ABSTRACT
This paper presents plate and shell models for multifield problems and proposes methodologies to refine structural models according to given accuracy and computational cost requirements. In multifield problems for multilayered structures, refined models are necessary to deal with many non-classical effects due, for instance, to the presence of large variations of properties among layers. In such research scenario, the Carrera Unified Formulation (CUF) is a well-established framework. Via the CUF, the 3D structural problem is reduced to a 2D or 1D variant. In other words, the 3D unknown variables become 2D or 1D, and expansion functions along the thickness or the cross-section of the structure define the order of the model, or computational cost and its accuracy. The CUF models proved to be able to detect 3D-like accuracies in multifield structural problems with very low computational costs. In the CUF framework, the axiomatic-asymptotic method (AAM) has been recently proposed by the authors to investigate the influence of each unknown variable on the solution of a given problem. Additionally, using the AAM, Best Theory Diagrams (BTD) have been obtained to read the minimum number of terms of a refined model for a given accuracy. The BTD generates guidelines to develop and evaluate structural models. In other words, via the BTD, a trade-off between accuracy and computational cost can be made. In this paper, mechanical, thermal and electrical fields are considered and BTDs are presented for various problems.
