Tensegrity design software with real world construction experiments
Anyone who has ever built or tried to build a tensegrity structure can tell you how difficult it can be. This comes from the fact that it’s structurally minimal, avoiding redundant parts, and it gets its integrity from the tension. That means that only when the tension is “finished” does it actually hold together.
The main goal for this project is to find ways to make the process much easier, so that many more people can do it and we end up seeing a whole lot more tensegrity structures in the world around us.
The software was built to open up the design space so that anyone can dream up more complicated and interesting pretenst structures. But the challenge to actually build them gets much bigger as a consequence.
When we think of building a house, we think of are made of stacking solid bricks on top of one another, maybe with something like cement trowelled in between to stick them together. Gravity holds the bricks down as we build the walls, and we lay down some thick beams horizontally when we need to support a floor. It all seems more or less like an exercise in stacking solid things. There are pushing and pulling forces throughout, but they’re hidden to us.
A bridge or a really tall building with some kind of skeleton structure goes a bit further, so then we need to start thinking about where all the forces come from and where they go. This is what civil engineers do for a living. How much weight can this bridge hold? How thick do the beams need to be to hold up this building? Maybe even: How does this building behave during an earthquake?
Civil engineers don’t just use concrete, because they know that concrete on its own will just crumble and fall apart. As we’ve all seen when we peek through the fence at a building site, the first thing they do is lay down a lattice of steel rods, and only later do they pour the concrete. It’s only the combination of the concrete and the pre-tensed steel rods that actually makes it last. Pre-tension is everywhere! It’s just mostly hidden.
Building tensegrity structures is generally quite difficult, because the end of every bar has to be connected to four or more cables which converge there.
We might think of it as top of a circus tent where the pole holding the tent up carries the canvas of the tent, only instead of canvase there are separate cables. The pole pushes against the pulling force of the cables. The hard part is making a durable connection which joins the cables to hold back the pole, and the reason it’s hard is because the connection is almost inevitably the weakest link.
The software-based design process doesn’t help at all for this because software lets us work in a kind of ideal world where a connection is just a logical thing. This virtual bar is pushing, and being held back by these virtual cables which are pulling, but all that the software cares about is the fact that they are connected, not how they connect.
To make building of pretenst structures easy, we have to somehow solve the connection problem. That means making connections which do not end up being the weakest part, and to make these connections simple and beautiful to behold.
Anyone who has ever built a tensegrity structure will know that there’s something very tricky about the process. Since all of the compression elements are floating, not touching each other, and tension is what gives the structure its integrity, it doesn’t really hold its shape until the last moment.
So while you are building, the structure is unstable and teeters back and forth instead of being stiff and solid. Only when you make the last connection and tighten it up does the structure go completely taut and hold its shape. On the one hand this can be frustrating, but it also goes to show how a tensegrity is a minimal structure, with every part contributing to the whole, so it actually gets to the heart of the matter.
Mostly this instability happens because we tend to build by progressively making more and more tension cable connections one by one until we get to the final tension cable is in place and ready to be tightened. This makes it hard to actually make the last tension connection as well because it must be done while the structure is almost taut. The tighter the structure, the harder this gets.
Now there is probably no way to completely avoid having incomplete tensegrity structures be slack enough to teeter until they are finished, but there is an approach which completely avoids having to tighten the tension lines to make the transition from slack to taut. The whole process changes when we turn things around and come from the other direction.
Instead of making individual tension connections one by one until the structure is taut, we can prefabricate the tension network, and only insert the bars later. All we need for this is bars which can be somehow lengthened a bit after they have been inserted and connected. It should be easier to lengthen bars in the near-final taut situation than it is to tighten cables. They need not lengthen very much at all, even a percent or two may be enough, depending on materials.
If the network of cables is constructed beforehand, all of the connections can be made while the cables are in a totally relaxed state (slack, just laying on the floor), which should make it very much easier to get the connections just right.
Building with prefab tension has its own challenges because the measurements must be made very precisely and an inactive tension networkcan make it difficult to keep straight which cable connects to which joint. It can become confusing, which is probably where the software comes back into play. It could be a kind of paint-by-numbers process, involving labeling things and following straightforward instructions.
The really dramatic advantages of prefabricating the tension becomes clear in the final stage because with the tension network finished, the only thing that remains to be done is insertion of the compression members. They would have to be some kind of solid bars with a “twist”. They must be able to lengthen slightly after insertion, so by some kind of wedge or screw mechanism.
One of the things that would be best avoided if we want the connections to be simple, strong, and beautiful, would be knots. Knots take up space, and introduce asymmetries, detracting from the aesthetic beauty, and probably even introducing weakness.
Braided rope is usually has something like twelve individual strands which interweave, and if a central core is not intentionally added, they are hollow. The idea of a hollow cable or rope may offer interesting options for connecting the ends at the rather complicated multi-cable connection points. The hollow tension line could be extended around something solid at the end and therefore make a strong connection without any knot.
Braided rope (made often from extremely strong synthetic material like Dyneema) behaves somewhat like a chinese finger trap. Push it and it widens as the strands separate radially, but when you pull it tightens to become as thin as it can be. These attributes are used by sailors when they prepare rigging for their boats. For example, an eye at the end of a braided rope is made to feed back in through itself in such a way that the eye does not at all represent a weak section of the cable.
Also interesting is that braided rope can display a kind of muscle behavior if somehow the hollow center of the rope is forced outwards. Just like the muscles in our bodies, they pull by shortening while they thicken in radius. The thickening acts like a kind of lever, multiplying the force which pushes outwards perpendicular to the length of the muscle.
Perhaps this bulge effect could also be used to introduce the last critical bit of pretension in a tensegrity structure, which could then have the great advantage that compression members could be nothing but solid bars. Instead of lengthening bars, we would see cables bulge somewhat. It’s an option worth exploring.
Either way, we will be able to create the entire network of cables or ropes while in a slack state, because they only later have to bear the pretension.