Slow Flyer

27 February 2017

Removed the board from the Slow Flyer on 25 February 2017 and used it in a F15 SG.

1 July 2016

I wanted a simple small delta with WLTOYS receiver board and a direct drive 8mm motor and hubsan prop pusher. Unfortunately, the FS TH9X transmitter had its throttle full on, and until such time I get it fixed or found a replacement, I am stuck with WLTOYS transmitter. So I stripped away the motor and board from the failed delta and did this BD5 instead. The curved shapes are free hand, the dimensions were what I had left over of the 2mm Daiso foam from the failed delta project.

The Hubsan propeller is only 55mm, maybe I made the BD5 too big, will it fly....

Well, if it does fly, it will be a slow flyer. On the bright side, the board is a V949, so I have the direction control on the right hand stick; feels more natural. And the wings are simple curled 2mm foam sheets glued at a dihedral angle after making appropriate concave cuts to the roots. The wing halves are hotglued and stiffened with strips of PVC sheet. The failed delta had the elevons hinged one at the top and the other at the bottom, I did a samehanded pair of wing halves! hahaa....

On 26 June 2016, the field was windy but I was there. The wings are too flexible, they curved up easily and it wasn't possible to fly. I shall stiffen the wings span-wise and try again because there seems to be enough thrust from the 55mm propeller.

I clipped the wings so it can fit in my motorcycle box. I have also added a carbon fibre strip to stiffen the wing. On 30 June 2016, test flights resumed. Test glide looks ok, but I still think the CG needs to be forward because it seems to sink more instead of glide. The wing does not distort out of shape now. The high thrustline makes the model dive, so the throttle was minimal during test flights. The fin and rudder was too large, it goes inverted easily and it doesn't turn with the rudder. The elevator is very sensitive even with small throws. I think the tall rudder is acting like an aileron, which is against the direction of intended turn. Back home, I reduce the fin and rudder's height, so it is ready for test flights. If that doesn't work, I have three options:

1. cut and hinge the rear fuselage below the tail to make that as a rudder
2. cut and hinge the rear fuselage above the tail to make that as a rudder and because the motor is glued to that, I get thrust vectoring.
3. buy and use that FS-i6 and use elevon mix. This seems to be simplest option.  

On 25 June 2016, I wondered and came up with this sketch.

Since my TH-9X transmitter is not working, these shall be controlled by the WLTOYs' V911 transmitter, that is, rudder/elevator/throttle. The receiver board will be hotglued through a rectangle hole in the top of the model. The servos and servo horns will be exposed. The battery will be routed forward of the board and velcroed to the nose.

For thrust, WLTOYs' geared motor and prop will be glued on to the wings, tractor-mode so I don't have to fabricate a pusher prop. The top half of the prop spin will be covered by the top and two sides; the lower half arcs through the bottom through a slot in the wings. I would prefer that the props be clear of the wings, but if I do that, the nose of the model which carries the battery and will have to sustain a lot of impact force will only be hanging at the top to the trough and structurally that would be too weak.

I could have chosen to use the 55mm prop, but I think it would be more interesting to have a boxy plane that has more thrust.

These are inspired by the RC 'Egg Planes'. I don't do them because they are a lot of work and a lot of foam. Thin foam sheets are not readily available off-the-shelf in Singapore. The Daiso foam I am using is 2mm thick. They are cheap, $2.00 for 5 sheets of 300mm x 450mm.

 I call this type of boxy planes, 'Cuboid Planes'. Simple to build accurately since they don't need extensive rolling or curling. The 90 degrees bends are easier to do and these are far and few apart. Cuboid Plaines are Profile Planes that uses straight edges and square bends to achieve the cuboid look. They are simple to design, just enlarge the section of the fuselage in height and width to house the propeller and blend the rest to suit.

I think the ducts in a ducted fan increases thrust, but the small fans reduces efficiency because the blades are of reduced span. Overall, propellers beat ducted fans, at least for model use. So I give away something by using a rectangular duct, but it is better than nothing! And not forgetting that I retained the full span of the propellers. Who knows, perhaps my square/rectangle ducted props might give out more thrust than a simple prop?

The square/rectangular duct has 2 horizontal and 2 vertical surfaces. They are far apart from each other, so structurally, that's a plus! The 2 horizontal surfaces are bi-planes, so they contributed to the wing area. The duct is short and goes directly to the horizontal tail surfaces, so the tail surfaces will be more effective; even though the prop stream is only to one side of the tail surfaces.

My idea of a F4 Phantom. Hoping this would be stable for a RET RC plane. Area above wings is dihedral. Motor at bottom lowers CG, partly offset by the receiver board which is mounted on the 'ducts'. The 'Duct' wing is bigger than the low wing. Anhedral tail is ok since it is above CG. Battery will be velcroed to nose for CG balancing. If CG balances alright anyway, then maybe the battery can be slung under the low wing? Instead of rudder at the rudder position, how about moving that triangular bit as a rudder? If I do that I get some indirect thrust vectoring?

