Physical correlation to FloWorks for air ducting
 

My primary interest has been using FloWorks to improve the flow rate of air through automotive intake systems (filter housing, filter, intake tubes.) While my goal of having the simulation show whether or not a design iteration would be superior relative to the last seemed modest, it has not been so easy.

I prefer to use FloWorks to simulate the design prior to making a rapid prototype. While using FloWorks for the past year or so I have had cases where the simulation indicated that CAD model A was surely better (less pressure drop for a given flow rate) than model B, but in making rapid prototypes of each this was not true. Sometimes the opposite is actually true, B noticeably better than A in the physical test, which means our simulation tool actually sent us in the wrong direction and lengthed development time!

In exploring the cause for this, I am focusing on the knowledge that often the situation exists where the flow outlet has non-uniform pressure and velocity. Also, each design iteration will affect these pressure and velocity profiles. Therefore, the boundary conditions are never the same at the outlet. Suspecting that the boundary conditions must be identical to compare iterations, I am setting out to build geometry in the model to baffle and then diffuse the air at the outlet, so that the pressure is uniform (.1% delta between min and max value.)

So my question is do others add geometry to their outlets to affect the boundary conditions in this way? What considerations or pitfalls are there in this technique?

Re: Physical correlation to FloWorks for air ducti
 

Perhaps this is not a boundary condition issue.

It would be interesting to run these in a different CFD solver to see whether it showed the same trend as EFD.Lab.

Looks like an old post, but I would say this is most likely a boundary condition issue. I've come across similar issues in electronics cooling applications, where to get the correct flow through the box it's been necessary to include the environment outside the box in the analysis, for example if an intake vent is very close to the floor, as a fixed stagnation pressure boundary condition assumes that flow approaches normal to the boundary, whereas in reality it does not. So, my advice would be to first extend the model to capture the way the flow approaches the intake and then judge.