Joining Wing Halves:

I glued the 2 cross joint spars into one wing half before joining the wings. Align the spars by putting both wing halves together with the spars, but only glue the spars on one side.
On final wing assembly glue the 2 halves and the other half of the cross joint spars. I used 1 hr epoxy for the wing halves because they were not exactly flat with each other and I wanted to fill the gaps. I used Titebond III for the spars.

The long spars which do not span the wing joint can be added independently of joining the wing. I added them beforehand. I first used PU glue for these spars, cleaning up the foam as it dried. But it still ended up a not very neat job with removal of excess glue necessary afterwards. That is why I used Titebond for the other spars.

Nose reinforcement:

One gets some hard landings on the slope so I added some reinforcement. This consists of a triangle of 2 laminations of 1/32” ply epoxied into a slot in the nose. The triangle is ~15 mm x 45 mm spanwise. I also added small pieces of epoxied Kevlar at the nose on both wing surfaces.

Ventral Fin, Tow Hook and CG Position:

CG position
We need to know where the CG will be before we design and position the ventral fin and tow hook. The only purpose of the ventral fin is to hold the model and help hand launching. The tow hook is not essential but if it works out might just make this model flyable at thermal fields too.

The rule of thumb for Tailless Wing CG position seems to be between 20% and 25% of Mean Aerodynamic Chord. My Halfbad is recommended at 21% and it flies well there. Here are the Sogwing parameters for calculating CG position.

Span: 118.7 cm
Root chord: 31.4 cm
Tip chord: 17.1 cm
Sweep: 35.9 cm

Note the chord dimensions include the elevons because of course these are also part of the lifting surface.

The resulting CG positions as a function of %MAC are as follows:

%MAC CG Position from nose (mm)
20 212
21 214
22.5 218
25 224

I'm lazy so did not calculate the above from first principles. I have 2 calculators for this which I've previously checked. This is the calculator I used;
http://fwcg.3dzone.dk/

It seems to be well vetted on RC Groups over a 3 yr period;
http://www.rcgroups.com/forums/showthread.php?t=973140&highlight=cg

I assumed the CG at 21% for starters and put the hook position at 16mm (5/8”) forward of the CG. I figure it is best to err on the forward side as although this hook will not be adjustable, the hook can effectively be moved back by adding material in the knee of the hook. This position also places the hook immediately behind the rear spar, so launch loads are transmitted directly to it.

Ideally the ventral fin needs to be at the CG, but with a tow hook there is not room. So it has to be forward of the hook.

The tow hook is 1/16” rod through the complete depth of the wing. It is anchored in 2 plywood laminations of 1/32” ply at the lower surface and 2 laminations of 1/64” plywood on the upper surface. A brass grommet in each of the upper and lower plywood anchors transfers the load to the plywood. The launch loads are carried as follows:-

Hook pull out load is taken by brass tube CA'd to the hook rod at the upper surface so the hook cannot pull through the grommet. Epoxy between the hook and the wing also adds pull out resistance.
Bending moment on the hook is reacted by shear between the upper and lower plywood anchors and the wing.
Shear on the lower surface is also reacted by the lower anchor. The anchors are epoxied to the wing.
At least that is the theory, but I won't be too surprised if it needs improvement in practice.

The ventral fin is just big enough to grip between thumb and forefinger. It is 1/8” balsa skinned with 1/64” ply. The fin fits in a slot in the lower anchor ply.

Balancing sling point: Where possible I balance my models by including a sling point at the CG. This allows accurate balancing in both pitch and roll. In this case it a piece of ply with 3 balancing holes at 21%, 23% and 25% of MAC.

Kevlar skin lower surface:

I skinned the lower surface after adding the ventral fin and tow hook. Trimming the Kevlar after the Polycrylic was dry ended up a much neater job than the upper surface. All part of the technique learning process.

Vertical Stabilisers:

The V. Stabs are 1/16” balsa skinned with ¾ oz/sq yd fibreglass each side and sealed with Polycrylic. I would like to have used 1 mm Coroplast but couldn't readily find any and it was not worth buying a sheet just for this. The shape and area were just scaled from my Halfbad.
I've found the following works well for lightly fibreglassing tail feathers.

1. Cut a generous overlap on the fibreglass (2+ cm).
2. Lay the glass on the balsa and coat generously with Polycrylic.
3. Smooth over with a paper towel and then with another paper towel, soak up as much Polycrylic as you can.
4. Repeat for the other side
5. Sandwich between plastic sheet, Mylar is best if you have it. Do not use food wrap as it will stick. I use page protectors or photocopy transparent plastic.
6. Weigh down heavily between 2 flat surfaces. I use 3/4” oak with a 4l paint can on top.
7. Leave a minimum of 24 hrs, preferably 48 hrs. Polycrylic needs air exposure to dry, so dries slowly when compressed between the plastic sheets. But it does dry eventually.
8. When removed from the sheets, trim the glass and round off the edges.

The finished stabilisers weigh 5gm each, before any colouring/covering. I'll decide how to finish them when I know what the balance looks like.

Servo Pushrods:

The pushrods are 0.032” wire with a Dubro #917 micro clevis at the servo end and an out of plane Z bend at the control horn. The advantage of an out of plane Z bend is that the pushrod can be inserted/removed from the horn by just twisting the wire.

