Reliability-based robust design optimization of polymer nanocomposites to enhance percolated electrical conductivity considering correlated input variables using multivariate distributions

In this study, reliability-based robust design optimization (RBRDO) for polymer nanocomposites (PNCs) design is conducted to secure the reliability of design conditions concerning the electrical percolation threshold and carbon nanotube (CNT) aspect ratio as well as the robustness for electrical conductivity. CNT diameter and length are known as following the lognormal and Weibull distributions respectively. To reflect the different probability distributions of CNT geometry parameters and correlations between these random input variables, Nataf transformation is employed. By performing several case studies with the first-order reliability method (FORM)-based RBRDO approaches, the objective function exhibited a noteworthy change according to the correlation coefficients and the reliability and robustness for PNCs were satisfied concurrently. Furthermore, the highlight of this work is to provide a generic framework for practical PNC design and multi-objective optimization to enhance electrical performance efficiently. [Display omitted] •In-house code is developed for percolation simulation of polymer nanocomposites.•The effect of tunneling resistance is considered in developed percolation model.•Reliability-based robust design optimization carried out with Nataf transformation.•Our design accounts for reliability and robustness in electrical performance.•Our generic framework provides informed guidelines towards better PNC design.