Introduction T basic hypothesis in the theory of plasticity is that there exists a scalar function of stresses which characterizes the yielding of materials. This scalar function, also called the yield function,/(crfj.) generates a closed surface in the stress space by a relation /(o^) = 0. When a material is subjected to increasing forces and torques, it describes a path in the stress space. When this path intersects the yield surface /(afj-) = 0, the material yields and becomes plastic. All the classical failure theories, e.g., those of Tresca, Von Mises, and Reuss, give conflicting predictions of the yield stresses of metals even for such simple cases of pure torsion or elongation. The reason for the conflicts in these theories is that these theories essentially form one-parameter yield surfaces. So fitting these criteria on tensile tests alone make these incompatible with the torsion test data, and vice versa. These theories also do not predict the failure under hydrostatic compression. Detailed references on these theories can be obtained in Ref. 1. The proposed theory, on the other hand, forms a multiparameter yield surface. Here a start is made from the basic yield point data from both tensile and torsion test experiments, as well as the hydrostatic compression test. Thus the present theory can predict yielding under hydrostatic pressure while posing no conflict between theory and, at least, the basic tension and torsion experiments. The formulation is flexible enough to satisfy a large number of the available experimental data on the yield point and subsequent flow for ductile materials, and includes both ideally plastic and strain-hardening solids. The yield surface generated is invariant with respect to rotation of coordinate axes, a necessary requirement for all such criteria.
Read full abstract