Percival Vega Gull

[Ron Gray]

1 hour ago, John Rickett 102 said:

it will be good enough for what is essentially a sport model.

What! 😲 I'd love to see a scale model of yours if that's the case John 🤣

[Richard Crapp]

  The F/S ones were metal screwed on to  the fuselage.

I used  'fairing compound' from Bucks Composites. I ran two layers of electrical tape along both edges to provide some definition then sanded in the shape. you can sand down to the tape for a fine finish .

 

As Ron said; Sports model ??  !!!!

[John Rickett 102]

Thanks for the compliments, unfortunately for this project I haven't been able to find any reliable drawings, most of the design has come from photographs - an inherently unreliable method as the pictures are rarely taken at right angles to the subject, so I doubt it would pass muster in a scale comp.

 

Thanks also Richard for the tip about fairing compound, I'll probably use that over the top of blue foam which will form the bulk of the fairing.

 

A trial fit of the 'test' foothold. This would definitely not pass muster as I'm sure the floor should be level with the ground, but I couldn't get it to fit any deeper as the inner flap was in the way. For the one further forward there is nothing in the way but installing one foothold level and one at an angle may look a bit silly so it may be best to keep the same (incorrect) angle.....I said earlier its a sports model!     

 

 

 

[Richard Crapp]

I meant to add that I filled in the blanks on my model with the Brian Taylor plan .

Probably the only plan artist that can be relied on for accurate shapes and outlines.

[John Rickett 102]

It was fortunate that Brian Taylor produced a plan for the Gull 4, I've not been so lucky. I have a book which describes the evolution of the Gull series including the widening, stretching and enlarging of the areas and control surfaces from Gull 4 to Proctor although there are no meaningful drawings. 

 

With the fuselage inverted in its cradle the wing centre section could be placed in position and checked for alignment before the dreaded no-going-back-from-here operation of gluing.

 

 

After allowing the glue to dry overnight a milestone was reached with the model being able to sit on its wheels unaided.

 

 

The next jobs are to blend in the line from the bottom of the cowl to the main spar and also make the wing fillets. Completing the footholds has had to be delayed until some gelcoat activator arrives.  

[stu knowles]

I am really impressed by your production of small fibreglass parts such as the foot steps on this model and the cylinder heads on the Se5.  Until you showed the way I would not have considered glass parts for anything much less than cowls and wheel spats.

 

Can I ask what resins you use for the moulds and for the parts themselves.    I have made mould for a large cowl, followed by the cowl itself and used skinning epoxy throughout - which is expensive.  I did wonder after if I should have used a polyester resin for the mould, which would have been much cheaper and then just used epoxy for the production part.

 

I would appreciate hearing your take on the different resins

[John Rickett 102]

Thank you for the compliment Stuart.

I use epoxy resins for everything, buying from Bucks Composites usually as Graham Buckingham trades at shows and provides prompt and helpful service when ordering online. His Formula 50 or L285 will cure in 24 hours at room temperature (20C), though probably best to leave for a couple of days before cutting back to a smooth finish with wet & dry cloth.

Stored in a dry, warm place part-used bottles or cans seem to have an indefinite shelf life. I try to be careful with measuring so that not much is wasted and really don't find it an excessive expense, particularly in the overall scheme of how much a model costs to make!

[John Rickett 102]

Four footholds have now been taken from the plug and the excess dry cloth removed. Each took about 15 minutes to lay up and then a wait of 24 hours before one could be released from the plug and the next one started.

 

 

The excess depth and length of the plug allowed for easier handling but most of the fibreglass needed to be cut away leaving quite small parts to be glued into the wing, this will be done once the wing is glassed. The combined weight of the castings is 5 gm, I’m well pleased with that as I doubt another method would have yielded a similar practical, lightweight result.  

 

 

 

I had assumed that the bases of the footholds would have been at the same angle, parallel to the ground, on closer inspection it appears they are different, the rear ones slope upward more than the front ones. The front ones can now be made deeper………an exciting revelation for an anorak modeller.   

 

 

Blue foam would be a suitable material for the wing fairings, or so I thought, however having cut some and tried to bend it to follow the rib outline I found it to be too stiff, it’s also quite difficult to sand without galling. Instead, the eurothane core recovered from Recticel flooring insulation (same as the cowl plug) was used. This cuts, bends and sands easily while still being light.

