wing loading, span, flying speed, landing speed, charts.

[OwenN]

I had some thoughts on benchmarking the model performance:

1) does it have a good flying attitude in the air and make smooth flying progress?

2) Can it flip, turn, roll, loop, do circuits in an expedient manner without:
(a) falling out of the sky,
(b) giving weird staggers, or
(c) using up a lot of airspace?

3) Is it fun to fly?
4) Does it take off and land easily without crashing or ground-looping?

If it can do all this, then it is performing OK.

I suspect the weird staggers bit may require work. You want it to claw back out of a maneuver in an expected manner
and resume flight in an expected direction. This is mainly down to the sophistication of the flight software.

This is really a blend of quadcopter and model aircraft.

 

[Quinnyperks]

I had some thoughts on benchmarking the model performance:

1) does it have a good flying attitude in the air and make smooth flying progress?

2) Can it flip, turn, roll, loop, do circuits in an expedient manner without:
(a) falling out of the sky,
(b) giving weird staggers, or
(c) using up a lot of airspace?

3) Is it fun to fly?
4) Does it take off and land easily without crashing or ground-looping?

If it can do all this, then it is performing OK.

I suspect the weird staggers bit may require work. You want it to claw back out of a maneuver in an expected manner
and resume flight in an expected direction. This is mainly down to the sophistication of the flight software.

This is really a blend of quadcopter and model aircraft.

Will it have tricycle landing gear? Or retracts for efficiency?

 

[Tench745]

Are you a software engineer by any chance? A lot of software engineers I know have this mindset of, “Build it with all the features and then debug until everything works together.”
You’re setting up a lot of hurdles to jump over all at once to achieve success here. -First time designing a plane, first time flying an RC plane, first time setting up a flight controller, first time picking out hardware, first time tuning a model, and first time flying FPV. It’s not impossible and you know your capabilities better than us, but it sure as heck isn’t going to be easy.
All I ask is that you take every precaution when testing to avoid collateral damage. Fly in the middle of nowhere, set your failsafes and test them. Work through the flight testing one piece at a time and go slow so you don’t risk the plane, yourself, or the safety of others.
Oh, and keep updating this thread! We want to take this journey with you.

 

[OwenN]

Ah - software, mechanical engineering!
I haven't entirely neglected aerodynamic common sense.
If I left all the heavy bits out, and trimmed the COG correctly, it probably would make a good glider. 10-15 oz/sq ft would do it.

I have a graveyard close by which is normally deserted, and has a 50 yard round area that is clear of all obstructions. I don't expect that it will instantly escape all control-I can test for that in my back yard.

It is not exactly legal to use, but if I don't overdo it, the police shouldn't chuck me off.

I can do some slow-handling tests, landing and takeoffs. It should handle in that respect rather like a quadcopter. Just resist the temptation to peg the throttle!

Re: getting it all to work together. I have videos on the matek flight controller setup, and I can find more covering all equipment aspects.
I am fairly good at following instructions, and test, modify, retest....

 

[CampRobber]

Are you a software engineer by any chance? A lot of software engineers I know have this mindset of, “Build it with all the features and then debug until everything works together.”

Lol, that's basically me. Compounding that, I have an excellent workshop but basically no flying field at this point. MORE AMBITIOUS DESIGNING AND BUILDING IT IS!

I'll second your advice 100%.

When you test software, nothing breaks even if everything is wrong. When you test airplanes, everything breaks if anything is wrong.

 

[OwenN]

Trying to narrow down everything that can go wrong is a bit of a problem. Here are some:

1) actuators too weak. I have sourced some metal gear units which do about 11 kg-cm, which should be enough.
I don't know how heavy they are -between 17g and 56 g presumably - the 25kg-cm units are at the heavy end.

2) Elevator horns that tear out. The ones I have brought look too light-I will search for some heavier ones.

4) Elevons that twist, disintegrate, or rupture near the horn mount. - I can incorporate some spruce slats?
They need to be split to fit the hinges. Maybe sandwich balsa with epoxy??
Do horns need through-bolting to take 5 kg-cm??

