Fully 3D printed sailplane model. optimized for 0.2 nozzle (weight reduction)

thingiverse

Overview This is same Walt's Fully Printed Glider model, but optimized for 0.2mm nozzle to reduce weight and tweaked a little bit to improve printing accuracy. Original design looks very interesting for me, but there were too many complains that resulting plane is too heavy. So I had to redesign some of the parts, keeping overall design unchanged. I am uploading Solidworks models so that you can add more tweaks if needed. UPDATE 14-Jun-2015: I finished printing the fuse and installing electronic components. Resulting weight is 636g including battery. Plastic parts are 320g. However I doubt this plane can fly, and weight is not the problem. The real problem is weight centering which is 57% of the wing horde (tail is too heavy comparing with nose). It should be 25-30% for a normal flight. In other words plane will crash right after take off. UPDATE 29-Jun-2015: I redesigned the fuse. I moved main fuse body farther to the nose, extending overall fuse+tail length by approx. 30 mm, while keeping wing-stab distance about the same. I also reduced tail parts diameter as well as walls thickness to reduce some weight. This all should move CG to 25-30% depending on motor and prop I just started printing these parts so perhaps I’ll be adding more tweaks to STL files. UPDATE 12-Jul-2015: Minor update just fixed few issues on fuse main parts. I finished printing fuse parts - everyting fits perfectly. Still stuck on vert stab redesign. UPDATE 15-Aug-2015: Finished designing vert stab. I preserved overall shape, but could not reproduce airfoil. No big deal, think. Since tail got narrower, I had to decrease rudder horn size to fix the tail. I also updated motor mount part. I am assembling the model right now. Will post updated photos and weight summary little later. UPDATE 25-Aug-2015: Maiden flight was not too long. Just 3 seconds :) It did a loop and crashed vertically on the nose. It might be caused by one (of more than one) of these reasons: Tail is still to heavy Horizontal stab was not set correctly in idle state and cause plane constantly pitch up There is huge backlash in stab mount. It plays for several degrees which can be too much for fully deflectable stab I am reprenting some of the parts that were broken on crash - wing tip, wing center, motor mount. Then I'll probably investigate whether I can use bowden for horizontal stab (to reduce backlash) UPDATE 14 Sept 2015: Second flight ended the same wayhttps://www.youtube.com/watch?v=7Otw3EAnxgw Not sure what happened. I removed all backlash on elevator, set elevator to 3 degrees down, double checked CG it was about 30%. I will have some consulting with r/c plane modeling gurus. Then will decide what should I fix in the model. UPDATE 25-Nov-2015: Made another design change Made nose even longer changed motor mount angle to 4 degrees down to address heavy propeller momentum few minor changes in fuse parts, mostly reducing weight Wing parts are untouched This should move CG closer to the nose, Now should be around 20% I am in progress of printing these parts. Will make another attempt UPDATE 04-Sep-2016: Finally took courage to make another flight attempt. As suggested I went to a field with a high grass. There I tried to throw plane several times from a small altitude. I was under impression that I throw it fast enough, but video shown that speed was too small. Obviously, plane wants more speed to fly. Otherwise it falls on a left wing. Motorized flights (thrust at 30%) were a little better, but I still could not understand how to trim it right. CoG was at 20-25%, but when I add more thrust plane raised its nose too much. Eventually I had to trim elevator until it balked to the rudder. No more room to trim. Fortunately it helped. Finally I added more thrust and it took off!! Though, the flight was not too long. I could not turn it back. Every time I tried to turn it tended to fall. Finally it raised its nose too much, did a loop and then I totally lost control. Another thing I found after crash is too hot motor. It definitely requires some cooling. Here is the video of my attempts. If you feel it is boring – go directly to 1:51 for the flight.https://youtu.be/BaLBQySvFkA Summarizing my findings: Plane needs speed to fly, it does not glide but fall immediately Rudder does not turn plane as it should No motor cooling CG is still not in place, needs to be moved a little forward. Not enough room for elevator to move (fixing CG should help) It looks like there is a huge propeller momentum pulling plane up and left What next? I have to regret my followers: I will not continue working on the project. Now it is time to move on. I have plenty of other projects to work on rather than stuck on this one. Though, I will be glad to see your prints and remixes. During the project I got plenty of experience in different areas: how to design plane models in Solidworks how to calculate essential plane parameters how to select proper material to print learned a lot of 3D priniting stuff: different slicers, nozzles, printing speeds and temperatures, and so on how to set up a model for its first flight I am still interested in the idea of a 3D printed plane. But next time it will be something different. I would prefer to have at least 4-channel control (add ailerons) and I like take off the ground on wheels. I have not selected a model yet but I have couple ideas in mind. Stay tuned. Equipment I used these electronic components Motor – Turnigy 2209 1050kv 28 turn 15A Prop – selecting between 10x6 Direct drive, APC 10x5 and SF 10x4.5 Servos – Hobbyking HXT900 Receiver – Hobbyking 6 channels (plan to change it with 3 channel receiver) ESC – Turnigy Plush 18A Battery – Zippy Compact 1300mAh 3S Here are my weight stats (for previous