FDS data visualization in Blender (with demo) – yes or no?

908 views
Skip to first unread message

Ian Fong

unread,
Apr 22, 2022, 9:45:29 AM4/22/22
to FDS and Smokeview Discussions

Hi everyone, new to the community here. 

I'm usually not much of a poster, but I have found a workflow to input fds data into Blender for data visualization. Initially I thought that this would be useful to many fds users, but after talking to many people I know, I'm second guessing myself. So I’d like to show you some demos(the demos are rendered at a low sample rate) and ask the people themselves whether this sort of thing is needed. 

My most recent demo I have made is the fds tutorial ‘hello world’ example; with the velocity mapped onto volumised soot density as a colour ramp. (ran out of time to do the full render - I wanted about 4-5 sec of footage).

https://drive.google.com/file/d/17wh0-7AnxodBtPJoo48Otv-qRCDbn1OC/view?usp=sharing

- not sure when google drive is gonna finish processing this one.

You can theoretically map most output as either volume or colour value => e.g. volumise temperature and colour it according to a density field; along with various other pulls which i'm just starting to explore.

For my initial render with just smoke; see below

https://drive.google.com/file/d/15DJY5oc8u7pVt0uGoPUlGViPgUy6Ehr5/view?usp=sharing

I have also recreated Jin’s 1970’s experiments within blender and found a correlation between Blender’s input density and Jin’s theoretical model. I have then used this to recreate the VTT 2007 experiments as seen below at 1 frame per second. (skip to around 1 min for the action). 

https://drive.google.com/file/d/1XkFA6jvzoCkYMxJbehDgQSYZSWyupjW4/view?usp=sharing

Id imagine that this could be useful to fire scientists, to find correlations and discrepancies in data. Also, id imagine this could be used in court reconstructions for fire cases to communicate effectively to jurors. Initially this was made for engineers to do aset/rset calculations (visibility criteria), but I haven't got a great response here... Also, this can be used to show off you’re doing something fancy. 

 So does the community need something like this? At the moment, the workflow is not very accessible (it requires piping data through a few software), so if there are enough people who need it, I can try to increase the accessibility of the workflow through trying to get direct data outputs from fds to be read by Blender; no promises though. 

Note: you also might need a decent graphics card and cpu – otherwise you have to skimp on a lot of things and wait a long time like me. Not sure what performance you get on a “good” computer though 

 

Cheers. 

wjxylsl

unread,
Apr 24, 2022, 3:29:33 AM4/24/22
to FDS and Smokeview Discussions
Good idea!

Emanuele Gissi

unread,
Apr 24, 2022, 4:48:30 AM4/24/22
to FDS and Smokeview Discussions
Hi Ian, what you did is great, I am very interested in learning your workflow and integrating it into BlenderFDS (https://blenderfds.org/).

I am not a full time developer (I am a full time fire chief), and I am currently busy in completing the port of BlenderFDS to Blender 3.
Then I want to add the DEM interpolation feature to qgis2fds (https://github.com/firetools/qgis2fds).
After that I am available in assisting you.

Is anyone else of this Community willing to collaborate?

Best,
Emanuele

Babak Bahrani

unread,
May 17, 2022, 2:58:50 PM5/17/22
to FDS and Smokeview Discussions
Hi Emanuele,

We had a previous discussion during the NIST Large Outdoor Fires Workshop in 2018 in D.C. regarding the new guideline (similar to FDS 5) for FDS 6.
I am willing to collaborate with the BlenderFDS project. 

Please let me know the next steps.

Best,
Babak

Ian Fong

unread,
May 17, 2022, 8:42:12 PM5/17/22
to fds...@googlegroups.com
Hi Babak,

I haven't really released much information on my workflow for a number of reasons, but I think I have found a middleground in getting it out there. I'm currently in contact with a number of people who are trying to streamline the process for anyone interested in the workflow. ie making it free. As such, I don't think it would hurt to just put the aims of the project online for free so it's out there.

