WimacWest

[Andre Germain]

Moved to a hobby farm early may and too busy with the property to get to rolling out the Nilex 700$ 15x300' geotextile (on the tractor's pallet forks in the image). I did finally get to mow a 300' grass section lined up with the prevailing winds. Property is 30' above surrounding farm fields, so you can fly below you!

30 seconds from my shop in the barn and same from the kitchen!

Hopefully I'll roll it out soon and get to fly for the first time this year.

[briankizner]

beautiful place. Congratulations

[Andre Germain]

Brian,

you are right on the money! See the 360 view of my *yard*, you got to use your scroll bar to see it all [to see it properly, save to disk and open it outside of your browser]. It was taken from the end of the runway, the highest point of the whole property and indeed the area. I had to decrease the resolution of the original 7000x1100 as it was 2.5MB. The shift from blue to white of the sky is because I should have removed the polarizer from my DSLR, or at least spun it so it wouldn't influence the image].

That silver thing is the observatory housing my telescopes. Two birds, one stone... dark skies for my telescopes, open skies for my EDFs, you can guess, I'm a very happy camper.

[Dwight Macdonald]

It looks like you'll be busy ... enjoy!

[Andre Germain]

Thank you Dwight, I believe I will be busy indeed.

On a side note - I checked the Wimac West club rule book, it states:

"It is mandatory to perform all takeoffs and landings from the side of the runway"

;)

There is a no fly zone here too! Neighbouring farm is 800+ feet to the east, and lots of sheep, goats, cows, dogs, cats, rabbits, and I don't want to disturb them in anyway. Only electrics here, including *loud* EDFs (a.k.a. leaf blowers).

[Andre Germain]

Finally rolled out 200' of Nilex woven geo-textile across my farm. Despite having mowed the grass as short as I could, I'll have to wait until next year for it to die and pack down when I'll unhook one side of the runway and pull it taught. Still, I took off my F18 to the east, hair pin back for a perfect landing, first ever on this soil. I was smiling ear to ear. Sun was getting low, bad lipos all around, managed another 8 T/Os and landings, all very good. Surprising how 15' width of runway can force you to nail them. There is one problem with this material however which I haven't seen mentioned on the web, it is highly reflective so the sun glares off of it at you - solution, hop over to the other side! See the photos here.

I'm quite certain this is the only private runway with an astronomical observatory 50' from it! Never mind the barn. I can shoot landings in the morning before going to work, or whenever the weather and winds are good.

I'd like to thank all those in these forums that encouraged me to leave Wimac with my leaf blowers, couldn't have done it without you.

Cheers and farewell, 20 years of Wimac under my belt.

[Andre Germain]

Isn't it always the case? Two years has gone by and despite having a runway 15 seconds from my barn, I've done little flying. Spring 20C weather got me to the runway with the F18, tarmac soggy and undulated, no matter. It's not like WIMAC, with 5 cats, one has to ensure the runway is clear. Chewy stayed by for 15 minutes, occasionally ingressing runway.

[Andre Germain]

And unrelated to this thread, I always hated using my charger as the voltage sensing connectors are hard to reach - eBay to the rescue, 3$ for 4 extenders, some plywood, a 30 amps 12 V supply (under the table) and charging is now a pleasure!

[Andre Germain]

The board of one voted to spend the monies to dig a 200 foot trench for drainage (30$ of diesel in all), as previous years' torrential rains saw a river on the runway with undulating waves from obstacles - quite surprising!. Work took 3 days. No flying activity was reported over the period.

Rumours of an FPV shelter is next.

Happy flying Wimac.

[Andre Germain]

Google finally paid for a set of satellite scans of Wimac West - the jet runway is visible from space!

https://www.google.com/maps/place/22060+Laggan+Glenelg+Rd,+Dalkeith,+ON+K0B+1E0,+Can ada/@45.4459432,-74.5614782,657m/data=!3m1!1e3!4m5!3m4!1s0x4ccebf95ed620015:0xf9b43 6b9a2b04c0b!8m2!3d45.445469!4d-74.5615801

And the property is for sale! We bought a 20 acre farm south of this place which will be more of a challenge for a runway [10 acres of maples], but I'm into quads now, so don't need them wings no more.

[Xavier]

Woaw ! Nice flying site :-)


Google finally paid for a set of satellite scans of Wimac West - the jet runway is visible from space!

https://www.google.com/maps/place/22060+Laggan+Glenelg+Rd,+Dalkeith,+ON+K0B+1E0,+Can ada/@45.4459432,-74.5614782,657m/data=!3m1!1e3!4m5!3m4!1s0x4ccebf95ed620015:0xf9b43 6b9a2b04c0b!8m2!3d45.445469!4d-74.5615801

And the property is for sale! We bought a 20 acres farm south of this place which will be more of a challenge for a runway [10 acres of maples], but I'm into quads now, so don't need them wings no more.

[zorba]

Google finally paid for a set of satellite scans of Wimac West - the jet runway is visible from space!

https://www.google.com/maps/place/22060+Laggan+Glenelg+Rd,+Dalkeith,+ON+K0B+1E0,+Can ada/@45.4459432,-74.5614782,657m/data=!3m1!1e3!4m5!3m4!1s0x4ccebf95ed620015:0xf9b43 6b9a2b04c0b!8m2!3d45.445469!4d-74.5615801

And the property is for sale! We bought a 20 acre farm south of this place which will be more of a challenge for a runway [10 acres of maples], but I'm into quads now, so don't need them wings no more.
You are using the runway for jets?
Don't you think is kinda short at 193ft?
The Wimac runway is 355ft and is still a bit short for jets.

[Andre Germain]

It sure is short! and only 15 feet wide, but you hone your skills when it is this tight. They're EDFs actually, but they still take up the entire runway to stop and most of it to take off.

I had a blast for 3 years and the skies here are better than WIMAC, fewer trees and its 30 feet up from the surrounding farm land, so you've got a better view all around. Of course, if you do drop one off the property, ain't no fun finding it in corn, but knock on wood, I haven't lost one yet.

[Andre Germain]

See below WIMAC versus my installation - both are the same heading, but WIMAC sure is wider and longer! But I sold the farm, pulled up the runway and the grass is already growing! The geo textile proved quite good for electrics, and had stripped the earth bare - when I rolled it up, the ground was black, no sign of life of any kind, yet within a month, grass blades began to pop up here and there.

[Andre Germain]

Although unrelated to this thread, I found the following article quite by chance, an enjoyable read; https://www.thedrive.com/the-war-zone/27604/confessions-of-an-a-6-intruder-pilot

[Andre Germain]

After three years of having bought the 20 acre farm in North Lancaster, and spending 2000$ for an excavator to remove 12 truck loads of stones/boulders from a 150 year old rock fence/pile, and a bulldozer to level the land, I finally put down the runway. You may say it is insane what with trees at both ends, but its the only flat land on the property, and at 180 feet long, 15 feet wide, I fly ducted electrics off of it 'sans probleme'. Arguably the only landing strip with TWO astronomical observatories near by! And that silo, planning on lifting the dome, putting an observation deck, nice place to fly planes!

[Mezri]

Very nice place...congratulations

[ImFM]

Beautiful views, landscape and plans for the tower, cool.

