Crash video: X8 with KDE 4014 into the water!

Discussion in 'Cinestar 8' started by Bryan Harvey, Jun 21, 2014.

  1. Gary Haynes

    Gary Haynes Administrator
    Staff Member

    Joined:
    Jul 6, 2012
    Messages:
    5,211
    Likes Received:
    460
    Tim one of the improvements made some time ago by Castle Creations was to implement auto erasure of the logs. From the current Phoenix Edge Lite manual. Might want to check if this is a configuration thing. The Castle's that I had some time back did not have this and each motor had to manually erased. It was a pain.

    DATA LOGGING
    Use Castle Link software to modify data logging settings. The capabilities are briefly described below. See the help files and on screen instructions in Castle Link and Castle Link Graph Viewer for complete details.
    *Duration of recording
    The ESC writes the recorded data to a limited amount of memory. Recording time varies with the data points selected and the sample rate. the ESC compresses data wherever possible; periods at IDLE or constant throttle do not take very much memory. Use the Castle Link software to select or remove values from the data logged by the ESC.
    Automatic Data Reset

    At power up, Automatic Data Reset checks the amount of memory used. If that value is more than the programmed limit, the controller will erase the entire data log. This ensures that the last few runs are saved to memory and available for analysis.
     
  2. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,383
    Likes Received:
    1,164
    I have to say that this is one of the daftest "features" I've seen. The whole point of capturing data is that it should be of the most recent events. So on a scale of zero to dumb, this is off the scale at the right hand end.

    Andy.
     
    Steve Maller likes this.
  3. Patrick Koegler

    Patrick Koegler New Member

    Joined:
    Feb 25, 2014
    Messages:
    11
    Likes Received:
    14
    Hey Guys,

    This is all good discussion, and I also received an e-mail from Bryan, so we'll work on getting the components back and analyzed.

    In terms of the shaft question - yes, absolutely, the shafts are hardened and then precision-ground to specification. They are incredibly strong and of course, designed to handle the loads put on them, and the same materials are used throughout our motor line, including the XF Brushless series for RC Helicopters that see peaks way over 240A and 12,000W outputs, even on 6mm shafts. In these cases, the shaft is under constant cyclic bending (as a pinion/gear mesh), so the loads are extremely high and far greater than in our multi-rotor applications. There's always a balance with hardening - too hard, and the material can become brittle. Too soft, and it can easily bend and will wear fast at the bearing/shaft interface, so we definitely have been through all this in the five years of designing and making motors to find the optimal design.

    For a shaft to break in this fashion, it would require some serious cyclic bending loads, most of which is soaked up by the electromagnetic forces themselves. In actuality, when the motor is energized, the magnetic forces are pulling on the stator in bi-directional forces, which naturally aligns the magnet bell to the stator and very minimal bending loads (if any) are on the shaft in this case. The shaft also broken below the location of the set-screw flats, which is surprising - where the material is a solid cross-section. There always could be a chance of a material void internal to the material - something we can't control and we're at the mercy of our suppliers for good materials (only our aerospace products are put through X-Ray inspection for this), but we do source from the best suppliers we can, so we do what we can to prevent any issues like this.

    On the other hand, if we look at this like material fatigue, what could cause something like this is a severe imbalance of a propeller - not just in weight, but in pitch-angle. It's seem tough that you would see this issue in-flight or hear the imbalance, but it is the lower-side motor in the X-8 and is already in turbulent airflow, so it may be more difficult to detect (?). If the imbalance was severe enough to shake the motor, it would impart cyclic bending loads on the shaft and also potentially be severe enough to be pushing the magnet bell into the motor housing? Again, the magnetic forces are amazingly strong and hard to believe this could happen, but anything is possible with these systems - that's why it's so important to fly safe and be aware of surroundings, as failures do happen that are out of our control. Even $2 billion military helicopters fall out of the sky on a regular basis from unknown mechanical or electrical failures (or so it seems in the news), so we always have to be aware of the chance of this occurring.
     
  4. James Cole

    James Cole Member

    Joined:
    Feb 26, 2013
    Messages:
    35
    Likes Received:
    1
    I think that most people here have a general goal, to fly safe.

