Elastoplastic Finite Element Analysis of Functionally Graded Beams

  • Valencia Murillo, Carlos Enrique (University of Guanajuato)
  • Gutierrez Rivera, Miguel Ernesto (University of Guanajuato)
  • Flores Samano, Nicolas (University of Guanajuato)

Please login to view abstract download link

The use of structures made of complex materials in advanced engineering applications has motivated the development of numerical models for predicting their mechanical behavior and thus optimizing designs at early stages. Such models must be capable of describing behavior under various mechanical load conditions, including linear and non-linear responses. One of these materials, widely studied in recent decades, is the so-called functionally graded material (FGM), which is known for its good performance under high-temperature environments or as an alternative to avoid stress concentration leading to delamination on composite materials [1]. When one of the components of the FGMs is metallic, recalling that FGMs are a heterogenous mixture of constituents, predicting the elastoplastic response of functionally graded structures may become particularly interesting for designers. Furthermore, these structures are often built with several structural elements, such as bars, beams, or shells. Exploiting the advantages of the finite element (FE) method, this contribution presents a FE model for the elastoplastic analysis of functionally graded beams within the framework of small strain plasticity and considering only material nonlinearity. In contrast to previous studies that adopt different beam theories [2,3], the present FE model is based on the displacement field of Timoshenko beam theory, and the Tamura-Tomota-Ozawa (TTO) homogenization scheme is used to compute effective material properties of the metal-ceramic FGM. To emulate elastoplastic behavior, a linear isotropic hardening law is considered, using an associative flow rule and the von Mises yield criterion. When material yielding is reached, the stress states at the integration points are updated using a return mapping algorithm [4]. Moreover, numerical comparisons are performed to verify the present FE model, from which good agreement in the results is found. Finally, various cases of study are analyzed to illustrate the performance of the present FE model.