A material test certificate arrives with the parts, gets stapled to the packing list and is filed. Three years later, when a joint leaks or an auditor asks for the heat, that certificate is the only thing that says what the steel was. Checking it takes about twenty minutes at goods-in. Finding out later that it does not match the parts takes months.
This page is the receiving side of the document: what to look at, in what order, and what a certificate cannot tell you no matter how good it looks. If you need the document explained from the supplier’s side, our material test certificate page covers the EN 10204 types and what we issue. This page is what to do with the certificate once it is in your hand.
Step 1: Check that the document is the type you ordered
The type is written in the header and it is the first thing to compare with your purchase order. Four types exist under EN 10204: 2.1 is a declaration of compliance with no test results, 2.2 is a test report with values that do not come from your heat, 3.1 is an inspection certificate with results from your heat signed by the manufacturer’s inspection representative, and 3.2 adds a countersignature from an inspector outside the manufacturer. A document headed “certificate of compliance” that carries no test values is a 2.1, whatever else it is called.
On a 3.1, the signature has to come from someone in the manufacturer’s inspection function who is organisationally independent of production. A sales manager’s signature, or a company stamp with no named person, is not that. On a 3.2, look for the second signature and the name of the body that sent the inspector: SGS, Bureau Veritas, TÜV, DNV or your own surveyor.
Check the mechanicals of the document too: certificate number, issue date, page count stated against pages supplied, the material standard and edition your order names, and no hand-written corrections. A corrected value has to be reissued with a revision number, not struck through and initialled.
Step 2: Trace the heat number to the parts you received
The heat number is what connects the paper to the steel. It appears on the certificate, on the part marking (stamped, engraved or on a tag) and on the packing list. All three have to agree. If the shipment carries parts from more than one heat, every heat needs its own results and the packing list has to say which parts came from which. A certificate showing one heat for a shipment of three is a paperwork gap, not a shortcut.
Ask how far the chain goes. For a forging there are usually two documents behind the certificate you receive: the steelmaker’s certificate for the billet, and the forging certificate that covers the heat treatment, the mechanical tests and the NDT the part went through. The heat number should be traceable across both. For oil and gas work the project specification usually fixes this: API Spec 20C, the closed-die forging standard for the petroleum industry, requires closed-die forgings to be identifiable by heat lot, and a certificate that cannot do that does not satisfy the order.
If the parts are small and unmarked, that is normal; the packing list then carries the mapping. What is not normal is a certificate with no heat number at all, or a heat number that appears nowhere on the goods.
Step 3: Read the values against the limits, not against the grade name
A grade name on a certificate proves nothing. What matters is whether each result sits inside the limit your specification sets, and whether the limit is printed beside it. A certificate that lists 585 MPa with no requirement next to it leaves you looking up the standard to find out whether it passed.
Chemical composition. Ladle analysis is taken from the melt before forging, product analysis from the finished part. On a forging, the product analysis is the one that shows what the process did to the steel, and it is the one to ask for. Both should be shown against their own limits, which are not the same: product analysis tolerances are wider than ladle limits.
Mechanical properties. Tensile strength, yield strength, elongation, reduction of area and impact energy. Impact energy is meaningless without its test temperature and specimen orientation; a value quoted at 20 °C against a requirement for −46 °C is not a pass. Charpy results should state whether they are single values or an average of three.
Heat treatment. The condition of supply (normalized, quenched and tempered, solution annealed, aged) plus the temperatures and holding times actually recorded. This is where a certificate often goes quiet: the condition is named but no cycle is given. For grades where the property depends on the cycle, such as F91, duplex and super duplex and the nickel alloys, ask for the chart or the recorded cycle.
Non-destructive testing. The standard, the method, the coverage and the acceptance level. “UT passed” is a sentence, not a report. A usable entry names the standard, the volumetric coverage and the acceptance level, and the report itself should be attached, with the indication record if the acceptance level requires it. The standard and the acceptance level are set by the order, not by a house default.
Dimensions. On machined parts, a dimensional report with actual values against the drawing tolerance, not just a tick.
What a certificate cannot prove
A certificate says what the supplier tested. It does not prove that the steel in the box is the steel on the paper, and mix-ups between grades happen in every supply chain. That gap is closed by positive material identification (PMI) on the parts themselves.