 

All but the curve around the leading edge was formed from the eurothane, the front part was completed with lightweight fairing compound mixed with fibreglass resin. Once cured this can be sanded but it’s quite a hard material, it’s easy to sand away the softer adjacent balsa so care must be taken.

 

The first intention was to apply filler to the surface of the eurothane and sand down to feather edges but rejected that idea given how hard the cured filler is, compared to balsa. Instead, the fairing was covered in 2 layers of 50 gm glass cloth, one layer just covering the fairing and a wider piece overlapping onto the fuselage and wing. This provides a resilient surface on the fairing and allowed for easier blending at the edges.

 

 

As the fuselage is to be covered in Ceconite and the wing centre section in glass cloth, I wondered how best to achieve a neat join between the dissimilar materials. A flash of inspiration was to mould fairings on top of the existing fairings. By being removeable, at least initially, the Ceconite/glass cloth join could be hidden under the top fairing and, being fitted after painting, would definitely look like a removeable item. Cutting the moulding into 3 sections (the same as the fullsize) and adding some small screws at about a scale 6” pitch, glued into pre-drilled holes will hopefully complete the illusion.

 

 

Here, the area to be moulded is masked off with release wax being applied - 6 coats, buffing in between, usually does the trick.

 

 

2 layers of 80 gm cloth made the fairing, it seems to be of adequate rigidity so not worth adding more layers when the other side is done. The fairing weighs 10 gm, it now needs a further coat of resin to fill the weave then it can be rubbed down with wet & dry paper and trimmed to final size, the black outline.

 

Given how easy this was to do, I’m motivated to do something similar for the wide straps which cover the wing fold.      

[John Rickett 102]

 

Returning to the filler, lightweight fairing compound is as the name suggests light, however the resin that has to be mixed with it, isn’t light. I tried a comparison test of cured filler versus eurothane. A convenient block of eurothane lying on the bench measured 3505 cc and weighed 95 grammes. I patted some left-over filler into a block which measured 2.45 cc and weighed 3 gm. My schoolboy maths calculated that, by volume, the filler is 15.5 times heavier than the eurothane. This doesn’t take into account the protective covering of fibreglass required, though I’m pleased now the eurothane was handy!

 

 

Attention has reverted to the wing outer panels to get the left panel to the same stage of completion as the other. Part of the process was to get the dihedral the same. Having set the right hand wing, it was a case of adjusting the left wing to be the same.

 

 

The coarse adjustment, if it could be called that, was bolting the inner parts of the hinges in place and then with the M4 bolts inserted, chocking the wing to the correct angle. The fine adjustment was to position the outer parts hinge on their ply supports and drill one hole through the ply and insert its M3 screw, then release the chock so that the wing took the weight. An error meant making another hinge part with the hole slightly offset, then drilling through the second hole to lock into position. Fortunately I only had to do this once, both wings now have a dihedral of 3 degrees, measured along the centre line of the ribs. Its actually difficult to know what the true dihedral of the wings is as the washout tilts down the leading edge and therefore affects the measurement – provided they are the same I’m sure all will be well. It’s possible that the dihedral is less than scale, however as a gyro (wonderful things!) will be part of the radio installation I’m not particularly concerned about a lack of lateral stability.

 

The picture above shows the model inverted with the bolts in position, lots of balsa will be required to fill the gap and perhaps some plastic padding as well to close the gaps completely.

 

 

A great tool to use for making sure all is aligned, is a laser level. I had bought one of these when carrying out some DIY electrical installation a few years ago and have found it invaluable for checking squareness during model building. Here, the wings are being checked for equal dihedral.

 

 

Before getting too far ahead with the wings, the landing light recesses needed sorting out. Although far from scale, modern LEDs provide a source of decent light intensity and can be powered by a small 3s lipo. I’ve used encapsulated lamps before, the type with a G4 (2 pin) base, these work well but are relatively heavy at 17gm for the 36mm diameter ones. This time I’m going to give surface mounted Cree LEDs a try. These lamps are supplied without reflectors so the plan is to line each recess with plasticard and  spray the whole area with chrome paint to enhance the brightness. Beryl Markham’s aircraft, or more accurately the one she borrowed for her Atlantic crossing, didn’t have landing lights fitted, nevertheless I think they are a nice touch when on an approach, so will be incorporated.

