Hardness testing is crucial for evaluating the mechanical properties of ceramics, especially in applications involving wear and scratch resistance. The widely used Vickers hardness test is often affected by an indentation size effect (ISE), resulting in hardness values deviating from their true values under lower loads. Moreover, significant discrepancies in hardness values obtained between the Vickers hardness testers and micro-/nano-indentation instruments at low loads for identical samples persistently cause a confusion. The conventional theories presume that the ceramic materials' indentation dimensions remain unaltered before and after unloading. However, these dimensions often exhibit variations in practice. The existing models and methodologies fail to universally accommodate diverse materials, making it essential to clarify these issues for the accurate assessment of ceramic material hardness. This study was to explore the origins of the Indentation Size Effect (ISE) and the patterns of indentation deformation via comparing the Vickers hardness data with micro-/nano-indentation results under varying loads, thereby furnishing a theoretical foundation for the standardization of ceramic hardness testing.

Three representative ceramic materials (i.e., Al₂O₃, ZrO₂, and (Hf_0.5Zr_0.5)C) were investigated to evaluate their load-dependent mechanical responses. After being sequentially sectioned, ground, and polished, the specimens were subjected to a series of Vickers hardness tests at different loads. The Vickers hardness values were determined via optically measuring the diagonal dimensions of residual indentations using a metallographic microscope. Simultaneously, micro-/nano-indentation tests were performed by a nanoindenter equipped with a Vickers indenter. The pre-unloading hardness values were calculated based on the changes in depth after loading. Subsequently, the post-unloading hardness values were obtained via measuring the indentation sizes after unloading using scanning electron microscopy (SEM). Through a comprehensive analysis of the variations in indentation sizes before and after unloading, this study could elucidate the characteristics of load-dependent mechanical responses in those ceramic materials.

The three ceramic specimens display pronounced load-dependent characteristics according to the Vickers hardness data. Three specimens exhibit significant hardness declines in a low-load regime (i.e., decreased by 3 4.44% for Al₂O₃, decreased by 22.85% for ZrO₂, and decreased by 29.44% for (Hf_0.5Zr_0.5)C), and stabilize at characteristic plateau values (i.e., 15.6 GPa, 12.9 GPa, and 18.6 GPa, respectively) beyond critical threshold loads (i.e., 3-5 N) as the load increases. This behavior is attributed to the predominance of elastic recovery mechanisms during low-load indentation, which results in measured values that exceed the true hardness of the materials. As the load further increases, the plastic deformation becomes a dominant mechanism, leading to stabilized indentation dimensions and hardness values that more accurately reflect the intrinsic properties of the ceramics.

A comparative study is conducted to evaluate the pre-unloading hardness values derived from depth changes during loading and the post-unloading hardness values calculated from residual indentation dimensions as using the micro-/nano-indentation system. This investigation reveals three distinct types of indentation behavior among the materials tested. For Al₂O₃, the Vickers hardness tester obtains significantly higher results than the micro-/nano-indentation tester at low loads. This discrepancy is since the indentation diagonal exhibits a rebound pattern after unloading, resulting in a minimal residual deformation and consequently an overestimated hardness value. In contrast, ZrO₂ demonstrates the opposite trend. Ceramic materials exhibit a localized microplastic deformation at indentation edges. For ZrO₂ under low loads, the phase transformation-induced expansion at an indentation periphery creates an extrusion effect, increasing residual indentation dimensions and resulting in extrusion-type morphology. The results of Vickers hardness values measured by the tester that are lower than those obtained from micro-/nano-indentation tests. For (Hf_0.5Zr_0.5)C, a stress is alleviated due to the formation of micro-cracks around the indentation. Since there is no significant change in indentation size after unloading, the testing methods both yield comparable hardness results.

A further analysis indicates that minor deviations in indentation size (Δd) have a significant effect on the low-load hardness calculations. For instance, in the case of ZrO₂ at a load of 0.01 N, a small Δd of only 0.24 μm can lead to a relative error of 20%, due to the extremely small indentation size of 1.20 μm, which results in a substantial 57% deviation in hardness measurements. The critical load is the threshold above, which the plastic deformation dominates and elastic recovery is negligible, leading to stable test results. The differences between the results by the Vickers hardness testers and by the micro-/nano-indentation testers can be attributed to several factors, i.e., a) indenter precision: Nano-indenters have sharper tips, allowing for more accurate contact area calculations at low loads; b) load range: Nano-indenters are designed for ultralow loads (< 1 N), but are more prone to ISE. The Vickers testers require loads of 5 N or greater to minimize errors; and c) test objective: Nano-indentation primarily reflects the properties of individual grains, while Vickers testing reflects a combined response of multiple grains and their grain boundaries. These factors contribute to the differences in hardness results obtained from the two testing methods, highlighting the importance of choosing an appropriate technique based on the specific material characteristics and testing goals.

For reliable mechanical characterization of advanced ceramics, an empirical minimum applied load of 5 N was established as a critical threshold in Vickers hardness testing. This threshold was essential for eliminating load-dependent artifacts and ensuring the reproducibility of measurements. This protocol effectively suppressed elastic recovery contributions to indentation work, obtaining the hardness values that accurately reflected a bulk plastic deformation behavior. Three distinct indentation morphologies were categorized based on the systematic analysis of dimensional recovery characteristics, a) stable type: This could align with conventional assumptions regarding indentation behavior; b) rebound type: This morphology resulted in higher Vickers hardness values due to the rebound effect during unloading; and c) extrusion type: This morphology resulted in lower Vickers hardness values. At low loads, elastic deformation could dominate, while plastic deformation could remain minimal. This scenario caused the calculated hardness that was greater than the true value. As the load further increased, the irreversible residual indentation size became significantly larger than the minor elastic recovery deformation, leading to a stabilization of the calculated hardness value. This phenomenon could be a primary reason for the load effects obtained in hardness testing.