MIKE ROACH’S SOPWITH AVIATION COMPANY

Mid-2007.  Having followed the building of Trevor Hewson’s Ivan Pettigrew designs Short Sealand and being at the first flight, when it was obvious that she was a beauty, I looked for a twin for myself. Ivan’s plans are simple, make lightweight models and are good value for money. His website has some interesting video clips of his models in the air (even a looping Blackburn Beverley, if that’s your thing) so I sent off for the plans for his 84” span Catalina. After a very short time, they arrived and I ordered what I hoped was enough balsa and hardware from Flitehook.

On looking round the ‘net for colour schemes I found lots of pictures at Airliners.net with my favourite being the the lovely version here, which has camouflaged upper surfaces with the remainder white, and sufficient weathering to keep my airbrush happy. The enlarged rudder is not on the plan, and the distictive gun blisters were not on the aircraft at this stage (although they were added later), so some decisions will have to be made later in the building programme.  There is also a very attractive American version in dark blue, white and yellow which would show up well and look good. Ivan’s plans are for “sport scale” models and having seen the Sealand look fabulous in the air and close-up on the ground, I must resist the temptation to “improve” the model!

I’m hoping to fly the model at Howard’s Field, off the grass, just like Trevor does with the Sealand, and also at Christchurch Harbour when conditions are right.  I wonder what the effect of salt water will be on the airframe.....

Pause for the sailing season to finish...

Dec 07.  Ordered two Hyperion Z2213-24 brushless motors, two 25amp SCs (400265). and a programming card (400232) to go with them, on Trevor’s advice.  This will be the same set-up as he has in the Sealand, so by Christmas I will be brushless and cashless.  Cut the wood for the tip floats, and will start building over Christmas. The motors arrived two days later, an amazing performance during the Christmas period.  Laid down the 7’6” main spar on a specially-bought 6’ pine shelf from B&Q.  The whole wing build depends on constructing the centre section on the spar, then propping it up and building the wingtips.  The section looks like a Clarke Y, so the process will be very easy despite the size of the beast. It certainly beats fiddling around with 1/32 square on an indoor Sopwith!  Ivan specifies a flat-bottomed section with a “NACA Cuff” on the outboard sections, to prevent tip stalling, but I’m going to build in 1/4” washout instead.

The tip floats went together easily enough and the tricky-looking scallop on the undersurface required no work at all, as it just formed itself when the sheet was bent to the profile. I had to keep telling myself that the floats take no real dynamic load at all and are just there to support the wing at rest, and to look pretty when being retracted. In other words, if they snap off, I’ve done something seriously wrong!

2 days before Christmas. Took advantage of the Queen on telly to nip upstairs and start the wing, having cut out all the centre-section ribs, decided on the position of the servos and the route of the snakes and the motor and SC loom.  In my experience, unless you cut the holes for snakes and wiring in the ribs before you start the building, you get a mare’s nest of rubbish splintered holes in the ribs, which have to be hidden by copious sheeting.  This is not going to be that sort of model... The wing section, although supposed to be Seelig 7055 is going to end up flat-bottomed (as are the tip sections) for ease of construction.

Would you believe the centre section was 0.5” wider than my 42” balsa? I’m very tempted to build it the width of a sheet of 1/16th and have done with it. A fancy scarf joint just to add 0.25” each side of a standard sheet is wasting time on a sport scale model, surely?  And who would know? Apart from you and me of course. I also find that Ivan’s spars are not quite parallel on the drawing. I shall build by the ribs I’ve cut, not by the plan!

28 Dec. The centre section went together pretty easily and I now have the tips framed up, but not after a lot of re-drawing and cut-to-fit. I found the ribs just didn’t match the spars, so have a complete new set on paper to add to the plan when I have time.