The goal is to leverage blender's relatively new ability to read VDB files - and as such to streamline the process; we want fds to directly output vdb files. This can largely be achieved and implemented using the openVDB library i believe.

This is probably the first hurdle to overcome to make the solution accessible. at that point; having a fully blender workflow would be making blenderfds recognise how to incorporate the new writer into fds code.

Also a general update and disclaimer: i've been in contact with some people regarding smoke visibility research - i'd hold off on using these visualisations for physically accurate smoke. When their findings are released - I'm sure we can incorporate specific wavelength absorption by smoke using blender's open shading language and create even more accurate smoke! after all, you can theoretically make a shader incorporating the beer lambert law with it.

Also, notwithstanding, you could theoretically get vfx crowd sims to work with the data - allowing simulations that interact with properties like temperature (not sure if fds+EVAC allows for things like this). However, i feel as though blender doesn't have that good crowd simulation capabilities at the moment - it'd have to be offloaded to a different program that can read vdb and bake the crowd simulation into a file blender can read - there are a number of programs capable of this, one in particular.

cheers

Ian Fong

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/601d46f6-0b27-4c35-af19-8a71abfad039n%40googlegroups.com.

Emanuele Gissi

unread,
May 18, 2022, 4:55:57 AM5/18/22
to FDS and Smokeview Discussions
Dear all, I am enthusiastic about these great ideas you had.
I propose you the following plan of action.

I am finalizing the release of the new BlenderFDS, hopefully in a few weeks.
I have to finish testing qgis2fds new interpolation feature and release it.
So I suggest having an online meeting the first week of July, for brainstorming and demo of current workflows.

I set up a doodle to choose together the best date:

The meeting is open to anybody willing to contribute or just get informed about the possibilities.

Thank you,
Emanuele

Emanuele Gissi

unread,
May 24, 2022, 1:53:40 PM5/24/22
to FDS and Smokeview Discussions
According to the availabilities you noted on the Doodle, I set the online meeting on Tuesday 5th July at 3pm CEST (9am EDT, 11pm AEST sorry).
Thank you very much! 

Emanuele Gissi

unread,
May 24, 2022, 1:55:09 PM5/24/22
to FDS and Smokeview Discussions
I sent you the Google Meet link.
If anyone else is willing to participate just answer to this message and I'll add an invitation.

MEDIENE YOUCEF

unread,
Mar 3, 2023, 5:48:00 AM3/3/23
to FDS and Smokeview Discussions
Hello everyone I'm a student in fluid mechanics and I'm working on a mini project on FDS. Basically I wanted to export the data on rhino 7 then on enscape but I can't. your work is fantastic and I would like if possible to have access to the workflow to export to blender. Thank you

Ian Fong

unread,
Mar 3, 2023, 5:59:27 AM3/3/23
to fds...@googlegroups.com
Hi Youcef,

Funny you should mention that, I'm actually making a video/tutorial outlining my process. Not sure where/when i'll post it but it'll be out soonish. ill link the video onto this page.

good luck with your project.

MEDIENE YOUCEF

unread,
Mar 3, 2023, 9:29:12 AM3/3/23
to FDS and Smokeview Discussions
well thank you very much

Ian Fong

unread,
Mar 4, 2023, 8:04:38 PM3/4/23
to FDS and Smokeview Discussions

if you have further questions, feel free to ask, i may have skimped over some of the finer details.

MEDIENE YOUCEF

unread,
Mar 6, 2023, 4:46:46 AM3/6/23
to fds...@googlegroups.com
Thank you very much for your tutorial. He is really helpful. I just have a question. How to change names and organize .q files? Thanks again for the amazing tutorial.

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 view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/71740536-9c1c-4723-8a6e-75079adef759n%40googlegroups.com.

