louvred fan / jet fan angle

[jim]

hi. as a general question, although can submit case file to issue
tracker if you deem it an issue...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:
VEL= - 1.2, VEL_T= 0.5, -0.3 (assuming zero thickness OBST with fan
property set as a single property for all faces)
but i find that if i want to extend this example to modelling a fan
blowing in the positive direction ie. VEL= +1.2 then i do not get the
angled flow effect.
i have played around with numerous combinations and looked at applying
the SURF to single faces...but not having much luck.

considered applying IOR=2 to force the direction, but this will not
work on an OBST, and VENT cannot be unattached to solid.

could you provide additional commentary on this to support section
9.1.4 in the user guide.

[dr_jfloyd]

Section 9.1.4 only discusses supply vents since VEL_T is only applied to supply vents.  In a supply inlet you have flow moving in a single direction in a duct that hits the louvers and deflects.  With a return air outlet you don't have this.  The pressure drop at the outlet pulls in air from all directions which is then redirected by the louvers as it enters the duct. 

[jim]

thanks.

is there a way to produce the effect im looking for, ie. angled flow
in +ve direction...this is in order to simulate multiple jet fans with
various orientations within a car park.

thanks

> > 9.1.4 in the user guide.- Hide quoted text -
>
> - Show quoted text -

[dr_jfloyd]

You cannot specify VEL_T for VEL > 0.  However, you could define the intake of the fan with multiple inlets and assign them the vector components.   For example if the fan is 45 degrees down and is only one grid cell thick it would look something like (where + is gas and O is fan obstruction):

++++++
+O++++
++O+++
+++O++
++++++

You could assign the left face of the upper left grid cell the x component and the top face of the upper left grid cell the y component. 

[jim]

thanks for the reply. but i dont really think that will be suitable for my application...each fan component would be 1 cell thick and i would not be able to ensure it is performing the same as the -ve direction jet fans without a lot of tweaking (angle is not 45deg).

 

is it really not possible to set the tangential flow for air for +ve flows? except by using multiple x, y SURFs to get a non physically representative resultant.

[dr_jfloyd]

Try this as an exercise.

Model one of your jet fans as a horizontal jet fan and look at the velocity vectors at the inlet.  See how far away you still observe the a strong directional flow towards the inlet.  My guess is that a fairly short distance away you will no longer see a strong directional influence of the entrance.  Then decide if given all the other uncertainties that will be present in your model how important the near-field of the entrance is to the overall results.

[jim]

the issue with my fan is that i know there are vanes at the 'exit nozzle' which blow the air at an angle of 30deg and at up ~25m/s towards the ground. the jets must blow downwards towards the ground in order to create sufficient air movement to flush the space of accumulated stagnant air (ie. CO clearance)

 

im not interested in the flows going into the fan. only the downstream flow ejected at an angle specified by contractor.

 

thanks

[dr_jfloyd]

Maybe I didn't understand your initial question.  If you have an exit flow, then VEL for the SURF is negative and you should have no problem using VEL_T.  If you are trying to specify an exit flow with VEL on SURF being positive, then you are actually specifying a suction and not a discharge.  The sign of VEL on SURF indicates incoming or outgoing flow and is not necessarily the same as the sign of the flow vector in the computational domain.

[jim]

right ok.

so when taking the example in 9.1.4 of the user guide i can only get this working in the -ve direction to the normal based on a single surface proprty and a multiple surface property. can you advise?

[dr_jfloyd]

I'm not sure what exactly you mean by -ve.  Create a very simple input file where you have the jet fan you are trying to model and indicate what you think should happen and what you see happening and post it to the issue tracker (http://code.google.com/p/fds-smv/issues/list)

[jim]

please see issue 1616.

 

thanks for any light you can shed.