Re: [ATM] 6-pole

["Roy Diffrient" <diffrient@toast.net>]

Per the definition given below by Nils Olof, "axial forces" in the members 
eliminates "bending" -- think tension and compression, with forces acting in 
the same direction as the members, trying to lengthen or shorten them. 
Bending implies perhaps cantilevered beams, with forces acting perpendicular 
to the length of the member, not axially.  As other posts here have said, in 
the usual 8-member Dob truss, the side triangles support the weight of the 
top end -- The top and bottom triangles, acting as cantilevered beams, 
typically will not even support their own weight at the deflection allowed 
by the side triangles while supporting the upper tube assembly weight. 
Those upper and lower triangles, however, do function to provide symmetrical 
resistance to horizontal forces pushing and pulling the Dob tube assembly, 
and provide a parallelogram linkage, eliminating tilt of the secondary 
relative to the primary as the structure deflects.  But that is really a 
carry-over from the Serrurier truss as used for an equatorial mount where 
the truss tube assembly rotates around the optical axis -- In that case a 
symmetrical truss structure can provide constant deflection regardless of 
the angle of rotation of the tube assembly around the optical axis. 
Practical considerations can drive the size and shape of truss members, and 
not just for handling -- structural collapse from buckling is also a concern 
for very slender members in compression.  This subject has been covered here 
several times before, and there is a rich resource in the ATM archives --  
Look up posts on trusses by Krajci and myself about three years ago for 
example.  Also, "Flexure of a Serrurier Truss", Sky and Tel, Feb. '94. 
Krajci's web site examines several truss structures, including 6-member 
trusses, which can work well.  But the simulations and calculations from 
Grape and other truss computation programs and formulae may not account for 
all the factors involved in creating the truss deflection -- It's difficult 
to identify and quantify all of them.  It takes very little slop in mounting 
those truss members to significantly affect the truss deflection.  The 
real-world result is that there is usually going to be more truss deflection 
than calculated.  Bottom line: Build it stiffer than you think you need to.

Roy Diffrient
_____________________________________________________________________

>Date: Thu, 10 Apr 2008 22:12:23 +0200
From: "Nils Olof Carlin" <nilsolof.carlin@telia.com>
Subject: Re: [ATM] 6-pole
To: "Arjan te Marvelde \(hetnet\)" <arjan.te.marvelde@hetnet.nl>,
<atm@atmlist.net>
Message-ID: <001301c89b47$51deac30$2646e755@dittda7685d135>
Content-Type: text/plain; format=flowed; charset="iso-8859-1";
reply-type=response

Arjan,

one definition of truss:
"A triangular arrangement of structural members that reduces nonaxial forces
on the truss to a set of axial forces in the member."

Reminds me a bit of "Recursion, infinite: see Recursion, infinite" but I
guess the triangles are essential.

In a practical telescope structure, my not-so-educated guess would be that
tube truss poles just enough to withstand bending forces might be so thin as
to bend easily or at least vibrate, or else be too thin-walled to take
handling well. Thus, in practice, structural rigidity might not be the
determinimg factor. BTW, in my only truss tube telescope, I use alu "T"
profile, removing the "vertical" part at the ends, fastening the remaining
flat piece with a screw with conical head. I like to do things differently
;-)

In my smaller 'scopes, I make a tube of square cross-section, with walls of
thin plywood right now some 3 mm of the cheapest stuff, but if I could find
some really thin model airplane plywood (a mm or so), I'd use it - I don't
expect it would bend much in its own plane, and anyway, it would form
something like a true Serrurier truss, FWIW, with a square frame a little
below midway supporting the trunnions.

Nils Olof


> This is precisely the point I try to make. You should really calculate how
> well the definition of your structure is (call it rigidity) as a function
> of the components and design. Tom Krajci had a link to a mechanical
> calculator program for truss structures:
> http://www.grapesoftware.mb.ca/
>
> What you see very often is overdimensioning of the trusses, in diameter
> and hence weight. Just take it a bit further and you can do with a single
> pole (I think it is technically not right to call this a truss) like Russ
> Porters' garden scope. Of course, your suggestion to use a hollow pole and
> put the optical axis in the center has been used on several occasions.
>
> The original statement I responed to, was that 8 trusses would overspecify
> the position. This is only true when the mounting points leave rotational
> freedom and, like a real truss should do, only define distance between
> mounting points.
>
> AtM
>
>> But the idea of a truss is that the members should not have significant
>> change in lenght during load (even if they are not very resistant to
>> bending) - a much less demanding requirement than stiffness against
>> bending or rotation as required with less than 6 members. Thus, the
>> primary requirement is sufficient cross-sectional area for the material
>> and loads. But whether a 6-member truss or a 8-member one with 3/4 as
>> large cross section and weight would be stiffer, I don't know. It should
>> be calculable - there might even be ready-made solutions to be found.
>
>



_______________________________________________
ATM mailing list http://www.atmlist.net/