Hardness test data processing is the systematic conversion of raw indentation measurements into traceable, standards-compliant hardness values, covering calculation, validity screening, statistical evaluation, uncertainty analysis, and scale conversion. Accurate data processing in accordance with ISO, ASTM, GB, and JIS standards is essential for material acceptance, quality control, and failure analysis. This article explains the complete data processing workflow for Brinell, Vickers, Rockwell, Knoop, and Shore hardness testing of solid materials.
Technical Overview
Solid hardness testing quantifies a material's resistance to localized plastic deformation by pressing a standardized indenter into the surface under a controlled force. The raw output of a hardness test—such as an indentation diagonal length, a penetration depth, or a durometer reading—is not itself the final result. Data processing transforms these raw observations into reported hardness values through a defined sequence of steps: validity screening, formula-based calculation, rounding, statistical summarization, uncertainty evaluation, and, where required, conversion between hardness scales or estimation of tensile strength.
The processing rules are not arbitrary; they are prescribed by international and national standards so that results from different accredited third-party laboratories are comparable. The principal standards governing test execution and data treatment include:
- ISO 6506-1 / ASTM E10 / GB/T 231.1 / JIS Z 2243 — Brinell hardness (HBW)
- ISO 6507-1 / ASTM E92 / GB/T 4340.1 — Vickers hardness (HV)
- ISO 6508-1 / ASTM E18 / GB/T 230.1 / JIS Z 2245 — Rockwell hardness (HR)
- ISO 4545-1 / ASTM E384 / GB/T 18449.1 — Knoop hardness (HK) and microindentation hardness
- ISO 14577 — Instrumented indentation testing (Martens hardness, HM)
- ISO 868 / ASTM D2240 / GB/T 531.1 — Shore durometer hardness for plastics and elastomers
- ISO/IEC Guide 98-3 (GUM) and ISO 21749 — Evaluation of measurement uncertainty
- ASTM E140 / ISO 18265 — Hardness scale conversions for metals
Modern practice integrates these steps into laboratory information systems and validated software, ensuring full traceability from the raw reading to the certified report.
Testing Samples
Metallic Materials
Steels, cast irons, aluminum alloys, copper alloys, titanium alloys, and nickel-base superalloys constitute the majority of hardness test samples. Specimens must be free of scale, decarburized layer, oil, and oxide. For Vickers and Knoop testing, the surface roughness should be low enough that the indentation diagonals can be resolved—typically a ground or polished finish with Ra no greater than about one-thirtieth of the expected indentation depth. Specimen thickness must be at least 1.5 times the Vickers diagonal, at least 8 times the Brinell indentation depth, or at least 10 times the Rockwell permanent depth, so that no bulging or marking appears on the reverse side.
Polymers and Elastomers
Thermoplastics, thermosets, cured rubber compounds, and elastomeric seals are tested by Shore durometry (Shore A for soft grades, Shore D for hard grades) or by IRHD methods. Samples require flat, parallel test areas of sufficient area to place the presser foot at least the minimum distance from any edge, and conditioning at standard laboratory temperature is mandatory because polymer hardness is strongly temperature- and time-dependent.
Ceramics, Glasses, and Coatings
Advanced ceramics, sintered carbides, optical glasses, nitrided layers, and thin hard coatings are evaluated by low-force Vickers or Knoop indentation (for example HV0.5 down to HV0.01). Data processing for these samples must verify that the indentation depth does not exceed roughly one-tenth of the coating thickness; otherwise the substrate contribution must be considered and the result flagged as composite hardness rather than coating hardness.
Welds and Heat-Treated Components
Transverse weld sections, carburized or case-hardened parts, and induction-hardened shafts are sampled by traverses of microindentations to build hardness-depth profiles. Here data processing also includes profile plotting and the determination of effective case depth, defined per ISO 2639 as the depth at which the hardness equals 550 HV1 for carburized steel.