3 Jan 08.  I got on with the tailplane and fin assemblies over the New Year, after having time off for an afternoon at Howard’s, flying the little indoor Spitfire and Walrus and then tightening the bicycle clips for 15 minutes with the Piper Cub. It was a challenge after not flying much since April, but all went reasonably well and apart from one slightly short stumble into the weeds at the north end of the main runway, I managed to get her lined up and in the right place for each landing. The only repair was a bent UC leg. A record.  As an aside, before flying I charged up the Cub’s 2S and 3S 3700 mAh LiPolys (for the first time since April) and they only took 500 mA before the charger cut out, so they keep their charge pretty well.  After 15 minutes flying they needed about 1500 mA each. Half an hour’s flying would be up agianst the bump stops, but 20-25 minutes seems well within their capacity.

The Catalina’s tail surfaces are symmetrical and taper along the planform, so although the actual balsa work is straighforward, there is a lot of careful packing up to do before the glueing begins.  The photos show the use of plain rectangular ribs which are planed and sanded to shape after the glue has dried.  Chamfering the spars before beginning work might make the “post building” work easier, but I have always done this after construction.

The square top to the fin is correct for the version I am modelling from the Airliners.net site, but can easily be reverted to the more usual curved top if necessary. The lower half of the fin will be sheeted later, then the slot for the tailplane packed up to provide one degree positive incidence.

The “Master” paid a visit and made some nice comments and reduced the size of some of the hurdles that remain in the construction sequence by drawing out the wiring diagram for the motors, SCs and the retract servo. Thanks Trevor. But he raised one hurdle much higher, and I’ve got to have another look at the way the floats retract, and how they fit into the wings once they have done so. It seems I should have stuck to Ivan’s plan rather than making assumptions and sheeting the outer panels. Where would we plan builders be if all modelling was as easy as ARTF?  Back to top

Sheeting the hull has proved fairly straightforward, although the end result does magnify the inevitable small errors that creep in.  I tried to stick to best practice: avoid joints by using the largest piece of sheet possible; where joins are necessary, support the joint; use a paper pattern to establish the shape of the area to be sheeted and finally, work to an achievable degree of accuracy. The planing section of the hull is sheeted with 2mm medium soft (20 grams per sheet of 3” x 36), which was easy to “persuade” round the twist and curve at the front of the hull (the bow, we should call it)

The rather sharp compound curves at the bow make it difficult to sheet in one go.  After a rather half-hearted attempt I took the planking route - a skill I have not tried since a Keil Kraft control-line stunter about 50 years ago! I used a variation on the method described by Henry Holcome in Model Airplane News (easily found on a search of www.modelairplanenews.com) where alternate strips are chamfered.  The bow section of the Catalina has a 1:2.5 ratio from formers A to D, so all the strips were cut 3mm wide at the A end and 8mm wide at the D end, thus forming a fan of strips along that part of the hull. I hope the pictures make it clearer than the words.

 After scrapping a failed version of the cockpit hatch, a break from fuselage work was called for.  All the control surfaces are now ready to cover and the fin/tailplane joint, something I had not been looking forward to, turned out quite well.  I’ve departed from the text again and installed micro servos for each aileron and hard-wired the leads together into a long “Y”.  The float retract mechanism is far simpler that Ivan’s but retains the central servo (although in an email he said he left out the operating crank on his Minicat and had a straight connection to the servo).  The system works but, like others, I find that the last bit of “up” closure is hard to achieve. I suggest that a slight change to the angle of the float operating wire (the one that is glued and sewn to the float leg) might impart more force to that part of the operation.  After all, the “down” action is quite easy to achieve and not really critical.

22 Feb 08. I took a long look at the two servos driving the ailerons and decided to take it all out and revert to the single servo in the centre section.  A GWS HS81 Micro (£10.99 from Channel 4 models) has about 3kg/cm of torque, so should be more than capable of doing the job. Fitting the servo, pushrods and linkages took a couple of hours and the end result is much neater, although my pre-cut holes in the ribs took a bit of a beating.  I also feel happier that there won’t be any interference from the leads.  I promise to put the retaining screw into the retract servo output arm! That servo is recessed into the pylon top. I don’t feel too happy about it, but I can’t lie the servo down because the wing section is not tall enough to enclose the servo movement.

