Physical and Product Testing

What physical testing is for

Physical testing measures what a cured compound does when you deform it, heat it, or expose it to something. The results decide whether a compound is acceptable, and they are the common language between a supplier and a customer.

There is one thing worth establishing before any of the individual tests, because it shapes how all of them should be read. Rubber test data is reliable for comparing similar compounds and much weaker as an absolute prediction of service performance.

That is not a criticism of the tests. It reflects the fact that a laboratory specimen and a real part differ in geometry, stress state, temperature history, and everything else. A test tells you that compound A is stronger than compound B under these conditions. It does not tell you that a part made from compound A will last eight years.

Keeping that distinction live prevents a great deal of misplaced confidence — and it is why service experience and field data matter alongside laboratory results rather than being replaced by them.

Tensile, elongation, and modulus

A dumbbell-shaped specimen is pulled at a controlled rate until it breaks. Three numbers come out.

Modulus is the one that causes confusion, and it is worth flagging immediately even though the reason is covered later. In rubber testing, modulus means the stress at a specific elongation. It is not the modulus of a materials textbook, and two compounds can have the same M100 and behave differently everywhere else on the curve.

Hardness

A spring-loaded indentor is pressed into the surface and the depth of penetration is read on a 0 to 100 scale. Shore A is the standard scale for most rubbers; Shore D is used for very hard materials and Shore 00 for foams and very soft ones.

It is quick, non-destructive, and can be done on a finished part, which is why it appears in almost every specification and why it is the most frequently measured rubber property.

It is also the most over-interpreted. Hardness measures resistance to indentation at the surface, over a very small deformation. Two compounds at the same durometer can differ substantially in actual stiffness at service strains and behave completely differently in service. The relationship between hardness and modulus is empirical rather than fundamental, and it degrades quickly once different polymers or filler systems are compared. Matching hardness is not matching a material — and since normal production tolerance is around ±5 Shore A, a match is usually a match within a band to begin with.

Two practical points the standards settle. Specimens must be at least 6.4 mm thick, plied if necessary, or the reading is depressed by whatever is underneath. And the two main standards read at different times — ASTM D2240 at one second, ISO 48-4 at fifteen — so the same specimen gives two different numbers, which is the clearest worked example of the standards divergence discussed in Module 9.

There is also a second hardness system worth knowing. IRHD, under ISO 48, uses a non-linear scale related to Young's modulus rather than a linear deflection scale, and it is preferred for small or curved parts where a durometer cannot seat properly. IRHD and Shore A numbers are close in the mid-range and diverge at the extremes.

Compression set

A specimen is compressed by a fixed amount, held at temperature for a set time, then released and allowed to recover. Compression set is the fraction of the deformation that was not recovered, expressed as a percentage.

Zero percent means it returned to its original thickness. One hundred percent means it stayed exactly as compressed.

This is the key property for seals and gaskets, because a seal works by pushing back against the surfaces it is sealing. A seal that has taken a permanent set is no longer pushing, and it leaks — even though it looks intact and passes a hardness check.

Compression set is strongly influenced by cure system. EV and peroxide cures give low set; conventional sulfur cures give high set. This is one of the clearest cases where a cure system decision made months earlier shows up directly in a property.

Tear

A specimen with a notch, a cut, or a specific geometry is pulled, and the force required to propagate a tear is measured. Different specimen types measure different things, which is why the die type is always stated.

Broadly, some geometries measure resistance to a tear *starting*, and others measure resistance to a tear that already exists *growing*. These are different properties and a compound can be good at one and poor at the other.

Tear results scatter more than tensile results, because the specimen preparation — particularly the cut or nick — has a large effect. Comparing tear values from different die types is not meaningful.

Abrasion

A specimen is abraded against a standard surface and the volume or mass lost is measured, usually relative to a reference compound.

Abrasion results are useful for ranking similar compounds and poor at predicting real wear. Service wear depends on the surface, the load, the temperature, the presence of contaminants, and the specific wear mechanism, and no single laboratory test reproduces all of that.

Aging and fluid resistance

Both work the same way: measure properties, expose the specimens, measure again, and report the change.

As a practical indicator, many elastomers are considered substantially aged once elongation at break has fallen by roughly half — the convention ISO 11346 is built around. There is no universal industry definition of significant aging, and different applications judge it on retention of tensile, compression set, modulus change, or dynamic performance instead. Treat it as a quick sense-check rather than a formal criterion. ISO 11346 covers estimation of life and maximum temperature of use from accelerated ageing data, and is the standard to cite where a lifetime figure has to be defensible.

Ozone testing deserves one note now: the specimen is held at a specified strain, and the strain level matters enormously because strain is what turns surface attack into cracking.

Low temperature

Several different tests exist, and they answer different questions.

These do not always agree, and in some cases they rank compounds in a different order from one another. Which one is appropriate depends entirely on how the part fails in the cold — whether it shatters, stiffens, or fails to seal. This is covered properly in the next part.