I tried to approach this design with organization and planning, things that I am usually too impatient for. However, my infatuation with the aerospace industry and the enjoyment I get from this hobby made it easier for me to accomplish these. In this article, I will provide you with some of the more important details of the process I used to design and build a Red Bull Air Race inspired plane.
Full disclaimer: This is the first plane I put serious effort into designing as a scratch build. Many of the techniques I used were new to me, especially making somewhat legible build plans. Admittedly, this build is not the easiest or most intuitive but I am always looking for ways to improve it. For any of you interested in building this plane, I am sorry for the headaches you are likely to encounter during the build. In the near future I will make a full build video. I am currently waiting on a new shipment of foam board. Thank you for your patience.
For those of you that don't want to read a long, boring article, I made three videos that cover all of this. They are linked at the bottom of this article along with my build plans.
I used a simple road map to help me with this project.
That last item on the list, improve, is something you can do at any point in this process. As I’ll show you, I made improvements after each of these steps and I can always find ways to improve the design further.
My design phase includes:
A. Creating a list of constraints/design choices
B. Making a sketch of the aircraft
C. Creating a 3D model of the aircraft
D. Using the 3D model to generate build plans
A. Constraints/design choices
A good way to prevent overwhelming yourself with a thousand ideas on what your plane could look like, is to make a list of constraints. These constraints are meant to define your design and limit your ideas to only those that meet them. You don’t need to have a lot and they don’t even need to be that specific. Flite Test already gave me one constraint; it had to be inspired by the Red Bull Air Races. To me, that meant the plane would be quick, it would have large control surfaces, and it would have a Red Bull inspired paint theme. Others will have a different idea of what that means. Either way, that one constraint significantly narrowed down the ideas running through my head. Another constraint of mine was that the plane would have a symmetric airfoil. This kind of airfoil is great for speed and aerobatic ability because it has low drag and it exhibits the same aerodynamic performance inverted as it does right-side-up. My other constraints were a high thrust-to-weight ratio and a note to myself to keep it simple.
Once you have constraints you can make a list of design choices. These are less limiting than constraints and are able to be changed or ignored as you see fit. I wrote down that it was going to have a unique design, easily removable wings, consider forward swept wings, consider winglets, and give it a silver, blue and red paint scheme.
Now that I had an idea of basic performance criteria I needed to decide on what the main features of the plane would look like. I wasn’t too worried about the empennage (tail) because I felt that depended more on the look of the fuselage and I would design them on the fly. For the fuselage, I drew inspiration from some of my favorite WWII aircraft, the P-51D, P-40, and Me-262. I started sketching the side view, followed by the top, of my fuselage. It was at this point that I started scrutinizing the design until I ended up with one that I was happy with. For example, I saw that the elevator was too close to the rudder and I didn’t want to have to make cutouts in the elevator to allow the rudder to move. So I drew another sketch with the horizontal stabilizer moved forward. On the second sketch, I saw that the hinge-line on my rudder was angled. Unless I wanted the plane to experience a pitching moment when I used the rudder, I would need a vertical hinge. I also wanted to move the wings forward some more because the weight of the motor in the front of the plane will be easier to balance out if it's closer to the center of gravity (CG). I made a third sketch with these changes and after a few moments I decided I was happy with this one. I used a similar process for the wings and I made sure that the drawings had a scale appearance to one another, I’ll explain why shortly. I even folded up the paper with my wing sketch on it so I could lay it next to the top view of the fuselage and get an idea of how the wing would look on the plane.
Before I could go any further, I needed to know how big this thing was going to be. I needed dimensions. What I did was pick a major dimension on the plane and visualize how it would look if it were the same size as something in my immediate surroundings. The dimension I chose was the length of the fuselage and the reference object I used was two sheets of foam board lying next to each other with their longest edges touching. I measured the width of these two sheets together and that was 40 inches. This was the dimension that would define the rest of the plane.
