Shop our Airplanes Products Drone Products Sales
Thread Tools
May 30, 2012, 08:00 AM
dalbert02's Avatar

Increase multi-copter efficiency with better propellers

Propellers available to us today were designed primarily for airplanes and I believe there may be some efficiency gains to be had with better designs. Propellers were "tuned" for moving aircraft and I don't believe the engineers envisioned that they would spend most of their time spinning in a stalled configuration hovering a multi-rotor aircraft. If you look at propellers designed by Aerovironment or NASA for slow moving aircraft, it appears that they are much wider and of greater pitch near the hub than we commonly see on offerings by Xoar, APC, Master AirScrew, etc. I am not an aerodynamic engineer, but there must be a reason why the professional props look very different than what we use. May anyone offer insight as to how we may improve propeller design for multi-copter use in an effort to increase flight times and/or payload capacity?

Attached is a pic from NASA's Helios, a solar powered, slow moving, long endurance aircraft.
Sign up now
to remove ads between posts
May 30, 2012, 08:11 AM
HerkS's Avatar
Prof Selig and his students have tested model props in his low speed wind tunnel at UIUC.

Its opening statement sounds a bit like your question.

"While much research has been carried out on propellers for full-scale aircraft,
not much data exists on propellers applicable to the ever growing number of UAVs.
Many of these UAVs use propellers that must operate in the low Reynolds number
range of 50,000 to 100,000 based on the propeller chord at the 75% propeller-blade
station. Tests were performed at the University of Illinois at Urbana-Champaign
(UIUC) to quantify the propeller efficiency at these conditions. In total, 79 propellers
were tested and the majority fit in the 9- to 11-in diameter range. During
the tests, the propeller speed (RPM) was fixed while changing the wind-tunnel
speed to sweep over a range of advance ratios until reaching the windmill state (zero
thrust). To examine Reynolds number effects, typically four RPM’s were tested
in the range 1,500 to 7,500 RPM depending on the propeller diameter. Propeller
efficiencies varied greatly from a peak near 0.65 (for an efficient propeller) to near
0.28 (for an exceptionally poor propeller). Thus, these results indicate that proper
propeller selection for UAVs can have a dramatic effect on aircraft performance."
May 30, 2012, 08:38 AM
Grad student in aeronautics
Sounds like you want a rotor and not a propeller
May 30, 2012, 09:45 AM
dalbert02's Avatar
Thank you Herks. I went to the website and read the document and also explored the data provided on various props.

If I am interpreting the graphs correctly, it looks like some of the GWS props perform rather poorly and Graupner Slim CAM props and APC props perform rather well, which matches my assumptions without tangible or measurable proof.

I apologize for my ignorance but I could not find definitions or explanations for some of the letters used in their calculations.

η (Greek eta - I think this is efficiency)
P, p and ρ (Greek rho)
n n
Cp (coefficient of power?)
Ct (coefficient of thrust?)

Thank you.
May 30, 2012, 09:55 AM
Suspended Account
yes, first one is efficiency,
rho is air density (kg/m3)
n is rotational speed ( rps not rpm),
Q is torque (N.m),
J is advance ratio (J= v/nD)
D is prop's diameter (m)
Cp is coefficient of power (Cp = P / (rho.n^3.D^5) )
Ct is thrust coefficient (Ct = T / (rho.n^2.D^4) )

there is a correction to make (mounting fixture drag correction) to the data:
May 30, 2012, 02:09 PM
B for Bruce
BMatthews's Avatar
The props on the small multi rotor copters are a LOT smaller than the ones you are looking at in most cases.

All else being equal a bigger diameter and slower turning prop will do better at static performance.

You mentioned running in a stalled condition. By using pitch values that are low for the diameter you can avoid running them in a stalled condition. For props in the 8 to 9 inch diameter range I found from static testing of motors that up to 4 to 5 inch pitch values run in a non stalled state. But once you hit 5 to 6 inch pitch the sound changes and the currect required to spin at a given RPM rises dramatically. To me that indicated that the flow over the blades became separated or "stalled". So stick with lower pitch values and you gain a lot of efficiency.

Due to the very low reynolds numbers for model propellers when you look at the first two inches out from the hubs it is pretty normal to find that the propellor makers reduce the real pitch over this portion to something well under the proper amount so as to work at creating nothing more than a low drag "connector strut" which holds the middle and outer portion of the blades which do the work.

If you want to try some testing it would not be that hard to make a true "helical" pitch prop and run it with a wattmeter and thrust guage. The watts used per oz of thrust is a good indicator of the final efficiency. For such testing you could use balsa or spruce wood props running at low to medium power consumptions so as to avoid blowing up the wood props.

From our own testing that a buddy and I did some years back involving various commercial plastic and wood props and trying to modify the airfoils and thickness to achieve optimum performance for the poor motors of the day I found that the airplane props are really not that bad. And that included making up a couple of different props from scratch that had the wider inner sections as you see in that picture. Seems like our props were too small to gain much from the wide inner areas. And the true tip to tip helical pitch with wide hub area produced great amounts of drag for props in the 10 inch size range.
May 30, 2012, 06:34 PM
dalbert02's Avatar
@ Yomgui Thank you, I appreciate you defining the symbols.

