How does FLUENT set the velocity conditions when we have a pressure inlet and pressure outlet?

- April 9, 2019 at 12:42 am
  
  Jamasp
  Subscriber
  
  In CFD simulation of incompressible flow in a pipe, when we have both a pressure inlet and pressure outlet we basically impose no velocity constraint on either inlet or outlet.
  
  My question is how does Fluent calculate the velocity profile at the inlet? In this kind of problems does the velocity gradient at both inlet and outlet is zero (Neumann boundary condition)?
  
  There is something about pressure inlet in the following link that is not clear to me:
  
  http://www.afs.enea.it/project/neptunius/docs/fluent/html/ug/node239.htm
  
  We specify total pressure at the inlet and use the Bernoulli's equation to relate static pressure and velocity inlet. Fluent documentation says "With the resulting velocity magnitude and the flow direction vector you assigned at the inlet". What is the resulting velocity magnitude at the inlet and how does Fluent calculate it? I don't understand this!
  
  Please see an example in the attachment. This simple simulation shows a parabolic velocity profile at the inlet. I'm wondering how this is calculated! Please let me know if there is any paper that can help me to understand this.
  
  Thank you,
- April 9, 2019 at 5:12 am
  
  Amine Ben Hadj Ali
  Ansys Employee
  
  What you specify is total pressure. Fluent will then iterate towards the velocity magnitude (or system mass flow) and static pressure. That is why it is recommended to start wuth good initial guess for the pressure field.
- April 9, 2019 at 3:01 pm
  
  Jamasp
  Subscriber
  
  Hi Amine,
  
  Thank you for your reply. It means that Fluent calculates velocity magnitude and static pressure at the inlet, which both are unknown. But how it calculates mass flow rate to determine the velocity at the inlet? What is the equation for that? Could you please give me more information about this?
  
  Best,
  
  Jimmy
- April 9, 2019 at 3:24 pm
  
  Amine Ben Hadj Ali
  Ansys Employee
  
  Velocity x density x facet area x density.
- April 9, 2019 at 4:52 pm
  
  Jamasp
  Subscriber
  
  But how the total pressure is related to this? Bernoulli's equation is what we use to relate to them. But we need to get velocity first to get static pressure. Am I right? Sorry, it is still confusing for me.
  
  Jimmy
- April 9, 2019 at 6:36 pm
  
  Amine Ben Hadj Ali
  Ansys Employee
  
  Iterative solver... That us why my remark regarding static pressure field ( first cell layer near inlet are used in the calculation...)....
  ...
  
  And I have just provided formula for Mass Flow.
- April 10, 2019 at 9:16 pm
  
  Jamasp
  Subscriber
  
  So the iterative solver gives us the velocity and static pressure only at the inlet using the information from the first cell layer near the inlet. Is that correct? I just want to clarify this.
  
  Thank you for your help.
  
  Jimmy
- April 11, 2019 at 4:08 am
  
  Amine Ben Hadj Ali
  Ansys Employee
  
  As a result if the iterative solution of the pressure/momentum field.
- April 11, 2019 at 2:36 pm
  
  Jamasp
  Subscriber
  
  Thank you Amine. Can you please share any references or papers regarding this? I would appreciate.
  
  Best,
  
  Jimmy
- April 11, 2019 at 4:25 pm
  
  Rob
  Forum Moderator
  
  Please look in the User's Guide and/or Theory Guide. We don't have references to the various solvers other than those in documentation which you have access to.
- April 24, 2019 at 7:41 pm
  
  Jamasp
  Subscriber
  
  Hi Rwoolhou. I reviewed the User's Guide and Theory Guide to learn about the iterative solver. I couldn't find more detail about the solver!
  
  Thanks,
- April 25, 2019 at 9:36 am
  
  veera
  Subscriber
  
  hi jamasp
  
  actually we know the , concept , number of unknowns are there , we need to have number of equations for finding that unknowns..so in cfd, for example , we need to find v, and p? means .. we need to equations ..yah it will take continuity and momentum(for example i am telling ) so it will solve both continuity and momentum ,, thats what i am understand-ed .. if any thing wrong sorry for that
- April 25, 2019 at 3:09 pm
  
  Jamasp
  Subscriber
  
  Hi Veera,
  
  That's true. I understand the concept of calculating V and P. But we need to know the BCs first to calculate them. My question is about Pressure inlet boundary that we define it by specifying total pressure. This means that we don't know V and static P at the inlet. The relation between them is Bernoulli's equation.
  
  I'm also interested to know how we get a parabolic shape of velocity at the inlet? Why isn't it fully developed? It's different from fully developed flow. Please see the picture in the first comment showing a simple pipe.
  
  I just want the get a clear answer about the velocity profile that solver calculated at the inlet. Can we specify uniform? if no, why?
  
  Thanks,
- April 25, 2019 at 11:44 pm
  
  Jamasp
  Subscriber
  
  Please see the following picture, showing the initial values of V and P and also in the first iteration. The initial values for V and P are zero. In the first iteration, both V_in and P_in are changing. What is the iterative solver that you are talking about to calculate these two? Are they part of the system and calculating with the interior? If it does, this means that we are solving the problem with ONE boundary condition! Am I right?