Now comes the tricky part, how do I apply this 40 inch fuselage to the rest of the plane and still have it resemble my sketch? Well, remember when I said that you should make sure that the drawings had a scale appearance? That's because you can measure the dimensions of the drawing and scale them using the major dimension you just chose. For instance, if I measure the length of the fuselage on my final sketch, it comes out to 4 and 3/8 inches (4.1875 in). If you divide 40 by 4.1875, you get 9.55. This number represents the ratio of the scale between my sketch and my final model. So now all I have to do is measure the dimensions I need from my sketch and multiply them by 9.55. That will give me the scaled up dimensions for my plane that I then put on a more detailed drawing.
You may notice that I changed my wing design a little on my detailed drawing. That's because keeping the root chord of my wing at 14.3 inches would require the use of more foam board than was necesary. To have a wing with a root chord of 14.3 inches I would need to fold the foam board along it's length (29.75 inches). I wanted to fold the foam board along its width (20 inches) to make my wings. So I modified my wing drawing to have a 10 inch (roughly) root chord.
C. 3D Model
With my design sketched out and dimensioned, I created a 3D model in Fusion 360. I wasn’t going to make the model exactly like the drawing, with curved surfaces but instead using flat surfaces while referencing my drawing. So why did I do this? Well, I had no idea how to make a set of build plans that would translate to a model that resembled my drawing. I had the idea that this would be easier if the plane was already built and I could just disassemble it and flatten in out to see what shapes are used in it’s construction. Obviously, that would require me to build the model using trial and error until it looked right and then tearing it apart just to get plans that would allow me to build it again. So, I did the next best thing I could think of and constructed a 3D model that I could essentially disassemble and provide a more accurate representation of the final geometry of the plane.
D. Build plans
So now I have this 3D model but still no build plans. With the 3D model, I can select all the faces and save them as sketches. These sketches were then saved on my computer as .dxf files. This file type can only be read by specific programs such as CAD, Adobe Illustrator, and Inkscape, to name a few. I used Adobe Illustrator which does require a paid subscription but they do offer a free trial period. I ended up liking it so much that I was happy to pay for it. Inkscape is a much simpler and free version of Adobe Illustrator. I tried using it but only after I had spent so much time with Illustrator. At that point I was already invested and I didn’t really give Inkscape a fair trial. Anyway, now that I had sketches of all the faces, I only had to import them into Illustrator and create build plans. In this program I could move, rotate, flip, and attach the shapes to each other however I needed to. It also allowed me to add and remove lines or draw shapes that I needed to convey important details. I even used Illustrator to create some of the images in this article.
I didn't want to release build plans that required you to figure out how to correctly print them. So I spent a lot of time figuring out how to save the above file as a PDF that was already tiled. All you need to do is make sure you have margins set to 0 on your print setup and hit print. I tried to make these plans as easy to understand as possible without adding too much clutter. On the last page, I added some helpful information. There is a guide to show you what certain lines on my plans mean, numbered notes that correspond to certain parts of the build that have a note tag next to them, a template for the popsicle sticks that the barbecue skewer goes through, and a comment describing how my tile numbers works. I added tile numbers to each page because I always manage to lose track of what page I have to tape to my foam board next. I then have to look at the image of the full plans to figure it out. On my plans, I put numbers near the upper left corner that have the row and column location of each page as they are shown in the full build plan.
The build might be my favorite part. It is during this step that I get to build my design for the first time. But not only am I testing the build, I am also testing the layout of my build plans. As I expected, this part of the process offered the best feedback and quickly made me aware of the mistakes I made during the design process. That’s ok though. As I told myself, you can’t expect to be prepared for everything, especially when you have little to no experience doing something like this. I was ready to make mistakes and I was almost exited when they showed themselves because it offered me the opportunity to troubleshoot and to learn.
The plane is broken up into five sections. Those are the nose, center body, canopy, tail, and wings. Most of the folds on this model are simple creases, similar to the folds used to make a wing. I made score cuts, ran a barbecue skewer through them to open up the cut and bent the plane into it’s final shape. This proved to be rather simple as long as I had something to help me form the final shape. Unfortunately, I did not think of that until after I started the build. So I used my 3D model to create sketches that represented the cross-sections of the aircraft at a few different points. These sketches were imported into Adobe Illustrator and put into my build plans as formers. I cut them out of foam board before continuing my build.
