Geometer

So Unit set 2 has some issues; this is why we iterate!

Due to scale, tolerances or some other issue units aren't matching up in our intended way. This is okay since we now have a benchmark to play off of but officially I'll declare Unit set 2 a failure.

This may be due to several specific decisions made transitioning from theory to applied theory, namely the abandonment of the bounding box operation which would result in slanted and non uniform termination points. Unit set 2 was instead constructed by placing a point at the origin, and with a set length line placed at various 15 degree increments in space and then extruding straight along the line the arm.

This approach seemed intuitively more dynamic since it would allow for rotation and surface continuity between units. This theory seems incorrect, although maybe not in its entirety. We may simply have printed too few "useful" components, namely those with self intersecting geometry.

Anyways, this is a setback and learning opportunity. Files will be compressed and re-uploaded in whole but should be used primarily in digital format for exploration and testing, not with a material investment. Its somewhat misleading when looking at the models digitally, they seem to work but will be misaligned by something like >5 degrees.

Batch one Units were generated in Solidworks but proved difficult to describe explicit dimensions for the number of desired Unit variations. A new methodology is being employed to generate the base forms in Rhino, smoothed and made functional in Solidworks with integrated labels denoting the Unit name (ex. "A" "B" etc.) and arm designation ("1","2" and "3".)

This will result in a spreadsheet with arm termination coordinates in relation to the origin (0,0,0) to allow for standardization but opportunities to scale, modify arm thickness and more down the line while preserving basic reference data.

Current specifications will be for arms to both terminate in female ports with an additional male union component. In theory this should work since all pieces will wedge against each other through triangulation. This is not our ultimate goal and serves only to expedite the physical testing process. We are still exploring joint possibilities which will allow for rotation, temporary or permanent adhesion, re-positioning etc.

Components with designations A-Z will be completed with the expectation that in practice certain Units will commonly not be used, this will be documented and those units will be culled from the directory and another will take its place. In this way we can work towards having a limited catalog of units with optimal possibilities for usage.

Our current preliminary estimate on this development cycle is as follows:

Base modules and spreadsheet completed 10/13/18

Fillets, imprinted text and STL exports by 10/15

First test prints 10/15-16 at reduced scale, initial test and then release.

Upon release we will need feedback and additional testing. Components will be released individually and as groups for ease of printing. At this point we are not worrying about structural integrity or optimization.

Rules should start basic and get more complex.

Simple:

Three Units creates a plane.

Four units create a volume.

Complex:

An example rule would be that a ring of A components with other components between that then ends with another chain of A components will result in parallel surfaces.

Since each Unit is named and leg assigned a number (Likely to be inscribed or printed) we can fully describe a system to communicate a particular collection of Units as an artifact.

An example would be a spreadsheet where A2 goes to C3, C2 to F1, C1 to D1, and so on to create functional closed volumes.

A Unit follows certain rules:

1) There is a central point.

2) From the central point three legs extend through space.

3) Legs follow a logical placement from one another (at 15 degree increments for example tbc.)

4) All legs terminate in a (multi-positional) connector which connects to other legs (This connector is of crucial importance.)

5) The connectors must be of a uniform diameter and surface continuity.

6) A bounding box contains all elements of each Unit and is of a set size, this determines scale.

7) The central point can be anywhere in the bounding box but legs and connectors must always touch the envelope of the bounding box. (This allows for tessellation at a higher order.)

8) A Unit has filleted or smoothed progression from the central point to the legs for aesthetic purposes and to add strength.

9) A unit must never exceed a certain material volume.

10) All above conditions describe one Unit which is then named and may not be geometrically identical to any other Unit.

Keeping in mind that Unit size determines resolution and possible complexity of designs we can assume we will need at least 34 unit types (arbitrary.)

Using the experiences of Precious Plastic each module should use a maximum of 180cm^3 or a solid 6x6x5cm cube. These values are not definite but help guide our design.

For now we are assuming that we will be using plastic and that it will either be injection molded on a small scale and molds exchanged between machines in a geographic area or that the units will be cast.

We can assume that the Units will be solid so some rough calculations lead us to:

Thirty-four units, three of each at a maximum of 180cm^3 = 18.36 Liters of material. (5 gallons)

If made via casting Smooth-On anticipated cost for Unit materials would be $500-620 for a full set for a maker space or design studio without bulk discount.

The above example highlights the importance of limiting the number of kit components variations! If we had 16 Unit types we would be potentially more restrained in our design possibilities but it would be cheaper to initially perform form finding operations.

Using the above example understanding the potential build volume should be calculated (tbc.)