What are the basic tensegrity modules that can be used to build something like an articulating bipedal system? A tensegrity description at the scale of the whole body must proceed differently than tensegrity descriptions of cells, tissues and organs. There are provisional levers and contingent fulcrums operating at the level of the vertebral body. They are being augmented and supported however by a tensegrity fascial mesh which encompasses all parts of the body, muscles, cartilage, ligaments, tendons etc. That’s where the tensegrity is – in the envelope.

Context: Tom Flemons Archive

Tensegrity modules for articulating bipedal systems

(Jan 12, 2016) The basic problem seems to be designing and building a plausible tensegrity simulation of the body. Over the years I have been working on this from various angles. My paper The Geometry of Anatomy  discusses how to create a minimal tensegrity description of the body. I really went at it from a geometer’s point of view. What are the basic tensegrity modules that can be used to build something like an articulating bipedal system? I looked at anatomy and tried to find homologous tensegrity structures that could perform similar functions. I looked for bilateral symmetries, chirality, and then tensegral linkages or couplers that would combine modules to create a working system. I emphasized connecting modules because it eventually dawned on me that a) the body as evolved is far too complex to model as a single tensegrity system and b) there is nothing in traditional definitions of tensegrity that involve articulating structures. Tensegrities model envelopes, hydrostatic or pneumatic structures that act like inflated membranes, whereas a joint should really been seen as a failure of a tensegrity – a disjoint. Only by combining separate tensegrity systems by means of tensioned linkages could I hope to even come close to the working of an articulating body.

There are probably many ways of going about this – what I’ve tried to do is only the beginning. I think the next step is to render similar attempts in the NTRT software and see where that takes us. I envision that as the accuracy of the modeling improves it will become possible to use a tensegrity simulation of the body to diagnose dysfunction. Imagine 3D photographing a patient and superimposing that onto a simulated tensegrity body. The tensegrity gets distorted to map accurately onto the patients form and in doing so the distortion tells us where the force vectors are and what needs to be adjusted to bring the patient back into a minimal energy configuration which is homeostatic balance. It would finally be an empirical methodology that would have some numerical data behind it.

I’m hoping that the next generation of biomechanical engineers could be trained in understanding tensegrity at a deep level and could turn this into a science. In the meantime all this research should also help us design a new generation of modular tensegrity robots – something that is already underway at NASA and several universities.

A tensegrity description at the scale of the whole body must proceed differently than descriptions of cells, tissues and organs

(Aug 10, 2015) Re: simple models versus complex models – there are several ways to look at this. If complexity is a matter of numbers of components my knee model involves about 12-16 struts to model essentially three bones not counting the patella which would require another six struts (to model a tetrahedral shape which the patella is) and double that number of tension components. So it’s maybe a complex enough model of the force vectors operating through the knee especially if I add cross-linked tension members to simulate the anterior and posterior cruciate ligaments which produces a rolling revolute joint which the knee is.

There are undoubtably a gazillion changes going on every second in the billions of cells the knee is composed of and the trillions of atoms those cells are composed of but my models don’t speak to that – they are concerned with the structural forces required to articulate a leg. We can model movement quite authentically using tensegrity without including fractal concerns.

In fact I would argue that the fractal tensegrity hypothesis leads to some faulty thinking. Yes it appears we can model atoms, molecules, cells, tissues, and even organs as nested tensegrity systems and we can include the fractal ladder that is built between these different levels of scale in our description. But it’s a jump in logical type to say that just because cells or tissues look and act in a certain tensegral way, that a fixed anatomical system based on rigid parts articulating across joints behaves in the same tensegral manner.

I think it is a lot harder to prove the case for tensegrity in vertebrate anatomy than it is at smaller scales in the body. Again it’s a question of envelopes versus interiors. A cell is an amorphous blob jostling thousands of other adjacent amorphous blobs in a matrix or array which forms an amorphous tissue and thence an amorphous organ (in the sense that one end of the liver looks pretty much like the other end of the liver). They are self-supporting only in as much as they are engorged with fluids, gels, and colloids which defines them as hydrostatic structures. (Internal pressure standing in for the compression components in the tensegrity model.) The body in contrast, has a relatively simple internal skeletal support system which allows for appendages, articulations and a variety of unique joints. We are only approximately bilaterally symmetrical and in every other plane we are highly differentiated structures.

Has no one made note of the fact that morphologically a vertebrate skeletal superstructure bears no formal resemblance to the fractal tensegrity systems from which is composed? This poses a problem for Biotensegrity that has never been addressed. A tensegrity description at the scale of the whole body must proceed differently than descriptions of cells, tissues and organs.

The argument I make is that the cosmos is organized by fundamental laws that reveal tensegrity principles of connectivity no matter what the scale or what the arrangement of space and matter. Looking at something at one scale doesn’t necessarily reveal a tensegral order but it’s there at another scale. e.g. turbulence in a flow of water is chaotic but the molecules making up the water are not.

This is why I suspect it’s a red herring to worry about joint capsule spaces between the bones and levers operating in the body. There are provisional levers and contingent fulcrums operating at the level of the vertebral body, no doubt. Rejecting this in the name of some pure version of Tensegrity just misleads us. By the same token, these aren’t exactly like the levers in our machines. But we can’t deny they exist. They are being augmented and supported however by a tensegrity fascial mesh which encompasses all parts of the body, muscles, cartilage, ligaments, tendons etc. That’s where the tensegrity is – in the envelope.

So back to my models for a moment, they do capture a lot of fluid lifelike motion to the extent that they are complex enough to model the fascial envelope as well as the bones inside it. The greater the ratio between the tension members and the compression members the more lifelike the model is likely to be. In other words the membrane is what holds everything in its place and the membrane is complex and continuous.

I’ve been trying to answer these basic questions in my papers and my models for a long time and I finally think I’m getting somewhere. It’s led me to focus more on tensegrity biomimicry and applying it to robotic systems. In terms of using these principles when building robotic and prosthetic structures, a lot of compromises and simplifications have to be made but that’s another email.

Finding tensegrity in the anatomy

(Sept 5, 2015) I’m feeling a bit discouraged at this juncture regarding the overall hypothesis of biotensegrity as applied to structural anatomy. It appears I’ve changed my mind considerably over the years and now I have a hard time finding tensegrity in the anatomy. I don’t think you find it in the joints because I don’t think tensegrity models joints (a caveat here though – I think I have found a way to model joints that on the surface at least appear to be clusters of tensegrity components – but it’s not clear at all to me that you can find the same geometry in the body). I think the only place to look for it obviously is in the fascia but I can’t see a way to test the hypothesis. Clearly fascia is a felt like substance that has multiple strands oriented in many different directions – some of them must carry loads similar to my helical tensegrity mass but that doesn’t necessarily prove that’s how it’s being done. Just because we are a tension based body doesn’t necessarily mean it is a tensegrity.

Related discussion: Layered Design for a Full-Body Tensegrity Model