It is very difficult to characterise the mechanical properties of biomaterials such as plant roots because of their heterogeneous properties, irregular shape and complex cellular structure. Particle image velocimetry (PIV) was evaluated as a novel approach to measure displacement fields along such materials being pulled in tension. PIV uses a cross-correlation technique to measure localised displacement between sequential digital images. The technique was applied to measure local surface strain rates from timelapse images taken during tensile loading. Load measured with a force transducer was converted to localised true tensile stress and strain along test specimens using data from PIV analysis. Tests using translation of a rigid wire were used to evaluate the effects of optical image distortion and assess the errors associated with using PIV for displacement measurements. Use of a digital single-lens reflex camera reduced systematic errors associated with lens distortion 6(.)5-fold compared to a compact digital camera system. Analysis using PIV gave rise to random errors in the strain calculated to be approximately +/- 0.5%. Two materials were subsequently characterised: a homogenous polymer (Vitron rubber) of constant diameter; and the primary root of maize, as a heterogeneous biomaterial. Axial strain and stress distributions along the materials were determined. The radial strains were measured by a separate image analysis method using thresholding. The maize root exhibited necking associated with failure of the stiffer outer sleeve of cortical tissue at axial strains of < 5%. Continued extension transferred the stress to the inner vascular bundle (stele), which separate tests showed failed at strains of about 10%. Mechanical properties varied with age of the tissue, with older cortical tissue becoming 50% stiffer, probably associated with cell wall thickening and deposition of lignin. Poisson's ratios were 0(.)38 for the Vitron rubber specimen, and 0(.)63 for the intact root. (c) 2006 IAgrE. All rights reserved Published by Elsevier Ltd.
|Number of pages||14|
|Publication status||Published - May 2006|