Week (2)

Overview

I’m thinking of my final project as an assemblage of different components that is meant to exhibit an integrated and working system. At the moment my project is made up of three major components.

1. Mechanical system

2. Touch interface

   • Analog
   • Digital
   • A set of rotating and embedded geometries

Design Challenge

How can we rotate two embedded dodecahedrons who share in a common center along different rotational planes?

This seemingly simple proposition will be the animating parameter for the first phase of the project. As I prototype, understand limitations and overcome challenges, I will revisit and revise this proposition.

My design process will first start with a geometry study.

For ease I will start with a dodecahedron since it is a complex platonic solid and offers a lot of geometric possibilities and planes to work with.

Thinking about fractile geometry in three dimensional space. I was able to produce three embedded Dodecahedrons based on their dual geometry, the icosahedron.

My project will start with the first and the second of these geometries. Maintaining the dimensions and center point.

With this starting point I will ask how can I rotate both geometries on two different rotational planes?

Using grasshopper to animate this movement

I used grasshopper to animate this kinetic motion so I can observe the rotation in real time. How does it look? Is it dynamic? Does it look interesting? I animated the pure geometry within the digital space before starting to think about the physical limitation and actual mechanics of the design.

Grasshopper process

How to express both geometries grasshopper

Add two curves on the GH canvas and set to geometries within Rhino Space/ Set multiple curves Add a point and set to the center point in the Rhino model Add a Rotate 3D component for each geometry and add a vector for the axis input. In this case I just used the Unit Y and Unit x but in the more advanced model I created my own vector by 2 point vector component. Add two number slides with range 0 - 360 degrees Merge

This model is limited in that you can only rotate one model at a time. To see both geometries rotating in tandem and also to auto rotate I used the plug-in Timeline.

How to set up Timeline.

Download and run timeline from food4rhino Add two construct Domains for each geometries Connect to Remap numbers and use the output R as the values for the angles in the Rotate 3D component. Slide the time line to see both geometries rotate together.

Notes on rotation

My immediate reaction was good. I like the visual effect of seeing these two geometries rotate along different planes. The movement feels stable and unstable at the same time which is a very interesting duality to work with. I also liked that I didn’t set any of the rotational planes at the typical X,Y, and Z.

Thinking of Mechanics: How to materialize this motion?

Realizing this embedded rotation is not easy mechanically since it has a lot of limitations.

If both geometries rotate with the same plane for rotation things would be much easier. Since both geometries would share the same axis of rotation. When the rotational plane differ, the geometries will eventually intersect and collide together or collide with the axis of rotation. See the following screen shot.

Imagine the out geometry rotating from the center point with a gear or bearing. The rotational plane will cut through the second geometry making this approach unfeasible.

This made me rethink the mechanics. The best solution I have so far is to have two separate drivers from each geometry that will connect to one of the 12 faces (call it the rotation face). Consider the following sketch. The diagonal axis for the outer geometry rotates through a mechanizing on the rotation face without the axis penetrating into the interior of the model.

The rotational face and plane directly opposite it in the outer geometry are the only openings a secondary mechanism can enter to move the second (smaller geometry)! If an axis enters from any other place it will intersect with the rotational sweep of the outer geometry**

Okay with this limitation we can start to think of the second rotation mechanism from either A. Rotational Face B. The face opposite to the Rotational Face

Staying with the rotational faces of the Dodecahedron I chose a diagonal axis. I’m imagining this can work with a bevel gear designed according to the geometry of the dodecahedron. Which will transfer the motion according to the needed angle!

Here is a photo from the internet of what I’m think about:

I did some more modeling to try to visualize the model and start to think of the structural challenges. I reoriented the geometries a few times. I liked the option where the second-smaller geometry is hung vertically along the Z-axis. Although I think diagonal axis are more interesting, I intuit that this will be more structurally sound.

The rotational plane for the outer shell remained diagonal which may offer structural challenges - but with the right martial (light weight) and reinforcing the rotational face things should be fine.

NEXT STEP Move from a sketch model to a working prototype Get into the details of the gear design/ research options Think of a fast and light weight structure to make - I’m thinking of soldering metal wire.

Start research on: User interface Analog option Digital option