3D Models Available Now. Click here for the 3D Catalog.

CAD Tutorial:  Using 3D/CAD to Help Design & Build Model Aircraft

A much more detailed CAD Tutorial is now available on CD. See the Products page for more info.

Since the late 1970s, I've been using computers to help me design airplanes. At first, I simply used it to collect and compare databases of known aircraft specs, as a reference. In the late 80s, i was one of several people who created "generic design" software, which, based on inputting a couple (literally) of parameters, such as "type" and "wingspan", would spit out some basic specs that would work. Obviously, these were just numbers, and not fancy graphics. In about 1995, I began using 3D exclusively, and quit programming.

Using 3D to Design Models

One thing worth mentioning is that every CAD program handles it's files differently...unfortunately. For example, a file that looks great in my native CAD program may look horrible when imported into another CAD program, and vice versa. Even within a single program, results can vary. Here are some lessons I learned about "Carrara", for example, using my King Air B200 fuselage as an example. All I did was convert the file to different formats, in different orders, within "Carrara", and rendered them. Look at the awful results that some conversion orders produce. I recommend that you do this same kind of test in your own software, so you can see the effects!

Research... The process begins.

The research part of your project can be as involved as you like. If you're simply copying an existing simple sport plane's drawings, and particularly if you've already flown a plane like it, maybe there's not much to do at this stage.

If I'm building a quick sport-type "fun" plane, I may not even make plans, and usually reserve the rest of the process outlined here for more serious scale or "big" projects. For these serious projects, I'm a fanatic at this stage.  The finished product will only be as good as your research and your drawings, so be very finicky here.  For this discussion, I'll be using our 1/6th scale super King Air B200 as the example.  There are many shortcuts you can take in this process, depending on how precise you want to be. My explanation is the long-winded one, for the truly ambitious modeler who is striving for perfection.

If you've decided to do all this work, then you must think that the aircraft is "beautiful and wonderful", right?  Keep that in mind when you go begging for documentation, and your love for the aircraft will show, as sincerity.

First things first... Create a folder on your computer called "(Aircraft Name) Project Files" and keep computer copies of everything you can in this one central location. You might also create sub-folders inside it, for "Photos", "Drawings", "Notes", etc.. You get the idea. Keep everything organized, because you're likely to build up a lot of "stuff", over time.

This is my folder for the B200 project.

Go and buy some good quality gridded vellum. (a fiber-reinforced drafting paper that is great for tracing)  The layout is tremendously helpful at keeping everything straight and in line. Usually, the layout markings will not show up on photocopies of the vellum drawings. You'll need some basic drafting tools, including rulers, straight edges, triangles, French curves, and a compass. Don't forget a BIG eraser! (In case it needs to be said, draw in pencil first, get everything right, then ink them at the end.)

Next, try and get everything you can from the actual aircraft manufacturer themselves. If they are no longer in business, you may have to start contacting museums and others who have archives. Don't hesitate to let them know exactly what you're doing, (briefly) and ask a lot of specific questions. If they're giving you drawings, what is their source? Are they actual manufacturing drawings, or drawings of their wind tunnel model, or simply "general arrangement" drawings? Often, actual aircraft manufacturers have no use for "3-views and cross-sections" of the type that modelers need. Their building process is quite different than ours, relying on each individual part being created and assembled on a factory floor... Maybe by computer, or if it's an older plane, maybe by hand.

Remember that many manufacturers have made the same basic model for many years, but that it may have been modified/updated several times. Decide on a particular version of the aircraft, and stick to this version in every step of the process, including documenting a particular actual aircraft.

Try and impress on them that you can only build a precision model of their wonderful, beautiful aircraft if they help you get the documentation you need to do it justice. Try and find a "fired up" sales rep., who is really proud of the aircraft and excited to see a model of it built. You might even offer to send them a "kit" when you're finished, in return for their help.  If you mention that you intend to compete in AMA-sanctioned scale competitions, they may see it as an opportunity to advertise their product, and be a little more willing to help you. If you're courteous and respectful, you'll be amazed at how much cooperation you get from people. The process we used for this initial documentation took almost two months!

