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.
The wing spar parts hit the U.S. shores about a week ago, cleared customs and then arrived in the bay area late last week. I picked up the parts from a San Leandro warehouse today and then we unloaded them and put them on the new tubing racks that were built a few weeks ago.
All the parts look in good shape, we'll inventory and weigh them in a day or two.
Tammy & Bob H finished building and installing the racks to hold the carbon fiber tubing today, in time for the tubing coming by sea freight that we expect at end of October/start of November.
We did a quick sanity check on the prototype airspeed indicator board, riding the trike and pacing with another bicycle since we don't have a way to display the ground speed of the trike setup yet. The airspeed indicator uses a small pitot/static tube with an analog output, then converts that to a digital signal for the display.
There is some digital noise that makes the airspeed pretty inaccurate near zero, but once it gets into the region of interest above about 5 m/s it is pretty stable and appears to agree within a few tenths of a meter per second with the cyclometer on the other bicycle (though it seemed to read slightly lower than the bicycle gauge). (Note on video #2, the trike rider Geoff is calling out m/s, the cyclometer on the bike is in mph -- the conversion is 2.24, so 6 m/s is about 13.4 mph -- we ran several more tests than the videos shown.)
In a few weeks when we have this hard mounted to the trike and also have a separate speedometer set up for the trike, we'll be able to log both and compare, and make adjustments if necessary.
Here's some photos and video:
Getting ready to test the prototype airspeed indicator
Geoff turns on the prototype airspeed/altimeter setup
Geoff built a breadboard version of the circuit for the ultrasonic rangefinder we'll be using as the altimeter for the DaSH plane. Here is video of a test:
We are building racks to hold the long carbon fiber tubes, the longest of which is 8.5 meters. Tammy got the first two vertical supports up over the last week. In about a week and a half we'll add 6 more supports like these, so they will be ready for the tubes coming by sea freight and arriving at the end of this month (October '12):
6 more vertical supports like these will be added to make the full set of racks for 8 m tubes
Welds:
We continue to practice welding both mild steel and aluminum, so that we can weld the superstructure for the trike and the trailer for the airplane out of steel, and the seat frame for the airplane out of aluminum. Steel we pretty much have down, but welding aluminum is a lot tougher, and we are just getting experienced at it.
We are getting better at, though, and by playing with the frequency and A/C arc balance settings, we have managed to make some acceptable welds. With a little bit more practice, we should be ready to go. Here is an example of a decent weld on the very thin tubing we are planning to use on the seat (1/2" OD, .035"wall thickness). (compare with a few blog posts ago when we were making lumpy messes...):
A good weld by Yohann on 1/2" OD, .035"wall Aluminum tubing
Wednesday, 26-Sept, the rest of our parts shipped by sea freight! This includes all the parts that were longer than 3 m (the Main Spar Center, Middle and Outer, the Rear Spar Parts, and the Horizontal and Vertical tail spars), as well as additional wing support parts that weren't ready for air shipping during our last two shipments.
On Wednesday we had a productive working meeting where we solved three problems in one go, working on the CNC Hot Wire machine, setting up a vacuum bagging system, and working on stop the chain skipping on the drivetrain.
The whole crew working on vacuum bagging and CNC hot wire equipment
Wes had machined longer support shafts and lead screws to increase the working area of the CNC hot wire from 74 cm (29 inches) to 127 cm (50 inches), and he installed the part with the help of myself and Bob H and the Geoff helped us get the hot wire running with the software, and confirmed that all axes moved as expected.
We'll cut some test pieces next week using the new setup, which has a large enough working area to let us cut full-sized wing ribs for the airplane (meanwhile, Zach H cut some of our first full-sized wing sections out of foam using a different CNC hot wire machine, and discovered a glitch in the .DAT file we were using. We don't have photos of that. He'll be cutting more parts in the future)
Re-assembling the CNC hot wire machine with longer lead screws
The finished CNC hot wire machine with 127 cm (50 inch) working area, getting ready to test
Ihab and Geoff worked on setting up a vacuum bagging system with adjustable vacuum pressure, which will allow us to vacuum bag composites and other items requiring vacuum bagging (such as gluing the plywood on the ribs).
Setting up the vacuum bag system
Vacuum bagging pressure monitors, filters and valves
The vacuum bag system in operation on a small sample section of horizontal tail rib foam
Ever since we have had the static test rig set up, we have had a problem that slowing or stopping pedaling caused the chain to jump the sprockets, something that would obviously be inconvenient if it happened in the air while flying. Our original setup had a cage from a derailleur as the idler pulley, and it included a spring that we thought might be part of the problem. We had a whole list of possible solutions, and getting rid of the spring was one of them.
Grant hacked together a new idler pulley setup from scratch, that didn't include a spring, and it works beautifully -- no more skipping. Now we just need to make a final design and figure out how to mount it on the fuselage, while providing for some adjustment.
