FDS simulation

199 views
Skip to first unread message

Cesar Becerril

unread,
Oct 24, 2018, 9:58:36 AM10/24/18
to FDS and Smokeview Discussions
Dear all,
I am used to LES simulation codes and discovering FDS. For now, I am trying to understand and play with one of the FDS examples, i.e. "circular_burner.fds". I have got some questions:

1. Why does the burner introduce a fluctuating mass flow and in a kind of gaussian shape and not in a flat like profile?
2. If I change the shape of this burner to a squared burner, I obtain a 0 mass flow introduction through the ventilation ( cf. &VENT XB=-0.6,0.6,-0.6,0.6,0,0, SPREAD_RATE=0.05, COLOR='BLUE', SURF_ID='BURNER' /). Am I doing a mistake?

The code is just below. In advance thank you for your help.
César

====================================================================

&HEAD CHID='circular_burner', TITLE='Test the mass flow rate from a circular burner' /

&MULT ID='mesh array', DX=1.0,DY=1.0,DZ=1.0, I_UPPER=1,J_UPPER=1,K_UPPER=1 /
&MESH IJK=20,20,20, XB=-1.0,0.0,-1.0,0.0,0.0,1.0, MULT_ID='mesh array' /

&TIME T_END=20. /

&REAC FUEL='PROPANE', SOOT_YIELD=0.015 /

&SURF ID='BURNER', MASS_FLUX(1)=0.02, SPEC_ID(1)='PROPANE', TAU_MF(1)=0.01 /

&VENT XB=-0.6,0.6,-0.6,0.6,0,0, XYZ=0,0,0, RADIUS=0.5, SPREAD_RATE=0.05, COLOR='BLUE', SURF_ID='BURNER' /

&VENT MB='ZMAX',SURF_ID='OPEN'/
&VENT MB='YMIN',SURF_ID='OPEN'/
&VENT MB='YMAX',SURF_ID='OPEN'/
&VENT MB='XMIN',SURF_ID='OPEN'/
&VENT MB='XMAX',SURF_ID='OPEN'/

&DUMP DT_HRR=0.1 /

&SLCF PBY=0.0, QUANTITY='TEMPERATURE' /

&TAIL /


Randy McDermott

unread,
Oct 24, 2018, 10:08:28 AM10/24/18
to FDS and Smokeview Discussions
SPREAD_RATE is used to mimic a point ignition that spreads out radially (it gives the classical t^2 heat release curve).  If you want a flat profile to exit the radial vent, set the SPREAD_RATE to 1000.

Don't use SPREAD_RATE for a square vent.


--
You received this message because you are subscribed to the Google Groups "FDS and Smokeview Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fds-smv+u...@googlegroups.com.
To post to this group, send email to fds...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/14ea84d0-7ed5-403d-8471-f31c5a8e855f%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Cesar Becerril

unread,
Oct 24, 2018, 10:55:24 AM10/24/18
to fds...@googlegroups.com
Dear all, dear Randy,

Thank you for your answer. I want to set a premixed laminar flame to extract some data as the adiabatic flame temperature and the laminar flame speed. So this script is for me a very good start to understand how fds works. I am now setting a square ventilation with a mixture of CH4 and air:

&SURF ID='BURNER', MASS_FLUX(1)=0.0184569608572, SPEC_ID(1)='METHANE', TAU_MF(1)=0.01
                   MASS_FLUX(2)=0.318754905207 , SPEC_ID(2)='AIR'    , TAU_MF(2)=0.01 /
&VENT MB='ZMIN', COLOR='BLUE', SURF_ID='BURNER' /

And with this configuration I would expect a flat flame front (which for instants is not the case, but lets say that the open boundary conditions elsewhere are not appropriate). I still have some questions: How or why is the mixture ignited? And why the ventilation produces a fluctuating flow?

Kind regards,
César


Randy McDermott

unread,
Oct 24, 2018, 11:14:28 AM10/24/18
to FDS and Smokeview Discussions
Have you read Chapter 16 of the user guide?

Cesar Becerril

unread,
Oct 24, 2018, 11:16:32 AM10/24/18
to fds...@googlegroups.com
Yes, Combustion. I surely missed something ?

Randy McDermott

unread,
Oct 24, 2018, 11:22:19 AM10/24/18
to FDS and Smokeview Discussions
Premixed combustion is not a typical FDS application.  You would have to use finite-rate kinetics and DNS.  There is no validation for FDS in this area.  You should be very careful in every step in your development process to ensure you are getting accurate results.

Cesar Becerril

unread,
Oct 24, 2018, 11:28:31 AM10/24/18
to fds...@googlegroups.com
Thank you for the warning I will be carefull with this kind of modelling. However, even if FDS may not be adapted to model a premixed case, I still don't understand why the mixture is ignited from the ventilation or why does the ventilation introduces a fluctuating flow?

Randy McDermott

unread,
Oct 24, 2018, 11:39:43 AM10/24/18
to FDS and Smokeview Discussions
By default FDS uses the EDC combustion model---mixed is burnt.