You might also consider a plastic display model as a source of information. Only you can decide whether or not to trust the company who made it. Some of these companies go to extremes, in search of precision, and their models are about as accurate in outline as you can get. I've also had many display models that were WAY off too, so be careful if using this approach.  Why not actually call the model company, and ask them how they documented their model? If they did it right, they're probably proud of their method, and will tell you. If they "don't know", or "aren't sure", that's certainly a reason to do some more research!

If you're desparate for measurements of an actual aircraft, here's a homemade device you could use (with a helper) to get cross sections from an actual aircraft. Just mark the fore/aft distance on a piece of paper, so you know where each cross section belongs. Later, you can scale the cross section down, or just input the numerical data into your CAD program and scale it there.

On the subject of physical models... There are some methods, both high and low tech, for transferring physical dimensions of an actual model into a computer. If you have a computer modeling system with this physical "tracing" capability, (Some use lasers, some use a 3D-sensing "stylus" to physically go over the model, get it's coordinates and transfer them into a valid 3D file, such as DXF, for you.) use it!

Most 3D/CAD programs can easily scale your data, should you change your mind about the size. Be aware that with some software, this can lead to problems later. Some software "remembers" your original scale, and always calls your new parts "110 percent", or "80 percent" or whatever. All software is different, and maybe this won't be a problem for you, but for that reason, I suggest always inputting measurements into CAD using the "actual size" measurements of the finished model.

There are several "starving hobbyist" approaches that can yield very good results, if you'll just take your time and be careful.  Let's assume that you have a plastic display model kit that is precise enough for your needs. First, figure out the "magic number" needed to convert your measurements to the scale of your desired finished model. You'll refer to it many times, so write it down. For example, you need to multiply every measurement of a 1/48th scale model by 8, if you're building a 1/6th scale model. Be as precise as you can.

Here are a few methods of producing drawings from a physical model.

1. Before assembly, you could trace every part you need onto graph paper, then simply scale up the layout to the scale factor you need, and redraw it. You may be able to scale these drawings up in your own computer, after scanning, or you might have a local copier/printer do it for you. (More on that subject in the next chapter)

2. After assembly, you could carefully mark off areas that you want cross sections for, by scribing or painting marks, or by using very narrow masking tape. You could then very carefully use a band saw to slice up the model. Then you can trace the cross sections as in the previous step.

3. Some modelers have used a semi-flexible wire, such as solder, to create outline sections. Simply press the wire in place on the outside of the part, (using reference marks made as above) then trace the outline of the wire onto graph paper. This is not a very precise method.

4. You could give the parts, or the model a couple of coats of wax, followed by PVA, and then actually do a fiberglass layup right on the part. You could then cut your fiberglass "copy" up, and use the graph paper method above.

I'm sure there are many other methods. Those are some I've tried. I've added a separate article showing how to use a plastic display model to produce larger scale plans. You can find it on the Design & Building page, or just click here.

Since you have a computer, use the internet to your advantage.  Do a search for the aircraft you're interested in, and download every snippet of information you can find, including specifications, photos, drawings, and even commentary on the flying characteristics. If you have a flight simulator program, you might want to see if you can get a version of the plane for it, and have some fun exploring the flight envelope.

If there are actual examples in your area, go make friends with the owner, or better yet, the pilot, and ask if they will let you come out and photograph their beautiful, wonderful aircraft, for the purpose of building a precision model of it. Again, these people are often proud of their aircraft, and may see it as an opportunity to advertise their company. If you get this chance, take full advantage of it, and take a LOT of photos, make notes and sketches, ask questions, etc.. These will all come in handy as you proceed with the building, and will help you avoid errors. In the case of the B200, I was able to do this, and it has really paid off. Using an actual aircraft to verify your measurements is vital if you're after precision.

Remember that it's a lot easier to "fix" something in CAD or on paper than it is to fix on a mold or a model! Attention to detail at the beginning will make the building process a lot more relaxing and fun, and takes away the stress of wondering "how it will come out". You should know this BEFORE you make the model!

