Experimental method to account for structural compliance in nanoindentation measurements

The standard Oliver-Pharr nanoindentation analysis tacitly assumes that the specimen is structurally rigid and that it is both semi-infinite and homogeneous. Many specimens violate these assumptions. We show that when the specimen flexes or possesses heterogeneities, such as free edges or interfaces between regions of different properties, artifacts arise in the standard analysis that affect the measurement of hardness and modulus. The origin of these artifacts is a structural compliance (C-s), which adds to the machine compliance (C-m), but unlike the latter, C-s can vary as a function of position within the specimen. We have developed an experimental approach to isolate and remove C-s. The utility of the method is demonstrated using specimens including (i) a silicon beam, which flexes because it is supported only at the ends, (ii) sites near the free edge of a fused silica calibration standard, (iii) the tracheid walls in unembedded loblolly pine (Pinus taeda), and (iv) the polypropylene matrix in a polypropylene-wood composite.