 

 

As much as I enjoy the summer months, being in the workroom when the weather takes a turn also has its attractions. Surely those who don’t build are missing something in the hobby, or perhaps not, given that building isn’t nearly as popular as it once was – I’m convinced though that it has a therapeutic value.        

 

 

With the dihedral set and (hopefully) confident that there is no further adjustment necessary to the hinges, completion of the wing sheeting and gap closing can continue. Here the wings are attached and the right wing gaps filled in. Some filler is required to get the sheeting level across the join the same.       

 

                         

 

 

 

 

 

[Richard Crapp]

I think the dihedral of the outer panels is 5 deg.

 

A freind has visited my subject aircraft in Australia and sent me 80 nice  photos. So anything you want ?

[John Rickett 102]

Thanks Richard. I’m unsure how similar the Gull 4 is to the Vega Gull. The Gull 4 evolved into the Gull 6 and then into the Vega Gull. Both iterations involved widening and lengthening, although I don’t know if the dihedral changed during the process, possibly not. I have a scale drawing of the Proctor, which was a further development and if this drawing is to be believed, the dihedral was 5 degrees when measured along the bottom of the wing. Of course, as the wing thickness tapers along its length, the dihedral would be less if measured along the centre line and even less if measured along the top of the wing.  From measuring along the bottom of the model wing, at the main spar position, the tangent is 4.33 degrees. When comparing the model to photographs it doesn’t look to be wildly out so I’ll happily live with the probable reduced dihedral.     

 

As exciting as it is to watch glue dry, there are jobs that ought to  be done during the interim to maintain progress. In order to keep the radio equipment out of immediate view,  yet still be accessible, the main components have been mounted on a removeable tray which sits at floor level below the instrument panel, the batteries are attached to the underside. Beryl, the instrument panel and the tanks are all held in place with small neodym magnets so it should be a relatively easy job to gain access to the radio gear when required.

 

 

 

A couple of wiring looms were made up using Ashlok connectors, principally to make the wing connections a single operation at the field, the connectors also ensure I don't make the silly mistake of crossover errors with flaps and ailerons.

 

 

There is a gap of approximately 3/8” between the ribs, this was closed by adding a ¼” profile to the centre section rib and 1/8” to the outer section, then gently swinging the wing and sanding down where necessary. A strip of paper acted as a feeler gauge to locate the exact point of binding, The top was done first, then the bottom to make sure I knew exactly where binding was occurring. A bit more may have to be eased later all the way round to allow for the overlap of the covering.

 

 

 

 

 

With the outer wing portions seated, gaps closed and footholds fixed into position, the underside of the centre section has now been boxed in completely plus a trial fit of the fixed half of the spats to make sure all is well.

 

[John Rickett 102]

With the bulk of the work on the fuselage out of the way which required it to be first right way up and then upside down, attention can be turned to attaching the tailplane and fin.

The laser level comes into its own once again for checking squareness. Here the fuselage is checked for level on the board, then by raising the tripod mast the tailplane can be checked.

 

 

 

Similarly, the fuselage is chocked up and levelled along its datum line (the upper stringer), then raising the tripod mast checks that the tailplane incidence is at 0 degrees – I’m quite pleased with that, better glue it while its looking good!    

 

#

 

[John Rickett 102]

The fin was aligned using the fool-proof laser level.

 

  

 

With the fin glued in position and the rudder structurally complete, I thought I’d have a try at using China Silk for the covering. During application I tried to keep the wrinkles at bay as earlier testing had shown the shrinkage of this material is not as great as other polyesters.

 

 

There are a couple of wrinkles that would not budge with a hot air gun and somehow, I failed to notice prior to covering there was a glue blob (the red circle) so the decision was taken to remove the covering and go back to something I know works well.

 

 

I still have off-cuts of Koverall so used some on the rudder – a much better result.

 

While I had the materials to hand it seemed sensible to cover the elevators as well.

 

 

Both the rudder and elevators have shrouded hinges, though the ailerons don’t. I can’t find a reference as to why that is, but thought the model should adopt the fullsize practice where possible.