When you're convinced that you've got all the documentation in your hands that you're going to get, then I suggest on additional step. Scan all the drawings you have, at a fairly high resolution, (at least 200 dpi) and put them in a safe folder on your hard drive, labeled "Originals". Don't ever alter these in any way!  They're your safety net, in case you screw up.  Instead, copy that entire folder, and rename it "EDITED Originals". These are the files you'll actually manipulate.

I've found it very helpful to do this little "quality check". If you have a CAD program, then you obviously have a computer, and if you can afford a CAD program, then you probably have one or more graphics programs that will let you scale images.   You'll probably have more than one set of 3-views, and you may be wondering which one is the most accurate. This is relatively easy on a computer, but would take a much longer on paper.

I first "cut up" these drawings in the computer, creating a SEPARATE side, top, and front view from each source I have.  When you "cut" these, clip them right to the edges of the aircraft.  For example, on the side view, clip the image right at the forward-most section of the nose, and at the very tip of the tail. Leave only a few pixels of "white space" on either end of the drawing. You'll see why in a moment.  Be sure and keep these organized, so you know which side view goes with which front view, etc..

Now, take one group at a time...say all of the side views, and using your graphics program, scale them all to the same size.  For the side view and top view, (facing the same direction) scale them all to the same LENGTH. For the front view, scale it to whatever wing span the top view had. The actual SIZE of these images doesn't matter too much at this point, as long as they're all the same SCALE.

Now you can compare them, using one of two methods: If you have a graphics program that will allow for transparency, you can make "slides" out of the images in the computer, and position them "stacked", to see if they match. If you don't have this type of program, then simply print them onto the thinnest paper you can, (or onto mylar) and stack them manually.  If they match, lucky you!  Probably though, you will find that they don't match, and you'll have to start figuring out why. You'll have to look at all the information you have, and perhaps even measure an actual aircraft, to figure out which ones are right.

As you can see here, we have some serious discrepancies between these two drawings!  When you've hacked your way through this process, you're ready for the next step, which is some preliminary planning on the design of the actual model.

Plan Ahead - Make Some Phone Calls

Eventually, you're going to want actual size plans from which to build your model, or your plugs and molds. There are many ways to end up there, and I'll outline a few ways here for getting these parts into your 3D/CAD software.

First though, think about the final output you can produce from your CAD system. Whether it's a graphic file, such as a JPEG, or a DXF file, or even a specialized file for CNC machines, be sure that it will work!  What I mean is this. Are YOU going to print the plans, or are you going to have someone do that for you? If you're planning to print them yourself, how will you do it? Do you have a huge plotter capable of that, and the right software to drive it? Or maybe you intend to print the plans on a kabillion sheets of standard size letter paper, and tape them all together? Again, does your graphics software allow for a precise enough method of doing that?

The easier way is probably to go to a local copier or printing business. So, what you need to do at this stage is make some calls.  Look under "Computer", "Copying", "Graphics", and "Printing" in the phone book, and talk to some of them. Ask them if they can do it, and if so, what type of file do they need from you. Will they be enlarging an image from paper, or will they be printing from a computer file?  If it's a computer file, what do they want? (JPEG, TIFF, GIF, BMP, PICT, PSD, DXF, EPS, and PDF are a few of the standard file formats.)

Ask them what it will cost. You probably only need black and white, so be sure and mention that. Most of these places have a rate per square foot, or per square yard, and you probably will be surprised what you will find after a few phone calls. For this type of work, in my town, I have found that the prices can vary greatly! It pays to make these phone calls, and maybe even go there and get a sample printed. It shouldn't be very expensive, for what we as modelers typically need. Be aware that some printers may refuse to copy a drawing for you, if it has a Copyright notice on it, or even if it seems like it MIGHT be Copyrighted. Generally, we're allowed to make "copies for personal use only", but that's a judgment call, and they may ask you about it, or insist that you get a Copyright release from whoever has the rights to the material. That's what they're supposed to do, so be nice. You'd want YOUR work protected, wouldn't you?

