Main Construction Page

September - 07

The staggered tape joins were laid up at the plane of symmetry and after a couple of days of extra curing the fuselage was demoulded. The part has some accessible edges that a thin nylon wedge can start off the mould, then the normal wedges at the flanges. The wing root was a bit resistant, but a couple of sharp hits on the mould and out it came. The 100g outer surface glass seems to be fine, just one very small area of delamination, with evidence of sticking to the mould a bit there.

The glue squeeze looked good. The inaccessible glue lines in the fin were quite tidy. The LE squeeze was also quite good, but the LE squish space, which is quite small, was full. There were a couple of places on the flanges that had squeezed a fraction of a millimeter thick layer of glue. Ignorable for the geometry.

Overall the shell looks quite accurate. We did later mock up the seat area and enjoy trying out the big pilot fit. More on this later.

August - 07

After some grinding and small mods to ensure glue gap there were the final preparations for glueing. The plasticine squish test was done and all looked good. Most glue gaps were about 2mm, except our joggle gaps, which, because of how we designed and built the joggles, normally have about 2.5mm at the LE or seam and about 0.5 to 1mm at the edge of the joggle.. This is because the joggle moulds were made using the origional plugs, just allowing for a translation down and back away from the LE. The result of this quite good for the glue squeezing. Not much goes inside the shell, it's forced into the external squish space in the mould flange.

Only the fin, the rear part of the fuse and the bulkheads are glued, the forepart is joined with glass tape later. We call it tape but it never is commercial tape. It's normally cut wet with fibres at 45deg. The glue used on joining the halves was about 900g.

The fuse moulds can unscrew from their steel frames. Three peoiple are enough to lift the port mould, containing the port shell, and lower it onto the starboard. All went perfectly to plan and was fun. Helpers in the photo are Robert Smits and Murray.

August - 07

As much as possible had to be done while the fuse shells were open. The rudder cable pulley mounts were finished - plywood bearing block for the main bolt, a 4130 bracket that is bolted on top retaining the cable and giving more stiffness and security. After a final alignment check these were glued in. For the rudder torsion tube bottom bearing mount bracket our idea changed a bit. A new bracket was made which was separated from the tail skid bearing bracket. The rudder contril stops are just cap screws locked to glassed in flange nuts in the glass bracket.

The linear bearing for the elevator push rod posed some interesting questions. The danger is that the ball cage in a normal linear ball bearing system will hit the bottom stop a lot and this will cause wear and maybe give a random control force input. A perfectly fitting linear ball bearing system might help, who knows. The bearings we made previously have a small clearance, so I experimented with very light springs to keep the ball cage off the stop. This was fun but seemed a bit complicated for a part that is inaccessable. So I tested some UMPWE bushes against the aero grade 6061-T6 tube at the deflections that our fin will have. The friction coef was 0.12 and the stick force increment due to friction even with fin at limit bending loads was small enough. So in goes the UMPWE bush.

A glass bracket for the tailplane front mounting bolt was made using a quickie mould for the port half. Also, some extra reinforcements for the shell at the top of the fin. Finally the agonising question of what to do about future proofing for the installation of TE probes etc later. Two 12mm sockets were made behind the fin leading edge. If the (highly expensive) 3 function probe mount is used we can either shift its front O ring position to shorten it or we can have the 12mm mount poking through our LE ! Our fin has carbon UDs at the LE in the design so the sockets have extra reinforcements with UDs going around defined hole locations. For the plastic tubing a set of well fitting clips were made and glued in at defined points.

July - August - 07

The "bellcranks" are held in aluminum brackets which are cut from 5058 plate and some extruded section. Murray bought an inexpensive Chinese mill which allowed some better features to these parts.

Martin at Airmaster made us some flanged housings for the 626 ball bearings. These clamp the bearings to the plate. A cap screw connects the bearing to a thread in the end of each bell crank. We have to be careful to install the assemblies without too much missalignment, another reason for the gig system. This was carefully checked while we had good access into the fuselage, then a final check with the mockup push rods. The control hardware was taken out while we prepared and glued the fuse halves.

July - August - 07

The control system elements behind the pilot are designed to bolt to a simple mounting geometry that will allow easy removal of the complete assemblies. Jigs for drilling and assembly were used so we hope this makes replacement or further development of the system easier. The system is connected to existing structure that is stiff or that can be made stiff. It remains to be seen wether this makes the control system more damageable in a crash. Some calculation and FEA was devoted to the effect on the control system under maximum landing loads. The bearing system (plain 626 ball bearings) recieve load and missalignments in hard landings. Sofar this all looks good. But maybe one day if the friction is not too high we could have a bearing system that is all done with something like DU bushes.

The mounting points on the spar bridge box are reasonably stiff raised sections with some shear connection at top and bottom and an enclosed nutplate aligned with the jig. The other mounting points are on the wheel box and the rear bulkhead, the rear bulkhead needing a lot of stiffness (holds the elevator bell crank) so a glass gusset is put behind each point. Glassed in flange nuts were used for these mounts.

