Pedal Powered Boats

[portacruise]

Originally Posted by Rick Willoughby

Mark
I will get some of the 1/4" and see how it goes. It is USD0.7/ft. I will do 4 layers over 1/4" spring steel if it fits.

I will measure the torsional stiffness before and after.

Rick W

Good info, Rick and Mark.

Rick, I think you have mentioned flex shaft failures occurring near the prop end and something of expected lifetime in a past message. Could you tell if the failure was due to bending stresses as the prop pushes forward or lateral stresses as the prop continuously adjusts to the flow? Or do you think it was it mainly torsional forces that caused failure?

All of my fiberglass/carbon flex shaft failures have been due to torsional stress 1/4 the distance or more away from the prop end. My off the rack tent pole shafts are not wound to resist torsional stress. The shaft cracks along the grain after a few hundred hours of use and becomes rubbery but remains intact enough to limp home. Have never had a clean break where a prop was lost in more than a decade of use.

Vic

Porta

[Guest625101138]

Originally Posted by portacruise

Good info, Rick and Mark.

Rick, I think you have mentioned flex shaft failures occurring near the prop end and something of expected lifetime in a past message. Could you tell if the failure was due to bending stresses as the prop pushes forward or lateral stresses as the prop continuously adjusts to the flow? Or do you think it was it mainly torsional forces that caused failure?
......
Vic

Porta

Vic
The last two failures have been down at the strut. I have two bearings in the strut to apply a bending moment into the shaft and the strut can also take thrust. Hence I do not compress the shaft just bending and torque.

The shafts fail due to the cyclic direct stress from the bending as they rotate. The shear stress is very low and constant. I have a safety factor of about 16 on the torque with the 8mm shaft at my cruising speed. It drops to about 7 in a flat out sprint. Even 1/4" shaft has a good safety margin but it feels rubbery with a 1:4 reduction.

The problem with my previous set up was that I mounted the gearbox a little higher than I laid out and then did not adjust the angle. There was no bending moment applied at the gearbox so it was not a nice even curve. I have since altered the angle a little but it will take a long while before I get enough distance on it to know if it will last. It is still not a perfect curve but it is close.

If you do not have a strut or the strut is not able to take thrust then the shaft breaks roughly in the middle because it is forced into a tighter radius in the middle due to the compression.

If you are using a rod that does not have any diagonal fibres it will be very weak in torsion. The orientation of the fibres has a lot to do with how a composite will perform. For torque, ie shear stress, they need to be diagonal at 45 degrees to the shaft axis.

[portacruise]

Yes, this technique suggested by Mark should give me the torsional strength needed for any shock load I will encounter. All my fiberglass flex shaft failures have been under shock load when the prop tangles in roots, branches or heavy weeds to the point of complete stall under electric power. If the shaft is fairly new, it can withstand the stress without any apparent damage, but that changes after several hundred hours and several tangle encounters.

I think this technique may work even with cheap pultruded carbon fiber Kite rod which I have tried unsuccessfully before. I used 0.18" carbon rod before which had very good spring and stiffness but less torsional strength than 5/16" fiberglass tent poles for my low power use. Probably I can go even thinner on carbon rod and top coat it with one or two layers of braided fiber and have a tiny, almost invisible and extremely lightweight low cost shaft that will hold up very well! Got to look around for the best resins to use with carbon fiber.... Maybe the idea can be adapted to the much higher power levels used in HPB?

Vic

Vic
The last two failures have been down at the strut. I have two bearings in the strut to apply a bending moment into the shaft and the strut can also take thrust. Hence I do not compress the shaft just bending and torque.

The shafts fail due to the cyclic direct stress from the bending as they rotate. The shear stress is very low and constant. I have a safety factor of about 16 on the torque with the 8mm shaft at my cruising speed. It drops to about 7 in a flat out sprint. Even 1/4" shaft has a good safety margin but it feels rubbery with a 1:4 reduction.

The problem with my previous set up was that I mounted the gearbox a little higher than I laid out and then did not adjust the angle. There was no bending moment applied at the gearbox so it was not a nice even curve. I have since altered the angle a little but it will take a long while before I get enough distance on it to know if it will last. It is still not a perfect curve but it is close.

If you do not have a strut or the strut is not able to take thrust then the shaft breaks roughly in the middle because it is forced into a tighter radius in the middle due to the compression.

If you are using a rod that does not have any diagonal fibres it will be very weak in torsion. The orientation of the fibres has a lot to do with how a composite will perform. For torque, ie shear stress, they need to be diagonal at 45 degrees to the shaft axis.

Rick

[markdrela]

Originally Posted by Rick Willoughby

I will get some of the 1/4" and see how it goes. It is USD0.7/ft. I will do 4 layers over 1/4" spring steel if it fits.

When you say 4 layers, is it 1K-tow sleeve or 3K-tow sleeve? With modest compaction, 1K tow fabric is about 0.005" thick, and 3K tow fabric is about 0.012" thick. So 4 layers of the 3K tow sleeve will make the 0.25" rod come out to 0.346" diameter. The last sleeve layer might be difficult to apply, and will have poor fiber angles.

