at the same time just seems logical to me that a force would always seek the path of least resistance, so rather than making a 90 degree turn follow a more similar direction. The thought of separating stresslines into groups of tension and compression ist interesting from a design perspective. I wondered how tension and compression forces relate to the S1 and S2 lines, so what I did is pluging the outputs of P1 and P2 into the respective vector display for S1 and S2 and coloring the vectors blue for compression (negative values) and red for tension (positive values). So when you look at the upper side of the surface S1 (SC_02), Tensors along S1 show compression towards the middle and towards the supports Tension. However the Principal Stress Display of the Mesh Visualisation Component for the upper side shows it the other way round, again Red/ Tension and Cyan/ Compression as it says in your manual. Did I miss out on something ? When I look at the lower side (SC_03) I find it more or less matching up (I am just decerning between negative and positive values) so that might make the difference in the size of the compression area. So, does this mean that the S1 and S2 lines are related to the upper and lower side of the surface ? One beeing predominantely in compression(upper side) and one being stressed(lower side) ? That would also explain to me why S1 and S2 swap when you change the side of the surface. I am sorry, many questions... If you have time to explain, would be great. Also, maybe you have a book or article in mind which would explain those things more in depth....
Many thanks,
Philipp…
returned to GSA, it is solving. You might have to reset result scales using the GSA button.
Cheers,
Jon
Checking Input Data - this may take some time.
________________________________________
Data checks commenced at 23/08/2017 4:59:18 PM.
Checking input data for static analysis by GSS.
Shortest element (element 9) is 5 m long.Longest element (element 1) is 6 m long.
Data checks completed at 23/08/2017 4:59:18 PM. No errors or warnings found.
Analysis commenced at 23/08/2017 4:59:18 PM.
Analysis by Gss Static analysis
Initialising results modulesSolving for displacementsSolution statisticsSparse Parallel Direct 12 active nodes 14 active elements 2 analysis cases 24 degrees of freedom Minimum degree ordering 90 terms in stiffness matrix Maximum stiffness is 4.804e+009 at node 4 in direction z Minimum stiffness is 3.132e+008 at node 2 in direction yy Condition number of the stiffness matrix is ~ 5328. Maximum relative error in displacements will be 2.4e-10 percent. Factorization in 109 msSolving for element forces and reactionsCalculation completeAnalysis completed with no errors
Analysis completed at 23/08/2017 4:59:18 PM.Analysis time: 0.172 seconds
…
ted a picture of in your post. The reason is that sound has larger wavelengths than light.
With a light rendering model, energy can be said to reflect specularly, relative to their geometry, because the wavelength of light is inifinitesimally small relative to any object you might have modelled. With sound, energy may travel and reflect diffusely, or move around objects, depending on the scale of those objects. Think of the fundamental equation of frequency to wavelength - speed of sound = frequency X wavelength. Using that, you can see that a wave in the 125 hz octave is about as tall as a human being (or maybe a little taller) and would easily move around your body, not being reflected at all. A wave in the 1000 Hz. octave band is as big as your forearm, and might reflect specularly from your torso. A wave in the 4000 hz. octave band is about as long as your index finger, and might reflect off of your torso, or even your head.
Similarly, if you were to model the seats explicitly, it might be relatively accurate at very high frequencies (say 4000 hz. and above) but that is a very small part of the answer. Consensus in the field is that the most accurate way to model the seats is with a flat plane, raised to about shoulder height, and then with scattering coefficients applied to represent the varying effects of geometry on sound. I tend to use low coefficents below 250 hz. (say around 30%) and high coefficents above 250 Hz.(90%).
Absorption depends on the seat which was chosen. This is often a good area to use for a model calibration based on measured reverberation time.
Arthur…
ces are distorted (second). What is going on?
Surfaces in the second are a rhino cage edit of the surfaces in the first image. They were originally all closed polysrfs exploded just to input into grasshopper.
In the definition attached, each surface is compared to an original (its the small box in the far left of the top image) The point there is the ability to select for more than just the 6 faces of a cube, but find the closest match to more complicated inputs. In the second image, distorted surfaces are being compared to a distorted original.
If I have my math right, two parallel unitized normal vectors should have a dot product of 1, and the further away from 1 their dot product the further away from parallel, no? Why does it fail when I leave the comfy land of 90 degrees?…
Added by Peter Stone at 2:39pm on January 28, 2015
e to constrains, I HAVE to do it like this (I can't 3D print everything or opposite).
First
I have no idea how to make the panels, without so many duplicate Edges, Faces etc.
Second
I can't figure out how to align the triangle panels to fit in the construction, so it can be assembled ideally without glue. This problem is both conceptual (I can't figure out how to do it fiscally) and grasshoper-wise - I don't know how to organize data list and produce a global movement, so that the triangle parts do not intersect with each other, BUT intersect the 3d printed construction part (where they fit fix in or just fit and can be glued).
Triangles will be milled out from 3mm Plexiglas, BUT I will not have an option to mill at an angle, so only 90° edges.
3D printed parts will be executed by a high level production powder printer, so it should hold good.
