One for the physicists -
It would seem obvious that the lower rotor in a Y6 (or any vertical rotor set-up) is working in a whole different way to any rotor in a hexa set-up. The fact that it's pushing down "pre-pushed" air suggests, intuitively, that it's not working as hard as it's upper neighbor and therefore not generating as much lift.
So is a Y6 weaker than a hexacopter, with every other spec identical?
Special Instructions for Alternate Frames
Prop Arrangement - http://code.google.com/p/arducopter/wiki/AC2_Props_2
Arduino-Based Setup - http://code.google.com/p/arducopter/wiki/Programming
A few build examples & just a sample of folks who've done it:
These are just the first three I pulled up, there are many more in this community.
Thanks for you help Mike. Ill see what I can glean from these links.
If one uses the simple calculation found on the MK site to determine battery endurance it soon becomes clear that the quad is the most efficient. If one simply adds four motors and converts it into an X8 the endurance drops off dramatically. So we can assume that the increase weight of the extra motors is not offset by any aerodynamic performance despite the fact that the motors are more efficient at lower lift values. Xaircraft V8 is one such example
The conclusion is right, but I think the reason is not. It is not about the weight of the motor. It is a linear use of the available current, additive for each motor, but two factors throw it off.
The lift gains are not linear with the X8, because the props on an X8 are mounted coaxially, you lose 20% of the thrust
The additional lift and more efficient do still help ... but most of these X8 arrangements are built using the same motors as their Quad builds... and are not intended to push the hover throttle range down into a more efficient range, but rather to add more lift to be capable to carry a larger payload. So they usually keep mostly the same hover throttle (usually peek efficiency ~40% throttle [prop && motor combo, not just one in isolation] and simply add more lift at the same lift-per-grams and lift-per-watt ratio. But by going with a coaxial arrangement, they suffer that 20% penalty. A full spread octo arrangement adds a bit in weight, but I suspect that is not why the X8 is often preferred. It is two other aspects which I think usually have folks building to a X8 rather than a full octo... first, that the full octo requires more space for the same prop size... that is, it must be a lot larger if you want the use the same props all around and not overlap them (in fact, you should have something like a prop radius space between the props for greatest efficiency, I believe, there is still a penalty if they don't overlap) and also many X8 are built for video work, and it is harder to get a good prop-free picture with an octo vs an X8.
Just some thoughts...my impressions from my own calculations.
If one studies the lift curves of various props and motors it is quite amazing how there is so little difference in performance. Lift/weight is almost a direct function of watts until you get into quite unrealistic combinations. So the effect of having the lower motor running in downwash seems to be almost self compensating. It simply adjusts its rpm until it is absorbing the correct current. What is very apparent is the relationship of stability against motor distance from the center point. As more turbulence is introduced the stability falls off to a greater extent on larger models. That is very apparent on super stable imu's such as Wookong M. What is also very apparent is that the most efficient props. are not the best for stability. Lighter props. with less dia. and higher rpm seems to work much better. It is always a little disappointing when one finds that the theory does not match up with reality. I hate having to do a complete concept change.
@ Denny: Actually, the theory has been around for 80 years or so and works rather well. In essence, these genuinely are rotary-wing aircraft, with the accent on the wing part. Look at it this way - a sailplane has a long wing span for better efficiency. In this use case, low wing loading is better. However, because of the long span and light loading, it will be less responsive to the controls and more influenced by minor air currents. This might translate into "less stable" for the pilot. It's like anything else in engineering - a compromise.
Do you have a link to the "MK site" you mentioned? It's always good to review someone else's data.
Just as a note on efficiency, almost all of these garden-variety model airplane propellers are dreadful at static thrust efficiency. See this linked paper and note that for low advance ratios (J), most fall below 30%. Just as a comparison, full size helicopter rotors reach 70% or more.
I have a far more reliable way of doing things it's called suck it and see. I have a simple device that measures static thrust. Once I have created a lift curve for each motor/prop. combination and I know the AUW then I divide that figure by the number of motors (which in my case is always 4 as it is the most efficient.) I then know how much each motor will draw. Then it's 54 x lipo/ total current.
That allows for 10% losses which seems to be very accurate.
Try this it works every time.
With the emphasis on having 6 motor as opposed to 4 for redundancy I think the butterfly concept would be better than a conventional Tri. coax. config. As seen at Minsoo Kims Worldwide Multicopter Shop.
Agreed they all suck at static thrust hence my own designs, however it is not that straightforward. Static thrust is almost a function of absorbing all of the available power into the largest thrust dia. The reality is that this does not fit in with the control functions that need very high response speed so a compromise has to be reached. My most efficient design of blade is swept forward. It is however totally unacceptable due to the danger aspect. All of the designs benefit from construction in autoclaved carbon with diadetically laid tows. blade stiffness and lightness are the key factors. not possible with plastic, wood or wet lay-up epoxy carbon. That is why F1 cars are mostly glued together with autoclaved components.