In general, the fiber angle theta (from the axial) that you get is given by the relation
sin(theta) / sin(45) = D / Do
or
theta = arcsin[ (D / Do) / sqrt(2) ]
where D is the sleeve diameter and Do is the undeformed sleeve diameter. So for the 3K-tow sleeve the outer 4th layer whose effective diameter is
D = 0.25" + 7*0.012" = 0.334"
will have
theta = arcsin[ (0.334/0.25) / 1.414 ] = 71 deg
Ugh. That won't work. You'll never be able to expand the last sleeve layer that much. And the above relations ignore the fact that deforming the sleeve away from 45 deg thickens it, so the problem will actually be worse. With 1K-tow sleeve it should be doable.

If this is a problem then you probably want to start with a smaller diameter rod. Or maybe use a larger sleeve for the outer layers.

[Guest625101138]

It is 3k. I was aiming to get the overall diameter to 8mm. If that only requires 3 layers then that will be my limit. The 1/4" steel will handle the torque. I want the CF to increase torsional stiffness and provide corrosion protection. Even a single layer will be worth a try.


I was thinking of using heat shrink to compress it and set the epoxy well.

[portacruise]

How does one go about balancing an RC prop or even one of the folding props which several people have made on this list. Any vibration seems exaggerated when one of these is mounted at the end of a long flex shaft. Is it a dynamic process or just weighing, and how is weight added or removed?

Thanks,

Porta

[Guest625101138]

Originally Posted by portacruise

How does one go about balancing an RC prop or even one of the folding props which several people have made on this list. Any vibration seems exaggerated when one of these is mounted at the end of a long flex shaft. Is it a dynamic process or just weighing, and how is weight added or removed?

Thanks,

Porta

Vic
I expect you will find the vibration is more to do with the angle of inclination than balance. If the shaft is flexible enough to allow the prop to quickly lift to align with flow there is very little vibration. That is why you want a shaft that is reasonably compliant in bending but not compliant in torsion.

The only time I notice vibration with my strutted props is when I turn.

I do not even check static balance these days. I take care with symmetry but as much for hydrodynamic force balance as mechanical balance. The speed of rotation is so slow with a pedal prop that any out of balance force is small compared with the hydrodynamic forces. I have used bent shafts that are a bit annoying because they give uneven loading and vibrate a bit at certain speeds but still useable.

If you set up a rigid strut to force the curve so the prop is aligned you should find the vibration goes away.

[markdrela]

What Rick said. Mass-balancing a prop is critical for an electric airplane, but not for an HPB. The imbalance loads scale as RPM^2, which is tiny for the HPB by comparison.

Far more important is to balance the pitch. ACP props often have mold flashing at the hub, which can skew the prop off-axis if the inappropriate part of the hub is used as the indexing surface. Or maybe the prop warped during cooling after the injection molding.

A good quick way to check the prop pitch is to first point one blade straight up, and make a sighting with one eye, parallel to one blade bottom near the tip towards a distant point. Then without moving your head rotate the prop 180 degrees and sight the other blade. Rotate 180 degrees again and sight again. Repeat a few times.

With some practice and a good eye, you can easily spot a pitch mismatch of less than 1 degree this way. Fancy prop angle measuring gizmos are not needed. It helps to have another person rotate the prop while looking parallel to the shaft, so that the rotation is exactly 180 degrees each time.

[portacruise]

Thanks for your suggestions, Mark and Rick.

My flex fiberglass shafts are run at 400-700 rpm off a hex quick connect at the motor and with a quick release pin at the prop; shafts are slightly under 48" long with 8" diameter to 12" diameter props being run according to motor used and voltage of the power supply. One of the things that helps vibration under steady straight runs sometimes is flipping the APC prop by 180 degrees. Also there are the 6 different positions at the hex shaft and if I try each one, usually at least 2 result in significantly less vibration. With some APC, I get a near perfect drive without any vibration whatsoever at maybe half of the hex positions.

The APC props are bored out from 1/4" standard to 3/8" flex shaft diameter. I have noticed that the molded circle on the hub and standard 1/4 hole don't always match up on the new props. I use a reamer to enlarge the existing hole from both sides and then chase out the middle with a 3/8" drill.

Some of the vibration would appear to be consistent with your suggestions for the cause. So thanks for that, I will try to make sure the hole is centered and pitch consistent on both blades. Glad I didn't go out and buy a blade balancer.

I do get vibration on turns or when starting with quick acceleration of heavy loads but that I expected as the untethered prop/flex shaft reaches equilibrium.

Vic

Porta

Originally Posted by markdrela

What Rick said. Mass-balancing a prop is critical for an electric airplane, but not for an HPB. The imbalance loads scale as RPM^2, which is tiny for the HPB by comparison.

