> Turns out, a slit really isn’t the best shape for the hole because if you don’t have the right kind of rubber, it can rip too easily. A circle is better because there’s equal pressure around the hole and it’s less likely to rip. I wish I had considered this sooner, but that’s what it takes to learn and that’s why I’m sharing this here.
Given the level of detail in the article, I'm led to believe that the next moment hasn't happened yet..
What you really want is a slit with a pair of small circular holes, one on each end of the slit.
o--------------o
All the benefits of the slit without the tiny points where all of the stress is focused, which is what causes the ripping. Adjust the size of the holes as needed.
Common design trick for sheet materials, and should be straightforward to cut with the same machinery.
I'd take it a step further and and go with a slot. Closer to the structure of the 2 circles with the ease of a slit. I am surprised the operator/programmer of the machine didn't suggest something like your idea or mine, since it is common practice for them to make engineering designs more manufacture-able/structurally sound.
(Source: I used to program water jets and was constantly re-designing parts sent by engineers/designers.)
Also, I would definitely move towards die-cutting once he settled on a final design, but with his sense of tinkering, I doubt he'll ever have a 'final' design.
The same trick is employed when repairing windshields in cars. If there is a small cut you drill a round hole at both ends to make sure the cut doesn't tear further and then you fill the whole thing with epoxy.
Yup, in shoemaking this is the "first trick" when doing some designs with straight-cut lines. Holes at the end prevents ripping completely (unless the leather is bad quality) and you can go on with your desired design.
I came here to make a similar digression but clearly am not the post to :) interestingly though this is one of the reasons for serifs on fonts! In stone carved writing serifs help remove corner stress and cracking.
Buckminster Fuller's approach to domes was that they should be made from factory-made components made to tight tolerances, made from modern materials - Fiberglas panels, aluminum connectors, and neoprene gaskets. Such parts are cheap when made in bulk. Domes made that way hold up very well. Some radomes in the Arctic have been surviving winters for 50-60 years now.
The hippie dome movement, as led by Kahn, encouraged do-it-yourself dome making from "natural" or even found materials. This doesn't work as well. A dome has a huge number of edge joints, all of which are exposed to weather. Making them leakproof is hard. Making them leakproof without tight tolerances is very hard. The best solution the hippie movement ever achieved was to nail shingles over the whole dome. Windows presented even worse problems.
A more subtle problem was discovered later. Although the basic geometry of a dome is simple, what happens to a dome when sunlight heats one side but not the other isn't. Large internal stresses develop, and this cycling from relative expansion tends to cause leaks at joints.
The rubber connector is a nice part. It should grab tighter as the struts are bent from flat to the dome shape. Getting the last struts connected, though, might be a pain.
I've interpreted Kahn's main objections to be that the dome shape is difficult to efficiently partition for human living. Any conventional appliances placed against walls would leave large gaps. You could put some plumbing and electrical back there, but it would still be inefficient use of space.
That, and the fact that conventional house designs are effective with natural and found materials, and you have geodesics beat. That said, I'm constantly looking at this problem because I feel that there must be some important concepts that could be derived from geodesics without the dome aspect.
There’s a great 99% Invisible podcast episode called “Derelict Dome” that tells the story of a geodesic dome restaurant built in the 1950s.
“The glass windows heated the restaurant up like a greenhouse, so the owner installed fiberglass over most of the dome, blocking the ocean views. It leaked constantly, and was difficult to maintain.”
This is a really nice example of how a "simple" problem can take years and a great deal of tinkering to find a workable solution. I've been thinking about--and building--bookshelves for a lot of years, and only recently came up with a design that I really kinda like: http://www.tjradcliffe.com/?p=1499
The optimal balance between imagining and prototyping is a fascinating problem, and seems to depend sensitively on the problem domain.
Prototyping even simple components is relatively costly in time, materials and tooling, but gives you knowledge.
Imagination is free--it can often be done in time that would otherwise be wasted--but it cannot create knowledge, and almost everything we imagine is wrong.
