Postings about the DaSH Human Powered Airplane project. Our goal is to have fun and learn while building an HPA on a budget of no more than the price of a new car by keeping the design simple and the build time to a minimum.
Craig's Mile Long Flight
(1.62 km, 4 mins duration)
We had our best testing day ever. 7 pilots, 8 flights. Once again Craig completed the longest flight ever for DaSH, this time flying from the start of the smoother part of 32 L, to very near the end, for a 1.62 km (1.01 miles), 4 minute duration flight that only stopped because he reached the end of the runway.
We had a successful return to flight after fixing the damage from our crash in June. The new nosewheel assembly with larger wheel in front worked great, as did our new joystick that has better spring feel for the individual pitch and roll axes. This helped us concentrate on only lateral movement without disturbing the pitch trim, which we tried to control only with trim buttons except for takeoff rotation, gust responses, and landing flare.
We revamped our wing catching procedures, and did more pilot training, and added a brake to the main landing gear wheel, and that helped make the landings go much more smoothly than in June. We had one inconsequential touch of a wingtip at the end of the flight with no damage, and had no runway excursions. A big improvement from June.
Ariana Peck, the second female to fly DaSH, had our longest flight so far by anyone other than Craig, at 850 meters, and 2 minutes, 24 seconds duration.
These are the statistics on all of our flights from Saturday 11/5:
(Note: all measured from first takeoff to final landing for those with multiple touchdowns)
Flight # 1, 5-Nov-2016, Alec Proudfoot, 355 m, 48 secs
Flight # 2, 5-Nov-2016, Craig Robinson, 1.62 km, 4 mins 0 secs
Flight # 3, 5-Nov-2016, Ariel Herrmann, 600 m, 1 min 26 secs
Flight # 4, 5-Nov-2016, Marshall Gusman, 570 m, 1 min 18 secs
Flight # 5, 5-Nov-2016, Ariana Peck, 850 m, 2 mins 24 secs
Flight # 6, 5-Nov-2016, Geoff Bower, 80 m, 11 secs
Flight # 7, 5-Nov-2016, Geoff Bower, 660 m, 1 min 25 secs
Flight # 8, 5-Nov-2016, Greg Thomas, 700 m, 2 mins 0 secs (Note: Geoff's first flight was a short hop due to the chain coming off of the chainring)
So far, among all of our testing days, DaSH has flown for over 11,000 meters (7 miles), and over 27 minutes duration, in 22 flights by 8 different pilots.
Here is a video playlist of all the videos we have from November 5th:
First Flight at Moffett Federal Airfield -- Alec Proudfoot piloting photo by Wink Saville
Video of first DaSH Moffett flight June 25th
360 degree VR video of first DaSH Moffett flight June 25th
Craig Robinson on his 1.3 km flight at Moffett Federal Airfield
We had 5 successful flights at Moffett near the end of June, including the longest flight so far (1.3+ km). The last of the 5 flights ended in a shallow dive into the pavement instead of a smooth landing, doing some damage to the nose wheel and the front of the fairing, which will be repaired to make the airplane better than before (more details below).
For this third round of testing, we moved to Moffett Federal Airfield. Moffett has two very long runways -- Runway 32 L/14 R, which is 1.55 miles (2.5 km) long, and 32 R/14 L, which is 1.73 (2.78 km) miles long. Unlike our previous test site, Half Moon Bay Airport, which has only one runway that had to be shut down entirely during our early morning tests, Moffett's two runways allow the airport to remain operational for incoming/outbound flights while we were testing. This is convenient because it allows us to operate as long as winds are favorable, rather than having to shut down earlier in the morning. It is also closer to our home base, only minutes away from our workshop.
Because it remains an active airfield during our testing, we had to undergo training a few weeks before our tests, so that we could learn how to properly operate and follow airport procedures. We had 18 crew members successfully complete the training ahead of our test flight date.
Day 1, June 25th 2016
On Saturday, June 25th, I (Alec) flew first to test out some of the changes we had made since the previous time. We were checking to make sure the changes to the idler pulley worked to help keep the chain on (we switched to pulleys with side channels), and also ended up testing out the new main gear wheel.
When we were wheeling the plane out to the runway, we had another failure of the tread of the blow molded plastic main wheel. So we switched out to a new, slightly heavier (by 150 grams), main wheel that has a grey rubber tread. We also had some issues with the avionics sensors and couldn't get airspeed and rpm readings to work. After fiddling for about 15 minutes, we decided to move on and fly as-is.
