The animation and visual effects industry is producing advanced software capable of generating realistic behaviors faster than ever by using algorithms that approximate the physics of the real world. There is an opportunity to utilize these software to support performance-based conceptual design for complex simulations such as Computational Fluid Dynamics (CFD). This paper investigates a method of performing wind flow simulation using an animation software that implements an Eulerian based smoke solver. These simulations run orders of magnitude faster than the similar simulations in dedicated high-end CFD applications. The paper compares the animated simulation results to a benchmark case with measured wind-tunnel data. The results indicate that at certain points in the animation, the accuracy is very high. However, the challenge lies in predicting the best frame at which to stop the animation. The paper ends with a discussion of how this challenge might be tackled.

View Full Paper