|
[Sponsors] |
![]() |
![]() |
#1 |
New Member
Join Date: Jan 2024
Posts: 23
Rep Power: 2 ![]() |
Hello,
I am simulating a 2d room K-omega with simple foam, with one inlet source in the middle of the room. in the results, i see some of the air leaves the flow and go towards the walls, although walls are far away from the inlet source. what i want to do that i don't want the walls to have any influence on the simulation. I tried different BC for walls like slip and fixed value (0 0 0 ) but no changes as well. Also, same happens with outlet some of air leave the flow and go towards the outlet again. I want to see the flow coming out of air source without any part of air attracted by any other boundary conditions. any way to solve these issues ? |
|
![]() |
![]() |
![]() |
![]() |
#2 |
Senior Member
M
Join Date: Dec 2017
Posts: 692
Rep Power: 12 ![]() |
"Air leaves the flow"? What exactly are you trying to do here? A explaining sketch of your model would help.
|
|
![]() |
![]() |
![]() |
![]() |
#3 |
New Member
Join Date: Jan 2024
Posts: 23
Rep Power: 2 ![]() |
the attached pic, left side of the pic shows what is happening right now in the simulation.
on the right side that's what i want to have, that the air should leave the inlet in it's direction without being interrupted by walls nor outlet. I want it to go directly without any interruptions until it reach the floor. |
|
![]() |
![]() |
![]() |
![]() |
#4 |
Senior Member
M
Join Date: Dec 2017
Posts: 692
Rep Power: 12 ![]() |
The flow moves according to the laws of conservation of energy, momentum and mass. Assuming your model is correctly implemented you can see from the results, that you either have to increase the initial velocity (i.e. decreasing the inlet cross section) or produce more volume flow at the same cross-section. This way the bulk flow will reach the floor.
What you are currently doing is messing with a correct model (wall = no-slip BC) to produce some (probably wrong) imagination you have in mind for what the flow "has to do". Where do you think the flow is going if not to the outlet? The reason we do CFD is to see how the flow actually behaves and not to produce what we think it is doing. |
|
![]() |
![]() |
![]() |
![]() |
#5 |
New Member
Join Date: Jan 2024
Posts: 23
Rep Power: 2 ![]() |
I do know, that what i am trying to do is not related to reality at all, but for some research and presentations needs i want to do it the way i explained above.
the goal is not reaching the floor because i already reached the floor in my case I am trying to stop any influence from outlet or walls on my air flow. |
|
![]() |
![]() |
![]() |
![]() |
#6 |
Senior Member
Yann
Join Date: Apr 2012
Location: France
Posts: 1,169
Rep Power: 27 ![]() |
I don't kow if I properly understood, but it looks like you need to define your walls as patch, with the proper boundary conditions on it. For instance fixedValue on p and something like pressureInletOutletVelocity on U.
As long as you have a wall (aka solid surface) with a slip or noSlip condition on it, it will behave as a wall, preventing the flow to get out of the domain. |
|
![]() |
![]() |
![]() |
![]() |
#7 |
Senior Member
M
Join Date: Dec 2017
Posts: 692
Rep Power: 12 ![]() |
You will have a higher pressure in that corner no matter what you do, it will divert the flow into into other directions. /Edit: unless you do what Yann proposed.
If you want unrealistic results, grab an AI image generator and tell him what it should look like :-) |
|
![]() |
![]() |
![]() |
![]() |
#8 | |
New Member
Join Date: Jan 2024
Posts: 23
Rep Power: 2 ![]() |
Quote:
I kept the inlet BC as it was before, but now the air doesn't go further than the inlet face. walls in BC K,omega,nut, and alphat are as zero gradient, and calculated. I think the patch idea, is near to what i am trying to do but the air going out of the inlet face totally changed although Inlet BC same as before. any suggestion? |
||
![]() |
![]() |
![]() |
![]() |
#9 |
Senior Member
Yann
Join Date: Apr 2012
Location: France
Posts: 1,169
Rep Power: 27 ![]() |
A screenshot would probably help to understand better what you are describing.
|
|
![]() |
![]() |
![]() |
![]() |
#10 |
New Member
Join Date: Jan 2024
Posts: 23
Rep Power: 2 ![]() |
okay let me share a screen shot, I repeated again the simulation, and i made the walls as patches with zero fixed velocity and zero gradient pressure.
you can see in the screenshot that the air going into the room and then the whole flow sucked by the outlet. I want to have no influence from the walls neither the outlet on the flow pattern, any suggestions for that? |
|
![]() |
![]() |
![]() |
![]() |
#11 |
Senior Member
Yann
Join Date: Apr 2012
Location: France
Posts: 1,169
Rep Power: 27 ![]() |
Hello,
Yes, on the walls, try using fixedValue on p and pressureInletOutletVelocity on U. |
|
![]() |
![]() |
![]() |
![]() |
#12 |
New Member
Join Date: Jan 2024
Posts: 23
Rep Power: 2 ![]() |
Thanks for your reply!
and what about outlet should i define it same as walls, or something different ? and then we can expect the air flow to be similar to the attached screen shot ? |
|
![]() |
![]() |
![]() |
![]() |
#13 |
Senior Member
Yann
Join Date: Apr 2012
Location: France
Posts: 1,169
Rep Power: 27 ![]() |
||
![]() |
![]() |
![]() |
![]() |
#14 |
New Member
Join Date: Jan 2024
Posts: 23
Rep Power: 2 ![]() |
now outlet is zero gradient in velocity and fixed value of pressure uniform 0.
so basically i will define both walls and outlet like Pressureinletoutletvelocity in U and in P zero fixed value. |
|
![]() |
![]() |
![]() |
![]() |
#15 |
New Member
Join Date: Jan 2024
Posts: 23
Rep Power: 2 ![]() |
I modified walls as you mentioned, but now the whole simulation just messed up.
I have weird behavior from the walls with crazy velocity range , and air doesn't enter from inlet. Inlet BC wasn't touched or modified. Any clue ? |
|
![]() |
![]() |
![]() |
![]() |
#16 |
Senior Member
Yann
Join Date: Apr 2012
Location: France
Posts: 1,169
Rep Power: 27 ![]() |
OK then maybe try inletOutlet for the velocity on the bottom and right wall, with an inletValue of (0 0 0) to prevent backflow, and keep the pressureInletOutletVelocity on the top and left wall. Still with fixedValue for the pressure.
|
|
![]() |
![]() |
![]() |
|
|
![]() |
||||
Thread | Thread Starter | Forum | Replies | Last Post |
SimpleFoam does not converge, velocity bounded | Andrea23 | OpenFOAM Running, Solving & CFD | 0 | September 28, 2021 09:44 |
Treatment of unimportant walls in turbulent simulations | Gerhard | OpenFOAM Running, Solving & CFD | 3 | March 1, 2021 03:02 |
Boundary Conditions k-omega-SST with slip walls | shock77 | OpenFOAM Running, Solving & CFD | 6 | October 23, 2020 16:57 |
simpleFoam - pressure | igor.leo93 | OpenFOAM Post-Processing | 5 | August 24, 2016 06:12 |
multiphaseEulerFoam loosing fluid with zeroGradient for p on walls??? | petr.f. | OpenFOAM Running, Solving & CFD | 0 | November 28, 2012 06:33 |