Jet-fan and FDS

[cfd-fire]

Gents,
could please help me.
I did real measurements for jet-fan and I did CFD analyses in FDS v.
5.5.3.
I received very strange results. For example: the velocity for free
stream jet 25 m far from the jet-fan outlet was about 12 m/s (grid
size 10 cm) and about 21 m/s (grid size 5 cm). The experimental result
was 1,2 m/s .


Thanks for your help.

[cfd-fire]

[F-Sim.de]

http://www.fds-usergroup.de/assets/files/Tagungsband/FDS_UG_4.Band_web.pdf

Page 92. Gregor Jaeger did some validation work and found that (the
pre-versions of) FDS 6 will do much better in simulating jet-fans.

By the way, it would be much better to measure the length of the core
zone instead of picking out one single value.

Boris

[cfd-fire]

In the our simulation the core zone is 25 m long.
We did velocity measurements in whole air stream not only in centre
line.


On 23 Maj, 13:13, "F-Sim.de" <I...@F-Sim.de> wrote:
> http://www.fds-usergroup.de/assets/files/Tagungsband/FDS_UG_4.Band_we...

[Gregor]

I have a lot of question.

Can you simplify the case and post it in the issue tracker? I'll take
a look.

[Kevin]

Keep in mind that FDS models a fan only as a pressure drop. There is
nothing to account for the vorticity generated by the rotating
blades.

> > > Boris- Hide quoted text -
>
> - Show quoted text -

[cfd-fire]

Of course Gregor.
I will do it.

[Koolifant]

@Kevin: Isn't there a possibility to ad something like a 'vorticity
source' on the outlet of jetfans in FDS?

[Kevin]

For a RANS (Reynolds Averaged Navier-Stokes) model, you essentially
increase the numerical viscosity that represents the subgrid-scale
motion that you cannot resolve. This basically smears out the flow
field, decreasing the velocity along the core. For LES, it is more
difficult. Hard to explain, but suffice it to say that the dynamic
form of the Smagorinsky LES model in FDS 6 is better equipped for
turbulent jets like a fan.

> > > - Show quoted text -- Hide quoted text -

[cfd-fire]

@Kevin: Could you please suggest us: how are we can model the jet-fan
in FDS?
I agree with you explanation about RANS model but the results of
calculation based on
k-e model are closed to experiments.

[Kevin]

Try putting some obstructions in front of the fan.

[cuub]

Grzegorz,
You're right. When using default settings of FDS 5, velocity vectors
are not so similar with those measured in reality. I compared them
with many velocity characteristics I got from manufacturers of jet
fans. But by turning on some FDS 6 options, You can get better
results, especially when You consider the core of the stream. There
are no long streams with high velocities. But in my opinion there is
still problem with the whole profile of velocities (shape). I noticed
that when You do an average profile (in time) for example for 10, 20
seconds, it gives quite good results. The stream is moving left, right
(because of turbulence model), so the average (in time) profile is
quite similar to those I got from manufacturers. But probably real
profile of jet fan is more stable (not so many turbulences), but this
is a question to You (as You made measurements).
Pozdrawiam,
Kuba

[cfd-fire]

Kuba,
the profile is much more stable. There are some turbulence but move
left and then right is non-physical.
Please keep in mind, that for number of cells depend on Reynolds
number. For good DNS we need Ncell = Re^(9/4)
for LES 10 times lower. But still a lot.

Pozdrawiam