X-Mozilla-Keys:                                                                                 

Bill Freeman wrote:
Mike,
Thanks for the clarification.   I had no idea that the mount was so

integrated,

snip for ID

         SNIP

Gary Moir writes:

 1) Bolts thru thin tube need bushings to allow full
    torque w/o crushing the tube.

 
   Sorrry to disagree.   For shear loads the only requirement of the
   nut is to keep the bolt from falling out.  A clevis pin retained
   by a cotter pin is just as good here.  My Long EZ has four
   main tubes, each with two 1/4" bolts attaching it to the airframe.
   Even with larger engines, there have been no problems.
 

 2) Structural bolts need tight hole tolerances, maybe reamed.

 
    Excellent idea, I did all mine motor mount bolts
    that way, reamed with close tolerance bolts. Can't
    hurt, but I'd bet it'd be OK without it.  Might show
   some wear in a few thousand hours, tho.

 3) Tubes in sockets introduce punch loads & stress concentrations and need
    adequate L/D ratios.

 
   True, but typically the actual loads in these many-tube mounts are
    pretty low. If the wall thickness of the doubler tubes is in the
.050-0.090
    range, probably not a problem.  First approximation.

4) Bent tubes aren't trusses, they're bending frames, and change
   the design requirements for thickness and end attachments.

    Agreed, but most of these tubes are pretty short and most bends 
   are small.  Simplifying assumption. Nobody is using rod ends on the
   ends of the "truss" members - but this means they are  stronger
   than a true truss would be.   Of course, any serious bends need to be
   triangulated.
 

        SNIP

Don't forget to inspect the weld! Surface flaws can be detected with

good dye

penetrant (preferably UV) and are most critical, but cosmetically covered

internal

flaws are worse because you need good X-rays to find them. (I've done

safe-life

fracture mechanics analysis for 3 NASA flight safety reviews and lots of

parts.)

Certified welding is worth the price when it's your neck.

 OK, inspect your welds, and get a good welder.   However, 
 we are *not* talking about a highly optimized structure here.  
 I recommend back of the envelope calcs for a 10G download,  
 and assume that this will cover quite a few simplifying assumptions 
 and some minor welding errors.  Typcial welding practice in A/C 
 tube trusses has much more weld area than the adjacent tube 
 (tubes are thin walled) so the weld is called on for much less 
 stress than the parent material.  All the failures I have seen were
 in the tube adjacent to the weld, not IN the weld.  If these were 
 thick walled tubes, this would be very different.  Basically the
 minimum attainable bead size is much larger than the parent
 tube thickness.

There are eight or nine other tubes in the stock
mount that can be done away with.

Don't be too casual about cutting and discarding. Analyze all the cases.

Here's Moir's

rule: Any structure that is Optimized for an incomplete set of loads will

fail when

exposed to the omitted load cases.

 Agreed,  the key word is "optimized" --   these things are
 *far* from optimized.  My guess is that I could design a 
motor mount that weighed 3-4 lbs and would work.  I'd 
rather avoid the X-rays, FEA  etc and have a 10 lb
mount which doesn't need to be X-rayed and can be 
bolted without fear of local tearout, because the wall 
thickness is 2X what is actually needed, there are a
few "unnecessary" brace tubes,  etc.  Over-designing is 
easier, and within sensible limits can be fine.

OTOH, we'd have never gotten to the moon and back with that
philosophy! :-)  I get paid for gov't stuff that DOES need to be
that carefully optimized, but it sure costs a lot and takes a
REALLY big computer.

No need to disagree, I was just keeping it kinda simple.
There are many levels of engineering sophistication, and 
it is sometimes difficult for everyone to be on the same 
page as to how much effort to optimize a particular structure 
is warranted. Save 5 lbs on a 1000 lb plane is OK, but saving 15 
lbs is better, and there needs to be some thought on what it
will cost in time and money to save that particular pound.  I
 push for a bit more structural efficiency than some, perhaps
 a bit less than others.
 
                      SNIP
 
 Take care.
 
 Bill
 
The Aircraft Rotary Engine Newsletter.              Powered by Linux.

http://home.earthlink.net/~rotaryeng/            http://www.linux.org