The limit that catches people out is carbon. Handheld XRF cannot measure carbon at all: the element is too light for the technique. That means a handheld XRF cannot separate 316 from 316L, because the difference between them is 0.07 % carbon against 0.03 %; it cannot separate 1045 from 1020; and it cannot confirm that a low-temperature steel really is fine-grained and impact-tested rather than a look-alike grade. The same limitation applies in varying degrees to sulphur, phosphorus, aluminium, silicon, magnesium, boron and beryllium.
Spark optical emission spectrometry (OES) does measure carbon, and it is the method to specify when carbon is the deciding element. LIBS sits between the two. On duplex and super duplex, nickel alloys and sour service work, project specifications commonly call for 100 % PMI of the finished parts, with the report attached to the release documents. If your specification says PMI, say it in the enquiry: it changes the inspection route and the price.
Our own practice is to re-verify material identity with a handheld XRF before packing, on top of the spectrometer analysis of every heat, so the steel that ships is the steel on the certificate. Where a customer’s specification requires carbon to be verified on the finished part, we carry that out by OES.
When something does not match
Do not accept “we will send a corrected copy later”. The certificate has to travel with the goods, because at that point it is the only evidence of what the goods are.
Write the mismatch as a document-level non-conformance and state three things: which field, by how much, and against which requirement. “Yield 355 MPa, order-specified minimum 415 MPa (illustrative values), certificate 3.1 dated …” is actionable; “certificate wrong” is not.
The usual resolutions, in order of cost: reissue the correct document type; retest from the delivered part and add the product analysis; accept by concession (deviation) with the deviation recorded against the heat; replace the parts. Whichever route is taken, file the certificate and the packing list as one pair, because separated neither is evidence.
The ten-point receiving checklist
| # | Check | Fail looks like |
|---|---|---|
| 1 | Document type matches the purchase order | A 2.1 or a certificate of compliance supplied against a 3.1 requirement |
| 2 | Signatory is the inspection function, not sales | Company stamp, no named person, or a production manager’s signature |
| 3 | 3.2 carries the outside inspector’s countersignature | Single signature on a document the project calls 3.2 |
| 4 | Certificate number, date, page count, standard edition | An edition older than the one the order names; missing pages |
| 5 | Every result printed with its specification limit | Values listed alone, so pass or fail cannot be judged |
| 6 | Product analysis on a forging, not only ladle | Ladle analysis only, on a part where the grade matters |
| 7 | Impact energy with test temperature and specimen orientation | A Charpy figure with no temperature |
| 8 | Heat treatment condition and recorded cycle | Condition named, temperatures absent |
| 9 | Heat number on certificate, part marking and packing list | Numbers that differ, or no heat number at all |
| 10 | NDT report attached, with standard and acceptance level | “UT passed”, with no standard, coverage or level |
Related Reading
- Material Test Certificates: EN 10204 3.1 vs 3.2
- Forging Quality Control
- API Spec 20C and NORSOK M-650
- Forged Materials and Grades
Frequently Asked Questions
Is a 3.1 certificate the same as a mill certificate?
Effectively yes. The mill certificate from the steelmaker covers the billet; the 3.1 you receive for a forging covers the forging, its heat treatment and its tests. Both use the EN 10204 numbering, and ISO 10474 uses the same 2.1, 2.2, 3.1 and 3.2 labels internationally.
Can I verify a certificate without opening the packing?
You can verify the document completely: type, signer, limits, traceability statements. You cannot verify the parts. PMI is the step that needs the parts in hand, and it is the only step that detects a grade mix-up.
How long should verification take at goods-in?
About twenty minutes for a shipment with one heat: the document checks are visual, the data checks are against the limits you already have, and the heat numbers are a three-way match. Shipments with several heats take longer, because each heat needs its own results.
What if the specification names PMI and the certificate does not mention it?
A certificate does not replace PMI; they answer different questions. Ask for the PMI report as part of the release documents, and specify the method (XRF or OES) in the purchase order, because only OES answers the carbon question.
Published by Zhejiang Longrui Forging Co., Ltd.
Technical reference information. Applicable specifications and inspection requirements must be agreed for each order. Questions about this note: sales@longruiforging.com