DYNAMIC ELASTIC-ANISOTROPY AND NONLINEARITY IN WOOD AND ROCK

Ultrasonic techniques are used to characterize the anisotropic and non linear elastic behavior of wood and rock and to interrogate the struct ural properties of these materials. For elastic anisotropy two types o f experiments are performed, namely qP-wave velocity measurements on s pherical samples in about 100 directions of propagation rather regular ly sampled in space, together with S-wave birefringence measurements i n a few directions on additional samples. For nonlinear elasticity, ac oustoelastic experiments were conducted, consisting in P and S-wave ve locity measurements under controlled confining pressure. The experimen tal results show that wood exhibits much larger elastic anisotropy,but much weaker elastic nonlinearity than rock. For instance, the deviati ons from isotropy in wood can reach 70%, whereas in rock, typically, i t can hardly exceed 20%. In contrast, regarding nonlinearity the incre ase of P or S-wave moduli per unit confining pressure in wood is alway s smaller than 30, in the radial direction, or 10, in the longitudinal and transversal directions, whereas it can reach roughly one to a few hundreds in rock. These contrasted behaviors can be simply explained by structural considerations. Thus, the exceptionally strong elastic a nisotropy of wood is due to the strict structural alignment of its con stituents, that is to say to the preferential orientation of the anato mical elements (tracheids, fibers, ray cells, vessels etc.) for 'textu ral' anisotropy, and to the cellular wall organisation for 'microstruc tural' anisotropy. In comparison, rock only exhibits a rough statistic s of the orientation distribution function of its constituents, mainly the grain minerals, the pores and the cements. In contrast, the strin kingly strong nonlinear elastic response of rock, a well-established c lassical observation, is due to the presence of compliant mechanical d efects (cracks, microfractures, grain-joints etc.). Such features are practically nonexistent in wood which explains its weak nonlinear resp onse. (C) 1998 Elsevier Science B.V. All rights reserved.