Geometer is a new design tool and form language mediating between the physical and digital world.
It is made to be understandable, communicable and dynamic.
-No black box, encouraging active participation and goal oriented design.
-Components open a channel for dialogue around relationships and patterns.
-Built-in system that encourages strength and freedom of form without compromise.
There are three components to this project:
(1) A kit of robust physical and digital components with unique names and specific geometry which can be combined in a modular fashion.
(2) A system of communicating the interactions of these components.
(3) A way to easily physically manufacture complex assemblies made up of modules, in a variety of materials reliably of considerable size.
We hope that this will be an efficient and approachable design to build process for all.
Snap together versions of set 6 suitable for printing on a Form 2 printer, best printed at half scale. Modules can be made form a variety of materials (we used grey pro) and the snap together part is made from the Flex resin.
Mates definitely help speed up the digital form finding process.
An example of a wall light and then its transformation into a desktop model.
And then using the same structure to inspire a table, initially too small but slowly growing to have correct proportions.
There are overlaps and several gaps which presents us with a unique opportunity. We can make a small insert module to take up these 1/4-.5" gaps and such or we can follow along with the natural troubleshooting method of our modules by inserting more modules and expanding dimensions until things line up correctly. Fortunately we can more or less cut and past and reorganize until things work but its definitely an interesting design process. You have an intention and find a module which will approximate it but due to its constraints it naturally guides the form into a specific form factor or increment thereof.
Using the mate tool in Solidworks assembly is very fast for form finding, although it is a little annoying to undo a mate if you change your mind often. A quick start file will be created for those with Solidworks 2017 or higher once module files are finalized.
We still haven't gotten to try out Mindesk for Solidworks since we do not have Solidworks 2019.
The above assembly took less than five minutes to put together.
We have been printing more physical half scale modules that snap together to ease the design process.
We are starting to notice an error in alignment, when trying to close angled sections it seems that the modules are too long, but to create parallel structures we need to standard length. Playing around with it some more we decided to add a shortened piece as a bridge for these angle errors.
Some other modules were developed to allow for easier execution of symmetrical structures and for making leg terminators. These are modules L-Q being printed today to add. There is also a short terminator for single leg ending. as needed in the below example.
We are debating adding graphics to communicate the module name and leg numbers as well as a sort of indexing system to determine angle between modules. Either this could be a saw tooth mate at fifteen degree increments or it could just be two graphical indicators.
With this modification you would be able to express ex.: A2-15-C3 when module A, arm 2 connects to module C arm 3 at 15 degrees clockwise from aligning marks. The argument becomes whether hydro-graphics or pad printing would make sense, embossing.... etc. How to print on such a structure? How much information and precision is needed?
To add to the complexity, these are half scale modules so the above "Desk" would be very small in fact; how do we develop a sense of scale? A half scale person model is out of the question, far too big. Yet the dilemma is that we also can't make the snap together modules 1:1 scale due to costs associated...
Definitely a challenge. This is where VR really helps out, its unfortunate its such a fringe practice.
Nearly all mold designs are done now so we are getting ready to send them out for a quote for production elsewhere due to time constraints. At this point we need to just start designing things and then move to production in time for our final presentations.
It's very inefficient to 3D print every joint connector for the snap together sets.
Since this type of medium seems like it will be so popular, I made a mold.
With this you can connect twelve modules together with a single casting. I think that expanding on this and making a tree which has enough connectors for an entire "set" will make sense. The end user can cut them apart to save a little labor and organizational nightmare.
If we decide to really scale up production an alternative design will probably make sense. This mold only looks complex to save some some material on the printer.
Once a casting is made the master can be used to make more molds exponentially.
Making molds for the actual snap together modules will be a little more annoying.
Right now I am thinking a block mold with a five piece insert assembly for each end point... Definitely not the simplest of setups. It would likely require the precision machining of small metal components. At that point we'll be so close to automation that I might as well go the extra step and make a machine with jigs and everything.
Jason is sick and can't bring materials to school, so while additional hydrocal is on the way we decided to try and cast some modules out of concrete.
Definitely not the cleanest method haha.
The mix was a little too watery I think, but I left them for only a day and then de-molded them.
Ouch! Where I used pipe clamps to hold the modules together the molds cracked.
However, as I took the molds apart it didn't look too bad...
There were plenty of air bubbles but all modules were actually pretty stable!
The 3D print adds a really cool texture, almost like a fingerprint.
All in all, I would call this progress. We definitely need to maybe use a more flexible mold material in combination with tape/flexible straps for actual concrete. Otherwise we need to find a material which changes less dimensionally such as hydrocal. A vibrating jig or air evacuation system might also help.
After some scaling issues were corrected we were able to make some snap together modules; an entire module set is printing currently. On the Form 2, at half scale all modules take around ~15 hours to print and the connectors takes about another four.