Abstract
Topology Optimization (TO) aims to determine the optimal material distribution within a structural domain to satisfy predefined performance criteria. In its multi-material formulation, TO combines different mechanical properties and costs, resulting in lighter, stiffer, and more economically efficient structures. This work presents a unified Multi-material Topology Optimization (MMTO) framework based on the Smoothing Evolutionary Structural Optimization (SESO) method, implemented in MATLAB and coupled with ABAQUS. The framework enables the analysis of complex structures under realistic boundary conditions. Uncertainties in loading, elastic modulus, and volume fraction are incorporated through the First-Order Reliability Method (FORM), considering normal and lognormal random variables and a reference reliability index of β=3.0. The main contribution is the identification and calibration of a compliance amplification coefficient (δ=1.77), establishing a deterministic–reliability equivalence. This equivalence enables reliable compliance to be estimated directly from deterministic optimization, with deviations below 5% relative to explicitly reliability-based solutions, as validated through Monte Carlo simulation. The framework was verified using benchmark problems and applied to a hydraulic transmission test-bench adapter support with four cast irons and a long-span bridge. The results demonstrate reductions in structural weight, compliance, material cost, and computational effort while maintaining target reliability.

This publication has 22 references indexed in Scilit: