It's official. Guinness World Records validated the flight (see link under Flight 1).
Two flights were achieved to demonstrate that multi-rotor aircraft can stay in the air carrying a payload for a long period of time. While 20 minute flights were considered long in 2014, these demonstration flights show that 1 hour flights can be the norm. Imagine the increase in commercial applications when it is common for ships to stay in the air for an hour or more carrying different payloads. It's just a matter of finding good rotors (Tiger), flight control electronics (3DR), batteries (Panasonic Li-Ion), ESCs, and removing unnecessary weight. Once industry realizes the importance of performance, we will also see a jump in rotor and ESC net-lift efficiency (lift after it gets itself off the ground first).
Flight 1 – Guinness World Record for Longest Electric RC Multicopter Flight (Duration). Pending Guinness review.
- Ship Name, Rufous
- A hover that stays within view of the stationary video camera
- 1 ½ hrs (97 minutes 23 seconds) on 8/21/2014 @ about 4 meters above open level ground (well above ground effect and without use of thermals or updrafts)
- Hover, while not impressive nor useful, does require more energy than typical flight speeds and as such, a good first test.
- Pixhawk Flight Log (Time, GPS, Altitude, Speed, Volts, Amps): Guinness Flight
- Video of the Guinness record
- Guinness Record Book
- Time Lapse (GoPro) of the Guinness record.
- Ship Secrets
Earth view of the hover breaking the world record.
The officials: Michael Allen (expert in UAV flight control; Cloud Cap Technology), Brett Faike (UAV extraordinaire; teaches UAV tech at local schools; gave FPV demo after flight), Forrest (pilot in training), and Kirby Neumann-Rae (editor of the Hood River News).
Celebrating surpassing the old record of 80 minutes (ahhh ... hey pilot ... a little less champagne and eyes on the copter ... it's not time to prune the orchard yet!). Bud, not shown, was taking photos. Hope you got some champagne Bud!
Co-pilot, Dr Carie Frantz ... yes this amazing looking lass is single, is an unreal outdoor bad-ass, and has a great job in case you were wondering. Thanks for keeping pops sane Carie.
Posting the final time (for the final application to Guinness, I requested that 4 seconds be taken off because some blades of grass were brushed prior to the ship finally giving up and setting down. An interesting note on the battery:
o The Li-Ion batteries used are rechargeable (flight was on about the 8th recharge).
o Running Li-Ion batteries down to where the ship drops out of the sky does not hurt them if the ship mass and battery voltage is engineered to invoke that event before voltage runs down too low.
o A few days later, Flight 2 was made using the same battery pack with the ship carrying a camera (see Flight 2 next).
o Li-Ion are low C discharge (not 30 or 20 or 10 ... think about 1 C). The chemical barrier will break down if amp draw is too high. This causes the ship to loose altitude. And if there isn't enough altitude for the battery to recover, crash. As an example, using a 4S5P battery pack (about 15000mAh), Rufous hovers and flies at moderate speed at about 8 to 10 amps depending on payload and flight demand and does great even in moderate winds. Rufous has even flown 60 mph (97 kph) on the same Li-Ion battery pack. However, get up to 15+ amps turning a sharp high-speed corner and the battery chemical barrier locks up in about 2 seconds (yes I've had to rebuild a few times before I figured this out). So caution. Do the calculations. Try to design a safe ship that stays in the air at 2x the amp usage at hover.
Flight 2 – Performance demonstration flight (actual flight versus stationary while carrying a payloads)
- Time aloft goal of 1 ½ hr; so far 82.9 minutes
- Distance traveled goal of 10 km; so far 14.4km (9 miles)
- Elevation gain (cumulative) goal 300 meters; so far 378 m (1242 ft)
- Speed (max) goal 18 m/s (40 mph); so far 9 m/s (20 mph)
- Speed (typical) so far 2.3 m/s (5 mph)
- Payloads GoPro Hero 3+ and 3DR Telemetry Module
- Flight Log (Time, GPS, Altitude, Speed, Volts, Amps) Performance Flight
- Video coming soon
[note: will try to beat 90 minutes carrying a camera when I get back from Sri Lanka]
Earth View showing the flight path of the performance demonstration over my orchard.