[Andre Germain]

Busy on the farm, finally got around to putting together a drone with a Pixhawk and OpenHD on an RPi3 (AirPi) and RPi4 (GndPi) with Mission Planner.

Call me crazy, but... http://watchobs.com/flight3.html

And to think that I was ahead of the game with my first RPV back in late 80s, wanting to down feed video and such, back when it would have required a miniature generator onboard, but as per the usual, life and CAE got in the way, only nearing retirement will I work it fully, after decades of everyone else having developed what I would have enjoyed doing.

[briankizner]

I remember the old plane well.

[Andre Germain]

And I remember you well Brian! :)

The first RPV with its two 60s on the top with little room for fingers in between the two props flew at the old field in Dorion. I couldn't find any indication the field existed last time I visited the area! Call it nostalgia, first roots, but that field was the best, there was an innocence to the place and time, and such names as Roger, Willis, Yvon and many more marked it.

[Andre Germain]

I was born 40 years too early! it's the golden age of DIY - low cost, fast, CNC, open source.. heck, maybe there's far more to come with A/I.

Anyway, got around to working the 4th RPV in OpenSCAD (code!). Lots more to do, but just as a teaser, it'll have two swiveling props on the front of the wings, and one ducted fan in the rear... why? firstly, I like touch&gos, and its near impossible in FPV, the runway comes at you like you were in the space shuttle, so the flight controller will keep the AOA just below stall when props and fan will handle the remaining lift and control required when well below stall speed. I selected high lift airfoils from online database. It's just too much fun! I was going to use foam, but with LW-PLA you can 3D print these things, but I'll probably need a hybrid in the end. I do however plan on laser cutting plywood for the landing gear and wing root structures for strength and durability.

When I add more components to the CAD, including all the internal parts, I'll post some more.

[Ricardo]

Hi Andre, amazing using openScad for this...can you send the code? I am learning it myself and would like to see how you did it!

[Andre Germain]

Ricardo, attached! I'll post more zips as I work the project further in the next few months.

I am by no means an expert in OpenSCAD, but I get by. There is still plenty more to do. Mounting holes, internal spaces to remove in the wing and fuselage, structural stress members in the wing, etc. The box I added is the plywood that will take the forces of the landing gear and wing. I'd say it's 20% completed, and that's already 10 hours. The advantage with OpenSCAD is that it is parametric, so its easy to change things and of course you can make copies or replace/modify parts. I've been using OpenSCAD for a few years for well over 100 parts by now, mainly for my two astronomical observatories, but rarely have I made such large parts and am thinking of getting a second larger printer, having my own printing farm to speed things up.

What convinced me was most particularly what this guy has achieved, it's phenomenal: https://www.youtube.com/@renerosentraeger/videos and it's also when I found out about LW-PLA. I was going to go old school with foam and balsa skin, but I'm always in for trying cutting edge methods.

[Andre Germain]

Ricardo,

I did some more work that introduces the shell of the fuselage. It has complicated the code much and had me toss the use of polyhedrons for the tail and nose as they can't be 'shelled', at least not easily, with the 'minkowski'. I also put a cutout to show the shell. Lots more to do.

[Andrew Fernie]

Glad to hear you are still active, Andre.

OpenSCAD sounds interesting. I have to try it sometime.

If you are interested in a more traditional 3D modelling program there is a free (for hobbyist use) license available for Fusion360. There are a few limitations:
1. can only have 10 models in an editable state at the same time, but easy enough to switch them between editable and read-only, so never a problem in practice for me.
2. can't save drawings as pdf.
3. can only export to STP

Another option is to get the hobbyist license for Solidworks. It costs about US$100/yr, although you can get it for half price through EAA which means that you get EAA plus Solidworks for about $100/yr. The only hassle I have with it is that its file format can't be opened by a regular-license Solidworks. A non-issue for personal use, but if you want to share your model it is limiting so you really need to export to STP.

Both work fine for me within the limitations listed above. Putting aside the cost and hassle with output format I prefer Solidworks as the constraints between parts ("mates") are much more reliable than the "joints" in Fusion360. They seem to work as I expect, allowing motion within the constraint (or not). With Fusion360 there is a lot of praying involved, and the prayers are not often answered positively. I am forever defining a fixed joint between two parts only to have the two parts separate when I move one of them.

There is also FreeCAD, which works well for part design, but the constraints are implemented through add-on "assembly workbenches" contributed by users and are not officially supported. I haven't spent enough time with them to figure out how they work. FreeCAD is supposed to be able to import SCAD files, although they need to be CSG format rather than the SCAD format you posted.

All three are parametric.

Andrew

[Andre Germain]

Hi Andrew,

thanks for the considerably detailed reply, much appreciated.

My main PC is the beautiful and still effective Acer Iconia 700W slate from 2010, but at dual core 2.66Ghz, both Fusion and FreeCAD are too slow, whereas OpenSCAD works very well. I didn't like FreeCAD and gave it up 3 years ago, and Fusion was frustratingly slow to even give it a chance. I didn't even try SolidWorks for the pricing (did not see the hobbyist fee). I have two other PCs that are very recent and have a lot more horse power, but they each live in their own astronomical observatory - not about to Fusion on those! I'm finding more reasons to upgrade from the slate, but it's hard to part with it, for being so sweet a unit. I dock it and plug it via USB3 into a Wavlink 5K hub driving two large 4K displays, a nice setup - switching over to CAE's laptop takes a few seconds.

About OpenSCAD - it's been very effective for nearly 100 parts, most for the telescopes and observatory, but this RPV is pushing the complexity that is inherent with OpenSCAD as the model gains more components.

Regards
P.S: just bought a very capable laptop :)

[Andrew Fernie]

Time for an upgrade, Andre. PC years are kind of like dog years, but more so!

You mentioned laser cutting some plywood. Do you have your own laser cutter now?

[Andre Germain]

Andrew,

I had optical bench rails I had picked up at Stellafane 20 years ago (an Astrofest in Vermont) which were perfect for putting together a large ultra low cost CNC table - I believe I had sent you a video of it. It can take a Makita router, an Everlast plasma cutter and an 80W laser (10 optical) - I may also add a drag cutter. All three work very well. I cut plywood with the laser, and its best to make multiple passes instead of a single one - the cuts are very good so it'll come in handy with the RPV.

http://watchobs.com/CNC.html

Regards
P.S: the new laptop will be here this Sunday.
P.P.S: Andrew, you're like me, you tinker both at work and home for decades, if only we had what's available today when we were 18! I need another 150 years to do what I want to accomplish...

[Andrew Fernie]

Exactly! When I speak with friends about retirement some say that they have no idea how they will keep themselves busy. Not a problem for me!

[Andre Germain]

For those following, I've changed the OpenSCAD solution so much using far simpler and better method - I'll post it when it's further along. I've been working hollowing out the wings. Where are the ribs you say? it's not easy to print such as the wing is printed vertically along its length, instead one relies on the skin which is actually an inner and outer skin with internal webbing done by the slicer app and printer. Here, ongoing internal view.. more to do.

