6DOF Micron-resolution adjuster design 1

[stuff]

I am starting to design a sensitive alignment device to carry an optical part and position it with 6 degrees of freedom relative to an optical system. It must stay with the product as periodic realignment will be required. The spec details are:

Optical part dims: 25mm x 20mm x 5mm
Range: 2mm all axes
Resolution: 1 micron
Environment: controlled atmos but 1g motion up to 20Hz
Qtys: 200 off in batches of 20

So it is a lovely design problem really. I am after advice on successful design practice. I am keen to control all backlash, maintain alignment over two years, allow adjustment with simple tools and to provide a neat locking solution.

I may well use a differential screw adjuster as found on micrometers (any design references anyone?). I am also keen to employ flexures as these seem prudent in this case.

I hope this makes sense, thanks in advance for any help.

Matt

 

[sreid]

Buy this book ($20 cheap)"Precision Mechanics." It specifically adresses the kind of device you need to design.

http://happinesstech.biz/index-Precision-Mechanics.html

There is also an excellent book on flexure-"Flexures."

http://www.amazon.com/exec/obidos/t...f=sr_1_2/104-7262443-3194360?v=glance&s=books

 

[tlee123]

I agree strongly with sreid. Buy Dave Kittell's book, "Precision Mechanics." You'll learn so much, plus it's a fun read.

If this is new territory to you, look into "Exact Constraint" by Blanding. ISBN: 0791800857. $70. It's really good.

Back to the problem. Do you really need adjustment in all 6 DOF?

If you do, look into Stewart Platforms (also known as hexapods). 6 legs meet at 3 feet. Maybe you could use a preloaded turnbuckle for the adjustment on each leg giving you 6 DOF. With the right preload and low mass, you may be able to eliminate the need for a troublesome lock. A problem with the Stewart Platform is that there is no pure adjustment in Cartesian coordinates (x,y,z,theta-x,etc.)

Good luck.

Tom

 

[EdgeDesign]

Many thanks guys, the Kittell book looks good and is flying over to me real soon.

I'd already checked out the hexapods, thanks, but it is likely the lack of cartesian adjustment will be a problem as you suggest.

Our alignment process is very likely to require a settting of the rotational DOFs and then the translation (and possibly fine iteration thereafter).

Matt