Reliability-based optimization of evolutionary topology and automated generation of strut-and-tie models for 3D structures

Resumo

This paper addresses reliability-based topology optimization (RBTO) coupled with the Smoothing-ESO (SESO) method for automated generation of optimal strut-and-tie models. The proposed approach handles with the generation of truss-like designs for three-dimensional problem, addressing the design of a single corbel and a deep beam with two openings. The reliability analysis is performed using the Reliability Index Approach (RIA) via First-Order Reliability Method (FORM) by inserting into the Topology Optimization (TO) approach the follow random variables geometry, volume, strength and compliance, considering the limit state functions maximum displacement and maximum von Mises stress imposed in the optimization procedure. The automatic generation of optimal 3D strut-and-tie models was obtained through the derivatives of the von Mises stress fields by finding the force paths where compression and tensil predominate, respectively, in the direction of the struts and ties for reinforcement insertion.