[Andre Germain]

I got to test out printing LW-PLA, that is low weight PLA, which will expand (foaming) at higher temperatures. When printed at 200F, no expansion occurs, and the layers are easy to separate (not good for R/C), but at 250F with half the extrusion rate, the layers bond so well that I was not able to separate the layer with any reasonable force. You do have to print slower. Certainly looking promising for R/C.

I'm printing the right wing inner panel at the moment - was to be 36 hours on a 0.4mm nozzle, so I switched to 0.8mm nozzle and doubled layer height, its down to 9 hours. Patience!

I've attached yet another OpenSCAD revision, very much different and improved.

[Andrew Fernie]

Hi Andre. I have heard that the LW-PLA can be sanded and painted better than standard PLA. Do you think that is the case?

[Andre Germain]

I stopped the printing after 2 hours because some of the surface was messing up, but otherwise the wing section is very strong, light and shows much promise.

Andrew, when using the expanded temperature range, the surface is matte and appears sand-able. They say its also very paintable, looks likely. I've painted PLA with Tremclad flat black (!) on an optical tunnel that holds a cold mirror for the telescope, and that worked very well - educated guess is this material will paint even better.

I'm so encouraged by the LW-PLA that I'll be purchasing three reels of 1kgs this week (50$ each).

[Andre Germain]

I switched to UtilMaker for more control of the Cura slicer settings and the surface defects are no longer. Printed the largest section I could on a Creality CR-20S, 10" of span, 1/4 of the wing. 180 grams. Balsa + monokote sheeted foam would be lighter, stronger and survive bad landings better, but for certain LW-PLA is fine for fuselage and probably tail surfaces.

[Andrew Fernie]

Hi Andre.
I have built and flown one of these https://www.rcgroups.com/forums/showthread.php?3110758-36-Northern-Pike-3D-Printed-Cuda-Clone-%28It-Flies-STL-s-Posted%21%29 and looked at a bunch of designs from 3dlabprints. They all seem to use thinner skins and more complex interior stiffeners. It seems like that is how they keep the weight under control. Have you considered that type of approach?

[Andre Germain]

Andrew,

yes I have considered it, I'd worry going less than 0.8mm skin as although its quite strong, you could puncture it with relative ease. The skin on those in the link are so thin, egg shells. My inner structure, at the same 0.8mm, quickly adds up and is a significant part of the total weight. See in the photo a cross section of the wing where the rod to connect the wing and fuselage has structure 100mm into the wing to explain that the inner and outer skin are 0.8mm. I still have to work the infill and add more layers around the rod for structural strength. I wanted honey comb, as per military A/C however the extensive transverse ribbing with variable filling circle radii is intriguing in the model you linked - I'll work that in this weekend and try it out (such variable infill are not native to Cura, will need to use a different slicer engine).

I did consider the first reel of LW-PLA for experimentation. The first few runs were to setup the slicers for R/C, followed by structural tests, and then further weight reduction, my flights are mostly touch'n'go's than flight! that punishes the structures. If I can manage VTOL or below stall CTOL, it should lessen such, more reason to shave off weight! 0.6mm nozzle is a logical next step. I'm musing of a hybrid wing where the leading and trailing edges are LW-PLA and the center is foam (skinned).

NEWS FLASH: the slicer had predicted 180gr, but I weighed it at 152gr. I plan for total wing weight under 450gr, so it achievable (the section I printed is ~1/4 of the full wing).

[Andre Germain]

I changed the wing from a skin shell to a single layer skin and used quarter cubic filling - as expected the surface in between the inner supports is softer, but overall the strength across the span and chord, and resistance to twist are improved, and the weight has dropped from 152 to 114 grams! That's quite viable and is under my target weight. Furthermore, as I was using Minkowski smoothing to create the shell and no longer need to, the trailing edge is sharper and thinner, not round as it used to be. Smells like a winner.

[Andre Germain]

Found the data on eSun LW, turns out 260C is just right for 220%.

I bought a 400x400x400mm build volume FDM printer to ease R/C build ups.

[Andre Germain]

Finally got working on the 3D printed FPV gen 4. On sabbatical, so more time to work it.

Took 3 weeks to get anything printing well enough on the larger Mingda as it was often under extruding the expanding PLA-LW despite online tools showing velocity and flow distributions throughout the 9 to 13 hours of gcode was fine. 1 kg of PLA-LW and 20 failed wings and fuselage parts went into the recycling bin. I had a hunch that it was the Ultimaker Cura 5+ that was the cause and switched back to the tried and true Creality 1.2.3 which uses an older Cura and the parts came out fine!

The airframe is close to 1 kg, for a 1.1 meter airframe, its a tad on the heavy side, but the wing is very high lift and it will have vertical thrust component. All parts are PLA-LW except for the control surfaces - I found them too rough and flexible, so they went on the Creality CR20pro as PLA, with 2x the weight as penalty. The PLA-LW right wing was printed with a 0.8mm nozzle for a skin of the same (220g), but on the left wing (175g) and fuselage dropped it to 0.6mm. The next airframe I'll nozzle it at 0.4mm with a skin at same. This would have cut 300 grams from the airframe.

Working the plywood into the airframe for the landing gear and wing spars.

Stay Tuned

[Andre Germain]

Here she is missing only the ESC to Lipo silicon AWG14 wires which I am receiving this Saturday. I have few parts lasered, but those I did on my DIY large CNC table were super well cut so will do far more parts that way in the future.
The motor tilt mechanism are not present, as it will first fly as a normal A/C, followed by adding in the flight controller and fpv gear (OpenHD). Only then will I begin to work in the near-VTOL components.

[Andre Germain]

Got it ready to fire up - amazing torque, tracks nicely. Flying weight is 1.8kg, and the thrust is 1.6kg (measured). Using DriveCalculator with the Turnigy outrunner, 25A ESC and three bladed prop predicted ~800 gr of static thrust each, so right on the money. VTOL is certainly possible once I throw a 55mm fan in the tail.

First flight as soon as weather permits, and to be safe, I'll do it from the asphalted country road with a spotter rather than my short yard runway with trees at both ends.

[Dwight Macdonald]

Nice work Andre! What is the tail fan for?

[Andre Germain]

Dwight!
thanks.
The goal is below stall landing speed to render a 'less' exciting fpv experience! The two props will eventually swivel upward and a lift fan at the back will help with lift, pitch and yaw control. But since there's plenty of thrust, I'll venture into the full VTOL realm as well. It's a long term project, and there will be variant of the A/C as I go. I've been interested in RPVs since the 80s, but my job got in the way! no time...

[Andre Germain]

Yesterday I high speed taxi tested the FPV4 on an asphalted country road - very useful. Was very squirrely, so reduced the NWS throw and switched to using exponential on NWS as well. Tried a few hops (in ground effect), pitch looked nice, but couldn't go for a flight as one aileron servo seized up.

Today, after addressing a few other problems, I took it for a flight on my runway and oh no! she was tail heavy and due to the short tail and diminutive horizontal stabilizer surfaces it was exceedingly hard to keep her airborne - pitch deviated continuously up and down. I thought about dropping her gently into the soy field, but elected to try the runway - bad decision... ack, too low, can't see it behind the lone tree at the east end of the runway, I let go of the sticks for 3 seconds hoping.. but alas, when I saw her again she was diving for the ground at 70 degrees nose down - WHAM. Total loss of the hull.