 

 

The shrouds are simply strips of 1/64” ply each with a strip of balsa on the inside, sanded to a triangular section.

 

 

Unusually for a light aircraft, the fuselage/tailplane/fin junction has a fairing. Percival went to some lengths in reducing drag on the Vega Gull. Now a four-seater and although wider (to accommodate the extra seat) and heavier than the Gull 6, with its increase in centre section area to maintain the same wing loading and increased empennage to keep the same handling characteristics, it was only marginally slower than the Gull 6. Improvements included enclosed control pushrods and a revised undercarriage which allowed slimmer spats. This obviously paid off as, with its range and cruising speed of 158 mph, the type became the aircraft of choice for many record setters and record breakers in the heady years preceding the second world war.

 

 

For interest, in 1936 Charles Gardner, flying a Vega Gull, won the King’s Cup Air Race around Britain with an average speed of 164 mph on the last (312 mile) leg of the course. The race distances varied each year but in 1936 it was 1,380 miles. In September of the same year, though after Beryl’s flight across the Atlantic, Charles Scott and Giles Guthrie won the Schlesinger Portsmouth to Johannesburg Air Race. Beryl Markham was aware of the aircraft required to attempt her crossing of the Atlantic and fortuitously had as a benefactor, Lord John Carbery, one of Kenya’s Happy Valley set and an acquaintance of Beryl from her upbringing in Kenya. He was about to take delivery of a new Vega Gull and was willing to lend it to her (its recorded that it was his suggestion she attempt the crossing) on the understanding that it would be returned in time to allow him to also compete in the Schlesinger Race. The die was cast.

 

Replicating the tail fairings was similar to the wing fairings, eurothane was again used as the basis. After sanding to the final shape one layer of 25 gm fibreglass was put on to provide a hard surface.  There are three fairings for the fin, one extends over the frontal area and another either side. It seemed logical to make them as three items as well.

 

 

Parcel tape was applied to the front part first and given one coat of release wax to ensure the parts would separate easily later.  Small strips of 50 gm cloth were then applied and left to cure for 24 hours after which the cast was removed.

 

 

The fairings are only attached to the tailplane or fuselage, not the fin. I presume that’s because pitch trim is achieved by varying the incidence of the tailplane, hence the fairings can only be fixed to one part of the structure. There is a substantial gap around the fin so for the model, 6 layers of masking tape were stuck to the fin plus the parcel tape, to try to make the result ill-fitting!

 

 

Both the fin and tailplane were partly covered in fabric offcuts, which could be stripped off later, with parcel tape on top to allow forming of the fibreglass covers. Two coats of release wax were applied for good measure, wax isn’t really necessary with parcel tape although makes lifting of the cured, delicate parts a little easier and only takes a couple of minutes to put on.  

 

 

The next step was to rub down to a smooth finish and fill any blemishes with cataloy paste. The desired result was a fairing which looked as if it’s fashioned from metal on top of a fabric covering.

 

Here, the front part immediately after moulding…….

 

 

….and once the rough edges had been trimmed.

 

 

The fin side mouldings didn’t come out right first time. Once marked up for trimming, it was obvious the tabs, which cover the gap around the elevators, didn’t extend far enough back or low enough.

 

 

When making the rudder shrouds (attached to the fin) I only made them deep enough to cover the fin during its construction, I hadn’t thought about the extra bit below the fin to the bottom of the rudder. The first idea was to add ply/balsa fillets the same as the upper parts, then thought it would be neater and possibly lighter to extend the fibreglass tabs to the bottom of the rudder.       

 

Here, the fin side fairings have been extended, trimmed and fitted round the elevator joiner (which is concealed under the fin).

 

 

The elevators have been made in two halves and are joined (screwed together) in the small space that exists under the fin. Making the side fairings removable, as with the full-size, solves the problem of assembly and future access to the elevator horn. Just now the fibreglass parts are not attached, once painted the front fairing can be glued on.  

[J D 8 - Moderator]

 Heck, I have one of those laser levels, never thought of using it for model work. 

[Richard Crapp]

'Percival went to some lengths in reducing drag on the Vega Gull. Now a four-seater and although wider (to accommodate the extra seat) and heavier than the Gull 6, with its increase in centre section area '

So do you think the increased span came about by inserting extra down the middle of  the cockpit thereby leaving the outer wing panels the same?