    I really don't feel like anyone is bashing anyone here, a lot of us are analytical thinkers and love the challenge of troubleshooting a situation (even though I'm going bald and gray from it) and want to investigate all potentials for failure and examine all possibilities. Suggestions that say, the motor shafts may not be hardened are just that, suggestions, just like the suggestion that the esc had a failure or brown out, possibly from being inside those Gryphon arms with little airflow and alot of potential for vibrational damage.

    I believe everyone here is in the same boat, getting to the bottom of someones copter loss and helping them recover from that loss with a better understanding of what happened and how to avoid the problem in the future. These are the Lambos of the RC world, highly temperamental, very technical, and subject to failure. We can't thank KDE enough for taking time away from engineering to discuss these suggestions as to how this copter went down. From the sounds of it Patrick comes from a very technical engineering background and we all welcome his input into solving this mystery.
     
  5. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,383
    Likes Received:
    1,164
    Thanks for coming on to the forum and addressing the issues, Patrick.

    I think James summed it up nicely. We're a concerned, responsible bunch of professionals and the forum is a place for serious information interchange. That's always been Tabb's goal from the outset and remains both Gary, Brad's and my goal -- and that of our Chief UAV Grief Counsellor, Steve. :rolleyes:

    Andy
     
    Dan Walker and Steve Maller like this.
  6. Bryan Harvey

    Bryan Harvey Member

    Joined:
    Nov 3, 2012
    Messages:
    227
    Likes Received:
    18

    For the record, we did not mount the ESC's in the arm. They were mounted on board the mainframe - running quite cool I might add.
     
  7. Howard Dapp

    Howard Dapp Active Member

    Joined:
    Aug 18, 2012
    Messages:
    977
    Likes Received:
    111
    5mm - 6mm hardened steel shaft for the weights and loads we're inflicting on these motors. I remember making this argument here on FF in the past.
     
  8. Dave King

    Dave King Well-Known Member

    Joined:
    Dec 24, 2012
    Messages:
    2,711
    Likes Received:
    311
    Hello Patrick

    Thanks for the reply. My question to you is how do you determine what amount of load and stress it takes before the shaft breaks? This sounds like a theory versus practicality learning experience. These motors are new and to be honest it might be that nobody really knows how much stress the X8 configuration can take before it takes a failure like this. To me its like a race motor. Until you race it and put it through tests you may not know what the weak point is. To give you an example I used to race cars for 20 years. I raced 4.6 Ford Modular engines that came with cast iron pistons and rods. When we rebuilt to motor we upgraded the pistons and rods and used the best products available from the most reputable manufacturer in the industry. We used and installed the parts based upon what they recommended. Mind you this manufacturer has over 50 years of racing experience. We put the parts in like they mentioned and found out that the rods started failing at 500 rwhp. We went back to the manufacturer and they had to develop a beefier combination that they said would be good up to 800 rwhp. Well, we put the parts back in, and as soon as we started making 750 hp the rods failed. The manufacturer had to go back and make a set of billet aluminum parts before the engines could take what we were trying to do. On the other hand everyone (racers, and manufacturers) thought the stock crankshaft wouldn't live past 800 hp. It wasn't until people started pushing it and testing that it was found out that the crankshaft could actually live up to 2000 hp. Theory is great but it needs put to the test.

    So my point is that we have very well might have found the weak link in these motors based for this specific combination. Like Howard said they very well might need a larger shaft for the configuration. I wouldn't be quick to dismiss it because event the most experienced, knowledgeable people in the industry have to make an educated guess as a starting point and then see if it works in real world applications. I completely understand the cost and manufacturing concerns that go into this type of decision but for I for one as a consumer who risks everything when I fly will not put my copter in the air until I know 100% for sure that the shafts are strong enough. I also disagree that 2 billion dollar military helicopters fall out of the sky on a regular basis. Do you see commercial airlines fall out of the sky on a regular basis? Airlines wouldn't be putting people in planes if there wasn't a 99.999999% chance that they will have a safe flight. I think the odds of a plane crash are less than winning the lottery twice in a life time.

    Thanks for the discussion.