May 07

We may need a tail skid extension for winch launching, don't know yet, the calculations for the early part of the winch launch are difficult. So we include the brackets and bearing areas for a removable tail skid now as this is easy. In the bracket above, this is combined with the bracket holding the rudder torsion tube bearing and rudder control stops. This is really just another rudder hinge, like the bottom hinge on a normal glider, but the access to mount the bearing sleeve is poor, so we have an oversize tapered hole molded in the bracket. The bearing sleeve can be glued in later to match the other rudder hinges. I think I would better like a bolt on metal bearing mount / control stops combined, as long as I can get to the bolts and can see a way to fit it with the correct hinge alignment.

May 07

The rudder pulley will be mounted on a glass stiffener bracket. The cable plane was set and transferred to the centre of the bracket form. So pulley alignment should be good. The port bracket we just have to mirror from the starboard one and do an alignment check before gluing in after closing the fuse halves. The area where we made the bracket was first protected with low tack tape on the shell and some packing tape, then the completed form was covered in packing tape. The local vacuumed was sealed onto the packing tape. Seemed to work fine.

April 07

I'd like to add as much stiffness to the fuse shell as possible before demolding. So the seat flange, bulkhead stiffener, canopy sill box section go in now. The control column bulkhead, front wheel box and brackets for the pedals and canopy hinge bracket, or most of those probably should also go in before demolding the fuse shell. I've added what will be a 12mm return to the seat flange to try to stiffen it, and we will probably have some small bulkheads between the flange and the fuse shell. We'll do those later after the controls, seat belt anchor and belly hook brackets are in.

We left the styro core in the canopy sill box to save time, and to positively locate the path of the rudder cable sheath. The port side nylon sheath slides in a masking tape tube to enable fitting after joining the fuse halves. The rudder control system and cable route was decided quite late so our prototype has the cable penetrate the carbon UD in the "cow horn" extensions at the top part of the fuselage.

March - April 07


Based on the mockup and some final drawings, the jigs were made for the vertical bell cranks behind the pilot and the rudder torsion tube. The later allows the rudder horns to be completely in the fuselage. A pulley will be used so that the rudder cable arrives at 90deg to the hinge axis. The steel is 4130, mostly 0.88mm thick. Most of the input and output arms are tube deformed to a flattened oval so a yoke is easily formed to accept the spherical rod end. Yoke doesn't need to be welded on. Gyu Tae has taken a job in Australia so we gave the welding job to Eddie Larking. Looks good.

The bell cranks have a turned fitting welded in each end. A 6mm rivnut goes into each of these to bolt to a 626 ball bearing. Like the APIS I think. As we assume the welding will give some distortion hence bearing misalignment we trued up the ends of the bell cranks in a lathe. So the flange faces on the rivnuts should be parallel. The airbrake and aileron cranks sit together so they need to be the same length.

Jan - Feb 07
Joshua Daisey arrived from the USA to help build. A keen glider pilot and AP mechanic teacher with some construction experience.. Everything looked good but a bad mountain bike accident put him in hospital and meant he had to return home. I hope he gets to return later.

Nov - Dec 06


We decided to make a mockup of most of the controlls in the fuselage before closing the shells. Much of the hardware will also be built and installed before closing. The control system had sofar been designed using calculations for the kinematics, sized by load costraints and drawn with 3D sketches in Autocad. So the mockup is partly a final kinematic tryout. The 'bell cranks' were made with aluminium tube glued together but are generally true to shape. The push rods use mostly real hadware and bearings. We tried for 2 straight rods to cnnect each control function to the cranks instead of the usual 3. This was difficult but did simplify the system. We still have to finalise the rudder control design, either a Libelle style gimble or pulleys to bring the cables orthogonal to the hinge line.

Nov 06

Julien worked on the demoulded wings. The flashlines were ground off and TE was cut and ground. The wing root/root rib was ground flat. The aileron cutout was made using the lines made by the tape marks in the mould. Then the glue at the rear spar/ aileron cutout was ground smooth. Small inspection holes for the aileron bell cranks were cut, moulded seats for these were made and then the seat, with cap, was glued in. There was some glue ooze from the wing closing restricting the aileron bell crank which was ground back to give safe clearance.

There were some spots of delamination of the surface 100g glass on the wing and this does seem to peel too easily. It looks like we will have to peel this off the wing, which will lead to extra work. Looking at the wing weight, we are already 36kg, so are heavier than planed. One person can lift a wing, but it's not really easy. Not quite a "Sparrowhawk moment".

Oct 06

A couple of days later the wings were demoulded. On removing the upper mould we could see that the LE glue squeeze was good and was never exceeding it's available squish space, which we previously thought might be undersized. Overall our result is very good considering the learning we have had to do. There was a small delaminated patch of surface glass (just visible on the upper port), and some tiny occasional pitted spots in the surface glass. Still wondering on the exact cause for that. Now our attention turns to the fuselage, things to do before closing those shells.

