How to Verify a Replacement Part Is the Same Material

Why this matters

A part can match on every dimension, every rating stamped on the box, and every photograph, and still be made of something else. Dimensions are what suppliers list and what techs check, so a material change is the one substitution that passes every ordinary control a shop has. It then fails on a schedule set by chemistry rather than by wear, which puts the failure on a different visit from the install and usually on a different tech. The output of this procedure is a short written record. Its job is not to prove what a part is made of, which is often impossible in a truck. Its job is to make you state how confident you are, so nobody downstream mistakes an assumption for a fact.

There is a sibling article on verifying a replacement matches spec rather than just fitting, and it owns ratings, dimensions, and duty limits. This one owns composition only.

Before any destructive or chemical test

Three of the methods below can hurt you, and one of them is popular in shops and should not be.

• Never burn an unknown polymer to identify it. Fluoropolymers release toxic decomposition products when overheated, and halogenated plastics such as PVC give off hydrogen chloride. You cannot rule either out before the test, because ruling them out is what the test was for. Burn identification is a laboratory method with fume extraction, not a shop method, and it is off the list here entirely.
• Spark testing on a bench grinder means eye and face protection, a cleared area with nothing flammable within reach of the spark stream, and a guard set to the wheel. Sparks travel a lot further than people expect.
• Chemical spot test kits contain acids. Chemical-resistant gloves selected for that specific reagent, eye protection, and the kit's own safety data sheet read before it is opened. Employers must keep a safety data sheet available for every hazardous chemical in the workplace under 29 CFR 1910.1200, and hand protection has to be selected on its performance characteristics for that substance under 29 CFR 1910.138(b), which for an acid reagent rules out the ordinary nitrile disposables in the truck.

If the part is coming out of live equipment, isolate before it comes out: lock, tag, and prove zero pressure at a gauge for a mechanical or pressurized path under 29 CFR 1910.147, and for anything electrical de-energize under 29 CFR 1910.333(b)(2) and prove dead with the live-dead-live sequence in NFPA 70E-2021, 120.5.

The record, and why each field is on it

Six fields. The whole method is filling them in, in this order, because each one narrows what the next one has to answer.

The evidence field is the one that earns the record its place. Everything else can be filled by a person in a hurry with plausible text. Evidence forces the honest answer.

Establishing what came out

Work down this list and stop at the first source that answers. Do not skip up the list because a lower rung is faster.

1. A marking on the part itself. Alloy designations get stamped or cast into pressure and structural parts. Pipe and fittings carry printed markings that include the material and the standard they are made to. Elastomer parts sometimes carry a compound code. Wipe it and read it under a light before you assume there is nothing there.
2. The manufacturer's parts documentation for that assembly. Slower than a phone photo and definitive in a way no field test is.
3. Your own service record from the last time this position was touched, if your parts-used field carries material and not only a dimension. Most do not, which is a fixable problem and a good reason to fix it today.
4. A field test, which gets you a class or a family and almost never gets you a grade.
5. An assumption, which is legitimate as long as it is written in the evidence field as an assumption.

Narrowing a metal

A magnet separates families, not grades. Carbon steel, cast iron, and the 400-series martensitic and ferritic stainless steels are magnetic. Annealed 300-series austenitic stainless is essentially non-magnetic, but cold work raises it, so a formed elbow, a swaged end, or the area beside a weld on genuine 300-series can show noticeable pull. Treat weak magnetism at a worked area as inconclusive rather than as proof of the wrong alloy, and test on an unworked face.

A magnet does not tell 304 from 316. Both are austenitic and both fail the magnet test the same way. The difference is molybdenum, which is what buys 316 its chloride resistance, and the practical field method is a molybdenum spot test kit: a drop of reagent on a cleaned surface produces a color reaction when molybdenum is present. That distinction matters exactly where chlorides are present, which is coastal air, pool and spa equipment, water treatment, and anything seeing de-icing salt.

File a corner before trusting color. Plating and coating hide the base metal completely, and a bright nickel or chrome surface tells you nothing about what is underneath. A small file cut in a non-sealing, non-structural spot shows base color and gives you a feel for hardness at the same time.

Weight in the hand is a real test at the extremes. Aluminum against steel against a copper alloy is unmistakable to anyone who has handled all three. It is worthless between brass and bronze, and worthless between alloy grades.

Definitive identification exists and is available. A handheld X-ray fluorescence analyzer reads alloy composition in seconds and is normal equipment at supply houses, fabricators, and testing labs. For a pressure-retaining or structural part, the paperwork answer is a material test report from the manufacturer, and asking for one is routine, not exotic.

Narrowing an elastomer

Elastomers are the family where a wrong material behaves best on day one, which is why they cause the most trouble. Rings that look identical can be near chemical opposites: nitrile is at home in petroleum oils and destroyed by phosphate esters, EPDM is at home in water, steam and glycols and destroyed by petroleum oils.

