Alfred Vink relates the story of Aviation Club Leiden’s 2019 winter group build.
In November 2019 no one could have known what 2020 had in store for us. In fact, it was business as usual at my local club LLC (Aviation Club Leiden) here in The Netherlands. A club with great history, founded as early as 1936, and still going strong.
The weekly November gatherings included the usual beers, snacks and exaggerated flight line talk, but also discussions about what to do for the yearly winter project. Previous years had seen build projects ranging from a Zagi flying wing to a Piper J-3 Cub, and from a Puddle Star to a Quido. But this year it was going to be something different; we needed something to appeal to a ‘larger’ crowd, while still being affordable, and it had to be something that many of us wouldn’t likely start on their own.
Enjoy more RCM&E reading in the monthly magazine.
Click here to subscribe & save.
Two members of our club (father/son) had been experimenting with a big solar glider over the past year. Moreover, they hold the Dutch endurance record with their glider with a flight of over 10 hours, and they came up with the idea to make a wing with a span of roughly 7 ft in the style of an F5E class glider.
Basically, what they envisioned was vacuum bagging a foam wing, modelled after the Horesji Q12, but with 16 flexible solar cells (SunPower C60) on top and covered with glass fibre and epoxy, with optional ailerons and flaps added later. This immediately appealed to several members as it ticked all the right boxes learning new techniques, solar powered, affordable and free choice of fuselage and setup.Thus the decision was made – the traditional Build Saturdays in January (or so we thought at the time) would see the birth of 12 new Solarwings.
Cutting The Foam
After weeks of preparation by our fearsome build leaders, Ernani and Mirko, 12 club members were eager to get started on that first Saturday. Supplies and tools in hand, checklists were ready made and handed out, and after the first briefing we all jumped at the chance to cut some foam.
We made several foam cutters by creating a spring-loaded frame with a length of resistance wire all hooked up to a laptop charger, which worked like a charm.
The foam of choice was the standard sheets of dense insulation available from the local hardware store. The sheets were first cut into blocks of the appropriate size by means of the hot wire and using a guide to make sure some cuts were square, and others had the correct angle to create the required dihedral for the wing. Each wing half would consist of four blocks glued together, making a span of seven feet in total.
There were four types of blocks, each with a right and left variant. Each individual block was shaped with a hot wire, with the help of sets of 3D printed ‘ribs’. These ribs were made up of an upper and lower part, with extended trailing and leading edges, and two pilot holes with which they were pinned to the sides of the foam block. Both ribs had their tops covered by a small strip of self-adhesive aluminium, thus making a smooth surface and protecting the printed rib from being distorted by the hot wire.
After cutting the upper part the printed rib was switched for its counterpart and pinned into the same holes. Cutting the lower part then resulted in a perfectly shaped profile. Well, after a bit of a learning curve that is – practice makes perfect!
Next step was creating the two wing halves by gluing the four blocks together to get the appropriate shape and dihedral. A small one degree at the wing root and another five degrees at around a third of the span. PVA turned out to be perfect for the job.
Birth Of A Wing
Although we were very pleased with the result so far it could hardly be called a wing yet. Next step was sanding the foam, especially the LE, the transitions between the blocks and shaping the wingtips. This all had to be done with extreme care as the TE was very fragile and foam dents easily, leaving unwanted marks. Also, never mark anything on the wing with a marker or a pen because as small as the mark might be, it will expand big time when in the vacuum, leaving a nasty stain.
Not only the wing blocks were glued together, but also the upper and lower left over parts of the original blocks, creating a perfect mould for transporting the wing; they would also be used later in the process when vacuum bagging the wings.
At the root side a balsa W1 rib was added with three holes – two pilot holes to line up the two wing halves and one to accommodate the wing joiner tube. These ribs were CNC cut by one of the participants in the project. Just inside of the rib, two balsa/ply blocks were glued into the wing and drilled for mounting the wing to the fuselage.As everyone had a different fuselage all the Solarwings were now starting to look different. My fuselage of choice was the Sapphire, available from Hyperflight, which is rightly suited for gliders of this size.
