Direct drive SBC

[RJW]

AD,

We figured this out. Positive discussion of a shared problem is fun and always fruitful. At the very least it helps us see where we are making mistakes (the big-beam idea). Often it leads to really cool stuff. If what we found in this discussion is correct then all of us (even the LS guys ) now have a cheap way to check our motors for resonance.

Nathan,

I agree about the “should work” principle. I’ve done it myself too many times. As soon as I get the inverted junkyard 305 smoking I’ll hook up our garage “torsiometer” and see what happens. It reminds me that I should have a crappy motor set up on a test stand at all times. It would be handy for testing all kinds of stuff.

If anybody is interested in discussing the 180-degree V8 crank idea maybe we could start a new thread? I’d like to talk about the mechanical advantages/disadvantages of fitting one to an otherwise stock V8.



Thanks again,

Rob

[Autodidact]

A question; I think there are two basic types of direct drive designs, a solid extension and one that has a very flexible coupling. They might both be good depending on the size and mass of the propeller, i.e., a large diameter, heavy prop might need a flexible coupling to keep the resonant freq below range, and a smaller, lighter prop might work well with a solid extension. So, if using a solid extension, which do you guys think is best: one that bolts solidly to the crank flange, like Wittman's drive? Or one that has splines driven by the crank coupler shown below AND an extension that fits into the pilot bushing hole?

[Natty Bumpo]

AD I like your ideas for how to accomplish the direct drive. they closely reflect my own thoughts. I propose to test the relative merits of a hard spline to spline arrangement vs. a sprung center similar to a clutch disc. It seems to me that an engine instrumented as you and RJW have been discussing would allow a fellow who understood to discern any benefits of either or neither setup. Good propeller data could also be gathered concurrently. Another poster has used an automotive wheel hub on the bell housing and has a lot of hours on it with no signs of ill effects. I wonder what would happen If I put an axle cv joint in the bell housing. It would actually simplify the problems of joining the two.......

Would it also have any positive or negative affect on the whole vibration mumbo jumbo............?

[Autodidact]

Another poster has used an automotive wheel hub on the bell housing and has a lot of hours on it with no signs of ill effects.

Is that the guy with the Wittman 215 Buick conversion? His handle is WittmanW10 or something close to that, can't remember it right off.

I wonder what would happen If I put an axle cv joint in the bell housing. It would actually simplify the problems of joining the two.......

Would it also have any positive or negative affect on the whole vibration mumbo jumbo............?

As long as it was tight, I don't think it would make a difference except that it is kinda heavy and would add a small amount to the inertia - but not much. I suppose that since it won't be operating at different angles, you might get away with no boot or grease, maybe? Would also make the shaft much easier to fabricate.

I've decided to calculate the polar moment for two different wooden propellers - a 68" and an 80" one - and see how their radii of gyration compare as a percentage of the propellers actual radius. If they are similar (and I think they might be) then I could use the result to quickly estimate the inertia of most old style props, if the mass is known. Then I will be able to design a torsion shaft that is strong enough and flexible enough to handle the engine torque and keep the natural frequency below engine idle speed excitations. It can be done, but the question is, "How long will it have to be?", if it is too long, then that's why I think a springloaded hub of some type could be added to the flywheel to increase the flexibility. We'll see in a couple of days, maybe.

Concerning the torsiometer (is that what they're really called?), if it can work and be made to work by regular guys, I think it would be a great tool for not only direct drive experiments, but possibly for those wanting to do gear reductions as well. Regardless of what your calculations say, it's always necessary to test them and you need a way to take the measurements.

[RJW]

I prefer a splined coupling. Roller bearings on one end of a shaft and hydraulic bearings on the other end is generally not a good idea. A spline will take care of minor differences in alignment and runout.

I prefer a stiff coupling over a soft coupling partly because of decreased complexity. Also I’m guessing a stiff coupling might end up being lighter (think of large diameter, thin-walled tube like a drive shaft). Which coupling to use will ultimately be decided by the moment of inertia of the prop and where resonance in the whole system occurs. My preference is stiff everything, light prop, and tune the natural frequency above the operating range. I would still prefer a stiff coupling if the natural frequency occurred at a low enough rpm where the engine wasn’t producing enough power to hurt anything. If the whole system resonated at say 1200rpm then it would shake a bit as you powered through this rpm but nothing would break. The difficulty with this though is that the second mode vibration might end up in the usable power range.

If the “torsiometer” works (I made up the word) then I’m sure a workable cheap and light coupling could be found. Stiff coupling and light prop would be my preference if it can be done.

Rob

[Toobuilder]

Do a Google image search for "Corvette drive shaft coupler". There are several different products available from the stock rubber, to poly, to solid aluminum. These hande much abuse day in and day out and might just match what you are looking for.