attempt) Total for the wing, including carbon rods - 231g Fuse printed parts – 129g Total with electronics – 636g I am uploading weights table as an Excel spreadsheet. Refer to second sheet for weight centering calculation. Changed parts description (comparing to Walt's version) Wing panels: reduced walls thickness to 0.4mm. This will force exactly 2 perimeters (external and internal) when using 0.2mm nozzle. These 2 perimeters are sticked together very well giving extra strength (when printing original STL there was a distance between outer and inner walls and therefore printed model strength was quite poor) added couple additional 2mm ribs every 20mm - this helps keeping wing shape more accurate, bending resistance removed hooks to mount sections. Original hooks were too easy to break. Instead I added thickening on the top and bottom to increase gluing area Wing tips: It is now hollow with ribs instead of solid. same thickening at the bottom to increase gluing area Wing center links: Original file had number of holes that suppose to reduce weight. But these holes require couple perimeters to be printed around them which add some weight. It appears that model without these holes and 5% infill weights less than original part with holes. Original model had a bug around rear screw hole. I could not manage to fix it, but it does not make an impact on result quality (small piece of extra plastic could be easily removed after printing) I added small membrane inside screw holes. It is needed to correctly print ceiling and hole above. Membrane can be easily removed with a drill I do not like supports, but they work really well for this part. If you do not like supports you can print this model using bridges. I added couple walls and supporting bridges at the solidworks file, but they are disabled by default. These walls should be removed after printing. Horizontal stabilizer: Same trick (as wing tips) worked here as well. I redesigned stab to be hollow with ribs Fuse: I completely re-developed fuse. My fuse version is twice lighter than original Decreased shell thickness to 0.6mm – this is enough to keep its shape, but stay at lower weight. Reduced main parts thickness and added big holes in order to reduce weight Redesigned servo mount, changed servo layout so that straight carbon push rods can be used instead of Bowden. Servo mounting holes are designed to fit HXT900 servos Motor mount is designed to fit my Turnigy 2209 motor. Mounting holes are 16 and 19mm and compatible with many other motors Fuse shell parts are now mounted with special hooks. No screws. For front section I designed a bayonet-like mounting Tail body moved little further to the nose to fix CG. Fuse made little longer at the nose by 40mm. Wing mount shifted closer to the tail (fixing CG). To preserve wing-stab distance tail was extended a little bit Tail: Tail diameter decreased to 20mm, thickness to 0.8mm Tail length extended to keep wing-stab distance Horizontal stabilizer and link parts remain the same Spinner: I did not have 40mm spinner, so I just designed a printable one. I took collet, spacer and nut from other broken spinner. Even more weight reduction There is also possibility to reduce wing weight even more. The idea is to reduce wall thickness to 0.2mm - same as nozzle size. But this required couple extra ribs. Unfortunately I got this idea after I printed my wing. I printed couple sections (just to evaluate the approach) but not the whole wing. Wing section weights about 9g, wing tip is about 10g. This all gives projected wing weight about 150g. Surface of the printed part is not quite smooth, so certain degree of sanding is required. I uploaded these files with Experimental* prefix. Please send me a note if this approach works for you. Instructions I used PLA as it gives more strength than ABS. Unfortunately PLA is heavier than ABS, but layer adhesion is more critical here. Print at maximum filament temperature and lower speed fot better layer adhesion and overall quality of printed parts. 0.2mm nozzle was used to print most of the parts, layer height - 0.15mm. Wing panels and wing tips: 1 perimeter, no infill, 3 layers for top and bottom. Wing Center Links, Fuse Main parts, Tail parts: 2 perimeters, 5% infill, 3 layers for top and bottom. It is better to print fuse shells with 0.3mm nozzle and 0.2mm layer height Spinner is also designed for 0.3mm nozzle/0.2 mm layer. I used 2 perimeters, 100% infill, 3 layers for top and bottom. Here are couple hints You may split spinner base into top and bottom parts so that both parts are flat. Slic3r does model split out of the box You may print spinner body upside down (nose part down). Printing will be more accurate, but you’ll have to use good model cooling. Rudder horn and ridder shaft are printed with 100% infill, nozzle size and layer height.does not matter. Rudder shaft is a small piece of plastic 3mm diameter, You may use any suitable material here. I decided to print it, I used slic3r's cut feature to split it along the part length. I was not happy with how Slic3r generates supports, often it produces holes in the main body. So I used Cura to slice parts that require supports. Slic3r works well for parts without supports. To glue parts I used Dichloroethane (glue for plastic). Remember that dichloroethane is very toxic and you’ll need good ventilation. CA glue should work as well. For better parts positioning I added small holes on the glue sides. Use toothpick or 5mm piece of the filament as a pin. I am uploading Solidworks files for all parts I changed. In certain cases I used self-made supports, but they are disabled by default. You can re-enable them if you want. In order to orient parts properly I added ‘output rotation’ section in fuse Solidworks file. Enable corresponding element while exporting a part.