FDS Simulation Taking Excessive Time

243 views
Skip to first unread message

Romaissa Hachana

unread,
Jan 19, 2024, 6:19:02 AM1/19/24
to FDS and Smokeview Discussions
Hello FDS community,
I'm currently running an FDS simulation and have noticed a significant increase in simulation time, even after switching to a more powerful computer. I'm using FDS version 6.8.0 on a system with Intel(R) Core(TM) i5-8250U CPU @ 1.60GHz   1.80 GHz and 8 GB RAM . Simulation is taking much longer than expected. (using &TIME T_END=5400.0)

Any guidance or suggestions would be greatly appreciated. Thank you!

dr_jfloyd

unread,
Jan 19, 2024, 7:34:16 AM1/19/24
to FDS and Smokeview Discussions
Depending on what you are simulating and the number of cores you have available to run FDS, FDS can take a very long time. Days or weeks of time are possible.

Are you running out of RAM? 
If this is a case where the timestep has suddenly decreased, do you have inputs that cause some change at that time or do simulation results show some suddent change in behavior that might impact the timestep?
Have you created conditions with high velocities?

Romaissa Hachana

unread,
Jan 20, 2024, 8:10:03 AM1/20/24
to FDS and Smokeview Discussions
AM runnig a simple simulation just when i change the time end to one hour or more it takes me long time like 2 weeks , is it from the input or i change something to boost it 
thank you 


&HEAD CHID='room_concrete', TITLE='room_made_with_concrete'  /

&MESH IJK=216,216,96, XB=0.0 ,38.0, 0.0, 38.0, 0.0, 17 /

&TIME T_END=5400.0 /

&MISC RADIATION=.FALSE.,TMPA =25./

&MATL ID='BRICK', CONDUCTIVITY=1.7, SPECIFIC_HEAT=0.9, DENSITY=2300. /
&SURF ID='BRICK WALL', MATL_ID='BRICK',THICKNESS=0.1 /

&REAC FUEL='METHANE', SOOT_YIELD=0.01/

&OBST XB= 2.9, 35.1, 2.9, 3.0, 0.0, 15.0,SURF_ID='BRICK WALL',COLOR='GRAY', TRANSPARENCY=0.5 / Side (A)
&OBST XB= 2.9, 3.0, 2.9, 35.1, 0.0, 15.0,SURF_ID='BRICK WALL',COLOR='GRAY', TRANSPARENCY=0.5 / Side (B)
&OBST XB= 2.9, 35.1, 35, 35.1, 0.0, 15.0,SURF_ID='BRICK WALL',COLOR='GRAY', TRANSPARENCY=0.5  / Side (C)
&OBST XB= 35, 35.1, 2.9, 35.1, 0.0, 15.0,SURF_ID='BRICK WALL',COLOR='GRAY', TRANSPARENCY=0.5  / Side (D)
&OBST XB= 2.9, 35.1, 2.9, 35.1, 15.0, 15.1,SURF_ID='BRICK WALL',COLOR='GRAY', TRANSPARENCY=0.5 /top

&HOLE XB= 14.0, 24.0, 2.9, 3.0, 0.0, 2.0/
&HOLE XB= 2.9, 3.0, 14.0, 24.0, 0.0, 2.0/
&HOLE XB= 14.0, 24.0, 35, 35.1, 0.0, 2.0/
&HOLE XB= 35, 35.1, 14.0, 24.0, 0.0, 2.0/


&VENT XB= 17.5, 20.5, 17.5, 20.5, 0.0, 0.0, SURF_ID='FIRE' /

&SURF ID='FIRE', HRRPUA=555.555, COLOR='RED'  /

dr_jfloyd

unread,
Jan 20, 2024, 8:15:57 AM1/20/24
to FDS and Smokeview Discussions
You have 4.5 million grid cells in a single mesh with a simulation time of 5400 s.  I would expect a runtime of weeks for this. Run this on a machine with multiple cores and use multiple meshes. There is discussion and examples of this in the FDS User's Guide.

Romaissa Hachana

unread,
Jan 20, 2024, 8:19:12 AM1/20/24
to FDS and Smokeview Discussions
Thank dr i appreciate your assistance 
Reply all
Reply to author
Forward
0 new messages