How can it b possible? Flow from low p to high p??
 

Maybe there's something I'm doing wrong and I'm oblivious to it, and I'd like to ask if you can spot what my mistake is.

Why is it that when I use the following boundary conditions in a straight 3-D pipe

1) Pressure inlet: 300,000 Pa (total), 268684 Pa (gauge) 2) Pressure outlet: 200,000 Pa

I keep getting reversed flow at both boundaries and there seems to be a continual divergence in my values of pressure to either 0 Pa or 5,000,000 Pa, rather than convergence. I don't understand what error I could have committed!

Please advise if you can.

Thank you very much!

Re: How can it b possible? Flow from low p to high
 

Hi, I have experienced the continuous pressure rise at some times, having a mesh where some nodes are accidently not merged - i.e. the mesh is not continuos. Have you checked that? Which program are you using?

Re: How can it b possible? Flow from low p to high
 

I used Gambit. :( How should I check whether it is continuous or not?

Re: How can it b possible? Flow from low p to high
 

hi in FLUENT, u have to give guage total pressure(static+dynamic)at inlet and staic pressure at outlet. so there is no question of high pressure at outlet, b'cos total pressure at inlet is definitely more than static pressure at outlet. so u just check up this error. otherwise if u know velocity at inlet u give that value.

Re: How can it b possible? Flow from low p to high
 

Start by looking at your (non-converged) solution. Is there some place where the flow stops, or where you have a large pressure gradient in the imposed flow direction? If so, it may be where your mesh is "cracked". Try to use the Connect Faces command to see if there are adjacent faces that do not connect in your mesh.

Re: How can it b possible? Flow from low p to high
 

I have tried connect-faces on all the faces, but Gambit reports "no actual connection done". I suppose this means that the mesh is not cracked.

I've just notied: Does it matter if I clicked on "init" choosing pressure inlet, and I get a negative value for z-velocity? Because I thought this was simply a matter of sign convention...I don't know whether this shows that my flow will definitely be reversed though...