    Dave
     
  9. Gary Haynes

    Gary Haynes Administrator
    Staff Member

    Joined:
    Jul 6, 2012
    Messages:
    5,211
    Likes Received:
    460
    Dave I think you answered your own question. "Theory is great but it needs put to the test." Just like your racing days you are the tester. Just a guess but if one of our type of motor manufacturers had to do the testing that the FAA requires for an airline engine none of us would likely be flying anymore. And the FAA doesn't allow an airline to pick any motor off of the shelf and bolt it on as we do with our copters. We are operating so far into the experimental category that in reality nearly every flight is a test flight ala Wilbur and Orville. :)
     
    Steve Maller likes this.
  10. Dave King

    Dave King Well-Known Member

    Joined:
    Dec 24, 2012
    Messages:
    2,711
    Likes Received:
    311
    Agreed :)

    This is exactly why I built the X8 and have been rolling it and testing it the way I have. I use my flat 8 Tiger U5 for all my jobs, and when I find time I test the X8 at my local park. So far I have about 35 flights on my X8 but to me its not enough for me to say ok this is a tested product. I didn't put a camera on the X8 until I had about 20 flights on it. I probably wouldn't feel confident with anything less than 75 flights before I would consider it good and reliable enough to use for jobs. This spring I tested the flat 8 relentlessly and turned down jobs because I wanted a lot of testing under my belt since I had all new equipment (motors, power board, flight control board, new firmware etc) on the copter. If we the testers find the weak point we either have to find another part that will work or the manufacturer has to make the adjustments.

    Lastly I''m not saying the shaft failed, I'm just saying keep an open mind and don't automatically discount anything.
     
  11. Gary Haynes

    Gary Haynes Administrator
    Staff Member

    Joined:
    Jul 6, 2012
    Messages:
    5,211
    Likes Received:
    460
    And this from the newline on military crashes:

    More than 400 large U.S. military drones have crashed in major accidents around the world since 2001, a record of calamity that exposes the potential dangers of throwing open American skies to drone traffic, according to a yearlong Washington Post investigation. Since the outbreak of the wars in Afghanistan and Iraq, military drones have malfunctioned in myriad ways, plummeting from the sky because of mechanical breakdowns, human error, bad weather and other reasons, according to more than 50,000 pages of accident investigation reports and other records obtained by The Post under the Freedom of Information Act.
    Defense Department officials said they are confident in the reliability of their drones. Most of the crashes occurred in war, they emphasized, under harsh conditions unlikely to be replicated in the United States. Military statistics show the vast majority of flights go smoothly and that mishap rates have steadily declined over the past decade. Officials acknowledge, however, that drones will never be as safe as commercial jetliners. The Post’s analysis of accident records shows that the military and drone manufacturers have yet to overcome some fundamental safety hurdles: 1) A limited ability to detect and avoid trouble. Cameras and high-tech sensors on a drone cannot fully replace a pilot’s eyes and ears and nose in the cockpit. 2) Pilot error. Despite popular perception, flying a drone is trickier than playing a video game. 3) Persistent mechanical defects. Some drones were designed without backup safety features and rushed to war without extensive testing. 4) Unreliable communications links. Drones are dependent on wireless connections that can be fragile. Records show that links were disrupted or lost in more than a quarter of the worst crashes.
     
  12. Tim Gould

    Tim Gould New Member

    Joined:
    Nov 3, 2012
    Messages:
    19
    Likes Received:
    0
    Gary - thank you for the info on the CC speed controllers. I was able to find where you can set the data log to automatically reset. This will be used going forward. I wish I had known to make that change when I programmed the ESCs.

