Accurately Predicting O-Ring Swell

For more than 60 years, engineers specifying seals and customers using them have accepted the arbitrary rule that the maximum amount O-rings will swell in most applications is a 25% change in volume.

The reason behind the rule was basic. Engineers analyzed how O-rings behaved when squeezed in a gland and discovered a 75% gland fill was typical, leaving 25% free space for thermal expansion and fluid-induced swelling. This rule was never actually tested or proven, but simply considered conventional wisdom and, therefore, heavily relied on by engineers when designing and selecting O-ring seals.

Out with the old
But today, high-performance sealing technology is more important than ever. OEMs and consumers demand leak-free systems, long life, safety, and environmental protection. And given the sophisticated test and analysis tools available, experts began to question why engineers still relied on a general rule of thumb.

To verify or disprove this rule, Simrit conducted extensive testing to establish volume-swell relationships between common seal materials and industrial fluids. We also developed design tools that accurately predict O-ring swell in actual applications.

Chemists tested VMQ (silicone), CR (chloroprene), NBR (nitrile), FVMQ (fluorosilicone), FKM (fluorocarbon), and EPDM (ethylene-propylene) in various fluids, including high-aromatic gasoline, jet-turbine lubricant, phosphate-ester hydraulic fluid, and standard reference test oils. Results showed that the maximum volume change for O-rings in most applications is actually 40%, not 25%. In total, 114 different experiments indicated there is a mathematical relationship that can predict the actual swell of O-rings installed in glands.

New industry standard
Obviously, results showed the old rule of thumb was flawed. The relationship between free-volume change typically reported on material data sheets and actual volume change of an O-ring in the gland is:

Squeeze-volume change
= 1.958 + (0.648 × V) – (0.0156 × V × S)

where V = free-volume change in fluid (from material data sheets) and S = % squeeze in the gland.

Results and examples are shown graphically in the Volume Change graph. Using initial results as a springboard, the company then developed a first-of-its-kind relationship design tool that provides much-needed, accurate linear-dimension-change data. The design tool is based on the equation:

% linear dimension change
= [(1+ (C/100)) / – 1] × 100

where C = % volume change from ASTM D471 immersion tests.

The results provide data on the relationship between volumetric and linear-dimension changes for the test materials in virtually any sealing application. Assuming that a rubber seal is homogeneous and swells equally in all directions, the volume change is simply the cubed version of a linear dimension change.

Tests concentrated on positive-volume change, not shrinkage, which is a valid concern with some fuel-resistant nitrile (NBR) rubbers exposed to fuel. And as of yet, Simrit has not applied the data to material modeling for FEA calculations.

The new data, verified by Underwriters Laboratories, provides new information about specifications and opens applications to alternative materials. The data eliminates the guesswork in seal selection and lets OEMs pick the most cost-effective, application-specific seal available.

With this new tool, customers can, potentially, reduce warranty issues and expand their material options for low-temperature applications, as well as reduce costs and improve long-term durability.

Underwriters Laboratories will also use Simrit’s research data to develop new standards for materials used with ethanol fuels. For example, UL has developed new limits for UL157a requirements for gaskets and seals used with gasoline/alcohol blends containing more than 15% ethanol. The new standard allows for free swell in these fuels of up to 40% based on Simrit’s data and verification testing.