5) Front wing fastenings that are inadequate. I propose to bond in brass tube to take M3 screws, and use lower strength loctite
on screws, so they can be removed. the fuselage needs to be built up to spread loads away from the screws.

6) Through-pivot-pins bending and jamming on vertical rudders. I was planning on 3mm dia piano wire, but I can do load calcs on that.
There is a rudder top and bottom. Also check for substantial tubing bearing mounts, and fuselage local structure.

7) Actuator clevis, turnbuckles, cross-pins that are inadequate. - don't know- I would have thought 2mm threaded would be adequate.
I will look at various clevis and pin system- probably over 1 mm pin sizes? They don't include loading guidelines when they are sold.
Metal sector arms, horns, clevises, spring-snap pin clips of some kind? - cotter??
The plastic sector arms seem to be paired with quite high torque actuators.

I am planning on putting actuators where they are to be used, and balancing things up by shifting the battery-that is about 300g.
A problem area is the location of the forward spar-this may clash with battery location. - it is just ahead of the planned COG.

A major revision in this area may be needed. Maybe I shouldn't skin the wing until I have all the internal components.
I could possibly include full fuselage loops as part of the spar-2 layers of 2.5mm ply, epoxy bonded?

8) Badly specced wiring and connectors -I would think that 30 amp appliance wire(fine multi-strand), copper cored, tinned, would be good?
Tidy solder joints, shrink sleeves, proper commercial end plugs.
I have a severe hatred of bare brass connectors of any kind, They should at least be tin plated!

9) Side support of the fuselage where the top covers slot in. -I could bolt the covers on securely- the coaming structure makes them very strong. Without the covers, the fuselage has poor resistance to twisting. Maybe spruce spars along the upper edges?
Anyway, this can be manually tested.

10) "Unplanned disassembly" of the wing structure. I need to establish the extreme wing loading to test to.

Turning the wing at right angles to the air flow at about 100 mph sounds like something that could happen.
(especially with borderline-stable COG placement-party trick? ) This also affects actuators, so I should check this case as well.

I can just use the basic thrust calcs to establish a load(Cd = 1). On a fairly straight wing, spanwise loading tend to be elliptical.
I can research how a delta wing loading should be distributed. loading can be gathered into two main load points each side to estimate
actual distributed loading, then I can apply test loads to those points, suitably spread out to avoid puncturing the wing,
and see if anything lets go.
Testing from the top put the worst loading on fuselage components, and can be resisted by a large flexible pad on the underside, which is flat. Maybe 100 mm wide?

I shall do bending moment calcs and bond in 1/16 balsa doubler plates, epoxy bonded and coated, first,
for any suspected points where concentrated loads need to be fed into the skin.

11) Vibration affecting electronic components.
Most to be rubber mounted-wrap in stiffish foam rubber sheet, 8mm? , clamp-brackets over the top, or rubber-band lugs.
Rubber band sound good-I don't want too many screws.

Exceptions are actuators and motors.

Any other ideas on things that need to be examined and tested?

 

[OwenN]

I just had a thought-what about 1/16 ply?
-midwest birch ply seems to be good, but it seems to be all sold out for now.

I can cut it into strips to sandwich the hinges, maybe it is the correct thickness for between the hinges,
and can be laminated up for the elevator horn mounts.

Also good for fabricating the more complex front elevators, which need ribs and skins, and need to transfer the twisting load.
Skin is available in 1/32 inch as well.

 

[OwenN]

I did the load calcs for maximum deceleration, and it is 30 Gs, so force = 30 kgf on the wings.
30 kgs distributed evenly over the area of the wings sounds sustainable.
It is not actually evenly distributed, but that is a good worst loading case.
The battery then weighs 11 kg, so better restraints would be Velcro straps. I will look for some small ones with metal end rings.
I don't thing the rings need to be welded. -It is not up to carabiner-level loading.

