T-37 the hard way

[Retired_for_now]

So, local teacher whose son flies WC-130s asked about something for him (while looking at a T-38 model). He's a heavy driver so trained in the T-37 and T-1 - but no models of either out there.

If you can't find what you need on the 'net, you can design your own model from scratch.
Step 0 - collect pictures, diagrams, blueprints, etc.
Step 1 - get a 3-D modelling program
Step 2 - search out the documentation, if found go to step 4
Step 3 - get another 3-D package and return to step 2
Step 4 - study documentation
Step 5 - begin inputting measurements and shapes
Step 6 - reset measurement units and other defaults
Step 7 - re-input measurements and shapes
Step 8 - ask (dumb) questions of other forum members to find the function you need buried 4 layers deep in the menus (it's always easy if you already know how to do it ...)
Step 9 - draw some more
Step 10 - attempt to print test parts
Step 11 - search web for print document formats and drivers
Step 12 - attempt to print
Step 13 - reset paper/page size
Step 14 - attempt to print
Step 15 - score, cut, bend, fold, and glue
Step 16 - assess the result
Step 17 - squash and toss
Step 18 - return to step 4

Or, you can just start drawing on a piece of paper or favorite graphics program (one you already have and know how to use). As a minimum you'll need to find or draw a 3-view of your subject showing top/bottom, sides, and head-on profiles all to the same scale.

Iteration One (blockheads anonymous)

The top and bottom views give us the top of the aircraft fuselage, the belly, and the wing and horizontal tails. For now, just lay them out on your page. The side profile gives the fuselage sides and vertical tail. For a model with gentle curves you can probably just lay them out on the page and start cutting. However, any curved outlines will increase the size of the part needed to wrap around them (review your geometry on arcs of circles versus chord lines). If you think it's needed, stretch the parts a little bit (only in the direction needed).

For the T-37 I left the side profile as is and figured I'd just cut the top/bottom surfaces short where the tail pinches together anyway. I did add in a windshield stretched to match the side profile view to keep the graphics from getting too far off. Top and bottom surfaces of the wings and horizontal stabilizers just glue together flat (like an Ojimak design). Glue a couple of extra strips of card inside to make a wing spar if you think you'll need it.

Now you need to attach the parts (WARNING-some contemplation required). For the first iteration tabs are fine. I drew tabs along all the edges of the bottom surface and bent them up to glue on the sides. The sides have tabs along the top edge only to attach the top surface - with a slight change at the back end. The vertical stabilizer is an integral part of each side piece but I cut the forward portion of the fin away so I could close it up and to allow the top piece to slip underneath.
Wings slip through a slot in the sides. The engine pods are glued up from appropriate sections of the top and side views, slipped over the wing and edge glued to the fuselage.

The result is a rectangular cross-sectioned box (think nobi) that still looks a lot like what we're going for.





So far, we have the start of an easy model. Easy means the parts fit well and the assembly is obvious or you have clear instructions. Again, this is only a start as we're missing the connections (tabs hand drawn in at this point), there are no instructions (or even labels like part numbers, top, left/right, etc.), and the fit is a work in progress (hand massaged and some trimming after assembly).

We share a lot of material - and it's often not clear just how developed the model patterns are until you take a look. This is not yet up to my "sharing" standards.

Yogi (workmanship reflects the worker as a design represents the designer)

[willygoat]

Looks like a good start!

[thorst]

Yogi,

I have rarely seen this kind of design attempt so well executed as on this model - really good work!

If I might add one aspect to the first part of your post:
In times of Pepakura and similar programs, I often get the impression that complex shapes can only be mastered with these programs. That's not the case. You can achieve the same, if not better, results with just a ruler and a compass or their digital analogons. It requires just once to think about how it works, and then a bit of time. But it's doable, if you don't believe, take a look at my shuttle.

Thorsten (who only said this to encourage people without the money (or the instructions) for the said programs to try it out "the old way")

[Uyraell]

Yogi, you've shown brilliance, here, and achieved that which I thus far can not, due to lacking the necessary skills.
Looking at your "test build" it is very clearly a T37and a very reasonable representation.
It occurred to me that a BAC Strikemaster could be done in the same way you've used to your T37, it being of a similar vintage.

