Hey Philipp, is there a reason for which you picked populate3d to create the 3d points than to make it a consistent grid? Not sure if that makes any change because now I'm struggling with my code being stuck into a massive setvars batch loop
There is no particular reason other than showing that probe points can be arranged not just as a grid but in any configuration in 3D space. The grid definitely has its advantages in most cases.
Not sure how... Also be aware that butterfly only shows a moment in time not a sequence. Simulations which show cause and effect might not as straight forward.
Hi @PhilippGalvanDesign, nice to see you are still active even on your older videos. I am currently trying to design a kind of "oven" that has to be put on top of an open fire. For this, I want to research if certain holes (size and locations) and shapes inside would improve the airflow. The goal would be to regulate where the hot air circulates to regulate the temperature in the oven. Do you have any tips? For now, I just keep watching your videos and try to apply them to my topic! Btw I would really like to thank you for uploading these! Without your videos, I would have never considered researching this using butterfly.
@@thibotytgat3980 hey glad to see you using my videos. I think you might want to look at one of my later videos about indoor airflow and heat transfer. Not sure if you can simulate and oven (maybe) but for sure it can give some insights. I would probably suggest to also use other testing methods, including logic and common sense to make sure it really works as calculated. If you work on a real prototype you probably even want to create mockups. Good luck and let us know where you got to.
hello Mr phillip, thanks for the tutorial, the solution to remove the very high values is by using mesh inclusion and cull pattern to select only the points outside the geo, thanks again
I started with Wind rose as well, but then I listed the vector in the Wind Boundary Profile. I think that would be a good way to get the direction based on the location. Also for the velocity, the wind profile has calculated it base on different height, although it's better to take the velocity at reference height 10 m.
Hey Philipp, first of all, thanks for sharing your knowledge. I have a question for you. I want to calculate the pressure on trees located in roof garden, so that I can use the suitable anchors. which software is the best?
To be honest, if this is a real project I would hire a company who can do a real wind tunnel test with a scaled or real size model. Other than that I would say Butterfly I's a good option as long the parameters are set properly. You can locally measure the pressure. Just a note of caution: if you are thinking of adding an actual model of a tree you need to be careful. It will be very hard and time consuming to calculate a snappy hexmesh for that... Plants are tricky in CFD analysis. They are better calculated as porou massing but for that I haven't got my head around in Butterfly CFD. But Eddy 3d has this function. See here: ua-cam.com/video/vPuB74am6cM/v-deo.html
Hey Philipp, I am unable to hide the mesh and the snappy mesh to just view the vectors and the buildings, how to do that? Thank you for the video btw :D
Every component in GH can be shown or hidden. Right click onto the component and turn off preview. Also the mesh component can be turned off completely if you don't need it. Any geometry inside rhino itself need to hidden there. I suggest to watch a GH intro video to understand Preview etc... Thanks for watching.
5:32 I've resolved this by silly rotation of the whole model xD But it seems like this issue has appeared not so long ago and there are no solutions for it yet, except updating the Rhino :(
Yeah, I agree i think it's a bug in the code. It's a bit of a shame that such a complex system fails on a relatively simple issue. But, let's see it as a chance to work together as a community and to solve it. That's the whole point of this channel. :) Thanks for taking the time to test the issue.
@@PhilippGalvanDesign right! By the way, since i am here, i have noticed a very crucial thing: i have tried to calculate several simulations in a single night using a datatree with different _cases in each and didn't manage with it, because it turned out, that the Solution node is constantly refreshes itself, so it never reaches the second branch and infinitly computes the first one. So, in order to avoid this, it is required to set "_interval" to "-1". The another important thing i've discovered -- there is a built-in butterfly node with skipped probes (the points inside the building, which were not calculated properly), but this doesn't resolve the problem until the end: you would have to detect and weed out the corresponding values(3.525425e+4) anyway. And the last thing i've found: there is a node, which allows to use several CPU cores during the simulation :D
Hello Philip!, the direction is set by three values for example [1,-1,0] in this case, the node does not perceive an angle of 45 degrees, it turns out that the problem is in the python source code, it turns out that it is not possible to set any angles
Yes agree, there must be something wrong in the code or not developed yet. But in any case thanks for your suggestions and taking the time testing for the community. That's the whole point of this channel that we all work together. It's highly appreciated. Cheers
Butterfly can only analyze one wind direction at a time... Although you could more than one inflow but that would not simulate a cyclone. Butterfly works like a physical wind tunnel. Maybe you need to modify your question. Instead of wanting to simulate a cyclone which is pretty unpredictable maybe think of what problem you want to solve first.
