Using only 4 out of 6 propeller mounting bolts

[rv7charlie]

This may have some bearing on the issue......something interesting I found when bending T3 aluminum in a pan break.....the traditional method of setback of the clamp did not produce the proper bend radius because the hard material would lift and bridge to the clamp and the radius would be sharp........

so perhaps the stiffer IVO prop when torqued .....the fibers are bridging (or warping?) between the bolts and not contacting as well as a wood prop....obviously there is a difference. Just an idea.

No; the knurling actually bites into the composite surface of the blade, to increase resistance to movement. The knurling was one of IVO's attempts to stop blade movement; the initial design had smooth faces on the hub plates.

The critical difference between the IVO and traditional props is that the IVO doesn't use drive lugs mated to close tolerance holes in the back of the blade.

edit: The IVO isn't stiffer; it's quite a bit more flexible than carbon or even wood props.

 

[rv7charlie]

As a corollaries to this question:

What does engineering 'best practices' say to do about friction in the joint when calculating the load that can be carried in shear by a bolted joint?

Why do cantilever wings use *close tolerance* bolts to secure the wing to the fuselage?

Why do wood wings typically have metal bushings inserted in their bolt attach points?

 

[kubark42]

The critical difference between the IVO and traditional props is that the IVO doesn't use drive lugs mated to close tolerance holes in the back of the blade.

This got me curious to actually do the math. I'm using the existing propeller attached to the MZ35 two-stroke on my AC-5M:

mu = 0.2  # Stiction coefficient, taken from emperical tests for wood on aluminum, c.f. https://www.coursehero.com/file/p6l7pr84/Aluminum-on-Smoothened-Wood-Note-I-calculated-the-acceleration-and-coefficient/
d_2 = .100  # Propeller hub OD
d_1 = 1 *.0254  # Propeller hub ID

# Calculate radii
r_2 = d_2/2
r_1 = d_1/2

# Let's assume that the crush plate evenly distributes the bolt forces
num_bolts = 6
F_a = num_bolts*12500  # 18N-m torque on an M6 (5mm minor dia), using https://www.engineeringtoolbox.com/bolt-torque-load-calculator-d_2065.html

# Calcualte theoretical torque coupling
# Eqn. 11, https://x-engineer.org/automotive-engineering/drivetrain/coupling-devices/calculate-torque-capacity-clutch/

T = 2*2/3*mu*(r_2^3-r_1^3)/(r_2^2-r_1^2)*F_a

---

T = 1051.4

So the theoretical maximum torque which can be transmitted is around 1050N-m. Of course, the stiction coefficient could be off significantly, but I suspect the orders of magnitude are still correct.

I have a 18.3kW engine spinning the prop at 2280 RPM, which means around 75N-m of torque. However, the average engine torque is really only delivered during the cylinder's power stroke, which in the case of the single-cylinder MZ35, is a single pulse lasting about 45 degrees of the rotation. So we can make a crude average and say that 360/45 = 8, and therefore peak torque is 75*8 = 600N-m.

This gives a safety factor of around 1050/600 = 1.75. This is reduced somewhat further by the prop loads mentioned by @wsimpso1, however those loads are small relative to the engine's power pulse. So it's reasonable to say that the safety factor is north of 1, but only by bare margins. This, combined with IVO prop's experience, makes me think that friction alone is not responsible for holding the prop in place. There might be some bolt shear which helps prevent propeller walk under certain peak conditions.

 

[Map]

Another piece of information:
when calculating engine mount loads, the FAA requires the use of a factor of 2 for the actual torque for 4-cylinder piston engines, or a factor of 4 for 2-cylinder engines. That should tell you something.
For turbine engines, the torque factor is only 1.25. I would use that for an electric motor.

 

[Map]

As a corollaries to this question:

What does engineering 'best practices' say to do about friction in the joint when calculating the load that can be carried in shear by a bolted joint?

Why do cantilever wings use *close tolerance* bolts to secure the wing to the fuselage?

Why do wood wings typically have metal bushings inserted in their bolt attach points?

The bearing strength is the problem with the latter. Wood and composite have much lower bearing strength than metal. That is somewhat improved by bonding metal bushings in. It also minimizes wear when the joint is repeatedly assembled.

 

[rv7charlie]

The bearing strength is the problem with the latter. Wood and composite have much lower bearing strength than metal. That is somewhat improved by bonding metal bushings in. It also minimizes wear when the joint is repeatedly assembled.

And...Why won't simple friction work for a bolted wing joint, instead of close-tolerance bolts in reamed holes? Shouldn't simple friction keep the joint from moving?

 

[kubark42]

Why won't simple friction work for a bolted wing joint, instead of close-tolerance bolts in reamed holes?

You typically don't want truss joints to transmit torque. It's much harder on the joint, and it makes analysis harder. Much better to pin the joint and let it rotate freely.

When you are working with wing joints which are designed not to move, you frequently (maybe always?) use close tolerance bolts. I have had to use liquid nitrogen to shrink the spar bolts sufficiently to reinstall the wings on an Piper Arrow and a Beechcraft Musketeer.

 

[Marc Zeitlin]

When using the same size bolts (3/8") and torqued at ~30-50% less torque than an IVO, would a wood prop have *more* friction against the smooth flange than the composite IVO blades have against their hub flanges, which are knurled and bite into both faces of the blades?

This is an interesting question. So I'd need to be more rigorous and actually do some math (which I'm loathe to do), but here's my theory.

Standard hubs - either fixed pitch, or C/S - are one piece. This means that ALL the bolts (and the surfaces under them, across the flange) that clamp the prop to the flange are involved in reacting the torque pulses - IOW, the moment arm for reacting the torque is the full width of the hub.

But the IVO props have 3 piece hubs (or however many blades they have, I suppose) which means that the clamping of the bolts is very localized, only to the area directly around the bolt, and does NOT span the whole flange face. So the reacting moment arm is substantially smaller than with the standard props. With the very large torque pulses of the 4 cylinder, large displacement lycomings, the small moment arm is not sufficient, even with the somewhat higher clamping force and friction, to withstand the pulse. If the moment arm is only 1/3 of the standard props, then even 1.3X the torque will not be enough to keep the blade from moving on the hub - 300% or the pressure would be required, not 130%.

_IF_ IVO bridged across the center of the prop hub to effectively make the multiple blade sections into one rigid section across all the bolts, then it would work the same as standard props.

That's my current working theory, and I THINK it holds up.

 

[rv6ejguy]

A data point here. My IVO has 1/2 inch thick 6061-T6, knulred plates on each side of the blades so bolt forces are quite well spread out. 60 lb./ft. on two 1/2 bolts per blade. Blades have never moved. Has been totally reliable in my application for a long time now.

 

[rv7charlie]

As far as I know (haven't followed IVO in recent years), IVO is still OK with using their prop on engines with reduction drives (higher frequency torque impulses), but quit selling for use on Lycs over 20 years ago, specifically because torque peaks from bigger 4 cyl engines exceed the clamping force available in the hub.

Again, 55-60 lb constant speed props are constrained perfectly on the hub with only 4 bolts *using drive lugs* that are near-interference fit.

 

[proppastie]

Why do they not have drive lugs?

 

[rv7charlie]

Why do they not have drive lugs?

No, that's *my* question. ;-)