CatsEyes Aerial Photography

Cat's Eye 2

Page 3: airframe

OK, here it is. The Cat's Eye 2 airframe.

Figure 5: Cat's Eye 2 airframe
Figure 5: Cat's Eye 2 airframe

Isn't it beautiful? And so simple! Well, OK, there are a few little extras and accessories missing, such as the camera mount. And landing gear. And empennage. And wings. Details.

When considering the airframe, I wanted to go with a fully modular approach, so that if any particular part of it isn't working out (or is damaged in a crash), I can replace just that part easily. I particularly liked the GWS Slow Stick's approach of using a "stick" as the main fuselage structural element, with all of the other needed elements mounted in various positions along the stick. This not only results in easy building and swapping of components, but allows you to move some components forward or aft as required, for example to adjust the CG.

The actual Slow Stick's fuselage boom "stick" is a (roughly) 3/8" square aluminum tube. It's walls are quite thin and the whole thing quite flimsy. It works fine for the Slow Stick, a light park flier, but I wanted something a bit sturdier. Luckily the local hardware store sells 1/2" square aluminum pipe, with 1/16" thick walls, that will do nicely.

The fuse boom "stick" forms the basis for the whole airframe. Now that that's taken care of, it's just a simple (!) matter of adding the rest of the necessary components.

Figure 6: Airframe and components

Figure 6: Airframe and components

Below I will describe each of the "modules" that are attached to the fuse boom. Before I go any further, however, I should of course add a big caveat. So far (Jan. 2011) this is mainly just a CAD model, and any and all of it is subject to change without notice. In particular, I still have to go through and calculate where the CG will be, and this may necessitate moving things around a little. Or a lot.

Design considerations

The main design criteria I was looking for are as follows:

  1. Cameras with unobstructed view forward, down, and to the sides
  2. Hand-launch (fuselage at CG to grab onto)
  3. Modular
  4. K.I.S.S.: 3 channel

Camera mount

Figure 7: Cat's Eye 2 Camera Mount

Figure 7: Cat's Eye 2 Camera Mount

I have covered the camera mount and the test apparatus fairly well on page one. Just a note here on how the mount is actually mounted on the plane.

I decided to mount the pan servo ahead of the pan bearing, as it's conveniently out of the way there and doesn't interfere with any of the other elements (motor mount, wing saddle, landing gear). Other than that, it's pretty much the same as the test apparatus. I am considering a slight modification to allow a boom to extend forward to allow the mounting of things to appear in the field of view of the cameras. For instance, a simple upside-down "T" in line with the horizon would provide a visual indication of the aircraft's pitch and roll. I could also mount various sensors and indicators if I'm too cheap to go with an on-screen display (OSD) option.

Most camera mounts I have seen for FPV have the camera mounted on top of the fuselage. For FPV this adds to the impression of being "in the plane" and adds to the realism of flying the plane. However, my primary focus is photography (no pun intended), and thus do not want anything in the way to muck up the photo. I also want to be able to shoot straight down without anything in the way. That was the reason for having the bearing on top.

Motor Mount

Figure 8: Cat's Eye 2 Motor Mount

Figure 8: Cat's Eye 2 Motor Mount

Yup, the Cat's Eye 2 is going to be a twin! Since the camera has to be out front, and I don't want to be shooting through a prop, that means the prop (or props) has to go somewhere else. I could have made it a pusher like the Cat's Eye 1, but with the fuse boom being just under the wing, prop clearance would be a problem unless I mounted it way up high on a pod. Somehow I never liked that solution. So I've decided on a twin. This makes it very easy to keep the props clear of everything.

Usually with a twin the motors are mounted on the wings. However, there's no law that says this has to be the case. Since there are no ailerons, I was not going to have any wires going to the wings to plug and unplug every time I put the wing on, and I wanted to keep it that way. So I decided the motor mount could go on the fuselage this way.(Thanks to Dan McLeod for this suggestion!)

As usual, a lot of my design decisions have to do with what I have "floating around," in this case a pair of TowerPro 2408-21T brushless motors. They pull about 27oz static on a fully charged 3S LiPo, going down to about 23oz for a half-discharged pack. Say 25oz for easy figuring, between the two of them I should get 50oz of static thrust. Unless this thing ends up a lot heavier than I expect, that should be plenty.

