Hello, I hope your Sunday afternoon is treating you well. I had many questions throughout the past couple days regarding the robot design. I answered most of them in the comments section of KS, and on the individual photographs on G+. But there were three that I felt deserved special attention. PAUL: “… Could you talk a bit about the rationale behind your design? What made you chose that particular number of wheels? What made you chose that particular offset? What prevents it from sliding off? How are you controlling the motors? Are you using any rotary encoders or pressure sensors for getting some feedback in regards to string tension/slippage?”
Paul - Sure, happy to.
First, it’s a long, thin design because of our earliest experience – dating back to 2002 and working with something as simplistic as Lego. We learned that we have better balance and that we didn’t need “control” because gravity would do a lot of our work for us. From there, we established a couple aspects: top/bottom/front/back and how we place various components (like heavy batteries) determine how the thing moves. Also, we discovered early on, that we have a built-in moment arm – the longer we made the ‘bot. That helped with guiding the string to the right spot.
The number of wheels has been a kind of random. We’ve used a little as 4 and as many as 10. That’s one of the things that has frustrated our efforts. There is a delicate balance between the friction (based on the number of wheels) and that mass of the equipment (batteries, motors, communications, etc.,) needed to climb. We have not found that balance yet.
The offset is based on the maximizing amount of time that the ribbon is in touch with the wheels (and, conversely, minimizing the distance between wheels). As it is, it’s only got significant traction on 1/3 of a wheels’ surface, at any given rotation.
The Ribbon rides in a U groove that holds it in position on the outside of the wheel. Tension itself keeps the ‘bot on the Ribbon. We do things called “hang” tests, where we ensure that there is enough friction to that ‘bot can simply hang on the string, in mid-air, without assistance. That determines how many wheels we have/need. We’ve got a brain and comms to the ground, and we are able to control speed and direction of the motors through that. We do not have ongoing strain gauges running to notify us of changes in tension. This ‘bot needs to work through a variety of tautness of the string – from just a few pounds to more than 1000. (In the real world, the wind can pick up or drop off quickly, and that changes the tension of the string constantly.)
UNCLEVER TITLE: “… it's making me think that the tether would weave in between them and thus the elevator would hold tight onto the the tether as a result of the tether's own tension... Maybe I have the wrong impression, but wouldn't the tension be too high for that to be a practical design? Or wait, duh. Lunar Space Elevator, not Earth Space Elevator. So I guess in that case the tension might be much more manageable. I don't know.”
Unclever – See my explanation above, for part of that answer.
Yes, you’re right, we will have substantially less tension on the Lunar System – eventually. But we are a long long long ways from building that system. For the time being, we are stuck here in a 1G gravity well, so we need to make it work here. ;-) Some people think we are trying to leap from the 1 mile high system we built several years ago, and leap directly to a Lunar Elevator in one jump. That’s kinda insane. Nope, we’ve got a complete roadmap laid out – this test, then some record-breaking cubesat designs, then a couple LEO/MEO experiments, two tests we’d like to fly on the International Space Station, and then a big one around the Moon. Then, about 8-10 years from now, we MIGHT have the first stages of a Lunar Elevator. But we have to test everything (at this stage) in 1G.
I’m not trying to build the Lunar Elevator yet. I’m only trying to satisfy the requirements I set for myself for the Kickstarter. 2km would be pretty simple. The thing that has made it so challenging is the 7km stretch goal… (I really wish I’d never posted that damn stretch goal!)
ADAM: “… I really don't see any reasons for using a string instead of a ribbon. Could you please outline the pros and cons of each?”
Adam – We have a philosophy on our team called “No cheater solutions”. Meaning that there are lots and lots and lots of way of building this robot. However, if the long-term goal is to build a bot that teaches us something useful about building a Lunar Elevator, then that limits our options. For example, it would be effortless to put enough string on a single spool, and place a motor on the spool, and then climb. However, that solution won’t work on the long-term Lunar Elevator because it damages the string. And because the string we are talking about is a 250,000km long, billion dollar asset, we’re not going to build anything that damages it. Therefore, we don’t use solutions for this KS robot, which we can’t use in space.
That’s a long-winded necessary preamble. You see, it’s that philosophy that guides our design choices. And with that in mind, let’s move on to answer your question. The Earth Elevator was designed to be a wide, flat ribbon. (Lots of reasons for that. Read our MANY publications on the topic.) However, there are design constraints of the Lunar Space Elevator Infrastructure that preclude this. Namely cost, mass, and whichever initial rocket we use. So, we simply don’t have room/money/lift to build a Lunar Ribbon… so we default to a string. And, back to our earlier ideals, we threw out all (some) of what we’d learned building robots that can climb a ribbon, and have to start over and learn how to build robots to climb string (within the self-imposed and self-limiting constraints mentioned above).
So it’s not a question of pros and cons of ribbon versus string; we like the ribbon very much and would choose it if it were an option. It’s not. We have to design for a string.
Take care, mjl