The servos are quite far forward, so the distance between the servos and the horn is quite long for an unsupported rod. So the rods run through an 1/8” dia. Teflon tube glued to the wing top surface. I wrapped some thread at the ends of each Teflon tube to improve adherability to the wing.

The control horns are cut from milk bag ties and glued into a slot in the elevons.

I tried just holding the servos in place with only double sided tape, but it would not hold the servo securely under the reaction torque. So I made the top of the servos flush with the wing surface with balsa glued to the top of the servo, and fastened them down with a 1/64” ply cover plate, slightly overlapping the servo opening, glued to both wing and servos. The servos are wrapped in masking tape before glueing.

Hatch Cover for Battery and Rx:

The forward position of the battery is such that I could not recess it completely within the wing thickness without compromising the strength of the nose. So the hatch needs to accommodate the battery proud of the surface. The hatch is built up layers of 1/16” and 1/32” soft balsa. It is built up on the wing with the battery in place so it smoothly conforms to the required shape of the wing and proud battery. This is more accurate and provides a neater fit than carving from solid block. The hatch is held in place by a 1.3 mm short CF rod at the front end and a Velcro type fastener at the rear.

The nose and tail pieces at front and rear of the hatch to provide streamlining are also fabricated in the same fashion as the hatch. The rear of the nose piece has a 1/32” ply bulkhead in the front of the battery recess and also includes a female tube (insulation from 14g house wire) to house the hatch CF pin.

I fibreglassed the hatch with 1.5 oz glass as it will be handled a lot and it is a bit thin at the edges to blend nicely with the wing.

More pics of the nose piece in front of the hatch during fabrication.

I've often thought that aeromodelling is a building exercise in which all the angles involved have to absolutely accurate. Attaching the V stabs on a tailless wing is no exception. The stabs have to be square to the wing in the vertical plane and parallel to the wing centreline.

This is the procedure I used to to ensure the stabs are aligned parallel. It is complicated by the fact that the wing tips where the stabs are installed are behind the trailing edge of wing root. So one cannot measure from the wing tip to the centreline without extending the centreline rearwards.

1. On a flat surface mark a straight line at least as long as from the wing nose to the rear of the wing tip measured along the wing centreline.
2. Lay down the wing (upside down as the ventral fin gets in the way otherwise) with its centreline on the drawn line.
3. Measure the distance from wingtip to drawn line and wingtip to wingtip at both leading and trailing edges of the tip.
4. Repeat the above but from the other wingtip.
5. Average the tip to centreline and tip to tip measurements.
6. Determine the differences between tip to centreline at all 4 points (LE and TE at each tip).

In my case the differences (mm) were as follows:

Left Right
Tip LE 597.5 595
Tip TE 597 593

From the above it is deduced that I need to add 2 mm to the right tip TE to ensure the stabs are parallel. There will still be a 2 mm difference between left and right half spans, but as this is only a 0.3% difference it is not worth bothering about.

The stabs will be attached to a 1/32” balsa airfoil shape shims glued to the foam wingtips. The main reason for this is that I don't have a fast acting adhesive to attach the balsa stabs to the tips. We need a fast acting glue because it is difficult to clamp the stabs to the tips. But it is easy to clamp a shim to the tips using tape.

The 2 mm difference is dealt with by shimming the right airfoil shaped shim at the TE by a small 2 mm shim. The gap between the airfoil shaped shim and the foam is then filled in with 1/32” stock sanded to fit.

The stabs will not be attached until after covering both wing and stabs, as the covering process is much easier with these as separate pieces.

Covering:
Before deciding on covering, I made a preliminary balance check. If the model is very tail heavy and needs up front ballast, I did not want to add more rear weight from the covering. The wing area behind the CG exceeds that forward of the CG, so covering will result in a net movement of the CG rearwards. It turned out that the preliminary balance was almost perfect at 21% of MAC. So I was not not overly concerned about the effect on balance from covering. Note covering on the wing is really only cosmetic, but does add some strength to the stabilisers.

I decided to use Coverite Microlite, which I like and have had success with on several models. It is light, advertised at 0.6 oz/sq yd (though I suspect heavier) and easy to apply. I tested application to ensure it does not melt the foam under the Kevlar. It does not at a medium iron setting (#5 on a Hobbico iron).

The only problem I had, that I had not experienced before with this covering, is that it tended to curl inwards, towards the adhesive side, after removing the backing. I suspect this may be a function of humidity. I had not covered at this time of year before when the humidity is relativity low and static may be higher.

For complex shapes I use a paper template to cut out the optimum covering shape. This enables some experimenting with the shape before cutting out and applying the covering.

A functional advantage of the covering I realized after applying it is that it seals the cut edges of the Kevlar. I did not overlap the Kevlar at the edges as it is not easy to do neatly.

The covering added 18 gms to the wing and 1 gm to each stabiliser. So I reckon the covering is roughly 40% heavier than advertised.

In retrospect covering would have been easier if I had masked the area where the pushrod tubes are attached, then removed the covering local to the tubes so they could be attached afterwards. Covering around the tubes/rods was fiddly. I also realized that I could have buried the pushrod tubes beneath the wing surface.

As noted previously I had already checked the wing tips for being parallel. So the only remaining alignment check prior to stabiliser attachment is to ensure they are perpendicular to the wing. It turns out no adjustment was needed for this and a big thanks to Ray for cutting the edges of the foam square. I attached the stabilisers with medium CA and used kicker to ensure they adhered rapidly. As noted previously clamping the stabilisers to the wing would not be easy.