[Autodidact]

torsiometer
An obsolete term for an instrument for measuring ocular torsion, cycloductions, and cyclophorias. ...

I'm sure you did make it up, but I found this definition on google somewhere; since it's obsolete, I say we appropriate it!

[Autodidact]

I graphed a 64" prop and an 80" prop. The radius of gyration came out to 32.5% of the prop radius
for the 64" prop and 32.6% for the 80" prop. Using a density for birch of 39lb/ft³, I got 6.5lb for
the 64" prop and 13.3lb for the 80" one; I think this is actually pretty close because I was reading
an article on fixed pitch wooden props for RVs on the Van's AF forum that said they weighed about
12lb. Something interesting was that the Inertia for the 64" prop was 1.3 times that of the 30lb
flywheel (in spite of it's 6.5lb weight) and for the 80" prop it was about 4.2 times the Inertia of
the flywheel. Here are the graphs I drew up - I sliced the props into 1" sections:

[Jay Kempf]

I graphed a 64" prop and an 80" prop. The radius of gyration came out to 32.5% of the prop radius
for the 64" prop and 32.6% for the 80" prop. Using a density for birch of 39lb/ft³, I got 6.5lb for
the 64" prop and 13.3lb for the 80" one; I think this is actually pretty close because I was reading
an article on fixed pitch wooden props for RVs on the Van's AF forum that said they weighed about
12lb. Something interesting was that the Inertia for the 64" prop was 1.3 times that of the 30lb
flywheel (in spite of it's 6.5lb weight) and for the 80" prop it was about 4.2 times the Inertia of
the flywheel. Here are the graphs I drew up - I sliced the props into 1" sections:

Wow AD,

That is awesome. It makes my head hurt to see that on paper sideways. 1.3 vs. 4.2 is an eyeopener for engine/crank/bearing design to hold onto different radius props.

[RJW]

Great work AD! Pounding this stuff out by hand is tough but rewarding. It makes me wish I had paid more attention to my calculus teacher! And now I’m too lazy to write a program to do this stuff for me.

Your results are telling. Even with a wood prop it seems it will be pretty difficult to use a stiff coupling and at the same time avoid resonance. Maybe we could get away with a stiff coupling using the 64-inch prop. Doesn’t look good for the 80-inch prop though. An 80-inch prop is what I have been wanting to use. Might have to go to a soft coupling. Rats!

I can’t tell at the moment how many blades we are talking about. Are your results for two blades?

Keep at it. I’ll try to get a junk engine running on a stand in the next couple weeks so we can start testing some of your math.

Thanks for the good work!

Rob

[Autodidact]

Wow, thanks! I didn't think it would be so interesting!

Even with a wood prop it seems it will be pretty difficult to use a stiff coupling and at the same time avoid resonance.

Maybe not, at least for small diameter props. When I misused the single pendulum frequency equation on the crankshaft, the error was that it assumed that the position of the first mode node did not change; in reality, it would move out along the extension toward the prop and so the natural frequency wouldn't drop as much as my result said it would - instead of dropping from 300 to 203Hz, it would probably only drop to about 230-240Hz or so.

I’ll try to get a junk engine running on a stand in the next couple weeks so we can start testing some of your math.

Oh oh, now I'm getting nervous!

Only two blades.

Do a Google image search for "Corvette drive shaft coupler". There are several different products available from the stock rubber, to poly, to solid aluminum. These hande much abuse day in and day out and might just match what you are looking for.

At first, I thought I liked metal springs better, but now that I think of it, if the rubber "squib" (edit: ok, it's called a "guibo"? what in the... OK no, it's a "giubo", sheesh!) is used to put the frequency below idle, then as the rubber wears out the frequency will tend to get even lower (unless the rubber hardens and becomes brittle?); it might actually work better. I'll bet PTAirco has been over this territory already.

[Jan Carlsson]

Nice work AD. I checked this one in US units. it is a 70 inch diam Beech prop, output test only.
it will rotate around the Z axis



Mass properties of Luscombe ( Part Configuration - Default )
Output coordinate System: -- default --
Density = 34.96 pounds per cubic foot
Mass = 10.24 pounds
Volume = 0.29 cubic feet
Surface area = 6.67 square feet
Center of mass: ( feet )
X = 0.00
Y = 0.00
Z = -0.01
Principal axes of inertia and principal moments of inertia: ( pounds * square feet )
Taken at the center of mass.
Ix = (1.00, 0.00, 0.00) Px = 0.30
Iy = (-0.00, 1.00, 0.00) Py = 7.65
Iz = (0.00, 0.00, 1.00) Pz = 7.77
Moments of inertia: ( pounds * square feet )
Taken at the center of mass and aligned with the output coordinate system.
Lxx = 0.30 Lxy = 0.02 Lxz = -0.00
Lyx = 0.02 Lyy = 7.65 Lyz = 0.00
Lzx = -0.00 Lzy = 0.00 Lzz = 7.77
Moments of inertia: ( pounds * square feet )
Taken at the output coordinate system.
Ixx = 0.30 Ixy = 0.02 Ixz = -0.00
Iyx = 0.02 Iyy = 7.65 Iyz = 0.00
Izx = -0.00 Izy = 0.00 Izz = 7.77

And by using only half propeller, but with half hub area. CG 21,6% from center.

