Quality Assurance and Non-Destructive Testing

Every lot ships with a dimensional report and a visual inspection record; UT and MPI are run in-house to whatever acceptance level the drawing names, and a new die gets a full first-article report from a CMM in a 20 ± 2 °C room before serial production is released. Material certificates to EN 10204 3.1 are issued on request for the heat actually used, process control runs to ISO 9001, and 12,000+ t a year ships to drawings held at ISO 8062 class C as-forged and ±0.2 mm on machined critical dimensions.

Which NDT method finds which defect

The method has to match the defect, and the defect has to match the process. A forging defect is almost always one of four things: a lap or fold from the die, an internal void or pipe from the ingot, a crack from thermal or mechanical stress, or a coarse grain / mixed structure from the heat treatment. Each of those sits in a different place and needs a different probe.

Specifying "NDT to the relevant standard" without naming the method is the most common reason an inspection is repeated at the customer gate: the supplier inspects for what the process produces, the customer inspects for what the drawing implies, and the two are not the same test.

What the inspection report contains

A report that states "inspected, passed" cannot be checked by anyone. The version that can be checked lists the instrument, the resolution, the datum scheme and the ambient condition, so that your incoming inspection can reproduce the measurement rather than re-litigate it.

On a CMM report the datum scheme matters more than the numbers. A part aligned on its draft walls measures differently from the same part aligned on its die face, and both are defensible until the drawing says which is authoritative.

First article: what is actually verified

First article inspection on a forging is not a dimensional check of five parts. It is the point where the die, the process card and the drawing are proven to agree, and it covers four things: every drawing dimension on the first five parts, the material certificate for the heat used, the NDT result at the specified level, and the process parameters recorded during the run (billet temperature, die temperature, stage sequence, heat-treatment cycle).

The value is in the disagreement. A first article where every dimension is nominal usually means the die was measured, not the parts. What a buyer should look for is the spread across the five parts and the one or two dimensions sitting near the limit, with a note on what will be done about them before serial.

• Every drawing dimension, on five consecutive parts, from a CMM in a controlled room
• The spread across the five parts, not only the deviation from nominal
• Material certificate to EN 10204 3.1 for the heat actually consumed
• NDT at the level the drawing names, with the coverage area recorded
• Process parameters recorded during the run, so the serial parts can be compared against a baseline
• A list of dimensions near the limit and the correction planned before serial release

Which standard governs which acceptance decision

Four standards do most of the work on a forged part, and confusing them is how a part passes at the supplier and fails at the customer. Each one answers a different question.

Traceability from heat to shipped part

Traceability fails at the point where a lot is split, not at the point where the certificate is printed. A heat number stamped on a bundle is not traceability if the bundle is broken and the parts are mixed into a bin with a different heat.

The convention that holds is to mark the heat on the part or on a tag that travels with the part through every operation, and to record the marking in the process card at each step. Where the part is too small to mark, the tag travels in the box and the box is not opened until the parts are packed for shipment.

• Heat number marked or tagged at the billet stage, before any cutting
• Marking recorded in the process card at every operation, not only at the end
• Mixed heats never share a bin, even when the grade is identical
• Certificate issued against the heat actually consumed, not against the grade ordered
• Any substitution in grade or mill flagged in the quotation before the order is accepted

Frequently asked questions

Which NDT method should I specify for a steel forging?

For internal soundness, ultrasonic testing to ASTM A388 or EN 10228-3, from a machined face where the geometry allows. For surface and near-surface cracks along the parting line and fillets, magnetic particle testing to ASTM E709 or EN 10228-1. Penetrant replaces MPI on austenitic stainless and on aluminium, where MPI does not work at all. Specifying the method and the acceptance level together is what makes the result comparable between suppliers.

Do you issue EN 10204 3.1 certificates?

Yes, on request with the quotation, with tensile and impact results for the heat actually used rather than for a generic grade. A 3.1 certificate is signed by the manufacturer independent of the production department, which is the difference between a 3.1 and a 2.2 declaration.

What does a first article report cover?

Every drawing dimension on five consecutive parts, measured on a CMM in a 20 ± 2 °C room, with the datum scheme stated. It also carries the material certificate for the heat used, the NDT result at the specified level, the recorded process parameters, and a list of dimensions close to the limit with the correction planned before serial release.

Why does the datum scheme matter so much on a forging?

Because a forged part has draft on almost every wall, and a draft wall is a poor datum: a small axial shift changes the measured position of every feature referenced to it. Alignment on the die face, or on a machined datum that is itself referenced to the die face, gives a measurement that both parties can reproduce. Best-fit alignment can rotate a real deviation out of the result, which is why the report states the method.

How much does temperature affect the measurement?

For steel, roughly 0.012 mm per metre per degree Celsius. A 300 mm part measured 10 °C away from the reference temperature is out by about 0.036 mm, which is comparable to a tight tolerance on its own. That is why CMM work is done in a 20 ± 2 °C room and why the report states the condition rather than assuming it.

What happens if a part is out of tolerance?

The report states the feature, the measured value and a disposition, and the customer decides whether to accept by deviation or reject. Rejected parts are either reworked where stock allows, or replaced. The useful outcome is the one where the cause is corrected on the die or the process card, so that the same deviation does not return in the next lot.

All figures are typical reference values for the materials and processes named. Final acceptance criteria are fixed by the drawing, the applicable standard (ISO 8062, ISO 2768, ASTM E140, EN 10204, ASTM A388) and the agreed inspection plan.

Engineering: wuq268022@gmail.com · +86-512-6808-6677