Thanks for great videos! For a more real example I’d imagine I would need a larger model with more context buildings around. Would you by your testing think that that will take a long time? Depend on computer and mesh accuracy of course, but my feeling is that it gets unrealistic to actually run this in a workflow, locking my cpu for hours. What’s your feeling?
Hi Alex, I saw this already being used in a larger models. Actually just recently by a student using my script or better to say her version of the script. I think it is important to work economically and with an appropriate massing model. Same for heatgain or outdoor comfort using a complex model with all details will definitely cause long computation. My experience though is that working with Rhino CFD, a commercial plugin, in my professional job it was not faster then butterfly and openfoam and I have to say the later feels actually more reliable and versatile and direct. But in general CFD computations can take very long, doesn't matter which software.
But aside, I'm glad you brought it up. Your question is actually a very good topic for one of the next videos: Testing butterfly and open foam on a real large-scale example. Let's try it and we will see how good it copes. Thanks for the question and watching my videos.
HI, I really appreciate all these wind series. I am pursuing an optimization process of form, facade design, location, and so on. I recently got interested in wind optimization of defining a form of roof. Do you think it works with Galapagos for optimization? I think there are some butterfly components are having actual values like wind speed.
Well it could work, I haven't tried butterfly with Galapagos. Although, the rewriting could be an issue since this never worked for me properly in butterfly. Also keep in mind that butterfly writes huge amount of data onto your hard disk. Also the re-meshing every time probably requires a very simple model probably. It would be very interesting though... Keep me updated.
That would not do much. The residual values only show how accurate the simulation is. Every iteration it becomes a more accurate model. I would first define the goal you want to achieve. What do you want to optimize? The roof shape, or its orientation? What are the criteria for the optimisation. Less turbulence? Slower wind speeds or faster? Do you want to create downdraft etc...
@@PhilippGalvanDesign I think it`s best solution better than Eddy3d , i tried a lot Eddy3d for wind but it`s only work in it`s building geometry and not accept in calculation any another geometry or Brep , Ploy surface or solid , did u try Eddy wind annual calculation in any another geometry
Hey, can you share the script and which ladybug u have used ? I am using latest version of ladybug, some components are not working. Can you help me with this ?
Hey, I use ladybug tools 1.1.0 (occasionally 1.2.0) and the legacy tools if I need to. Which components don't work on your end? Normally that means you installation is corrupt. This script uses 'butterfly' which is a separate plugin and installation.
@@PhilippGalvanDesign hey, I have installed butterfly seperately and all the butterfly components are working fine. But, the ladybug components are not reading. majorly says " no module named ladybug.analysisperiod, no module named ladybug.epw....and so on
@@PhilippGalvanDesign rest of the components I found in ladybug and redid the script, but some components like LB UTCI comfort, heat map, etc cannot be found.
You can find a more advanced version of script here: 1drv.ms/u/s!AroqbdH7xD3LjfM5knnmyLcEo5zmsQ?e=B208vL Regarding you ladybug installation : which ladybug version have you installed? did you follow the installation guide? github.com/ladybug-tools/lbt-grasshopper/wiki like for example did you ran the bat file in your user installation folder? let me know
3D probe and vector is a nice idea,thanks for sharing
Thanks, Yes I think it becomes a really good script now. It just needs a bit more refinement still.
Exactly I was surprised that we could do it in this way.
Hey Philipp, is there a reason for which you picked populate3d to create the 3d points than to make it a consistent grid? Not sure if that makes any change because now I'm struggling with my code being stuck into a massive setvars batch loop
There is no particular reason other than showing that probe points can be arranged not just as a grid but in any configuration in 3D space. The grid definitely has its advantages in most cases.
It was important to me to show this because it was not clear for all. The arrangement as 3d point cloud or 3d grid has some advantages.
Phillip, can you please show supersonic flow through a tube/pipe? Like a shock tube?
Not sure how... Also be aware that butterfly only shows a moment in time not a sequence. Simulations which show cause and effect might not as straight forward.
