On Fri, 29 Jan 2010 10:08:07 -0800 (PST), Oliver Arend
wrote:

Inspired by a couple of articles on Bob Hoovers blog and the fact that
it's likely that my first homebuilt will be slow and VW-powered, I was
thinking of making a wood propeller.
snip
I probably made a beginner's mistake and went to the local home
improvement store (Bauhaus, a German chain) to look at the different
solid wood boards they offered. The good side: The choice is huge. The
bad side: I have no idea if the stuff is any good.

Woods available a
- spruce
too soft. although the tensile strength is ok the crush strength in
the hub will not be sufficient.

- birch
probably ok.

- beech
probably very ok

- oak
probably not the best choice since the wood has chemical interactions
with iron.

- eucalyptus (which has neat coloration, btw)
eucalyptus (sold as "australian oak" in australia), probably euc.
deligatensis and about 3 other eucalyptus types can be successful
propellor woods.
read my caveat that comes later though.

- paulownia
very soft new zealand wood suitable as a spruce substitute for ribs
and the like. probably too soft to make a good prop wood.

I guess spruce is widely used in aviation,

spruce is used in other areas of wooden aircraft construction but it
usually isnt suitable for propellors.

The problem I have with it is the huge number of
knotholes, which would weaken the prop.

glad to hear of your concern! it will save your life.

propellor woods need to be denser than the woods used for
construction. wooden construction typically uses woods in the order of
25 to 35 pound per cubic foot density.
my guess is that propellor woods are best in the 45 to 65 pounds per
cubic foot density range.
propellor woods need to be perfect because of the stresses they are
subjected to. in cruise the tips will be flying around mach 0.8
props are made in laminates about 10mm thick (or less) so that any
blemishes and weaknesses can be identified and removed. this makes
certain that the prop is of adequate strength and has no stress
increasing voids or splits within it.
laminating also allows the weight of each blade to be evened up if
necessary by end for ending heavier ended laminates.

as bob hoover says, try as you may, your first prop will be a junk
item. your second prop will probably be usable though.

my first prop was in eucalyptus deligatensis, was all glued up and
shmick (...an australian term for bonzer, which is an australian term
for pretty near perfect)
I carved the back faces by the old practise of tenon saw cuts almost
down to the face and chiseling out the waste then rasp and sanding to
the final shape. the rear faces looked bloody beautiful. all the
cutaway glue joints were broken and found to have impressive
strengths.

work interrupted and I had to leave it for a few months.

on my prop the back face is recessed. note that the front face is the
datum. to do the recess I set up my saw bench to the depth needed and
passed the half carved blank back and forth across the circular blade
to cut out to the depth. something about the eucalyptus makes it
difficult sawing and the rear face was left less than perfect.
work interrupted again.

when I bought my knee mill I realised that by putting the front face
of the prop face down on the milling table I could traverse a cutter
to correct my datum problems with the rear face of the hub. in
practise this worked well.
however.....
while I was sitting there traversing the cutter (manual mill controls)
I noticed a dark line in one of the joints that shouldnt have been
there. I cleaned up a thin automotive feeler gauge and probed the dark
line. to my horror the feeler could be pushed in full depth.
doesnt end there though.
the feeler could be slid along quite a way.
then I noticed another dark line.
the feeler could be slid in full depth and moved quite a distance.
then I noticed another dark line. ...and another. ...and another.

I went back and tested to destruction the joints in all the cutaway
chips. all showed perfect glue strengths.

to admit that this caused some consternation is to understate the
confusion totally. how could joints just a few millimeters apart be so
different? I even located chips that came off right adjacent to some
of the cracks. the jointing in the chips was perfect while the actual
blade was utterly incompetent.

I mulled over this for some weeks before deciding that the bandsaw
would answer my questions.
it would also prevent the folly of trying to repair the half shaped
blank.
when I had bandsawed the prop into six inch lengths I went to town on
destructive testing to understand what had gone wrong.

I deduced 3 things.
1...the wood was not fully cured (had not lost all its surplus
cellular water)and the idle period during summer had seen the wood do
some additional slight shrinkage. this was evident in the very slight
cupping of the faces I had tried to glue.

2...the wood showed a trait of eucalyptus that I was referring to
previously when I made mention of a caveat. occasionally the wood
shows a waxy surface that is hard to detect and quite happily ruins
attempts to glue the wood.
I'm told that cabinet makers will wash down the surface with MEK
thinners and Acetone thinners to dissolve the waxes from the glue
surface. evidently when dried off these will glue satisfactorily.
I've yet to try this but be warned!
also glue squeeze out gives absolutely no hint of surface adhesion
problems.

3... the smaller size of the chips meant that the wood had more
surface area to evaporate water from and the distortions that resulted
in cupping in the blank didnt exist in the chips. these differential
shrinkages can create enormous shear forces in the glue joint.

the waxiness in the eucalyptus was so pervasive in places that the
glue had cured without bonding to either adjacent surface!

I hope something is learnt from all this.
I'm goint to try making prop number two in a few months time.

on the airfield a few months ago a commercial wood prop was
investigated for a paint crack. it too was found to be a dry joint for
half a blade. it doesnt just bite us amateurs.

Stealth Pilot