By the way, the blue things on top of the motor mount in Figure 8 are the ESCs. My abplogies for the strange state of the "wiring" in Figure 8. I never really did figure out how to (if you can) do wiring in the CAD package. Obviously, the motor wires will be connected to the ESC wires.


Figure 9: Cat's Eye 2 Undercarriage

Figure 9: Cat's Eye 2 Undercarriage

The criterion that I want to be able to shoot straight ahead means the camera has to be mounted on the front of the plane. The criterion that I want to be able to shoot straight down means that I can't have anything underneath the camera mount to obscure the view. This lead to the problem of how to locate the landing gear.

A number of solutions occurred to me. First, don't bother with landing gear at all. This was what I did with the Cat's Eye 1. The problem with this plane is that the camera is out front and sure to hit something on landing (unless I want to hand-catch it, something I have never done and don't want to start now). To prevent this, I have to keep the cameras mounted up fairly high off the ground. With a tall fuselage behind it, though, it will just tip over on landing, right onto the cameras!

Another solution is to mount two wheels widely-spaced and allow the camera to shoot down between them. I have toyed with this idea, but it would require the wheels be mounted on sturdy booms and attached in a sturdy way to the rest of the plane, and I had trouble with this design.

The solution I adoped was to mount the landing gear just behind the camera. I'm going with one big wheel, as shown in Figure 8, above. (For a while I was toying with the idea of calling the plane the "Big Wheel." However, I think the associations are less than flattering with that name (I think of a large, plastic 1970's child's toy). And besides, after testing I may have to dispense with that idea and go with something else.) The bigger the wheel, I reasoned, the less it should be affected by small bumps in the "runway" and thus the less likely it should be to tip over on landing. So goes the theory.

There is obviously, a balance issue with only one centre-mounted wheel. However, my plan is to hand-launch the plane, and on landing the plane should stable while there is airspeed to keep the wings "flying". After that, yes, it will tip over; by then it will be slow enough that that won't present a problem.

Behind the big wheel is a thing I rather whimsically call the "gondola." Its purposes are 1) to provide a good, big area to grip for hand-launch, and 2) to provide a "tray" for the electronics and the battery. The tray will be long enough that I can move the battery fore and aft as required to balance the CG. Of all the modules, this is probably the one that is most likely to undergo a redesign, as it is dependent on where the CG falls out (and what needs to be moved around to correct it), and on what the mounting requirements are for the electronics, which has yet to be determined.


Figure 10: Cat's Eye 2 Empennage

Figure 10: Cat's Eye 2 Empennage

The empennage is envisaged to be a fairly direct copy of the Cat's Eye 1's tail. In keeping with the modular approach, I will have one "module" for the horizontal surfaces (horizontal stabilizer and elevator), and one for the vertical surfaces (vertical stabilizer and rudder). The servos for the control surfaces are mounted directly to the relevant module, allowing the module to be removed without undoing the linkages. Also shown here is the tail skid.

As with the Cat's Eye 1, the surfaces are 6mm Depron, reinforced with a carbon fibre bar of roughly 1/16" x 11/64". This has worked well on the Cat's Eye 1, and my Long-Nose Getter, so I'm continuing with the "tried and true."

The differences between this and the Cat's Eye 1 are the mounting to the fuselage boom, having the horizontal stab mounted under the boom, and making the surfaces a bit larger to reflect the larger overall size of the plane. The Cat's Eye 1 didn't need a tail skid, so that necessitated putting the horizontal stab on top of the fuse boom to prevent damage on landing. In the Cat's Eye 2, the tail skid will keep the empennage well away from the ground.

Up next

As you can see I have a preliminary design more-or-less in the bag. I still have a few details to work out. In particular, I am going to have to try and estimate weights and moments for all the components in order to determine where the CG is going to fall, and there may be some adjustments (possibly major ones) required depending on the results.

I also have to finalize the electronics in order to determine the mounting requirements.

And of course the "design" is never completely finished until the thing is built -- I always seem to be fiddling with one thing or another at the last minute.

However, this will certainly serve as a working design, and certain things I can start building in parallel with the rest of the design work. I will be posting information on the "build" as it proceeds here.

Coming soon: The build log.