
[Sponsors] 
July 8, 2020, 12:25 
Initial conditions  Pressure

#1 
New Member
Blegh
Join Date: Jun 2020
Posts: 16
Rep Power: 4 
Hello I'm a student who is new to CFD. I hope you guys can help me.
I want to simulate flow through a duct. Lets say my inlet velocity equals 100m/s, my reference pressure equals 101300Pa (~1bar) and my outlet pressure equals a delta of 0. Now what confuses me are the inital conditions. So initial conditions are the values in each cell of the domain at the very beginning of the simulation. When I change my initial velocity from 0m/s to 100m/s all it does is converge quicker, since it needs less time to adapt. Am I getting that right? Lets say inital velocity and pressure both equal 0, if now I want to change the initial velocity to 100m/s would I have to change the pressure as well in order to get the same physics? In my understanding each cell now is initialized with the values of 100m/s and a static pressure of 1 bar (pressure delta of 0). Before however it were 0m/s with static pressure of 1bar. Now thats a difference, isn't it? If it iterates itself up from 0m/s to 100m/s my static pressure drops as the velocity increases so that when it has reached 100m/s the corresponding pressure now is < 1bar and therefore no flow occurs cause it's smaller than the outlet pressure. Again, am I getting that right? Last question: My inlet velocity of 100m/s must have a static pressure highter than 1bar in order to make a flow possible (since the outlet pressure equals 1 bar). How does my programm calculate the static pressure of my inlet velocity? Sure it must have a pressure higher than 1. But how exactly is it determined? Thanks a lot in advance! 

July 9, 2020, 13:19 

#2 
New Member
Blegh
Join Date: Jun 2020
Posts: 16
Rep Power: 4 
little push


July 9, 2020, 21:56 

#3 
Senior Member
Lucky
Join Date: Apr 2011
Location: Orlando, FL USA
Posts: 5,112
Rep Power: 60 
Well, you are correct that you need to provide two initial guesses: one for pressure and one for velocity. Both need to converge eventually. If you guess both correctly, it tends to converge faster. If you guess one incorrectly, it tends to take a little longer to converge. If you guess both incorrectly, longer than that. Your mileage may vary. So yes, different guesses leads to different solution histories and convergence behaviors. Hopefully, they all converge to the same correct result.
At boundaries, you have boundary conditions. When you have a "velocity inlet" it quite apparent that you specify a value for the velocity on the boundary (which is a Dirichlet type boundary condition on velocity). Actually, you also apply boundary conditions for all the transport variables (pressure and temperature if applicable). What you have is a Neumann boundary condition for the pressure (zero gradient). The rest is correctly solving the linear system of equations that results (i.e. matrix inversion / gaussian elimination). 

July 9, 2020, 22:17 

#4 
New Member
Blegh
Join Date: Jun 2020
Posts: 16
Rep Power: 4 
Hello, first of all thank you a lot for the answere this was very helpful.
When I have a inlet velocity boundary condition then I can only decide on temperature and velocity (as well as Turbulence specification but lets neglect them here). My inlet will be 100 m/s, my outlet will have a delta pressure of 0 (= static pressure of 1bar). Now, how to apply the correct pressure value at the inlet? And what is the key driver for that, "initial pressure"? So lets say my initial conditions are a velocity of 100m/s and a pressure difference of 0. Could it be that this is too low? Now both my inlet and outlet pressure are actually the same, right? Then a better option would be to increase the initial pressure? 

July 10, 2020, 14:19 

#5 
Senior Member
Lucky
Join Date: Apr 2011
Location: Orlando, FL USA
Posts: 5,112
Rep Power: 60 
Initial conditions are initial conditions. Boundary conditions are boundary conditions.
The pressure at the inlet is determined by the boundary condition which for a velocity inlet should be zero gradient condition. You can set you initial condition to anything, it doesn't change the way the boundary condition is calculated. What's initial pressure? Introducing new nomenclature without defining it doesn't help but it sounds like you are using Fluent. The initial /supersonic gauge pressure in Fluent has two purposes. When the flow is supersonic it becomes the fixed static pressure needed because the flow is supersonic. The other is if you use the "compute from" feature and let Fluent automatically calculate the initial values the it utilizes the value you put there. This is implementation specific to Fluent because the dropdown box for "compute from" is a boundary. So although the initial gauge pressure box appears in a boundary condition box in the interface, it's not a boundary condition. And although the initial pressure appears in the boundary condition box, it's also does not affect the boundary. If you don't use the compute from feature then the value in the initial gauge pressure has no meaning. It never gets used. 

July 22, 2020, 13:27 

#6 
New Member
Blegh
Join Date: Jun 2020
Posts: 16
Rep Power: 4 
sorry for the late reply, thanks a lot for your help! I am using StarCCM+ btw


Tags 
initial conditions, inlet and outlet, pressure and velocity, star ccm+ 
Thread Tools  Search this Thread 
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
[solidMechanics] Support thread for "Solid Mechanics Solvers added to OpenFOAM Extend"  bigphil  OpenFOAM CC Toolkits for FluidStructure Interaction  686  December 22, 2022 10:10 
decompose dependent solution  arionfard  OpenFOAM  3  December 10, 2018 10:36 
pressure in incompressible solvers e.g. simpleFoam  chrizzl  OpenFOAM Running, Solving & CFD  13  March 28, 2017 06:49 
Floating point exception error  lpz_michele  OpenFOAM Running, Solving & CFD  53  October 19, 2015 03:50 
Orifice Plate with a fully developed flow  Problems with convergence  jonmec  OpenFOAM Running, Solving & CFD  3  July 28, 2011 06:24 