Photo taken by Brett Faike, a local multi-copter flying legend, of the flight area (man-cave on left and princess palace on right). He treated the witnesses of the Guinness record to FPV flights afterwards using two sets of goggles so we could "ride" along (keeping the ship within visual range and over my farm, of course).
Flight #2 (with camera and telemetry payloads), Elevation Profile.
Flight #2 (with camera and telemetry payloads), Watt Profile.
Flight #2 (with camera and telemetry payloads), Speed Profile. The wind speeds were about 2 - 4 m/s from the north..
Flight Team
- Design/Engineering Forrest Frantz
- EE/Flight Engineering Jim Frantz
- Test Marty Frantz
- Pilot Forrest Frantz
- Copilot Carie Frantz
Acting Guinness Judges/Timers
- KirbyKeumann-Rea Hood River News Editor
- Michael Allen Cloud Cap Technologies Flight Control Software
- Brett Faike Multi-Rotor Extraordinaire Software Developer
Ship Summary: 1.65 kg (w/ camera & telemetry)
- Frame Quad Carbon tube w/ open front for photography 62 g
- Flight Controller 3DR Pixhawk + 3DR Power Module 34 g
- Motors T-Motor MN3508-29 380KV 328 g
- Propellers T-Motor Carbon 16 x 5.4 116 g
- ESC Afro 20A 31 g
- Batteries Rechargeable Panasonic NCR18650B 4S5P 930 g
- Camera GoPRo Hero 3+ 76 g
- Telemetry 3DR Radio V2 19 g
- Wires & Misc Long wires runs were magnetic wire 54 g
Design Elements of Note:
- The electronics platform (EP) was eliminated to save weight (previous EP added 8 grams). The battery doubled as the EP. The flight controller (FC) was bonded directly to the battery, which was bonded directly to the frame. The battery has a strong electro-magnetic field around it. 3DR deals with this issue by putting the sensitive elements, the GPS antenna and Magnometer, separate so it can be placed away from the battery. In this case the GPS/Compass was bonded to the starboard-fore motor arm away from the battery. This is a testimony to the excellent design of the FC as it did its job that close to a large battery. Even though this worked, it is not recommended nor fully tested.
- The ESC have their heat sinks removed. To ensure cooling, the ESCs were placed directly under the tips of the propellers to get positive prop wash. The loss in lift is much lower than the loss in weight. To help protect from shorts, spray the ESC with electrical silicone.
- Wiring in right sized. The normal EE rule of 3% tolerated loss is put aside. The rule is replaced by physics of electrical transmission. A larger wire area cross-section produces less heat that is measurable in watts. And a larger wire area cross-section weighs more thus taking more watts to lift it. Thus watt usage can be calculated for all wire sizes and thus optimized. Some wire insulation is weak or can break down in sunlight. Magnetic wire was used where possible. But multi-wire strands are recommended to prevent wire breakage due to stress aging.
- Li-Ion batteries are about 50% more efficient (watts hours/gram) than LiPo for multi-rotor flight.
- Metal screws/nuts/washers replaced with nylon. The only exception are the propeller screws which are aluminum.
- Metal screws/nuts/washers and clamps replaced by bonded parts.
- Gussets, screws/nuts/washers, and plates replaced by continuous masts and bonding.
- Metal spacers replaced by short nylon spacers.
- The most efficient propellers and motors were used – Tiger T-Motor.
- Props and motors statically and then dynamically balanced.
Ship Performance:
- Vibration Average Score 0.05 gs.
- Stability Score: 0.14 degrees.
- Photography see test photos
Photo taken from tripod (baseline) with GoPro Hero 3+ at med 7m.
Photo taken from Rufous. The camera was hard mounted, so now need to work on camera isolation that works.
Test Flight Highlights: Prior to “Flight 1” the ship crashed three times.
Crash 1 – Tried saving 2 grams by using Single Ply Carbon Skin Nomex core sandwich panels for motor mounts. The Nomex core on one motor sheared causing a cascading event where all of the motors sheared. Picture the ship moving away in an uncontrolled fashion with one motor suddenly floating away and quickly followed by the other three motors as the frame and battery kept flying in a deep descent.
