cell design and uneven loads [was: Re: ATM Glass Advice]

["Jeff Anderson-Lee" <jandersonlee@sbcglobal.net>]

If you look at it, there is a difference in tolerance with uneven loads.

Assume you have a 100mm bar and design for even loads but end up
with a 1mm error in positioning of the fulcrum (i.e. 49mm:51mm). That
leads
to a force ratio of 51/49 instead of the designed 50/50 or about a 4.0%
error.
If you have the same bar with 3:1 loads and the same 1mm positioning
error (e.g. 24mm:76mm) you end up with a force ratio of 76/24 versus
75/25
or greater than a 5.5% error in the ratio (which is greater than 4%).

If you consider the forces (assuming for simplicity a 100g load) you end
up with
a 51g:49g split instead of 50g:50g in the first case, or about +/- 2.0%
error
in the loads.  In the second case you end up with 76g:24g instead of the
designed 75g:25g.  That's a +1.3% difference on the heavier load, but a
-4.1% difference on the lighter load (once again, greater than 2%).

Thus, there is a difference in the tolerances.  However, if you stick
with load
ratios of less than 3:1 you might be willing to live with that change in
tolerance.

That said, I'd be interested to see your designs for 19 and 21 point
cell along
with your proposed support plans.

Jeff

----- Original Message -----
From: "Mutalib Abdallah" <omegatroid@hotmail.com>
To: <jandersonlee@sbcglobal.net>
Sent: Thursday, January 17, 2002 1:32 PM
Subject: Re: ATM Glass Advice


> All of the configurations I mentioned to you are easily made with
levers and
> triangles.   The tolerances for unequal loads are no more precise than
for
> equal loads.   I assume you know how levers work, so that you can
figure out
> where on the lever to put the balance point.
>
>
> >From: "Jeff Anderson-Lee" <jandersonlee@sbcglobal.net>
> >To: "Mutalib Abdallah" <omegatroid@hotmail.com>
> >CC: "atm mailing list" <atm@shore.net>
> >Subject: Re: ATM Glass Advice
> >Date: Thu, 17 Jan 2002 08:14:50 -0800
> >
> >I haven't played much with uneven loading, and generally stick with
cell
> >patterns that are easy to break down into levers and triangles to
spread
> >the loads.  Numbers that don't factor into a product of twos and
threes
> >might be harder to build although some could work better if you could
> >build them.  Uneven loads ARE possible with just triangles and
levers,
> >but make tolerances more crucial if you are going to keep the weight
> >distribution correct.
> >
> >That a 20" mirror of 7/8" plate glass could be supported on 18 points
> >seemed good enough for me.
> >
> >Mutalib Abdallah" <omegatroid@hotmail.com> wrote:
> >
> > >Did you try any cells that use the center of the mirror as a
flotation
> > >point, such as the 19 point?   It should work slightly better than
the
> >18
> > >point.   Also, there is a 21 point cell that looks like an 18 point
> >with an
> > >inner ring of three points.   Plop should be able to study all of
this
> > >easily.   Be sure to allow plop to adjust the angles as well as the
> >loads of
> > >the points.
> >
>
>
>
>
> _________________________________________________________________
> Send and receive Hotmail on your mobile device: http://mobile.msn.com
>