I can't believe it's been two months since I posted an update on this one. I am really a slow builder. So OK, let's start the weekend with some progress information, and by summing up the work so far. I am going to call this the tales and tribulations of a clumsy modeller...
1. Defining accurate drawings
This has been like searching for the Graal, but in this case I finally finished by finding it, I believe.
As mentioned in earlier posts, I was trying to define the right profile for the Gulfhawk, from drawings found here and there, and from photos. In the end I decided to CA draw my own plans, some initial results were shown earlier. In the process, I had to solve the riddle of the landinggear cinematics. When you start laying articulation points on a scale drawing, you soon realise that there are not many options that will result in the LG looking as it is on the ground, and having everything fit into the fuselage when retracted, taking into account the oleo extension...The probllem is that without manufacturer data, you have to trial-and-error until you reach a satisfactory drawing. The good news is that I now know how to rebuil the landing gear from scratch ! The other good news is that I also have understood the shape of the fuel tank and of the keel box at the undercarriage level !
Anyway, I drew a few frames in the process. This is where I had arrived at after a few week-ends of work, and things still needed some adjustments to fair each frame into a smooth continuous curve. So this is not yet the definitive drawing for the F3F, plus there are some mistakes like axes having moved on one of the frames as on frame C3bis....Do net expect to be able to uses them as is, not even considering this is a mid-res photo and not a vectorised drawing ...
After many hours of work, I was growing frustrated that achieving these drawings was taking too much time from my sparse modelling time over the week-end. So I decided I had to do some more search of drawings. Having read on Papermodelers that the Cleveland drawing was the best around, I ordered it. It is good, but not really as accurate as claimed, and oversimplified in many areas, but then this was for balsa construction back in the 30's or 40's. I then stumbled upon a Comet drawing on the net. I finally found a link to a vectorised one somewhere on a RC-forum, than a CD with vectorised Comet drawings on eBay....More time has elapsed, but I had drawings I could rely on. Unlike many other Comet drawings, which are also very simplified for balsa flying models quick production, this one is a really detailed one to produce an accurate 1/12 replica of the Gulfhawk. It shows how popular this aircraft must have been in its time with US people. And being vectorised, it can printed easily to 1/32 scale. Eureka !
To get a glimpse of the Comet drawings (sheet 1 of 3) see below my fuselage surgery.
2. The engine ignition ring
Another example of my clumsiness...
The Vector engine is nice, but not without its faults. One of them (more below with some pics) is that the ignition ring is not totally convincing, and more importantly I do not know how to butt-glue ignition wires, nor trust the structural integrity of such joints in subsequent handling. I have no idea how you guys who do it succeed. Plus the two ducts leading to ingnition cables to the magnetos in the back of the engine are set up at a polar angle whereby they interfere with the cylinders...So I had to redo an ignition ring, with the cable stubs drilled for inserting the ignition wires and having a strong joint. On the Gulfhawk, the ring appears to have a regular distribution across the circumference, and is behind the rocker arms. Be careful to check this point because it varies from one aircraft to another, or rather from one engine version to another.Attempt one
: an ignition ring in plastic. Using platsic rods, I achieved this :
Not bad, but I still had the problem that I could not drill cables holes in the plastic stubs. The solution was to use some thin (0,7 mm) brass tube, soldered !Attempts 2, 3 and 4
: Soldered brass tube. A 1,2 mm brass ring, with 0,8 brass tubes stubs butt soldered. This is my best result
(attempt 3) :
Diagnostic : butt soldering is not strong enough, because I am not that good and because the heat from one soldering point softens the previous ones. Plus the distribution of the stubs is not regular enough...The solution was to drill holes in the ring for the stubs, to have some mechanical aid in holding the stubs in place. The problem is that doing some hand measurements resulted in me drilling 17
So, on to attempt 5 :
why not drill the holes in a straight tube, and bend it in a circle afterwards. Plus let's find something else than the soldering iron. Here Wojtek's thread came to the rescue. So I lifted his idea of using a heat grun with soldering paste, rather than a soldering iron. The ones in DYI stores being just too big, I ordered the small one from Proxxon. Still more time elapsed ...
In the process I have also experimented that "Patafix" (the English/US equivalent is Blutack) is an excellent medium to hold the parts to be soldered together in place, and is almost insensitive to heat. So the idea is to drill holes regularly (a lot easier along a straight line); insert the stubs, paste them with soldering paste, and heat the whole set-up with hot air gun. It worked ! BUT ... Who is the genius said measure twice before cutting once?
Not me !
Where did I read that the ring circumference was 22 mm? On my own drawings, it says 17 . So if you start with the wrong circumference, you end up with the stubs being too fart apart, you stupid !
. And with your mechanical background in everyday's work, how could you expect that a tube weakened by 18 holes drilled into it would bend in a nice circle, even when reinforced by soldered stubs ?
End of attempt 5 :
Edited by MostlyRacers, 18 June 2011 - 09:45 AM.