Tolerance to wind

[KDog]

Hi All - so I'm sitting in the airport on my way home from a holiday in Far North Queensland where I had hoped to do a bunch of filming, but the weather had other plans (also why I can't post any footage or flight logs yet).

Had a bit of an 'oh ****' moment when I put my MP up and after ascending for a few seconds it then got into the wind. It was getting chucked around pretty bad and well outside of the gimbal stabilisation tolerances (pic is a phone screen grab from the 2 seconds that I shot before I shat myself and stopped filming - in hindsight I should have kept filming), but it got much worse than this. BOM speed said 25/h winds in the area, but obviously the gust speed and the delta are the killer here. Essentially I had almost no manual control for about 10 seconds as the aircraft control systems tried to keep it in the air - a job they ended up doing well enough for long enough to get out of dodge and out of the wind.

Question is, how much can the mavic tolerate, not in terms of wind speed but in terms of de-stabilisaing gusts. At what point of pitch/roll will the aircraft stabilisation systems lose out and the aircraft become out of control? If out of control due to wind, how likely is it that the systems can re-stabilise?

Before anyone tells me I'm an idiot, published wind speed and ground gusts were within tolerance and I was flying it within regulations so there was no safety risk to people.

Keen to hear any real world experiences like this?

Cheers,

K.

 

[KDog]

 

[Domagoj]

I get a lot of high velocity wind warnings in DJI GO app but footage always look stabilized and Mavic seems to handle it preety good.

 

[Logger]

... It was getting chucked around pretty bad and well outside of the gimbal stabilisation tolerances (pic is a phone screen grab from the 2 seconds that I shot before I shat myself and stopped filming - in hindsight I should have kept filming), but it got much worse than this. BOM speed said 25/h winds in the area, but obviously the gust speed and the delta are the killer here. Essentially I had almost no manual control for about 10 seconds as the aircraft control systems tried to keep it in the air - a job they ended up doing well enough for long enough to get out of dodge and out of the wind.

Question is, how much can the mavic tolerate, not in terms of wind speed but in terms of de-stabilisaing gusts. At what point of pitch/roll will the aircraft stabilisation systems lose out and the aircraft become out of control? If out of control due to wind, how likely is it that the systems can re-stabilise?

I dont know exactly, but I suspect you will get blown away downwind way before you see the flight control system losing control. I do not recall anyone around here reporting their Mavic going out of control and crashing, due loss of flight control system in strong or gusty wind. Some of us have had these things up in pretty strong wind, far exceeding recommendations too.
If you post the log from DJIGO one of use can put it in HealthyDrones or Airdata and see what sort of wind strengths it thinks you were experiencing.

Your photo suggests wind from right to left as indicated by calm water in lee of the land. Aircraft will roll to the right to hold position. Tilted horizon is very common in this scenario especially when you start crabbing to windward which will further increase roll.
Where you in sport mode or P mode?

 

[KDog]

I get a lot of high velocity wind warnings in DJI GO app but footage always look stabilized and Mavic seems to handle it preety good.

I too have experienced high wind velocity warnings. As I said in my OP, the issue wasn't high winds so much as gusty conditions (i.e. Inconsistent wind velocity) which the aircraft was struggling to deal with. Think if a 26km/h wind that the flight systems were compensating for were to suddenly stop. This causes overcompensation which in this instance was violent.

To be clear, this isn't a 'high wind warning' discussion which has been done to death; I'm specifically interested in any experiences of temporarily (or permanently) losing stable control of the aircraft due to gusts.

Thanks.

K.

 

[KDog]

I dont know exactly, but I suspect you will get blown away downwind way before you see the flight control system losing control. I do not recall anyone around here reporting their Mavic going out of control and crashing, due loss of flight control system in strong wind. Some of us have had these things up in pretty strong wind, far exceeding recommendations too.
If you post the log from DJIGO one of use can put it in HealthyDrones or Airdata and see what sort of wind strengths it thinks you were experiencing.