The new wheel, with grey rubber tread
My first flight to shake out the changes went great -- I flew almost three times as long as my longest flight before, going 636 meters in 1 minute 20 seconds. I tried getting higher at the start of the flight and got up to about 2 meters, then in the second half of the flight was back to about 1/2 to 1 meter height. The pedaling got easier there and I was able to continue on about 30 seconds beyond when I initially thought I would need to land because of fatigue -- I think I must have trimmed the plane slower, but given the balky airspeed sensor, we couldn't tell for sure.
Alec's flight from chase bike
Alec's flight in 4K video (Slight SloMo ~80% speed)
First flight at Moffett Federal Airfield
The plane was now positioned in the approximately halfway down runway 32 L, and Geoff Bower took over. He had a good flight to the end of the runway of 700+ meters long (exact distance not determined due to GPS recording failure, but using visual marks, it was 700+ m).
Geoff's flight, from multiple angles
Geoff Bower's flight in 4K video
(Slightly Slo-Mo ~ 80%)
360 Degree VR video of Geoff's Flight
Now we turned the airplane around and Craig Robinson got ready to fly in the other direction. We were lucky because it was a very low wind day, so flying in each direction was a possibility. But when he went to take off, he made it about 1/2 meter before the nose wheel failed and the takeoff run was aborted.
It turned out that the turntable bearing we were using on the nose wheel was fairly weak, and some combination of landings in February at Half Moon Bay Airport (particularly the last one where we went off the runway), and the two landings on the day at Moffett, had spilled enough of the ball bearings out that it couldn't function anymore. Unfortunately, we didn't bring any spare parts for that part. But we had some back at the shop, a 15 minute drive away.
Nose wheel break on Craig's first attempt
Once we got it fixed, the wind had picked up a bit. Craig had a long flight, over 1.3 kilometers, our longest so far, but he got bounced around a bit by turbulence. We probably wouldn't want to fly in any higher wind conditions than this.
At the end of the flight, he had a left crosswind and had to put left rudder in to try and keep the wings level, and this caused the plane to roll off the right side of the runway. When it hit the gravel off the runway, the front wheel stopped abruptly and the bearing failed again, spilling ball bearings out. We had a spare this time, but due to the wind conditions, decided to go back to the setup area and break down for the day, and fix the front wheel bearing.
Craig's 1.3+ km flight at Moffett from Chase Bike
Craig's 1.3+ km flight in 4K video (80% speed)
Craig's 1.3+km flight in 360 VR video
Craig up high during his 1.3+ km flight
Working on the nose wheel bearing
Day 2, June 26th 2016
After repairing the nose wheel bearing the previous evening, we were out of spare bearings, so we bought a different kind of caster wheel at a hardware store as a backup.
First up to fly today was Greg Thomas, who'd had a couple of flights back in February at Half Moon Bay.
Setup went quicker the second morning because we were practiced, and because there were fewer planes on the ramp and we got to set up under a large bank of lights, which really helped.
Greg's first flight was short, 125 m and 18 seconds, as he had an unexpected landing early on, and due to the angle with the runway, wisely decided not to try and takeoff again.
Greg's first flight, 125 m, 18 seconds
Once again, we had the same issue with a crosswind -- on landing the pilot had to yaw to keep the wings level while the wing chasers ran to catch the wing cords. This caused the plane to yaw towards the crosswind direction and roll off the runway edge, then stop abruptly on gravel. With no brake, the pilot had no way to stop the plane until the wing catchers caught up with it. (We'll fix that next time, see below.)
Once again, this broke the turntable bearing on the nose wheel. We were out of spares of that part, but in less than half an hour were able to fit the different, slightly heavier caster wheel we had bought at the hardware store the night before into place to make the plane operational.
Greg's second flight had a bit more wind and he had to correct a couple of times on to try and keep the plane on centerline. Unfortunately, our current joystick doesn't have a good detent feel or enough spring to differentiate the pitch and roll axes, and he inadvertently put in some down pitch while focusing on correcting the roll. The plane did a shallow dive to the runway and he wasn't able to recover in time. It hit at too steep of an angle for the nose wheel to roll, and there was some damage to the nose wheel assembly -- the carbon tube broke off -- and there was some damage to the front of the fuselage fairing. Greg was unharmed. The prop also hit and sustained some damage, but should be easily repairable.
We're looking at this as an opportunity to fix some things that needed fixing anyway -- we're going to replace the turntable bearing that was causing us so many issues this time, and replace it with a setup with two bushings. The fairing nose was always a bit bumpy and too wide, we are going to replace it with a molded Kevlar nosecone that will be about 80 mm narrower, and also smoother, hopefully cutting down on drag.
