Numerical instability

Torbjørn Knutson

unread,

Apr 13, 2015, 7:21:55 AM4/13/15

to fds...@googlegroups.com

Hi.

I'm having a problem with "numerical instability". I thought it had something to do with too much turbulence in the model's air-/smokeflow, but whatever I do it still stops after about 80-90 seconds (out of 300 s)

My efforts to fix it:
- Split meshes (till a total of 4 meshes)
- Increase vent sizes (air out) for HVAC
- Incresae vent size (air out) for smoke hatchet
- Increase vent size (air in) for HVAC/smoke hatchet
- Decrease flow rate for smoke hatchet
- Change location of vent (air in), further away from smoke hatchet
- Decrease number of fires (burners) from 2x350kW to 1x700kW
- Remove smoke hatchet completely

I've added both the FDS- and the Pyrosim-file. I really hope someone can help me with this.

Kind regards
Torbjørn

ss_simplified_twomeshes_7_oneburner.psm

ss_simplified_twomeshes_7_oneb.fds

dr_jfloyd

unread,

Apr 13, 2015, 8:36:48 AM4/13/15

to fds...@googlegroups.com

Since you are specifying the flows explicitly for all of your vents, there really isn't any need to use the HVAC model. You could just use SURF with VEL or VOLUME_FLOW.

Your fresh air duct has a 0.3 m diameter with a flow of 15 m^3/s. That is resulting in a > 200 m/s flow in the duct. This might be your problem.

Randy McDermott

unread,

Apr 13, 2015, 8:55:22 AM4/13/15

to FDS and Smokeview Discussions

Running your input file with the latest code I get the following message:

ERROR: SURF burner cannot be applied to a thin obstruction, OBST # 73

Please fix this. You can't apply a burner to a thin obstruction. Then, we are about to release 6.2.0 (we have been testing bundles since Friday, hopefully release in the next day or so). Please wait and run your new case (no thin OBST with SURF) with 6.2.0 and if that does not work submit an Issue. Thanks.

--
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/b162ce47-bde2-43de-9abb-cb747a28208a%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Torbjørn Knutson

unread,

Apr 16, 2015, 12:53:13 PM4/16/15

to fds...@googlegroups.com

Hello again.

Thanks for your feedback, guys!

I thickened the slab with the burners. Also, seeing as I'm pretty sure it's the HVAC that's causing all my problems, I removed the HVAC system and used two surfaces with a total volume flux of 15,3 m3/s and a "fresh air-wall" with -15,3 m3/s. Upon the error: "cannot attach velocity boundary to thin obstruction", I thickened these slabs as well, but even with a thickness of 1,0 meter, the error message remains!

Any suggestions as to what I'm doing wrong?

Torbjørn

mandag 13. april 2015 14.55.22 UTC+2 skrev Randy McDermott følgende:

Running your input file with the latest code I get the following message:

ERROR: SURF burner cannot be applied to a thin obstruction, OBST # 73

Please fix this. You can't apply a burner to a thin obstruction. Then, we are about to release 6.2.0 (we have been testing bundles since Friday, hopefully release in the next day or so). Please wait and run your new case (no thin OBST with SURF) with 6.2.0 and if that does not work submit an Issue. Thanks.

On Mon, Apr 13, 2015 at 7:21 AM, Torbjørn Knutson <torbjorn...@gmail.com> wrote:

Hi.

I'm having a problem with "numerical instability". I thought it had something to do with too much turbulence in the model's air-/smokeflow, but whatever I do it still stops after about 80-90 seconds (out of 300 s)

My efforts to fix it:
- Split meshes (till a total of 4 meshes)
- Increase vent sizes (air out) for HVAC
- Incresae vent size (air out) for smoke hatchet
- Increase vent size (air in) for HVAC/smoke hatchet
- Decrease flow rate for smoke hatchet
- Change location of vent (air in), further away from smoke hatchet
- Decrease number of fires (burners) from 2x350kW to 1x700kW
- Remove smoke hatchet completely

I've added both the FDS- and the Pyrosim-file. I really hope someone can help me with this.

Kind regards
Torbjørn

--
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+unsubscribe@googlegroups.com.

ss_13_simple.fds

ss_13_simple.psm

dr_jfloyd

unread,

Apr 17, 2015, 10:41:21 AM4/17/15

to fds...@googlegroups.com

The exact error message is:

ERROR: SURF Slanted roof, vent cannot be applied to a thin obstruction, OBST #83

ERROR: SURF Slanted roof, vent cannot be applied to a thin obstruction, OBST # 201

The 83rd OBST is:

&OBST XB=26.082,26.24,22.3987,22.5523,9.5,9.5, SURF_ID='Slanted roof, vent'/ Obstruction

The 201st OBST is:

&OBST XB=39.354,39.512,20.7083,20.862,11.8333,11.8333, SURF_ID='Slanted roof, vent'/ Obstruction

These are thin obstructions.

To unsubscribe from this group and stop receiving emails from it, send an email to fds-smv+u...@googlegroups.com.

陳佑任

unread,

Apr 17, 2015, 11:28:31 AM4/17/15

to fds...@googlegroups.com

How about your MESH ?

從我的 iPad 傳送

To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/8b0300a0-f5c3-474e-a61a-af0771f433e6%40googlegroups.com.

Torbjørn Knutson

unread,

Apr 21, 2015, 8:11:51 AM4/21/15

to fds...@googlegroups.com

Hi.

You are thinking a different mesh might help? I'm open to any suggestions.

Kind regards

Torbjørn

--
You received this message because you are subscribed to a topic in the Google Groups "FDS and Smokeview Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/fds-smv/OVvb5hB9jyo/unsubscribe.
To unsubscribe from this group and all its topics, 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/B3AEE984-7C2C-4946-9602-CFE9C9B0287F%40gmail.com.

Randy McDermott

unread,

Apr 21, 2015, 8:14:01 AM4/21/15

to FDS and Smokeview Discussions

Jason's suggestion was to remove the thin obstructions. Have you done this?

To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/CAF_M1HrPOboYkMLzgsKKLb2YrD_GoX-_AF6h7bSY%3DxueCqRBkg%40mail.gmail.com.

Torbjørn Knutson

unread,

Apr 21, 2015, 8:16:58 AM4/21/15

to fds...@googlegroups.com

Yes, but now the numerical instability-error is back.

To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/CAC6wax3gtKiLPYzwZLYsCufDFOcfYE%3Dci32QoRUG1sU%3Dzv624w%40mail.gmail.com.

Randy McDermott

unread,

Apr 21, 2015, 8:18:12 AM4/21/15

to FDS and Smokeview Discussions

If this occurs with 6.2.0, then please simplify your case as much as possible and submit an Issue. Thanks.

To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/CAF_M1HoA%3D17vtNWi8%2BmWn2E_oabvx7k4XM9UftvomKmqGWTooQ%40mail.gmail.com.

Connor McIntosh

unread,

Apr 29, 2015, 9:37:27 AM4/29/15

to fds...@googlegroups.com

Hi Torbjorn,

I had a similar issue some time back. I solved my issue through changing the pressure iterations and refining my mesh, at the expense of some computing time.

C

Torbjørn Knutson

unread,

May 4, 2015, 11:23:42 AM5/4/15

to fds...@googlegroups.com

I got it to work.

Thanks for all your help.

Reply all

Reply to author

Forward