Material Traceability 2026: EN 10204 Types 2.1, 2.2, 3.1 and 3.2, Heat Numbers, and What a Hardware Factory Can Honestly Trace

"Can you send the material certificate?" is asked on almost every serious hardware enquiry, and it is answered with four genuinely different documents depending on who is answering.

The differences are not shades of thoroughness. They are differences in who performed the test, whether it was performed on your material, and whose signature is on the page — and a buyer who does not know which one they received cannot tell whether they have evidence or a letter.

The four EN 10204 document types

EN 10204 is the European standard for inspection documents on metallic products. It defines four types, and the two axes that separate them are specific versus non-specific testing, and who validates the result.

The line that matters runs between 2.2 and 3.1. Everything at 2.x tells you what the manufacturer's material generally is; everything at 3.x tells you what your material specifically is, tested and signed by somebody who does not report to the person who made it.

A 2.1 is not worthless — it is a contractual statement that the order was met, and for a great many products it is proportionate. It is simply not a test result, and a specification asking for 3.1 is not satisfied by a well-presented 2.1.

What a heat number actually traces

A heat number identifies a single melt at the mill. Every kilogram poured from that melt shares one chemical composition, and a 3.1 certificate for that heat states the analysis and the mechanical properties measured on it.

That is a strong chain while the material is large and few: a plate, a bar, a coil. Each piece can carry the number stamped or etched, and the certificate follows the piece.

Where it weakens is exactly where door hardware lives. A single melt becomes tonnes of bar; the bar is cut into hundreds of thousands of small parts; the parts go into bins. Once a lever is one of forty thousand identical levers in a bin, it does not carry a heat number and nothing short of segregating the entire batch will put one back on it.

So the honest statement for small hardware is this: traceability to a heat is a property of a *batch that was kept separate*, not of a part. If a project needs it, that has to be specified before production, because it changes how the factory handles the material — and it costs, because separated batches cannot be pooled.

Two more places the chain breaks, and neither is dishonesty

**Finishing.** A part is often plated or coated by a specialist outside the factory. The base metal's heat may be known; the plating bath's chemistry is a separate record kept by a separate company, and a certificate covering both is two documents rather than one.

**Multi-part assemblies.** A mortise lock is a case, a forend, a latch, a follower, springs and fixings — frequently from different suppliers and different melts. A single "material certificate" for a lock is a category error; what exists is a certificate per component for the components that have one.

Neither of these is a factory being evasive. They are the shape of the supply chain, and a supplier who says "yes, full traceability" for an assembled lock without qualifying it has told you something that cannot be true.

Material traceability is not RoHS, and not REACH

These three get requested in one sentence and answered as if they were one thing. They are not.

A RoHS declaration says nothing about tensile strength. A mill certificate says nothing about restricted substances. Asking for "the certificates" and receiving one of them is not a partial answer to your question; it is a complete answer to a different one.

The REACH row carries a trap worth naming: the candidate list is revised periodically, so a REACH statement is true as of a date. A statement with no date on it is not usable, and one from three years ago should be refreshed rather than filed.

What to ask for, in the order that gets an answer

Name the document type explicitly. "EN 10204 3.1 for the lock case" is answerable; "the material cert" produces whatever the factory has nearest.

Name the component, not the product. See the assembly point above.

Say whether you need it before or after production. A 3.1 on a specific batch has to be planned; a 2.1 can be issued against an existing order. Asking for a 3.1 after the goods have shipped and been mixed into stock is asking for something that no longer exists.

Say what the requirement comes from. A specification clause, a customs requirement and an internal quality policy lead to different documents, and a factory that knows which one you are answering can tell you whether a cheaper document satisfies it.

Five checks on a certificate you have been sent

Most bad certificates are not forgeries. They are real documents for something else, forwarded because they were the nearest thing to hand. These five checks catch almost all of them and take about two minutes.

**Does the material grade on the certificate match the grade on your order?** A certificate for 201 stainless attached to an order for 304 is the single most common mismatch, and the two are visibly identical in the hand. Our stainless pages cover why.

**Does the quantity or batch on the certificate plausibly cover your order?** A 3.1 naming a heat of two tonnes cannot be evidence for an order that was drawn from a bin holding five different heats.

**Is the issuing party named, with an address?** A mill certificate is issued by the mill. A document issued by the trading company that sold you the parts, restating figures it was given, is a transcription rather than a certificate.

**Is it dated, and is the date before your production?** A certificate dated after the goods shipped was written to answer your question rather than to record a test.

**Does the document say which standard the testing followed?** A table of numbers with no method beside it cannot be checked by anybody, including the person who wrote it.

If a certificate fails one of these, the useful response is a specific question rather than a rejection. "This is dated after our shipment — is there an earlier one for the heat our parts came from?" is answerable, and the answer tells you what kind of supplier you have.

What we can say about our own material

We publish the material and finish of every product as we supply them — the alloy family, the finish code, and where a model is supplied in more than one, all of them.

What we do not do is publish a blanket traceability claim, because the honest answer depends on the component and on whether the batch was kept separate. Our documents page lists test documents as available on request, and "on request" there means we tell you per order what exists for those exact components rather than sending a document that looks like the one you asked for.

Where a project needs EN 10204 3.1 on a specific component, that is a conversation before production rather than a box ticked afterwards, and the cost of batch separation is stated rather than absorbed quietly and then discovered in the price.

Why we are this careful about a piece of paper

Because the failure mode is delayed and expensive. A borrowed certificate, a 2.2 presented as a 3.1, or a heat number attached to a part that was never segregated all pass inspection on the day.

They fail later, at an audit, on a project where somebody checks — and what fails is not one document. It is every other number the supplier ever published, because a buyer who finds one paper that does not hold up has no way to know which of the others do.

That is the same reasoning as the dash where a dimension is missing. A missing document costs an enquiry; an invented one costs the account.

That is why the five checks above are written as questions to send rather than as grounds to walk away. A supplier who answers them precisely is worth more than one whose paperwork was tidy the first time, because you now know what happens when you ask. A precise answer to an awkward question is the cheapest quality signal a factory can give, and the only one that survives an audit.