Testing Items
Typical measurands and deliverables processed in a hardness laboratory include:
- Brinell hardness values (for example HBW 10/3000) with mean and scatter
- Vickers hardness values across macro (HV10, HV30) and micro (HV0.5, HV0.1, HV0.05) force ranges
- Rockwell hardness values on regular scales (HRA, HRB, HRC, HRE) and superficial scales (HR15N, HR30T)
- Knoop microhardness values (HK0.1, HK0.01) for brittle materials and thin layers
- Shore A and Shore D durometer readings for polymeric and elastomeric products
- Martens hardness (HM), indentation hardness (HIT), and creep parameters from instrumented indentation
- Hardness traverses, hardness-depth profiles, and effective case depth measurements
- Scale conversions (for example HV to HRC, HBW to HV) and estimated tensile strength
- Expanded measurement uncertainty statements accompanying certified values
Testing Methods
Data Acquisition and Validity Screening
Before any calculation, each raw reading is screened against the geometric validity rules of the applicable standard. Indentations must be spaced so that strain-hardened zones do not overlap—typically at least three diagonal lengths between indentation centers for Vickers and at least 2.5 diagonal lengths from the specimen edge, with larger factors for light metals. Brinell results are valid only when the ratio of indentation diameter to ball diameter lies between 0.24 and 0.6. Any indentation showing asymmetric shape, edge lifting, or surface marking on the reverse side of the specimen is rejected and excluded from the data set. Ambient conditions are also verified: standard tests are performed at 10 °C to 35 °C, while referee tests are conducted at 23 °C ± 5 °C.
Hardness Value Calculation
Valid readings are converted to hardness numbers using the formulas defined in the standards, with the test force F expressed in newtons and lengths in millimetres:
- Vickers: HV = 0.1891 × F / d², where d is the arithmetic mean of the two diagonals
- Brinell: HBW = 0.102 × 2F / [π × D × (D − √(D² − d²))], where D is the ball diameter and d the mean indentation diameter
- Knoop: HK = 1.451 × F / l², where l is the long diagonal
- Rockwell: HR = 100 − e for diamond scales (A, C, D) and HR = 130 − e for ball scales, where e is the permanent depth increment expressed in units of 0.002 mm; digital machines apply this algorithm automatically
- Shore: the direct scale reading taken after the specified dwell time, with force-geometry conformity verified per ASTM D2240 or ISO 868
Rounding and Reporting Rules
Standards prescribe how calculated values are rounded. Vickers and Brinell hardness numbers are rounded to three significant figures. Rockwell values are reported to 0.1 HR unit on digital displays, and Shore readings are recorded to the nearest scale unit. The test report must state the scale designation (including force, indenter, and dwell time, for example HV10/15), the number of valid indentations, the individual or mean values, the arithmetic mean, the scatter range, and the machine identification and its verification status.
Statistical Treatment of Replicate Results
Because hardness is a locally measured, inherently scattered property, replicate measurements are mandatory. Common practice requires a minimum of three and typically five valid indentations per specimen. Data processing computes the arithmetic mean, standard deviation, and coefficient of variation. Individual readings that fail geometric validity criteria are removed before averaging, and genuinely anomalous values may be screened with a statistical outlier test such as the Grubbs test, applied only when a physical cause is suspected and documented. For durometer testing, the median of at least five readings is reported because the median is robust against localized surface defects.
Measurement Uncertainty Evaluation
Quantitative uncertainty is evaluated following the GUM framework. Type A components arise from the statistical scatter of replicate indentations. Type B components include the error of the test force, the geometrical tolerances of the indenter, the resolution and calibration of the diagonal-measuring system, machine repeatability demonstrated during direct and indirect verification per ISO 6507-2 and ISO 6507-3 (and the equivalent parts of ISO 6506, ISO 6508, and ISO 4545), and temperature effects. Because Vickers hardness is inversely proportional to the square of the diagonal, a relative uncertainty in diagonal measurement is doubled in its effect on HV and therefore dominates the budget at low forces. The combined standard uncertainty is multiplied by a coverage factor k = 2 to obtain the expanded uncertainty corresponding to approximately 95 % confidence, which is stated on the certificate.