Greg's 2nd flight and subsequent crash, from Chase Bike
Greg's 2nd flight and crash from multiple angles
Damage to fairing nose after the crash
Broken nosewheel tube
Damage to pilot seat
It survived quite well, the mesh took some of the impact load
Damage to propeller tip
Rib damage from prior ground loops
Ground handling damage to vertical tail
After reviewing the damage, we have a plan for fixing the nose wheel assembly, building a new, smoother fairing nose and repairing the damage to a few of the ribs that came from some of the semi-ground loops on earlier flights. We also have come up with a multi-part solution for ending the off-runway excursions that were causing us problems, which involves a hand brake on the main gear so the pilot can slow the plane, longer wing cords for the wing catchers, and different procedures on how to catch and stop the wing.
Craig Robinson pilots DaSH during his 1 km long flight
Craig's 1 km flight filmed from Chase Bike
On the weekend of February 21st & 22nd, we had our second test flight weekend, and it was a great success.
We entered the weekend with a few main goals:
- Verify that our fix to the vertical tail mount worked
- Verify that changes to the tail surface slew rates improved handling
- Test that our fixes to prevent the chain from falling off worked
- Get successful readings from the powermeter and the new airspeed sensor
- Have successful flights with as many new pilots as possible
We came out of the weekend meeting all those goals and more. We had myself and 5 other pilots available and ready to fly. In our best case scenario, we thought that 2 or 3 people each day might get a chance to fly, and there would be just an outside chance for all pilots to get a flight in on the weekend. In the event, only I flew the first day before weather conditions and the condition of the airplane stopped flying for Day 1, but on Day 2, all 5 of the remaining pilots got a chance to fly, with most flights of significant distance and duration. The longest flight was just under 1 km (994 meters) and 2 minutes 24 seconds duration, only ending because we were running out of runway on which to fly.
This bodes very well for our future flights, that we hope to be of longer distance and duration. And having 5 brand new pilots do so well in their first flights speaks well of the handling qualities of the airplane.
Here are some photos from the day, followed by short descriptions of the flights from Day 1 (Feb 20) and Day 2 (Feb 21), with videos of most flights.
Assembling the airplane before dawn
Greg Thomas flies DaSH (453 m flight)
Craig Robinson flies DaSH
Craig Robinson taken from Chase Bike
Geoff Bower flies DaSH
Closeup of DaSH flying
Wheeling DaSH out to the runway
Ginger Kroft after her successful flight
Alec Proudfoot preparing for flight
Day 1
On Day 1, I was slated to fly to plane first, the main goals to be making sure our tail mount fix worked and that handling was improved given our adjusted tail slew rate settings. We also wanted to make sure our chain guide fixes prevented the chain from popping off as frequently as it did back in December, and we also wanted to make sure the airspeed indicator and power meters worked.
I was somewhat out of shape, and had been suffering from a cold, so I wasn't in as good of a condition to pilot as I was during our first flight back in December. We also had more challenging conditions, a right 45 degree crosswind from 3 to 6 kph (compared to 3 kph straight down the runway back in December).
However, the first flight went pretty well, albeit a somewhat short one at 125 m and 18 seconds. We didn't realize it at the time, but I landed a bit sideways due to the crosswind, and that damaged the main landing gear wheel somewhat.
My second flight was longer, 178 meters and 26 seconds. The takeoff run was long because of the bumpy main landing gear (which we still hadn't discovered), and I just ran out of gas at the end of the flight. Geoff later determined that flying in the runway 30 'uphill' direction is roughly 50 Watts extra effort compared to flying level (depending on pilot weight), so that, added to my poor physical condition made a longer flight problematic.
But I was very happy with the performance of the plane. It handled great, the tail mount fix obviously worked, and the chain was staying on. The airspeed indicator had worked pre-flight, but an auto-shutoff feature prevented us from getting reliable airspeed data from our new sensor (the old, less reliable sensor was still working and recording data). A software fix by Rodrigo overnight cured the auto-shutoff problem for the next day.
We were going to do more flights, and get more pilots going, but three things happened almost at the same time: - the crosswind picked up to 10 to 12 kph, right at the limit if it was down the runway but probably too much for a crosswind, a support wire fitting for the kingpost snapped, and we had a radio receiver/servo failure in the tail. The latter two might have been partially related to the higher wind conditions. So we were done for the day.
