上海科技大学知识管理系统

Inverse problems arise in a wide range of practical systems and processes and, thus, form an important and well-known area of research in mathematics and physical sciences. The basic nature of an inverse problem is mathematically ill-posed and, as such, finding a unique, accurate, solution is challenging. A prime example of inverse problems in mechanical engineering is the inverse heat transfer problem where the desired solution is known and the boundary and initial conditions, or even material properties, are not. While the heat conduction problem has attracted the bulk of research due to its generally linear governing physics, inverse heat convection presents a greater challenge. The inclusion of fluid mechanics and the nonlinear nature of momentum and energy equations, combined with the previously mentioned ill-posed nature of the inverse problem, presents a challenging research topic. In this paper, a search-based inverse heat convection methodology, developed earlier, is tested against experimental data. The problem considered is a wall plume flow generated by a heat source, such as an electronic component or a fire, located on a vertical surface. This method shows that by measuring the temperatures at a limited number of points downstream of the heat source, the heat input and location of the source can be estimated within an acceptable range of accuracy. This paper continues the earlier work by including experimental data in the prediction of heat source strength and location. It is shown that the methodology introduced earlier is quite robust and can be used for a variety of important practical applications. © 2022 Elsevier Ltd