louvred fan / jet fan angle - problem with direction

[gforney]

Please complete the following lines...

Application Version:5.5.3 serial
SVN Revision Number:7031
Compile Date:fri, 29 OCt 2010
Operating System:Windows 2010

Describe details of the issue below:

i am having touble setting angles on 
louvred fans. 

i can get the correct angle im looking for when following the example 
in the user guide based on 9.1.4, but i cannot get the same angle of deflection when
trying to model the flow in opposite direction.
the application is for jet fans suspended from a ceiling in a car park, they face various
directions (+ve and -ve to the normal and all must be set with a deflection towards
the ground)

i have attached 2 very simple input files to explain what i mean:

120410_NEG.fds shows the sort of tangental effect i want to see and is based on the
following SURF properties:-
VEL=-19.0, VEL_T=5.0,-10.0

120410_POS.fds shows an attempt to create the tangental effect i want in the +y direction,
but i do not get any angular deflection. It is based on the following SURF properties:-
VEL=19.0, VEL_T=5.0,-10.0

the effect i want to create is the same downward deflection as 120410_NEG, but in the
opposite direction. cant seem to get it to work fort some reason. i realise specifying
VEL=+number 

Original issue reported on code.google.com by berto57 on 2012-04-12 13:11:04

---

• _Attachment: 120410_NEG.fds_
• _Attachment: 120410_POS.fds_

[gforney]

Please carefully read the User's Guide (I also mentioned this in my post in the forum.)

Section 9.1.1

"The easiest way to describe a supply or exhaust fan is to create a VENT positioned
on a solid surface with a SURF_ID with some form of specified velocity or volume flow
rate. The normal component of velocity is usually specified directly via the parameter
VEL. If VEL is negative, the flow is directed into the computational domain, i.e.,
a supply vent. If VEL is positive, the flow is drawn out of the domain, i.e., an exhaust
vent"

Your second case has VEL > 0. The User's Guide states that VEL > 0 means flow is removed.
 Why are you expecting it to inject flow when the guide says it will be exhausted?
 Change the sign on your VEL and see if that solves your problem.

Original issue reported on code.google.com by drjfloyd on 2012-04-12 13:58:07

[gforney]

I missed the POROUS line when I first looked at your case so your positive VEL usage
is correct.  The overall issue is as I indicated in the forum.  VEL_T does not get
applied when VEL>0.  In FDS 6 POROUS will be replaced by the new HVAC functionality
and this issue would not exist. 

Original issue reported on code.google.com by drjfloyd on 2012-04-13 11:46:58

[gforney]

There is a potential work around for your issue.  Do all of your fans lie in only one
or two orthogonal directions?  That is are all your x direction fans only +x or only
-x but not combinations of both and same for y?  If this is the case, then just reflect
your geometry about the axis to make the positive fans negative.  For example if your
fans are pointing +x and -y you could reflect about the y-axis: 

Make all
XB = x1,x2,....
into
XB=-x1,-x2

This can be done fairly quickly by pasting your input file into EXCEL and doing text-to-columns
expanding on , and =.  Then paste in formulas to turn x into -x.  Then use CONCATENATE
to reassemble the input line with the = and ,

Original issue reported on code.google.com by drjfloyd on 2012-04-13 12:19:14

• Labels added: Application-FDS

[gforney]

hi. no they are a combination of directions, including at angles to the normal. 

Original issue reported on code.google.com by berto57 on 2012-04-13 15:17:31

[gforney]

In that case you could use two POROUS OBST lines for the fans in the +x and/or +y direction.
 Make one POROUS OBST to provide horizontal flow (this would be the input line you
have now) and make the second POROUS OBST to provide the vertical flow.  For example
say you have a 30 degree down angle and a fan exit velocity of 10 m/s.  

You would arrange the porous obst to look like:
 --
|
|

And assign a velocity of 8.7 m/s [10 m/s * COS(-30)]to the horizontal flow and -5.0
m/s [10 m/s * SIN(-30)] to the vertical.  The vector sum of these two flows would be
10 m/s at a 30 degree down angle.

If you have an angle in the x/y plane as well, then add a third POROUS OBST for that
flow component

Original issue reported on code.google.com by drjfloyd on 2012-04-13 15:43:52

[gforney]

ill give that a shot, was trying to avoid that way of doing it if possible though as
its a lot more set up across all the various fans. 
i will report back how this performs. 

Original issue reported on code.google.com by berto57 on 2012-04-13 15:48:16

[gforney]

this does not perform well relative to the correctly angled vents. there are unusual
resultants jetting between the fan plates.

are you saying this issue does happen with FDS 6...can you tell me more about that?

Original issue reported on code.google.com by berto57 on 2012-04-17 15:53:07

[gforney]

The POROUS feature has been deleted in FDS 6 and will be handled by the HVAC model.
   You can look at the current version of the jet fan case in the repository to see
how that would be done.  The end result of this in terms of VEL_T is that you would
make the jet fan vent one cell thick and then you could apply VEL_T to the fan outlet
without the thin OBST limitations.  

On your right angle vents, try setting the FREE_SLIP=.TRUE. on those vents.  That will
keep wall friction from interfering with the flow field at the face of the vents. 
That should get the flow to behave correctly.

Original issue reported on code.google.com by drjfloyd on 2012-04-18 01:39:20

[gforney]

Closing this issue since HVAC in FDS 6 will allow for VEL_T for a jet fan.

Original issue reported on code.google.com by drjfloyd on 2012-05-18 16:56:01

[gforney]

Issue 1647 has been merged into this issue.

Original issue reported on code.google.com by drjfloyd on 2012-05-30 13:04:44

[gforney]

btw, is there any way to apply this 'fix' when considering volume_flux rather velocity?

Original issue reported on code.google.com by berto57 on 2012-07-12 14:37:45

[gforney]

I'm not sure what you are asking.

Are you asking if you can use VEL_T with VOLUME_FLUX? Are you asking with respect to
FDS 5.5.3 or what will be in FDS 6?

Original issue reported on code.google.com by drjfloyd on 2012-07-12 14:44:30