3 Replies Latest reply on Mar 5, 2013 12:04 PM by Jared Conway

Water hammer flow simulation

Hi,

I saw a youtube video where someone simulated the water hammer effect. Can someone please explain to me how I can simulate this. I currently have a car muffler shaped tube with holes inside and i would like to see if pressure is disipated. Thank you.

• Re: Water hammer flow simulation

Flow Sim has the mechanics of solving a water hammer problem: time dependent, compressible water, but this is a very challenging calculation.

I would say that you could do it, but realistically I would try it only if the results that you were after did not have to be very accurate.  Why?  To get acccurate results, since water hammer involves a moving shock, you would need to constantly refine the mesh to capture the effect of the shock that is very rapidly moving down the tube.

Also see a recent reply of mine* where I explain that the time step is dependent on the smallest cell (or element) size, and also in this case the velocity of the fluid is also a factor.  Thus a very long solution time would have to be completed to get any length of physical time solved.

• Re: Water hammer flow simulation

I don't need to results to be very spot on, but I just want to know if my design can reduce the water hammer effect or not. I'm not sure how to setting up flow simulation for it since I haven't really used it before. Parameters that I know for sure is 7 L/min on the inlet and the outlet closes every 857 ms. My main goal is to see that this design is bettter than a straight tube. I tried running a steady state using 7 L/min on inlet and envrionmental pressure on outlet, but I don't see the flow trajectory going out of the midsection and through the holes surrounding. Thanks for your help

• Re: Water hammer flow simulation

Vincent, setting up a steady state is a good first step. Flow trajectories are good for visualizing the flow but they are dependent on the number of particles. Does a cut plot with velocity vectors enabled show flow inside? If it does, you're probably ok. If it doesn't, it might mean your mesh is too coarse and the holes have been omitted or you have geometry blocking the flow path. For the flow path check, make sure there aren't any interferences and use the treat coincidence as interference option to trace where your part is sealed.

A couple things you'll eventually want to look at:

1. as Joe mentioned, a transient study.

2. can you replace the perforated plate with a porous medium?

3. extending the inlet and outlet tubes to move the boundary conditions away from your area of interest