Inquiry usage of VI-SUIT | BlenderBim | IFC File | Glare studies | WIndow double face for DL studies

[Ole Marius Svendsen]

As requested, i could send some question here, and i am sure this also could benefit more people to have read some of the questions i want to bring up.

First question i would like to ask. When running either or of the analysis, like daylight factor, sDA, or glare. Will the computer run each simulation grid simultaneously per logical processors? I am working professionally where this is a high need. I can sit with well over 50 rooms where i need to have the results quick. If the VI Suits run one room at a day, it can take a rather long time to render the results. Per now i use both IDA ICE, and DL light which allows to run the grids per tread i have in my computer, for my case i have 31 processors i would like to utilize.

[Ole Marius Svendsen]

To continue with the message i want to write in the initial post (length restriction makes me have to devide this in two): 

For the glare analysis. We have in the BREEAM manual a glare value of 5% discomfort through the year. Can your glare analysis only simulate one spesific hour at a time? If so, i can not see how i can use the glare tool, as i can not for curtain tell at what time of day nor date to accumulate the requirement. Climate studio have made a good tool for this where each grid has a circle where it calculate the tresshold value for each hour of the entire year + in the view of the sub circles. Pictures attached under. With this tool, i can see the potential to find the actual 5% requirement. 

 

[Ole Marius Svendsen]

Next (3rd as the lenght recuirment here is rediculus).

I can not add the picture, so i would have to show climate studio, and its glare analysis.
https://www.solemma.com/climatestudio

[Ole Marius Svendsen]

Forth message with this nonsens of the lenght requirment:

next question. How do you see your VI SUit can work with the blenderBIM tool? With the blenderBIM i can import a IFC file from the architect firm, and thus the geometry to perform daylight analysis is there. In theory, should it not be then easy to add reflection values for each main surface, such as ex. floor = 20 %, wall = 50 %, ceiling = 70%? Have you though about any overlapping usage here? I do find it a bit diffecult to know for sure if i have added the correct reflection factor on all surfaces, both inside and outside. Can it be a possible feature to one day show reflection by colors? IDA ICE does this, and this makes a peerreview of a model far easier to go trough.

[Ole Marius Svendsen]

Fifth (yet again, the lnght requirements makes this a hassel to write:

Furthermore as i can see with both the climate studio and DL light, the transparency factor for windows is depending on manual deleting the one of window's two surface (exterior and interior). If you keep both surfaces then the calculation will be faulty as light need to pass two surfaces with ex. transparency factor 70 %. I know blenderbim can by clicking quanteties of windows get all of the object and merge surfaces close to 1 cm as an example. WOuld this be a needed quicktool that should be implemented if importing an IFC file? It would be smart to stream line any user errors such as forgetting the two faces for window objects. 

On Friday, 27 October 2023 at 14:51:56 UTC+2 Ole Marius Svendsen wrote:

[Ole Marius Svendsen]

So to summerise, in the image i am attatching, this is want i wanted to write as a long post.

On Friday, 27 October 2023 at 14:51:56 UTC+2 Ole Marius Svendsen wrote:

[VI-Suite]

Hi.

I didn't even know there was a length limit here. Other people have written longer messages on here so not sure why that is.

Numerical calculations are multi-core. 

Glare analysis is currently based on a Radiance generated image and that image can be generated multi-core on OS X and linux but not on windows. The glare functionality you seem to be looking for is from the new dcglare component of Radiance, and I have not implemented that yet. v0.8 maybe.

I don't use Blender BIM but if it exposes whether a material associated with a mesh face is a floor/roof/window then you can write a small Python script to set the lighting characteristics based on that designation. The parameters you would look to change are:

blender_material.vi_params.radmatmenu (e.g. setting this to 'plastic' would set a radiance plastic material )

and then the properties of that material type e.g. blender_material.vi_params.radtrans would set the transmission of a glass material. If you turn on Python tooltips in Blender it will show you the Python address of any menu.

