Designing/making a VTOL (aka V-280 Valor, Bell XV-15, or the V-22 Osprey)

[CampRobber]

 

[JasonK]

@CampRobber - looks like you got about 50g more thrust out of your motors/props then I did when I was testing (I got about 400g). Care to share your testing process?

 

[CampRobber]

Load cell & shunt / voltage sense wired to a computer. I'm not sure what you mean by "maximum"; this config is overpropped. 3s/2280kv/6x3 has a 72 mph free stream velocity, which equates to 2242 grams of thrust and 344 watts of kinetic power output. This whole curve took place at PWM=1000...1500; it would have gone higher, but exploded at some point. I suspect in your test you had voltage sag or something.

 

[JasonK]

interesting... I only got around 400g on 3s on the A-pack motor with a 6x3 prop. 3s with 6x4.5 prop gets really hot for me.

I don't have a load cell, I just built a rig to thrust test on a scale and pushed it to 100% on the throttle after calibration of the ESC.

here is a picture of roughly what I was doing:
https://forum.flitetest.com/index.php?threads/thrust-measurement-device.64317/

I find it very interesting that your getting as much thrust at 1500ms as I was getting at 'full'. I might have to revisit my test rig.

Then again, in the FT Dart video, they talked about using 2, 3, and 4s on the H-pack motors, but mine get to hot to touch in about 5 seconds on 3S at 'full' (with the supplied props), so I only run them at 2s.

 

[JasonK]

https://www.ecalc.ch/cgcalc.php?dee...0;4;2;0;0;0;0;2;0;0;0;0;3.5;0;0;0;0;12;25;10;

 

[JasonK]

I just build a thrust test rig similar to @CampRobber 's up there and still am maxing out at ~400g of thrust on my 1806 2280kv Power Pack A motor on 3S. I am still building the current & voltage sensing circuits so I can setup a fully automated test (parts are sitting on my bench, just need to wire it up and write some updated code to read all the sensors).

 

[JasonK]

the F pack motor only claims (Over 1Kg/motor on 4s and most prop choices) at 4 s and it is 2300kv (so basically the same KV but 33% more voltage), so I question the A pack motor getting 2kg of thrust on 3S.

 

[Pieliker96]

the F pack motor only claims (Over 1Kg/motor on 4s and most prop choices) at 4 s and it is 2300kv (so basically the same KV but 33% more voltage), so I question the A pack motor getting 2kg of thrust on 3S.

I do too. Thrust can be estimated from the mass flow rate and velocity of the air. Taking the given freestream velocity and x-sectional area of the prop disk (and air density), I calculated the theoretical maximum thrust of an 1806 2230kv w/ a 6x3 prop to be 720g (freestream v = 72 mph). Of course, it won't ever reach this due to rpm drop from loading of the motor, variations in the freestream velocity along the prop radius and error in the x-sectional area of the moving air in the prop disk's plane.

For what it's worth, I do have some data on the RS2205 2300kv. With a 6x4.5 Bull Nose prop on a fully charged 3S (12.6v) and installed on a wing which effectively reduces thrust (NOT a test stand, which is the ideal case), I got 750g of static thrust.

 

[CampRobber]

I calculated the theoretical maximum thrust of an 1806 2230kv w/ a 6x3 prop to be 720g (freestream v = 72 mph).

How are you calculating that?

 

[CampRobber]

(1.23 kg/m^3)*(1.00*(6"/2)^2)*((3.7*3*2230/minute*3"))^2/(9.8 m/s^2) = 720 grams

(1.23 kg/m^3)*(3.14*(6"/2)^2)*((3.7*3*2230/minute*3"))^2/(9.8 m/s^2) = 2.26 kg

 

[JasonK]

the spreadsheet here:
electricrcaircraftguy.blogspot.com/2013/09/propeller-static-dynamic-thrust-equation.html

give me 1053g for a 3s charged at 3.6 volts per cell, however my bench test only puts out ~ 400g, if I change the kv in the equation to 1500, it gives me about 450g for that same setup (which seems to be close to what I am getting.

my lipo is sitting at 11.32 unloaded right now which works out to around 3.7v per cell (I just quickly tested using a battery that had been set to storage, will do more detailed testing when I get the full bench setup working...)

also looking at seeing if I can add an optical RPM meter to the setup, so I can get the real RPM, not just the calculated. (would also like an IR temp sensor aimed at the bell, so I can watch for thermal issues, but this is lower on the priority list).