Your photo suggests wind from right to left as indicated by calm water in lee of the land. Aircraft will roll to the right to hold position. Tilted horizon is very common in this scenario especially when you start crabbing to windward which will further increase roll.
Where you in sport mode or P mode?

Thanks - the photo was just a screen grab of about 15 seconds worth of pretty hairy flight - I've experienced tilt compensation before but this was all over the place (not holding steady but violently pitching and rolling to the point that I could see aircraft body in the frame). I was in p-mode, not sure what effect this would have on stabilisation rather just speed and aircraft pitch limits.

Will get the log up as soon I get a chance.

 

[Logger]

So if you fly in turbulent air in the lee of an obstacle it will get tossed around somewhat. I am not quite sure how you would know that you have temporally lost control though. It is the flight control system that is really flying the aircraft and not you. You are simply guiding it around the sky. If it really was losing control you could pull the logs and might see messages like this, that occurred as a Mavic tumbled from the sky after being shot down. I very much doubt you got anywhere near this though.

271.782 : 1403326306 : 15344 [L-RC]craft ctrl failed!!!
272.044 : 1404504102 : 15357 [L-FDI][CTRL]: fault off, horiz_ctrl_fail
272.144 : 1404954206 : 15362 [L-FDI][CTRL]: fault on , tilt_ctrl_fail
272.924 : 1408464476 : 15401 [L-FDI][CTRL]: fault off, tilt_ctrl_fail
273.802 : 1412416532 : 15445 [L-RC]craft ctrl failed!!!

I asked about flight mode simply because P-Mode allows us to derive the Average wind and Max Gusts from the log with Airdata like this.

If you were in S Mode this feature is not available.

 

[KDog]

So if you fly in turbulent air in the lee of an obstacle it will get tossed around somewhat. I am not quite sure how you would know that you have temporally lost control though. It is the flight control system that is really flying the aircraft and not you. You are simply guiding it around the sky. If it really was losing control you could pull the logs and might see messages like this, that occurred as a Mavic tumbled from the sky after being shot down. I very much doubt you got anywhere near this though.

271.782 : 1403326306 : 15344 [L-RC]craft ctrl failed!!!
272.044 : 1404504102 : 15357 [L-FDI][CTRL]: fault off, horiz_ctrl_fail
272.144 : 1404954206 : 15362 [L-FDI][CTRL]: fault on , tilt_ctrl_fail
272.924 : 1408464476 : 15401 [L-FDI][CTRL]: fault off, tilt_ctrl_fail
273.802 : 1412416532 : 15445 [L-RC]craft ctrl failed!!!

I asked about flight mode simply because P-Mode allows us to derive the Average wind and Max Gusts from the log with Airdata like this.
View attachment 14380
If you were in S Mode this feature is not available.

I should clarify - I meant loss of manual control - meaning the flight control systems had the conch and my inputs were not really having much of an effect. I'm actually very impressed that it maintained flight in those conditions.

As you've said, the flight logs should tell a good story - if I'd been filming the whole time it would tell a better story yet. Ultimately I'm 100% sure that it was the wind - when I bought it down it started behaving again and subsequent flights have been fine. What I'm really interested in from an analysis perspective is how close to death I was. Not sure if the flight logs will enable this analysis?

 

[hychewright]

I get a lot of high velocity wind warnings in DJI GO app but footage always look stabilized and Mavic seems to handle it preety good.

Same here.

 

[FatherXmas]

I do not recall anyone around here reporting their Mavic going out of control and crashing, due loss of flight control system in strong or gusty wind.

Had that happen yesterday - it wasn't very windy at all before I took off. I took off, went straight up to 35 feet so the precision RTH could get its bearings. The MP hovered in place a few seconds then a huge wind gust hit it, drove it sideways into a tree which flipped it upside down and it came crashing into the parking lot, hitting a concrete curb. So, today it's on its way to Thunderdrone for some major surgery (gimbal, broken arm, cracked top cover).

