[zrjtju]

hi everybody
I am wondering whats the difference between laminar simulation and DNS. what i cannt understand is that when we simulate laminar flow, certainly there is no turbulent model used. does that mean that we solve the NS equation directly? or some special methods are used to prevent small perturbation to grow?
I am simulate laminar flow using code developed by some individual. my problem is that if I dont use any turbulent model, the flow field eventual become messy. the residual would grow up as well. so can somebody please tell me the difference between them and recommend me some papers or books
all the best

[FMDenaro]

Originally Posted by zrjtju

hi everybody
I am wondering whats the difference between laminar simulation and DNS. what i cannt understand is that when we simulate laminar flow, certainly there is no turbulent model used. does that mean that we solve the NS equation directly? or some special methods are used to prevent small perturbation to grow?
I am simulate laminar flow using code developed by some individual. my problem is that if I dont use any turbulent model, the flow field eventual become messy. the residual would grow up as well. so can somebody please tell me the difference between them and recommend me some papers or books
all the best

DNS is a simulation performed by solving NS equation without any turbulence model. It requires to use grids fine to get the smallest structures of the flow. An CFD code can formally run in DNS mode.
Laminar flow is only a particular regime of the flow, which is, of course, solved in DNS mode.
In your case if you have laminar flow and see residual growing, that sounds more a numerical instability

[zrjtju]

hi
thanks，very much. that's very helpful. Can i understand it this way, as long as my numerical schemes are stable, whether it's laminar or DNS depends on my grid? of course the boundary condition and the accuracy of my schemes are taken into account.

[FMDenaro]

however, laminar flows are always solved in DNS formulation ...

[Simbelmynë]

1. Laminar flow with large grid spacing will probably give you a solution that is laminar, even if there is a possibility for turbulence.

2. I think it is important to make a distinction between a laminar solution and a laminar solver.

3. A DNS resolves all temporal and spatial scales of turbulence. Given that the flow is known to be laminar then a laminar solver can be said to be DNS since you do not need to resolve any turbulence at all.

[Cesar Cabrera]

Hello!

I am trying to simulate the jet from a laminar flow pipe. I have not been able to model the geometry to define the two phases of fluids, water and air

Can anybody help me?

thanks in advance

César