How to Read Degradation on a Polymer

Why this matters

A failed seal, hose, or plastic housing carries a readable record of what killed it, and the record is fragile. Most of it lives in the first few thousandths of an inch of surface and in the part's current shape, and both are gone within seconds of a tech wiping the part on a rag and flexing it to see if it is still any good. What is left is a piece of black rubber that tells you nothing, so the replacement goes in identical to the one that failed and the clock restarts.

The examination below is ordered by what each step destroys. Every step wipes out evidence the later steps do not need but the earlier ones did, so running them out of order costs you the finding, not just time.

Before the part comes out

If the polymer part is part of a pressure boundary, relieve the pressure and verify it at a gauge or an open bleed point before you break the joint. A seal that has hardened and set can hold a joint together well past the point where the fastener load has relaxed, and it releases all at once. That is stored energy in the sense 29 CFR 1910.147 covers: isolate the energy source, lock and tag it, then verify the isolation by attempting to observe pressure at the gauge before you separate anything. Thermal energy counts too, so let a hot system cool rather than trusting gloves.

Assume the residue on the part is the fluid it ran in. Read Section 8 of that fluid's safety data sheet for the gloves and eye protection it calls for, and handle the part with those on. Degraded elastomer holds absorbed fluid and releases it when you squeeze the part, which is precisely what step 6 asks you to do.

Step 1: Record the part in place, before it is disturbed

Photograph the part installed, with enough of the surroundings in frame to show orientation, which face was toward the fluid, which was toward the outside, and where any heat source sits. Note the position of the worst damage relative to those.

What the next steps destroy: everything about position. Once the part is in your hand you can no longer say which side faced the heat, and half the interpretations below depend on exactly that. A hose degraded on the bore and a hose degraded on the cover are two different findings with two different fixes, and thirty seconds after removal nobody can tell them apart on a black hose.

Step 2: Look at the surface without touching it

Chalk, bloom, deposits and surface film are loose. They survive a photograph and do not survive a glove.

• Chalking: a dull powder over the exposed face. Surface photo-oxidation, usually sunlight.
• Bloom: a waxy or greasy film that reappears after wiping. Additives migrating out of the compound, often normal, sometimes a sign of overheating.
• Deposits: mineral crust, varnish, or fluid residue. Note the color and whether it is only on the wetted face.
• Gloss change: a matte patch on a glossy part marks where something attacked or where something rubbed.

Step 3: Map the cracks before you handle the part

Crack geometry is the single highest-value observation on an elastomer, and handling adds cracks and opens existing ones so you can no longer tell which were there.

• Fine, parallel cracks running perpendicular to the direction of stretch, only where the part is strained: ozone attack. Unstrained regions of the same part are clean. This is why a hose cracks on the outside of a bend and nowhere else.
• A network of shallow cracks over the sun-facing side, with the shaded side clean: ultraviolet photo-oxidation.
• Cracks radiating from a fastener hole or a clamp line: mechanical stress, often overtightening, sometimes stress plus a chemical agent.
• Fine cracks in a clear or rigid plastic, often shimmering under angled light: crazing. In service this is usually environmental stress cracking, which requires three things at once, a susceptible polymer, applied or residual stress, and a chemical agent. Remove any one and it stops.
• Deep cracks on the bore or wetted face only: chemical attack from the contents, not from the environment.

Step 4: Measure before you flex

Dimensions are destroyed by handling, because flexing a set or swollen part changes its shape permanently.

Measure a cross-section thickness or a diameter in the damaged zone and again in a part of the same component that was not exposed, or against a new part of the same nominal size. You are looking for direction, not precision.

• Larger than nominal means the polymer absorbed fluid: swell.
• Smaller than nominal means something was extracted from it, usually plasticizer, which is why the part is also harder and probably cracked.
• Flattened in the seal groove with no recovery is compression set: the material lost its ability to spring back, so the seal load fell to nothing while the part still looks physically present.

Commonly quoted compatibility guidance treats volume swell above roughly 10 percent as unacceptable for a dynamic seal and tolerates more in a static one, but that threshold belongs to the elastomer manufacturer's own compatibility table for the specific compound and fluid, so use theirs for a selection decision and use yours only to establish direction.

Step 5: Hardness, which wipes the chalk off

A durometer reading is close to non-destructive, but taking it means pressing a clean flat area, which removes the surface deposits from step 2. Use the Shore A scale for elastomers and Shore D for rigid plastics.

Two constraints make the difference between a number and a guess. The method (ASTM D2240) requires a specimen at least about 6 mm (0.25 in) thick, so stacking thin sections or reading a small O-ring cross-section gives a reading dominated by whatever is underneath. And you need the nominal value to compare against, which is on the part specification, not on the part.