FDS is an LES code.  It is transient.  Depending on the boundary conditions, the flow is usually fluctuating.

Cesar Becerril

unread,
Oct 24, 2018, 12:29:58 PM10/24/18
to fds...@googlegroups.com
I completely agree with you. However, boundary conditions in non compressible solvers usually impose “hard” values.

dr_jfloyd

unread,
Oct 24, 2018, 3:04:10 PM10/24/18
to FDS and Smokeview Discussions
FDS USer's Guide

7.9 Background Noise

FDS initializes the flow field with a very small amount of “noise” to prevent the development of a perfectly
symmetric flow when the boundary and initial conditions are perfectly symmetric. To turn this off, set
NOISE=.FALSE. To control the amount of noise, set NOISE_VELOCITY. Its default value is 0.005 m/s.

The moment fuel enters the domain at your VENT, the noise means that there will be slightly different mixing in the different grid cells over the burner. The small differences in heat release rate that develop will quickly result in the fluctuations you are seeing.

Cesar Becerril

unread,
Oct 25, 2018, 3:20:25 AM10/25/18
to fds...@googlegroups.com
Dear dr_jfloyd,

I have already set this noise parameter to false (i.e. &MISC NOISE=.FALSE.) but it does not change the fluctuating character of my vent.

Thank you for your help

Randy McDermott

unread,
Oct 25, 2018, 6:27:42 AM10/25/18
to FDS and Smokeview Discussions
Please send us the input file.

Cesar Becerril

unread,
Oct 25, 2018, 7:45:02 AM10/25/18
to fds...@googlegroups.com
Please find it below (simulation ran using FDS6.7.0):

&HEAD CHID='circular_burner', TITLE='Test the mass flow rate from a circular burner' /

&MESH IJK=40,40,40, XB=-1.0,1.0,-1.0,1.0,0.0,2.0 /

&TIME T_END=20. /

&REAC FUEL='METHANE' /

&MISC NOISE=.FALSE., NOISE_VELOCITY=0.0/


&SURF ID='BURNER', MASS_FLUX(1)=0.0184569608572, SPEC_ID(1)='METHANE', TAU_MF(1)=0.01
                   MASS_FLUX(2)=0.318754905207 , SPEC_ID(2)='AIR'    , TAU_MF(2)=0.01 /

&VENT MB='ZMIN', SURF_ID='BURNER' /
&VENT MB='ZMAX', SURF_ID='OPEN'/
&VENT MB='YMIN', SURF_ID='OPEN'/
&VENT MB='YMAX', SURF_ID='OPEN'/
&VENT MB='XMIN', SURF_ID='OPEN'/

&VENT MB='XMAX', SURF_ID='OPEN'/

&BNDF QUANTITY='WALL TEMPERATURE' /
&BNDF QUANTITY='MASS FLUX WALL' /
&BNDF QUANTITY='MASS FLOW WALL' /

&DUMP DT_HRR=0.1, MASS_FILE=.TRUE. /


&SLCF PBY=0.0, QUANTITY='TEMPERATURE' /
&SLCF PBY=0.0, QUANTITY='MASS FRACTION', SPEC_ID='METHANE' /
&SLCF PBZ=0.0, QUANTITY='MASS FRACTION', SPEC_ID='METHANE' /   <=== Fluctuating
&SLCF PBZ=1.0, QUANTITY='MASS FRACTION', SPEC_ID='METHANE' /
&SLCF PBZ=0.0, QUANTITY='MASS FRACTION', SPEC_ID='AIR' /

&TAIL /

Randy McDermott

unread,
Oct 25, 2018, 10:19:25 AM10/25/18
to FDS and Smokeview Discussions
Couple of things:
1. SCLF gives you gas phase quantities.  Even if you put Z=0 plane, it is interpolating gas phase data to that plane.  Not good practice to use SLCF on a surface.
2. The value of the mass fraction at a wall cell in FDS will change to make the total flux of species equal to what you specified in the boundary condition.  So, what you are interested in is that the flux is constant.  And it is, as you can see in a BNDF of MASS FLUX WALL for METHANE.

&BNDF QUANTITY='MASS FLUX WALL', SPEC_ID='METHANE', CELL_CENTERED=T /

Note, do not use a BNDF for MASS FLOW WALL.  This output requires a DEVC.  (I've added an error trap for this in next release.)


Cesar Becerril

unread,
Oct 25, 2018, 10:27:18 AM10/25/18
to fds...@googlegroups.com
Dear Randy,
Thank you for your answer, this is really valuable information for me. However, I cannot use BNDF with only MASS FLOW, I obtain an ERROR indicating that this output is not appropriate for BNDF.

Cesar Becerril

unread,
Oct 25, 2018, 10:30:08 AM10/25/18
to fds...@googlegroups.com
Sorry, I've just understood, bad question.
Thank you very much for your help!
Reply all
Reply to author
Forward
0 new messages