Hardness Scale Conversion and Strength Estimation
Where a conversion is contractually required, data processing uses the tabulated empirical relationships of ASTM E140 or ISO 18265, which are alloy-family-specific for steels, nickel alloys, copper alloys, and cast irons. Converted values are always reported as approximations with an indication of the source table, because no universal conversion exists. For carbon and low-alloy steels, an approximate relationship Rm ≈ 3.4 × HBW (in MPa) is sometimes used for tensile strength estimation, but it must never replace a direct tensile test for acceptance purposes.
Data Recording and Digital Traceability
Complete data processing includes the retention of raw readings, force–time records, dwell times, indenter serial numbers, reference block values, operator identity, and software version. Automated hardness systems export machine-readable results that are version-controlled and audit-trailed, enabling periodic interlaboratory comparison and proficiency testing to confirm the long-term validity of the data pipeline.
Applications
Processed hardness data support decision-making across the entire product life cycle:
- Heat treatment verification: confirming that quench-and-temper, annealing, and carburizing cycles achieved the specified hardness range and case depth
- Incoming material inspection and supplier certification against purchase specifications
- Weld procedure qualification: hardness traverses across the heat-affected zone to screen for excessive hardening or softening
- Failure analysis: mapping local hardness to identify grinding burns, decarburization per ISO 3887, hydrogen embrittlement, or fatigue-softened regions
- Production statistical process control: trending mean hardness and scatter on control charts for machined and stamped components
- Research and development: instrumented indentation per ISO 14577 for local mechanical properties of new alloys, coatings, and additively manufactured materials
- Field assessment: Leeb rebound hardness per ISO 16859 with ASTM E140 conversion for on-site evaluation of large structures where stationary machines cannot be used
FAQ
How many indentations are required for a valid hardness result?
Most standards require a minimum of three valid indentations per specimen, and laboratory best practice uses five. If one indentation fails the validity screening rules for spacing, edge distance, or geometry, an additional indentation must be made so that the reported mean is based on the required minimum number of valid readings.
Why do replicate Vickers measurements sometimes scatter widely?
Scatter typically originates from insufficient surface preparation, inhomogeneous microstructure, indentation size effect at very low forces, or poor optical resolution of small diagonals. Data processing will reveal this through the standard deviation, and the corrective actions—repolishing the specimen, increasing the test force, or verifying the measuring system—should be applied before results are reported.
Can Vickers results be reliably converted to Rockwell C values?
Conversions performed with ASTM E140 or ISO 18265 tables are approximations valid only for the specific alloy family and metallurgical condition for which the table was established. Converted values must be labeled as such and cannot substitute for a direct test when a specification calls for a specific scale.
What level of data scatter is acceptable?
There is no single universal limit; acceptance depends on the specification and the uniformity of the material. As a benchmark, well-prepared homogeneous steel tested at HV30 commonly shows a coefficient of variation below 2–3 %, while heterogeneous materials such as cast iron or weld zones naturally exhibit larger scatter that should be reflected in the reported range and uncertainty.
How is measurement uncertainty expressed on a hardness certificate?
The expanded uncertainty U is reported with the hardness value, for example "650 HV10 ± 12 (k = 2)", indicating approximately 95 % confidence. It combines the statistical scatter of the measurements with the calibrated contributions of force, indenter geometry, and length measurement as evaluated under the GUM methodology.
Does dwell time affect the processed result?
Yes. Longer dwell times produce slightly lower hardness readings, particularly in soft metals and polymers. The applied force and dwell time form part of the scale designation and must be recorded and reported so that data sets from different sources remain comparable.