I have seen a video of an aerobatic model doing a flip-stop, so it is not uncommon. That one must be using a flight controller and an unstable trim as well.
You would think that "software enhancement" would be against the rules in aerobatic competitions, because the flight path could be programmed in, and the patterns would be "too perfect".? -you could add a digital wifi upload channel, and fly the thing entirely
off ground-based computer programs....- using a laptop as a controller would be a bit of a giveaway!!!

 

[OwenN]

Here is a proposal for a front spar with a large hole for the battery.
By my estimate, it needs to take 3 n-m,(2.2 ft-lb) which is not a lot. Would 1 layer of 3-ply 2.5mm birch plywood,
model grade, be enough ?
The root fillet would project through the wing skin, then be faired in front and back.

It is a difficult shape to analyse approximately.
By eye, the critical load is in shear in the 12.5 x 2.5 section, giving a shear force of 60 N
and stress of 2 N/mm(sq) Considering birch ply is good for 9.5 n/mm2 in panel shear, then 2 should be ok in cross-sectional shear.
This is partly across the grain and partly parallel to the grain, depending on the ply orientation.
I can improve this by using 2 layers of 1/16 inch 3-ply, bonded at right-angles at the surface, wrt. the surface grain.
This reduces the prospect of thin sections bulging and delaminating under this odd loading.

( panel shear is the shear parallel to the plies, associated with bending, and occurs over the whole section.)


Also, the battery would need to be hard-mounted. Is this normal for batteries?
I can't get more than a couple of mms padding all round inside a 70mm wide fuselage, and still get the skin, ribs, spar membrane, and keel-stringers in there. the battery is 30mm x 42 mm cross-section (2.75, 1.2, 1.7 inches).

I can soft-mount the receiver and flight controller, and the escs.

I found the velcro straps. they are one-piece, so would need suspended loops each side to thread through.
I will look for those, next. They do have a loop on one end.

 

[OwenN]

Looks like loops and pivots don't come separately. I can cut the loops off some straps, and use undercarriage pivot brackets, make my own pivot brackets from sheet metal, and/or bend up end loops from piano wire. That should be strong enough, even not end-welded.
Looks like I will be getting some sheet aluminium after all. -model grade, 1.2 mm? - or see the local boat builder. He makes "tinnies" !

 

[leaded50]

Rockwell XFV-12

 

[OwenN]

That's a neat looking aircraft-Very tempting to change to a similar configuration!

The body looks really bulky in relation to the wing area, though.- mine is more Heavy Bomber -proportioned. - like the Hustler??

The strong reverse dihedral wing is not so easy to make. Is this one of those "super unstable" aircraft?

It would make a good ducted fan model-plenty of fuselage volume.
Ducted fan is not a very efficient use of the available power, though.
You want say 3-inch props, 4 bladed, paddle, and spin them at 60,000 rpm?

Is it beneficial to build it like an axial compressor, with several stages of rotor and stator blades, and reducing cross-section?
You could get a very high speed air jet, which does not dictate top speed like a prop.

Matching the multiple stages would be tricky-different blade length, pitches for each stage, special blade cross-sections.
Does anyone make prop stages like that?
Also, you would need to support the fan axis both ends, with a motor shaft extension.

I hade a look at Wiki, and it was an unsuccessful attempt at a ducted vertical lift system. It is still an interesting looking aircraft, though.

 

[OwenN]

I looked at ducted fan power.
Basically, air velocity and power are inversely proportional to the prop diameter.
1/2 dia = 2x the velocity and 2x the power for the same thrust.

Looking at the plane above, the tip fins are about 1/6 the wing area each, or 1/6 of the fuselage side area each.

 

[Tench745]

That's a neat looking aircraft-Very tempting to change to a similar configuration!

The body looks really bulky in relation to the wing area, though.- mine is more Heavy Bomber -proportioned. - like the Hustler??

The strong reverse dihedral wing is not so easy to make. Is this one of those "super unstable" aircraft?

An FYI, "reverse dihedral" is called anhedral. On heavily swept wings have a natural stabilizing effect like dihedral, so anhedral is often added to make the aircraft more maneuverable.
I suspect any characteristics of being "super unstable" has more to do with the small wingspan than the anhedral.