If you can produce a "vector" graphic, (EPS or PDF, for example) and your local printer can use it, that's ideal. The reason is that "vector" drawings scale up or down to any size, with no loss in quality. You've seen what happens when you enlarge an image by "scaling". The lines become thick, and you tend to lose accuracy, because you don't know whether to use the "inside" or "outside" of the line for a reference. If you're producing the more standard "raster" graphic, (like a JPEG) then make it at the greatest resolution you can, (at least 200 dpi) so that you don't get the "jaggies" when it's blown up.

If you've gotten this far, and have some actual size drawings in your hands, then ask yourself a question that may save you some reading here, and a lot of time on the computer...

Ask yourself if you really have a need to transfer this material into the computer. (and back again to paper, later)  Each of us has our reasons, but rest assured that it's a lot of work, so be sure that this really is important to you. Otherwise, you already have drawings in your hand, so just transform them into plans yourself, and make a plane!

Here are the plans for my "J-45" aerobatic design.
I'm tracing over the enlarged CAD version, editing as I go.

I use CAD for several reasons, although I don't make EVERY model this way. After several years of CAD use, I have a "library" of parts built up, that includes model engines, fuel tanks, servos, linkages, wheels, etc.. So, one of my reasons for using CAD, when there's any doubt about it, is to make sure that everything will fit into the model, and to visualize where it will go. Your CAD system may even allow for you to determine the center of gravity, or perform some function such as a "virtual wind tunnel". Maybe you want to use it to create more precise parts than you feel qualified to draw on paper. Or maybe you just want to create some cool images.  You decide.

Enter the Data into your CAD System

Ok, so you're ready for CAD!

If you've got a very expensive input system, such as those mentioned earlier, you should obviously use it. I'll outline a few more methods for those of us in the lower and middle price range of CAD systems. I must assume that you're familiar with your computer, have read and understood the book that came with your CAD system, and understand basic graphic operations. I can't possibly explain those things here, as each subject deserves it's own book. So, if I'm losing you here, that's ok. Just go read the book, get up to speed, and then come back. I can't possibly predict what system you have, what version it is, or how well you know it. So, these have to be general guidelines, rather than something specific, like "press the X key". I'll assume a few things about your system, so we all have a place to begin...

I assume that you can import graphics into your CAD system, one way or another. Some CAD programs allow you to use an imported (raster) graphic as a background, and some will even let you trace right over it on screen. If so, this makes your job somewhat easier. Other programs will allow you to import a "vector" graphic (such as an EPS file) and use it on a "flat plane", as a cross section. You may be able to use a side, top, or bottom view, to describe the "envelope" into which your cross sections can fit. Any of these options will help.

Depending on your original source material, the importing process can be simple or a big pain. Here are a few images showing the steps I had to take for every useable cross-section on our B200. The example here is a nacelle cross section. The original drawings were at 1/20th scale, and don't look very precise, once scanned at 600 dpi. Some work was required.

This is the original section after scanning. There's a lot of "junk" to remove, before this is really useable in a CAD program.

After removing the background lines, and a little cleanup, one pixel at a time, the section could then be imported into Adobe "Illustrator" to be converted to a vector format.  Any jagged edges, gaps, etc. must be smoothed out, because "Illustrator" will trace these exactly, creating a very jagged surface, with too many Bezier curves to be practical.

This is the vectorized result, after converting the image in "Illustrator", and saving it as an EPS file. It still needs some small corrections before use. It can then be imported into CAD software for extruding parts into 3D.

If you only have the outlines, and no cross sections, you have more work to do. You could create the "envelope" mentioned above, and gradually create the cross sections manually, using your eyes and some good drawings and photos as references. This is "ok", but not as precise as actually importing the cross sections you already checked for accuracy, on paper.

I use this method.  In my CAD program. I create three ultra-thin pieces of "virtual paper", one for each of the X, Y, and Z axis.  Unfortunately, not all CAD systems use the same identifiers for these. Some for example, use "Z" as the vertical axis, while others use "Z" for the horizontal "depth" axis. The "papers" look like the image below, and will serve as "drawing holders" for the side, top, and front views you created earlier.

