Aerostat envelopes are generally bodies of revolution with length to diameter ratio ranging from 3 to 5. Drag coefficient for this class of bodies can be obtained using empirical formulae or co-relations based on experimental studies. However such formulae are valid for specific class of envelope shapes only and result in errors of around 30% compared to the values determined by numerical methods. The motivation for the present study arose from the need for a simple but generic methodology to

more »

... timate coefficient of drag as a function of envelope geometry, thus eliminating the need for running numerically expensive CFD codes each time the shape is altered during an optimisation exercise. In a previous study, the envelop shape was parameterized in terms of six geometric coefficients, and a shape generation algorithm was developed to generate various possible shapes satisfying manufacturing and few geometric design constraints. An empirical co-relation for envelope drag coefficient was developed, but it was not amenable to coupling with an MDO process, since it required detailed geometric data about the envelope shape, especially the coordinates of several points at the nose and trailing edge, and the grid density in these regions. In the present study, around 600 feasible shapes satisfying the user-specified volume and length constraints were generated using the shape generation algorithm. The flow patterns over these shapes were studied using FLUENT TM CFD Package and a better corelation was obtained by fitting a quadratic response surface using Design-Expert TM package. The current methodology uses only the values of six design variables to determine the drag coefficient, thus making it easy to integrate with a multi-disciplinary optimization algorithm for determining optimum envelope shape. Nomenclature 2 ρ a = Density of air, m 3 /kg ρ he = Density of helium, m 3 /kg σ max = Maximum hoop stress per unit thickness, N/m RSM = Response Surface Methodology I.