My profound Thanks to you Yogi, for sharing with us lesser-skilled mortals the steps by which you achieved the test model. In doing so you give hope for the as-yet unskilled among us.

Kind and Respectful Regards Yogi, Uyraell.

[Retired_for_now]

willygoat, thorst, Uyraell,
Thanks for the kind words. Point of this is to show just how easy it is to get started - and what an accomplishment it is when our more talented designers produce their (accurate and detailed) masterpieces.
Yogi (still fiddling away ...)

[Retired_for_now]

For iteration next, pick out a part of the model to refine.

In this case, I decided the engine pods were next. I wanted to more closely model the way the pods wrap the engines rather than just block them in. First step is to reduce or eliminate the flat side plates of the pods. Since the top and bottom of the pods are angled the information needed is captured by the top and bottom views. I thought about cutting down the side plates but decided to skip that step and massage the top and bottom of the nacelle instead.

So we try just leaving out the engine nacelle end plates and bend the top/bottom nacelle surfaces to the wing. This required slitting the parts where the pods bend down to the intakes at the front to allow them to take the shape and meet the wing surface. I also squeezed in the top of canopy while I was at it and added some insignia graphics. The result:



Different, but not really better on the engines as they've lost the chunky look that characterizes the tweet. The problem is projection - the curved surface of the actual nacelle is wider than what you see viewed from the top. So, I stared at the result a while, sketched a few curves, and decided to stretch the width of the top of the nacelle by 30% (bottom part looked OK as is).
Third try:



We're getting closer, though the pattern is looking more cluttered and confusing (unless you're the one adding all the tweaks).

Yogi (still slitting and trimming ...)

Just a note on the horizontal stabilizer - it's built up by gluing a doubled strip of card inside one half of the stabilizer. The strip sticks out far enough to go through a slot in the tail and into the other half of the stabilizer. Basically a short spar to connect the surfaces without cutting a really wide slot in the vertical tail to insert the complete part.

[ltla9000311]

I love your step by step process in the original post! Soooooo like me.

That is a great looking model so far. Keep up the good work. I can't wait to see more pics!

[Nh3ave2009]

Yogi, you are always my Hero!!

[charleswlkr54]

I like this one! Will it be available for download?

[Retired_for_now]

Todd, Dennie, Charles,
Thanks. Charles see post 1, last line of text ...

Next step would be to round out the nose. The tweet is pretty distinctive looking - a wide, flat tadpole. If we keep the sides and bend in the top and bottom edges we'll have to narrow the top and bottom fuselage parts to avoid looking like roadkill. This will also require a close look at how
the revised parts flow toward the aft.

So, I reduced the width of the top by 20% and the bottom by 10%, figuring on curving the sides to make back the width. Rounding the sides and top brings up another factor - a surface that curves top to bottom and front to back forms a complex curve. Sheet goods like card are good at forming simply curved or conic section surfaces (think flats, cylinders or cones). A complex surface like a sphere can only be approximated by stacking conics or using petals (either of which we are trying to avoid to keep this simple). The front of the airplane is the toughest as it curves in to the front and curves top to bottom.

To account for this I made a couple of slits on the top part (just aft of the windshield bow and aft of the rear canopy fairing). For the nose I simply relied on incrementally shaping (crushing) the card with a fine tweezer.

Comparing the results:





You can see what a difference a subtle change makes in the model's appearance. Just a little curvature in the surfaces and no sharp corners at the joins. I think it looks better from the canopy forward. However, since we reduced the width of the top and bottom the required curvature needs to come from the sides - which is difficult as they are complex surfaces. The ultimate result is that I didn't get enough curve into the side panels and the resulting model has lost the T-37's signature look - wide and flat like a little sports car (or a tadpole ...).

So the next try will keep the original top and bottom panel widths and reduce the height of the side panels. This should allow a relatively slight curve in the top and bottom to form the shape better without requiring a radical curve in the sides. We'll see ...

Yogi (somebody stop me before I print again!!!!)