CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Question about explicite and implicite time integration theory (http://www.cfd-online.com/Forums/main/106084-question-about-explicite-implicite-time-integration-theory.html)

Anna Tian August 17, 2012 15:56

Question about explicite and implicite time integration theory
 
Hi,

Usually we prefer explicit treatment to the convective term but implicit treatment to the diffusive term. From some papers, I know the reason of this difference is the stability. But I'm wondering how to derive this. How to derive that for stability reason we use different treatment?

FMDenaro August 17, 2012 16:27

Quote:

Originally Posted by Anna Tian (Post 377549)
Hi,

Usually we prefer explicit treatment to the convective term but implicit treatment to the diffusive term. From some papers, I know the reason of this difference is the stability. But I'm wondering how to derive this. How to derive that for stability reason we use different treatment?


The question is more complicate ... the non-linear term, when discretize in time in an implicit way leads to a non-linear algebric system. Then you have to face by using some techniques... (for example you can see the book of Ferziger and Peric).

Generally the diffusive part is linear and is better suited for implicit discretization. Furthermore, the stability constraints are more critical for the the diffusive term (e.g. first order implicit 1D equation has the limit alpha*dt/h^2 < 0.5) than for the linear one. But the Fourier stability analysis give some response for the linear case.

I suggest you to derive the stability regions for the 1D equation du/dt + a du/dx = alpha d2u/dx^2 using implicit and explicit discretizations


All times are GMT -4. The time now is 17:00.