When I got to the tail section, I started experiencing some difficulties. For starters, I hadn’t thought how I would attach the horizontal and vertical stabilizers. I settled on a way to mount them and updated my build plans. Not too long after, I noticed a huge flaw. I had two independently moving elevators. I knew this was going to happen because I remember deciding against making the cutout in the fuselage to allow the movement of a single elevator. What I didn’t think of was how difficult it was going to be to get these two surfaces moving in unison. I spent some time trying to solve this issue only to discover that it would be a waste of time. Even if I did figure it out, I wouldn’t trust it to be reliable and it was an unnecessary design choice. I had to redesign the plane with a single elevator. I also had to start the build over. Fortunately, this gave me a chance to use my updated build plans. The second build went by more smoothly.
In order to join the nose to the center body and the tail to the center body, I had to use a few bevel cuts. Because a lot of the edges that made up these joints were angled, they didn’t make a flat surface to join to the edges of the other section. Bevel cuts easily solved this. Also, after joining these three sections, I went over all the seems with hot glue in order to make them stronger and to make them cleaner.
The wings weren’t too difficult but they employed some techniques that were new to me. From the start, I knew I wanted a symmetric airfoil and that was going to require a very shallow bevel on the trailing edges to allow them to meet up at more of a point than just the thickness of two sheets of foam board. I also made each wing from two separate sections. The the inner wing section had a varying thickness. It was the thickest at the root and thinnest at the tip. The outer wing section was the same thickness as the tip of the inner. One thing I am uncertain of is the strength of the outer sections. The spars I used were very thin and I could flex these sections enough to make me doubt their strength. In a future iteration, I might use something that has more stiffness than foam board for the spars.
You may have noticed that the spar in the outer wing section is not in line with the aft spar on the inner section. It is supposed to be but I didn't even notice that I glued it on the wrong side of the crease until I put the two sections together. Fortunately, this did not affect the joining of the two wing sections. The build plans have guide marks for the placement of these spars to help you avoid this tiny mistake.
Now that whole body and wing were built, I had a few other things to figure out. First, I had to figure out how to hold the wings on. I wanted them to be removable which I figured would require the use of rubber bands. However, I had nowhere to attach the rubber bands. To solve this, I cut out 12 small squares, about 1.5” x 1.5”, and used four 1/4” dowels that I cut to be 1.5” long each. I glued the squares together into stacks of three each. Then I made a hole through the middle of them for the dowels to slide through. These squares gave support to the dowels and acted as a larger surface to glue onto the inside of the fuselage. They hold on to the ends of the rubber bands that stretch over the top of the wing.
I also had to figure out how to attach the removable canopy. After some tinkering, I glued a popsicle stick on either side of fuselage, just behind the wings. These had holes drilled into them that lined up with holes on the sides of the rear portion of the canopy. A barbecue skewer slides through these holes to secure the rear portion of the canopy. For the forward portion of the canopy, I used more popsicle sticks that stuck out from the front of the canopy section and slid under a 1/4” dowel. This dowel was glued across the inside of the nose using the same squares that I used for the dowels that held onto the wing rubber bands.
The last thing was the motor. I had to figure out how to mount it. All I did was make a piece of foam board that would fit snuggly into the nose and act as a platform that would hold up what resembles a small power pod. To get the size of this platform right, I went back to my 3D model and modeled this setup. All the squares used for the dowels and the pieces for the motor mount were then added to the build plans.
Here is the moment of truth. All of my hard work has led me to this moment. Now I get to see if the idea I turned into a physical model will fly or end up in a mangled pile of foam board and electronics. I feel a lot of anxiety just before maddening a new plane. None of those compared to the anxiety I felt right before the maiden flight of the plane that I spent weeks designing and building. By-the-way, if you're wonder why my plane has a light blue color on it, it's because I chose the color scheme from the sugar free Red Bull can. That's what I drink so I thought it was fitting.