Crash 2 – Really stupid piloting. I cut power after a landing but forgot to disengage the motors (throttle lower left). As I bent over the ship, my belly pressed the throttle to full and I got a face full. Luckily wearing protective goggles and thick clothing. Sometimes it takes a hard lesson to learn – always disengage the motors and the first thing one does when approaching the ship is to push the red button on the Pixhawk.
Crash 3 – Had been suspicious of one motor mount bond (didn’t sand all surfaces on all the mounts). The suspicions came true as one motor left the motor mast and the ship impaled itself into the lawn (photo below) after some high speed runs that loosened the faulty bond. This crash broke one of the motor masts. The motors are held on with nylon screws that are sized to shear off without harming the frame. This works most of the time, but not this time.
Anyone interested in beating this record (the 100 minutes mark is itching to be broken), friend me and I'll gladly pass along what you need to do to satisfy Guinness. I only ask that we are gentlemen and set records using engineering and piloting skill minimizing the use of ground lift (e.g., please avoid a hover off of a hot tar roof above a heated wall or on a side of a hill/ridge or in ground effect). We want to show industry what is possible under normal flight conditions. I also ask that you fully disclose your ship so we can all learn from what you achieved. Thanks.
If you have any engineering questions, I'll will answer them below.
Comments
On motor tests ... do I have test data to share?
I have many Excel files with motor test data. The problem with sharing the data is context of the test. The following are examples of how the purpose and context of the test varies from day to day.
- comparisons between motors (hold the ESC and props constant) but the next day of comparisons might require a different ESC or an ESC gets operated on to determine its true weight.
- comparisons between props (hold the motor and ESC constant) but in the following week, the motor might change as more efficient motors rise to the top.
- picking the best ESC ...
- picking the best 4 props out of a set of 24 supposively identical props ...
- picking the best distance between props ...
I've also refined the test procedure over time, which adds to the confusion. I'm still learning how to do this.
So no, I really don't have files to share. Without the context and reading of the notes, the data is good for me but might be misleading to others.
I've never seen such an informed response by a manufacturer before. Impressive support.
On the topic of net-lift efficiency. Most motor manufacturers state an efficiency (grams thrust per watt) of between 10 and 20. So lets say the test results were:
- 500 grams thrust
- 10 watts
This would be a motor with an attractive efficiency number of 50! But this motor could be horrible. Let's day the motor, prop, prop adapter, leads, and structure required to support it weighed 500 grams. It's Net-Trust per watt drops to zero (500 grams thrust less 500 grams mass all divided by 10 watts = 0).
So let's apply this to the 475KV motor in the discussion above. Their spec for 50% power for 13" to 15" props were: 16.6, 16.3, and 16.2. But when you account for the mass of the motors and props, this drops to 11.0, 11.7, and 12.7. These numbers compare favorably to the motors used on Rufous, 13.0.
But remember to not trust data from the mfgr as it applies to your ship. The test results are highly dependent on test method and the actual props used, which is why you can sort of use mfgr numbers to pick a motor from a mfg but not to compare between mfgrs.
I was considering these KDE motors. Here an answer from the KDE designer to a question I asked about updating my Tiger motors 3515-400KV with something with more efficiency. One drawback of KDE motors is the obligation to use their own ESCs (other ESCs will not drive their motors coroectly apparently):
"
Hello Hugues,
Thank you for contacting us here at KDE Direct and consider our products for you needs. We appreciate your business.
Let’s break down the math to see what’s need for you requirements. You have a vehicle that weighs 7Kg in a octo arrangement. That 7000/8 to get the thrust needed to carry the weight at 50% throttle. 7000/8 = 875 grams of thrust needed at 50% throttle for the requirements.
We have 2 motors that would fall into this category running a 6S battery on 15” props. Ether our 3520-400 or 3510-475 motors, here are both the motors and they’re performance charts.
http://kdedirect.com/KDE3520XF400.html
http://kdedirect.com/files/KDE_Direct_3520XF-400_M-R_Performance_V2.pdf
http://kdedirect.com/KDE3510XF475.html
http://kdedirect.com/files/KDE_Direct_3510XF-475_M-R_Performance_V2.pdf
The weight of the 3520-400 with wire leads is 245g, but remember your going to end up shortening the wires for your needs. The weight for the 3510-475 motor with wire leads is 175g, again you’ll end up shortening the leads, they’re super long.