Ready For Covering
Next up was testing and matching the solar cells into somewhat equal sets of eight to be fitted on top of the wings. All sets were created using a template as there had to be a difference between right and left, and some gaps between the cells needed to be slightly larger due to the dihedral. When all was approved the cells were soldered into sets. But more preparations had to be done to the wing before the sets could be placed on the top, and all wires had to be put in before any covering was applied.
Besides the wires for the solar cells, servo wires had to be embedded into the wing if fitting the optional flaps and ailerons. For the cells a short wire was run at the root from the underside to the topside and a long wire was run all across the underside of the wing to the far side of the cells and brought to the topside. At the required position of the wing servos a small circular hole was made, just smaller than the actual servo; the servo wires were fitted in the groove, ending in a small roll in the created spaces. Rolling up the servo wire like this and covering it with a bit of foam or tape, making it level with the surface, prevents the foam edge around the servo from collapsing when the vacuum is applied.
The groove with the wires was filled with epoxy, mixed with white micro-balloons. After filling with the correct amount, cover the groove with painter’s tape and turn the wing over; the epoxy mix now spreads evenly against the tape leaving a smooth finish, flush with the wing, once hardened and the tape removed.
Several small strips of fibreglass fabric were cut by rolling out very thin double-sided adhesive tape on the glassfibre and then using a rotary cutter to cut along the edges. These strips were used as reinforcement all along the wing’s LE. The same method was used for cutting wider strips of Kevlar, to be placed on the wing as hinges for the aileron and flaps.
At this time the wing is ready to accept the solar cells by using some epoxy mixed with a small amount of gap filler before sticking them on top of the wing. This is to prevent the cell breaking during the vacuuming process. Lastly, both sides of the cell set were soldered to the wires protruding from the wing.
Let’s Cover It
Next, the tricky part that none of us had ever been tempted to try on our own. Luckily, we had our checklist guiding us through the process step by step. Covering the wings with glass and epoxy turned out to be nothing short of a test in time management and some of us failed miserably. In our clubhouse four large tables are available, so only four wings could be done at a time and at any moment no more than 10 people were present. Still, at times the stress could be felt from miles away…
First, two Mylar sheets had to be cut to the shape of the wing and covered with wax on one side. The Mylar doesn’t stick to the epoxy so when used as the outer layer it results in a very smooth finish of the wings after vacuum bagging.
The tables were then covered with thin plastic sheets, on top of which the different materials were laid out 50g (top) and 80g (bottom) fibreglass fabric, with several carbon roving strips used as reinforcements near the root and across the wing. All were cut to approximate size and taped along the edges to the plastic. At this point all coverings were rolled with epoxy, just enough for all covering to appear equally wet. Next the Mylar was placed on the glassfibre, with the waxed side towards the fibre, and dry rolled once again. It was then cut along the edges with a rotary cutter and turned over. After that the plastic was removed and the carbon roving and reinforcements were placed on the wet fibre. The final step was giving the fibreglass LE strip and the Kevlar hinges a brush with some epoxy.
Now the covering can be put onto the foam, making sure the Mylar follows the shape of the wing nicely and leaving a small gap at the leading edge. When both sides are done, the LE gap is covered by peel ply taped to the Mylar.
With all the wings ready to go the makeshift vacuum was prepared. Each wing was placed on some plastic sheet, with paper towels draped over the TE. Just aft of the TE a spiral cable wrap is taped to the plastic and connected to the hose of the vacuum pump to help create an even suction across the full length of the wing. The plastic is then folded over the LE whilst using sealant at the edges of the plastic to create an airtight bag.
