Fastener Inspection: What Hardness, Chemistry, Metallography and Dimensional Checks Actually Tell You

A custom bolt can look perfect and still be wrong. The finish is even, the threads clean, the head marking says 10.9 — and none of that tells you whether the steel is the grade you ordered, whether the heat treatment actually happened, or whether the thread will fit its mating nut. Those questions are answered by inspection, not appearance.

This guide explains the four core checks used on custom fasteners — hardness, chemical composition, metallography, and dimensional measurement — what each examines, how it is done, and what it can catch. It then shows you, the buyer, how to verify a shipment on arrival and read the test report that comes with it.

At HyHaven, all four checks — plus profile (contour) and roughness measurement — are done in-house at our Wenzhou works, and test reports are available with the order. But the point is bigger than our lab: once you know what each test proves and what it cannot, you can judge any supplier's report, including ours.

1. Hardness testing: did the heat treatment actually work?

For a quenched-and-tempered bolt, hardness is the fastest indicator that the steel was heat-treated into the class on the drawing. Strength itself is proven by tensile testing, but hardness tracks strength closely in alloy steels and can be checked on finished parts — the workhorse of batch inspection.

How it is done. An indenter is pressed into the surface under a defined load, and the depth of the indentation gives the value. Three scales cover most fastener work:

• Rockwell C (HRC) — a diamond cone under heavy load; the standard scale for hardened high-strength bolts (classes 8.8, 10.9, 12.9).
• Rockwell B (HRB) — a steel ball under lighter load; for softer low classes such as 4.8, where the C scale reads too low to be meaningful.
• Vickers (HV) — a diamond pyramid covering almost the whole range and, at low loads, able to measure a thin surface layer — useful for checking decarburization on a cross-section.

What it catches. Per ISO 898-1, each property class has a defined hardness window — roughly 23–34 HRC for class 8.8, 32–39 HRC for 10.9, and 39–44 HRC for 12.9. A batch below its window was not hardened and tempered as specified — or is the wrong steel altogether; a batch above it is brittle and risks cracking under the head or snapping during tightening. Both failures are invisible: a bolt stamped 10.9 that tests at 8.8 hardness is functionally an 8.8 bolt with a misleading marking, and hardness testing is how that gets found. (For how hardness, tensile strength and proof load relate, see bolt strength explained.)

Its limit. Hardness proves heat treatment, not chemistry — a wrong alloy can sometimes be hardened into the right range. It is never the only check.

2. Chemical composition: is the steel what the order says?

How it is done. The standard tool is optical emission spectrometry (spark OES): a spark vaporises a trace of metal from a tiny spot on the sample; each element emits light at its own characteristic wavelengths, and measuring them gives the percentage of carbon, chromium, manganese, silicon, phosphorus, sulfur and other elements within minutes.

What it catches. It verifies the material grade itself — the one thing no dimensional check can see. Our main production material is 40Cr, a chromium alloy steel of roughly 0.40% carbon and about 1% chromium: the carbon makes it hardenable to classes 8.8–12.9, the chromium improves hardenability through the section. A spectrometry reading confirms the batch actually is that steel, rather than a cheaper plain-carbon substitute that passes every visual and dimensional check, then fails to reach strength — or reaches it on the surface and stays soft in the core. It also flags out-of-limit phosphorus and sulfur, which make steel brittle. (More on this grade: 40Cr steel for high-strength bolts.)

Its limit. Chemistry says what the steel is, not what was done to it. Correct steel with skipped heat treatment still fails — that is what hardness and metallography catch.

3. Metallography: what does the inside of the steel look like?

How it is done. A sample is cut, mounted, ground, polished and etched with a mild acid. Under a microscope, the etched surface reveals the internal structure — the grains and phases that decide how the part behaves.

What it catches. Metallography is the deepest of the four checks — it confirms what the other three can only imply:

• Heat-treatment structure. A correctly quenched-and-tempered bolt shows a uniform tempered structure through the section. A part never heat-treated, or treated unevenly, shows a different, softer structure — direct evidence, not inference.
• Decarburization. If carbon was lost from the surface during heating, the thread keeps a soft skin that wears and fatigues early. A cross-section shows that layer directly, and microhardness can measure it.
• Grain flow in cold-headed parts. In a cold-headed bolt, the grain flows around the head and under-head fillet, following the part's shape — one reason cold heading is strong for standard geometries. In a machined part, the grain is cut through. Metallography makes this visible — and it is where harmful inclusions or internal cracks show up.

Its limit. Destructive and slow, it is a sampling and audit tool — used to qualify a process and verify batches, not to check every piece.

