We are interested in using a momentum source to simulate a pump in a time dependent flow system. It looks the information in the manual does not serve our purpose as it is not making any difference to the velocity !

How can a momentum source be implemented? The option in the 1. UDF is one possibility I believe. How involved is this option?

2. Is there another simpler method?

Thanks

HDZ

Go to boundary conditions. Select your fluid zone. Select 'Source terms'. Define your momentum source terms in the different directions. Done.

I have done that. I have defined a limited volume of a long pipe where the momentum source has been applied but I did not get any difference.

A long pipe with an inlet velocity of say 2 m/s is set. over a certain volume of this pipe a momentum source is set in the direction,as you have suggested but somehow no difference was noticed.

That is why I equired about UDF?

Have you tried the method you suggested? Are we doing something silly there?

I just tried it, a pipe with 2 m/s water flow and a momentum source in the middle. It worked just fine, I got a nice pressure jump.

Don't forget that the momentum source is in N/m3, so if your zone is small the source term becomes pretty big.

Thank you very much. I will try it. I was aware of the units.

I assume that you have seen an increase in velocity aftter the momentum source, i.e. more than 2 m/s. Please confirm.

Many thanks

HDZ

You would only see a velocity increase if the static pressure decreases a lot. The resulting density reduction would lead to a velocity increase (in compressible flow). However, I 1) used water and 2) I modeled a momentum source (not sink) so the pressure across the zone will rise. Two reasons not to see a velocity increase.

It all makes sense now.

However it turned out that the problem we are trying to solve is a little different.

Now what would happen if you apply that momentum source to a fluid at rest? Will it give it some velocity similar to a pump?

The problem at hand is simply recirculating flow to a tank.

I guess so. Flow is a result of pressure gradients, so a local increase in static pressure should result in a velocity in the direction of the lower pressure.