I traveled to my flying field with the hopes that no one was there because I had no idea what this thing was going to do once it was in the air. Fortunately, it was just my wife and I at the field. I walked out to the runway with the plane in hand while my wife filmed. I pushed through the intense anxiety, throttled up the motor and tossed the plane. Almost immediately I knew something was wrong. It pitched up and veered left toward the pit area. I tried to control it but it ended up turning back toward my wife. Now I was worried. In the video, you can hear me telling her to watch out. Now the plane was uncontrollably pitching up and almost stalling. Fortunately, it had enough power to prevent in from falling out of the sky and landing on my wife. I eventually got it to turn away from her and go toward the other runway. I was trying to get it to turn right but all it would do was something resembling a poor attempt at a knife edge. Once I had it over the other runway I throttled it down and let it drop quickly to avoid flying it into the trees. I saw it hit the ground nose first. It wasn’t a violent impact but I’ve seen enough of these to know that the nose of my plane was going to look like an accordion. I embarked on the walk of shame. To my surprise, the plane looked relatively unharmed except for a bent propellor. I picked it up and inspected it with disbelief because I wasn’t seeing any damage other than the prop. I walked back to the pit thinking to myself, this is why I didn’t want anyone else here. The funny thing is, my wife had no idea that I lost control of the plane. She also didn’t know that I was very close to dropping the transmitter and tackling her out of the way of the falling mass that was awkwardly, yet elegantly, flailing over her head. I couldn’t figure out what went wrong. My wife noticed that I didn’t have a sufficient hole to allow the air flowing over the esc to escape the plane. I had a large opening in the nose of the aircraft to allow air to flow over the esc but I had not made one for it to escape. I cut out a hole in the bottom lower portion of the center body and went out for another attempt.
Still nervous, I prepared for another launch. This time, I had a little more confidence after the most recent lesson in flying under duress. I throttled up and tossed the plane. Again, it pitched up like it was tail heavy but for this time I was prepared to counter it’s tendency to veer left. I also trimmed the ailerons right just before the flight. Despite this, it was very stiff in turning right. I got it to make what looks like a stable right turn. The only way I got it to turn was by using the rudder. I decided to abort the flight and I turned it back around and brought it down slowly. It began falling a little too fast out of the turn but some throttle made it pitch up and almost stop it’s decent right before reaching the ground. I walked over and inspected it. There was no damage. I wanted to check the surfaces and try one more time
At this point, I was confident that I could handle the plane even though it had a mind of it’s own. Without thinking, I launched it one more time. As soon as it left my hand I knew it was going to get away from me. I immediately aborted the flight and brought it down about 20 feet away from me. On this landing, the tip of the right wing hit the ground with some force and that pushed the left side of the nose harder into the ground. The left side of the fuselage got a slight crease in it from the impact but other than that, the plane was still in good condition.
On the way home, I couldn’t help but notice how excited I was that I didn’t destroy the plane. I also noticed how satisfied I was despite the failed flights. I told my wife that this was the most excitement I’ve had at the flying field since the first few times I ever flew. That was probably because I had a brand new and very meaningful experience. I was reviewing the flights in my head and making assumptions as to what went wrong. I was troubleshooting and I couldn’t wait to review the footage to analyze them more. From the videos I noticed that something was definitely causing it to pitch up whenever I increased the throttle. At first I thought it was just tail heavy but the video showed me that the problem could also be in thrust angle. I went into my garage and measured it. It was angled upward by at least 2 degrees. That’s not a lot but it’s equivalent to as much as 3% of the thrust pointing straight up. If my motor/prop combo produced 740 g of thrust at full throttle (which is actually a conservative estimate), 25 g of that thrust was pulling up on the nose (thanks trigonometry). For a 930 g (2.05 lb) aircraft, including battery, that seems insignificant, and it would be if it was closer to the CG. My propellor is about 16 inches away from the CG and 25 g is about 0.06 lb. 16 in x 0.06 lb = 0.96 lb-in. So what does this mean? I’ll put it this way; it’s like placing a 0.96 lb weight (almost half the weight of my plane) 1 inch in behind the CG. Small forces become more significant when they get farther away from a fulcrum. When I was increasing the throttle, I was increasing the amount of force pulling up on the nose. To fix this, I cut a small piece of foam board and wedged it between the top of the firewall and the top of the nose.