27 Feb 08 At last, a nice day and enough structure built to put it all together and admire it! As seen, it weighs 1lb 14 oz: let’s say 2lb 6 or 7 oz with the ailerons, floats, rear sheeting and cowls finished, making a target weight of 4 lbs seem reasonable. That will give a wing loading of about 11 oz/sq ft, according to Ivan’s website, and thus a very similar flight pattern to Trevor’s Sealand. 

For years I have used the very lightest filler on the market. It keeps changing its name, but is now called “One Fill Pollyfilla” and I got mine from B&Q in Christchurch. It blends well with balsa or Depron, dries quickly and sands very easily. The fin base now looks as wide as it should and had I thought of it earlier, I could have used the fairings to camouflage the rudder and elevator control rods.  As it is, they are “lost” when looking from the front, but pretty obvious from the side.

Back to Top

7 Mar 08.  I found the cockpit hatch rather tricky, and was going to abandon all the balsa work and make a foam “black windows” version, but for some reason found myself cutting 1/16th square from a sheet of hard balsa that had been in my spares box for a decade.  A few minutes later the outine of the cockpit was looking good, so after a couple of hours (and super-glueing my fingers to the wood at least twice) The version recommended by Ivan had taken shape. The fit lines are terrible and will need some disceet filling, but as an “objet” it works, and has the distinctive squinty look of a Catalina office. The pilot and his co-pilot are actually Harry Hawker and TOM Sopwith from one of my earlier models, carved from pink foam and painted up.  I will have to make something more warlike before fitting all that glazing. 
Two pairs of small magnets provided by Andrew Tubb complete the fittings, and the hatch snaps neatly and firmly into place, yet is easy to release.  Another hurdle overcome and the structure 95% complete. The Solartex is calling!

8 Mar 08.  Finally started the covering process, with some Litespan on the elevators.  These, together with the rudder, ailerons and wing, can be covered off the model. Litespan is funny stuff and quite difficult to get a smooth wrinkle-free surface even on large models. It’s worse on small ones!  I apply the Balsaloc to the film, not the model: this keeps the iron clean and saves a tiny amount of weight, so you feel virtuous when doing it. I use a small domestic travelling iron and start on the nylon setting, but first I wipe a glue stick round the edge of the structure.  This grabs the covering sufficiently so that you don’t need three hands to get started.  Then I tack round in the usual way and seal all the edges, still on the nylon setting.  The curved edges need special treatment that can be seen in the photos.  For this and the next stage I turn the iron up to “wool” and go over the sealed edges again, then the curved parts, gradually shrinking and sealing as I go.  It still seems very much a matter of chance whether you get a wrinkle-free surface, though.  The iron settings of Nylon, Wool and Cotton equate to tack, shrink and far too hot.

14 Mar 08. Glass-cloth covering.  I have never attempted this before, and was very reluctant indeed to get involved in the epoxy method as described in many a modelling magazine.  I’m strongly allergic to cyano (but not to the odorless type, thank goodness) and mildly so to epoxy, and wish to continue using the latter for as long as possible without too many adverse effects.  But the hull needs some pretty good waterproofing, and Trevor’s glassed hull stands up well to landing on damp grass, so I have a limited range of options. Ivan’s preference of using doped-on silkspan appealed, but I couldn’t find a source of the material, and although I like the smell of dope, neither Wendy nor the dog does!

I had read on the Ivan build thread about using acrylic varnish instead of epoxy as an adhesive, and talking to the ever-helpful staff at Channel 4, was introduced to Rustin’s varnish, which they had used with 0.6 oz glass cloth to cover the fuselage of an ARTF airliner.  I was impressed.  Here are the results.