I was surprised how the fuselage and wings broke apart for PLA-LW - not at all as bad as I had thought. A balsa airplane would have been worst off, only a foam skinned A/C would have been better off. One of the prop alum spinner assembly sheared off, an indication of the force of impact.

So back to the design.. to lengthen the tail and nose to address pitch stability, control over CG position and better yaw tracking on ground, will also print thinner for lower weight, increase wing span, use less plywood, switch to Dubro hinges, no more carbon fiber spar - far too much work to include them, instead the wing will be two outer panels and a center wing box with fuselage sitting on it....

Stay tuned.. shouldn't take more than two weeks.

[Dwight Macdonald]

To bad you can't do a simulation flight on the design before printing and flying, to check the weight and balance.

Sent from my Pixel 6 using Tapatalk

[Andre Germain]

Oh but I did think about it, and there are freeware CFD out there, and assumed it was a steep learning curve, and there is the question of Reynold number, as the CFDs I saw were tailored to full size A/C. Perhaps I should have looked into it further. Truth is, I've designed and built quite a few original A/C, and have generally been successful, but with this one, I was worried about a few points, and sure enough, they came to be (a problem). Also, if that tree wasn't there, I would have landed her and resolved the problem, but the smarter thing would have been to gently let her into the soy field, the plants are at their maximum height at the moment.

[Andre Germain]

Completed the upgrades, should start printing Friday.

1) Tail and nose lengthened (longer nose more effective in CG control [less or no ballast])
2) Nose gear further forward
3) Elevator is single piece, constant chord, thus one servo
4) Horizontal stabilizer is a single piece, sitting on fuselage
5) Wing is two outer panels and a central wing box
6) Fuselage sits on center wing box - no spars
7) Wing increased from 44 to 48 inches and constant chord
8) Rear fan are pop out panels (waiting for when the fan will be installed)
9) Single tunnel for wing wiring (ESC, servo) - used to be two hollow carbon spars each wing

Lift area has increased by 20%, and printing at 0.4 iso 0.6mm nozzle will reduce the weight despite the dimensional size increase of the A/C. The build up will also be far easier as no spars, parts sit on others, less need of plywood and epoxy, ect.

[Michael O'Bree]

A fascinating project and thread…… look forward to the continuing story with other experts chiming in.:cool:

[Mezri]

Very exciting project...

I dont have big experience with ''inverted wing-profile elevator'', but I can tell that the response is not linear like a flat profile or at least symetrical-profile (9% thikness)...With 2 motors blowing the elevator, the down force might vary a bit with the thrust.

The CG might be reviewed forward (28-25%) ...

The thrust of the motors is a bit low...to be able to hover, the thrust recommended is 1.5 to 2 times the total weight...the 2 motors are too close to the CG, so they might do the hole job. The position of the fan at the back will force to get the lightest one in the market :D...If you permit, my predictions of the total weight (Fan+Auto pilot+etc) 2.5 to 3 Kg...For that I think it is better having a 2x2Kg thrust motors and the lightest 60mm Fan...

For the 2 motors, Nower days you can find easy the thrust needed with the minimum weight using FPV mtors. The ''T-Motor'' F80 or F100 are powerfull...and the advantage that you can buy a CW and CCW threads, so your props will never loosen...Luckyly your prop was loose with the crach, otherwise it will loosen slowly with the vibrations and fly away...

Very good Project...Please keep posting for a successfull project. :)

[Mezri]

I just tried the thrust in calculator Ecalc,
The T-motor F80, 2500Kv, 39gr,
4Cell lipo, 50 or 60A ESC,
6x3 APC prop will give a thrust of 1.4kg under 38Amps...

And With The F100, 1350Kv, 66gr,
4Cell lipo, 60Amp+ ESC,
9x4 APC prop, 2.1kg under 49Amps...

Sorry, thoses ones dont have CW and CCW treads, but have a grip prop plate and Nylock nut :D

[Andre Germain]

Hi Mezri,

it was originally not intended to be VTOL, only to land at say 60% of stall speed. I fully expected VTOL performance to be marginal, but exploring it will provide much information.

About non symmetrical and under camber on horizontal stab - been doing it for 30 years, works great. The lift is of course more pronounced, and generally the center of pressure moves more forward with higher AOA, but on a horizontal stab, it's less problematic than on the wing - this one is a classic Clark Y. The airfoil of the wing is the CH10 with over twice the lift of a typical symmetrical wing, and has quite strong CP movement, which certainly didn't help when I was reticent to put more lead in the nose for the first flight (CG margins) - my bad. But hey, much of the reason for doing this is exploring, and today's hardware makes it a snap - wish we had the same back in mid 80s when I made my first RPV (it flew video NTSC down feed and 8mm film), twin 60s, 10 foot wing span.

I'm printing parts for the RPV4.1 now, Prusa slicer instead of Ultimaker (Cura), as usual, trial and tests are in order.

As for the motors you suggest, the current requirements are way beyond what I have presently in this A/C. But they would be more suitable for my next project: two identical wings slapped onto a fuselage fore and aft, with two pivoting motors on each wing - that'll be VTOL for certain. A quad copter with wings essentially.

Regards
P.S: before EDF foamies, the prop in the nose jets I built were too heavy to fly with near scale wings, so I dropped the trailing and leading edges permanently, which made them fly more safely, but of course forget aerobatics, but I've always been a circuit kinda guy (even on PPL!). The airfoil looked like an F18 in landing configuration, so cool! An added bonus is drag was quite high, which allowed nailing landings as throttling was all you needed. I have a YouTube video showing us doing touch and gos on the taxi way at Wimac (no one was there except the two of us). Practically impossible to do with any other fixed wing A/C: (27s mark) https://www.youtube.com/watch?v=cNKb8nfMkOo

[Mezri]

Wow, Very nice landings...:cool:...The carrier landings are really cool...
Waiting for more posts on this project...:)

[Andre Germain]

The second version (modified) is ready, but can't fly as missing one spinner from Amazon.

The earlier version was 1.1m at 1.8kg printed at 0.6mm, this one is 1.2m at 1.6kg printed at 0.4mm. The nacelles are shorter as I had noticed vibrations were exciting them.

IMHO 0.4mm skin is fine for a park flyer, but not a workhorse that will live punishing landings in its lifetime. I'm likely to bump it up to 0.5mm and move to using wing crisscrossing hollowed out ribs so it can be printed in vase mode, thereby ridding of PLA-LW extra oozing which can't be avoided in its expanding mode (retraction is of no help), and do away with voids (travel), e.g: one continuous print. I doubt it will save weight over cubic infill as the ribs are going to be two nozzle wide whereas infill is only one. It's a lot more work to CAD such a wing (and fuselage, tails), a winter project! I'm trying out FreeCAD as after looking up the tutorials it's not as bad as I had first thought, and with the huge developer base, its likely there will be a plug in to do such R/C methods. OpenSCAD I'm very fast with, and although I can even put out such a vase printed wing, it's better for simpler parts.

Stay tuned for the 1st flight.

P.S: what's with landing gears, wheels, hinges and threaded pushrods on Amazon? very expensive. I got a roll of TPU to try my hand at making tires - I find that square profile tires are horrible for tracking on my runway, so need rounded profiles.