 Lovely work you are doing there John.

 

Just for interest here is a picture of the Gull 4 with tail removed. You can see the winding mechanism to the trim jack and a few other interesting bits.

[John Rickett 102]

Thanks Richard, we're getting there slowly.

According to my reference information, when the Gull 6 evolved into the Vega Gull the cabin was widened by 9 inches and the centre section by 15 sq.ft, this to maintain the same wing loading at 14.5lb/sq.ft. The changes brought the span to 39ft 6ins.  

[John Rickett 102]

The surface mount LEDs for the landing lights arrived safely across the high seas from China so thought I’d see how bright they were and also think about fitting them in the recesses. I was quite satisfied with the brightness especially if they were doubled up in series.

 

 

Each LED is rated 3w at 3.7v. By using a 1000mAh 2s lipo (7.4v, which provides the ideal voltage) no regulation is required. The current drawn is 0.94A, which doesn’t tally with my schoolboy science formula of W=IV so I imagine the stated 3w is optimistic, however the battery should power the lights for many flights between charges. Given that the lights are only on for a minute or so per flight, the little battery should be fine.

 

At first, I wired a couple using red/black wire but thought it intrusive so tried again using white wire. Although far from scale, I wanted to include LEDs on this model and had the brainwave that adhesive backed silver foil would make a simple reflector, but having tried it decided it was too amateurish so went back to the original idea of chrome sprayed plasticard. With the basic set-up complete, the LEDs were stuck with double-sided tape so that a test of the lamp wiring and control could be conducted. It worked but melted the double-sided tape quite quickly…a Plan B was required.

 

With some fairly thick litho plate (0.33mm) a backing plate was made and the LEDs stuck this time using JB Weld, which claims to be able to withstand 288C. That seems to work as a heatsink, leaving the lights on for 2 minutes did generate heat, but not enough that the backing plate couldn’t be held in the hand.

 

Having lined the recesses with plasticard, each area was masked and then sprayed with gloss black paint. The instructions for the chrome paint I have is that best results are achieved when sprayed over a glossy black base.

 

 

Once the paint had dried, a dusting of the ready-mixed chrome paint was applied.

 

 

 

I think the process works quite well, this was done using a cheapo non-branded Chinese airbrush.

 

 

For small areas which includes lettering and striping these airbrushes are adequate. Provided they are cleaned properly after use, something which applies to all types and makes, they offer a way of getting good results for little outlay. I’d recommend using a compressor to provide the air, not cans, as even a small gun like this will consume air quickly. My one came from Machine Mart many years ago, I think it was the smallest in their range, its more than adequate for powering this and a touch-up gun which is used for larger areas.               

 

I bought a vinyl cutter about three years ago although up to now had only cut paint masks and vinyl lettering, having read that Richard Crapp had cut shapes in plasticard for his current project, I thought I’d try the same.

 

A simple bit of design and hey presto, the machine will cut the acetate landing light cover and the plasticard rim, both of these are for the right-hand lamp.

 

 

The flap boxes had been made, hinged and latches installed but I’d done nothing with how the little latches would be operated. There is only a small (3mm) bent wire tab which should emerge through the sheeting, the slots for which would surely wear very quickly unless reinforced. To strengthen the 3/32” balsa covering, a slotted piece of 1/32” ply was glued on the underside of the balsa and then the topside covered in 19gm fibreglass overall with a couple of smaller pieces of 50 gm cloth where the slot is.

 

 

 

After being left for a few hours to semi-cure, the slots were cut through with a sharp knife and then, once fully cured, tidied up with a small file. Hopefully now the reinforcement will prevent the slots from widening and looking a mess.

 

 

The next concern is how to prevent paint from getting into the slots when that time comes round….

 

Having been surprised in how glass cloth as light as 19 gm can impart strength, it gave me confidence in covering the underside of the flap boxes in the same weight material; the upper is open structure so will be covered in fabric. The flap boxes seemed to be quite rigid prior to covering so only required something as a base for the finish. 6 gm of resin was all it took to get to this stage and that also included covering the areas for the latches and servo cover.