This last step should be employed throughout the process. You will almost always have improvements to make after testing the first iteration but making improvements as you go will make the process more efficient. It will also help you avoid larger hurdles that can set you back. If I made the two separate elevators on my first build work, there is no telling what trouble that would have caused me during the first flight. The uncertainty of that setup led me to make an improvement on the spot. However, I definitely had improvements to make after my test flight. The first improvement was to fix the thrust angle. After that, I decided to make the motor mount sturdier while closing up some of the unnecessary holes. I also made a tapered nose piece to streamline the nose better.
The last big improvement I made was to the wings. The wing tips may have been stalling too soon so I decided to add small winglets. These were only 1.5" tall and as long as the chord at the wing tip. I made them this way because I noticed that my Turnigy EFX racer had very similar winglets. Since I want this plane to be a racer, I thought it made sense to give it similar winglets. Their low profile would keep drag down and they would help my wing tip stall characteristics.
Center of gravity
I wanted to address something that I found to be one of the more frustrating aspects of this project. Determining where to put the center of gravity can be a very difficult process. I was digging through my text books on aerospace vehicle design and aeronautics in hopes of finding an easy solution. I knew how to calculate the location of the center of pressure and I knew that the location of the CG is usually near, or on, this point. However, this calculation requires experimental knowledge. You need to know the lift characteristics of the airfoil. Since my airfoil was not one of the many listed in the National Advisory Committee for Aeronautics (NACA) database, I didn't have any of this data. It turns out that you can't find the right location for the center of gravity by using theory alone. You have to test the performance of the aircraft as well. This is something that most of us are already accustomed to anyway.
I decided to use a more practical approach and assume a good location for my CG until I could test it and make adjustments. Usually 1/3 the chord length, or 1/3 of the way behind the leading edge of the wing (for straight, rectangular wings) is a safe place to establish your center of gravity. If your plane has swept wings you need to find what's called the mean aerodynamic chord (MAC). To do this, draw two straight lines, equal to the length of the root chord, above and below the wing tip. Then draw two straight lines, equal to the length of the tip chord, above and below the wing root. Connect the ends of these four lines with two straight lines. The point where these lines intersect is the location of the MAC (see the images below). The MAC is what you use to determine the placement of your CG.
I decided to initially place my CG at 1/4 the MAC because this is usually a good approximation for the location of the aerodynamic center (center of pressure). The aerodynamic center is the point at which all the aerodynamic forces are balanced. Generally, the closer to this point that your CG is, the better your aircraft will perform. However, this is not always true. It is a good idea to keep you CG located somewhere between 1/4 chord and 1/3 chord. Moving it beyond either of these ranges can have some very unfavorable consequences.
You can perform this step after your build but I prefer to wait until I know my design works before I waste time painting it. Don't get me wrong, if my design did not survive the test flight I would rather watch a beautifully painted plane get destroyed than some monotone thing that ends up resembling a crumpled pile of cardboard, but I don't need to waste the time and resources. However, I had trouble making enough time to get out to a flying field to maiden it for over a week so I decided to use that downtime and paint it anyway. It's ok if you deviate from your plan a little bit, as long as it doesn't disturb the flow of your process. You may have noticed that the blue on my plane is lighter than that on a standard Red Bull can. I chose this color because I wanted it to have my own personal touch. These are the colors on the sugar free Red Bull can. That is what I drink almost every day. Also, because I am CptCrazyFingers, I call my plane the Crazy Flyer 1 (CF-1).
I probably spent a little over 4 weeks working on this project and it was so much fun. I learned a lot of new things and experienced new challenges. It gave me a very different perspective on this hobby and I thoroughly enjoyed it. It also gave me a deeper respect for everyone at Flite Test, and in the community, that provide us with quality foam board airplane designs and build plans.