Blender is a mesh modeller and the VI-Suite works based on that mesh. Imported IFC data therefore has to be converted into a mesh for the VI-Suite and how, or how well, Blender BIM does that I don't know. For the VI-Suite what mesh geometry you see in the 3D viewport is the geometry you can simulate.

Hope that helps.

Ryan

[Ole Marius Svendsen]

Thanks for the answer.

I do think the  dcglare is needed. Only Climate studio seem to offer it, so no other tool can be used for glare and BREEAM.

For material with BlenderBim, when importing an IFC file, the window as an example is a block. Within that block you have a material for the frame, and for glas. Here it is then possible to use directly the predefined IFC material. Though, how would you come about selecting the sub-material with the current V0.8? 

In general would it be possible for a V0.8 use maby a different standard name beside plastic? It can be misleading. Also to add your own material and save it with the correct name (what image it has as a layer is not relevant for these studies), with its correct light reflection value seem to be a bit harder. Is it the roughness that is the light reflection value i want to use?

For the mesh with IFC, it seems to be working well. You have en each component faces, vertex etc which you can edit. Though again all of the objects are locked, so you need to get into the edit panel for wall, window, floor etc.

[Ryan Southall]

I'm afraid I'm not familiar with IFC, so I don't really understand the material designation you get with IFC. Ultimately, if you have a valid Blender mesh, with Blender materials associated with the faces, then these materials can be given the lighting characteristics you deem appropriate e.g. for a single pane of glass of double panes of glass. 

'Plastic' is a Radiance material class, and the VI-Suite follows the Radiance material class designations, but you can obviously call the material whatever you want in the VI-Suite and you can create your own materials with their own properties. It is maybe easier if you try it yourself and this video https://youtu.be/9YVPVPgmjQ4 I did for my students explains the process of lighting analysis with the LiVi component of the VI-Suite. 

Cheers

Ryan

[Ole Marius Svendsen]

Thank you, will take a good look into the video. Parden difficult question, appreciate you taking some time.

[Lars O. Grobe]

Hi,

speaking of glare (and maybe other luminance-based assessments) - it would be cool to have an interface to raytraverse sampling, but that would be a rather significant development effort...

Best, Lars.

Am 27.10.23 um 15:35 schrieb VI-Suite:

--
You received this message because you are subscribed to the Google Groups "VI-Suite" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vi-suite+u...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/vi-suite/ea7fd408-715a-49b2-9334-6959ac248b25n%40googlegroups.com.

[Ryan Southall]

If ray traversing is a just an additional step in a conventional radiance command sequence, it may not be too bad. 

I have now a basic dcglare implementation, although not yet committed to github. Is ray traversing a significant improvement over that?

To view this discussion on the web, visit https://groups.google.com/d/msgid/vi-suite/361b66e7-53ab-4509-a32b-c6f65a090545%40gmx.net.

[Lars O. Grobe]

Hi Ryan,

I guess it would be a slightly different module - it implies installing the raytraverse package, passing evaluation zones (e.g. floor polygons), and processing of zonal daylight metrics.

So far, it has not been implemented in any GUI, but it would offer an opportunity to overcome the limitation of luminance-based metrics to entire building zones without sacrifying accuracy and speed. Basically, it aims at reducing the number of rays needed to evaluate glare, e.g. to image regions with steep gradient - it is "imageless" in that it renders sparse rays, but without losing contrast information. This aims to avoid assumptions such as glare being correlated to high eye illuminance (which is not the case for high-contrast situations in rather dark environments) or glare sources occuring only in the visible sky. The method is described in several publications listed here:

https://data.snf.ch/grants/grant/179067

I have changed the subject, since this would imply some significant effort. I was just reacting since the objective of such glare evaluations was stated.

Best, Lars.

Am 31.10.23 um 18:03 schrieb Ryan Southall:

[Ole Marius Svendsen]

Evening, i have taken time to review the Youtube clip. The workflow seems fairly okey to replicate, and also i can indeed see overlapping methodology by using BlenderBim for geometry usage. When it comes to compiling results i have some further questions.

With the node-based workflow, how would you come about simulating 10 rooms in one go? Would the Livi Geometry by default simulate all 10 rooms, or is this tool only ment for one room at a time? 