Direction, again, is what you want. Harder than nominal means heat aging, oxidation, or plasticizer loss. Softer than nominal means fluid absorption or chemical attack. A shift of a few points is noise on a curved or thin part; a shift of ten points or more on a properly supported flat section is a real change.

Step 6: Flex it, which destroys the crack map

Now bend, stretch, or twist the part.

• A part that cracks audibly and shows a lighter-colored crack face has a brittle skin over sounder material underneath, which is the signature of a surface-first attack such as UV or ozone.
• A part that tears easily all the way through has degraded in bulk, which points to heat or chemical attack rather than exposure.
• A part that stays folded has lost its elasticity entirely.
• A rigid plastic that snaps with no yielding where the same material new would bend has embrittled, usually from heat, UV, or extraction.

Step 7: Section it, which ends the examination

Cut a clean cross-section through the worst area with a fresh blade and look at the cut face, ideally with magnification.

Depth is the finding. Degradation confined to a thin layer on one face means an external agent attacking that face. Degradation all the way through means the whole part was above its temperature limit, or the fluid permeated it completely. A discolored band on the wetted side, sound core, and sound outer face means the fluid did it and it had not finished.

This is last because there is no undo. Cut a part in half and you cannot photograph it installed, measure its free dimensions, or hand it to the manufacturer for a warranty claim.

Two seals, one machine, opposite mechanisms

A service call on a pumped-fluid skid produces two failed elastomer parts on the same day.

Part one, a static gasket at a hot flange. Nominal thickness 3.0 mm. It measures 2.6 mm across the seal face and 3.0 mm on the tail that sat outside the groove. That is a loss of 0.4 mm on 3.0, or 13.3 percent, and it did not recover after the part sat overnight. Durometer on the thick tail section reads 82 Shore A against a nominal 70, so it is 12 points harder. It cracks with a light-colored crack face when flexed and the cut section shows the discoloration running all the way through.

Harder, thinner, degraded in bulk, at a hot location: heat aging with plasticizer loss and compression set. Fluid swell would have gone the other way on both measurements. The fix is an elastomer rated above the actual flange temperature, so the first job is measuring that temperature rather than ordering the same part.

Part two, an O-ring in the pump end, same machine, same day. Nominal cross-section 3.0 mm, measures 3.5 mm, an increase of 0.5 mm on 3.0, or 16.7 percent. Durometer reads 58 Shore A against the same 70 nominal, so 12 points softer. No cracking, no set, and it is slightly tacky.

Bigger and softer means absorbed fluid. That is an incompatibility between the elastomer family and the process fluid, and it needs a different polymer family, not a higher temperature rating. Note that this part's swell of 16.7 percent is above the roughly 10 percent commonly quoted ceiling for dynamic service, which is worth calling out explicitly rather than letting it pass inside a tidy result: by that guidance this compound was outside its window from the day it was installed.

Two identical-looking black rings, two opposite mechanisms, two different corrective actions. A tech who wiped both, flexed both, and wrote "seals worn" replaces both with the same part and sees both again.

Confirming the finding before you order anything

• Get the fluid list, not the fluid. The process fluid is rarely alone. Flush chemicals, system cleaners, thread sealant squeezed into the flow path, and residual assembly lubricant all touch the same seal, and the incompatible one is often the cleaner used twice a year rather than the fluid that runs daily.
• Measure the actual service temperature at the part. Nameplate and setpoint are not the seal's temperature. A surface reading at the flange during operation settles a heat-aging call.
• Find an unexposed sample of the same material. A tail outside the groove, a section under a clamp, or the spare in the parts bin gives you the nominal you have been comparing against, and it costs nothing.
• Check that the replacement is the same compound, not the same color and size. Elastomer families are not distinguishable by eye, and a substitution made at a counter is invisible on the invoice. Ask for the compound designation in writing when the failure you just diagnosed was a compatibility failure.
• Keep the part until the replacement has run a full season. If it fails again on the same clock, the old part is the control sample, and by then you cannot get another.

References

• 29 CFR 1910.147, control of hazardous energy, for the isolation, lock, tag and verification sequence before breaking a joint that may hold pressure or stored mechanical energy
• ASTM D2240, durometer hardness, for the scale selection and the minimum specimen thickness that make a hardness reading valid
• ASTM D395, compression set of rubber, for the property behind a gasket that does not recover
• Elastomer manufacturer chemical compatibility tables and temperature ratings for the specific compound, which are product-specific and the only authoritative source for a selection decision
• See related: The Elastomers and What Attacks Each One; How to Check a Material Against the Fluid It Will Touch; How Seals and Gaskets Fail