Any ideas?
best,
cuki
…
File) 2. I have designed a curved Trichordal-Truss from one curve in Rhino.
The Truss is lying in the XY direction and the footer is placed on the zero point.
3. And now my problem:
I want to put the Truss-object on the feet, move 90°
(from the XY axis to Z axis, see sketch 1).
4. Then copy / move the truss to all 36-points of ellipse (see sketch 1).
5. Align the 36 trusses with the center of the triangle .
pointing at the center of the ellipse (see sketch 2).
6. Using a slider to change the position of the 36-Trusses at der ellipse.
Variable distance between Truss and ellipse (see sketch 3).
Thanks for you Answer.
Best regards
Noureddine…
multiply of variants from Galapagos, to have a chance for better analysis and comparability after. I also would like to use more then one solution in my design after.
In old topics i found kind of 3 solutions.
1.Change Galapagos to octopus ( what don t really want to do, i am kind of happy with Galapagos)
2. Use Slingshot! and MySQL Database ( it s look a little bit too complicated from the first view)
3. Use Colibri and Design Explorer Platform (looks kind of pretty way to solve my problem)
So i tried to add Colibri components to my definition , but have some mistake in the Colibri Aggregator after adding the Genome "An item with the same key already been added". I think it comes because for some steps i am using the "Gen Pool" and not a normal slider. Is it a way to connect Gen Pool and Colibri (i really prefer to have it, then a lot of sliders in some cases)?
And the second question (if i will get it solved with gen pool), could i somehow controll the recording process? For example i would likte to record only variants wit fitness over 90% or start recording just after 20. generation and record till the end?
I also opend for all other possibilities to reach the same goal (record/save/bake multiply variants from galapagos)
…
starting as soon as possible.
We're offering challenging projects, insights and contact to leading industry companies, project responsibilities according to abilities and initiative, great work environment and laid-back atmosphere, room to play and evolve,...
Our ideal candidate:
- is passionate about construction, engineering and (computational) design
- is proficient in Rhino / Grasshopper / (GH-)Python
- knows his ways around the Adobe Suite and MS Office
- has a current work permit for Germany
- is a German speaker (other native speakers also welcome, with excellent English skills)
- has an architectural background (Student / BA / MA /...), ideally with work experience
- is interested / has experience in digital manufacturing and prototyping
- will be able to join us shortly
We're looking forward to your applications / inquiries / CVs to: mpelzer@fat-lab.de
View our past projects here: www.fat-lab.com
(Current projects, unfortunately, are non-disclosed)
…
hen you determine their position and number by the "precision_" input. Their number is equal to "precision_ - 1".They are essentially mesh edges, not curves. To hide them, in Rhino application menu choose: Tools -> Options -> View -> Display Modes -> choose your current display mode, and uncheck the "Show mesh wires":
b) How do you change the x- and y-scale of the chart? For instance displaying every 10 degrees of azimuth?
Can you be a bit more precise?You would like to change the labeling of the azimuth directions from 30 degrees step to 10 degrees step? If this is so, you can not do that.
c) My input surface is a hyperbolic paraboloid, facing south symmetrically. How does the component calculate its tilt and azimuth?
All Ladybug Photovoltaics components calculate the amount of AC energy generated by a planar (flat) surface. Tilt and azimuth angles are calculated based on surface normal at 0.5, 0.5 surface parameters. So you can not use the hyperbolic paraboloid as the _PVsurface (or _PV_SWHsurface) input, as it will yield incorrect results. You need to planarize that hyperbolic paraboloid surface first.I attached below an example with default grasshopper components, but the size of the panels is not equal. If you want them to be equal use some paneling Grasshopper plugin like Lunchbox or Paneling tools.…
50 and reduced the 'cell size' slider to 0.5. When the 'Azimuth' angle is changed to 180 +- 90 (dawn or dusk), the points are widely dispersed, reducing the density and increasing the number of cells in the "sparse grid". Under these conditions, the number of cells was ~2000 and the Profiler time for 'Boundary' went up to a full minute or more each time 'Altitude' or 'Azimuth' was changed.
So I created this code to benchmark some alternatives and found two interesting things:
'Boundary' surface performance (v.1) is not linear. As the number of surfaces goes from 1000 to 2000, the time per surface goes up dramatically.
I tried three alternatives for creating a rectangular surface at a given point that are all substantially faster: v.2, v.3 and v.4. For 2000 points, v.4 is 150 times faster than v.1 !!!
Performance of v.2, v.3 and v.4 are similar and all scale up very well. To benchmark beyond 2000 points, I recommend disabling the VERY SLOW v.1. At 5000 points the 'Pop2D' component takes ~11.3 seconds but v.3 and v.4 take less than one second to generate 5000 surfaces!
See boundary_2015Nov19a.gh attached.
So I replaced the 'Rectangle' and 'Boundary' components in my sun reflection model with v.4 in focus_2015Nov19b.gh (also attached) and the performance is amazing.
I'm sure someone has mentioned this performance issue with 'Boundary' on the forum before but as with many things, I didn't realize what a major obstacle it can be until I discovered this for myself.…
Added by Joseph Oster at 9:16pm on November 19, 2015