When it comes to showcase the results, in your video it gives DF for one room, however how would this be when applying the simulation to several zones? At most at my work we simulate Maby 100 at most, where we most of the time are only interested in the numerical value. Would the last node you used to show the numerical value not be a bit limited as it shows a picture, insted of making a table of the results?

Another element i would like to ask. When compiling the graphical results, each box with its normal shows the value. Is there a function to transform the graphical results into lines of values with interpolation between them? Much like whats shown in the image below:

One of the more interesting usages here, is that the graph on its left side can be dragged to lock under/above value and change the graphical view, like the image under. Here we wanted to see where 1,1% average daylight ends, as a mean to estimate where working zones could be located. Would this be hard to compile in a potential new realese?

On Friday, 27 October 2023 at 16:14:49 UTC+2 rgsou...@gmail.com wrote:

[VI-Suite]

You can have multiple objects with the light sensor for multiple rooms.

I have experimented with interpolation but never liked the look of it so have not included it up to now.

To get a coloured cutoff set the legend levels to 2 and then, for your example, set the legend max to 1.2 and the legend min to 1.0.

[Ole Marius Svendsen]

Good afternoon. Again, thank you for your reply. Would it be considered unpolite if i asked for the interpolation feature, or other aspects? I am aware you are doing this project whenever you have time, so i dont want to come across as ungreatful.

I did a test. Imported an IFC file. Seemed rather easy to get hold of all materials as many of the building materials has it by default. I made a grid for a test room and made sure to assign the plastic + its reflectance value. I also made sure that the window with its two faces is merge into one to assure you do not get two layers of light transmission calculations.

However, it seem for me that when running the simulation, no material nor shading is considered. Would you have an idea as to why? Would it be to the nature that all walls, windows etc. are blocks, and VI-Suit neglect merged objects?

[VI-Suite]

My first step in debugging a scene is to create a Blender camera, position it to view the scene, select the camera in the LiVi Simulation node and press 'Preview'. This will open up a Radiance rvu window that allows you to see the scene from the perspective of Radiance. You may have to alter the exposure of the view if it's all white. I would also recommend you start by creating your own simple scene to get a feel for it.

It looks like the walls in your import are not being exported to Radiance so check that you have applied the desired Blender materials to the geometry faces. It is a purple material that seems to be applied to the walls. Check this Blender material has the correct LiVi material characteristics.

As you are now using the VI-Suite, make sure you are familiar with the user manual https://blogs.brighton.ac.uk/visuite/documentation/.

It is not impolite to ask for features, as it is also not impolite if I do not say yes. Interpolation is quite a bit of effort for only a change in the appearance of results, not in the results themselves. As such a low priority for me. There is probably a way to do it with Blender's geometry/simulation nodes as LiVi results are stored as attributes within the result geometry. If I figure out this way, I will post here.

Ryan

[Ole Marius Svendsen]

Thank you for your guidance. I have played with the preview function, and i can confirm it works when i make a blank mesh. However, when i add a predefined window from the IFC creation tool, next to the blender box, and assign some materials into the frame, and glass, none of these meshes will be visible with its preview function. Regardles of what i compare the box, with the window, it seem to have all the same settings for blender materials.

I have been reading your manual, and for me it seems like i am following the steps to assign correct values within the VI-Suit options. May i ask for assistance. If you open the same test file i have made, are you able to activate the VI-Suit material? If not, would you have a suggestion towards what i can do to progress further?

-

[VI-Suite]

Ahhhh. You have a forward slash in the name of the object, which is not a legal character for LiVi simulations. Never thought anyone would put a forward slash in an object name, but if that is what Blenderbim produces I'll put a warning message in the vi-suite-log file.

[Ole Marius Svendsen]

I am currently out running an errand with my wife. I will as soon as I get home test changing the name. Though is it possible to have the "/" as a legal name for your plug-in? As I want to import a IFC model, all 1000s of objects will contain this character. To change all of the elements sounds like a hinder for many to use VI-suit.

