Hardness Is Not the Same as Strength: The Correct Use of Fastener Hardness Testing
Executive Summary
Hardness testing is fast, cheap, and non-destructive, so it is often treated as a substitute for strength testing—but converting hardness to strength has explicit conditions and an error band. This article explains the physical relationship between hardness and strength, the limits of GB/T 1172 conversion, the logic for choosing among Brinell, Rockwell, and Vickers, and where hardness testing truly belongs.
Table of Contents
- The Cost of "Replacing" Tensile Testing with Hardness
- The Physical Relationship Between Hardness and Strength
- The Applicability Boundaries of Conversion
- How to Choose Among Brinell, Rockwell, and Vickers
- Where Hardness Should Be Measured
- The Four Places Where Hardness Testing Truly Belongs
- Conclusion: Hardness Is a Sieve, Not a Ruler
The Cost of "Replacing" Tensile Testing with Hardness
On the production line, tapping a bolt head with a portable hardness tester gives a result in minutes, without a destructive tensile test—a practice very common in batch incoming inspection. It does improve efficiency, but the problem follows: converting a hardness value into a strength value and then using it for acceptance judgment already exceeds the range that the conversion relationship itself can guarantee.
In a real case, a batch of grade 10.9 bolts that passed hardness (with converted strength meeting the spec) repeatedly fractured at the head transition zone during tensile testing. The reason was that the head grain flow had been cut during cold heading, while the hardness test point was chosen on the bolt end face and did not cover the problem area at all.
The Physical Relationship Between Hardness and Strength
Hardness can roughly reflect strength because both stem from the same microscopic mechanism in the material—the ability to resist plastic deformation. When an indenter presses into the material surface, local material undergoes plastic flow; when stretched, the material likewise undergoes plastic flow. The two physical processes are correlated, so an empirical conversion relationship exists:
Rm ≈ k × HB (k ≈ 3.3–3.5, in MPa)
This relationship fits fairly well for carbon steel and low-alloy steel, but it remains a statistical curve with an error band typically of ±8% to ±10%. This means: a grade 10.9 bolt with a true tensile strength of 950 MPa may convert to 870–1030 MPa from hardness. If 1040 MPa is the acceptance lower limit, this bolt would be judged non-conforming—but it is actually conforming.
The Applicability Boundaries of Conversion
The reliability of the conversion depends on whether the material meets the preconditions. The following situations significantly weaken its validity:
| Situation | Effect on Conversion | Recommendation |
|---|---|---|
| Carbon steel, low-alloy steel, uniform structure | Conversion reliable | Can be used for batch screening |
| Carburized or surface-hardened parts | Surface hardness far higher than core; conversion seriously overestimated | Conversion prohibited; must test the core or run a tensile test |
| Non-uniform quench & tempered structure | Large scatter; converted value unreliable | Increase measurement points and average; confirm with a tensile test |
| Stainless steel (austenitic) | Significant work hardening; non-linear hardness-strength relationship | Conversion not recommended |
| Non-ferrous metals (aluminum, copper, titanium) | Relationship completely different from steel | Use the corresponding material conversion table, or run a tensile test directly |
| Thread roots, fillets, and other thin-wall locations | Indentation disturbed by adjacent material and curvature | Use micro-Vickers or do not test hardness |
How to Choose Among Brinell, Rockwell, and Vickers
The three common hardness test methods each have their own applicable scenarios; choosing the wrong method directly causes distorted data:
| Method | Indenter and Load | Applicable Object |
|---|---|---|
| Brinell HBW | Tungsten carbide ball, 3000/750/187.5 kgf | Large fasteners (M16 and above), castings and forgings |
| Rockwell HRC | Diamond cone, 150 kgf | Quench & tempered high-strength bolts (grades 10.9, 12.9) |
| Rockwell HRB | Steel ball 1.588 mm, 100 kgf | Low-strength bolts (grades 4.8, 8.8 soft condition) |
| Surface Rockwell HR15N | Diamond cone, low load | Thin-wall parts, carburized layers, small sizes |
| Vickers HV | Diamond pyramid, 1–50 kgf | General purpose; micro-Vickers for thread roots, thin layers |
- Indentation size should match the part scale: the indentation diameter should be less than 1/4 of the part's effective thickness, and also less than the available width at the thread root.
- Hardness range determines the method: HRC applies to 20–70 HRC, HRB applies to 20–100 HRB; readings outside these ranges are meaningless.
- Prefer Vickers in acceptance disputes: HV indentations are small and highly repeatable, adapting best to irregular parts and thread-root locations.
- Prefer Rockwell for batch screening: fast to operate and tolerant of surface roughness, suited to on-line production inspection.
Where Hardness Should Be Measured
The measurement location is the step in hardness testing most prone to problems and least often standardized:
High-Strength Bolts: End Face vs. Shaft Measurements Differ Greatly
For cold-headed bolts, the head material undergoes severe plastic deformation and its hardness is usually slightly higher than the shaft; if the head experiences different heating conditions during heat treatment, head hardness may instead be lower. As a basis for batch consistency judgment, the test point should be fixed at the center region of the bolt head end face or at a circumferential location 0.5d–1d from the head on the shaft, and standardized uniformly. As an aid to strength verification, measure on the cross-section of the unengaged thread portion, taking a reading at the quarter point.
Nuts and Surface-Treated Parts
Nut hardness is usually measured on the bearing face or the hex side; bearing-face hardness directly affects preload transmission and carries greater engineering significance. The hardness of zinc plating or zinc-aluminum coatings is completely different from the substrate, so a reading taken on a coated surface will deviate markedly from the true value; the correct approach is to measure below the diffusion layer, or on unplated specimens from the same batch.
A practical rule: the hardness test plan must write the "measurement location, number of points, and reading method" into the inspection document. For the same part, different measurement points can yield results differing by 40 HV.
The Four Places Where Hardness Testing Truly Belongs
The value of hardness testing is not in "replacing tensile testing," but in the irreplaceable role it plays at four specific steps:
| Purpose | Method | Criterion |
|---|---|---|
| Heat-treatment batch consistency screening | Rockwell HRC, sample 5–10 pieces per batch, 3 points each | Within-batch scatter ≤30 HV (approx. 3 HRC) |
| Fast incoming discrimination and mixed-material screening | Portable hardness tester, full inspection or high-rate sampling | Compare against the hardness band corresponding to the nominal class |
| Monitoring substrate change before/after surface treatment | Vickers HV, micro-indentation | Substrate hardness change before and after plating ≤5% |
| Failed-part hardness distribution analysis | Measure HV point by point along the cross-section from surface to core | Determine whether decarburization or local hardening exists |
Conclusion: Hardness Is a Sieve, Not a Ruler
There is a stable statistical relationship between hardness and strength, but no strict functional relationship. Treating hardness as a sieve lets you efficiently screen out clearly abnormal batches and detect heat-treatment process variation; treating hardness as a ruler—using it to produce a precise strength number and then judging conformity on that—is mistaking a statistical tool for a metrological standard.
The correct inspection combination is: hardness handles fast screening and consistency monitoring, tensile and wedge-load testing handle final judgment, and metallography with hardness gradients handle failure localization. Only when the three each do their own job can cost be controlled while real risks are not missed.