Far more important is to balance the pitch. ACP props often have mold flashing at the hub, which can skew the prop off-axis if the inappropriate part of the hub is used as the indexing surface. Or maybe the prop warped during cooling after the injection molding.

A good quick way to check the prop pitch is to first point one blade straight up, and make a sighting with one eye, parallel to one blade bottom near the tip towards a distant point. Then without moving your head rotate the prop 180 degrees and sight the other blade. Rotate 180 degrees again and sight again. Repeat a few times.

With some practice and a good eye, you can easily spot a pitch mismatch of less than 1 degree this way. Fancy prop angle measuring gizmos are not needed. It helps to have another person rotate the prop while looking parallel to the shaft, so that the rotation is exactly 180 degrees each time.

[MarkX]

You also have to be sure the hole is straight axially, not just central, or one blade will have more pitch than the other.
I do Injection moulding and toolmaking and it looks to me like the APC props are injected centrally and the sprue is then drilled away to produce the hole for the shaft.

[portacruise]

Vic
I expect you will find the vibration is more to do with the angle of inclination than balance. If the shaft is flexible enough to allow the prop to quickly lift to align with flow there is very little vibration. That is why you want a shaft that is reasonably compliant in bending but not compliant in torsion.

RICK, THANK YOU SO MUCH FOR YOUR COMMENTS. NOW I CAN SEE IT IS NOT A PROP BALANCING ISSUE.

I AM THINKING THAT IT HAS TO DO WITH THE THE WAY I HAVE MY FLEX SHAFT SYSTEM CONFIGURED. MY SETUP RESEMBLES A THAI LONG TAIL MOTOR SETUP WHICH BALANCES ON WHAT IS SIMULTANEOUSLY A FULCRUM AND PIVOT FOR STEERING. THAT LONGTAIL SETUP WITH THE PROP AT THE END OF LEVER ARM WOULD SEEM TO AMPLIFY EVEN MINOR VIBRATIONS, WHATEVER THE CAUSE. THE FIBERGLASS SHAFT IS UNSHEATHED AND HAS BEEN "TUNED" BY COMBINING DIFFERENT SHAFT DIAMETERS TO GET THE BEST GRADUATED BENDING COMPLIANCE. THAT IS QUITE A CONTRAST TO THE HPB VERSION WHERE THE SHAFT FEEDS OUT FROM A FIXED, SOLIDLY BOLTED DOWN GEARBOX, WHERE THERE IS MORE ISOLATION FROM WHAT IS GOING ON WITH THE PROP.

The only time I notice vibration with my strutted props is when I turn.

I do not even check static balance these days. I take care with symmetry but as much for hydrodynamic force balance as mechanical balance. The speed of rotation is so slow with a pedal prop that any out of balance force is small compared with the hydrodynamic forces. I have used bent shafts that are a bit annoying because they give uneven loading and vibrate a bit at certain speeds but still useable.

GETTING A BENDING KINK IS ESSENTIALLY IMPOSSIBLE WITH FIBERGLASS (ALSO MAYBE CF?). THE FAILURES ARE NOT CATASTROPHIC SHEARS, BUT GRADUAL WORSENING WHICH GIVES WARNING AND ALLOWS FOR A LIMP HOME MODE.

If you set up a rigid strut to force the curve so the prop is aligned you should find the vibration goes away.

I CAN SEE WHERE A RIGID STRUT IS AN ADVANTAGE IN A HPB BY ELIMINATING THE SUBMERGED GEAR UNIT. A RIGID STRUT IS MORE A LIABILITY FOR WHAT I LIKE TO DO WITH MY SETUP, BECAUSE IT COLLECTS DEBRIS WHICH IS BEYOND MY REACH TO REMOVE. ALSO, USING A STRUT WOULD REQUIRE DEALING WITH SEVERAL ISSUES INTRODUCED BY A REQUIRED RUDDER ASSEMBLY.

Thanks, see comments above.

Vic

Porta

[portacruise]

Good observations, MarkX. This is most probably the source of my vibrations. Since steering is done with my legs that are in constant contact with a steering tiller, any slight pitch error would be more noticeable than for an isolated, bolted down gearbox.

Porta

Originally Posted by MarkX

You also have to be sure the hole is straight axially, not just central, or one blade will have more pitch than the other.
I do Injection moulding and toolmaking and it looks to me like the APC props are injected centrally and the sprue is then drilled away to produce the hole for the shaft.

[spidennis]

I saw this boat at the link below and hope to get to the texas homebuilt messabout/bash this year.
http://mysite.verizon.net/ngc704/TKB...ut/gallery.htm

[xj35s]

Cool, I wish I lived closer. That boat with the sail and small cabin in the picture is like the one I want to build for pedal power. Not sounding too doable tho.

[portacruise]

The escapade is pretty close:

http://www.nauticraft.com/

Originally Posted by xj35s

Cool, I wish I lived closer. That boat with the sail and small cabin in the picture is like the one I want to build for pedal power. Not sounding too doable tho.