[saaghi]

Hi,

As far as I know using "static pressure" at both inlet and outlet should not give the right answer in a CFD simulation as these are not enough BCs to solve an internal fluid dynamics problem.

What is baffling me is that I have the experimentally obtained inlet and outlet static pressures in a pipe when a specific inlet velocity profile (unsteady) is applied. I ran a simulation in Fluent with these experimental data (Pin-static and Pout-static) as the BCs and surprisingly, the velocity profile I get is exactly the same as the one that has been used in the experiment!

I use pressure inlet and pressure outlet BCs in fluent and set the inlet dynamic pressure value as zero. Can anyone explain this? It is an incompressible viscous flow problem.

Intuitively, if we keep the inlet and outlet static pressures constant in a pipe in an experimental setup (let's say for a steady flow), how many different flow fields are possible?

Bests

[FMDenaro]

Originally Posted by saaghi

Hi,

As far as I know using "static pressure" at both inlet and outlet should not give the right answer in a CFD simulation as these are not enough BCs to solve an internal fluid dynamics problem.

What is baffling me is that I have the experimentally obtained inlet and outlet static pressures in a pipe when a specific inlet velocity profile (unsteady) is applied. I ran a simulation in Fluent with these experimental data (Pin-static and Pout-static) as the BCs and surprisingly, the velocity profile I get is exactly the same as the one that has been used in the experiment!

I use pressure inlet and pressure outlet BCs in fluent and set the inlet dynamic pressure value as zero. Can anyone explain this? It is a compressible viscous flow problem.

Intuitively, if we keep the inlet and outlet static pressures constant in a pipe in an experimental setup (let's say for a steady flow), how many different flow fields are possible?

Bests

I am not sure about your question ....
in the Hagen-Poiseuille solution in a pipe, the velocity field is determined by the pressure gradient and Re number. But that is not the only solution you can get for the fixed pressure gradient.

How do you set the BC.s for mass, momentum and energy equations? is the flow really compressible (I mean Mach >0.3)?

[saaghi]

I made a mistake, it is an incompressible problem and I just edited my post. Sorry for that.

I do not solve the energy equation, and I chose "pressure inlet" and "pressure outlet" BCs at the inlet and outlet boundaries and no slip BC at walls in Fluent.

In a steady problem where the pressure gradient is fixed (delP,static=A), let's assume Vin=B is one of the answers. What other values are possible for Vin? What other factors are involved in how fast the fluid will flow with this fixed delP,static?

[FMDenaro]

Originally Posted by saaghi

I made a mistake, it is an incompressible problem and I just edited my post. Sorry for that.

I do not solve the energy equation, and I chose "pressure inlet" and "pressure outlet" BCs at the inlet and outlet boundaries and no slip BC at walls in Fluent.

In a steady problem where the pressure gradient is fixed (delP,static=A), let's assume Vin=B is one of the answers. What other values are possible for Vin? values bigger than B or smaller? What other factors are involved in how fast the fluid will flow with this fixed delP,static?

For incompressible flows, the BC.s are indeed either in velocity or in pressure.
You can have the exact solution from the Hagen theory

[saaghi]

In fact it is not exactly a pipe, it is an irregular geometry of blood vessel and also the flow can not be solved as a laminar flow because of the stenotic regions.

[pranab_jha]

In your original question you mentioned that the dynamic pressure value was set to zero. How was this done? Please explain.

The 'Pressure Inlet' BC allows us to provide a 'Total pressure' at that boundary. Hence, you are actually providing a dynamic pressure value also. The 'Pressure outlet' BC is a static pressure condition. The velocity at this boundary will be adjusted as per the flow field obtained from the solver.

Hope this helps.

[saaghi]

At the "total pressure" field, I enter the static pressure value I have from my experiments which means I neglect the dynamic pressure.

Now the question is why do I get the right answer with just knowing the Pin,static and Pout,static? These are not enough unknowns to solve the problem!

Some people say that there are several possible answers and what I get is one of them, but why do I always get the right answer? What other velocity can give the exact same pressure gradient? faster or slower?

[pranab_jha]

Originally Posted by saaghi

At the "total pressure" field, I enter the static pressure value I have from my experiments which means I neglect the dynamic pressure.

Now the question is why do I get the right answer with just knowing the Pin,static and Pout,static? These are not enough unknowns to solve the problem!

Some people say that there are several possible answers and what I get is one of them, but why do I always get the right answer? What other velocity can give the exact same pressure gradient? faster or slower?

Interesting! I do not know all the details of your problem, but will still try to answer.
If the static pressure value is being specified in place of the total pressure, a fixed part of that pressure will be converted to dynamic pressure. You can check the static pressure drop between the inlet and the outlet from "report ---> surface integrals" and compare with experimental results. Also, please check if your solution is converged. Try reducing the monitors to 10^-5 or -6. Consider running a finer mesh to see if you are still getting the same result.

One more thing, check your experimental results to see if you did actually include the dynamic pressure at the inlet.

I would be interested in your findings.

Regards,
Pranab

[FMDenaro]

the mathematical problem for the incompressible flow is well posed with the pressure BC.s. The viscous part of the flow is taken into account by the tangential velocity