 

[Pieliker96]

(1.23 kg/m^3)*(1.00*(6"/2)^2)*((3.7*3*2230/minute*3"))^2/(9.8 m/s^2) = 720 grams

(1.23 kg/m^3)*(3.14*(6"/2)^2)*((3.7*3*2230/minute*3"))^2/(9.8 m/s^2) = 2.26 kg

NOTE: I assume in the following analysis that the propeller's pitch speed is equal to its efflux velocity. This is most accurate in the static case but does not apply once the velocity of the incoming airstream is no longer zero.

I used your figure of 72mph in post #43 as the efflux velocity. I was wrong, I realize now that I calculated the mass flow rate (0.719 kg/sec), not the thrust. Multiplying by the efflux velocity results in a thrust force of 23.1N (which, when divided by g = 9.81 m/s^2 yields more or less your 2.26kg of thrust (within rounding error, at least). However, this does not account for pressure drag due to the lower pressure of the faster moving air on the backside of the prop.

The dynamic pressure of a 72mph (32.2) efflux is 635 Pa. Assume the air velocity in front of the prop is 0 (this is true at infinite distance and in ideal circumstances). Since total pressure remains constant in isentropic flow, the static pressure behind the prop is 635 Pa lower than in the front. Multiplying this figure by the x-sectional area of the prop disk (0.0183 m^2) yields a force of 11.6N.
Subtracting this from the prior figure yields 11.5N (or, if you'd like, around 1.17kg) of thrust.

Of course, this is a simplification of things. It does not take into consideration the RPM drop under load that the motor experiences, which changes the efflux velocity, mass flow rate, and pressure difference. It is also general in that it represents the aggregate and average case - the efflux velocity will most certainly not be constant along the radius (in part due to aerodynamics and the lift distribution across the prop radius, but more obvious in that no air flows through the motor itself). A more accurate definition of thrust would be something along the lines of "The integral of mass flow rate times efflux velocity along the radius minus the integral of pressure times area along the radius".

I redid the calculations from some data on the 1806 datasheet, which lists a 6x4 prop pulling 460g at 15160 rpm (pitch speed = 25.7 m/s). This yielded a theoretical thrust from mass flow of 14.8N and a thrust loss due to pressure of 7.4N, resulting in a net thrust of 7.4N or ~750g.

Just for giggles we can calculate the efficiency of the 6x4 setup and apply that to your setup to get a rough idea of real-world thrust. The efficiency looks to be (460g / 750g =) 61%, which puts your theoretical thrust figure (WITHOUT accounting for the RPM drop under load!) at 0.71 kg. If we assume thrust scales linearly with RPM (which, it doesn't, I know, but with things being locally linear and all) and adjust from the unloaded 12.6v * 2280kv = 28730 rpm to something more reasonable like 17000 rpm (the motor spec sheet lists a 5x3 prop at 18510 rpm and a 6x4 at 15160, this seems reasonable), we get (17000 / 28730 * 0.71kg = ) 420g of thrust, which is much more in line with the existing data and what I would expect.

Regardless of the math, I can speak from experience and common sense when you claim to get over 2kg of thrust out of an 1806 on 3s and a 6" prop. And to get over 300 watts out of an 1806 requires putting more than that in. I have a little 3D plane with an 1806 and a 12A ESC that runs on 2S and pulls 300g. At most, I'm running 8.4v * 15A = 125W, and that's nearly enough to melt the firewall off the airframe with the help of some Florida summer sun. I can't imagine managing 300W of heat in such a small motor and the proppage it would take to load it to such a state.

It's a good exercise to extrapolate performance data from existing data and equations. With that said, it is imperative to cross-check the numbers you get with existing data and verify those numbers via more direct means. I believe you when you give a calculated efflux velocity, thrust figure, and power output. But unless those numbers can be verified through more direct means (you appear to have a load cell suited to that very purpose), we must keep in mind that most equations are useful generalizations.

 

[JasonK]

@Pieliker96, thanks for the info and confirmation that my data appears to be in the correct ballpark.

so, it does seem like this 2280kv motor only puts out ~1500kv of rotation at 11.1v with a 6x3 prop (which seems to match up on that spec sheet you have there). Is there any good way outside of testing to figure out how much 'under performance' one should expect from the ratings (the listed kv on the motor) and reality outside of bench testing?

Would it be reasonable to expect that the motor would be able to 'speed up' and get closer to the listed kv as the craft speeds up (still producing less effective thrust, but basically lossing thrust slower then the spreadsheet would show because the kv would be going up as the ambient airflow goes up).

 

[Pieliker96]

@Pieliker96, Is there any good way outside of testing to figure out how much 'under performance' one should expect from the ratings (the listed kv on the motor) and reality outside of bench testing?