@ Bruce, Thank you, I value your opinions. Currently I am running 2820/14 motor spinning 14x4.7 APC props on 4S3P A123 (6900mAh) or 4S2P 4300 LiPo (8600mAh) packs. The APC prop is the larges prop I could find available as a counter rotating version. From there I chose a motor that would be slightly overloaded at full throttle, in this case 36A max rated motor and 38A to 40A actual draw at full throttle.

I have been able to lift 14lbs but would like to increase my flight duration. Adding battery packs adds flight time with diminishing returns. Adding another battery pack adds a few minutes and yet another pack adds perhaps only 1 minute. Some people have been able to achieve 60 minute flight times without payload and Aerovironment claims 40 minute flight time with EO and IR payloads. How would you best tackle the mission of lifting 2-3kg for at least 30 minutes and an ultimate goal of 60 minutes? I seem to be chasing my tail and could use another perspective. I was thinking perhaps my answer lies in prop design, but I could be mistaken. Even using xcopter, I can't seem to break the 30 minute barrier with any payload.
Thank you kindly,
May 31, 2012, 04:05 AM
Registered User
Hmm, perhaps using a large fixed pitch rotor to carry most of the weight at the maximum possible efficiency and 3 or 4 smaller props to provide torque compensation and control? The mixing and stabilization would require a fair bit of onboard logic, though.
May 31, 2012, 08:55 AM
Grad student in aeronautics
Along the lines of what Brandano is saying, disk loading (thrust/area, where thrust = weight) will be one of the biggest factors determining your endurance.
May 31, 2012, 01:09 PM
B for Bruce
BMatthews's Avatar
That Xcopter app is pretty sweet.

For giggles I played around with it using a single type motor and battery pack and only varied the prop charactaristics.

For my example I chose a Turnigy G32-770 powered by a 4000 mah 3S3P pack on a 100 oz 3 rotor copter. I only offer this as passing info.

The telling point was the propellors used. I didn't get decent results out of the duration until I went with oversize props driven through a reduction gear setup. With normal fixed pitch props I could only get around 15 minutes. And in most cases less.

Once I started using the custom sizes and adding in gearing to keep the motor current under control I easily got up to 30 minutes using a set of 28 inch 5 blade rotors driven through a 4.5:1 gearing. The duration mounted quickly with both diameter and adding blades.

Going back to the Xcopter I picked out your AXI 2820/14 motors and 4S3P pack options and tried the same thing at 160 oz (10 lbs) AUW. With an APC 14x4.7SF prop the motor was over it's max current rating by quite a bit and total duration was a mere 9.3 minutes. Switching to the 28x6 5 blade rotors used in my previous example got me up to 21 minutes of duration.

So it would appear that your route to success is going to be custom made larger rotor like props with multi blades. The multi blades aids by allowing smaller rotors. The same results as with the 28 inch 5 blade rotor required a 36 inch diameter 2 blade rotor.

Of course this all means that your multicopter will need to grow in size. But it's not realistic to expect a 10 to 14 lb craft to be compact or to do well with the relatively small commercially made props that are available.
May 31, 2012, 01:28 PM
Registered User
Zen09's Avatar
One thing to consider, there are two types of propeller efficiency:
-prop's real efficiency (measured in % because it is = Pout/Pin). This must be as high as possible for (fast) airplanes, but has no meaning for a copter.
-prop's thrust efficiency (measured in g/W as a convention). This is the only thing that matters for copters because they only need thrust while the air flux velocity doesn't matter (that one multiplied with thrust = Pout).

The two efficiencies are not always proportional.

If you move a nearly infinite mass of air with an infinitely small velocity (with a given relation between the infinitenesses, but I won't go further) you get FINITE thrust with close to zero spent power, which means a very high thrust efficiency. That's why bigger diameter props tend to be more efficient statically.
May 31, 2012, 03:50 PM
Suspended Account
Hmmm... I have this motor in my calc, it follows quite well innov8tive's prop data, and so I guess your motor is really overloaded, more than what eCalc (xcopter) would compute (well, if you really have around 15 V on each motors).

For the prop, slow fly style props are good, you should be looking for high diameter and low pitch. I have doubt on the multiblade props... I think ecalc is optimistic here again.

Others have better endurance, but they also have a very lighter quad, so...

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Efficiency test : hexTronik DT700 and 24 propellers - GWS 9x5x3 added Mortimer Multirotor Drone Talk 110 Feb 15, 2014 12:30 PM
Cool Hot new Landing Gear for Multi Rotor Copters! Jason Cole Multirotor Drone Talk 0 Feb 28, 2012 04:14 PM
Help! propeller efficiency abcd123 Multirotor Drone Talk 9 Feb 22, 2012 06:04 PM
Discussion Where to buy GWS (or better!) propellers? Cheesehead Vendor Talk 3 Nov 09, 2011 07:56 AM
Question Has anyone flown commercially in the US with a multi copter in carry on? empire007 Multirotor Drone Talk 6 Jul 06, 2011 11:03 PM