抄録
In order to reduce the crosswind sensitivity of trains, we developed an optimization method for roof shapes with the aim of reducing the side force acting on the vehicle on various wayside structures. Our method consists of an evolutionary algorithm and a flow simulation. First, initial shapes are given randomly. Then the flow fields around the train on the each wayside structure are calculated and the aerodynamic force of the vehicle is evaluated. Next, new possible designing variables are set by the evolutionary algorithm, before returning the flow simulation stage. The process is repeated for a given number of generations. Cross-sectional shapes of the roof are designed by two parameters, radii of curvature of the roof and the shoulder sections. The side force is employed as an objective function. At the flow simulation stage, unsteady three-dimensional incompressible flow simulations around a train on each wayside structure are conducted, employing the Navier-Stokes equation as a governing equation. A 10th generation evolutionary calculation with eight individuals was conducted for a train without wayside structures on a trial basis, and we obtained the following results. The side force decreases as the radius of the curvature of the shoulder section increases. In the case of the minimum side force, the flow separation around the shoulder section is suppressed, and it results in weakening the vortex and reducing the pressure drop behind the train.
