Material Behavior Modeling
Material behavior and its incorporation into Finite Element (FE) constitutive models is often the most important aspect of any analysis, especially nonlinear predictive/failure analysis. One of the common threads that link ANATECH’s diverse business areas is material modeling. Useful analysis requires accurate predictions, which require advance material models which accurately model the stress-strain behavior of complex materials.
Most FE programs include a large library of material models, many of which are advanced nonlinear models. For example, elastic-plastic material behavior with combined isotropic and kinematic hardening. While classic plasticity theory provides an excellent approximation to the behavior of metals that are monotonically loaded at room temperature, these same models offer poor predictions of the cyclic behavior of both material and structural models. Also, many of material models available in general purpose FE programs do not allow the treatment of strain softening. This deficiency is a major reason why most general purpose FE programs fail to provide accurate or reasonable predictions of the observed response of concrete structures to cyclic loading.
Most general purpose FE programs allow the user access to a “user-provided constitutive model” where the user typically provides a FORTRAN coded subroutine that defines the stress-strain behavior of the material. ANATECH has developed and coded many such user material models. Such modeling requires advanced training and considerable skill in both mechanics and computer programming. Few FE program users have the skill or experience to develop such software. Most users are content to limit their consideration to those few material models generally available. Unfortunately, this limits the accuracy and value of their analytical results.
The following links describes some of the material models developed by ANATECH. In all of these material models, ANATECH took an empirical approach with the “Form Follows Function” philosophy in order that these models provide “engineering,” that is, useful solutions as opposed to simply programming an “established” theory.