Ian Fong

unread,
Mar 6, 2023, 2:40:34 PM3/6/23
to fds...@googlegroups.com
Thanks. I just renamed them manually by selecting the w files in orders of 10-1 (1-9, 10-99-, 100-999 etc) and just renaming them with the beginning number. If your up for the challenge, you could probably write a .bat file to rename them… or maybe try the chatGPT thing to help you write one if you don’t mind AI. 
To unsubscribe from this group and stop receiving emails from it, send an email to fds-smv+unsubscribe@googlegroups.com.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/CADfrUfWdiHMG-t2vqyd8AGe4V50e%2Bf8BPsE96fpws1RMxc0BEg%40mail.gmail.com.

MEDIENE YOUCEF

unread,
Mar 7, 2023, 4:26:30 AM3/7/23
to fds...@googlegroups.com
Hi it's me again, I'm sorry for the inconvenience, I was able to follow all the steps except that at the end I don't see the rendering on blender, I used the simple_case case in the fds examples. once again sorry for the inconvenience


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

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+u...@googlegroups.com.

--
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 view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/CAMaq3Rc6mK2xF_Fsgp2QGxSiMwph9Picef3cKzRBXEwrOyaPuA%40mail.gmail.com.
Capture d’écran 2023-03-07 102450.jpg
Capture d’écran 2023-03-07 102350.jpg

Ian Fong

unread,
Mar 7, 2023, 7:51:14 PM3/7/23
to fds...@googlegroups.com
It could be a number of things. It could be that your in material preview and not the IPR(render view). Have you checked to see if Houdini gives you output? Does smokeview give you your output? Are you still on frame 1 in your animation in blender?

They are just questions on the top of my head. I would need more information/screenshots to help-
To unsubscribe from this group and stop receiving emails from it, send an email to fds-smv+unsubscribe@googlegroups.com.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+unsubscribe@googlegroups.com.

--
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.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/CADfrUfWNdh0h2OJZn_vDp2JBjH6bJASS2gC2q_cYThfudb3xCA%40mail.gmail.com.

MEDIENE YOUCEF

unread,
Mar 8, 2023, 5:23:14 AM3/8/23
to fds...@googlegroups.com
Hello,
Smokeview gave me the results that I was able to verify on paraview, I was also able to verify that on houdini in which I was able to import the csv data and then redo the manipulation as in your video where I converted to vdb by taking care to put the right size of voxels and subtracting the csv-00000, I was able to save in vdb and I had 240 .vdb file. On blender I imported the vdb and by parameterizing as in the video I could not see anything (I checked frame by frame and in the render view). I did not recreate the geometry of the simulation and I just put a light source and a camera.
here are some screenshots of my settings in Houdini
Thank you for your help

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

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+u...@googlegroups.com.

--
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.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+u...@googlegroups.com.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+u...@googlegroups.com.
substract.jpg
file cach.jpg
vdbfromparticlewarning.jpg
particlevdb.jpg
tableimport.jpg

Ian Fong

unread,
Mar 8, 2023, 6:16:41 AM3/8/23
to fds...@googlegroups.com
something isn't right with your vdb from particles node in houdini. If you look at the output in vdbfromparticleswarning.jpg, there is one VDB contained as "surface" i.e. there are no attributes called density and flame in the vdb - hence blender is not able to read density or flame - they simply are not in your vdb; hence no output. What did you do for your empty vdb?

MEDIENE YOUCEF

unread,
Mar 8, 2023, 7:50:21 AM3/8/23
to fds...@googlegroups.com
thank you now i can see something in blender but it is not really the expected result. i am attaching the screenshots 

vdb1.jpg
vdb.jpg
render.jpg

Ian Fong

unread,
Mar 8, 2023, 5:47:16 PM3/8/23
to fds...@googlegroups.com
Hmm, what’s not expected? Visually it looks like your using eevee to render which is possible, but you might need a different shading setup for it to look good. Use cycles imo. If it’s the output that’s not right, you may need to rotate your vdb in blender to get the right orientation. Otherwise, I’d recommend checking your output in the scene view at every step in Houdini, just to make sure your getting the intended result.
To unsubscribe from this group and stop receiving emails from it, send an email to fds-smv+unsubscribe@googlegroups.com.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+unsubscribe@googlegroups.com.