• Colour is not identification. Black is every class. Some suppliers colour-code, many do not, and colour is a compound choice rather than a class marker.
• A durometer gauge measures hardness, not class. Shore A hardness is worth recording because a soft ring in a hard ring's groove will extrude, but two different classes at the same hardness are indistinguishable to the gauge.
• The controlled soak is the honest shop test. One ring from the new stock and one of known class, measured across the section with a caliper, dropped in jars of the actual service fluid at shop ambient, measured again after three days. Swell, softening, shrinkage or hardening in one and not the other is a compatibility answer even though it is not an identification. Manufacturers report the same property as percent volume swell under ASTM D471.
• The supplier can answer this in writing. If a listing, box, or packing slip does not name a compound class, the part is not identified. A supplier who cannot name the compound has told you something you needed to know.

Narrowing a rigid plastic

• Read the markings first. Piping products carry printed material and standard designations along the run. Moulded parts often carry an abbreviation in a recess near a gate mark.
• The float test separates the light polyolefins from everything else. Polyethylene and polypropylene are less dense than water and float; PVC, CPVC, acetal, nylon and the fluoropolymers sink. Use a clean unpainted chip with no filler or metal insert, and add a drop of dish soap to break surface tension so a small piece does not float on skin.
• Feel and sound. PTFE is heavy for its size, waxy to the touch and takes a fingernail mark that stays. Acetal is slick, stiff and rings when tapped. Polyethylene is soft enough to scratch with a thumbnail.

The record, filled in

A recirculating pump with an assembly of small elastomer seals, on a system carrying a petroleum-based fluid, in a plant with salt-laden outdoor air on the intake side.

• Location and part: pump discharge union, static face seal, and the four cap screws in that flange.
• Incumbent material: seal came out with no marking. The parts record said only a dimension. The equipment manual's parts list named the class as nitrile. Screws were non-magnetic on an unworked shank face and passed a molybdenum spot test.
• Candidate material: the seal in the assortment kit on the shelf, class not stated anywhere on the box, and replacement screws from general stock, magnetic on every face.
• Evidence: incumbent seal, documented from the manufacturer's parts list. Incumbent screws, tested, class-level only, which is enough here because the distinction that matters is whether molybdenum is present. Candidate seal, assumed, and the assumption is that an unlabelled assortment ring sold for general use is more likely water-service EPDM than oil-service nitrile. Candidate screws, tested, and they are a magnetic grade, so not 300-series at all.
• Duty: petroleum-based fluid on the wetted side, chloride-bearing outdoor air on the fastener side, ambient to hot in normal running.
• Accepted by: nobody, because two of the four evidence entries came back wrong.

The record did its work in the evidence column. The candidate seal was assumed rather than known, and against a petroleum fluid an EPDM ring is the one class that will not survive, so it was not fitted; a compound-labelled nitrile ring was ordered instead. The candidate screws were tested and are magnetic, which rules out the austenitic grade that came out and puts a chloride-sensitive fastener into salt air, so they were not fitted either. Neither of those calls needed a laboratory. Both needed somebody to write the word assumed instead of leaving the field blank.

Note what the record does not claim. It does not say the incumbent screws are 316. It says they are non-magnetic on an unworked face and molybdenum-positive, which is class-level evidence, and it says so in the same field. A reader three years from now can tell exactly how much that entry is worth.

What to own, and what to send out

Not every method belongs in a truck, and the split is about how often the answer changes a decision.

Worth owning: a small magnet, a caliper, a Shore A durometer gauge, and a few clean jars. Together they cost little, live in a drawer, and resolve the family-level questions that come up monthly. The jars in particular do more than their size suggests, because a side-by-side soak needs no calibration and no interpretation.

Worth owning if chlorides are in your work: a molybdenum spot test kit. Coastal service, pool and spa, water treatment and road-salt exposure all turn the 304-versus-316 question from academic into the reason the fastener failed.

Worth sending out: alloy analysis by X-ray fluorescence and any polymer identification beyond family. Both exist as a service, both are quick, and neither justifies a purchase for a shop that needs them a few times a year. Sending a part out also produces a document, which an assumption written on a job record never will.

References

• 29 CFR 1910.1200, hazard communication, for safety data sheet availability on test reagents and service fluids
• 29 CFR 1910.138(b), hand protection selected on performance characteristics for the specific substance handled
• 29 CFR 1910.147 for isolating and relieving a mechanical or pressurized path, and 29 CFR 1910.333(b)(2) with NFPA 70E-2021, 120.5 for de-energizing and proving dead on electrical work
• ASTM D471, standard test method for the effect of liquids on rubber properties
• See related: How to Verify a Replacement Part Matches Spec, Not Just Fit; How to Tell Which Seal Material You Are Holding