4. Dimensional and surface checks: will it fit and assemble?

How it is done. Each feature gets its fitting instrument: calipers and micrometers for lengths, head size and diameters; thread ring and plug gauges (go / no-go) for thread fit; optical image measurement for the profile of special-shaped parts; and a profilometer, tracing a stylus across the surface, for roughness.

What it catches. Everything about fit and assembly: thread diameters and tolerance class, head height and across-flats, shank length, concentricity and runout on machined parts, and the profile of non-standard shapes no single gauge can describe. On our machined work, dimensions are held to ±0.05 mm, and this check proves it part by part. The specified finish (black oxide, Dacromet, color zinc or zinc plating) should be even and complete, and roughness measurement confirms the surface itself — which matters for friction, and therefore for tightening. Cold-headed parts run M5–M24 and machined parts M8–M64 at our works; both go through the same discipline.

Its limit. A dimensionally perfect bolt can still be the wrong steel — fit is necessary, not sufficient.

The four checks at a glance

No single row proves a good bolt. Together they close the loop: right steel, right treatment, right structure, right size.

Incoming inspection: how to check a shipment when it arrives

You do not need a laboratory to receive custom fasteners intelligently. Work through these steps before the parts enter production stock:

1. Match the paperwork first. Confirm the report's grade, class, size, finish and lot identification match your order and the carton labels. A report from a different batch is worth nothing, however good its numbers look — traceability is the first check.
2. Read the hardness line against the class. Compare the reported values with the window for the class you ordered. Values outside it — or no hardness line at all on a heat-treated class — are questions to raise before use, not after a failure.
3. Check the chemistry line. Confirm the grade on the report is the grade on your order, and glance at the key elements — for 40Cr, carbon around 0.40%, chromium around 1%.
4. Spot-check dimensions yourself. Pull samples from more than one carton — top, middle and bottom, not just the box that opens first. Measure length, head size and major diameter with calipers, and run a mating nut or thread gauge over several pieces. You are verifying the batch behaves like the report says, not re-inspecting it.
5. Look at the finish. The coating should be even, with no bare patches, blisters or heavy thread build-up that stops a nut running on freely.
6. Quarantine problems, do not blend them. If something is off, keep that lot separate and contact the supplier with the lot identification and your measurements. Blended into stock, a suspect batch becomes untraceable — and unreturnable.

Common mistakes

• Judging a batch by appearance. Every serious failure mode — wrong steel, missed heat treatment, decarburization — is invisible. Appearance proves almost nothing.
• Accepting a report with no lot link. Numbers without batch traceability cannot be tied to the parts on your floor.
• Hardness-testing through the coating. A plated surface can distort a reading, and a surface reading can miss a soft core. Know where the measurement was taken before trusting it.
• Sampling only from the top box. Segregation happens in packing and transit. Sample across cartons.

FAQ

Which test reports can HyHaven provide with an order?
Reports covering the in-house checks described here — composition (spectrometry), metallography, hardness, dimensions, profile and roughness — are available with the order. Tell us at the quotation stage which reports your process requires, so they are prepared for your batch. (Our custom bolt quote checklist shows what to include in an enquiry.)

Does every part get tested?
No — and no honest supplier tests every part destructively. Dimensional and visual checks cover production broadly; hardness is sampled through the batch; composition is verified per material batch; metallography is a destructive sample test that qualifies and audits the process. What matters is systematic sampling, with results traceable to your lot.

Can I have the batch tested by a third party instead?
Yes — incoming third-party testing is normal practice. Agree the standard, sampling plan and acceptance windows in the order, before production rather than after a disagreement. Our reports give you the baseline.

What if my incoming check finds a problem?
Keep the lot sealed and separate, record the lot identification, and send the supplier your measurements or photos. Traceability makes a claim resolvable: with the lot identified, batch records can be checked against your findings.

Do these checks apply to both cold-headed and machined parts?
Yes — the same four checks cover both routes, across cold heading M5–M24 and machining M8–M64, in classes 4.8, 8.8, 10.9 and 12.9. The emphasis shifts: profile measurement matters more for special-shaped machined parts; grain flow is a cold-heading story.

The short version

Appearance tells you almost nothing about a custom bolt. Hardness proves the heat treatment, spectrometry proves the steel, metallography proves the inside, and dimensional checks prove the fit — and a buyer who matches the report to the lot, reads the hardness and chemistry lines, and spot-checks pieces across the shipment will catch nearly every realistic problem before assembly.

Preparing a custom fastener order, from an MOQ of 2,000 pieces? Send the drawing and tell us which reports your process requires via our custom fasteners page — inspection is built into the quotation.