Search Results

A common scenario in engineering design is the evaluation of expensive black-box functions: simulation codes or physical experiments that require long evaluation times and/or significant resources, which results in lengthy and costly design cycles. In the last years, Bayesian optimization has emerged as an efficient alternative to solve expensive black-box function design problems. Bayesian optimization has two main components: a probabilistic surrogate model of the black-box function and an acquisition functions that drives the design process. Successful Bayesian optimization strategies are characterized by accurate surrogate models and well-balanced acquisition functions. The Gaussian process (GP) regression model is arguably the most popular surrogate model in Bayesian optimization due to its flexibility and mathematical tractability. GP regression models are defined by two elements: the mean and covariance functions.

Shell-and-tube heat exchangers, commonly referred to as radiators, are the most prevalent type of heat exchanger within the automotive industry. A pivotal goal for automotive designers is to increase their thermal effectiveness while mitigating pressure drop effects and minimizing the associated costs of design and operation. Their design is a lengthy and intricate process involving the manual creation and refinement of computer-aided design (CAD) models coupled with iterative multi-physics simulations. Consequently, there is a pressing demand for an integrated tool that can automate these discrete steps, yielding a significant enhancement in overall design efficiency. This work aims to introduce an innovative automation tool to streamline the design process, spanning from CAD model generation to identifying optimal design configurations. The proposed methodology is applied explicitly to the context of shell-and-tube heat exchangers, showcasing the tool's efficacy.