Not getting as much thrust as I would expect when flying 37 kg AUW

ArduCopter
1

Hi,

These days I have been testing a heavy lifter multicopter, 37 kg AUW. I would like to share with you my concerns regarding what we saw after running several flight tests on the field, to see if you can help me with some tips that allow me to understand what is going on with the thrust, and how I can fix it. This is my setup:

  • FRAME: Gryphon Dynamics 1400SP
  • MOTORS: T-MOTOR U11II KV120
  • ESC: T-MOTOR ALPHA 80
  • PROPS: T-MOTOR 26” x 8.5 folding
  • BATTERY: TATTU 12S 32.000 mAh

It is important to highlight two aspects of the flight tests that are crucial for the next discussion:

  • Octocopter on a coaxial configuration (thrust penalization due to interference)
  • Flight tests were done at 2800 m above sea level (thrust penalization due to lower air density)

Another thing to notice is that I had completed two successful AUTOTUNE procedure in serie:

  • First, with 22 kg AUW (no payload)
  • Second, with 32 kg AUW (10 kg payload)

My issue, in summary, is: I am not getting as much thrust as I would expect, according to the manufacturer motor data, when flying 37 kg AUW (15 kg payload).

WHAT I WOULD EXPECT

This is the manufacturer motor data for the motor-prop combination it is used in my setup, i.e, U11 + 26x8.5:

image

This is my reasoning (please, If any of you notice any mistake in this reasoning, please let me know):

  • My setup should give, at sea level and in absence of interference (ideal conditions), 99.360 kg of total thrust.
  • What is the % of thrust losses due to the combination of the two effects: coaxial configuration + high altitude? This is the BIG question here. Any of you know any kind of predictive method to estimate this figure before hand? I would assume a total losses due to both effects combined in 30%. Is this a reasonable assumption? In this case, with 30% losses, I would have a total thrust of 69.552 kg.
  • Finally, let say that I would like to devote 70% of my total thrust for altitude control and the remaining 30% for attitude control. That means that my setup should be able to lift, safely, in the aforementioned scenario, up to 48.686 kg

WHAT I AM GETTING

First time I flew my setup with 37 kg AUW, it was obvious right away that the dron was getting trouble maintaining the altitude in both, ALTHOLD and LOITER flight modes. As soon as I lower the throttle stick a little, the dron tended to land, no matter if I applied 100% throttle (hopefully it was very close to the floor, so no damage to regret).

After reading about this issue, I concluded that the answer may be in the parameter which controls the throttle vs attitude priorisation, ATC_THR_MIX_MAN, which was set at his default value 0.5. I read here that somebody tried to lower this value (even going lower that the minimum value recommended, which was in fact 0.5), and it worked, so I set it at 0.35 and test it again.

This time it was better, we managed to lift very easily, at least apparently. We switch to ALTHOLD and LOITER several times. It is true that we noticed the drone this time a little bit shaky, in LOITER, but It may be explained due to the reduced attitude control priorisation. However, after 5 mins, I started to experiencing the same synthoms as in my previous flight: difficulties to maintain altitude and tendency to land. We decided to abort the mission at this point.

And one last thing, after both flights, the motors were cold, which I guess is because they were not working as much as one would assume, after noticing the problem with the altitude control.

Here I am sharing the logs from this last flight and some pics with comments:

CURRENT
150A average

image
image.png1033×268 40.2 KB

VOLTAGE
Failsafe was set at 42 V (3.5 V per cells). As you can see, we stopped before reaching that point

image
image.png1048×262 43.7 KB

RC OUTPUT
As you can see, the eight motors are performing quite similar and the signal don’t look saturated

image
image.png1048×259 98.3 KB

THROTTLE OUTPUT
Most of the flight the throttle output was almost 0.5.

image
image.png1048×261 40.4 KB

QUESTION

As you can see from the last picture, most the flight the throttle output was at 0.5, meaning that there was plenty of room for maneuver. Why then did it start to fail to maintain altitude properly after 5 mins?

Why is that happening? Any clue? Any recommendation?

Thanks in advance,

JM

2

That configuration is grossly overloaded at 37kg. The log indicates 80% of motor output at hover which is exactly what eCalc (a decent predictive method) predicts. 50% is the desired design goal.

3

I concord to that. Here you can visualize your issues (qiuck eCalc setup). First is your 26x8.5in setup and the second is just a test of what can be done to get closer to the 2x thrust requirement with your existing hardware (providing there is sufficient space om the frame for larger props). This might not be feasible at all but it shows that you are quite far off with your original setup on 37kg AUW.

eCalc - 26x8.5in.pdf (435.1 KB)

eCalc - 31.2x10.2in.pdf (466.5 KB)

4

Hi Dave,

Thanks for your answer. Yes, I understand that this setup is overloaded, and that the design goal is being able to hover at 50% throtle. What impressed me was that only high altitude and coaxial configuration is accountable for way more than 50% of effciency loss. Because as I said in my post, after a few minutes, 100% throtle was not enough to hold altitude.

You mentioned that the log indicates 80% of motor output. Where exacty are you looking in the log to reach that conclusion? I am not an expert in log examination. What I saw is the ThrO variable under the CTUN message, which indicates 50% of throtle output. Is that correct? I am missing something? Or maybe my Radio Control is not calibrated correctly?

Thanks in advance,

JM

5

Yes, the biggest props T-Motor offers would be your best choice. The 28x9.3 3-blade would also work at a bit less AUW. Or fly with zero payload with the current setup.

BTW-To answer your question about co-axial vs, a flat octo setup the thrust/weight ratio is 1.2:1 vs 1.3:1 respectively. And the difference between 2800m elevation and sea level is 1.2:1 vs 1.5:1 respectively.

6

Hi ronsand,

Thanks for your advise, i will keep it in mind for the next iteration. As I have just said to Dave, I am still a little bit impressed due to the fact that only high altitude and coaxial configuration is responsible of more than 50% of thrust loss. That made me thing that maybe there is some subtle pixhawk configuration in the pixhawk that is limiting the thrust available…

7

You can see it in the graph you posted. At hover all motors are ~1800. That’s 80% of the typical range 1000-2000. Ideally this would be 1500.

8

This is also backed up by the eCalc estimate of 79% throttle input at hover.
eCalc might not be perfect but after using it for many years (actually since the beginning), I can tell you that it will get you pretty close, or at least a good starting point :slight_smile:

9

Hi Dave,

I understand this. If motor range is more or less between 1000-2000, an average of 1800 means 80%. But why then the throtle output graph, under CTUN, is showing an average of 50%? Why is that difference?

image
image.png1033×268 40.2 KB