PSN-L Email List Message

Subject: Re: STS-1 leaf spring
From: Brett Nordgren bnordgren@..............
Date: Wed, 22 Mar 2000 17:10:36 -0500


Sean-Thomas,

Thanks for the quick reply.

At 01:55 PM 3/22/2000 -0600, you wrote:
>Brett,
>
>The only published dimensional info on the STS-1 seems to be in the paper:
>Wielandt, E. and G. Streckeisen, "The Leaf-Spring Seismometer: Design and
>Performance", Bull. Seism. Soc. Am., 72, 2349-2367, December 1982

Yes, thanks, I'd gotten that one from Bob Barns.  Those were the dimensions 
I was starting out with.  Just guessed at the mass (0.41 Kg), and spring 
width (73mm), and don't yet know what to assume for the side-plate mass.

>which has a sketch with some dimensions of the elliptic spring
>configuration. The spring, by the way, is firmly clamped at BOTH ends
>(to the frame and to the outside end of the boom) with the boom
>inside end pivot under the center of the ellipse. Both ends of the
>two-metal spring have to flex where they are massively clamped under
>head-touching-head rows of screws. Rumor has it that adjusting these
>screws is a dark art of the assembly process that can take weeks,
>and that there is a high rejection percentage (hence the high cost).

I'd suspected that there might have been two springs, possibly one Elinvar 
or such and one spring steel.  Since both are clamped, they could be 
arranged to be spaced a small distance apart, far enough not to touch.  I'm 
hoping that the tempco won't be too bad with a single steel spring as I 
have no idea where to get tempered Elinvar or Ni-Span-C .  It might turn 
out that they are adjusting the net tempco toward zero by slightly tuning 
the spring geometry, which I think would indeed qualify as a dark art.

>The transducers are mounted above the spring and above the pivot with
>clear plexiglass fixtures. THey use a proprietary LVDT that accommodates
>the rotation about the flexures, which are ~ 6mm Bendix cylinders.
>The LVDT coils and the feedback/calibration magnets are mounted to
>the frame; fine pigtails connect to the moving coils.

With such a short radius, I'd been wondering how they avoided 
nonlinearities, or even worse, jamming in the bores.

>When I first explored making a VBB vertical in '97, I tried to emulate
>a 2x version of the elliptic spring. I quickly realized that the
>flexure of the ends of the spring was innately involved, and decided
>to mount the spring with isolating flexures instead, and use it strictly
>for one-dimensional lift of the boom rather than a bending movement.

I agree, the fixed-end spring is a hard gadget to analyze.  Think I may 
have found an approach to solve it that's crude but fairly accurate, using 
Excel.  Broke the spring into 600 pieces, assumed an end force and bending 
moment and calculated the bending of each piece as they get stacked 
end-to-end, sort of a poor-man's FEM approach.  After stacking up all 600, 
that gave a trial location for the end of the spring and also its 
end-slope.  Then used the Excel "Solver" function to try different values 
and directions for the force and different end moments until (after about 
15 sec.) it would come up with set that had the spring arriving at the 
desired point at the desired angle.  Messy, but the results are looking 
quite reasonable.

>I can send you excerpts from the manual; most of it is taken up
>with schematics.

Anything you have the time to copy would be greatly appreciated.  Happy to 
cover any costs.

Thanks and best regards,
Brett
__________________________________________________________

Public Seismic Network Mailing List (PSN-L)


[ Top ] [ Back ] [ Home Page ]

Larry Cochrane <cochrane@..............>