From: "Richard Schwartz" 
To: "Orion105" 
Cc: 
Reply-To: "Richard Schwartz" 


If you mean twisting around the tube axis, the resistance to twisting comes
from the fact that the vanes are not co-planar with the optical axis.  
From
the end of the tube, you want see this (may the e-mail have mercy on my ascii art):

            |
            |
______|

                   _________
                   |
                   |
                   |

NOT this:

                 |
                 |
                 |
______          ______

                  |
                  |
                  |


In the first configuration above, any twisting of the secondary holder will
cause the vanes to change length a lot.  The change is proportional to the
twist angle, so the support is very stiff for twisting of the secondary
holder.   In the second configuration, the vanes change length only a
little
because the change in length is proportional to the square of the twist angle.

Note that the secondary support can only be as stiff as the tube!  That is
why you frequently find stiffening rings on the tube in the region of the
spider.

. . . Richard

----- Original Message -----
From: "Orion105" 
To: 
Sent: Thursday, February 20, 2003 4:15 AM Subject: ATM Spider design rules of thumb



Are there any rules of thumb regarding the length to width ratio of spider
vanes.  I see them made quite thin (10:1 or higher) and quite thick (4:1 or
so).  I guess you need width to stop rotation of the diagonal, but how much
is optimum?

I have seen spider vanes that taper in width markedly from the outside of
the UTA to the hub.  I have even seen a design that goes the other way.  Is
there any advantage in tapering?  It seems to me that resistance to
twisting is proportional to the width of the vane and so a tapered vane
would be only as good as its narrowest point.  Is this right?

Ado

--- BBBS/NT v4.00 MP