 

 

After 24 hours a light rub around the edges with a sanding block will cut the cloth. Using a sanding block rather than a knife will trim the cloth at the very edge leaving nothing to clean up afterwards. The next time fibreglass resin is required for another job, a second coat of resin will also be applied here, there’s no point in mixing resin especially for this one job as the small amount required would just result in waste – something I’m averse to!       

 

    

 

 

The model has got to the stage where its structurally complete, so time for an assembly to check the fit with all the main bits in position. Not much room to move around with the wings on so it won’t stay like this for long!

 

 

 

The undercarriage oleos from John Brooks are a work of art, it seems a shame to cover them up, I just hope my landings do justice to them.

 

While there is still a fair way to go, the next tasks are cockpit glazing and making up the control linkages, particularly the closed loop runs for the rudder and tailwheel, then attention can be turned to covering.              

[RottenRow]

33 minutes ago, John Rickett 102 said:

Each LED is rated 3w at 3.7v. By using a 1000mAh 2s lipo (7.4v, which provides the ideal voltage) no regulation is required. The current drawn is 0.94A, which doesn’t tally with my schoolboy science formula of W=IV so I imagine the stated 3w is optimistic

 

That current is about right John. Your LiPo is probably a little over 7.4V if fully charged, which will put the current and hence the wattage up a bit. It works out a bit over 3.5W per LED. If you wanted to reduce the power to 3W per LED you could add a resistor in the wiring; a 1 ohm resistor, rated at at least 1W, for each wing would do the trick.

 

They do look impressive.

 

Good to see the progress that you've made.

 

Brian.

[John Rickett 102]

Ok Brian, Thanks for that.

The LEDs seem bright, but may not be quite as good outside. I'd like to leave them as bright as possible though don't want to burn them out, I'm hoping that they will last the lifetime of the model, once the model is covered they won't be accessible without surgery.

Do you think that a full charged lipo will be doing them any harm, if so I'll add the resistors as suggested?  

[RottenRow]

Over-running the LEDs will reduce their life but as they will only be powered for short periods that won’t be a problem.

 

Also as the LiPo discharges its voltage will reduce, lowering the current and therefore wattage.

 

If you were to fit resistors, the best place to fit them would be near to the LiPo rather than in the wings, so they would be accessible at a later stage. Or fit them on the microswitches if you are switching the lights on with the flaps.

 

Brian.

[John Rickett 102]

Thanks Brian, it was food for thought.

I don't have any 1 ohm 1 watt resistors just now but do have some 0.47 ohm 3 watt so tried an experiment. To get a scientific (!) baseline the top wing was connected to a 7.4v lipo and the bottom to a 6.6v life battery. There is a noticeable difference to the eye although its not so obvious in the picture. To try to quantify the difference, I used a 35mm SLR camera with an exposure meter. The lipo powered lamp indicated 1/4000th at f8, the Life 1/2000th at f8. Clearly the exposure meter could detect the difference. I then connected both lamps in parallel and checked them with the lipo connected and with the resistors in series (near enough to 1 watt) this measured 1/5000 at f8. Measuring again without the resistors, showed 1/6000th at f8. From this I've deduced that inserting the resistors is similar to running on life batteries. 

I'll charge both the lipo and the life batteries, then tomorrow (its supposed to be dry and sunny here) see what the brightness is when outside.

It may be that a life battery is bright enough, in which case it will be a safer bet to leave in the model and save a few grams at the same time.   

 

      

Edited January 7 by John Rickett 102

[RottenRow]

Yes I’d think a LiFe battery would be the best option, both from the lower voltage and greater safety.

 

Brian.

[John Rickett 102]

I was pleased with the lightweight fibreglass covering of the flap boxes undersides so covered each of the flaps in the same lightweight (19 gm) cloth. Prior to covering it was possible to twist the flaps if enough force was applied, but it takes far more torque to put any kind of twist in now.  

 

 

Brushing out the wrinkles during cloth application seemed to give a flat surface, though during the curing process some wrinkles would creep in again. All that could be done was to sand flat, apply patches where the cloth has been sanded away and then blend in – effective but time consuming. I’ve since discovered that its better, once the cloth has been completely wetted out, to tug at the edges as if to stretch the cloth, perhaps only a millimetre or so is gained but its sufficient to take out any invisible slack and prevent the wrinkles forming.