The 3520-400 on a 6S battery and 15” props produces 1175 grams of thrust at 50% throttle. Keep in mind this at sea level, higher altitudes will produce different results. This is also coupled to our esc’s, different esc’s will also have different results.
The 3510-475 motor on a 6S battery with 15” props produces 1245 grams of thrust at 50% throttle. Again, same information above will vary the results.
Let’s talk about the pros and cons of one motor over the other. The 3520-400 motor might weigh a little more, but could be used on much larger system when going up in prop sizes. It’s really built to give performance on larger systems. It will be more efficient spinning a larger prop slower over the 3510-475 motor. The 3510-475 motor is a lighter motor and has more thrust over the 3520, but will not scale up to be able to spin a larger prop over 15”. You have to have the lager stator on the 3520 motor to be able to handle the heat and work load for a larger prop. So, it’s really up to you to decide which motor will fit your needs, ether one will work fine. If you never going to go over 15” props, stick with the 3510-475, if you ever decide to go over 15’ props, go with the 3520-400.
I hope this information helps you out, let me know if you need anything else, and I look forward to hearing form you.
Thanks,
Chris McVey
Lead Technical Support, KDE Direct
www.KDEDirect.com
"
Hi Forrest,
Nice work. I have a couple questions:
1. What do you mean by "net-lift"? I assume you are talking about Power consumption vs. Thrust/Lift.
2. Do you have any power vs thrust measurements for prop/motor combos from your project that you could share?
Thanks!
Gary - Is the link in your message the most efficient KDE in your opinion? And if so, what prop do you think is the most efficient with it?
What I might do is to test your recommendation plus one of their lower KV motors, which should be the most efficient.
1) Do not trust any spec from the mfg. I have yet to find one that is accurate. At best use them as a guide to picking between motors of the same mfgr. Most all mfg specs, for example, include ground effect. And none are based on net-lift.
2) Work on weight first. If the ship only flies for 10 minutes, then increasing propulsion efficiency by 20% only adds 2 minutes of flight. But if you shave weight aggressively, then small efficiency increases become huge. A 20% improvement on a ship flying 100 minutes results in a 20 minute longer flight.
Some of the KDE motors provide exceptional efficiencies at the optimal prop size and copter weight.
I and 2 of my friends are using the KDE 515KV motors with 13", 14" (me) and even 15" propellers.
http://www.kdedirect.com/KDE2814XF515.html
While not on record setting type quad copters, using the Turingy MultiStar LiPos they are proving to provide the excellent efficiencies they advertise with normal flight capabilities extend well beyond 40 minutes.
KDEs are truly superior motors, generally with even better specs than top of the line Tiger Motors and if you select the optimal one for a given size prop and weight copter it should provide excellent record setting capability.
Definitely worth checking out in any case.
Best Regards,
Gary
For any other mfg, in general, for efficiency, choose the lowest KV motor you can throwing a prop that is within 2" of the largest prop it throws. If you can find one of the right lift and KV < 450, throwing a large prop, you will probably end up being within 20% of the most efficient rotor (the T-Motor propulsion system used above). With testing, you can probably get within 15%.
Great job guys, just got a couple of questions for you.
Have you tested turnigy multistar quadcopter motors, and if you have how do they compare to your most efficient motors?
Thanks
If considering Li-Ion in auto mode, ditto the recommendation above. Try the mode and parameters using a LiPo of similar weight, enable CURR, fly, and then check the amp draw at various, speeds, pauses, and accelerations.
Li-Ion and Appropriate Use.
Test the C of the battery. For example, A 4S5P Li-Ion Panasonic battery pack might have about 15 Ah of battery life given the draw of your ship (lower draw rates derive slightly higher capacities). Let's say the ship draws 9 amps, so your draw rate is 9/15 or 0.6C. This give you about 40% margin above normal efficient flight speeds since 1C is generally safe.
Mapping in Auto mode usually means going in a straight line, slowing to a stop, changing directions, accelerating in the new directions, maintaining a reasonable flight. My problems have normally occurred when changing direction at speed (did not slow down).
To test, maybe try a similar weight LiPo, enable CURR and see how close you come to the C (amp) limit. The limit for Rufous was about 1C or 15 amps for more than 2 or 3 seconds.
-
1
-
2
-
3
-
4
-
5
of 6 Next