Actuation Ideas and Notes

Pneumatics

I had been thinking that Pneumatics would be hard to argue for in a mobile platform due to the poor energy density when a compressor is factored in. On the the other hand, maybe the energy density with compressed air is not so bad.

• There have been some compressed air cars in the news lately. Some from France and elsewhere. The basic idea is old, but the new versions store are at 5000 psi.
• Note that pneumatic muscles do not consume power to maintain an isometric force against a constant load (unlike motors, unless you have a non-backdrivable transmission). Depending what pressure you run the muscles at, it might even be possible to do some multi-stage high/low pressure system or at least to run the exhaust through a turbine for some energy recovery.
• For a large platform, it is not particularly difficult to take power off the engine to run a compressor (which is why pneumatic brakes, doors, etc. are often used on buses and similar vehicles).

-- MarkCutkosky - 06 Nov 2008

EPAMS

What about EPAMS? It seems that the big issue is to get high quality, thin stretchy sheets to improve durability and reliability. If the kind of effort that is going into Li Polymer batteries were applied to EPAMS we'd probably see some significant advances. Otherwise, how far can we go with using readily available materials and clever multi-layer designs? What if we use Dubowsky's concept of only dynamic pulsing? Are the energy/weight ratios good enough to be attractive in mobile manipulation applications?

-- MarkCutkosky - 06 Nov 2008

• EPAMS have several options for dielectric material, the choice of which affects COTS availability, manufacturing, durability/reliability, and performance.

• • Acrylic - VHB 4910/4905. Thinnest you can get is 0.5 mm unstretched, 50 μm stretched. Very good dielectric properties, very good thickness tolerances. Vulnerable to puncture and tearing. Manual pre-streching causes variation which induces fatigue failure in multi-layer structures. Some work that seems promising by Pei at UCLA on interpenetrating polymer network (IPN) to hold pre-strain in place and preclude need for rigid frame. Also, acrylic is very temperature dependent and viscoelastic, which makes it unsuitable for even normal outdoor temperature ranges.

• • Silicones - much better temperature performance, but overall less strain/stress. Can be cast, spun, or purchased as a film. Spinning seems most promising right now (casting hasn't been well-documented in literature). Lotz et al. at Darmstadt have made films down to 5 μm with a 3% variation in thickness. Lots of groups have done multi-layer folded actuators. With good tolerancing, can multi-layer materials be consistent layer-to-layer and not fail through variations in strain?

• Dynamic pulsing is more important for acrylic than silicone bc of viscoelasticity. I don't believe pull-in has been documented in silicones.

• Energy/weight depends heavily on frame material, and with non-prestrained materials or materials where strong frames are not required, the energy/weight is much more attractive. See Wax (Steve Wax from DARPA) and Sands 1999, attached, for interesting comparison of EAP to other technologies.
  • Proceedings_of_SPIE_1999_Wax.pdf: Wax and Sands, SPIE 1999

-- SanjayDastoor - 12 Dec 2008

Hydraulics

Hydraulics: This is the solution used on Big Dog and other large mobile platforms (e.g. systems from SARCOS). If you can afford and engine to run a pump, the rest of the system has excellent power/weight ratio. It is a challenge to find good, light-weight valves. Also, the most efficient hydraulic systems require fast-acting variable-displacement pumps so that you don't dump pressurized fluid wastefully. Waldron used an efficient hydrostatic system on the giant OSU DARPA hexapod.

-- MarkCutkosky - 06 Nov 2008