 

[SteveCA]

After months of flying my Mavic in high to very high winds, I finally found the Mavics limit. I have found that it flies fine so long as the wind gusts are 35 MPH and under. The key is landing where I sometimes have to wait until the strongest gusts subside for long enough to land.

A few weeks ago I flew in what was supposed to be 25MPH gusting to 35MPH. As soon as the Mavic got airborne, it took off until I was able to switch to sports mode. It took about 20 seconds as I was flying with gloves using my Ipad mini 2 attached to the controller. You know the drill, you need finger nails to make the switch. The Mavic in P mode could not hold its position and promptly flew down wind. I had to elevate to 150 feet to avoid some tall trees before switching to sports mode. Of course, the winds were even higher at 150 feet. By the time I got it into sports mode, the Mavic was 600-700 feet away. I brought it back at 3-5 MPH in sports mode but could not land for around 5 minutes while the strong gusts buffeted the mavic like a toy. After a few minutes the strongest gusts subsided and I promptly put her down very quickly. Not a soft landing (I landed in sports mode with VPS turned off) but everything was okay. Upon regaining my composure, I checked my wind app (sailflow) and found the the winds were actually 35 with gusts to 46 MPH.

That is too much for the Mavic. Up until that flight, I had routinely flown in 25-35 MPH winds and the Mavic handled those very well. The Mavic can fly pretty well in high winds but landing can be another matter so if you fly in very high winds, find a sheltered place to land and you'll be okay. I did not have a sheltered place to land that day as I was in an open parking lot at the beach. Lesson learned.

 

[KDog]

Had that happen yesterday - it wasn't very windy at all before I took off. I took off, went straight up to 35 feet so the precision RTH could get its bearings. The MP hovered in place a few seconds then a huge wind gust hit it, drove it sideways into a tree which flipped it upside down and it came crashing into the parking lot, hitting a concrete curb. So, today it's on its way to Thunderdrone for some major surgery (gimbal, broken arm, cracked top cover).

Tough break. Did you get any 'warning signs' of instability prior to the gust, or was just over and done with?

 

[TTFullTimer]

The control loop built into the Mavic has adjustments which many of us use to smooth out video etc. As we slow responsiveness in those hysteresis curves the Mavic will be much less able to control sudden perturbations like wind gusts or swirls. The control loop will have various characteristics like deadband, hysteresis, acceleration of the motors, ability to maintain performance when the gyros in the IMU are outside the working bands. And then you throw in the ability of the motors to change speed in some limited amount of time, the ability of the propellers to provide thrust when the air around them is turbulent.

We throw our Mavics into the air and the dadgum things fly so well we forget it is just an autopilot and an air vehicle with lots of limitations. Ever been in an airliner that hit an "air pocket" really just a big downdraft. I saw a drink cart hit the ceiling. Watch a dust devil and think about how you would provide power on four corners to resist that.

I know from piloting my various boats over the last 60 years currents and wind can make keeping a nice straight headway difficult at best. And the autopilot will wander too on a boat.

So I don't think there is one answer to how much wind. The best you can say is "It depends"

Personally I use my flag to determine it I will fly. If it is just wagging a little bit good flying. If it stands straight out, I stay grounded.

 

[KDog]

The control loop built into the Mavic has adjustments which many of us use to smooth out video etc. As we slow responsiveness in those hysteresis curves the Mavic will be much less able to control sudden perturbations like wind gusts or swirls. The control loop will have various characteristics like deadband, hysteresis, acceleration of the motors, ability to maintain performance when the gyros in the IMU are outside the working bands. And then you throw in the ability of the motors to change speed in some limited amount of time, the ability of the propellers to provide thrust when the air around them is turbulent.

We throw our Mavics into the air and the dadgum things fly so well we forget it is just an autopilot and an air vehicle with lots of limitations. Ever been in an airliner that hit an "air pocket" really just a big downdraft. I saw a drink cart hit the ceiling. Watch a dust devil and think about how you would provide power on four corners to resist that.

I know from piloting my various boats over the last 60 years currents and wind can make keeping a nice straight headway difficult at best. And the autopilot will wander too on a boat.