Hi @PhilippGalvanDesign, nice to see you are still active even on your older videos. I am currently trying to design a kind of "oven" that has to be put on top of an open fire. For this, I want to research if certain holes (size and locations) and shapes inside would improve the airflow. The goal would be to regulate where the hot air circulates to regulate the temperature in the oven. Do you have any tips? For now, I just keep watching your videos and try to apply them to my topic!
Btw I would really like to thank you for uploading these! Without your videos, I would have never considered researching this using butterfly.
@@thibotytgat3980 hey glad to see you using my videos. I think you might want to look at one of my later videos about indoor airflow and heat transfer. Not sure if you can simulate and oven (maybe) but for sure it can give some insights. I would probably suggest to also use other testing methods, including logic and common sense to make sure it really works as calculated. If you work on a real prototype you probably even want to create mockups. Good luck and let us know where you got to.
hello Mr phillip, thanks for the tutorial, the solution to remove the very high values is by using mesh inclusion and cull pattern to select only the points outside the geo, thanks again
Yes, Thanks! i realized that i have the option within the standard Grasshopper tools. Thanks for reminding.
I started with Wind rose as well, but then I listed the vector in the Wind Boundary Profile. I think that would be a good way to get the direction based on the location. Also for the velocity, the wind profile has calculated it base on different height, although it's better to take the velocity at reference height 10 m.
Good suggestion to check the wind profile!
Hey Philipp, first of all, thanks for sharing your knowledge. I have a question for you. I want to calculate the pressure on trees located in roof garden, so that I can use the suitable anchors. which software is the best?
To be honest, if this is a real project I would hire a company who can do a real wind tunnel test with a scaled or real size model.
Other than that I would say Butterfly I's a good option as long the parameters are set properly. You can locally measure the pressure. Just a note of caution: if you are thinking of adding an actual model of a tree you need to be careful. It will be very hard and time consuming to calculate a snappy hexmesh for that... Plants are tricky in CFD analysis. They are better calculated as porou massing but for that I haven't got my head around in Butterfly CFD. But Eddy 3d has this function. See here:
ua-cam.com/video/vPuB74am6cM/v-deo.html
Hey Philipp,
I am unable to hide the mesh and the snappy mesh to just view the vectors and the buildings, how to do that?
Thank you for the video btw :D
Every component in GH can be shown or hidden. Right click onto the component and turn off preview. Also the mesh component can be turned off completely if you don't need it. Any geometry inside rhino itself need to hidden there.
I suggest to watch a GH intro video to understand Preview etc...
Thanks for watching.
5:32
I've resolved this by silly rotation of the whole model xD But it seems like this issue has appeared not so long ago and there are no solutions for it yet, except updating the Rhino :(
Yeah, I agree i think it's a bug in the code.
It's a bit of a shame that such a complex system fails on a relatively simple issue.
But, let's see it as a chance to work together as a community and to solve it. That's the whole point of this channel. :) Thanks for taking the time to test the issue.
@@PhilippGalvanDesign right! By the way, since i am here, i have noticed a very crucial thing: i have tried to calculate several simulations in a single night using a datatree with different _cases in each and didn't manage with it, because it turned out, that the Solution node is constantly refreshes itself, so it never reaches the second branch and infinitly computes the first one. So, in order to avoid this, it is required to set "_interval" to "-1". The another important thing i've discovered -- there is a built-in butterfly node with skipped probes (the points inside the building, which were not calculated properly), but this doesn't resolve the problem until the end: you would have to detect and weed out the corresponding values(3.525425e+4) anyway. And the last thing i've found: there is a node, which allows to use several CPU cores during the simulation :D
Awesome stuff! I guess these are the topics to explore next.
Hello Philip!, the direction is set by three values for example [1,-1,0] in this case, the node does not perceive an angle of 45 degrees, it turns out that the problem is in the python source code, it turns out that it is not possible to set any angles
Yes agree, there must be something wrong in the code or not developed yet. But in any case thanks for your suggestions and taking the time testing for the community. That's the whole point of this channel that we all work together. It's highly appreciated.
Cheers
@@PhilippGalvanDesign Thank you! I have long wanted to try experimenting with the wind)
Is there any cyclone simulation
Not that I know at least not in Butterfly.
What you like to analyze?
@@PhilippGalvanDesignwind flow
@@PhilippGalvanDesign yes
Butterfly can only analyze one wind direction at a time... Although you could more than one inflow but that would not simulate a cyclone.
