DoE-based modeling of cracked 3D orthotropic panels with stiffeners

Abstract

This study uses the Design of Experiments (DoE) technique to develop analytical formulations for the stress intensity factor and other non-singular variables near the crack front of a stiffener-reinforced three-dimensional orthotropic panel. The method employs a 52 fractional factorial design, effectively reducing the number of finite element simulations required. The input variables consist of four parameters, each with five levels, which directly affect the process responses. The statistical analysis is performed using the Ellistat software, which offers powerful tools for Design of Experiments (DoE) and incorporates specialized algebraic-statistical approaches. These techniques are utilized for analysing the outcomes and formulating a mathematical model of the cracking characteristics for panels that are strengthened with five stiffeners. The findings of the analysis of variance (ANOVA) indicate strong statistical significance for all tested models, hence verifying their validity based on statistical evaluation criteria. In addition, the residual analysis demonstrates a normal distribution, which strengthens the reliability of the underlying models and fulfils a vital requirement for numerous statistical analyses. The analytical formulations that come from this study have enormous potential for practical use in the field of engineering, specifically for structures that are orthotropic and feature cracks. This study signifies a notable progress in integrating numerical simulation, statistical analysis, and mathematical modelling to develop pragmatic and dependable analytical instruments for the examination of intricate systems.