[FreeFall79]

Hi all

I need to be saved from throwing the workstation out the window... Joking but really frustrated with trying to get my BCs right. I am getting a converged solution but only when artificial walls are placed at my outlets.

I have a pic of my monitor points and they converge to the same value. I start with no pressure drop with the intent of lowering it and ideally getting the walls to drop down as well. I begin entering in negative pressures but the walls stay at 100% for face and area. I use a global initialization of zero to start since I am just trying to get a base point that makes sense.

My reference pressure is zero and my relative pressure set at -1. I should have vacuums pulling the fluid out of my outlets not stuff rushing in. To make convergence easier I have a slow mass flow rate of 0.1 kg going through a rather large volume.

I have consulted the modeling guide but that was not helpful. I have the most robust settings, but its giving me a non physical answer.

any ideas on where I could look?

[FreeFall79]

update

I tried to really perturb the model and but in large pressure drops at the boundaries(-100 psi relative pressure). It still will match up the pressures. I am absolutely perplexed.

[ghorrocks]

Putting some iron bars on the window will save you from throwing the PC out the window.

You need to set your reference pressure to be a representative pressure in your domain. Then the pressures are re-cast relative to that. You may well be getting a pressure drop between your points, you just can't see it in the huge scale of your diagram.

Please post an image of what you are modelling, the location of the monitor points and what the boundary conditions are. Also your CCL would help.

[FreeFall79]

ghorrocks you are a saint!

I have uploaded as you requested. I forgot to change the scale but even though its log scaled those pressures are all the same. I zoomed in to confirm.

My model is going to capture steamflow around this airsolid, entering form the bottom of the green highlighted region and exiting from 2 outlets (the tube protruding from the main body or the thin sleeve exiting from the bottom.

My approach was to start with zero reference and just small pressure drops at the outlet. I would bring the pressure up to my target followed by increasing the flow. Lastly I would raise up the gravity to where it needs to be and save. Then I could go back and use this as my initialization case for more accurate models like SST for turbulence model etc....

As I stated earlier, the outlets keep having flow enter into them despite all efforts so I can never really get started. Model behaves if there is a single inlet but I tried then switching on the other outlet and it goes wacky.

Let me know if you need anything else and thanks again.

[ghorrocks]

Are you sure your interfaces connect up correctly? Can you post the velocity vectors you are getting.

Also you are using local time scale factor. Only use this if convergence is difficult to achieve and if you have to use it always run some time steps as physical time scale afterwards in the final run to convergence.

You are also using upwind for advection. This is fine for debugging, but you will need to use a more accurate scheme for the real simulation.