Building a 1/16 scale P39N Airacobra

Discussion in 'Aircraft & Aviation' started by Leif Oh, Sep 21, 2004.

  1. Leif Oh

    Leif Oh Member

    6. Problems of achieving that perfect fit

    I was so happy with the way things had worked out so far, in my slow and ponderous way of doing things, that I immediately decided to glue the finished part of the back canopy on to the framework. Fitted beautifully, I thought.

    Big little mistake, though. Lesson learned here: Do not let yourself get carried away, and do not ever glue canopy parts on, other than as wholes, outer, middle and inner layers firmly joined!

    As it turned out, the outer layer, the first real covering stretch, did not want to join with the inner layer of the canopy at one little spot at the very back. Nothing doing by now. I could have fixed it by dry-fitting more carefully, and sanding the underlying structure. But now there will always be a little hump there on this model (not on the next I make, though, promise!).

    Otherwise everything seemed to fit very well indeed, and I was so happy about things going quickly for a change that I immediately (without taking photos even) started to add front parts of the outer covering as well, trying to remember Ron's and others' advice on how to achieve the best fit with joining strips.

    Soon, however, it became apparent that the framework, which I had made a tiny little bit of 1/2 a millimeter or so too thick, made the front parts not fit well at the bottom. The landing gear well simply stuck out too far below, and I couldn't sand away a lot of the internal markings there.

    For the first part glued on, I was so preoccupied with locating the proper places for making holes for the recharging plug, and for the switch (even managed to bungle that slightly, but not irreparably) that I glued it in at the top, before noticing the bad fit at the bottom.

    Here, I had to make and print small enlargements of the same pattern, and butt-glue them with a thin strip of copy paper on the back side. Worked fine, but stupid mistake. Do not ever glue anything in until completely dryfitting it, not one but many times!

    For the rest of the front parts, I was real careful about dry-fitting them, measuring how many millimeters were missing on each side, and then simply went back into the computer and added those millimeters, invisibly clone-stamped to the same patttern as the adjoining parts.

    Lessons learned here: 1. Do not make frame members too thick. Many unexpected things may get distorted further on. 2. Even some semi-serious mistakes can be repaired if you are prepared to go back into the computer and print new, adjusted parts.

    This later lesson was soon to come in handy, with a vengeance.
  2. barry

    barry Active Member


    Great thread, watching you put the errors right is even better than seeing it done perfectly the first time. Not that it didn't end up perfect !!

  3. bfam4t6

    bfam4t6 Member

    That looks absolutely amazing :shock: . I have been following the entire thread but I think is my first post in it. I agree with Barry. It is definitely (although probably not so much for you) fun to watch and learn from small little mishaps and inaccuracies. If I wanted to see a perfect construction of a kit then the instructions would suffice. Learning how to correct mistakes is very valueble. Keep it up. Also, you have inspired me to try enlarging a kit. I would like to get the P-39 eventually, but right now I am thinking about enlarging the Halinski F6F-3 Hellcat. We shall what happens I guess. I need to finish my current project first.
  4. Leif Oh

    Leif Oh Member

    Dustin, look forward to your Hellcat project very much - will be nice to compare notes. If you go that way, be sure to tell me in advance - there is an amended part for the Hellcat included in the P39N kit, and Halinski recommends to exchange it.

    I'll be happy to post it. In fact, that goes for everybody of course - anyone in need of the amended Halinski Hellcat part, just holler.


    PS. In fact, I just uploaded the amended part to the Parts Bin. It'll be there as soon as it clears the scrutiny of Ron & Rick.
  5. jrts

    jrts Active Member


    Fantastic work, oh to make errors that look that good.
    Look forward to the rest of the build, outstanding as always


  6. Jim Krauzlis

    Jim Krauzlis Active Member

    Leif, this is a great thread! :D

    Superb write-ups, excellent photo diary, great tips and techniques and an outstanding model. :D I should be so lucky to have my models come out as well with those "mistakes". :lol:

    I also loved the photo of your work area. It gave me a better appreciation of what size this model is...but, boy, are you do you do that? Very impressive.

    She's looking outstanding, Leif!
    Keep those updates coming when you can. :D

  7. Leif Oh

    Leif Oh Member

    Thanks Jim, and the rest of you guys! I didn't photograph the shelves were I store all those unbuilt models and unfinished ones, nor the drawers with the doomed projects, now did I...

    At least half the pleasure of this build is telling you about it - and to find out that many of us obviously grew up with the same kind of models, finally ending up here (I also failed to mention the Revell Constitution, didn't I). It's a small and comforting world in that respect, isn't it.

    More tomorrow, I hope and plan.