Through the observation of BD5 I think this F4 Phantom's rudder will not be effective but the triangle bit will be. Simplest option is still to buy and use the FS-i6 transmitter, choose elevon and either make the elevators into elevons or make the bottom of the duct into elevons. For the first option, the horizontal surfaces will be huge and the CG can be located aft. For the second option, the horizontal surfaces can be reduced since it has become a flying box and the rest is appendage, so the non-moveable tail surfaces should be reduced with the CG located forward.

13 June 2016

I got 30 minutes flying time on a 520 mah cell. The model went up sedately under power, and with the motor cut off, it came gently down. The flying speed is slow, I can catch the model easily when seated. The rudder response is good, coupled with the dihedral, I could turn very tightly, sometimes, it seems it can turn around along the fuselage length.

Obviously the incidence of the wing relative to the fuselage is too great and the model doesn't require that amount. As the model sinks slowly to the ground, the fuselage is pointing downwards at a noticeable angle. If I reduce the incidence, the fuselage will appear more in line with the direction of glide and there should be less drag. The high wing incidence relative to the horizontal stabiliser (which is mounted flat along the fuselage) also meant that the elevators have a large amount of down. If I reduce the wing's, less down elevator will result and better efficiency. The two wheels are still there although they could be safely taken out because landing would not be any problem without them and more drag reduction.

The soft wire connectors are great, I could tighten the pull strings to give the controlled surfaces more left/right or up/down trim.

8 June 2016

Yes it flew. In the end I managed to get the model trimmed at around 50% CG with drooping elevators and responsive left and right turns. 15+ minutes of power and glide is no problem with old WLToys' V977 520mah single cell. Main construction material were from bought from Daiso and hot-glued together.

It flies slowly, it was the slowest plane on the field. Short blips of power would cause the model to surge forward before slowing down again visibly. Landing is on the spot.

Structure

Assembled, the plane was placed in the car boot, but on arrival, the front cabane strut to fuselage was detached. On inspection, wood fibres were stuck to the hot glue, so I come to the conclusion that hot glue is 'stronger' and therefore perfectly ok. A tinsy bit of UHU POR from our Philippines freinds got the model repaired for flight testing.
The tip panels were bendy, pulling up on a dive has the tips bend upwards and too much down elevator has the tips flattening!
Fuselage is too bendy and it can't resist the left pull of the left sided elevator pull lines, I hotglued a line on the right to pull the tail.
There is no wing struts. I figured I would let it flex.
A single rubber band is used to secure the wing to the pylon mount.
My fake spoked wheels held up, perhaps because my landings were on the spot and not rolling landings. Whatever it is, I think I can adopt the spoked wheel making method used for my WLToys' planes.

Layout

CG: The CG was shifted from 75% to 50% and I did not foresee having to drop the elevator so much, I could add area to the tail since I can't add tail moment. I could also remove the pylon mount and hotglue it rearwards. On last flight I reversed the wing's planform. I placed the trailing edge infront and the leading edge behind. The wing's centre of area is shifted backwards relative to the CG, therefore causing the model to react as if it had a slightly forward CG.
Wing's angle and motor's downthrust: I thought it would. The fuselage stick pointed lower than the glide slope. If I reduce the wing's and elevator's angle to bring the stick fuselage along the glide slope, I will have to increase the downthrust angle by the same amount because now it flies ok power on and gliding power off.
Right thrust: I think it needs a bit more of right side thrust, the model turns more readily to left (could be the side pull of the elevator pull lines), but currently it turns left and right ok.

Mechanical

I tried correcting the pull pull conversion at the field by heating with a lighter (to cater for the large down elevator setup), instead of softening the hot glue, I softened the plastic at the same time!
The pull lines were doubled back and hotglued with heat shrink sleeves and not adjustable.
I could use smaller levers and horns.
Next time use thinner lines and braided lines would be better than monofilament lines. Current (thicker) lines does not cut into the plastic holes but they have largish bend radius.
Everything is hot-glued: the motor, the receiver, the undercarriage, the struts, the wings, the tail surfaces etc. At end of flying session, no problem.

Improvements

Here's an idea to replace the pull pull lever set.

A soft wire is bent into a V and hotglued to the WLToys' F929 F939 receiver board. The arms are in lined with the hole of the servo horn that is to be used and the line path to the rear. If they are too far off, then the pull in the opposite direction will be lesser.

Short plastic sheaths, such as a section of the plastic cotton bud, will be allowed to rotate freely on the arms. Even if they got snagged, it ought to be alright too because the pull strings will glide across the smooth surface. In the interest of efficiency, we don't want that, so a short plastic spacer (not shown) can be hotglued to the V arms. The short plastic sheaths can be made to look like tiny sewing bobbins, but I think a simple cylinder will work well as a pulley, the string will not run out of the cylinder because of the geometry and tension.

The pull string for a servo is one length. A loop of string passes through the hole of the servo horn that is to be used. To prevent the looped string from pulling out of the hole, we could tie it or, as suggested, to make a keeper from thin wire. This wire is kept in placed on top of the servo horn by the hole it entered (the other hole, not the hole occupied by the string) and the loop of string. Perhaps a twist or two before inserting the Z bend will prevent the string from sliding too easily. With the looped/hooped string 'caught' by the servo horn, one end goes around the pulley one and a half times and head back to the control surfaces.

