ingenuir@ix.netcom.com wrote:
Hello, All:
In response to a request from Paul Lamar, and others, for a Finite Element Analysis of
one of the motor mount configurations, I set up and ran the attached model with my FEMAP
program and ran it on NASTRAN. I still haven't checked the hand calculation procedure.
The GS1MotMt motor mount geometry is basically from the Glass Star motor mount in
Paul Lamar's GSLOADS.GIF. dated 1/12/99. Note the aft support is simplified to a single
near-vertical tube, since that is the only necessary reaction on that mount. This mount still
has one redundant reaction since it uses a total of seven tubes. However, the redundancy
is minimized by symmetry. Also, the front fitting for the two bottom-center struts may need
to be moved down to clear engine parts.
I applied three loads simultaneously, which is easy in finite elements, but messy by hand
calculations. The loads are:
1) A 6-Gz pullout (eyeballs down as we used to say) on a 350 Lb engine +PSRU +prop
+accessories.
2) Maximum torque on the prop of 400 ft.lb.
3) Maximum thrust of the prop of 400 lb.
In structural engineering parlance, this "maximum expected load condition" is called a
"Limit load". With a good stress analysis, stresses from it should be checked against
material yield strength and initial buckling of thin sheet.
> A factor of safety of 1.5 is required
by FAR25 for ultimate strength analysis using these stresses. (I know… experimental
aircraft use a different part of the FAR. I just don't have it. J ) Air Force experience
comparing analysis to actual test failures shows a factor of safety of 2.0 should be used if
you don't perform a proof test. On the other hand, 6-G's is more than a general aviation
rating. :-)
This is a very simple, stick model of the assembly. Only a single element is used for each
strut member and a single concentrated mass is used without any mass moment of inertia
or gyroscopic effects from the prop or rotor. This is OK for a first look static analysis, but
this class of analysis should only be used with generous factors of safety. In this case, the
tubes were taken as 0.625 in. dia., 0.049 in. thick AISI 4340 steel and peak stresses were
below 40 ksi in the region of the welded fittings.
Incidentally, I used engine mount data from a Lord motor mount that I used on an
AiResearch environmental control unit. That resulted in a small change of thrust vector with
increased load factors due to the CG location. A softer aft center mount could be used.
Paul, your comments on my previous e-mail about the absence of sideways loads in
normal flight missed the point that the mount system needs to prevent movement (i.e. react
loads) in 6 degrees of freedom for a statically stable installation of the
engine. Specifically,
the struts have to resist translations and rotations in each of 3 independent directions.
(Lateral load capability is only critical for crash L conditions. However, the FAA does
require analyzing them for ultimate strength.)
I'll second Jeff Spitzer's request for a CAD model of a rotary. I'm considering reverse
engineering my '79 RX7's engine, but fenders and stuff are sort of in the way. :-)
Please note, this case is an example only, it is not for design purposes. If anyone wants a
detail analysis of their installation, I can perform both static and vibration analyses, by FEA
or hand methods using their actual geometry. End fittings matter! No charge for initial
consultation, moderate fees for detail analysis for this group.
Ingenuir@ix.netcom.com Phone and fax: 310-378-7076
Hey great Gary. I love to see this kind of stuff.
Paul Lamar
The Aircraft Rotary Engine Newsletter. Powered by Linux.
http://home.earthlink.net/~rotaryeng/ http://www.linux.org
