September 23, 2010, 22:32
|
First Order Upwind X High-Order Upwind (CUBISTA) Schemes
|
#1
|
New Member
Vinicius
Join Date: Sep 2010
Posts: 1
Rep Power: 0
|
Hello,
I'm doing a software that evaluates the numerical solution of Navier-Stokes 2D. I'm using Sola Method's and when I was going to compare the First-Order Upwind (FOU) with the CUBISTA (A High-Order Upwind Scheme), used in the non-linear convective terms, some difficulties appeared. In the FOU scheme, the convergence is achieved normally, i. e., the residue go down to 10^-7. However, in the CUBISTA scheme, the numerical solution does not converges: the residue is oscillating (1 ~ 10^-1)! The same happens with the dt (time step), that is oscillating.
When I look in the simulation, everything is normal with the flow and apparently does not exists unstability.
What can it be? It's possible to compare FOU with CUBISTA (or also QUICKEST) using this strategy?
I'm using dt = 0.001, dx (step line-x) = 0.01, dy (step line-y) = 0.01.
Thanks!
|
|
|