Pulsatile Inlet & Opening Outlet
 

Hi everyone

I have a trouble. I run model as a straight tube. When I set the pulsatile inlet and constant pressure outlet, it will be overflow.
So, I change the constant pressure outlet to opening outlet. But the results are really strange. The flow is reverse. How could I get the normal flow direction(In→Out) at any timestep?

Thanks a lot for your help and time.
Best Regards

Attachment 38437

Attachment 38438

Attachment 38439

In the post processor draw a contour of the normal component of velocity on the boundary. Then use it to draw a contour of velocity =0. All the flow in one direction will be one colour, all the flow in the opposite direction will be another colour.

Sorry, you misunderstood what I meant. My problem is how to solve the flow direction correct (In→Out)?
At some timestep result, the flow direction is wrong (Out→In).
Thank you.

If you mean the backflow is wrong, so how to stop the back flow? Well, the backflow is a result of the conditions you apply. If you are applying a pulsing flow then it sounds like backflow is inevitable.

I think you need to go back to basics on this simulation. What are you actually trying to model? What do you know about the flow?

Sorry, I know backflow is inevitable. But as you could see the picture, my inlet is change into outlet. http://www.cfd-online.com/Forums/att...g-outlet-2.jpg
I want to try a pulsatile flow in a stright tube. I setup inlet boundary conditions as a function of timestep. Inlet is always be inlet, but this picture result inlet flow direction is oppsite. I need to get inlet flow direction is a fixed direction, how could I solve this problem.

Thank you in advance.

I think I found the problem. Is my expression too complex?
http://i.imgur.com/vPjTpjh.png
Plot it.
http://i.imgur.com/OQOEyWp.png
I really need this wave of inlet velocity. If my expression is incorrect, please help me.

Many thanks.

Can you post your CCL?

Why do you use Current Time Step (ctstep) instead of time (t)?

I misunderstand ctstep mean...
Thank you all. I use time(t) instead of Current Time Step (ctstep).
And it does work. The flow direction is correct.
Really pretty thanks!!!