CHAPTER 11
Hack the Flack: Make and Fly Your Own Design
Abenefit of the Brooklyn Aerodrome approach comes from getting to put something strange and fun into the air. This chapter takes you through the process of creating planes out of whole cloth from design to execution. You will have joined the select few who have designed, built, and flown their own remote-control (RC) airplane. It is assumed that standard Flack gear will be used (i.e., motor, servos, etc.) and similar building techniques. There are lots of ways to make airplanes; this is one of them.
The best part is that it is not that hard if you approach it methodically and with a bit of patience. This is not a good place to start as a beginner, but you don’t need to be an expert either. Basic building skills and piloting skills are all that are needed. That said, this is the deep end of the book, where you are expected to fill in all the details. I had three testers (thanks Ben, Lowell, and Andrew), and it went okay, but not rock solid like the first five chapters. This chapter lays out basic frameworks for novel flying wings. It is not going to work well for traditionally laid-out designs or canards—so it is limited (I find that I do my best work with constraints).
Two major approaches to novel aircraft are presented. One is incremental, based on starting with a small glider and working your way to a powered RC aircraft. The other approach works by removing material from known good designs to achieve the design objectives. And remember to send me a picture/video at [email protected].
Getting the Idea
What gets your creative juices flowing? Do you want a night flyer? Do you want something that looks like an airplane or not at all like an airplane? I constantly seek inspiration from the animal and vegetable kingdoms. The flying-wing approach I pursue is well suited for big creatures, abstract shapes—How about some flying lips?
Prototyping
Once you have some ideas, the key issues are whether the shape can generate lift and be controllable. There are computer programs out there that can reduce the risk of a new design by modeling the lift, control, and structural issues around your idea, but I have never used them because there is a simpler way—build a half-scale model glider and see if you can get it to fly.
Figure 11-1 shows some of the gliders I have built when trying out new ideas. Blue foam is an excellent material to “sketch” with, and I use subway cards as stabilizers. Getting a glider to fly can be a little tricky because you may not know what the center of gravity (CG) should be or appropriate trim. Below are some steps you can take to maximize the chance that your glider can be made to fly. The first glider to make is one with known properties that are easy to work with—the classic plank flying wing.
FIGURE 11-1 Various 50 percent scale gliders used to prototype new shapes.
A Plank
I strongly recommend that you build a plank to help calibrate your expectations as to how to launch, trim, and evaluate prototype gliders. It will take 5 minutes of your time once you have the materials. The steps include
1. Cut a 12- × 8-inch plank from foam or whatever you are using.
2. Bend up the last 1 inch of one edge to have 1/16 inch of up trim for reflex. This will be the trailing edge. A good way to do this is to place the wing on the edge of a table with 1 inch hanging off the edge. Take a ruler, and use it to force the bend on the 1-inch section.
3. Split the elevons in the midpoint, and keep cutting farther so that a business, playing, or subway card can be inserted vertically into the slit.
4. Add the business, playing, or subway card as a stabilizer in the center slit.
5. Launch the plane. Observe how it flutters to the ground in a very nonflying way. This is a plane with a very far aft CG. The next steps will predict where the CG should be.
6. Measure 2 inches back from the leading edge, and mark both the top and bottom with a line parallel with the leading edge. This line is the 25 percent chord point of the wing. Chord is the dimension of the wing that is geometrically parallel with the direction of flight. Chord does include elevons.
7. Flying wings work well if the CG is between 20 and 25 percent of the mean aerodynamic chord (MAC) of the wing. Twenty-five percent MAC is the point at which fore and aft split the surface area of the wing 25 percent/75 percent relative to the direction of flight. For the plank design of this example, the calculation is trivial—I just did it by measuring from the leading edge. For strange shapes, MAC can be quite difficult to determine.
8. Add sufficient weight at the nose to have the plane balance at the 25 percent chord point. Put the weight on the bottom side to give your finger a place from which to launch. The plane in Figure 11-2 balanced with three quarters and two pennies.
9. Launch the plane with a firm flick of the wrist, level in both pitch and roll. Or you can launch like you learned with the Flack. Do not throw it up. Level launch is the goal.
a. If the glider pops up, then reduce up trim a little.
b. If the glider dives, then increase up trim.
c. If the glider turns left or right, adjust with opposite elevon trim.
10. Keep at trimming the glider and improving your launch technique until the glider is flying at least 20 feet. This glider is your reference for trimming and evaluating future gliders. Some gliders will fly worse, some better, but the plank is helpful for assessing likely performance and viability of your idea.