    Best, Leif
  8. Leif Oh

    Leif Oh Member

    7. The big frontal misfit corrected

    There was no way the foremost part of the fuselage covering would fit the curvature of what was already added, and at the same time fit snugly around the circular front former. Everthing was askew, as a result of probably several factors: 1) Every little mistake or misfit so far, however small, would add up and compound, not really showing its ugly face until you reached this last stage where a tight fit was crucial; and 2) the front former was most probably misplaced by me (compare my misgivings from the early installments of this thread).

    The only solution now was to redesign the final covering piece itself (and no, I'm neither 3D-literate, nor thus equipped), to fit what was actually built so far. I tried several different ways, from measuring the required dislocations, to physically recutting a series of trial pieces, and then holding them up against the computer screen, set to a magnification that would correspond to real life size.

    I then used the various distorting tools in Photoshop (you'll find similar tools, I'm sure, in any good graphics program - they come in very handy in a situation like this).

    This is actually a rather good tip in its own right, I think. You can set the magnification (or zoom) of your graphic programme to correspond exactly to real-life size. Just hold up a real ruler against the ruler in the graphics programme on the screen, and adjust zoom until they correspond exactly.

    The figure differs for every computer model, and of course for every resolution, dpi, of the image at hand. For the screen I used it is, as an example, 60.4 percent for 150 dpi images. On my laptop, the figure is slightly different. Be sure to make a note of this figure for your own screen & resolution!

    Then you can hold up a piece you have adjusted manually (by scissors and knife) to fit well, and use the distortion tool of the graphic programme to get the part real close before printing a replacement part. The beauty of it, of course, is that the pattern will be preserved perfectly (the small distortion is hardly noticeable).

    After several nights of trouble with this single part I arrived at something that would pass muster, and I could finally close up the front part of the fuselage, and even mount the propeller again for a test run (compare for the first time I tried that big prop on). Very good feeling indeed!

  9. barry

    barry Active Member

    Thanks Leif when my mistakes look that good I will be happy

  10. Gil

    Gil Active Member


    Truly interactive, real time, graphical users corrective interface..., Entirely new application for PC's.

    Great job of documentation and a benchmark for ingenuity on the fly.

    Best regards, Gil

    P.S. And the P-39 looks terrific!
  11. Leif Oh

    Leif Oh Member

    8a. Main canopy tackled

    Now for the really fearsome task of making the compound curve main canopy (and I have to admit I didn't even notice that one part of it really was a compound curve, until well into the build - goes to show you've got to study what lies ahead over and over again).

    I've always looked on publicity shots of compound curve canopies (Mustang, Spitfire, etc) and wondered how those guys did that without vacuum-forming. There are different segments of the compound curve part, that much I had gathered. But how do you glue them together as invisibly as possible?

    After practising on the back canopy, the straight parts of the canopy went together rather well, although it certainly took some time. Made a little jig for the canopy to be able to build it in roughly correct shape from the start, and added one element at a time, until only the top middle element was missing. That's where I suddenly, to my shock and disgust, found out that it was really two parts making up a compound curve.

    Spent a Sunday on attempts to curve a piece of office OH film in the oven, above light bulbs, with hair-dryers, etc. All to no avail. Vaccum forming really is the only way if you wish to make a one-piece compound curved canopy. I realize that now, and will turn to more resourceful friends if I really need that in the future (there are no commercially available canopies when you build at odd scales).

    Finally I decided that I had to resort to the two-piece solution suggested in the kit. Adding one of the pieces, I found that the other really needed to be a little big bigger than drawn in the kit (one of the very few adjustments, not caused by my own mistakes, necessary so far).

    After glueing them on I found that the fit between them really was better than I had expected. Very happy and proud - and there wasn't even any glue in the joint between the two transparent parts. Big question - should I try to add a small string of white glue, or not?

    Couldn't resist it, of course. But the result was not good. The joint now became much more visible. Wiped it off in a semi-dry state with a tooth-pick, and luckily everything was back to square one.

    Lesson learned here: Joints between transparent part that are supposed to be invisible should not be glued at all. Just take extra care to mount them exactly right, and the joint will be as little visible as can be achieved. (Again, it would be most interesting to hear the experience of others on this point. Is there a good, invisible & transparent, glue for this?)

  12. jrts

    jrts Active Member

    Hi Leif

    If you had not pointed it out I would not have noticed it in the first place 8)
    A very fine bit of work you should be well proud of this build.

    keep building, a very good thread :D


  13. Gil

    Gil Active Member


    Overhead foils are made of Polycarbonate plastic. Polycarbonate is great for flast surfaces but is the least formable of the "hobby" plastics. PET-G or styrene sheet is recommended for pulling one-off canopies.

