Caveats concerning conjugate stress and strain measures for frame indifferent anisotropic elasticity (Q1267112)

The paper addresses the problem of formulating correctly invariant large-strain constitutive equations for anisotropic elasticity with a view to predict magnitude and directions of stress, even in the case of substantial material distortion. The author shows that it is not simple to write down a generalization of the small-distortion law via constitutive equations in the unrotated material frame or by using polar rates (so-called Jaumann time derivatives). Due attention should be paid to the duality between strain and stress measures. This is demonstrated on a worked-out example that considers the stress solution in an idealized fiber-reinforced composite (described in the manner of A. J. M. Spencer). The Cauchy tangent stiffness (linear relationship between rates of strain and stress) must distort the process, i.e. microstructure does affect macrostructure. A safe solution would obviously consist in working in convected coordinates, i.e. in using the second (symmetric) Piola-Kirchhoff stress together with its thermodynamical dual, the Lagrangian strain. However, one must also realize that such an approach requires accurate measurements of the large-distorsion material response in order to avoid potential instabilities in numerical implementations. The paper contains many interesting remarks.