is why silicone seals are used in silicone oil systems and not in gearboxes.

The rule is to test the actual fluid, not a family. Oil formulations contain additives that change the interaction, and two mineral oils from different suppliers can give different swell results. A 72 hour immersion at the service temperature, with a volume swell measurement, settles it.

When to use something else

Fluorosilicone is the first alternative. It keeps most of the temperature range of silicone and reduces mineral oil swell to 10 to 15 percent, at three to five times the cost. It is the standard choice for fuel and oil service where the temperature is high.

FKM, or Viton, is the second. It swells very little in mineral oil and handles high temperatures, but it loses its flexibility below about minus 20 C and it is much harder to mould. It is the material for an engine seal and the wrong one for a freezer.

EPDM and nitrile are the third and fourth. Nitrile is the classic oil resistant rubber and is cheap, but it has a narrow temperature range. EPDM resists water and steam and is poor in oil. Neither has silicone's temperature range, and both are the right answer when oil resistance matters more than temperature.

The decision is a table, not an argument: temperature range, fluid, cost, and whether the part has to flex. Silicone wins on temperature and flexibility, loses on oil, and the answer is usually a different material rather than a different compound.

Table 1 — Oil resistance against temperature range

Designing around the swell

If silicone has to be used in oil, the groove can be sized for the swollen dimension rather than the original one. A seal that swells 15 percent in service should be designed so that the swollen size is the correct size, which means a smaller groove and a looser initial fit.

The second option is to keep the oil away from the seal. A lip seal with a wiper, a double seal with a drain between them, or a boot that separates the oil from the silicone will all work, and all of them cost less than a change of material.

The third is to accept a shorter life. A silicone seal in mineral oil at moderate temperature will swell and then stabilise, and if the groove has room for the swelling and the pressure is low, it can still seal for years. That is a legitimate engineering answer as long as it is a decision rather than an oversight.

Testing oil resistance

The standard test is ASTM D471: immerse specimens in the test fluid at a specified temperature for a specified time, then measure volume swell, hardness change, tensile strength and elongation. The volume swell is measured by weighing in air and in water, which is quick and accurate.

The test fluid matters more than the test conditions. Using a generic ASTM oil number 1 when the application uses a specific hydraulic fluid will give a number that has nothing to do with the application. Ask the fluid supplier for the test fluid or use the actual product.

The acceptance limits should be set from the groove design. A seal in a generous groove can tolerate 20 percent swell; one in a tight groove cannot tolerate 5. Setting a limit without reference to the design is how a part passes a test and fails in service.

Table 2 — What to measure after immersion

The compression set after immersion is the measurement that matters most and is the one most often left out. A seal can swell acceptably and still fail, because the swollen material takes a set faster and loses its sealing force. Measuring the set on the swollen specimen is the closest thing to a service test.