    Patrick - The motor is boxed and will be shipped to you today. I have included another close up of the shaft. You can clearly see the edges of the flat spots for securing the bell to the shaft.


    castle link data.jpg P1030081.jpg
     
  13. Gary McCready

    Gary McCready Active Member

    Joined:
    Oct 26, 2012
    Messages:
    284
    Likes Received:
    42
    A 15,000 lb Global Hawk crashed near my house two years ago, $131 million up in flames. It was about 10 miles down the road, and fortunately in a marsh. Of course, I was the blunt of many jokes about it, being one of the few around here flying anything bigger than a Phantom.
    Interesting article in the Washington Post about drone crashes.
    http://www.washingtonpost.com/sf/in...s-mount-as-military-flies-more-drones-in-u-s/
     
  14. Patrick Koegler

    Patrick Koegler New Member

    Joined:
    Feb 25, 2014
    Messages:
    11
    Likes Received:
    14
    Dave,

    Absolutely, sorry if I'm coming off as discounting anything, that wasn't my intention. The goal here is to find the true cause, and if a 5mm shaft will help in this regard, for sure we'll look at incorporating the change in production. This is the first case of a potential motor failure from the market (referring to a <0.05% failure rate), so I don't want to jump to conclusions before we have a chance to find out the true cause. Increasing the shaft diameter may or may not be the proper solution, depending on what we can find out.

    For sure, I value the opinions and comments in these threads, so I keep notes and discuss all of this with production. We are always looking for ways to improve designs between production runs, so if a 5mm shaft is a good solution, we'll for sure incorporate the design change and any other items we feel improves the overall product to the end customer.
     
  15. Tim Joy

    Tim Joy Active Member

    Joined:
    Aug 2, 2012
    Messages:
    405
    Likes Received:
    72
    The worst crashes are when it is a total mystery, and I think this one is pretty much solved, and I have been thinking about a solution:

    Assuming that if one motor/ esc can fail in a way that causes a huge amp draw depriving the others, bringing the whole ship down, what about putting a high amp fuse on each esc's power connection? Something that would ONLY blow if it was drawing a fairly ridiculous amount of juice, like 85-100 amps on a machine of this size?
    After seeing first hand what a short against a carbon frame from ONE motor wire can do to the whole system I'm thinking this might be another safety feature that is worth exploring.
     
  16. Steve Maller

    Steve Maller UAV Grief Counselor

    Joined:
    Oct 30, 2012
    Messages:
    3,981
    Likes Received:
    807
    MK BLs have such a feature, where they'll shut down or limit current (can't remember which one) above a certain amount. Or they'll just fail. :eek:
     
  17. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,383
    Likes Received:
    1,164

    The Wright brothers did more systematic testing that we do, I think. :)
    "OK, let's lob it off the sand dune and see what happens....."

    Andy.
     
  18. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,383
    Likes Received:
    1,164

    I have a 150 Amp fuse on my LiPo testing circuit. It's something like this: http://www.westmarine.com/blue-sea-systems--anl-ignition-protected-fuses--P009_273_004_004

    Actually, for my test setup it's more like this: http://www.westmarine.com/anl-fuses/blue-sea-systems--anl-fuse-block-for-35-750a-loads--10093821 -- I don't think I'd want to fly with eight of those mounting blocks! :)

    But the fuse itself might be viable -- they're not crazy heavy.

    The hard part would be testing whether the fuse really did the job -- and whether or not we see current spikes more than 150 Amps. If we do, they might blow the fuse even during normal operations....

    Andy.
     
  19. Adam Paugh

    Adam Paugh Distributor

    Joined:
    Jun 29, 2012
    Messages:
    323
    Likes Received:
    78
    HI Steve,

    Yes, the MK Bl-Ctrl are set up with self-limitation which I believe is a factor of 0.1 past the user defined set point for max. amp/temp. With such a low factor it would seem as though this was designed as a limitation for overburdened systems (over max payload/incorrect CG) rather than in cases of motor/ESC failure.

    Greetings,
    Adam
     
  20. Bryan Harvey

    Bryan Harvey Member

    Joined:
    Nov 3, 2012
    Messages:
    227
    Likes Received:
    18
    posted this on the wrong board by mistake earlier:

    This crash has got me thinking mostly about what I could have done better as a pilot. As a flight instructor, I used to routinely cut the engine at the most inopportune times on my students, so that they were always sharp on recovery and flew with a plan of action at the ready for every scenario, always in the back of their minds. We ALL need to be able to practice simulated motor-outs if we want to be as safe as possible in this game. If commercial ops are going to be sanctioned by the FAA I suspect it would be mandatory. Does anyone know of a multi-rotor flight simulator that can do this?
     
    Dan Walker and Chris Dillenbeck like this.

Share This Page