Is it beneficial to build it like an axial compressor, with several stages of rotor and stator blades, and reducing cross-section?
You could get a very high speed air jet, which does not dictate top speed like a prop.

Generally, multi-stage fans in RC aircraft just waste energy. I don't know the specific physics of it, but I have heard of people trying multi-stage EDFs at an RC scale and it turned out more efficient to just run a single stage.
In a real jet engine you're using the multiple stages to compress the air and heat it for combustion. On an EDF you'd be wasting energy if there was any compressing between the stages.

 

[OwenN]

I have noticed that high-wings have anhedral and low wings don't. Possibly an interaction with the vertical center of gravity.
Maybe there is a pendulum effect assisting roll stability??
Re: Super Unstable:
It just seemed similar to a previous aircraft I saw on RC Groups??
The NASA-Boeing X-36
https://www.nasa.gov/sites/default/files/images/353809main_EC96-43641-8_full.jpg

RE: Multiple stages. Bypass air on big jets usually has more than one stage.
They are set-up to just move the air faster, not compress it.

I agree, adding extra blades or using contra-rotating props is not the most efficient method of generating thrust.

Just use 2 blades of a larger diameter, or slightly wider blades.

Those USA ww2 military 4-bladed props had quite narrow blades for their diameter.

I am not sure that these 4-bladed small paddle-props used on quads or ducted fans actually pump more air.

Maybe it allows better performance at zero speed, and high pitch angles. Otherwise, tip reverse flow tends to dominate.

Possibly high pitch narrower blades are unstable, if they are widely spaced, at zero forward speed and high rpm, or when ducted???

On big planes, multiple blades probably smooth the prop wash a bit and cut noise and vibration.
5 to 7 blades is a bit of overkill, though.

 

[OwenN]

Loading calcs results: max actuator = front elevator, 7kg-cm, so 11 kg-cm unit needed., 55 g each
rear elevons-2.6 kg-cm each.-use 2x 11 kg-cm units.
Rear rudder - balanced-can use a micro actuator., 17-20 g. wt, 1.5 kg-cm
Rear rudder pivot pin bending stress, 1/8 spring steel, = 6.5 KN/m2
Steel = 448 MPa = 448 MN/m2, so well under!
Nice through-bolted elevator horns found.

Passing Elevator torque into fuselage-check out pine dowels, 10mm dia.
Z= 2 x the beam-type moment (I) , M = 7 kg-cm, I = (pi/4)r4, Z = 9.8 x 10^-10 m^4

Actually M each side is 1/2 this, this much in total is resisted by the elevator horn connection.
(for twisting, M should really be T-same units.)

Pine shear parallel to the grain= 4.7 MPa, so dowel -type shear should be similar.

Stress = Ty/Z = 0.35 N-m x 5 x 10^-3 / 9.8 x 10^-10 = 1.8 MPa,

so a 10mm (3/8 inch) should be OK to use. I need to pick a really nice-grained dowel.

Splicing it into the elevator structure is a bit of a trick- use lap-sandwich joins in ply, preserve
80 % of the pin cross section, flatten sides, epoxy all round. I can't just epoxy the dowel into a bore.

The horn connection needs to be done with care as well- shave sides, sandwich with ply,
insert bonded spacers both sides. sandwich height over 10 cm.

Allow enough space around mounting drill holes to carry the torsional loading.
Crush loading should be OK with plastic plates both sides. Put loctite on screws.

The main pivot bearings are the elevator hinges, so support at the dowels is not needed.
They need to be well lined up with the hinge pins.

I need to keep reading the Lennon RC book PDF to find tidy details for the elevator slot area.

 

[OwenN]

Here is a revision of the XFV-12 with twin props, to my new STOL layout.

.This is semi-scale.
It abandons tailsitting, but should launch off the ground near instantly.
I need to test the blow-back scheme to get a nice launch angle of 45 degrees.