--
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.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+unsubscribe@googlegroups.com.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+unsubscribe@googlegroups.com.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+unsubscribe@googlegroups.com.

--
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.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/CADfrUfXZYGPGD1i-4Gxr95ZpbjQtc7jgAYDv7u337nV8Ky8zXQ%40mail.gmail.com.

MEDIENE YOUCEF

unread,
Mar 8, 2023, 5:53:12 PM3/8/23
to fds...@googlegroups.com
yes, I started again step by step and I was able to get a fairly good result, but it could be greatly improved. I thank you very much for your help

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

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+u...@googlegroups.com.

--
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.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+u...@googlegroups.com.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+u...@googlegroups.com.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+u...@googlegroups.com.

--
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.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+u...@googlegroups.com.

--
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 view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/CAMaq3Rf%3DTtwhDj4RHoZgd1pCEn3hSv4_HmtXdyQfCiKJ7oc%2BbQ%40mail.gmail.com.

Kartini Ganesha

unread,
Jul 21, 2023, 10:30:28 AM7/21/23
to fds...@googlegroups.com
Hello Ian

I'm trying to follow all of your instructions. I got stuck where I have more than one .XYZ file and I tried to save the extract csv file. It runs very slow, do you have any suggestions?

Thanks  

Ian Fong

unread,
Jul 21, 2023, 9:26:58 PM7/21/23
to FDS and Smokeview Discussions
Hey Kartini,

It depends on what your trying to do with the file.

if its for a proof of concept, maybe try reducing your mesh density in fds? but that will require you to run your simulation again.

if you have the budget, you could upgrade your cpu?

Otherwise, if your limited, id suggest you just wait it out. while developing the workflow, i had to wait approx 45 mins each run - but my computer was a potato back then.

i don't know your computer specs or what you are trying to simulate, but if it's struggling with this, you may also struggle with the blender part; it uses volumetric data which is taxing on the gpu/cpu. (not undo able, just you have to wait very long times for loading)

There may be an alternate solution on the horizon; where you can get your vdb directly - but I'm not sure of the details.

regards,

Ian Fong

Kartini Ganesha

unread,
Jul 23, 2023, 9:10:17 AM7/23/23
to fds...@googlegroups.com
Hello Ian, 

Thanks for the explanation. I utilize AMD Ryzen 7 6800, GPU RTX 3060, and RAM 16 GB. I thought it would be fine , but yeah I have to upgrade my computer. That's make sense to me. Considering my simulation time is 15 minutes, and the code generates plot3d 25 files / second with almost 300k meshes obviously it will go crazy. I will rerun with reduced mesh to see to logic. Thanks and good day for you

Kartini Ganesha

unread,
Jul 26, 2023, 10:16:26 AM7/26/23
to fds...@googlegroups.com
Hello Ian

I'm back again with a question. 

When I tried to connect between the imported table with vdbfromparticle, the vdbfromparticle cannot connect to the data from importtable. Do you know how to resolve it? For the additional note, I'm able to visualize the table data. 
IMG20230726160741.jpg

Ian Fong

unread,
Jul 26, 2023, 4:36:44 PM7/26/23
to fds...@googlegroups.com
In your vdb from particles node, I think you need to add attributes that you want the vdb to hold to your vdb.
To unsubscribe from this group and stop receiving emails from it, send an email to fds-smv+unsubscribe@googlegroups.com.

--
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/FjV1JuvnkHo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/CAGjs4KA4oHqueLRRuSmdUG7mV8dA28dw4s0UsS30emgawVNyoA%40mail.gmail.com.
Reply all
Reply to author
Forward
0 new messages