Fluid flow computation based on physical model

Abstract

We propose a fluid flow estimation method based on a physical model; in particular, constraints related to fluid equations are introduced. In previous researches into fluid speed measurement, studies based on numerical analysis and those involving actual measurement using sensors were completely independent of each other. We attempt an integration of these two approaches aiming at better accuracy and reliability. With the proposed method, images of capturing a fluid flow are used as input data, and the velocity field is estimated through the basic gradient equation. With the gradient-based estimation proposed by Horn and Schunck, the basic constraint that describes the relation between the spatiotemporal intensity gradient and the velocity is employed together with a constraint related to spatial smoothness. The latter constraint, however, is too general, and using a constraint related to the fluid equations of specific fluids seems more efficient. In the present study, an incompressible viscous fluid is considered, and fluid- related knowledge is represented by the continuity equation and the Navier–Stokes equations. Such use of object-specific knowledge is expected to yield physically adequate solutions. The efficiency of the proposed method is confirmed by simulations using numerical data and real images for the Karman vortex. © 1999 Scripta Technica, Syst Comp Jpn, 30(4): 41–50, 1999