The vacuum pump was a DIY project. A salvaged refrigerator pump was hooked up to an empty fire extinguisher to act as a buffer and controlled by a small Arduino board and a pressure sensor; we had it set to get a vacuum of 0.2 bar and kick in again when it reached 0.5 bar. It’s all a compromise between the strain you want to put on the wing and how often the pump has to kick in again in case of a small leak. The force equals to about 50-80 kg/sq.dm.
Paper towels help prevent wrinkles or bubbles, as air can be passed right to the end whilst the vacuum is applied. The wings were then placed back into their mould and weighed down to make sure they kept their intended dihedral and shape while the epoxy cured. All wings were placed in a Styrofoam enclosure with a heater set to about 30 degrees for the first 24 hours.
The bags were first opened after 48 hours. After removing the peel ply and the Mylar a super smooth surface emerged. Okay, so not all were equally good, but it was a learning experience; some wings showed some light spots over the cells, caused by not enough epoxy or specks of dust in that particular area. But, overall, we had very pleasing results. All that was needed was some light sanding of the trailing and leading edges.
Flight controls were next, with the flaps and ailerons cut out very delicately with a sharp knife without cutting through the Kevlar hinge. The best way is to run your finger over the Kevlar while cutting from the opposite side; you can then feel the knife just touching the Kevlar when you’ve hit the sweet spot. On the hinge side use a blunt knife to scratch the epoxy outer layer, then your control should be able to move freely. The servos were put in place by excavating the previously made holes with servo wire, cut to fit and wired up. This needs to be done with care as the wing is not much thicker than the servos and you don’t want to be cutting into the back end of your solar cells.
Last thing to do on the wing was to drill a hole for the wing joiner tube, for which we used a 3D printed jig to make sure we drilled at the correct angle. The joiner is made up of a thin carbon tube with piano wire ‘drilled’ into it; the piano wire expands the carbon tube just a hair, making it a very snug fit with our tubes.
With the wing finished we needed some way of optimising the charging process. The glider was meant to be flown with 2S or 3S packs and the output of the cells varies, depending on light and temperature. So, we designed a MPPT (Maximum Power Point Tracker) controller, with a buck as well as a boost function, which optimises the power we can get out of our solar cells. The MPPT consist of a few FETs controlled by an Atmega328p chip.
Our biggest concern however was the form factor. Lucky for us one of the project leaders was an electronics student and my son happens to be an electronics engineer, so they got together and came up with our own PCB design, which we ordered from our Chinese friends, as well as all the components. In the end the finished board measures just 25 x 54mm – small enough to fit in almost any fuselage.
The whole project was based around making a common solar wing, while everybody could go wild with their own ideas about the set up and fuselage.
Personally, I chose to go with a 2S set up with the Hyperflight Sapphire fuselage, with a carbon tail boom and an all moving tail. The stabiliser and fin are made of 3mm Depron, which after some light sanding and shaping underwent the same treatment as the wings. The rudder was hinged with a Kevlar strip as well.
Total weight of the different builds averages between 750 – 800g. I opted for a HobbyKing motor, the Turnigy L2210A-1650, in my setup. With a 8″ x 5″ folding prop it gives me 110W at 15A; ample power and a perfect match with the 20A ESC.
Finishing the build was pretty straightforward. It just came down to putting all the pieces together and setting up my radio. All this ultimately led to a beautiful day in April when my Solarwing had its maiden and what great fun to see it take-off without a glitch. After a total of around 10 build days she was finally able to show her true beauty.
Have a look at:
While we were all excited from the get-go, this project exceeded our expectations. A lot of time and effort was put into seeing it through till the end, but all of us in the hobby know the rewarding feeling of completing a project and seeing it perform way better than you thought possible. The whole thing was great fun and the Saturdays became days to look forward to, because what’s better than doing something you love with like-minded friends and learn something along the way?
It all just emphasises the added value of being a member of a local model club. You can find out more about Aviation Club Leiden here:
go to the Aviation Club Leiden web site
Enjoy more RCM&E Magazine reading every month. Click here to subscribe.