Comments on the Mroz multiple surface type plasticity models

Several questions are addressed regarding the role of number of surfaces in the Mroz type multiple surface models. Firstly, increasing the number of surfaces improves the accuracy of the plastic modulus function. but at the same time alters the translation behavior of surfaces. Two 90 out-of-phase axial torsion loading experiments are chosen to illustrate the number of surface influence on the model behavior. The stresses exceed the experimental levels with increasing numbers of surfaces from 5 to 100. Secondly, for proportional loading. the multiple surface models do not predict ratchetting (progressive plastic strain accumulation in one direction). However, these models predict ratchetting for general nonproportional loading. An "ellipse" shaped axial-torsional loading path has been considered where predicted ratchetting rates far exceeded their experimental counterparts. An explanation is forwarded to address these properties of the models. It is further demonstrated that the multiple surface model of Mroz and its modification by Garud produce identical stress-strain predictions when the number of surfaces exceeds a certain value. For infinitesimal loading increment. intersection of surfaces does not occur when using either the Mroz or Garud model, however, when finite loading increment is selected in numerical calculations the intersection problem arises in the Mroz model.