    Best regards, Gil
  14. Leif Oh

    Leif Oh Member

    8b. Main canopy (cont'd)

    Now for the clincher - final assembly of the canopy, including 1) cutting out the door windows from the built-up canopy assembly; 2) removing the temporarily spot-glued doors from the cockpit sides; 3) joining the cut-out door windows to their respective doors; 4) trimming and mounting the canopy assembly, now without door windows; and 5) mounting the doors, with freshly attached windows, one open and one closed but with a partially lowered window.

    Learning experiences here were several: Next time I'll be sure to build the canopy on as good a mock-up as I can accomplish. This time the canopy turned out to be just a little bit crooked when mounted in its proper place.

    Due to this, the painfully executed compound curve (two transparent parts mounted butt against each other with no joining glue) now shows a small but distinct creak. Pity.

    Fact is, by now I'm doubting whether I shouldn't have built it in place, from the inside out, to avoid this. But no, I don't see how I could have managed to get a tight fit that way. And it would have been impossible to clean up the inside of the transparencies.

    So, a really good mock-up it is going to be - the next time around.

    Also, to make the partially lowered window one has to think real hard about which part of the upper window really is visible in the lowered position. Bungled it twice, and its still not perfect.

    All in all, not a completely successful canopy build. But I'll scratch this up to learning experience - it was my first closed canopy, and my first compound curve canopy at that. And it's there. From now on, I'm really looking forward to some straight cardmodel building for a stretch.
  15. Jim Krauzlis

    Jim Krauzlis Active Member


    From what I see that canopy came out great! I'm sure having worked on this part so much you see lots of problems, but quite frankly they are not distracting at all, nor obvious. Like Rob, had you not mentioned the seam on the compound curve I would not have made anything of it...same with the forward edge of the canopy. Really a smashing job! :D I only wish I could do one tenth of a job, but I am emboldened to at least try this some day because of your outstanding description of the process.

    I have heard of folks using a copy of the canopy frame to mock up the curves and to use it to form clear styrene sheet in the oven to form the compound curves. I'm sure it's like most things, a bit of trial and error before you get comfortable with the right temperature, the right time and softness of the plastic. Some even use a homemade vacuum box to pull the softened plastic to the mold after coming out of the oven.

    Really, Lief, this model is coming along great, and I am very thankful you have taken the time to detail your build for us. Superb write-up and very, very informative photographic essays.

    Looking forward to the next update!


  16. silverw

    silverw Member

    Fantastic stuff, Leif :shock:
    I'm enjoying your thread Immensely!

  17. Gil

    Gil Active Member


    Great reporting on the travails of the build..,

    I'm now firmly convinced that when it comes to compound curves you either have to "cast it" or use a material which can be worked into "COMPOUND CURVES" period. On reflection it is now apparent that pulling a canopy would have been more time effective than cutting and fitting.

    Best regards, Gil
  18. Leif Oh

    Leif Oh Member

    9. Interlude - making the exhaust stack stubs

    It's funny how you enjoy thinking about how to accomplish even the smallest, and most obscure parts, in the best way possible. The exhaust stack stubs (12 of them, six on each side) are each made up of two parts, which in turn require butt-glueing in order to look as little cardmodel-like as possible. That means no overlap, and no visible joining strips, since you can look in through the exhaust pipes.

    Started by painting all the parts matt black on the back of the sheet before cutting them out. After all, if matt black ever is required it would be for the interior of exhaust stacks. Again, the old light-box came in very handy here.

    For the butt-joints I decided to try ordinary Scotch document tape for joining strips, holding the two joints butt together while the white glue dried. Worked very well, I think, and I shall be seriously contemplating using this method for the large-scale butt-joining I foresee when the wings eventually come to the fore (at 1/16 there is no way to fit these large parts into A4 or Letter size width papers, however long parts the printer can swallow).

    Neither these little details, nor the canopy before them, were supposed to go on, according to the instructions, until at a much later stage. But I felt unable to leave the centre-front section until everything was completed. Never mind problems of handling the model later on - now I feel a number of important bridges have been crossed, and I've got a reasonably good feeling going on from here.

  19. Jim Krauzlis

    Jim Krauzlis Active Member

    Great post, Lief!

    I have to agree with your comment about getting enjoyment from thinking through the sub assemblies. Not sure why, but the time I spend "day-dreaming" about the possible ways to accomplish the next few steps in the build are quite relaxing and entertaining, at least to me. I have to admit, sometimes the execution isn't as planned, but figuring out how to fix that is fun also! :D

    Anyway, your build is looking fantastic, Lief!

    Looking forward to more!

  20. Leif Oh

    Leif Oh Member

    Thanks Jim (and there goes my 200th posting, wouldn't you know - calls for a cup of coffee!). Talk to you soon again. Leif

Share This Page