Butterfly works like a physical wind tunnel.
Maybe you need to modify your question. Instead of wanting to simulate a cyclone which is pretty unpredictable maybe think of what problem you want to solve first.
Thanks for great videos! For a more real example I’d imagine I would need a larger model with more context buildings around. Would you by your testing think that that will take a long time? Depend on computer and mesh accuracy of course, but my feeling is that it gets unrealistic to actually run this in a workflow, locking my cpu for hours. What’s your feeling?
Hi Alex, I saw this already being used in a larger models. Actually just recently by a student using my script or better to say her version of the script. I think it is important to work economically and with an appropriate massing model. Same for heatgain or outdoor comfort using a complex model with all details will definitely cause long computation. My experience though is that working with Rhino CFD, a commercial plugin, in my professional job it was not faster then butterfly and openfoam and I have to say the later feels actually more reliable and versatile and direct. But in general CFD computations can take very long, doesn't matter which software.
@@PhilippGalvanDesign okay, thanks for the answer! Good to know
But aside, I'm glad you brought it up. Your question is actually a very good topic for one of the next videos: Testing butterfly and open foam on a real large-scale example.
Let's try it and we will see how good it copes. Thanks for the question and watching my videos.
@@PhilippGalvanDesign Hi Philipp. That next video sounds really great! P.d. Thanks for great videos!
Hey Philipp,
Hope you're doing well. I am unable to export the vectors and buildings viewport into jpeg, Is there any specific way to do that?
i'm using the command '_viewCaptureToFile'
to export whatever is on the screen as image
HI, I really appreciate all these wind series. I am pursuing an optimization process of form, facade design, location, and so on. I recently got interested in wind optimization of defining a form of roof. Do you think it works with Galapagos for optimization? I think there are some butterfly components are having actual values like wind speed.
Well it could work, I haven't tried butterfly with Galapagos. Although, the rewriting could be an issue since this never worked for me properly in butterfly. Also keep in mind that butterfly writes huge amount of data onto your hard disk. Also the re-meshing every time probably requires a very simple model probably. It would be very interesting though... Keep me updated.
Thank you for your reply. Do you think I should connect Galapagos with residual_values of Solution?
That would not do much. The residual values only show how accurate the simulation is.
Every iteration it becomes a more accurate model. I would first define the goal you want to achieve. What do you want to optimize?
The roof shape, or its orientation? What are the criteria for the optimisation. Less turbulence? Slower wind speeds or faster?
Do you want to create downdraft etc...
@@PhilippGalvanDesign I am trying to do it for roof shape and make it better flow so it would be faster wind.
Great Work
Thank you so much 😀
@@PhilippGalvanDesign I think it`s best solution better than Eddy3d , i tried a lot Eddy3d for wind but it`s only work in it`s building geometry and not accept in calculation any another geometry or Brep , Ploy surface or solid , did u try Eddy wind annual calculation in any another geometry
Hey, can you share the script and which ladybug u have used ? I am using latest version of ladybug, some components are not working. Can you help me with this ?
Hey, I use ladybug tools 1.1.0 (occasionally 1.2.0) and the legacy tools if I need to.
Which components don't work on your end?
Normally that means you installation is corrupt. This script uses 'butterfly' which is a separate plugin and installation.
@@PhilippGalvanDesign hey, I have installed butterfly seperately and all the butterfly components are working fine. But, the ladybug components are not reading. majorly says " no module named ladybug.analysisperiod, no module named ladybug.epw....and so on
@@PhilippGalvanDesign rest of the components I found in ladybug and redid the script, but some components like LB UTCI comfort, heat map, etc cannot be found.
You can find a more advanced version of script here: 1drv.ms/u/s!AroqbdH7xD3LjfM5knnmyLcEo5zmsQ?e=B208vL
Regarding you ladybug installation : which ladybug version have you installed? did you follow the installation guide?
github.com/ladybug-tools/lbt-grasshopper/wiki
like for example did you ran the bat file in your user installation folder?
let me know
very good man
Thanks!
can i have the script file? thanks
Please check this follow up video and shownotes, i posted a link for the script.
ua-cam.com/video/c9Xaspg5Dqw/v-deo.html
All my vector go through the geometry :c
You might want check out some of my newer videos where I explain how to get rid of the vectors inside geometry.
@@PhilippGalvanDesign oh, I'll do it, thanks a lot!