 

[OwenN]

I will redraw this in 1/5 scale, and add internal detail-this can be similar to my existing layout of ribs, spars, leading and trailing edges.

Re: anhedral - once in a banked turn, any pendulum effect is cancelled out, and this design may automatically continue rolling and diving. Positive control is needed to counter those features??

I think pulling air over the front wing with the elevator down may give enough static lift to get a jump-launch.

Otherwise, there is enough room to add a small thrust tube behind the cockpit area-needs an extra throttle channel?

Wings can be attached using ply doublers onto ply spar stubs. glue up with 15 minute epoxy, do before putting the top skins on the wing.
This makes sure any gaps are filled.

Control surfaces can be linked up using flexible sleeved push-pull cables.
Insert these into the wing at construction.

There are wing end pods to hold rudder horns. They can be connected each end of a "T" actuator arm.
Revert to 11 kg-cm actuators all round.

The front wing can be glued in place, and the whole fuselage top can be made removable, past the cockpit.
I haven't made an allowance for the FPV camera yet. Maybe merge the 40mm dome in a bit closer under the cockpit?

At this stage, I can easily adapt my wing and section design to this new layout.

It looks a little pot-bellied. Maybe thin the fuselage a little vertically?

 

[OwenN]

My scaling was a bit off-it turns out at 42 inch span, not 34, and 44 inches long.
This design looks a bit fresher than my old one, and solves several problems.

1) the load bay can be split into lower rear, and upper front, with covers which can be removed from the whole fuselage.
The sides between the two wings can be beefed up to suit, into 2 box-sections, one each side.
The rear equipment can be installed with the model upside-down, and then hang from the "roof" or sit on
velcro straps.
2) there are outer pods, originally intended for undercarriage, but they are good for concealing rudder horns.
The rudders look rather undersized for a model. I can increase then by 20%, and also increase the upper fin height.
3) the location of the motor pods make excellent fake jet pods-this improves the appearance.

I need to design some planked round surfaces to suit. - maybe cross-grain 1/16, sanded down.
I can also use balsa filler. That stuff is very light and quite strong. I use it on my static models.

Re:finish-I see that skinned models have a highly glossy surface. What sanding fillers and top coats are used?
That would have to be sprayed. Can I use monofilm as a surface finish? How smooth do I need to get the initial surface?
What can I use as a show coat? - It needs to be glossy, so in shows any imperfections when sanding.
I have a cheap airbrush on order. Can I use that to put on gloss coats? Do I need a full size spraying set?
thst is usually used on model 6 inch long. 44 inches is a step up!

this editor is still playing up! I have trouble turning of an overtype fuction. I have to close the website, go back to
Edge history, and reopen this website. - I have a query out with microsoft community.

 

[Tench745]

Reading your posts is a bit like talking to my brother (a civil engineer). There's a lot of talking through ideas that I don't completely follow, quick jumps between thoughts, and a lot of ideas to wade through.
So I just wanted to say again that I am following this one even though I have little to contribute most of the time.

You can use any kind of light filler on balsa. Sand smooth and seal with whatever sealer you like. I often finish models with spray paint; no spray equipment necessary and you can cover a lot of area quickly. I'll usually do some kind of sealer like water-based polyurethane. Then a coat or two of primer, sanding and smoothing between coats. Then a color coat, then a clear coat in whatever level of gloss you want.
Why does it need to be glossy?
I'm not a fan of super-glossy finishes because they dry slow, mark up easily, and show off ever little imperfection in the layers below it and make orientation difficult with the sun reflecting off them when flying.
You can use an airbrush to spray if you want, but if it has a small cup you will have to refill often. Remember that area quadruples when size doubles so you'll be spraying a LOT of surface compared to those 6" models (more than 53 times the area). Quality of finish depends a bit on your level of skill and the quality of airbrush.
Iron-on films can also be used and will create a moderately durable high-gloss finish. They will show any lumps or sanding crumbs that are below the film, so you'll want to sand things smooth and wipe everything down with a damp rag or tack-cloth to remove the sanding dust.