[Andre Germain]

1) Got the spinner hubs...rats! Turnigy outrunner shafts are 1/8", not 3mm as on the drawing that came with the motors! Fixed on the lathe.

2) Taxi tests... bumpy runway... and... the right wing cracked open on top and sagged. Ack! I knew 0.4mm was thin and the PLA-LW layers hadn't bonded well (ongoing slicer and printer issues).

I tore off the wings in the shop, and went back into OpenScad [OS] to draw up internal ribs for vase mode.. not finished yet (see image)... why OS? because the entire project is there, so for this drone, it will remain in OS.

[Andre Germain]

Reworked the RPV with more wing area, CG 1" further forward, flew it... ACK... severely out of trim roll + pitch, fought her.. got roll under control, insufficient pitch, had to hold full stick, brought it onto approach, 6 seconds to go, was looking stable, power was on, and it just dove suddenly to the ground 30 feet before the threshold, what the...

It behaved somewhat like the last version, which got me thinking this new FlySky radio might have a range problem... going to test it fully before the next flight.

I managed to save some of the parts, re-worked the CAD - nearly ready to fly.

About the wing, after weeks of testing with the printer, I've got a decent vase design, but two layers (nozzle never travels in a layer, continuous extrusion!), so much stronger at a reasonable weight (less than using cubic infill) on 0.4mm nozzle, so there is next to no stringing in the print. See photo. Very easy to pass servo and ESC wires as opposed to spars or printed ducts. My first PLA-LW 16" wing sections came out at 230g, then followed 150g, 120g and now 100g. My next will likely hit 75g! without significant loss of stiffness.

The spars are 10mm carbon fiber tubes, but the main and nose gear are light plywood, with plywood structures connecting the two as a PLA fuselage would never handle the punishment of repeated hard landings. The strength of epoxied light ply into/onto PLA is astonishing! I wasn't able to pull them out to test their strength, and I applied considerable force in doing so.

[Andre Germain]

Day 1: With a longer tail, 3 degrees more horizontal stabilizer incident, it's first flight was closer to trim on a gusty cross windy evening - but pitch was not neutrally stable, rather mildly unstable making for a trying first flight on such a windy evening - I did make it back onto the runway on the third approach. I'll have to revise the CG, as my first attempt to researching the CP of the CH10 airfoil was unsuccessful. The FlySky range was checked with RSSI (OpenTx) and was nominal.

Day 2: I just checked the Cm vs Alpha for the CH10 and Clark Y at Reynold number of 200K. The Clark Y has a progressive change whereas the CH10 is mostly flat (in the AOA range for R/C), which shows why the Clark Y has better longitudinal static stability (with same tail configuration) - but I thought the Cm vs AOA was supposed to have a negative slope for long static stab... probably because these are the Cm of the wings only, not of the wing and stab, nonetheless the trends are different enough that it could be the cause - more analysis to come.

Day 3: moved CG fwd 1/4", flew in mild wind, better... spent less of my abilities fighting the pitch divergence, still present but bearable. As it is, it's straddling the intermediate to expert pilot required, not appropriate for an RPV/FPV, but I'll throw the flight controller into it for the next flight to let it handle that aspect of flight. It also has poor spiral tendency and certainly needs dihedral (my OpenScad design has 3 degrees per wing, but in printing it I lost it because the instruction to remove the spar material was not tilted in the wing part whereas it was in the fuselage to compensate for dihedral, but it's the former that would end up in the print.. I digress - ack!)

Day 4: Installed the flight controller and set stabilize flight mode on a transmitter switch. Took off, flicked it on, what a strange sensation - sure enough it kept it level in pitch and roll and damped the yaw. Strange because I needed to use considerable stick deflections to take over, and then all the bad tendencies of the airplane would show up. However, approaches to the runway were rock steady, I mainly controlled thrust and nailed the center of the 17' wide runway with hardly an impact, surreal, 3 times out of 3. Despite that, I doubt I'll bother with adding the ducted fan in the tail to attempt below stall approaches - I rather use an inherently stable airframe to begin with. See attached flight path - may seem messy but it's the large tree at the end of the runway that blocks my view so I cross over my head, then do a circuit over the house and eastern farm field.

Day 5: Installing the AirPi + HD camera, and recommissioning the ground station - this means telemetry + recorded video. Flight will still be by radio sticks, but am preparing USB controls via RPi zero (Python) to PPM of the radio (OpenTx). Stay tuned!

All in all, it's been months of trial and error, gaining experience with 3D printed airplanes and more. I have to conclude that heavily cambered airfoils are not the best choice for powered aircraft for a full flight envelope. I'll design and print RPV5.0 this winter, a canard with two swiveling props per wing, a quad with wings. I'll also read up on aerodynamics of aircraft (already begun) and use CFD tools such as xflr5. I'd done fluid mechanics in Physics at McGill back in 1984, but that was for cylinders, cubes and the like in theoretical flow fields, Aced it though, loved it too.

[Andre Germain]

With the CH10 highly cambered wing having very poor pitch stability, when I added another 200gr of RPi Zero and Wifi dongle to allow for OpenHD video/Mavlink, I knew I was at the limit.. took off, had the flight controller on stabilize mode, and at about 50 feet above ground climbing out, it stalled, I recovered over the roof of the house, it stalled again, I recovered, and continued the flight as best as I could and landed it safely. The telemetry showed that the AOA was swinging high and even the flight controller moving the aileron would cause a wing to stall.

...I cut the wings off, printed a set of Clark Y airfoils, stuck them on, and finally flew it today. It flew uneventfully, proving the highly cambered CH10 wing was not well suited for powered drones. The landing was decent although fast due to the weight.

[Andre Germain]

Since spring I've spent so many hours perfecting 3D printed wings suitable for a tough tricycle job, and I conclude it's not adequate, it doesn't come close to what can be had with foam or balsa.

I was also wanting to try hybrids, 1) leading & trailing edge 3D printed with a central foam section, 2) thin skin of foam with 3D printed ribs and main spar.

To test such, I reassembled a DIY 4 axis CNC foam hot wire cutter I had made back in 2010 and in the same afternoon cut out a Clark Y wing section. Frankly, I had forgotten how good it could be, the wing is strong, needs no leading edge or even a spar (although both are easy to add), weighs less than half the 3D printed wing and is more resilient. It also took 5 minutes to cut as opposed to 26 hours and cost less than a dollar as opposed to 25$. I won't bother skinning it with balsa or econokote as I'm not looking for beauty, distance, and such - it's good enough for RPV testing of all sorts.

The photos show the two X-Y towers. The wire is quite high so the foam board needs to sit on an elevated platform, here a folding table, but I'll build something more appropriate.

[Mezri]

Nice job...old school is still viable :)...here is same with my Super cub wing...you still can save weight and make ailerons and flapls even better...You will notice that the gaps on the bottom, when hollow out the wing, will be a nice fit for a thin spares.

[Andre Germain]

Hi Mezri,

yup I did a lot of foam wings the old school way! my first was in 1984. Back in 2010 when I built the 4axis CNC, I used TJZoid which even today works very well - makes nacelles, fuze and wings. I'm currently working on cutting a shell, and will add a foam spar and ribs to try out that technique. Other methods are coming.

