THERMAL CONVECTION: AN EXTENDED MacCREADY THEORY AND A SIMUTATION TOOL

Abstract

Classical MacCready theory is based on a deterministic model. Due to uncertain information during a cross country flight this seems to be rather questionable. The present paper deals with stochastic models which are more appropriate to describe realistic convectional pattems. In a first step the classical theory is extended to the case that the intensity of the next used lift can be characterized bya random variable A. The optimal speed ring setting leading to minimal expected time then is l/E(l/A), where E denotes the expectation. As a full stochastic model, including both random intensities of ups and downs and their extension, we suggest a Markov jump process. An optimal strategy is unknown for this model. However, a simulation tool has been developed to compare the efficiency of different tactics under probabilistic convection profiles of the above type.