So I don't think there is one answer to how much wind. The best you can say is "It depends"

Personally I use my flag to determine it I will fly. If it is just wagging a little bit good flying. If it stands straight out, I stay grounded.

My understanding of the control curves we edit to smooth responsiveness is only dampening input responsiveness - not changing the actual autostabilisation characteristics of the IMU or control computers. The motors should remain minutely responsive to external factors regardless of how sluggish these input settings are.

That's largely irrelevant of course when the pitch and roll of the aircraft is outside of the tolerance of the gimbal. I take all the points re: turbulence and its effect on what is a very small aircraft, and also the points on when and when not to fly (at ground level the wind was pretty placid as it were) and the issue wasn't wind speed but wind variance.

My question remains not so much about my ability to control it, but for the flight control systems to handle pitch and roll outside of gimbal tolerance and recover after the same.

Maybe a clearer question might be "how many degrees can a MP pitch and roll before the flight control systems lose control and it's in a death spiral?"

 

[TTFullTimer]

My understanding of the control curves we edit to smooth responsiveness is only dampening input responsiveness - not changing the actual autostabilisation characteristics of the IMU or control computers. The motors should remain minutely responsive to external factors regardless of how sluggish these input settings are.

That's largely irrelevant of course when the pitch and roll of the aircraft is outside of the tolerance of the gimbal. I take all the points re: turbulence and its effect on what is a very small aircraft, and also the points on when and when not to fly (at ground level the wind was pretty placid as it were) and the issue wasn't wind speed but wind variance.

My question remains not so much about my ability to control it, but for the flight control systems to handle pitch and roll outside of gimbal tolerance and recover after the same.

Maybe a clearer question might be "how many degrees can a MP pitch and roll before the flight control systems lose control and it's in a death spiral?"

I have noticed that the actual control response changes. When i have it "turned up" forward acceleration is abrupt and stopping the forward motion is also abrupt. I was noticing close in today that when I commanded a fast forward move the time to get to speed was quite slow. And when I let up the time to stop was fairly long. In both cases the angle of flight was much more level than when it is going balls to the wall. So it is not just input it is control response. I have everything slowed down to make videos smoother.

Easy test on the pitch and roll if you had some soft surface for the Mavic to crash into. Just get two good sized fans. Mount one blowing up on one side and the other blowing down on the opposite side. Then hover the Mavic and start the fans. Observe when it goes bonkers.

 

[flyNfrank]

The way the propulsion system has been programmed, 98% of the focus has been placed on the aircraft's video and picture taking ergonomics. I have spent hour upon hour looking over internal flight logs from the Mavic. There is absolutely nothing that favors the weather conditions, and especially the wind. The amount of current usable volts with in the battery, is what dictates how the mavic responds to commands.

 

[gumaco]

I recently flew it in strong winds and like SteveCA says, it needs Sport Mode to be controlled in such situations or it flews away with the current. I sincerely do not understand why you make such a big deal about it. Simply, when you fear the wind is too strong, just take off with sport mode already activated.
Also, if I may, I suggest you to land with hand catch in such situations.

 

[KDog]

I recently flew it in strong winds and like SteveCA says, it needs Sport Mode to be controlled in such situations or it flews away with the current. I sincerely do not understand why you make such a big deal about it. Simply, when you fear the wind is too strong, just take off with sport mode already activated.
Also, if I may, I suggest you to land with hand catch in such situations.

Please take some time to actually read and understand the OP so that you can reply with something relevant in the future.

 

[gumaco]

Please take some time to actually read and understand the OP so that you can reply with something relevant in the future.

I apologize, I just wanted to help the community with my experience, anyway you're right 100%.

 

[SkunkWerx's]

When in P-Mode top speed is 22mph any wind or Gust above that will cause the Mavic to drift on the wind and seem as you have no or little control. Best move to make is put it in S-Mode .
When in S-Mode top speed is 38+mph any wind or gusts above that ---- You know what will happen , Bye Bye Mavic.
FlySafe -FlyFree