[fatirishman53]

I was in a debate with my advising professor about openFOAM and I fear that I was unable to make a compelling argument. I was hoping that someone might be able to answer this question for both of us:

The Navier-Stokes equations are equations with reference to components of velocity and contain a pressure gradient term. So, if, for instance, the boundary and initial conditions are set for a wall patch, why is there a need to set pressure boundary and initial conditions at that wall? Shouldn't all pressure conditions be implied or solved for using the velocity conditions?

My professor's concern is that of over-constraining the model.

I hope I conveyed this question correctly. Thanks, in advance, for your help.

[shipman]

Originally Posted by fatirishman53

I was in a debate with my advising professor about openFOAM and I fear that I was unable to make a compelling argument. I was hoping that someone might be able to answer this question for both of us:

The Navier-Stokes equations are equations with reference to components of velocity and contain a pressure gradient term. So, if, for instance, the boundary and initial conditions are set for a wall patch, why is there a need to set pressure boundary and initial conditions at that wall? Shouldn't all pressure conditions be implied or solved for using the velocity conditions?

My professor's concern is that of over-constraining the model.

I hope I conveyed this question correctly. Thanks, in advance, for your help.

Hi,

I just could not understand well your question? How to you set ''for instance, the boundary and initial conditions are set for a wall patch''???

Actually, in openfoam mostly initial conditions can be driven by pressure or velocity. If you set condition based on pressure (For example at the inlet if you set pressure as fixedValue, then you set the velocity as zeroGradient means that calculate the velocity according to fixed pressure at the inlet) For a wall mostly it is used slip or non-slip wall conditions. If you use non-slip wall means that at wall velocity is zero and pressure is defined as zeroGradient or another special pressure wall conditions.

Just i could not understand that by defining like this how you are over-constraining the model? By the way which solver do you use and what kind of case you try to solve?

Hope these helps for you.

BR

Baris

[fatirishman53]

Originally Posted by shipman

Hi,

I just could not understand well your question? How to you set ''for instance, the boundary and initial conditions are set for a wall patch''???

Actually, in openfoam mostly initial conditions can be driven by pressure or velocity. If you set condition based on pressure (For example at the inlet if you set pressure as fixedValue, then you set the velocity as zeroGradient means that calculate the velocity according to fixed pressure at the inlet) For a wall mostly it is used slip or non-slip wall conditions. If you use non-slip wall means that at wall velocity is zero and pressure is defined as zeroGradient or another special pressure wall conditions.

Just i could not understand that by defining like this how you are over-constraining the model? By the way which solver do you use and what kind of case you try to solve?

Hope these helps for you.

BR

Baris

Sorry, what I meant to say was this:
Let's assume incompressable flow. According to the incompressable NS equations, if all components of velocity are 0 (i.e. "uniform (0 0 0)"), then the pressure gradient (the only pressure term in the incompressable NS equations), is 0. So why must pressure conditions at this boundary be specified? The solver should have everything it need from the specified boundary conditions to solve for the pressure gradient. Wouldn't it be over-constraining to specify that pressure has "zeroGradient"? By OpenFOAM's definition, "zeroGradient" means "normal gradient of phi..." (pressure, in this case) "...is zero". This should already be implied by any algorithm that uses the incompressable NS equations to solve for pressure.

If it helps, we can refer to the icoFoam solver directly, but the above should apply for any solver that uses the NS equations to solve for pressure. I am currently working on turbulent flow within a channel, but this question isn't specific to any particular case.

As an aside, is there anyway to post equations? Or do you have to insert pictures of equations?

[elones]

Hi,
look at the code of solver,
Poisson's equation is solved in step of pressure correction.

OpenFOAM mainly used a pressure velocity coupling algorithms (SIMPLE, PISO etc.)

More info can be found for example at https://openfoamwiki.net/index.php/O...hm_in_OpenFOAM

Ondra