Modeling Challenge: Aircraft

Aircraft are some of the most fun things to model. For whatever reason, I find them easier to model than cars. Maybe because they are so big, you are excused from a lot of detail. To the left is a SR71 Blackbird I modeled in about 4 hours, and rendered in PV360.

You can submit images or models. If you submit a model, I might analyze it a bit. I’ll be nice. If you just submit an image, it can be rendered or just a CAD screen shot. Please give some details about it, what sort of plane it is or if its your own invention. What software you modeled it in, what was challenging, what you learned.

I plan to do a P58 Mustang, and maybe an F4 Phantom. I’ve already done an F104 Starfighter and an F5 Tiger, which I’ll include in the submissions. Maybe we can include space craft, in which case I have a couple versions of the Enterprise (1701 and 1701D). Anyone have any Klingon or Romulan craft?

The purpose of this is mainly to show off your stuff, and share ideas and techniques. If you usually just let other people model stuff, why don’t you send something in this time. Someone will learn something from your efforts. You don’t have to put a huge amount of time into it. Let’s see what you can do.

At least one regular reader has already submitted a nice model, I’m just checking to see if I can post it here. Rick, you out there?


The first entry is from Rick McWilliams. Rick spends his time on a sailboat going where the wind blows him and uploading SolidWorks data through a satellite link. So boats are kind of a thing for him. This model nicely integrates a boat hull into a plane. The top wing is not complete, but you can see there is some nice modeling going on here. Rick had some comments about his work:

The model has some useful concepts:
1. The first few sketches determine most of the geometry. Lines and curves from these are projected to become guide curves. Planes are mostly defined by these main sketches. Sometimes a short perpendicular construction line is added to make it easy to define a plane. The midline curve in the profile view is most useful to define the widest point in the fuselage. This curve is projected onto the edge line from the top view, and used as a guide curve for many surfaces. Make extras as the surfaces seem to consume these.
2. Boundary surfaces are often better than lofts when there are several sections. Lofts do not necessarily follow the sections. Lofts with just two sections usually work well.
3. Surface trims are often confused when a surface is changed. Mutual trims sometimes work very well and sometimes fail.
4. Smoothly intersecting surfaces may be generated by projecting profile or plan curves, and then splicing a planar curve. These are particularly useful for sensitive small angle intersections. Trimming small angle intersections is unreliable. I usually extrude a trimming surface from a plan curve.
5. Poor edge conditions are common. A surface that uses a curve projected onto a surface as a guide may not quite contact the surface and consequently trim or knit improperly. Extend surface can make the trim definitive. Alternatively create a trimming surface and trim one of the surfaces.
6. Boundary conditions for guide curves have unreliable results. Four sided boundary surfaces work better than three sided ones. I just hate the peaks and butt cracks that become obvious when the part is mirrored.
7. Once a group of surfaces is knit together fillets and rounds may be applied. I usually knit into a solid.
8. Split lines are wonderful. These divide surfaces into parts which can have different colors such as tinted clear for windows or colors. I use these last as they can interfere with features.
9. Configurations are very useful. Beware that constraints and dimensions may not be the same between configurations. This can cause features to do weird stuff. Sometimes I have to delete some configurations, and recreate them from an unbroken configuration.

The next series of images are from Robert Bilbrey. This looks like a lot of work. Robert also had some comments to go along with the imges:

I have been alternating car and airplane models for some time now.  I find the “antiques” in both categories more interesting than their modern descendants in their variety.  Today’s aircraft designs, especially, seem hostage to more mature rules defining speed and performance and mission and thus have a certain sameness.

The attached were modeled in SW-09 and 10 and all were  rendered in Photoworks.  One important reason I don’t like PV360 for airplanes is that a background image is not visible through a transparent surface.
Such is the trick I use to simulate a “motion blurred” propeller.   I would be interested if anyone has a more effective or elegant solution to this problem.  Flying planes with static propellers are just plain ugly.  I sure hope this is fixed in the 2011 version of PV360.

The British WW2 Spitfire flying in formation was modeled and the F-16s were of the image (obviously).  I made the Nazi Folke-Wullf F-190 responsible for the World Trade Center tragedy.  I hope this historical liberty does not offend.

For those not familiar with pre-1945 planes, the twin engine job is a Beechcraft 18 dating about 1937 and used extensively well into the 1950s,  The prop blur worked pretty well on this one.  The twin-boom is a Lockheed P-38 WW2 fighter.  The biplane is a Stearman-Boeing PT-17 dating from about 1937 and used as an Army Air Corps trainer before and during WW2 ( yep, there was no independent US Airforce in them thar days. )

Beech 18 over Denali


P-38 (PV-360)



Very nice work. It is hard to make renderings fit well into the background.


Entry #3 is from Mark Reader.  The rendering is great. It’s good to have a little fun with what you do, and this one looks like Mark had some fun with it.

Mark Reader’s toy Lear


Matt Lombard’s -51d Mustang

And an F5 Tiger


The next entry is a glider from Steve Farner. Click the pic for download. Here’s what Steve had to say about his model.

Attached is my first attempt at surface modeling (what little I know about this topic is from your “Surfacing and Complex Shape Modeling” book – one of the best purchases I ever made). It is of a fairly historic sailplane, the Darmstadt D-36 Circe, and the intent was to create a fuselage model mold suitable for CNC fabrication of a mold for a 1/4 R/C model. The wings and tail surfaces are really more cartoons than well made parts…

The fuselage model was built from 3-views drawings and cross sections inserted as sketch pictures, and the necessary sketches derived from these. (I included this part with all its warts intact. You can see all my suppressed features, features that didn’t work, etc.)

Over, I was pleased with the way it turned out. I still have a lot to learn about cleaning up the little details and discipling myself to use better modeling modeling practices, but I was able to generate the g-code for the mold based on this model without too much difficulty. I should be able to lay up the actual fuselage pretty easily…


Here is another from Rick McWilliams. I think this is my favorite so far. Great looking model. very nice! Click the image to download an eDrawing assembly.  Makes me wanna FLYYYYYY!

Rick McWilliams personal Z plane
Here are a couple from Sooraj Rajendran

Leave a Reply

%d bloggers like this: