Aerodynamical vs. geometrical natural vent area

[quantumfds]

Hello,

I am trying to model a smoke extraction unit located horizontally on the roof (2 × 2 m). The blades open 90 degrees and are aligned parallel to the smoke flow from the room to the outside.

According to the manufacturer, the discharge coefficient is 0.7, which seems rather high to me. I am unsure whether the blades actually help maintain the flow through the opening or whether they create additional blockage.

The grid size in the model is 0.2 m, and I am not sure whether the area defined in FDS should be the geometrical area or the aerodynamic (effective) area e.g. if the grid is able to catch the effect. Alternatively, should surfaces in FDS be interpreted strictly as geometrical openings, meaning that aerodynamic openings need to be increased in size in reality to match the simulated flow? In this example, an aerodynamic area of 4 m² would correspond to approximately 5.7 m² of geometrical opening.

Thanks.

[mcgratta]

The first thing that I would do is set up a simple example where you create an opening in a compartment ceiling that is nominally the same size as the actual vent. Push air into the compartment at a rate that is sufficiently high to create an over-pressure of, say, a few dozen Pa. Record the volume flow through the vent and the pressure differential. Calculate the discharge coefficient using whatever formula the manufacturer is using. If you are close to 0.7, then I do not see what else you can do other than running a full-scale experiment on the actual vent. If your coefficient is very different, then we can talk some more, and you can provide your simple example case.

[drjfloyd]

There is also the approach used in the FM/FRPF data center cases. Mock up a simple, small compartment with the resolution you plan to use in the full scale simulation. Resolve the details of the vent structure as best you can on that grid. Do as Kevin suggested with creating a small over pressure and compute the discharge coefficient. If your coeffiicent is too high, you could try placing a plane of particles in the vent opening and tuning the drag coefficient to get the correct loss.