Toward accurate evaluation of bulk hardness from nanoindentation testing at low indent depths.

[Display omitted] • Nix-Gao model combined with pileup corrections are critical to improve the accuracy of bulk equivalent hardness from nanoindentation testing. • Evaluating bulk hardness by hardness ratio or changes at a reference depth will cause quantitative errors as large as 60%. • Constant strain rate and constant loading rate tests result in comparable hardness for materials with a weak strain rate sensitivity. • The curvature in Nix-Gao fitting is not simply caused by nanoindentation testing procedures. • Size dependent dislocation obstacle strengths are essential for accurate microstructure- and nanoindentation-predicted strength change. Estimations of bulk hardness from nanoindentation are frequently subject to considerable uncertainties due to indentation size effects (ISE), pileup effects, and potential influence of surface quality or test methods. This study examined materials science principles of nanoindentation test methods to enable accurate prediction of bulk hardness for a series of high purity Fe and Fe-(3–25 wt%) Cr alloys. These materials were tested in as-annealed and thermally aged (100–900 h at 475 ℃ to produce Cr-rich α' precipitates) conditions. Nanoindentation with a Berkovich indenter at constant strain rate (0.05–0.5 /s) and constant loading rate conditions provided comparable bulk equivalent hardness (H 0) extracted by Nix-Gao model, indicating a weak strain rate sensitivity at room temperature. Results from electropolished and fine mechanically polished samples gave comparable measured hardness. Pileup corrections produced a 5–14% correction to H 0 which agreed with the experimental bulk Vickers hardness within ∼10% for most tested materials. The microstructural model-predicted and measured strength values agreed for aged samples. A derived analytic expression demonstrates that an ISE error, associated with inappropriate methods such as hardness ratios or changes at a reference depth, would be as large as 60% in estimated bulk hardness for the investigated Fe-Cr alloys. [ABSTRACT FROM AUTHOR]