Paul,
I am not sure I agree with the idea of small sections of tubing
welded together. I had a few projects inside the GE turbine
components plant in Wilmington NC where I got to see lots of
turbine
parts made. Early on, the interior spool sections of the
turbines
were cut from a large solid block of Inconel. This was
incredibly
expensive so they experimented with small sections welded
together.
None of the welds were sufficient for the stresses involved
partly
because the welding itself altered the hardness in the
location of
the weld. The solution was to use Inertial welding. Each
rotor is
machined and broached to hold the blades. One part is
fastened in a
large (100 ton and up) flywheel and the other is fixed to a
large
stationary hydraulic ram. These parts are adjusted to within
.002
inches in any axis, then the flywheel starts spinning at 10
to 20
RPM. After the spinning starts, the part on the ram is
forced into
the mating surface of the flywheel mounted part, causing the
entire
approximately 3 inch mating surface to be one large weld
with the
heat being spread out over a large area avoiding localized
heat issues.

I have welded lots of aluminum, and I would be very
concerned about
the welds and the adjacent metal being able to withstand the
forces
placed on them during flight.

Why wouldnt you make a plug from wood or plastic on the
lathe for
the inside and the outside and use it to make a sand casting and
cast the mold yourself? Its not difficult to make a mold.
You need
sand and some bentonite clay. Look up any local directional
boring
contractor in your area for the bentonite. We use the
bentonite to
condition the "mud" when we drill to keep it thick and hold
the hole
open for the pipe. After you make the mold, melt the
aluminum and
cast the part. Put it on a lathe, clean it up, and make your
part.

Kevin Alderman

I think you are referring to these parts.

It all depends on the stress. These are very low stress parts.

There is no tension load as there is a 1/2 inch bolt that goes
through it
holding all the parts together. Tension load capabilities on
a half
inch bolt is about ten tons :).


Paul Lamar




   I thought the first round of these emails were referring to making
the case for the PSRU, not the mounting brackets. Sorry if I missed
something.

   The loads on those mounts should be mostly in shear, which should
be fine. The PSRU on the other hand sees tension, compression axial and
torsional loads in differing amounts on every part of the unit. I would
be very concerned about welding together aluminum pieces to make one of
these.

   Kevin Alderman

   The planetary PSRU case sees no torsional loads unlike a spur gear
case.
   Using needle thrust bearings on the first bulkhead this about
   1000 pounds max thrust. The tube containing the ring gear is about
six inches in
   diameter. If the wall on that part is .250 inch the area of the
material
   is about pi times 5.75 time .25 or  4.52 square inches.
   Therefor the tensile stress is 1000/4.52 or 222 psi. Since aluminum
is good
   for 40,000 psi 222 psi is near nothing.

   The other load applied is if you are in a spin of one radian per second
   and the prop weighs 40 pounds there is a 650 ft pound bending
   load applied to the over all gear box.  See the attached Skyranch
jpg's.

   Lets say it is only 4 inches in diameter on average and the wall
   is .25 inch. This is a cantilever round tube beam 15 inches
   long subject to bending   The load on the end of the beam is 650
foot pounds
   times 12 or 7800 inch pounds divided by 15 inches or
   520 pounds up or down  load. The load is always 90 degrees
   to the direction of the rotation in calculating gyro loads.

   http://www.engineersedge.com/beam_bending/beam_bending9.htm
<http://www.engineersedge.com/beam_bending/beam_bending9.htm>

   The moment of inertia of a 4 inch round tube with a .25 wall is

http://www.engineeringtoolbox.com/area-moment-inertia-d_1328.html
<http://www.engineeringtoolbox.com/area-moment-inertia-d_1328.html>


   /I_y  = π (d_o ^4  - d_i ^4 ) / 64/ See the attached jpg.

   pi X (256-150)/64 = 5.2

   Going back into the beam bending web site and plugging that number
   in gives a stress of 219 psi and the deflection is 0.00001 inches.
   Also inconsequential.

   Over the years we have gone over this several times in reference
   to the stiffness of the NC machined adapter plate. In fact Jerry
Hey built
   a static test rig.  See the jpg's.

   Check my arithmetic and go over the steps as I did as I may have
   made a mistake.

   Paul Lamar

   -- Its not the adapter plate that concerns me, its the aluminum
weld joint between the various aluminum pipe sections. I have seen too
many aluminum welds crack and fail due to fatigue to bet my rear end on
a welded attachment like this. I see the test rig has what looks like
either a steel pipe or a DIP attached to a flange. If it is DIP, usually
they thread these and do not bother trying to get a weld to hold, if it
is steel they may have welded it to the flange. The plate looks like the
machined plate made from a single piece of aluminum, again with no welds.

Our aluminum dump beds on our dump trucks would always break at the
welds. Always. The aluminum handrails we installed on marina docks would
break at the welds, the pipe never gave any problems. Even if a boat or
golf cart hit the railing, the welds would break and the pipe would bend.

My PSRU if I were going to make one from scratch would be turned from a
solid piece or cast, not welded.

KA


I can see how that could happen when welding railings.

Normal practice with aircraft welded parts is to normalize the part after
welding. An ordinary self cleaning kitchen oven can do that with aluminum.
To restore the T6 hardness you normally  send it out to a heat treater.
At 200 psi stress it is hardly necessary.

Many aircraft parts are welded. Piper cubs fuselages are all welded.

Paul Lamar


    The Rotary Engine News Letter. Powered by Linux.
    ACRE NL web site. http://www.rotaryeng.net
    You Tube http://tinyurl.com/beqqxas
    Copyright 1998-2014 All world wide rights reserved.