PhysOrg News Monday, Oct 29 2007
random 9:07 pm
I often enjoy the news items from PhysOrg, a top-notch science news site. Here are some from just today:
Scientists found a clam that lived for 400 years. A non-sentient clam gets to live 400 years and we humans typically drop dead at around 80? Doesn’t seem very fair.
A Japanese Institute is taking robotics to the next level, creating a system that learns through gestures rather than just executing pre-programmed routines.
UC San Diego scientists found that the T4 virus contains a molecular motor with twice the power density of an automobile engine. My thought is, “that’s it?” Scaling laws should enable molecular motors with thousands or even millions of times the power density of an auto engine.
Lots of people are keen to modify their appearance surgically. 48% of women were interested, 23% of men. As the procedures lower in cost and increase in elegance and utility, more people will sign up for cosmetic surgery. Of course, I take this to mean that many people will embrace enhancement-oriented surgeries and implants when they become available 10-20 years from now. (We already have artery-cleaning, bio-powered micro-robots, after all.)
A UK scientist has brought a 53 million year-old spider “back to life” by scanning tiny fossil details with x-rays and reconstructing a 3-D digital image. This is made possible by recent advances in scanning resolution.
October 30th, 2007 at 5:13 am
I am not aware of any current treatments that make use of artery cleaning biopowered microrobots. Can you provide a source for more information?
October 30th, 2007 at 7:44 am
Here.
October 30th, 2007 at 1:17 pm
“My thought is, “that’s it?” Scaling laws should enable molecular motors with thousands or even millions of times the power density of an auto engine.”
I think this business of scaling laws for motors needs a bit more thought. Firstly, one should say that for heat engines - like auto engines - scaling laws very much don’t help you when you go small. A heat engine needs to sustain a temperature difference between a hot reservoir and a cold reservoir, and the law of thermal diffusivity tells us that the time we can sustain a given temperature difference scales as the length squared, so as you make a heat engine smaller you have to run it faster and faster until (before you get to the nanoscale) you exceed a maximum frequency which scales like the speed of sound divided by the size.
This is why biological molecular motors are not heat engines at all, but isothermal motors driven, not by temperature differences, but by chemical potential differences. Essentially they operate by rectifying Brownian motion, so this means the relevant energy scale is set by kT. For this reasons power densities don’t get to be much more than the one described (why they come out at roughly the order of magnitude of an auto engine probably is more than coincidence but I’d need to think about that more).
I think the scaling you refer to comes from Drexler’s electrostatic motor design. This is fine (apart from the difficulties actually building one described in my old post “6 challenges”), but it’s not really fair to compare the power density of an electric motor with, say a petrol engine. In an electric motor you’ve still got to get the electrical power to the motor, so there will be Ohmic losses in the wiring (scaling, of course, as the inverse size squared), and you’ve still got to find a source for the electricity. Maybe you’ll use a fuel cell, in which case you will probably be limited by how fast you can supply the cell with fuel (again, with very unhelpful size scaling for flow rates and viscous losses that come from the Poiseuille formula). Finally, as I describe elsewhere I am deeply sceptical that you will be able to run a mechanical device at high power densities without leaking energy into vibrational modes (this leakage looks very obvious on the animations posted on http://machine-phase.blogspot.com/).
Sorry to reply at length but I thought I would try and cheer you up with a technical discussion after all that pesky cultural studies stuff.
October 30th, 2007 at 1:40 pm
Thanks Richard. This clears things up a bit! Of course I’ve read your “6 challenges” post and think it makes many important points. I do see the vibration you’re talking about in the Machine Phase animations, and look forward to NanoEngineer being released to the public so that more people can run simulations and look at the issues in closer detail.
I thought it was suspicious that a biological motor would have similar power density to an auto engine, hence my comment. I am getting the power density estimate from the electrostatic motor numbers. Question: what do you think lowering the temperature to only a few K could accomplish for nanoscale robotics? Could the Ohmic losses be circumvented by using superconducting circuitry? Could resonant vibrations be at least partially avoided by decreasing the temperature?
It does cheer me up to understand this, thanks.
October 31st, 2007 at 9:58 am
Decreasing the temperature is always going to be good for MNT. Superconductivity doesn’t necessarily save the day for getting your electricity around without losses, though - remember that superconductors have a critical current density above which they revert to being normal conductors, so you’ll hit this problem as you try to decrease size and increase power density.
I looked up a very handy review article about biological motors, which gives the following power density values: (all in ergs s-1 g-1)
Actin polymerization 1e9
Microtubule polymerization 5e8
Myosin II 2e8
Kinesin 7e7
Vorticellid spasmoneme 4e7
Typical passenger car engine 3e6
Striated muscle 2e6
Bacterial flagellar motor 1e6
Thyone acrosomal reaction 1e5
Limulus acrosome reaction 1e4
Eukaryotic flagellum 3e2
It’s a bigger range than I would have guessed. I think this must be because, though the energy per power stroke is constrained to be of order kT, the effective frequency can change over quite a few orders of magnitude.