Hi,

I read the paper of A.G. Kravchenko. Journal of of Compt. Phys. 1996 412-423. Until now, I cannot generalize out the advantage and disadvantage of the B-spline Mehtod.

Anybody can compare this method with Finite Volume Method briefly here.

Also, I do need Ref. 18 ( M.R. Malik, T.A.Zang)'s Paper 1985, J. Comput. Phys. 61. Can you do me a favor?

Many thanks.

peter.

Hi Peter,

In B-splines, the knot functions have compact support as compared to the spectral basis functions (sine/cosine for fourier spectral schemes or chebyshev polynomials for non-periodic domains). These schemes have high spectral like resolution like compact or Pade schemes. More over the flexibility of zonal embedded computational domains can be achieved readily.

For complex geometry DNS/LES, such a scheme can be more useful than standard spectral method.

Hope this helps Mayank

Thanks Mayank!

I read much those days about Dynamic Subgrid-Scale Turbulence Models ( Beaudan & Moin). Literatures that I have compared much about SGS model with Fixed-coffiecient Smagorinsky model n' without model, (2D and 3D; Numerical dissipation analysis). (using Upwind-biased, 5th order n' 7th order).

Then, I went back again to A.G. Kravchenko n' Moin 1999

"Numerical studies of flow over a circular cylinder at Re=3900". I found what I need