A day after the blade test, we cut out new full-sized templates for the main wing root chord on a ShopBot.
The finished templates
Later the following week, we started welding practice. We'll be welding the trailer and trike in steel, and also the aluminum seat on the airplane. Welding aluminum is hard, especially with this thin walled tubing which is very easy to blow out. But when we got it to weld, even though lumpy looking, it was surprisingly strong (we hope to get a lot better at these welds before making the actual seat, and if it doesn't work out we may revert to lashing with Kevlar cord instead).
One of many examples of a blowout when attempting to weld
Any ugly but effective and surprisingly strong weld of 1/2"Aluminum tubing with .035"wall
In mid-September, we finished the new blades and did a final trim on the trailing edge, then mounted them on the static test rig and took it outside to test the blades. (The new 3.66 m (12 ft) diameter prop is too big to operate inside the building. The old 3.05 m diameter prop (10 ft) just fit inside.)
The finished blades, with adapter sleeves on the end to fit into our currrent prototype static rig drivetrain
Once we mounted the blades, we took it outside and tested it. We found one issue right away, the chord of this blade is so huge that at the nominal angle of attack, the back of the blade was hitting the chain, so we had to de-pitch by about a degree to make it work (we had calculated the prop shaft length to make sure there was ample toe clearance, but hadn't considered the prop hitting the chain).
The blades seemed to work well at light loading, and once everyone was in place, I ramped up the power and one blade spar immediately snapped near the root. We don't have a power meter on this setup, but I would guesstimate it was only around 200 or 250 W when the prop spar gave way, much lower than our 750 W desired load level (1000 W with safety factor).
This prop had been built using a carbon fiber ski pole in place of the custom-designed prop spars that we have on order, so that we could get the new 3.66 m prop built sooner instead of waiting for a couple of months. These ski poles were slightly heavier than our designed spars, but also slightly smaller diameter. We were hoping they were going to be strong enough, but weren't sure (and also didn't know the fiber composition and ultimate strength of the fiber/epoxy combo used).
On inspection after the break, it looked like the ski poles were made of 4 layers (possibly 5 -- it's hard to tell) -- one unidirectional axial layer, two torsion layers, and one completely circumferential layer that would add almost nothing to the bending strength. (in contrast, the prop spars we designed have spar caps of high strength carbon fiber, with varying number of layers from root to tip).
Since the break was a relatively clean one near the root, we can repair these blades by putting a carbon or aluminum support shaft inside, and then several more layers of carbon on the outside, tying at least one layer into the inner prop skins so they can carry some of the load (we need to add more layers anyway to add about 2.5 mm to the OD to make it fit with the prop hub designed for our custom prop spars).
We'll do that repair in a week or two so these blades are usable, but we also accelerated the manufacture and shipment of the proper prop spars that we custom-designed. They are now being air freighted ahead of the shipment of the rest of the parts that is going by sea and will take about a month to arrive (Ed. Note: The prop spar parts actually arrived on Wednesday). Sometime after we receive these, we'll build a new prop of the new design with the smaller chord size (still TBD if the first version of this will be a built-up style or a foam and fiberglass construction like this one was).
A couple of weeks ago we finished building blade A and blade B of prototype prop #2 for the DaSH project.
The blades were finished by covering with two layers of 1.4 oz/yd^2 fiberglass cloth infused with MSG laminating epoxy.
One blade glued on spar, the other ready to be glued
Both blades glued, with some test sections with various covering material in front
Sanding the blades before glassing
Finish sanding
Finish sanding
Glassing the blades
The finished blades with partially trimmed trailing edges, drying.
The blades came out quite heavy because of the large size of the chord (the weight was roughly 720 grams per blade, or about 1450 grams for the prop, when we are shooting for something in the 800 to 900 gram range for the entire prop).
We have already designed new props for lower climb power, which has the effect of decreasing the chord size, and will make the maximum chord a bit more than half of the chord of this current design, which should make the foam weight about 1/3 as much as these blades, and the glass and epoxy weight about half as much.
In addition to building the blades, we also lubed the CNC hot wire machine and cut a test piece to make sure it was working properly, before we disassembled it to lengthen the working area. (previously, one of the x-axes got partially stuck).
While we were working, we were visited by a gopher snake:
Yesterday we cut out all the blade B prop sections, and today we sanded the carbon fiber spars, and finish sanded all the individual prop sections, weighed them, and then glued the sections for prop A onto the spar. After the Labor Day holiday we'll glue the prop B sections on the other spar, then blend the intersections between sections, and fiberglass the surface.
Some photos from today's work sanding and gluing blade sections:
Sanding one of the smaller prop sections after spackling all the divots
Sanded sections waiting to be weighed
All the prop sections sanded and ready to glue to the spar
Gluing the prop sections to the spar
Blade A with all sections glued to spar, with 67 mm tip to be added to end of foam
