At the suggestion of Rana and Koji, I have worked out some design parameters for a Stewart platform to be used as a vibration isolation device or as a platform for characterization of suspensions. I have made some initial guesses about the following design requirements:
 linear travel: 40 microns peak to peak (based on SOS design requirements in LIGOT950011)
 angular travel: 3 mrad peak to peak (based on SOS design requirements in LIGOT950011)
 payload mass: 5 kg (wild guess of mass of loaded SOS)
 payload moment of inertia: 0.01 kg m^2 (wild guess)
 bandwidth: 500 Hz (suggestion of Rana and Koji: ~kHz)
From these assumptions, I have worked out:
 peak actuator force: 0.88 kN
 minimum radius of top platform: 15 cm
 minimum radius of bottom platform: 30 cm
 minimum height: 26 cm
The combination of high force, high speed, and ~micron travel limits seems to point to piezoelectric actuators. PI's model P225.80 would meet the peak pushpull force requirement, but I have not yet determined if it would meet the bandwidth requirement. Apparently, typical piezoelectric actuators can exert a greater push force than pull force; wonder if one could use an actuator with a smaller force range than the P225.80 if the actuator is biased by compression. (Is this what is meant by a "preloaded" actuator?)
I have attached a PDF explaining how I worked out the actuator force and platform dimensions. (I'll try to dice up this PDF and put the contents in the Wiki.) I also have a plant model in MATLAB with which I have been playing around with control schemes, but I don't think that this is ready to show yet.
Here are some tasks that still remain to be done for this preliminary case study:
 select sensing technologies: integrated linear encoders and/or strain meters, inertial sensing, optical levers, etc.
 study joints: Koji and Rana suggest flexures; I need to propose the joint geometry and material
 study internal modes of the platforms and actuators themselves
 build noise budget
I'd like to ask for input principally on:
 appropriateness of my design assumptions
 piezo actuators currently in use in the lab
Edit: I also added a Mathematica notebook with the inverse kinematics (mapping from platform state to leg lengths) of the platform.
