The influence of the ball material on the Brinell hardness

A Brinell hardness tester with carbide and steel balls was calibrated and thereafter the hardness of ten samples of steel STN 41 1600 was measured by four appraisers. The uncertainty of calibration and of the hardness measurement were calculated. The results obtained with carbide and steel balls were confronted using uncertainty analysis, measurement systems analysis (MSA), analysis of variance (ANOVA), unpaired t-test and Z-score. The calibrated hardness tester is not considered satisfactory because the values of relative expanded uncertainty of calibration (relative maximum permissible error) and maximum error exceed the values permitted by the standard for both balls. The statistically significant difference between the uncertainties and relative maximum error obtained by carbide and steel balls was not confirmed neither at calibration nor at measurement of steel sample hardness. Measurement system analysis confirmed some influence of used ball materials on the results of individual appraisers on the capability of the hardness measurement process (indices %AV, %GRR and ndc) in accordance with the results of the unpaired t-test.