Inverse design methods for radiative transfer systems

Radiant enclosures used in industrial processes have traditionally been designed by trial-and-error, a technique that usually demands considerable time to find a solution of limited quality. As an alternative, designers have recently adopted optimization and inverse methodologies to solve design problems involving radiative transfer; the optimization methodology solves the inverse problem implicitly by transforming it into a multivariable minimization problem, while the inverse design methodology solves the problem explicitly using regularization. This paper presents the details of both methodologies, and demonstrates them by solving for the optimal heater settings in an industrially relevant radiant enclosure design problem. (C) 2004 Elsevier Ltd. All rights reserved.