I still prefer the simple cut as it's so easy to make another if one is damaged, whereas ribbed wings takes a lot more work. Hollowing out isn't so good for smaller wings as you lose a lot of strength - even though most of the mechanical force comes about from the surface material, I suspect it's because the foam is composed of cells, and small wings hollowed out leaves far fewer of these and their binding force. However, if one skins the foam, it's no longer an issue of course.

Regards
P.S: nice work and nice setup! and I use machine shop blocks too for weighing it down!
P.P.S: when I used a bow, I also spatially marked my templates, but I used cardboard, as I didn't have a bandsaw at the time. Used needle to hold it onto the foam, often we were two on each side, no bow, just pulling on the wire as we went.

[Mezri]

Yes You are right that the skin give it the needed strength ..but in my case, to make my plane light, it is almost no skin :D ...the thin spares with a smal carbon tubes were enough. the wings were maintained by a fonctional struts.

[Andre Germain]

Mezri,

beautiful work - I don't have that patience anymore. Back in the early 90s, I made a Tomcat and an F18 from scratch all in balsa: bulkheads, stringers, skinned with 1/32", swings wings on the F14 and tail hooks on both. They were very light and beautiful, but I could never do that again. Sadly, oil infiltrated them over time, and when I stored them in the barn, a mouse dug into the F18 and lived in it over the winter.

Anyhow, my first draft of the canard. I already foam cut the wings, vertical tails and canard, glued them up. Working the fuselage to print 3D and will do glide tests. Finally it will be three outrunners, one large in the tail, two on the canard, swiveled. I had originally wanted two motors on each wing and swept the wing to move the CP back, but now that I switched to 3 motors, I don't need to sweep the wing - ack!

Regards
P.S: I gave up on using foam shell with ribs: as I suspected, at these small chord lengths, foam is far too weak unless its whole, or so my experimentation showed yesterday.
P.P.S: why the wing and canard on top? ease of construction! they're both flat bottomed airfoils, so no need for complex fuselage mating, and can use dowel and elastics. Indeed, having the canard lower lessens its turbulence on the main wing, but then it leaves less space for prop ground clearance.

[Mezri]

Nice project...keep us posted on the tests...May be one day a canard scale project : Long-EZ.
Thank You Very Much for sharing the evolution of the project.

[Andre Germain]

There are of course so many canards, but despite this they aren't wide spread in use, for reasons you can easily Google. This one caught my eye and although I won't be placing the props on poles as this one, it's basically the configuration I'm adopting.

https://robbreport.com/wp-content/uploads/2021/01/1-13.jpg?w=1000

[Andre Germain]

I printed the fuselage and stuck the wing and canard on it with rubber bands and threw it around the yard. Quite pleased! tracks well, stall is benign, although it flattens out and drops gently flat. Not worried as adding a few degrees of incidence to the canard will ensure a nose down tendency at stall. I didn't even bother installing the top of the fuselage.

I'm going to cut full span elevons into the wing and attach a rear facing outrunner to do hand launched flight testing in the coming weeks. I've also changed the fuselage somewhat and will instead wire cut it in foam, also will switch to a single taller vertical stab. Foam is 40% of the weight for similar strength, just not worth continuing with PLA-LW.

:)

[Andre Germain]

Test flew it, very nice. Stalls gently, flat, easily recovered with power. Will continue modifying it - it's a test platform for the final version 5 for next spring. Flying weight of this hand launch is 2 lbs.

https://youtube.com/shorts/KXbdCnoBbIQ?feature=share

[Mezri]

Very Nice and stedy :)

[Andre Germain]

Before I hung up the RPV4.3 for good (100% 3D printed), I flew it on a very gusty day with no stabilization and landed it while a combine was passing right next to my runway - I wasn't about to do a go around and startle my neighbour in his combine, so I was committed. And yes, I'm not lined up because I fear that plane eating tree. See YouTube video link (Rpi v2 camera downlinked via OpenHD).

https://youtu.be/aM7UEeMtK9k

To give you an idea of the size of that John Deere monster (a factory on wheels): 24 metric tons, 35 feet wide, automated (follows GPS assisted tracks), internet connected... all for over 500,000$

I also spun out an RPV5.2 test airframe (see photo), quick and dirty, to test CGs and such. I reused the wing and canard, but CNC cut foam fuse parts and a single tall rudder. Flew it under mostly gray skies with moderate but severely turbulent wind. It flew well at 800 gr (1.2 meter wing), but coupling between yaw and roll was excessive as it rocked about. I conjecture the swept wings, dihedral and short tail moment are responsible - I'll work that out on RPV 5.3!

The sun poked out at times, surprisingly this RPV wanted to be a sail plane, it kept aloft with zero power for minutes, hardly losing altitude - I had to push her down otherwise.

The saga: https://watchobs.com/flight3.html

[Kamille Nielsen]

Nice project...

[Andre Germain]

Had a rather calm day to test the RPV5.2 further - the roll/yaw coupling is obvious. Removal of dihedral will reduce it as will increasing vertical stab moment. I also added thin aluminum extensions on the elevon to give more control - much nicer!

I enjoyed flying it around the property, even though it's meant to hone a better design. I've been flying EDFs for so long, I miss props, so much less stressful, getting more than 5 mins per battery at that. Have a look.

https://youtube.com/shorts/nYuhk0AkKUA

https://youtu.be/yG0xqBBVFB4

https://youtube.com/shorts/xuwu8Wc43sI

The 'dutch roll' will be corrected when I remove the dihedral and increase yaw moment.

[Andre Germain]

When I bought the incredible Bambu X1 Carbon 3D printer, the Creality CR-20S pro I had on hand was no longer required. I took it apart and used some of its parts to re-do my DIY foam cutter as it was originally using kitchen drawer bearing slides which wobbled about 1 mm, leading to a less ideal surface on parts. With the extruded rails, and lots of new green and black Bambu printed part, the wobble is no more. Another bonus is the minimum wire height is lower, with more travel on the horizontal axis than the original wood carriage. I also am trying out a weight constant tension wire system (see black part with cone, tubing and pulley) as compared to the original spring tension which varied too much.

[pierre sarault]

Andre, Bravo for the experiments in Design and build on your posts. A WIMAC member for years I've designed a few of my planes and thought 3D CAD was my answer to several hand-building and time-consuming limitations as well as the limited ability to easily replicate the designs.
Thanks, you have shown with your work and the posted videos what can and can't be done to a great extent.

I just got a Flashforge 3D printer from a friend (who designs components and products with Solidworks for a Company) and I was planning to subscribe to SolidWorks or Fusion 360 and start designing using 3D. In an earlier post you described that thus stuff should have happened lots of years ago...totally agree as in the 70's and 80's I was designing and building my own 18 ft racing sailboats (as models for larger projects)...life is too slow and too fast at the same time!
Andrew Fernie's posts/ replies to yours are also appreciated interesting.

Thank-You much! Pierre Sarault MAAC 60547

[Andre Germain]

Merci Pierre :)

Although it's going to be quiet for winter - my shop is a garage separate from the house, not heated unless I fire up an oil furnace. I promise lots more in spring.

