drag calulation over a square cylinder
 

i am using MAC method for incompressible flow in a 3-D channel with a square cylinder in it. the pressure and velocities fields have been computed. how do i calculate the drag on the cylinder ? how to extrapolate the values of pressure etc. on the cylinder surface as the MAC algorithm uses staggered grid? thanking u in advance. sex00

Re: drag calulation over a square cylinder
 

(1). The drag on the cylinder comes from the pressure drag and the viscous drag. (2). Since the pressure gradient normal to a wall is zero, the effect of the exact location where the pressure is evaluated will have little effect on the result. (except around the corner where the pressure variation is large). (3). In any case, if your are interested in the accurate evaluation of the pressure and the tangential viscous stress, then, you should just refine the mesh near the wall. (4). The normal force on the wall will produce the pressure drag, and the tangential viscous stress on the surface will give you the viscous drag.

Re: drag calulation over a square cylinder
 

To get the drag just integrate the pressure along the upstream side (Pa) and on the downstream side (Pb). drag=Pa-Pb. I hope this is correct and our friend John.C.CHien accepts this.If pressure had builded up to larger values normalise it with some pressure at entry.

By the by I also use MAC algorithm.I need to know how u r discretising the convection term asap bye

Re: drag calulation over a square cylinder
 

dear senthil, the convection terms are discretised using a weighted average of second upwinding and space centered scheme (Hirt et al. a numerical solution algorithm for transient fluid flows ,LA-5852 ,Los Alamos Scientific Laboratory Report,1975) however for LES it seems that either pure central differecing or pure upwinding works best!

also, i am interested in knowing how to integrate the pressure (procedure)! and the extrapolation procedure! i too know that i have to integrate the pressure! sex00

Re: drag calulation over a square cylinder
 

Hi, Senthil- could you tell me please few words more about accuracy of this method ? There is similar way for model-researches called Jone's method in which you integrate pressures on plane located in 5-10% of chord-line downstream but tangent stress term and energy disipation are ommited cause of difficulties of measurement shear stresses on whole hull. How does it look like for finite volume method which gives anyhow approximate disribution of pressure in cells but over whole shape ? Is it worth ? Maybe the way you are thinking about is taking into account (implicite way) the shear stresses and they influence on the wake and next on the total forces? Thanks

Michal