FIGURE 11-2 Plank flying wing with 1/16-inch up trim and nose weight.
Designing and Building a Novel Design
Building your own design glider is a little more open-ended because I can’t possibly know the degree of your genius. But this is how I proceed:
1. Draw the initial design in top view on graph paper with a 24 × 48 grid outline. Scale it to what you think will be a 50 percent scale glider, and cut it out of foam. The grid will really help with scaling to a full-size model.
2. All flying wings require some reflex or up trim to be self-stabilizing in pitch. Figure 11-1 shows how I have bent the trailing edge up a little bit on various shapes. Sometimes an elevon needs to be explicitly cut, or the trailing edge can be just bent up as with the plank.
3. Add a stabilizer as far back as possible from the leading edge or where the design calls for it.
4. If possible, determine the point on the wing that splits the wing area 25 percent/75 percent in the direction of flight. Add enough weight to the nose to achieve that weight distribution. I just pick a point one-quarter of the way back on the chord of the wing and add weight until the glider balances on my fingers at that point.
a. Launch the glider with attention to elevon adjustments for flight path.
b. If elevon trim is not working or looking too extreme (see reference plank build), then adjust nose weight in very small increments (one penny) as follows:
i. If the glider is diving, then remove nose weight.
ii. If the glider is climbing and stalling, then add nose weight.
c. Keep working with nose weight and elevon trim to optimize stability and glide path.
d. If you like what you see, consider building a full-size version.
5. If the shape is too complex to determine the CG point (e.g., the Banana or Carrot), then apply the “that looks about right” (TLAR) method of adding nose weight. I have never built a glider that did not need nose weight. This makes it harder to know whether the CG is correct, so I fix the elevon trim at something believable (again, consult the reference plank) and add or remove nose weight to figure out what the CG should be in one-dime increments.
a. If the nose is rising, then add weight to the nose.
b. If the glider is diving, then remove nose weight.
c. If the CG adjustments are not working, then adjust elevons further. Keep trying things until you get the glider flying well.
d. If you like what you see, build it full size.
6. Assess the directional stability of the glider. If it is falling off left or right, add more stabilizer area to try to keep it flying straight. Remember that this may move the actual CG away from the design CG.
7. Be forewarned that actual glider CGs tend to come out too far aft for an RC airplane. Factor this into your anticipated weight distribution in the full-size build.
Building the Powered Version
If you have a glider that flies about as well as the plank, then there is an excellent chance that a powered version will fly as well. The next step is to size your creation. But please read this entire chapter before proceeding. Many factors influence scaling to a powered prototype that have to be considered simultaneously, and books are inherently linear.
Scaling Up from the Glider Proof of Concept
This book assumes that the same basic materials that I used for the Flack will be used. These include foam, speed control, motor, servos, and battery.
Wing Area
The Flack is sufficiently powered and has enough wing area that almost any shape you can get flying as a glider likely will work if you stick with the wing area and overall weight of the Flack. An easy way to calculate wing area is to draw the design in SketchUp, select its surface, and use the Entity Info menu, which will pop up a window showing the surface area of the design.
Another way to measure wing area is to draw a 2-inch grid on the design and count how many squares there are. For partial squares, just guess how much of a square there is—the measurement does not need to be exact.
The Flack has 425 square inches of wing area, including elevons, so try to size your design to match that. Significantly more wing area (100 square inches or more) will make the airplane slower, assuming no changes in weight. Significantly less wing area will make the airplane faster. Look at Chapter 8 for an idea of how surface area affects performance across the designs.
Also consider the natural shape of your raw materials. Blue foam comes in 24- × 48-inch rectangles. Keeping wing spans under 48 inches will mean that joints won’t be needed.
Weight
The Flack is very happy with a flying weight between 15 and 22 ounces. But the delta-wing design generates lots of lift, so don’t assume that you have the same weight-carrying capacity because you have the same wing area. If your design is very “un-wing-like,” then a good move is to keep everything as light as possible. It is hard to define what “un-wing-like” means, but some examples include
1. A flying superhero—this has actually been done.
2. Flying the carrot pointy-tip first.
3. A flying doily with lots of holes in it.
If the test glider flew less far than the plank glider by 50 percent, then you likely have a shape that is “un-wing-like.”
That said, almost any flat shape will generate some lift, so keeping it light increases the chances of success. The minimum weight possible is around 13 ounces for Flack-class designs just because of the equipment and foam. Often what makes an airplane heavy is the weight added to achieve a balanced CG. Also, careful placement of the battery and motor can really help to keep the overall weight down.
