Turbulence research by numerical simulation

Abstract

Our task is to describe recent progress and present prospects for research in turbulence using numerical methods. This is quite a challenge, because we know that Liepmann has not been a strong advocate of computational turbulence, or "compulence," as it has sometimes been called -- the incompressible version presumably being "incompulence." We want to recognize several objectives in exploring turbulence by numerical methods, and to order these in priority as follows: 1. To gain understanding and insight into the physics of turbulence, so as to complement the insight obtained from experiments and analysis; for example, by helping the experimentalist to understand what it is that has been observed experimentally, and by helping the analyst to explore solution space in more detail. 2. To provide special "data" for guiding, evaluating, and calibrating simpler predictive models of turbulent flows. 3. To predict turbulent flows by numerical simulation.