Oct 06

Many tasks remained before closing the wings. Most critical was the final
check of the glue gaps by closing the moulds to squash small spots of plasticine. Low tack masking tape protected the surfaces. The gaps looked good, matching the spot measurements done earlier, and no grinding was required. Most of the glue application was approximated from these gap measurements. The spar glueing flange amount however was calculated for each meter of span so we had an accurate amount to avoid a big mess in the wing (this method did seem to work). Finally, after careful preparation the glue was applied and the
wing closing was quite easy. The glue application time was a concern in
the planning, but it took only 2-1/2 hours and for the second wing this was just two guys. The glue squeeze visible through the inspection holes looked good. The starboard spar web glue looked a little low at the front so Gyu Te made an aluminium tube aplicator so we could syringe some glue in at the last minute. Special thanks to the wing closing helpers: Murray, Julien and Gyu Tae.

Oct 06

The aileron bell crank brackets were made. On one side an aluminium bracket bolted to the wing main spar web, on the other side a moulded glass bracket to glue to the bottom shell. A teflon cable sheath was taped to the shell to allow later development of a tip drag rudder. For the wing tip joiner a part rib was made to later provide the bearing for the joiner tube.

Oct 06

The rear shear fitting at the wing root was fabricated. A wooden block for the shear flow, a nut plate and a foam form. Then a glass shear wrap around that. A PVC closeout allows connection to both the inner and outer laminates.

Sept - Oct 06

The new rear spar was made with rebates for the carbon hinge plates to glue in and allowance for local reinforcements.
The old wooden mould was used but some mods were required. Once made the new rear spars had their glue clearance checked and were glued in. Linear bearings for the aileron were made up using the design developed earlier, but with acetal balls, then fitted and taped to the wing shell. The aileron push rods have side loads at the wing root so a 3 wheel bearing will go there. There is enough room to mount this externally where it can be maintained. After testing the side loads on the push rod we added another linear bearing about 300mm from the root to increase the buckling safety.

Sept 06

The aileron bell crank and push rod positions were mocked up to consider the best mounting options. A slightly non standard,
bolted metal half bracket vs a fully composite bracket with shear loads going direct into the shell inner laminate. The leading edge
joggle was checked for glue gap using the joggle moulds to squeeze the plasticine (they were first translated back into register
with the plug surface). Some local reinforcements were added to the shell at the aileron cutout, rear spar area and at the hinge
mount points. Also did the PVC closeout and shell reinforcements for the exit rod to the aileron horn. Now the shells are ready for
the rear spar, but first the rear spar has to be rebuilt with the new dimension for glue and the new hinge mount design.

Sept 06


Julien Dorol arrives from France. Some carbon UD test hinge plates were laid up. Some UD was wraped around the mandrel
first as we wanted to try pressing in some split self lunricating DU bushes. Sofar this looks promising. Julien finishes the small
mods, reinforcements, lock nut mountings on the brake boxes. The brake push rod was aligned, the bearings glassed in place
and then the brake box glued in. The brake push rod can be fitted or removed after the wing closing without surgery.

August 06

The brake box ends were modified and some stiffeners and bearing hardpoints were aded to the sides. A glueing flange was
added at the bottom of the box at the arm spindle locations for better load spread into the bottom shell and better glueing. A tube
flange for the push rod seal was made and fitted. The bellows will be short, with a sliding nylon bush in the end. The metal
hardware for the wing was painted with etch primer then enamel. The bearings were pressed into the brake arms. A simple
sleeve and collar made this easy. The brake arm spindle holes were drilled in the brake box with a basic jig. Then some MDF
templates of the brake leaves were made from the drawings and some mods to the shape of those were made so they fitted
propperly. Then some top leaves were made in aluminium. These may yet prove to be 2nd generation templates. The basic
movent and clearance for the arms and leaves was checked. Also tweaked the top leaf angle just as the cap will close.

July 06

There has been more refinement to the design of the control system and internal hardware. In the end we have kept with fairly simple ideas and used 4130 steel, some folded sheet and tube. Engineer Gyu Tae Kim is shown TIG welding the brake arms.
The bearing misalignment due to heat distortion seems to be small. Martin at Airmaster (Kiwi company makes adjustable pitch prop hubs) did the bearing housings for us on his Mazak CNC. He also made all the threaded aluminum fittings for the ends of the push rods. These, with a spherical rod end, will be used almost everywhere.

June 06

The root ribs were jigged to control the glue gap and finally glued in.
The calculations, design and buld process for the rear shear fitting at the root rib were reviewed.

Feb 06

Visitor Samuel Dupland from France helped fix and prepare the main workshop and our old lathe,
needed for making some small parts. The root ribs were laid up. with a molded inspection hole.
Some FEA showed the strains working around the hole(s).
We also needed to check and modify the glue gaps to the wing shell a bit.
The root rib push rod exits will be made later. They can be solid plastic bearings with glass fiber bearing
housings patched onto the root rib. For the nylon ball type linear bearings used everywhere else we made
our own nylon cages. Some drawings and experimenting with cage geometry was useful.
A step drill was used to give a retaining step so the balls don't fall out.
The cage has a sliding contact with the outer sleeve. Unfortunately the
outer sleeves will probably have to be aluminum with WT=1.47mm.

Main Construction Page