Quad Arms - CF Square Tube vs CF Round Tube vs 3mm CF Sheet

[mmeyer]

I'm hoping to make a CF version of my new design after the success of the wood version. The only issue I have at the moment is what the best material for the arms is. I guarantee this quad WILL take a beating (words cannot describe how many arms I had to make for the wooden version) and therefore need something as rigid as possible. I like the idea of the 3mm CF plate like the blackout quad, it seems to take such a beating. After reading Cyber's Hex post about the thrust column efficiency I'd like to be able to use round tube. On the other other hand the simplicity and ease of building for the square tube is very enticing. At the end of the day though i just want something that is as strong as it can be, so out of these 3 options, which one is the best?
P.S. if there are any other options feel free to tell me them too

 

[Cyberdactyl]

If you want a tried and true design, 2.5mm or 3mm booms is probably the way to go. They certainly don't have the TCE of round carbon fiber, but round booms require a semi-elaborate, semi-circular, opposing mounting clamp system. Not to mention the added mass and the clamp system crowding an already tight space.

There's also one method that I'm surprised is not used more, that's the dirt cheap 10x10mm square carbon fiber booms epoxy filled with fir or poplar. The method is quite popular in ultra strong CF plate composites. They are FAR easier to make and replace than cutting CF plate, and are probably 80% as strong as a 3mm x 1" wide CF plate boom. The square tube is 2g per inch. Wood filled 10x10mm square tube goes to around 3.5g per inch. A 1" wide 3mm CF plate is in the same ballpark just shy of 3g per square inch. But the square booms cut the thrust obstruction by 55% over a 1" wide plate .. IF .. the ESCs are not mounted on the square boom, even on the side, vertical.

Lastly, vertical CF plate has not been accepted, probably because of the extra effort required to mount the motors. However, there's really no extra effort in mounting vertical plate to the frame at the size of a mini, as a matter of fact, I've found the huge advantage it actually requires less fasteners.

I would have thought TCE would be much more a design factor in the minis in that the prop size is so small, the boom width is more critical factor. I'm pretty sure the current crop of minis such as the Blackout are merely the single-step evolution of what the designers began flying back in 2012. I'm pleased to see the warp-quad at least provides the option of keeping the TCE high, at least comparable to a 10x10mm square tube.

 

[mmeyer]

I think initially i will go with 3mm CF plate as this is what i had wanted to do. I need this thing to be as strong as possible.
This pile of wood represents about 70-80% of the booms broken from my flying in the past 4 or 5 days. I tend to push it a little too far a little too often oops

It needs to be as 'me proof' as is physically possible.
Also could anyone make a rough guestimation as to how much stronger 3mm CF plate would be than 3mm ply. Are we talking like 20% or closer to 80%?

 

[Cyberdactyl]

3mm CF plate is so much stronger than 3mm plywood, it's not even close.

I guessing here, but 3mm CF plate probably equates to 1/2" - 5/8" marine grade plywood, or in the neighborhood of a 4 to 1 strength ratio.

If you use 3mm x 1" wide flat booms for a mini quad that has the boom length around 3-4", there's no freaking way you're going to break them unless you dive straight down into concrete from a considerable height, and even then you'll probably only chip the boom. The CF will be the least of your worries.

 

[Tritium]

Riddle me this: Why cant you put a very cheap thin round tube over a square one to help with the TCE? That would give a round cross section AND eliminate the clamping issues with a round boom.

Thurmond

 

[Cyberdactyl]

Sounds like an idea to try. While it would be a marked improvement over a square tube in that the flow would be smooth, with very low turbulence, you still have the problem with 'form drag', or the planform cross sectional area.

Why I've begun to believe a vertical plate is ultimate method, is the form drag is almost zero. First, a vertical plate with even modest stiffness will not oscillate if the motor and props are properly balanced. Second, while the 'skin friction' is rather high, skin friction is a very low variable in the over-all drag of the component if the surface is reasonably smooth. Skin friction can be further reduced by introducing a thin 'boundary layer' on the leading edge and partially down the side by applying an appropriate 'tuned' texture for the expected air velocity, much like the dimples on a golf ball.

The trick of course, is transferring voltage to the motors without disrupting the desired plate width which would be as thin as the material will permit. In the case of carbon fiber, I fully expect a small mini multirotor in the range of a 250-400mm would be 1.5-2mm x 15-20mm plate of around 85mm length. That's what I'll be shooting for in a couple weeks anyway.

 

[Craftydan]

CD,

Not to dispute perfection, but you are remembering the thrust column has a rotational component . . . right?

 

[FinalGlideAus]

I read test data years ago from sailboat makers showing the drag from a round mast was horrible and no real gain from even a square mast. This was at lower speeds which most Multicopters still fly at. Of course an airfoil shape mast provided a improvement but since we do more than hover it becomes complecated. 3mm carbon plate flat is the most durable but has the highest TCE (even though the loss is minimal). Vertical plate presents technical and durabity issues in the construction of the copter and then you loose the efficiency in forward flight. Round and square carbon tube are not durable at all unless you add shear web material like a wooden dowel but that then brings the weight right up high and the grain of the wood is totally wrong to work well as a shear web. Round tube has complexities in mounting motors and doesn't yield the savings in drag as much as first thought. The simple answer is use flat plate and take a 1 or 2% performance loss due to TCE or use square tube with dowel and take a slight weight penalty.

 

[Cyberdactyl]

Perpendicular movement to the thrust, I certainly would agree there would be increased drag. I address that in the post with the chart of suggested transitional drag of a round, vertical flat and horizontal flat boom.

But as I've said, reducing downward thrust obstruction should be far and away the location of highest concern in regard to drag, since it receives unrelenting and constant high velocity air. Vertical plate booms are even better for racing mutirotors where they are tilted and traveling largely in the direction of the z axis, or very close.

As to the rotational air or off-axis thrust that is off the z-axis and not directed outward (as in larger octocopters, etc) I believe that would be truly minimal, but that is certainly something to consider if one wants to absolutely minimize the thrust obstruction. However, I believe and slant of the boom, if indeed it was a factor, would be no more than 2-5 degrees to compensate for the form drag induced by a 'spiraling' thrust column.

But again, you are right FGA, the increase in efficiency is going to be minimal over a round boom. That's why I say that vertical flat booms must at least match the mass and strength of round to make it worthwhile. My contention is blocking ANY thrust is a bad choice IF it can be "easily" avoided without paying a price in increased mass or strength.

 

[FinalGlideAus]

One thing is for sure, multicopter design is still very much in its infancy.

 

[Tritium]

How does all this play out with the motors and props mounted upside down under the arms rather than traditionally?

Thurmond

 

[Cyberdactyl]

How does all this play out with the motors and props mounted upside down under the arms rather than traditionally?
Thurmond

I can offer this up for perusing.