2 June 2016

 I decided on bellcranks (without any bell or crank, a pivoting lever). I have this waste plastic piece for mobile sim card and 2.54mm pitch electronic pins.

1) I hot glue the pin bank to the edge of the plastic. I used the black plastic spacer and the copper pins as drill stops.



2) I drilled a series of holes with a 0.9mm drill bit on a dremel rotary tool. I didn't have to 'centre punch' and electronic bank guided the drill bit well. The holes went clean through.

3) I cut the drilled plastic with a pair of shears. I used the tips of the electronic pins as a guide. I removed the hotglued pins with alcohol.

4) I cut the drilled plastic to shape and I have my bellcranks and horns (maybe I shall re-do the rudder and elevator horns, but it really doesn't matter much).

To keep the pivot and levers in their correct position, I used a paper puncher and made many discs of the plastic. I will make a hole through these discs with the same 0.9mm diameter drill bit on a dremel by hotgluing them to a scrap holding piece, drilling, and removal and cleaning up with alcohol.

30 May 2016

I throw out the previous attempt, the wood kinked easily and I can imagine them breaking easily at a later stage. I think that 2mm foam sheets from Daiso would probably be as light, but much easier to do, so I did a bigger version with 2mm foam sheets. I also want the CG to be behind the 50% wing chord to get a short nose, so I chose to have a large tail and long boom.

Centre panel was two sheets with two 2mm diameter rods for the leading and trailing edges and an upright 10x3mm wooden strip for spar.
Tip panels, fin/rudder and elevator (with a 2mm diameter rod at the leading edge) as of 2mm foam sheet.
The left over 10 x 3mm wooden strip was the fuselage.
Hot glue was used throughout, motor was hotglued and the receiver board hotglued to 10x3mm wooden strip which was glued to the fuselage.
The main undercarriage was a thin wire bent and heat shrink sleeve applied to provide bigger contact area for hotglue. A failed pair of wheels was installed on the wire undercarriage. Tail skid was a triangular piece of 2mm foam.

The wing mount was a T-section made up of 2 pieces of 10x3mm wood strips. They will be glued to the fuselage with 10x3mm uprights later.
Before I continue with the pylon wing mount, I want to link up the control surfaces to the WLToys' receiver board. I don't want to use pushrods because it would be so long and heavy, so I am thinking of using pull-pull for the tail surfaces, either with bellcranks or pulleys which can be mounted on the fuselage, to the rear of the receiver board.

This is a learning experiment with short nose and lifting tail before I continue on my R. Sommer's Monoplane.

Yes, this is too big for a basement carpark flyer.

31 December 2015

Daiso is selling a packet of 10 wooden sticks of 2mm diameter x 910mm length. The wood is light and flexible. It is not springy like bamboo and retained some curves after bending it dry by hand. I don't know what tree or plant they came from, but their mechanical properties are useful to modelling.

Yesterday morning, I decided to try my hand at making a slow flyer.

I started with the wing.
I marked out the single piece into 4 segments.
The two ends were connected so that the single piece became a loop. The ends were connected with CF rod, heat shrink and Daiso's superglue.
I wetted the loop at the 25% and 75% markings and started to pull the loop at those markings.
Then the loop snapped, so I re-joined the ends to make a smaller loop.
This therefore became the elevator.
I wetted the smaller loop but didn't want to pull the loop anymore, so I pinned the joint and the centre marking down onto a corkboard (Daiso's), forming an oval.

Another piece of wood was cut approximately in half, wetted and bent so that the area is slightly less than half elevator.
This is the fin/rudder.

A third piece of wood was made into a loop, and pinned to the corkboard like the elevator.

























With the wing, elevator and rudder pinned, I sprayed the frames and left them to dry out for a few times.

I wanted to leave the frames pinned for 24 hours but didn't resist the urge to remove them and played with the frames.

If I want a simple rounded tipped wing, I should pull at the 25% and 75% mark. I pulled but the frame snapped. Thang was telling me I should have used CF rod because I wasted so much time.















I am not going to throw it away, so I transferred the length of the snapped side and cracked the other side of the wing frame.



Now I have to figure out how to do the rest of the model.

The objectives of this model are:

1. Slow flying. This means it has to be light. I shall use the Hisky 7mm geared motor that Thang gave me (3gm?), a WLTOYS F929 939 receiver board, and V911 batteries. I will use a curved airfoil at an angle of attack near the stall point. The elevator shall be lifting foil and the CG shall be set rearwards.
2. Small. The model shall be small so it can made single piece and be resilient.
3. Cheap. Using Daiso's wooden sticks, some CF rods, paper clip, hot glue, superglue, plastic covering.
4. Easy. The model shall be easy to build, setup and adjust.
5. Fast. The model shall be fast to build.

The targets:

1. Fly in basement carpark.
2. Fly at walking speed.