Momentum modeling of air supply devices

Hello SimScaler
For indoor climate analysis the correct modelling of air supply diffusors are essential. This means air flow rate an inlet velocity need to be accurate.

Since the geometry of diffusor cant be modelled directly (unpractical) simplification are been made. One of these simplification is called the momentum method, described here: In this paper a jet diffusor with 84 round nozzles (picture below (a)) is in different methods simulated (b,c,d,e) and the results with measurements compared.

As results of those 4 different methods (b,c,d,e) simulated and compared. Only (e) the momentum method correctly predict air flow. Now keep in mind that the surface area of the momentum method is the same as the gross area of the diffusor (a). The effective area of the diffuoser (nozzles → white area in the picture) are way smaller.

What i try do understand is how to model the momentum method within simscale. From the paper:

In the momentum method boundary conditions for the continuity equation and momentum equations are set separately so that the Inlet area can be chosen freely. Volume flow rate Qv and momentum Qu, of supplied air belonging to the actual diffuser are specified for an opening with the same dimensions as the area occupied by the nozzles (gross area): The momentum method can be regarded as setting infinite nozzleslslots as shown in figure 2.4.

Since the simplified diffusor (e) has a much larger effective surface area (= gross area) flowrate and velocity needs to be decoupled. If i understand correctly, this is not possible with an existing boundary condition in simscale? Any ideas on this topic?


Check this out. Momentum Sources | Advanced Concepts Documentation | SimScale

1 Like

Thanks for the answer! I guess thats the only option to de-couple flowrate and velocity, except if you have CFD-code that can do that out of the box (like in the paper described). I used a momentum source in one of my “development” project. I placed the momentum source in front of the air supply to accelerate the air flow. It is a possible solution, not a nice one but better than nothing.

I just wondering how people solve the issue with complex air diffusor. You cant really model them exactly and the simplification needs to be done correctly.

Cheers David

Yep, for example, I work with the simulation of spray drying chambers. The inlet diffuser has swirl vanes and discharges in form of an annular swirling jet into a drying chamber. The inlet diffuser is almost impossible to simulate. Therefore I use a simplified geometry and calibrate the inlet swirl to the prescribed boundary value. See the paper here: ChemEngineering | Free Full-Text | Low-Cost Eddy-Resolving Simulation in the Near-Field of an Annular Swirling Jet for Spray Drying Applications | HTML.

My question is are you using a grid “fine enough” to capture the behaviour of those 88 mini nozzles? Isn’t it better to use a simplified geometry and prescribe the turbulent behaviour produced by them?

If the momentum source you can set in SimScale is not enough, then OpenFOAM is your solution.

1 Like