Due to high utility of the sandwich structures in the vast spectrum of applications, this research focuses on the practical solution to protect them from damaging consequences of thermal shocks through parametric assessment of its time-variant thermoelastic performance. To this end, current research proposes a novel approach by taking the advantage of coupling the exactitude of the analytical solutions with the agility of the machine-learning methods to predict the accurate results by lower computational efforts. Case study of this research is a sandwich sector plate with functionally-graded carbon-nanotubes reinforced (FG-CNTR) nanocomposite face-layers exposed to the thermal shock loading. As another novelty, this study considers a radial-variant elastic substrate as a mechanism of resistance against the destructive consequences of thermal shock loading. As all machine-learning based regressors require the set of datapoints, this study employs harmonic format of the differential quadrature approach (HDQA) to find the system’s response toward thermal shock in the background of general elasticity theory at definite design- points. Laplace transform integrated with the extended model of Dubner and Abate’s scheme to translate the time-variant response of the system from time to Laplace domain and vice versa. Validity of the solution used in this survey is confirmed through comparing its outcomes with those of the published articles. As a valuable finding, it is strongly suggested to enhance the system’s thermal shock resistance by using the substrates which their reaction’s dependency to the radial coordinate is expressed by higher-order polynomial estimator functions.
Read full abstract