Abstract

In the present study, a new optimization method called natural construction (NC) method was presented. This method was inspired by constructal law and it is an effort to extend this theorem by eliminating its predefined constraints. The proposed NC method was implemented on evolutionary optimization of embedded high conductive continuous strips in a cold plate to cool a uniformly heat-generating surface. These strips guide the generated heat to a localized constant temperature heat sink. The proposed method is a straightforward approach with a low computational cost which works based on gradual growth of the constructal cells and unlike the majority of optimization methods, does not need any special assumption about the initial shape of strips configuration or constructal cell size. Moreover, since in this method constructal cells are added gradually, the effect of each element on maximum or mean temperature can be assessed at each step while the proper amount of required area fraction of high conductive material will be evaluated accordingly. The NC method is a general constraint-free optimization method that can apply to a wide variety of engineering problems and will be very useful in different disciplines.

摘要

在本研究中，提出了一种称为自然构造 (NC) 方法的新优化方法。这种方法受到构造律的启发，它试图通过消除其预定义的约束来扩展该定理。所提出的 NC 方法用于对冷板中嵌入的高导电连续带材进行进化优化，以冷却均匀的发热表面。这些条带将产生的热量引导至局部恒温散热器。所提出的方法是一种具有低计算成本的简单方法，它基于构造单元的逐渐增长而工作，并且与大多数优化方法不同，不需要关于条形配置的初始形状或构造单元大小的任何特殊假设。而且，由于在这种方法中构造单元是逐渐添加的，因此可以在每个步骤评估每个元素对最高或平均温度的影响，同时将相应地评估所需的高导电材料面积分数的适当数量。NC 方法是一种通用的无约束优化方法，可以应用于各种各样的工程问题，并且在不同学科中都非常有用。