Regards
P.S: Gosh, I don't remember my MAAC #, joined MAAC & WIMAC circa 1984, and haven't been a club member since 2015 when I moved out to Eastern Ontario on farm acreage.

[Andre Germain]

blush... my MAAC # is displayed with everyone of my posts.

[Andre Germain]

Tested out my new Bambu X1C with AMS on printing PLA-LW and TPU. The PLA-LW came out much more clean and solid, however that won't address the 2 to 3x heavier parts than when done with foam - still, interesting how a premium FDM can perform. Also tried out TPU rubber for an R/C wheel with the hub in PLA - not bad!

[Andre Germain]

Sliced the wing to remove the dihedral and flew it today - lovely! no more Dutch roll, only a hint of it near stall. No video - wife didn't want to brave the cold and mist :(

On the CNC foam cutter, moved from spring to hanging weight (over the side of one tower by slotted pulley), constant wire tension, perfect - bye bye springs. Also tried a 0.8mm iso 0.4mm nichrome wire, worked but a tad too stiff and harder to get kinks out of it.

Amazon'ed a scissor lift for foam stock height control - see attached.

[Andre Germain]

Downloaded an OpenSCAD propeller file - adjusted sweep and chord, and printed a 7x4.5 3 bladed prop in PLA at 0.1mm layer, 100% infill (rectilinear), threw it onto the test canard airplane, and it gave out more thrust than the 2 bladed 7x5. It vibrated some so needs to be further sanded while testing it on a prop balancer.

The prop is flexible yet stiff, similarly to the prop I had on the canard. I wouldn't recommend PLA, better to use PETG, PLA-CF or nylon - I'm waiting for those spools. Even better is to use a resin printer as you can go far smaller layers, but the strength will likely be less than with an FDM.

Now I would not recommend doing such for gas or nitro jobs, but small electrics are probably a good place to test - wear safety glasses just in case!

I'm not too worried about efficiency as one can sand, and if you print with the front of the prop up, it's the back side that is less smooth due to necessary support structure where its less critical. My main worry is the root at the hub for strength, but by increasing the cord thickness (within reason) and adding fillets there, it mitigates.

Regards

[Andre Germain]

What the heck!

https://youtube.com/shorts/jz75qsnxEc4?feature=share

Hilarious bit not video'ed... I was belly landing it on the snow covered raised septic field and didn't like that I was going to overshoot a tad, so advanced the throttle... the belly contacted the snow for 5 feet and airborne it went... a Snow and Go!

[Andre Germain]

Anyone got a dead (or otherwise) Turnigy 9x transmitter they don't need anymore? Mine needs a replacement throttle potentiometer (brass shaft bent).

Regards

[Andre Germain]

I sliced the wing and glued it on the bottom sides of the fuselage. As expected, roll authority is improved as the wing is about at the CG vertically. Pitch trim changed requiring lots up - the canard downwash is reaching the wing now so likely the cause (and a tad reduction of main wing lift). Pitch authority is sizably improved - that is unexpected.

She flies quite nicely now, despite a gusty day, the video was a very pleasant flight - I even flew her down on the decks (not shown) between my position and the obstacles on my property (trees, house, etc). Every change I did to the airframe has improved it such that I'm more certain of the final design for next spring's RPV.

https://youtu.be/DnnnE4olrc0

Funny thing happened on the first throw of the day - the wing being lower, I hit the back of my head when I tossed it, and down it went to the ground - no damage. I thought I should check controls when I readied to throw it again but too excited, threw it... no elevator response, vlam into the ground, canard broke in two, left wing snapped off pulling a large chunk of fuselage. Both elevons wires had pulled out. Argh. Taped/glued it back together, 1 hour later I flew the video.

[Andre Germain]

I got to try out the modified (low wing) RPV in the yard again, many flights on many warm December days.

One of the best airplanes I've flown: docile, maneuverable, predictable, easily flown low and slow (~10 feet AGL) around you, benign straight and flat stall. It was easy to keep it on the edge of the stall, hardly moving forward and descend from 300 feet to landing almost vertically, surreal. At high power I jinked it around obstacles easily, showing its unusual but graceful swept wing and canard.

My love affair with conventional A/C is over!

Such a joy to fly.

I should kit it and start a business :)

[Andre Germain]

Took a while, here's RPV 5.4 ready to fly, 1.8kgs. CNC foam fuse + wings, 3D printed PLA-LW elevons, rudder & nosecone. Once flight trials are over, I'll install the 'avionics' and retest the flight controller then move to fpv HD. Some notables:
I coded an RPi for joysticks to put out PPM to the trainer radio port, complete with trim switches, using A320 stick and 3 lever throttle unit, very smooth and fast response.

New hinge design, laser cut plywood such that when slid into slots in the 3D printed control surfaces, I slide a pin to lock it in place - all circular shapes to boot.

To strengthen the foam fuze, I glued laminated foam boards inside from the dollar store, adds great stiffness and resistance to twist, no need for plywood (although ply is better).

When all that's done, I'll move to vertical thrust: I dropped the design of tilting all motors, will only tilt the aft, whereas the two forward out and in front on the canards will be permanently pointing up (haven't installed them of course).

It actual already flew - I was doing taxi test and it rocketed up, clearly tail heavy, it swung up and down while I nursed the controls and thrust to prevent it going up and to attempt a gentle ground contact... 120 feet later and never getting higher than 10 feet, it nosed in a tad hard and snapped the nose off. A couple of hours later, all fixed. But to lessen the weight I'd have to add, I sliced off 2.5" from each canard.

Reminds me of the hand thrown precursor - in May I threw it much too hard and the battery slipped aft - it nosed straight up to about 25 feet, did a somersault and slid back tail first... just as it was about to hit the ground, I max thrusted it and it impacted tail first will less force than otherwise and caused minor damage. I laughed it was such a crazy gyration.

[Andre Germain]

...better this time, still tail heavy, and the gear is wanting.

I moved the main gear forward to lessen nose wheel loading, stiffened the steering arm. Will shave off another 1 inch of the canards, and add a spar across the wings as the flexing was probably causing pitch issues as lift would vary with the flexing coupling to pitch.

[Andrew Fernie]

Thanks for the update, Andre. It is great to see someone designing and building.

I really like canard designs, and think of the Rutan Ez series as the classic small aircraft example, and the Piaggio P180 as the classic somewhat larger example.

I have another one for you to try. I went to the AeroMontreal conference in May and there were some exhibits, including a radio controlled reduced scale flight test vehicle of the Horizon Aircraft Cavorite X5. They are based in Lindsay, Ontario, NE of Toronto. It is an eVTOL for takeoff and landing, but flies on the wing in cruise.

The main wing is forward swept, canards are swept back. The control surfaces at the back are interesting - looks like that between the two slanted rudders and two elevators between the rudders you would get pitch, yaw, and some roll. There are also control surfaces on the wings and canards.

The really interesting part is the eVTOL mode in which the front and rear halves of the wing and canard skins move apart to reveal a series of lift fans. Lots more images if you go to their site.

It would be complicated, but if anyone is up to building a model with a configuration like that, it is you!

p.s. the airframe is carbon fibre and looked amazing.

