Inspecting hardened parts: what to ask for and what it proves

“Send a full inspection report” is the most common quality instruction on a tooling RFQ and one of the least useful. It costs money, it delays shipment, and it frequently proves the wrong things.

Here is what each measurement actually establishes, and what a proportionate package looks like.

Dimensional: CMM vs hand tools

A CMM report proves position, form and relationship — true position of a hole pattern, flatness across a face, perpendicularity between datums. It is the right tool when the tolerance is geometric rather than a simple size.

Hand measurement — micrometer, bore gauge, height gauge — proves size, and proves it faster and often more accurately than a CMM for a simple diameter. A good bore gauge beats a CMM on a precision bore.

Asking for a 40-dimension CMM report on a part with three critical features buys you 37 measurements nobody will read and a delay while they are taken. Mark the features that matter and ask for those.

The thermal issue. Steel expands roughly 11-12 µm per metre per °C. A 400 mm part measured at 30 °C reads about 45 µm longer than the same part at 20 °C. On a ±0.01 mm tolerance that is four times the band.

So: for tight-tolerance work, specify the inspection temperature and require that parts normalise before measurement. “Inspect at 20 ± 2 °C after minimum 4 hours soak” is a real requirement on a precision die and it is routinely omitted. In a tropical shop without climate control this is not a formality — it is the difference between a meaningful report and a fictional one.

Hardness: what it proves and what it does not

A Rockwell C test proves the bulk material reached the specified hardness. That is genuinely important — it catches a missed cycle, a wrong grade, or an undersized quench on a thick section.

What it does not prove:

• Case depth. On a nitrided or carburised part, the C-scale indenter punches through a thin case into the core and reads something in between. Case depth needs a micro-hardness traverse on a cross-section, or a superficial scale (HR15N, HR30N) for a quick surface check.
• Through-section uniformity. A reading on one face says nothing about the core of an 80 mm block. If section size is marginal for the grade — see S7 in heavy sections — ask for a reading on a cut face of a test coupon run with the batch.
• Surface integrity. A bulk hardness test cannot see a 10 µm white layer. See below.

Practical ask: hardness on the part, at a stated location, plus the heat treat certificate. The certificate gives you the cycle — austenitising temperature, quench medium, temper temperature and number of tempers — which tells you whether the double temper you specified actually happened.

Surface finish

An Ra value from a stylus trace proves texture amplitude. It does not prove:

• Direction. A turned surface with Ra 0.4 and a ground surface with Ra 0.4 behave completely differently against a seal, because the lay is different. If lay matters, specify it.
• Waviness. Ra filters out the long-wavelength component. A surface can read Ra 0.3 and still have 5 µm of waviness from a chatter condition. If flatness of the surface matters, ask for Wt or a profile trace, not just Ra.
• Integrity. See below.

What a standard package misses

The dimensional report, hardness test and finish trace — the usual three — all pass a part with thermal surface damage. White layer is 5-20 µm thick, changes no dimension, can be very smooth, and sits above material the bulk hardness test cannot resolve.

If the part is fatigue loaded, add a nital etch for thermal damage. It is cheap and non-destructive. The white layer article covers the detection methods in full.

If the part was wire EDM’d on a load-bearing edge, ask whether the recast layer was removed and how.

A proportionate package

One-off tooling, moderate load:

• Hardness reading with location stated
• Heat treat certificate
• Material certificate
• Dimensional check on the marked critical features only
• Photographs of the critical faces

Production die or mould:

• All of the above
• First article inspection on the full dimension set, once
• CMM report on critical features for each batch
• Nital etch on cutting edges or fatigue-loaded surfaces

Fatigue-critical or regulated component:

• All of the above
• Metallographic section on a first article
• Documented process controls: tool change intervals, cutting parameters
• Micro-hardness traverse if there is a case

Ask for it before the order, not after

The cost difference between planning an inspection and requesting one after the parts are cut is large. A shop that knows a nital etch is coming will manage tool wear on the finishing pass differently. A shop asked afterwards has already made whatever surface it made.

The same applies to the heat treat certificate — it has to be requested from the heat treater at the time, and retrieving it three weeks later is sometimes impossible.

List the proof package in the RFQ alongside the drawing. The RFQ checklist has it as line seven for exactly this reason, and there is a notes field on the form to put it in.