With these, and your basic "envelopes" created, you can gradually refine the 3D model until it matches all three of the drawings, to the limits of your CAD program's accuracy. Here's another image, with my Super King Air B200 model aligned that way. Even if you lack cross sections, you can, with enough effort, create a very accurate CAD model. I've done it many times, so I know you can do it.  Just be aware that this is a process that can take several weeks or longer. You won't do this in an afternoon!

You can "slide" any of these paper "plan holders" along their axis, in order to help your drawing of specific parts. For example, you could slide the side view to the center, as in the image above, to concentrate on the fuselage, then slide it outboard to the nacelle position when working on the nacelle, landing gear, spinner, prop, etc..

My biggest reason for doing this is to correct whatever drawings I've selected, using CAD and taking advantage of the 3D point of view. When I'm happy with the outline, I print out a high resolution (300 dpi) version of my altered drawings, and take those to the printer. If they're good enough, "as is", great. Usually though, I take these new enlargements and use them as a guide, tracing over them onto vellum, and produce my final building drawings by hand. That's how the King Air is being built, as you can see in this image of my drawings.

This is one of the Raytheon drawings after some correcting in CAD.
At 109 inches of wingspan, small errors in a drawing can become large errors in a model.

The next step is simply "Keep going until you're done." For some of us, this means just getting the outlines right, to create accurate exterior plans. For some of us, it means "virtually" installing RC gear, a cockpit, passengers, and on and on. You decide what your CAD model needs to be, for your particular purpose.

Editing

When you're finished, check everything again. It wouldn't hurt to have a friend or two take a look at your work too, before you "finalize" it. Sometimes it's very hard to edit our own work...

Read this silly little phrase: "After you mow our lawn, let's go to the the park and have a picnic with the kids.".

Many people won't notice that the word "the" appears twice in a row, in the sentence above. So, if you didn't notice an error in your work, maybe someone else will.

Check Everything You Can Think Of (Twice, at least!)

When you believe that your CAD model is finished, I suggest printing out some letter-size copies of it, either as "mini-plans", or as 3D images from different viewpoints. Leave them on your desk, and compare them to your original drawings and images, and see if you notice any problems. Relax, and give this a few days to sink in. By now, you should be very familiar with the aircraft you're modeling, and you'll start to notice small anomalies.

Is everything that's supposed to be symmetrical correct?  Are the angles of the surfaces, dihedral/anhedral, thrust line(s), landing gear strut length and angle, etc., all correct? If not, fix them. If so, buy yourself something nice.... You deserve it!

Since the next step is to create full-size "plans", you have to make a decision about what is "frozen" in your design. Real aircraft manufacturers do the same thing.  As certain design elements are finished, they must be "frozen", so that the next person down the line can depend on what you gave them. For example, the guy designing the landing gear must know in advance how much space there is to put it in.

So, with at least the outlines of the aircraft "frozen", you're ready to make plans. Since you've already talked to the local copier/printer, you know what format to save your files in. Get them to the printer, and get ready for finishing the paper version. Of course, if your CAD program will directly output all the parts for your entire "kit", you can skip the next step, which is to refine the plans on paper. The rest of this discussion assumes that your CAD program will NOT produce kit-style plans, and that you will have to do some drawing. Next, I'll discuss the method for producing balsa parts in a traditional "kit" form, and also the process for making plugs and molds.

I like molds, even though it's a LOT more work. Why? On a scale model, in particular, there is a tremendous amount of detail work involved. Depending on the aircraft, there may be panel lines, rivet details, doors and hatches, and many other scale items to reproduce. If you do this the traditional way, and then have an accident with the model, you experience a lot of stress.  Molds, on the other hand, enable you to do the work ONCE, and then forget about it. You can include all the detail you want, and reproduce another part, such as a fuselage, almost overnight. Granted, you still have to paint or otherwise finish it, but that's a LOT less work than building it from bits of balsa. Another advantage of molding in these details is that there is virtually no weight penalty.

So, onward to the drafting board!

Refine the Plans

Let's assume that you've been able to get (or make) good 3-view images, but have had to "guess" about the actual cross section shapes. You can't make a plug, mold, or flying model without some cross sections from which to create formers and bulkheads, so this has to be done. If you haven't solved this problem during the CAD process, how will you solve it now?

