Hi,

Is this statement correct? In a flow simulation, the resulting velocity and pressure profile within an arbitrary pipe should be VERY similar between using pressure inlet / pressure outlet BC's vs flow rate inlet / pressure outlet BCs?

In other words... If I ran a flow simulation using 1 GPM inlet flow rate and environmental pressure (atm) outlet and my surface goal was pressure drop between the inlet and outlet and this yielded a 2 psi pressure drop.

Now if I set up the same flow simulation with (exact same geometry and mesh) with a inlet pressure of 2 psi higher than the environmental pressure outlet with a surface goal of max volumetric flow rate at the inlet... I should yield 1 GPM inlet flow rate correct? And these two simulations should yield very similar if not exact velocity/pressure profiles correct?

I find running these two simulations (using the result of one sim as a boundary condition of the other) a good way to verify the simulation results.

However, I don't seem to be getting similar results and the velocity and pressure profiles are different.

Thanks,

Chris

Hi Chris,

In theory you are correct in thinking that comparing a velocity inlet/pressure outlet project and a pressure inlet/pressure outlet project should yield the same results. In practice, this isn't really the case. There is no analytical equation that perfectly predicts flow behavior (you will make $1,000,000 if you find an equation that does though) so the iterative approach CFD takes leads to multiple possible solutions.

Having only pressure boundaries on a flow projects creates a less stable system to solve. That combine with the fact that there are going to be several "solutions" that fit the boundaries you set within the tolerance of convergence leads to situations where intuitively the results should match but the solver math didn't make it happen.

You can help bias the solver towards solutions through the use of more goals but unless you tell the program what you are expecting to see with a boundary definition (adding a flow inlet or outlet) the goals just serve to make the solver keep solving further until it can meet the additional goal criteria.

In short, you want to give the solver the best chance at giving you a correct solution. If you have data for a flow boundary to add instead of using two pressure boundaries, that would give you a more stable solve and a more reliable result.