MAT 594X_Stackable and Nested Objects W/ Grasshopper

281

1

8

Introduction: MAT 594X_Stackable and Nested Objects W/ Grasshopper

Using affine transformations, the project assignment is to parametrically model stackable or nesting objects, which ultimately needs to be printed and tested.

I am working on a simple pendant lighting fixture design. It will have a pattern for the light to go through and create playful shades.

Inspired by Echinus Esculentus, AKA sea urchin.

Supplies

Rhino and Grasshopper

Rhinoscript

Step 1: Preliminary Result With Stacking

I might need to change the geometry of this lighting fixture so it can stack better. With that, I will also try nesting it.

This geometry itself is designed purely in grasshopper with no RS (rhinoscript). I created a series of polygons with the same number of segments and the same origin point on X & Y.

Z coordinate varies. Then, I lofted it, converted it into mesh. I triangulated this mesh and used a series of components from Weaverbird (wb) library: wbFrame, wbMeshThicken, and wbCatmullClarke subdivision smoothening.

There does not seem to be naked edges, but there are some problems with the geometry. See the detail picture on the upper edge. I first thought that this is due to the wbCatmullClarke smoothening algorithm, but I noticed that there is a subtle crease there before applying that algorithm. The crease is present as soon as I loft the polygons.

Sliced a mini and patern-free version (7cm in diameter) of this. What is interesting is that, if I place this in the same orientation as the rendering, the print time is 17 hours, however if rotate it one pi, the print time is around 8.5 hours. I am guessing that the discrepancy is due to the differences in support structure that needs to be built.

Question:

Doe that mean lofting polygons from parallel planes is not a reliable way to generate geometry for print? What are the reliable ways? Are there certain surface, solid, or brep creation algorithms that land themselves better for printing purposes?

Update:

My new theory about the crease I mentioned earlier is that if I increase the segment number of the polygons that make up the geometry, it could help smooth out that crease. Will try that later today.

Step 2: Nesting and Embracing the Segmented Polygon

Ok, in addressing the crease problem that I had mention before, my theory was to increase the segment parameter of the polygon component. After all, the smoothest shape of all -a circle- is about having infinite segments. The problem was, I could not have infinite segments and I could not use circle, instead of a polygon.(Somehow the loft operation does not work the same)

So, I increased the segment up to 100. The crease was still there. (see the red geometry on the screenshot take from Rhino).

Then, I decided to embrace the "polygon"'s true nature and I decreased the segments for a deliberate polygonal look. In this version the polygons are octagons. It worked quite nicely, I think, but, it will still take forever to print.

Sliced a mini version (7cm in diameter) of this (for LED maybe?). The print time is around 7 hours with a support structure.

Oops, I realized that the shape has deformed with the scaled iterations. My guess is that the reference point for the scale transformation does not match the center point of the geometry. I'll fix this in the next step!

Step 3: Why! Circle Indeed!!!

I a not sure why I assumed that the circles could not be lofted into a similar geometry. Just replaced polygons with circles and lofted, et voila! Completely smooth.

Sliced a mini version (7cm in diameter) of this (for LED maybe?). The print time is around 7 hours.

Step 4: 1st Attempt at Printing W/ Creality Ender 3

First trial: not quite success. Several problems

1. The support structure is so dense that it is impossible to peel away. This is a matter of slicer settings. I need to find the maximum stability with minimum material in support structure.

2. The bigger problem is that it got stuck on the bed. I used brim and the structure was extremely dense. There was no way it was coming off, so I heated the bed, 50, 60, 70, and finally at 80 degrees Celsius, it started loosening (but also deforming). Even in this temperature I had to use the spatula really hard for it to come off. As a result, I dented the bed. (See the photo). This may effect my future prints, so I will look into ordering this part as an extra.

Be the First to Share

    Recommendations

    • Pets Challenge

      Pets Challenge
    • Stone, Concrete, Cement Challenge

      Stone, Concrete, Cement Challenge
    • Backyard Contest

      Backyard Contest

    8 Comments

    0
    mengjiazhu
    mengjiazhu

    1 year ago

    Nice pattern! I had similar problem when doing surface morph. My shape is irregular and closed surface, so the morphsurface won't wrap. Only when I select an open surface it will wrap. Wondering if you have solved your morphsurface problem?

    0
    solenk
    solenk

    Reply 1 year ago

    Before applying the surface morph component, you need to right click on the base surface and pick "reparameterize". Try this to see it will work.

    0
    soniahashim
    soniahashim

    1 year ago on Step 4

    This work is so beautiful! I appreciated how clearly you documented your process of laying out the troubleshooting with the creases and the urchin inspired design. I'd be curious to learn more about your design process and understand better about how you decided to loft geometry and came up with the original polygonal shapes for the cutouts (so I can try it too). Thanks for sharing!

    0
    solenk
    solenk

    Reply 1 year ago

    Thanks, Sonia. It was not like I had a clear picture of what I wanted to do from the get go. I would be lying if I said I started with the urchin idea. What I did know was that I wanted do some sort of lamp shade with some patterned perforations on it. With that, I started playing around in GH. I knew that loft was a very common way to create surfaces from curves, so that is what I tried. While I was tweaking the parameters and getting different shapes, I got the shape (that I eventually used), which made me think of sea urchins. The pattern design relies heavily on the weaverbird library components, so it's a bit of a shortcut. I did not sit down and 'design' it. I included a snapshot of the GH patch in step1.
    So, the process is: come up with a general loose idea of what I want to do. Start playing with code. The results inspire new ideas. With those, play with code again. So, the whole thing is shaped as a non-linear combination of tweaking code and getting inspired by results and giving it some direction. A lot of my decisions are determined by the capabilities of the components and some of things I already knew due to having previously played with GH and several libraries.
    Sorry for this really long reply. I did not intend this to be this long when I started typing it :)

    0
    jas0126
    jas0126

    1 year ago

    I really like the patterns you created. They make the objects graceful! Looking forward to see these objects printed.

    0
    solenk
    solenk

    Reply 1 year ago

    Thanks! I added the 1st attempt at printing as step 4. It was not quite a success story.
    :-(

    0
    merttoka
    merttoka

    1 year ago

    Solen these look awesome! I especially loved the version at step 3. Did you render the models in Rhino or in some other program?

    0
    solenk
    solenk

    Reply 1 year ago

    I rendered in Keyshot. The educational version costs $95, but I am currently using it in free trial mode (15 days, I think). It is a very good software for quickly get decent results.