Imagination can be fun. Prototyping is generally considered less so, although the Maker community is working to remove that unwarranted stigma. We never talk about "leaps of prototyping" the way romantics talk about "leaps of the imagination", but maybe we should. One good prototyping experience can take you a lot further toward knowledge than almost any amount of imagination.
what a delight to wake up and see this on the front page. thank you everyone for your kind comments and thoughtful suggestions. this project has been very close to my heart and i'm pleased to see this story become inspiring to others.
http://domekit.cc/kits has more information, pictures and links to buy one of our kits on Amazon.
I think, with the advent of 3d printers, cnc mills, etc, software people like myself are finally starting to grasp the complexity and difficulty inate in the production of physical objects.
This realization comes through the same long and expensive process of trial and error that you experienced, but often without a successful ending like yours, so this is great to see!
I've been going down a similar path, trying to create a smaller-scale connector for a regular dodecahedron. So I can fully appreciate how impossible these connectors are for CNC Milling/resin casting or injection molds.
I love the kits. I didn't see an FAQ anywhere- what's the minimum number of inexperienced people it would take to put together the small/big domes and how long does it take?
at a casual pace with the help of a 3-5 friends, it might be about 15-30 minutes to raise a small DOME KIT. a big DOME KIT might take a bit longer, say 45 minutes with 5-7 people. we've seen highly-coordinated groups complete the challenge in less than 10 minutes.
I love stories where persistence over time pays off. Thanks for posting this.
This story reminds me of Sugru's story. If you have not yet read "A partial visual history of sugru", it is well worth your time. Of course, this has been posted to HN in the past.
thanks, this is a great read. I got a multicolor pack of sugru as a gift 2 years ago and I never would have guessed how much of a labor of love it had been. And yeah, I thought it was pretty awesome right away.
I really enjoyed this writeup. The author's persistence and commitment to a problem he found interesting reminds me of the energy I had before I worked at a big software company.
I find it strange that you have designed your own dome CAD software, toyed with 3D printing and other techniques, yet your final product is a ready-made kit that only comes in two distinct sizes.
I thought you are aiming for a solution allowing everyone to build the dome they want. You could be supplying the connectors (vertices), and users would add the wood planks for edges (btw, would aluminum tubes be much heavier or costly?). Maybe you could also provide ready-made planks of varying sizes; that still would allow for many variants. Have you considered this?
I really enjoyed the article.
It reminds me the sort of iterative search for an elegant solution I often had to do during my design engineering MA.
I didn't appreciate "design thinking" at the time, it was too much manual work for me for something that seemingly resulted in the same output.
Now that I am working full time for a software startup, I realise actually how much I miss this way of working.
I love getting lost deep in a problem space and making something just a tiny bit better at every iteration.
As a developer I constantly have to remind myself that done is better than perfect, at least until your company finds the one thing thats worth improving on. Product designers often don't recognise but working as a designer is such a privilege. Thank you for posting this. It made my day.
You can click through to the author's dome-selling site, they seem to be aimed as a kind of group-fun, beach toy or team-building type of fun project. And it seems the author uses them as a combination point of instruction and ice-breaker while teaching classes in design.
Moreover, though it's an area of history/design I'm not familiar with, it seems they are keeping alive a certain dream of Buckminster Fuller, making geodesic dome kits available to those who want them and allowing for the possibility that it may at some point trigger a worthwhile structural use of the idea.
"I’ve found there’s two main reasons why people build geodesic domes. Some folks do it for shelter and others build them for pure enjoyment of the shape itself."
Total kudos. Now please move onto creating a dome that can serve as a shelter. I believe, to start with, the connectors simply could be made from a thicker rubber, but of course move will be needed. Thank you for your effort.
Given the level of detail in the article, I'm led to believe that the next moment hasn't happened yet..
What you really want is a slit with a pair of small circular holes, one on each end of the slit.
All the benefits of the slit without the tiny points where all of the stress is focused, which is what causes the ripping. Adjust the size of the holes as needed.Common design trick for sheet materials, and should be straightforward to cut with the same machinery.
Enjoyed the article.