Previously, my normal preference for covering a largeish model was fabric, especially if the original had been fabric covered. Having now experienced the strength, rigidity and minimal weight gain of really light glass cloth and resin, I think it’s the way to go in the future for those parts, like flaps or fully sheeted tails. The flaps on this model are sheeted in 3/32” balsa but I believe now I could have used 1/16” and achieved more than enough strength.

Here a flap box has had a coat of resin brushed on, rubbed down with 180 grade wet & dry cloth (used wet) and then given a further coat of resin using the credit card squeegee technique.

 

 

After allowing 2 days to cure, 400 grade wet & dry cloth soon produced a surface to which a coat of primer/filler could be applied.

To further keep down the weight, the insides of the flaps and flap boxes will only be given a light coat of colour, if someone wants to try and peer into the flap recesses and find fault, that’s ok with me!                  

 

All the control linkages have now been made up. In trying to keep with Edgar Percival’s quest for a clean design which minimises parasite drag, the flaps and elevator rods are internal. The elevator rod is 3mm, that’s a bit of weight right at the back where it’s not wanted, but with the length of the rod being 7” and the elevator area at about 112sq ins, I didn’t want the chance of flexing to occur. The flaps and aileron rods are 2mm, which I think is sufficient given their short lengths. None of this has any scientific basis but gives me peace of mind. Ignorance is bliss!

 

 

 

The servo covers are sheet balsa covered either side with glass cloth.

 

 

Rudder and tailwheel are both closed loop using 80lb rated, nylon covered fishing trace. The ferrules supplied are large so instead I’ve used 2mm od brass tube. By squeezing the tubes oval, two cables (0.75mm od) will go through the hole. The ferrules were then nipped with side cutters and a spot of thin cyano applied. The cables run in 2mm id plastic tube, using tube allows for the cables to be fitted once covering and painting is completed.    

 

 

The tops of the cabin doors and aft cockpit have single curvature windows, on the model the radius is about 1”. I thought at first that 0.5mm PETG clear sheet would easily follow the curve.…until I tried it, then found that there is a lot of spring in the material which would be difficult to hold while glue was setting but also that the spring probably would, over time, distort the doors.

Another plan B was required. A phone call to Sarik confirmed they could vac form the windows if I provided suitable plugs. Their suggestion for ‘one-offs’ was for a plug to be formed from plywood, which had been coated with Upol Easy 1 or Isopon P38 filler sanded to a smooth finish. They further suggested that a plug should be raised a small amount above the finished height (a piece of wood fixed to the base) to allow for the fillet during pressing and that the sides should be bevelled slightly so that the formed sheet will release from the plug.

Armed with their suggestions a couple of plugs were made. Its possible that on the real aircraft, the front and back curves are the same, but on the model they have come out slightly differently, it’s not a lot, but an unknown to me whether the formed shapes would readily adapt to a slightly different shape. If the curves had been the same just one plug would have been needed but I decided to play safe and make them handed. A DIY electric planer produced the rough shape which was then finished with an orbital sander.

 

Here, the two plugs prior to coating with P38, are lined up with the doors.

 

 

The coated, primed and sanded plugs.

 

 

 

Its been reported by she who observes such things that all the surfaces in the kitchen have taken on a faint grey colour - who would have thought that P38 dust would travel far! 

In discovering how difficult it was to curve the side windows, I thought I had better check that the curved portion above the windscreen panels would accept the PETG sheet. The conclusion was - with difficulty, so the decision was made to make a plug for that part as well.

 

 

The plugs are all ready for a coat of gloss paint, once the weather cooperates.   

 

Fortunately, the five panels which make up the front windscreen are completely flat so should be straightforward to glaze. As an aside, Percival introduced a blown curved windscreen later on, as fitted to the Proctor, though some owners didn’t like the visual distortion and had their aircraft retro fitted with the flat panels windscreen. 

 

[RottenRow]

Excellent progress John.

 

Where do you buy 19g glass cloth from please? The lightest that I found was 25g, which I have just used to cover my Magister wing.

 

Brian.

[John Rickett 102]

Thank you Brian, I get most of my supplies from Bucks Composites. I've just checked the site though and the lightest advertised just now is 25g. You could try giving Graham a call to see if he's going to restock. There is an ebay seller here asking £12.15 for 5m of 18g - that seems a good price.