Probably. One of the problems with doing so is it requires a well-characterized motor and prop. The motor, with torque curves and such for finding the loaded RPM and the prop for finding its aerodynamic properties (which can be used to find thrust and load). From there there are some calculations that can get you thrust, although there are some factors that either have to be measured or solved (ex. prop efficiency). I'm sure CFD is very useful for this stuff, but it's a bit out of my wheelhouse at the moment. For the average hobbyist, just testing it is the simple answer.

@Pieliker96, Would it be reasonable to expect that the motor would be able to 'speed up' and get closer to the listed kv as the craft speeds up (still producing less effective thrust, but basically lossing thrust slower then the spreadsheet would show because the kv would be going up as the ambient airflow goes up).

Yes. As the forward speed of the craft increases, the angle of attack of the prop blades decreases, which reduces load on the motor and thrust. This allows the motor to spin to a higher RPM. As the prop's pitch speed is approached, the load on the motor tends towards zero (but doesn't reach it due to prop drag). This allows it to come close to its unloaded rpm. If the airspeed is somewhat higher than the pitch speed (enough to counteract the small amount of drag), the motor will effectively be backdriven by the freestream, allowing it to attain higher rpms than when unloaded.

 

[CampRobber]

Regardless of the math, I can speak from experience and common sense when you claim to get over 2kg of thrust out of an 1806 on 3s and a 6" prop.

I think you misunderstood me. I wasn't claiming it could do that. My point was just that since the theoretical output is so high, it means that in the real world some other factor will be the constraint, like the prop flying apart or the esc melting or whatever. You're talking about this configuration melting on 2S ... he's running it on 3s and not melting it. Something is fishy with his test. My test only went up to pwm=50% or something; 450 grams is only as fast as I was willing to run those motors, not some specific limit.

 

[JasonK]

I think you misunderstood me. I wasn't claiming it could do that.

This looks like a claim of ~2kg of thrust to me:

3s/2280kv/6x3 has a 72 mph free stream velocity, which equates to 2242 grams of thrust and 344 watts of kinetic power output.

You're talking about this configuration melting on 2S ... he's running it on 3s and not melting it.

I ran these motors at 3s with a 6x4.5 prop at the airfield, when my son did a full throttle most of the fight with it, the motor was barely handle-able for at least 30-45 seconds after the flight and the 3D printed firewall was heat warped (this was in the southern summer, which would have exasperated any thermal issues).

Something is fishy with his test.

my test gave very close results as the spec sheet linked above and I have done it with multiple different force sensors now at this point and have been getting consistent results. Honestly, at this point the only test that has disagreed with the ~400g of thrust at 3S with 6x3 is yours.

I would like to believe that the thrust(s) that you were getting were correct as it would mean my project would be rather easy to do as I could easily go to 600g AUW, but based on my tests 400g AUW is as far as I think is reasonble for this to work.

 

[Pieliker96]

My point was just that since the theoretical output is so high, it means that in the real world some other factor will be the constraint, like the prop flying apart or the esc melting or whatever.

Your theoretical output is based on a prop with zero inefficiency and a motor with infinite torque. The hardware is not the limiting factor here, the laws of physics are.

You're talking about this configuration melting on 2S ... he's running it on 3s and not melting it. Something is fishy with his test.

That powertrain runs on 2s with a (quite overpropped) 7.25x4.7. It pulls more current than a 6x4 on 3s. What matters here is the power consumption of the motor (Volts * Current), not the voltage alone. I've found 125W (15A * 8.4v) to be around the limit. The motor's data sheet lists the maximum power for a 6x4 on 3s at 125W (11.3A * 11.1v). I can verify from past experience with a mini FT-22 that they run decently hot on 3s and a 6x3 - and I would expect a 6x4 to run even hotter.

 

[CampRobber]

I would like to believe that the thrust(s) that you were getting were correct as it would mean my project would be rather easy to do as I could easily go to 600g AUW, but based on my tests 400g AUW is as far as I think is reasonble for this to work.

Oh, I think 400g is a totally reasonable weight limit. I'm kinda still aiming for 250g.

 

[JasonK]

I have a first pass on all of the 'body' parts. this is 107 of foam and wood spar. I put everything I had handy on the scale and hit 335g without the motors, escs, flight controller, receiver, and servos. My goal of 400g AUW looks potentially in the realm of happening... I still have a bit of paper I could strip (4 strips of foam to mount the tail in the fuselage) and the main fuselage probably bigger then it needs to be. I might redesign it so the battery is just strapped on the bottom of it vs having it hold it inside, which should cut down some of the foam weight (not really that much) but also make putting the battery on/off easier by having it external.

 

[JasonK]

Got a paint job on, mounted the tail to the body, got the flaperon and V-tail servos installed and connected to their surfaces.
next up is to redesign my motor pods for the new motors and servos... fish the rest of the wires through the wing, mount and configure the flight controller, then start hover tests.