Mass properties of Luscombe-Inertia ( Part Configuration - Default )
Output coordinate System: -- default --
Density = 34.96 pounds per cubic foot
Mass = 5.12 pounds
Volume = 0.15 cubic feet
Surface area = 3.33 square feet
Center of mass: ( feet )
X = 0.63
Y = 0.01
Z = -0.01
Principal axes of inertia and principal moments of inertia: ( pounds * square feet )
Taken at the center of mass.
Ix = (1.00, -0.01, 0.02) Px = 0.15
Iy = (-0.00, 0.96, 0.27) Py = 1.77
Iz = (-0.02, -0.27, 0.96) Pz = 1.84
Moments of inertia: ( pounds * square feet )
Taken at the center of mass and aligned with the output coordinate system.
Lxx = 0.15 Lxy = -0.01 Lxz = 0.04
Lyx = -0.01 Lyy = 1.77 Lyz = 0.02
Lzx = 0.04 Lzy = 0.02 Lzz = 1.83
Moments of inertia: ( pounds * square feet )
Taken at the output coordinate system.
Ixx = 0.15 Ixy = 0.01 Ixz = 0.01
Iyx = 0.01 Iyy = 3.82 Iyz = 0.02
Izx = 0.01 Izy = 0.02 Izz = 3.88

In aluminium, with a reduced hub, it weighted 51 lb at first, still seems heavy. but the hub thickness is large.

Mass properties of Luscombe ( Part Configuration - Default )
Output coordinate System: -- default --
Density = 174.80 pounds per cubic foot
Mass = 42.46 pounds
Volume = 0.24 cubic feet
Surface area = 6.26 square feet
Center of mass: ( feet )
X = 0.00
Y = 0.00
Z = -0.00
Principal axes of inertia and principal moments of inertia: ( pounds * square feet )
Taken at the center of mass.
Ix = (1.00, 0.00, -0.00) Px = 0.88
Iy = (-0.00, 1.00, 0.00) Py = 37.97
Iz = (0.00, -0.00, 1.00) Pz = 38.17
Moments of inertia: ( pounds * square feet )
Taken at the center of mass and aligned with the output coordinate system.
Lxx = 0.88 Lxy = 0.06 Lxz = -0.00
Lyx = 0.06 Lyy = 37.97 Lyz = 0.00
Lzx = -0.00 Lzy = 0.00 Lzz = 38.17
Moments of inertia: ( pounds * square feet )
Taken at the output coordinate system.
Ixx = 0.89 Ixy = 0.06 Ixz = -0.00
Iyx = 0.06 Iyy = 37.97 Iyz = 0.00
Izx = -0.00 Izy = 0.00 Izz = 38.17





Jan

[Autodidact]

Thanks Jan, my inertia units are slugs/in² and my brain hurts when I think too much so I haven't tried to convert your data to something I can compare to mine, but I noticed that the half prop CG is 21.6% of the radius and if I'm correct in thinking that this is the radius of gyration, then the Inertia of this type of propeller (I call them spade blade props, after the playing card symbol) is quite a bit lower than the 1925 porpeller from an old NACA paper that I used as a model. This makes me think that it could be used with a solid (torsionally stiff) crank extension and that would be great since that would be the lightest way to go, but not necessarily the cheapest.

It's a nice looking prop too.

A 4 or 5 blade small diameter prop with this type of low inertia blade, like you and John (Synergy) are talking about, and with very light weight blades would work well on a fast airframe, I think.

[Jan Carlsson]

AD,

The fixed pitch wood prop above have a fat hub, I just played with the shape to match a DH moth prop hub, so that clump is in the calculations and move the "CG" inboard as well as the slim blades.

By the way, now I have to ask, I often see the units slugs, and even use it sometimes, but what the hellsinkii is it? where does it come from?
I would be surprised if it to be found in the SI unit system. :-)

[Autodidact]

...,but what the hellsinkii is it?

I think SI is much easier for engineering but I've thought in pounds and feet all my life and it's difficult to get it out of my head.

Actually, slugs is lb/32.2ft/s², and I converted that to lb/in/s² and I'm not sure what to call those, inch worms?