Abstract:
A genetic algorithm based inverse analysis is done to predict unknown parameters in a trapezoidal extended surface for
satisfying a given temperature distribution. An inverse method is adopted to estimate six unknowns involving thermal,
surface, and geometric parameters, which helps to identify feasible fin materials, necessary dimensions along with other
requirements. Various controlling parameters of genetic algorithm along with random measurement errors have been
investigated. Fin efficiencies have been also compared. For satisfying a prescribed temperature distribution, this study
shows that many feasible materials exist which may satisfy a given temperature profile, which shall be useful in selecting
any material from the available choices depending upon the relevant dimensions, convective and surface requirements.
This study also shows that fin dimensions along with the coefficient of thermal conductivity influence the temperature
distribution more than other parameters. The maximum variation in the efficiency among the predicted parameters is
found to be within 9%.