Videos of Alec's Flights:
Alec Proudfoot Flight #1 from Chase Bike, 125 m
Alec Proudfoot Flight #2, 178 m
Day 2
On Day 2, the intention was to get as many of the remaining pilots (all of whom are more athletic and better suited for longer flights than I) flying if possible. However, I thought at best we would get 2 or 3 pilots in the air.
But we lucked out and it was a near nil-wind morning, with wind usually below 2 kph, and the highest reading we saw was about 3 kph. That meant that we could fly the airplane in either direction down the runway, saving precious time. (it would take 20 or 30 minutes to walk the plane all the way back down the runway to try another upwind takeoff, had there been wind)
Fixes were made to the tail controls (servo and receiver) and the broken kingpost support wires, and the plane was ready to fly again the next day. (The servo & receiver failure was caused by high power going through the receiver due to high load when the wind came up -- rewiring avoided this problem in the future, so that the power line for the servo no longer is routed through the receiver, but comes directly from the battery. The Kingpost is not necessary for flight, and in fact will be removed entirely during record runs to reduce power, but is used to hold the wingtips up prior to flight to help make ground handling easier. We replaced the broken steel wire fittings with braided Kevlar cable.)
Everyone got to fly. Conditions were perfect. We couldn't have planned out a better day if we'd tried, and we learned a lot and all the pilots got some significant flight experience.
Greg Thomas went first. His first flight was 114 m long and had a long takeoff run. We noticed the clicking sound and discovered the damaged main landing gear wheel from the previous day (the tread had a large lump on it, though the wheel was still functioning), and were able to swap it out in about 5 minutes with a new wheel. His second flight was fantastic, 453 meters, up to near the end of the runway. (Again, this was in the 'uphill', higher power direction)
We turned the plane around, adjusted the seat, and then Craig Robinson flew almost the length of the runway, with a flight of 994 meters, only stopping because he was running out of runway. He could have kept going much longer if he had the room.
Control in both flights was quite fluid and worked well, with no problems.
In quick succession, we turned the plane around and Geoff Bower flew 601 meters in the runway 30 direction. Ginger Kroft, the first female pilot of DaSH, did a 290 m flight down runway 12.
Then we were running out of time, and needed to reposition the airplane back to the taxiway halfway down the runway, in order to get the airplane off the runway before the closure time ended (Half Moon Bay airport runway was closed down for us, and we had the runway from 6:30 a.m. to 9 a.m.). Flying the plane was quicker than walking it, so our final pilot of the day, Ariel Herrmann, did a short hop back towards the middle of the runway (actual flight about 15 meters, with a long takeoff roll and landing roll), ending up adjacent to the taxiway, where we then walked the plane off the runway to end the day.
We had found that our new circular chain guides on our top and bottom chainrings worked well and prevented the chain from skipping off (which happened back in December during test runs mostly from flexing of the fuselage on the rough pavement -- the chain bouncing smoothed out during flight).
However, on about half of the flights this weekend, the chain now came off of the idler pulleys instead. The chain still worked, it was just rubbing against the wood washers/spacers for the idler pulleys instead of the pulleys themselves, so that was adding extra drag to the drivetrain (interestingly, on both Greg and Geoff's long flights, the chain was off the idler pulleys the entire time...).
Ariel's flight was just a short hop because he realized that the chain came off for him as well, and he didn't want to put too much power into the system in that condition.
We'll fix this by changing the idler pulleys to be the type that have sides to them (similar to our circular chain guides for the main chainrings) before our next flight.
One other issue that came up was we had significant dew both days. This affects the performance of the plane in 2 ways -- added weight, and decreased airfoil performance. We don't know how much this affected the performance of the plane (roughly, it looks like 30 to 50 extra Watts (not counting uphill or downhill power adjustments as described above) were needed by each pilot compared to predicted). We expect much of this is due to dew, some of it due to unfaired parasite drag items like the wing mounts, and some because we flew without the pilot door covering so there was extra fuselage drag. We'll work on fixing these things (the dew part will be a matter of weather conditions) before our next flights.
That's is for now, we hope to continue improving the drivetrain and decreasing the drag on the plane during out next test flights, and once we have longer flights, and have gathered more power data, we hope to move to the desert for even longer flights, hopefully sometime later this year or early next year.
Below are videos of some of the rest of the Day 2 flights as described above...:
Greg Thomas Flight #2, 453 meters
Craig Robinson 1 km flight from Chase Car
Geoff Bower flight, 601 m, from Chase Car
Ginger Kroft flight, 290 m, from Chase Car
Ariel Herrmann, short hop to reposition plane on runway