Control
The glider prototype should have forced the design into being statically stable, but this doesn’t mean that there is a way to control the aircraft. Many a design has failed because I could not work a way to put elevons on it. Some failed efforts include
1. A flying tadpole
2. A flying candy cane with handle forward
This is not to say that it cannot be done, but that it is challenging.
Sizing Elevons
Chapter 9 is a good place to get a sense of how elevons should be designed on novel shapes. Keep the area similar to a Flack’s elevons, which are approximately 100 square inches total. Another metric is to have the elevons be one-fifth the surface area of the entire wing. Reasons to have bigger elevons include the elevons not being in the prop blast, which reduces control authority, and having elevons that are oddly shaped.
Elevons also can be too big, which can result in servos that are overwhelmed with the aerodynamic forces on them or an airplane that is overly twitchy on the controls. If you don’t occasionally use full control throws (e.g., landing, launching, and acrobatics), then consider making the elevons smaller if the control throws are at 45 degrees or less.
Sizing Stabilizers
I do not test gliders to evaluate stabilizers generally. I use an expired subway card to provide yaw stability and do my serious thinking when the full-size design is being built. The Flack flies just fine with one stabilizer, and with deft piloting, it can be flown with no stabilizers at all. Generally, stabilizer size is more than sufficient for aesthetic reasons.
The Flack with a single stabilizer uses 32 square inches starting 2 inches from the CG and ending 10 inches away to provide yaw stability. The Flying Heart uses a single stabilizer starting 12 inches from the CG with a surface area of 60 square inches—it has more stabilizer area than probably is needed, but it looks good.
The one case where stabilizer area gets tricky is if there is a significant destabilizing vertical area ahead of the CG. Figure 11-3 shows a Flying Tadpole prototype that had a large dome in front of the CG. That surface area had to be compensated with a sizable vertical stabilizer for stability, as shown. The size of the stabilizer was determined by taking the vertical area of the Tadpole body and making the stabilizer 1.5 times that area. During flight testing, the stabilizer area was reduced gradually until the plane became slightly unstable. The area of the last stable configuration was used.
FIGURE 11-3 Flying Tadpole with big destabilizing nose and compensating tail.
Structural Considerations
Once the basic airframe has been cut out and the elevons cut and hinged, it is time to decide whether reinforcement is needed. Some type of reinforcement almost certainly will be needed for the motor mount and servos. If the final airframe will be stiffened by covering, the prototype can use Coroplast as a proxy stiffener. There are no hard and fast rules, and experience dominates—I have made many overly floppy prototype airplanes.
Build with strength in mind. My prototypes crash pretty hard about half the time on first launch. Generally, this occurs because I got the CG entirely wrong. I am getting better about this, but it still happens. Coroplast is your friend in such situations.
Motor-Mount Attachment Points
The motor generates a lot of force on the motor mount when flying and particularly on landing. It needs to be able to take abuse. The standard motor-mount attachment from the angle stock to the airframe is not always possible, however. I have sandwiched the motor mount between two sheets of blue foam and zip ties, as done with the battery, or used high-stick tape to attach it to adhesive plastic film.
Servo Attachment Points
Servos need solid mounting as well. High-stick tape on plastic film can be strong enough on its own to hold servos down.
General Stiffening
The standard stiffening agent for blue foam is either a layer of Coroplast or a layer of plastic film. Both work really well to stiffen, and they provide a solid surface for motor mounts and servos. Another route for stiffening was used for the Bat, as shown in Figure 11-4. It was doubled blue foam, and it worked very well and has endured many flights, firm landings, and transportation. Packing tape was used to stiffen the opposite side of the motor mount. Use the Flack’s stiffness as a guide to determine whether your design is stiff enough. Long, thin wings are going to need some help. Look at the shapes in Chapter 9 for guidance on what does and does not need stiffening.
FIGURE 11-4 Details of both foam doubling for stiffness and short-rod servo installation on the Bat.
Placing Equipment
It is a very good idea to experiment with different placements of equipment to attempt to achieve the expected CG as determined by the glider. The heaviest items are the most useful for this, and these include the battery (3.4 ounces), followed by the deck (2.5 ounces), with the motor/prop/motor mount coming in at 2.1 ounces. The rest of the components are less than an 1 ounce each, so they are less likely to have a big impact on airplane balance. Remember that the glider has determined the design CG that needs to be achieved.
Almost all of our designs at Brooklyn Aerodrome would fly better with the motor on the front, but our desire for safety means that the hard metal bits need to be surrounded by foam. Please keep safety a major component of your designs as well.
