Stochastic Optimization of Convective-Fin Design

In the design of convective fins, stochastic variations in fin dimensions have traditionally been handled by the use of safety factors. Often this process results in a multiplication of safety factors and thus an overly expensive design. This paper presents a probabilistic approach that not only analyzes the probability of system failure but also uses this analysis to synthesize the optimal design. Four methods of varying accuracy and difficulty are described and compared. The method based on the RMS approximation for the variances appears to be most useful for design purposes.