I know this design is incomplete until I can prove it flies reliably. I waited a long time to finish this article because I was trying to offer a better design. Unfortunately, I didn’t have much time to make it out to my flying field and I had the problem with my warped foam board. I won’t get a new shipment of foam board until August 13. If I ordered it sooner, I would have rebuilt the plane and put it through more flight testing. I will do this when I get more foam board and I will update everyone with more details on it’s performance. I will also make a complete build video if anyone is interested in having better instructions on building it.
I’ll end with a video of my last attempt to test it’s flight characteristics. I thought my back yard was large enough to do a powered glide test. I wanted to throw the plane and see how it would glide after making a lot of modifications. I was going to use throttle to control the glide slope and prevent the plane from taking a hard nose dive. As I found out, this was not a good idea. Because I didn’t have a lot of room, I was nervous and I didn’t give the plane enough throttle to prevent it from dropping like a rock. Fortunately, this little hiccup didn’t damage the airframe. It wasn’t a total waste of time because it showed how durable the plane was and it showed me a possible weak spot in the wings. I slowed the video down to show how much the wing flexes when it hits the ground. Interestingly, there were no signs that the flexing ever happened. However, I don’t think the joint between the inner and out wing section (where it flexes the most) is as strong as it should be. I am going to try and build the next wing with a stronger spar between the two sections.
Thank you taking the time to look at my Red Bull Air Race inspired plane. Please feel to offer constructive feedback. I know there are a lot of you out there that have much more experience with this than I do so please enlighten me.
Don't forget to make sure that your margins are set to 0 in your print settings before printing these.
Updated build plans (8/9/18)
As MikeRobey pointed out to me in the comments, I may need to have a rightward thrust angle. I figured that it would be immune to torque roll from the motor because it has a significant wingspan. However, it's probably safe to assume that it's affected a little by torque roll. So I added a slightly shorter motor pod side plate to the build plans. I didn't remove anything, I only added this one piece with a note explaining what it's for. This side plate is about 0.08 inches shorter which will translate to about a 2 degree rightward thrust angle. I will continue to make updates to this article. I will not delete anything already here but I will label the updates and offer an explanation.
Updated build plans (8/13/18)
I realized that the support pieces used to hold the long dowel across the nose are smaller than those used for the wing rubber bands. I have updated the plans to include these 6 smaller support pieces. I also added markings on the front of the center fuselage piece to show where these supports should be placed. Always make sure that this placement will work before gluing it down. It's possible that my tolerances were too small and you may end up with misaligned pieces. On that note, I am also looking into redesigning the way the canopy is held onto the plane. I want to use techniques I've seen in other FT builds such as creating some sort of friction fit or doublers on the front of the canopy piece that fit into cutouts in the nose.
Check out my second article on the successful test flight
Updated build plans (8/14/18)
Another update to the build plans to improve print quality. I had to decrease the size of each tile by 0.1 inches because the dotted outline on the sides of each page was just barely being cut off by the printer. This wasn't affecting the way the plans fit together, it just made it more annoying to do so. All the pieces are the same size I just got rid of some more white space to allow the dotted blue line that goes around each tile to show up in the print. Also, I am currently trying out a new way to hold the canopy on that doesn't require any popsicle sticks. Only foam board and a barbecue skewer.
Updated build plans (8/14/18)
I am so sorry but I noticed a problem with all versions of my build plans. On the center fuselage part, I don't have blue lines (indicating a score cut) on the two middle cuts. I also have lines that are on the sides of the center fuselage, at the back of the wing cutout, that don't belong. I am terribly sorry if this caused any grief. Here is a correct copy of the build plans.
Updated build plans (8/23/18)
I had some flaws that required fixing and I also made some changes to the design. I added pieces that help the front of the canopy piece catch without using popsicle sticks or a dowel. The popsicles sticks that go through the back of the canopy were left in. I also added some angle guides that help keep the leading edges of the wings at the proper angle while folding them.