Flight Testing
Nothing could be finer than giving a new design a huck into the wild blue yonder. These are the steps for a new design’s first flight:
1. Do yourself a favor and verify that surfaces move the right way both in the studio and on the field before your first launch. The 3D Banana was pounded into the ground by reversed aileron control.
2. Verify that the CG mirrors that of your glider. Make it 1 inch forward of that location because the gliders tend to have aft CGs. A plane with a too-far-forward CG is flyable, whereas a plane with a too-aft CG tends to be uncontrollable.
3. Have ¼ inch of up reflex with your elevons.
4. Have someone who knows how to launch launch the plane. Have that person practice on a Flack if they are not sure.
5. Consider different ways of launching the plane. The Banana planes are both launched from the middle with an underhand toss.
6. Fly over tall grass if at all possible. It really cushions crashes.
7. There are two schools of thought about whether to apply power on first launch. It depends on the wing loading of your design, how fragile it is, and how soft a spot it will land on. If there is knee-deep wheat fragrantly wafting before me, I will do my first launch without power if the design has a remote chance of gliding. Here in Brooklyn with hard-scrabble dirt or asphalt, I always launch with power—this generally results in a crash, which is why prototypes need to be built tough. The steps for each approach are as follows:
a. An unpowered launch needs enough airspeed to test the aerodynamics. Launch just like you did when learning to fly the Flack with a pilot on the controls. Level, firm, and set to land 10 feet out. Make adjustments and repeat until that throw is controlled. Then move to the powered launch approach.
b. For a powered launch, the first rule is to launch at full throttle. I have failed to do this and crashed. Keep at the launching, making adjustments based on the following.
8. Fly the plane like a beginner. Just get level flight to 50 feet out and land. Then start pushing maneuvering.
9. If the plane has anything wrong or is difficult to control, then cut power as much as possible, and get the plane down.
10. Be ready with up elevator. If it is very hard to get the nose up, then consider
a. Moving the CG back ½ inch
b. Making the elevons bigger
c. Getting a more powerful motor
11. Be ready for the airplane to be very squirrelly in pitch. If it is, then the CG is likely too far back. Move it 1 inch forward.
12. Be ready for the airplane to tuck or dive uncontrollably. This can be from a too-far-back CG or an airframe that is too soft and curving in flight. The latter produces a huge pitching-forward moment that can overwhelm the elevons, as happened with the 2D Flying Heart.
13. Video the flight tests. It can be very useful to have a record of what actually happened. In addition, if it works, you can put it on YouTube.
14. Don’t fly too much the first time out. Go back to the lab and think about how the airplane is flying.
15. As you get to know your new design, try small changes once you have it roughly controllable and flying. Move the CG ¼ inch in various directions to see what it does. Increase/decrease elevon throws, stabilizer sizes, and elevon sizes.
New Shapes by Morphing Old Ones
Figures 11-5 and 11-6 are examples of creating a new shape by slowly morphing a known shape into the desired shape incrementally. It is an excellent way to adapt to extremely radical designs. The Flying Chandelier started out as a traditionally laid-out plank wing with a tractor motor. It flew well, as expected.
FIGURE 11-5 Flying Chandelier with planned wing outlines in electroluminescent wire.
FIGURE 11-6 Servo and other gear installation shown for the Flying Chandelier.
Incremental Refinement
Next, the top and bottom edges of the Chandelier arms were cut out on the wings, and the airplane was test flown with the CG moved back a bit to compensate for the loss of wing area. Flight performance was slightly degraded but acceptable.
The Chandelier needs four arms, so the vertical stabilizer was fashioned out of foam, and a bottom arm was made out of Coroplast to withstand being landed on. This version was flown successfully as well with increasingly degraded aerodynamics.
The goal was to make this airplane a night flyer with electroluminescent wire (el-wire) outlining the shape of the arms. The wings eventually would have had clear material where indicated in Figure 11-5. Flight testing pointed out that the design needed to be made bigger to slow it down and make it more controllable. Ultimately, this design was abandoned for lack of interest, but the strategy of incremental refinement worked perfectly and pointed the way to needed design changes as well as validating the overall idea.
Conclusion
This chapter ties together the sorts of knowledge and techniques needed for creating your own designs. It is very fun and rewarding—there is nothing like launching a completely new shape with weird aerodynamics and seeing it work. After my first test flights of the Flying Heart, Bat, Manta Ray, and many others, my hands shook from the excitement of the event. I have to keep first flights short because I am afraid that my nervous thumbs will crash the plane. It is an awesome experience to have. Send me a picture/video of what you come up with at [email protected].