On the global optimum for heat conduction
Yang Liu,Liang Chen,Zhen Li,Jianbin Du
DOI: https://doi.org/10.1016/j.ijheatmasstransfer.2022.123381
IF: 5.2
2022-12-01
International Journal of Heat and Mass Transfer
Abstract:We demonstrate the sufficient and necessary conditions for the global optimum for the heat conduction problem, using the functional variation principle, and prove that the best performance can be guaranteed. Numerical verifications show that no matter how the initial design changes for given boundary conditions, optimized results converge into the solution with the exactly identical objective function. We identify the numerical features for the global optimum, that is, the temperature contours vary linearly, and the temperature gradients are of the same magnitude with directions pointing to the heat sink. We propose an ODE-driven level-set density method for thermal design. Numerical results present hair-like structural morphogenesis with countless regularly curving fibers forming an independent rather than hierarchical system. However, countless hair-like structures are manufacturing unfriendly for practical engineering applications due to their topological complexity and length scales. To address this issue, we introduce a simple numerical strategy through which the optimized solutions become remarkably simplified without much loss of performance. The present findings are theoretically meaningful, also, the method employed herein can be useful in practical engineering.
engineering, mechanical,thermodynamics,mechanics