Ed K. modified my stock oil cooler which I had retrieved from a
wreck. He cut off the stock tanks which included the thermostat and
replaced them with the tanks  you see. It no longer matters which way
the oil flows through the cooler.     While he was at it, he  cleaned the
core which was found to be full of debris.  Finally he pressured  tested
the unit.  I have placed it in front of the firewall.  Air
will be ducted from the right nostril.  For those who are  interested, Ed
can build a NEW oil cooler using cores that are identical to the  stock
Mazda. He can shorten up the cooler a bit and make it wider  as well,
which should make it  easier to fit under the cowl.  Jerry

That seems to be a pretty good place for an oil cooler. That is where
Chuck Dunlap mounted his. Use a smooth interior surface duct
to feed air to a wedge shaped diffuser on top of the oil cooler.
Chuck used rolled up sheet aluminum.

Paul Lamar

Probably not a big deal (much structural margin), but that mounting  method
imparts bending loads (buckling?) on members that are already heavily
loaded in compression.   Welding anything to the tube at mid span  seems
like it would add a discontinuity which would make it more likely  to fail.
 Especially since the mounts are essentially cantalevered from near
mid-span.


Matt-

Good point Matt. Also you might set up a vibration mode in the tube  that
might lead to fatique at the joints.

Paul Lamar

I thought about that but in this case the loads will be very small
and the convenience great.    In addition to the oil cooler (which is
rubber shock mounted) I have also mounted the coils on  a bracket
welded to the engine mount.  Although I have not done it yet, I
expect use the motor mount to support the radiator (and louvers) as
well.   This probably shows the limits of my imagination but I can
not think of a better way to do it. I like the heavy structure of the
engine mount.  I do not like attaching the oil cooler, radiator, or
coils to the engine because of vibration.  At one point, a couple of
years ago I experimented with attaching the oil cooler to the sheet
metal  firewall, but that came with it's own set of compromises. As a
matter of practice, in homebuilts, it is not unusual to see brackets
for whatever, welded to the engine mount.  Of course that does not
make it "right."  I am not worried about it.  Jerry

Tropical storm Ophelia is bearing down upon  us here in Wilmington.  So
far, not too bad, winds gusting  to 38 mph.  KILM is still VFR, though.
When we lose power, I'll be out of the loop for awhile.

I agree with Matt.  Welds on truss members compromises them to an
unknown degree; especially with respect to compression.  Then adding an
external moment makes it worse.   If  you must attach to the truss
members, I would suggest designing a structure that is attached using
clamps such as these.  Use the ones  with high temp. silicone rubber.
Build a substructure that clamps to the truss members near the truss
joints  and it will have less effect on the structural integrity of the
mount.

Aubrey

 You show two short cantilevered arms in your picture. If you were to
run a slightly thinner support all the way across, connecting the two
mount struts, you might possibly be better off.

   What you need to be concerned about is buckling. You can easily
stand on an empty pop can, but if you just tap the side, it crushes
easily under your weight. A cantilevered support on the mount strut is a
lot like tapping on the side of the pop can you are standing on. It is
likely to put just the right amount of kink in the strut to cause it to
buckle under load. You won't know it is a problem until you pull back
hard on the stick and the engine suddenly hinges under the fuselage.

    If you instead connect the two struts with a single oil cooler
support, it would reduce the amount of twist you are putting on each
strut, and could actually help stabilize the center.

   This is all conjecture unless you do the structural analysis or test
load the mount with the same simulated g load on the oil cooler support.

Bill Dube'