Airfoils are another issue. It's worth a bit of research, to find out two things. First, what airfoils are used on the actual aircraft? Second, are they suitable for a flying model, at the scale you intend to build it? Airfoils for full-size aircraft are not always suitable for models, and taking this lightly is one way to end up with a difficult to fly, easy to crash model. There are a number of airfoils that are proven to work with models, and you should do some research on it. The chances that anyone would ever notice that your model doesn't have "scale" airfoils is very slim. Besides, what's more important... Safety and good flying qualities, or proving that your airfoils are "scale"? Once you have this information in hand, you'll find a number of free, or inexpensive computer programs that will accurately print airfoil templates. These can be used to make balsa ribs in traditional construction, or as templates for making foam wing cores.

Back to the cross sections. I'll use the fuselage as an example. If you have accurate 3-views, you're at least half way there, because you at least know the height and width of the fuselage at every point. You have to decide, based on the shape of the fuselage, how many cross sections are required. It's my opinion that with a typical aircraft shape, a very accurate fuselage outline can be produced with about 15 cross sections. Use more if you're seeking more detail, or if the aircraft has a lot of shape changes along it's length.

Another thing to consider at this point is the assembly itself.  Will the fuselage be a one-piece unit, or will it be "splittable", perhaps in consideration of shipping? Maybe you want the nose cone to be removable, for access to your RC gear. Some items, such as a passenger window, can be molded-in outlines, and cut out later, while others must actually be built separately.

A typical example is a jet model with a removable canopy. If you make a fiberglass fuselage, you'll never get parts to fit properly if you make the whole thing in one piece and try to cut them out later. A better way is to make separate fuselage and canopy plugs, (that connect firmly for sanding and shaping) and each of these parts needs a mating "flange" that you must build in. Each of these parts also needs some blocks or other inserts, to hold the bolts that will connect them. Think about these things now, because you need that information to go to the next step.

Mark all the positions you select on the plans, both on the side and top views. At the least, you can now draw a "box" at each of your cross section marks, indicating the height and width. You can then use a French curve, compass, or just skill, to round the corners of these boxes, add whatever indentations and protrusions exist, etc. It's not easy, and it's not fast. I suggest drawing each cross section on a separate piece of vellum, so that you can stack them and see how one shape will blend with the next. You might consider hand carving a small "approximate" version of the fuselage out of foam, getting the shape close, by using your eyes, and then slicing it up and looking at those slices. Of course, any photos,drawings, and measurements you have of the actual aircraft are the best sources of information.

Once you're convinced that the SHAPE is right, then it's time to consider the STRUCTURE of the model. Hopefully, you'll have enough modeling experience to know how to properly design a strong, lightweight structure to support the shape you have now drawn. If not, do some studying. The easiest way to do this is to browse model magazines for plans and construction articles, to see how these things have been done by others. Ask some of your flying buddies if you can look at some of their old plans or kits. Don't be afraid to ask for help, and save yourself some headaches later.

So, now you have the exterior shape right, you've decided where any "splits" or "breaks" need to be for assembly purposes, and you're almost ready to make parts. Now, think about two different aspects of the assembly process. First, how will YOU assemble either the plug, or the "kit ? Then, how will the model be assembled, once you've created all the parts? Are you going to sell a few kits to your friends?  How will they assemble the parts? Are you going to write a good set of instructions? How will the assembled model be broken down for shipping, or for trips to the flying field? How will you fill the fuel tank, start the engine, check your RC linkages, and so on? All these things should be part of your design process.   What's the point of a beautiful fuselage, if there's no way to install your RC equipment? You don't want to randomly cut holes in it, when you could design a nice hatch just as easily. Take your time, and again, have some other people look at it. An experienced modeler might think it's an "easy" kit, whereas a beginner might be totally lost as to how your design works.

When you're convinced that you have all these things covered, it's time to cut out some parts.

You can view my previous tutorials on plug and mold making, by using the links on the Design and Building page. To continue on to the making of the plugs and molds, using the King Air B200 as the example. click here.

- Mike James