[mrxD]

hey all
i'm working on matlab code for instationary channel flow with given boundary condition:

Inflow: u(t=0) = 0, v(t=0)=0 and P= f(t) > 1
Walls: u = 0, v = 0
Ouflow: free outflow dudx = 0, dvdx = 0 and P = 1

at the end i need the velocity of u.

i failed to modify the programm "MIT18086" because of the given boundary pressure.

or for the simplicity i could assume that the pressure is linear decreasing from in- to outflow and use the equation du/dx+dv/dy=0 to calculate v. and then insert to the navierstokes and try to calculate that but then i have a problem with v*du/dx and u*du/dy

anyone ideas or solution for this problem?

thanks

[mechiebud]

Originally Posted by mrxD

hey all
i'm working on matlab code for instationary channel flow with given boundary condition:

Inflow: u(t=0) = 0, v(t=0)=0 and P= f(t) > 1
Walls: u = 0, v = 0
Ouflow: free outflow dudx = 0, dvdx = 0 and P = 1

at the end i need the velocity of u.

i failed to modify the programm "MIT18086" because of the given boundary pressure.

or for the simplicity i could assume that the pressure is linear decreasing from in- to outflow and use the equation du/dx+dv/dy=0 to calculate v. and then insert to the navierstokes and try to calculate that but then i have a problem with v*du/dx and u*du/dy

anyone ideas or solution for this problem?

thanks

hey could you please send me your program of channel flow I need it as college assignment.

[Alex C.]

Originally Posted by mechiebud

hey could you please send me your program of channel flow I need it as college assignment.

You will learn a lot more from your assignment if you do it yourself.

[mechiebud]

yes I know that. But since I missed my classes I am unable to start this one. The lid driven cavity and alternate directional implicit problem I solved myself.

[mechiebud]

yes I know that. But since I missed my classes I am unable to start this one. The lid driven cavity and alternate directional implicit problem I solved myself.

[kaya]

Originally Posted by mrxD

hey all
i'm working on matlab code for instationary channel flow with given boundary condition:

Inflow: u(t=0) = 0, v(t=0)=0 and P= f(t) > 1
Walls: u = 0, v = 0
Ouflow: free outflow dudx = 0, dvdx = 0 and P = 1

at the end i need the velocity of u.

i failed to modify the programm "MIT18086" because of the given boundary pressure.

or for the simplicity i could assume that the pressure is linear decreasing from in- to outflow and use the equation du/dx+dv/dy=0 to calculate v. and then insert to the navierstokes and try to calculate that but then i have a problem with v*du/dx and u*du/dy

anyone ideas or solution for this problem?

thanks

Originally Posted by mechiebud

hey could you please send me your program of channel flow I need it as college assignment.

if you guys get stuck somewhere specific on the development satge, I think I can help but I am not sure if I got your problem.

[mechiebud]

Thanks kaya. I am doing a 2D laminar flow on a channel using SMAC. I am unable o initialize the pressure.Could you please guide me through that.

[kaya]

Sorry I can't help you with SMAC since I never worked with that.

I think you're not using MIT18086? According to its manual, it doesn't use SMAC but a projection method for pressure coupling. I think it solves the momentum equations and finds an intermediate velocity ( which doesn't satisfy continuity at this stage). And with a pressure correction ( which is found by taking divergence of this intermediate velocity --> poisson equation) the intermediate velocity is updated to the u^(t+1). I saw that they put the matlab code in the course's webpage.

[kaya]

see the page 4
http://math.mit.edu/~gs/cse/codes/mi...vierstokes.pdf

[mechiebud]

Thanks a lot for your response. Yes I have seen that code but I need to do it using SMAC. Since I am new at CFD I am facing certain certain challenges. Anyways I'll try to put more effort.Thanks again..

[mechiebud]

Hi Kaya, Could you please give e some idea regarding the boundary conditions. Like what pressure and velocity boundary conditions should I use?

[kaya]

maybe no slip at the walls with neumann pressure and periodic at the inlet outlet of the channel ?

[mechiebud]

In periodic boundary what initial values of u,v and p should I give at the left and right boundaries to initiate my simulation?

[kaya]

you can initialize your velo. field with analytical poiseuille flow solution which you can find somewhere in the middle of Frank white's text book. should be located on the top of a left page .. and then you can specify a constant pressure derivative

[ss477]

Were you successful in changing the boundary conditions to achieve chanel flow