Pratt Computer Aided Construction Summer 2014

This discussion is for Grasshopper-related file sharing, questions, and general tomfoolery in CAC Su14.

Replies to This Discussion

Permalink Reply by Qin Shasha on July 11, 2014 at 8:16am

Units:

These Pictures are showing systems and units what I interested. I would like to try to combine a curve line system to a straight line system. But I am not sure whether they work or not that is why I selected many pictures even though I know some of them may be useless.

I sketched two pictures that how to develop unit. I hope get you some suggestions so that I could start to make the unit in grasshopper.

Permalink Reply by Brian Ringley on July 16, 2014 at 9:55am

Hi Shasha

For "Picture 2" applied to a cellular grid like those pictured above I would recommend lofting from rebuilt Voronoi cells to a series of circles. You'll see in the definition that the circle sizes are limited to those which can be inscribed within their respective Voronoi cells. This can be done by measuring from the centerpoint of each cell to all of the cell vertices and edge midpoints and then basing the circle radius off of the t value (curve parameter space) of the shortest curve in the list per cell.

In some instances manual surface building may be required, or you may need to adjust rebuilt Voronoi cell curve seams In Grasshopper to get them to align properly with the curve seams of the circles for a well-crafted loft.

In terms of the ribbing of the units I'm sure you could align those with the Voronoi cell vertices which would be very cool. Of course you would also want to create gradations of things like unit height and unit size relative to cell size.

See the image below and the attached Grasshopper definition for reference.

Attachments:

loftingFromVoronoi.gh, 14 KB

Permalink Reply by Qin Shasha on July 11, 2014 at 10:15pm

Hi Brain,

This is the model what I made.

1. What I beginned with the geometry which is too much work. Is that any simple way?

2. I could not figure out how to do the opposite seeting so that I just copy that and change the data.

3. The shape is not nice as what I drew before. Could you help me to make the curves much nicer and having more control for the start/end points?

Thank you

Shasha

Attachments:

Shasha_Unit_7_11.gh, 19 KB

Permalink Reply by Brian Ringley on July 12, 2014 at 4:12pm

Hi Shasha,

It looks like you're trying to understand the geometry of the classic Yin Yang form. As you can see on this site, the Yin Yang is formed by two circles within a larger circle. In the attached definition I use this to create a module that allows you to control the Yin size and the Yang will size correspondingly. You can also adjust the height of each using a patch component with a centroid Z translation, although there are other ways to do this depending on precisely what you're looking for. You may also want your base unit to perform in other ways, for example opening up in the middle in which case a gap would need to be modeled between the Yin and the Yang.

The important thing is for you to explain your concept further - for example, is this model to be your base unit? If so, what grid will you be applying it to? Do you want to use something like some of the Voronoi system images you posted above? Please think about these things as they will help you push your design further.

Attachments:

YinYangSample.gh, 17 KB

Permalink Reply by Peter Liu on July 12, 2014 at 11:35pm

Hi!

I'm interested in the unit and relationship of koru fern spiral and the hierarchy of feather of bird's wing.

What I am thinking to make is to apply the spirals onto a tree-branch system then apply it to a wing-shape surface. I'm not sure if it is a correct strategy. Currently I tried to make a 2D spiral surface (haven't made volume yet) and still looking for the way to make the tree-branch system. If it is possible, please give me some advice.

Thanks,

Peter

Permalink Reply by Brian Ringley on July 14, 2014 at 8:45am

Hi Peter,

A sketch of how you envision the spirals being applied to the wing-form could be helpful here. Are you thinking of using the same organizational logic (spiraling feathers - in which case overlap becomes important)? Additionally, how will the spirals gradate across the surface? And how will the combination of the spiral units and the wing form lend itself to flexibility in the fabricated piece?

I'm having trouble envisioning the combination of branching and the wing form. If you're interested in pursuing L-systems I would recommend a Grasshopper add-in called Rabbit.

Permalink Reply by Peter Liu on July 17, 2014 at 12:29pm

Here is the problem that I encountered at this moment. I've figured out the variation of curve and try to loft them. I linked both of the curve list to merge then do the loft as usual, but it just doesn't give me the surface; also, there is no error bubble pop up.

attachment is the file I am working on, problematic section are in the red group.

thanks,

Peter

Attachments:

fibonacciSpiral 0717.gh, 47 KB

Permalink Reply by Brian Ringley on July 17, 2014 at 3:28pm

Hi Peter

The respective data structures of your inputs don't match. In terms of branching, you're fine because all of your inputs are consistent in the respect that there is one item (or "leaf") per branch (shown as "N=1"). However, your inputs' tree path integer counts vary. The D1 input of the "loft base curve" has 7 integers (expressed as {a,b,c,d,e,f,g}) while D2 has 8 (expressed as {a,b,c,d,e,f,g,h}). The good news is that if only one of the integers in a tree path varies (in your case {f} of both D1 and D2) then you can "simplify" the data with a simplify component or by selecting simplify within the right-click menu of the input to reduce both input integers to simply {a}. Please refer to the image below, and I highly recommend watching at least the first video in this data tree series by David Rutten.

Permalink Reply by Dane Borda on July 14, 2014 at 11:18am

Permalink Reply by Dane Borda on July 14, 2014 at 11:20am

Permalink Reply by Jeremy Hill on July 16, 2014 at 1:35pm

HI , I have two Brep's, A array with a hole, and Half sphere's made by using the hole's edge and center points, splitting the sphere in half and re-scaling in the Z. the Issue I'm having is that I can't join the two together right?