[Andre Germain]

Thank you Andrew, most interesting. I'll study it some more.

In my case, I'm trying to integrate and perfect too many technologies almost at once: OpenHD, Mission, Flight controllers, USB stick to ppm, foam constructions (CNC), 3D printed parts, the list goes on, so it's a given things will break. Take this past weekend - I sliced off another 1" from each canard, and it flew better, but still had thrust to pitch coupling, most undesirable to piloting - one wants sudden thrust changes to have no immediate effect on pitch. This occurred on the EDF Phantom (greatly), the twin boomed pusher RPV, and this latest RPV. You simply don't get this when a dirty prop on the nose blows over wing and tail, but the three mentioned blows (or pulls) air mainly over one surface - the Phantom was an interesting case as it blew forcefully air right under the stabilator, causing enormous pitch moments.

So, this weekend on final approach, nursing that pitch moment, 5 feet from the ground it plunged into it, and I heard a loud prop slicing into something. Fortunately stuff broke off cleanly, easy to fix. But the vertical fin had detached and the rudder was sliced through - I suspect the fin tore off (it's a thick airfoil) contacting the prop and ensuing plunge.

I'm switching to shorter chord longer span canard as per the last design (since they got quite short by slicing for moving the CP back), and using epoxy for the fin to fuse instead of UHU.

I'm rather vexed as the last canard flew beautifully, but this one requires some piloting skills.

So yes, the job is slowing progress on the RPV! BTW, my colleagues are working the P180... I didn't force my position to work it too :)

[MichaelLevy]

I fixed the F-4 pitch-up issue easily by angling down the exhaust tubes about 6-7 degrees. No more pitch up when applying power on take-off and during flight.

[Andre Germain]

Brilliant!

Mixing throttle stick to pitch works too (as I did), but the nozzle skew has the benefit that it compensate using the real world air velocity. The techno geek would probably throw in a rate sensor, and a simple flight controller to accomplish the correction, but again building into the airframe the necessary compensation is best.

Back in 1992, I worked with an ex RAF pilot that flew the F4 when we both were involved with the F4 simulator in Neuburg Germany. When I flew the F4 foamie, my thoughts kept going back to my discussions with him - he never indicated there was any such tendency in the airplane, in fact he liked the F4 tremendously and said that it flew quite graciously despite its husky and heavy look. I just checked Google images for F4 nozzles, and quite a few show descending nozzles! I had originally thought it might just be a Reynolds number affair of foamie vs real McCoy.

[MichaelLevy]

I can't take credit for inventing the nozzle mod,it is described in the RC Group forum. I got tired of having my F-4 pitching up and did some research. As Pikestaff commented on post 641 "Do the mod, it is needed. Period"..."And it is scale."
https://www.rcgroups.com/forums/showthread.php?1179826-Nitro-Planes-FMS-AirField-70MM-F-4-EDF-Jet/page43

[Andrew Fernie]

Hi Andre,
I noticed you mentioned OpenHD. I am putting together a RubyFPV system. I got the firmware loaded this weekend and have a 720P 30FPS image coming through nicely. RPi4 and WiFi card on the ground and the camera plus WiFi card on the aircraft. Now I need to get the telemetry from the flight controller into the camera. It will be good to compare notes on the two systems.
Andrew

Thank you Andrew, most interesting. I'll study it some more.

In my case, I'm trying to integrate and perfect too many technologies almost at once: OpenHD, Mission, Flight controllers, USB stick to ppm, foam constructions (CNC), 3D printed parts, the list goes on, so it's a given things will break. Take this past weekend - I sliced off another 1" from each canard, and it flew better, but still had thrust to pitch coupling, most undesirable to piloting - one wants sudden thrust changes to have no immediate effect on pitch. This occurred on the EDF Phantom (greatly), the twin boomed pusher RPV, and this latest RPV. You simply don't get this when a dirty prop on the nose blows over wing and tail, but the three mentioned blows (or pulls) air mainly over one surface - the Phantom was an interesting case as it blew forcefully air right under the stabilator, causing enormous pitch moments.

So, this weekend on final approach, nursing that pitch moment, 5 feet from the ground it plunged into it, and I heard a loud prop slicing into something. Fortunately stuff broke off cleanly, easy to fix. But the vertical fin had detached and the rudder was sliced through - I suspect the fin tore off (it's a thick airfoil) contacting the prop and ensuing plunge.

I'm switching to shorter chord longer span canard as per the last design (since they got quite short by slicing for moving the CP back), and using epoxy for the fin to fuse instead of UHU.

I'm rather vexed as the last canard flew beautifully, but this one requires some piloting skills.

So yes, the job is slowing progress on the RPV! BTW, my colleagues are working the P180... I didn't force my position to work it too :)

[Andre Germain]

Hi Andrew,

I looked into this 4 years ago, and bought a year later. I went simple, OpenHD of course, Air Pi 3a+, Gnd Pi4 8GB with monitor (video + OSD), Pixhawk, Windows 10 slate with Mission. Mavlink of course. I had it flying in various modes. I upgraded the camera this past year, far better, but haven't tested in flight yet. I prefer capturing video on the Gnd Pi RAM, hence 8GB otherwise it's too short recording time.

OpenHD evo is a lot better than the previous version, particularly the joystick has much less latency and more frames per second, but lacking trim, mixing and gain, I opted for an Rpi I coded and stick + thrust levers. I programmed trims and other nifty features, ppm out to the FlySky OpenTx transmitter. What I'm looking forward to is the next gen OpenHD on their very own PCB.

Peter Scheer also put a system together recently, we can converse on the subject. He will eventually come and test it out at my farm.

Regards

[MichaelLevy]

Brilliant!

Mixing throttle stick to pitch works too (as I did), but the nozzle skew has the benefit that it compensate using the real world air velocity. The techno geek would probably throw in a rate sensor, and a simple flight controller to accomplish the correction, but again building into the airframe the necessary compensation is best.

Someone answered you on RCGroup. In his opinion,the thrust down is more advantageous than the transmitter throttle/pitch mix. I don't know if your F-4 is still flyable. Mine is, the only issues I have with it, I don't know if it's related to the throttle mod, is the excessive banking on turns translating in altitude loss in low speed tight turns on landing, so I keep the circuit wide to prevent the wing from dropping. I don't have rudder installed but it would definitely benefit by opposite rudder mix.
https://www.rcgroups.com/forums/showthread.php?1179826-Nitro-Planes-FMS-AirField-70MM-F-4-EDF-Jet/page123

[Andre Germain]

My F4 is alive and well, as is the F14. F18 was retired having flown far too much. Despite being one of the few flyers to have a runway in the yard, at two homes to boot, I get very little flying done as too busy with chores.

I don't recall a particularly annoying tendency on the F4 when banking, but then all these tiny winged airplanes have less than ideal tendencies we learn to work with. What I do recall is the thrust pitch coupling prevents one from landing it well, and since I've been using mixing, I look forward to diverting the nozzles down to see if that improves landings.

I won't bother replying on RCGroup - he sounds far too intense, no interest in yet another meaningless internet exchange - I've flown plenty of airplanes with heavily off axis thrust lines and blown surfaces to prove that the latter is at least as significant, if not more.