--
You received this message because you are subscribed to the Google Groups "VI-Suite" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vi-suite+u...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/vi-suite/2ddb36db-129f-4a6e-9255-292c55d68d4en%40googlegroups.com.

[Ole Marius Svendsen]

I am happy to share with you, that changing the "/" did activate VI-Suit materials.

I have done a test to compare the results with a current project where i have used IDA ICE and its daylight module with VI Suit. I have some questions.

I have taken the same room, with same room geometry, and surface reflection. Both models have light transmission value of 70 % for the window.

With the VI-Suit preview, i do belive i have now a functioning DL model:

However, after comparing the results i was expecting somewhat same results as i do in IDA ICE, which is my main software and i do consider myself an advance user of it with my current experience.

For VI Suit, i have made sure that the glas material is Transmission with a value of 0,7, which then should be the same as IDA ICE 70 %. I have made sure to delete the extra surface which comes native when opening, or editing an IFC file. This should mean that there is not two layers of 0,7 transmission layers. Still with the results i have, there is a difference on 3,5 in numerical average between the two models. Hight of window is the same, same with sill Hight and width. I have the other layer needed from the obstruction in front of the window. The two model has the same high of the calculation field.  

What would you advice that could be the error i have in VI-Suit? My only though is that somehow i have inaccurate usage for the window glas property.

[VI-Suite]

Make sure you are using higher accuracy and your test file from before was set to transmission in the Trans type material option, so you were modelling the window transmission rgb colours you set i.e. blue. Set Trans type to transmittance if you want to use the transmittance value.

[Ole Marius Svendsen]

High accuracy has been used for the big IFC file. 

For the transmission value and its rgv colors, i do belive i have had it accurate:

[Ryan Southall]

It would be if the dot was in the middle but you have a 70% blue, not a 70% white. 

[Ole Marius Svendsen]

Thank you so much for this feedback. Was not even aware it is two Components that can interfare with each other. If i recall your manual, i dont think this is highlighted enough as a source of error.

I am delighted that i am not getting much closer with my results. After tweaking the Ida Ice modell to make sure all imported object has its right placement (and changed one reflection Component), with VI-Suit i have the following results as shown underneath. Yet there is a roughly a 15 % difference between the two softwares. I am emotionally invested now to learn your plugin well, so pardon if i intrude and extend this tread for such a long time.



As i have a 15 % difference This led me to a new curiosity. With the same as the glas material, the plastic material option has 1) Roughness, and 2) Specularity. It seems also as these can change the overall results. Is this accurate? And may i once more ask a question, how big impact do these settings have, and for when should either one be tweaked? What would the recommended value be for both?

[VI-Suite]

Those are parameters of the Radiance Plastic material and there are plenty of online resources about the meaning of Radiance material parameters. Whether a pair of values is accurate depends on whether you have accurately captured the nature of the real materials in your scene.

[Ole Marius Svendsen]

Evening Ryan, i will do a deep dive into reading your science paper and other papers i can find. I would like to express gratitude for the guidance you have given. I hope that me asking questions over a longer duration is not too intruding. Even my employer has emphasized it can be considered as a fully functioning tool in our repertoire. 

I think as soon as i have done some reading, and further testing between IDA ICE and VI-Suit, i would like to share my experience in the OSarch community forum and my own LinkedIn. I hope you will receive my interest and wish to inform more of VI-Suit in a positive manner.

I do think this tread can be seen as partially resolved with regards to all the initial question i brought up, and its shift of focus into daylight simulation.

Lastly Ryan, based on my novice hands on attempt to use your tool, may i ask if i can forward some minor featurs (not featurs simelar as the initial question on interpolation view of results)? If yes, where can i send them too, in case i am overextending the intended purpose for this google group.

[VI-Suite]

Before you start posting about the VI-Suite anywhere I'd be grateful if you started spelling it correctly.

Feature requests can go here but open a separate thread.

Good luck with your testing.

Ryan

[Ole Marius Svendsen]

Morning Ryan, i will make sure of that. Appolegize for misspelling.