Modular F-104 70mm EDF Jet development and testing
- Thread starter telnar1236
- Start date
telnar1236
Elite member
I tend not to with the exception of SAFE built into the receiver of my E-Flite F-16. My experience has been that I don't notice too much of a difference and that in cases where I crash, a gyro probably wouldn't have saved the plane regardless. I do want to look into flight controllers that have autopilot capability, though, since it would be nice for planes to be able to fail safe into a holding pattern instead of an intentional crash if I lose signal.
telnar1236
Elite member
The CAD for the first modular prop plane is coming together now too. I ended up completely throwing out the original prop design I posted because of how hard it would be to make it compatible with tricycle landing gear and came up with this instead.
It will have tricycle landing gear and 3 channel control (throttle, ailerons, and elevator) plus rudder steering on the ground. The wings, fuselage sections, and vertical stabilizer are compatible with the F-104, although I don't plan on testing it in that configuration. Because of the delay with the F-104, I'll probably actually get around to testing the modular prop setup before the jet trainer configuration.
It will have tricycle landing gear and 3 channel control (throttle, ailerons, and elevator) plus rudder steering on the ground. The wings, fuselage sections, and vertical stabilizer are compatible with the F-104, although I don't plan on testing it in that configuration. Because of the delay with the F-104, I'll probably actually get around to testing the modular prop setup before the jet trainer configuration.
Last edited:
We have many windy days here in Kansas City. I'm getting more involved using flight stabilization systems (6-axis) in order to increase my flying time. My current setups on the Flea Fli and the Rapier use the A3 Pro from Motion RC. It does make a difference for me.
How long does it take your printer to print a f-104 fuselage? How about a wing panel? How do you join the upper and lower wing panels, or is the wing solid?
How long does it take your printer to print a f-104 fuselage? How about a wing panel? How do you join the upper and lower wing panels, or is the wing solid?
telnar1236
Elite member
Huh, I actually find gyros most intrusive when it's windy. It feels like it makes the plane fight the wind instead of fly with it. No single correct way to enjoy flying.
I tried to optimize the fuselage sections to take less than 12 hours each to print so that I could start one going before work, then come back and start another one over the night. Some of them do take closer to a full day, though. I printed both wings at the same time in a roughly 26-hour print, but to get a better surface finish, you could print one wing at a time in about 12.5 hours. Using two printers, it takes about a week to print the whole thing, so about two weeks with one printer. The wings on the F-104 are printed standing up with the wing root on the bed. The leading edge, flaps, and ailerons all glue on. For the F-104, the internal geometry is explicitly modeled, but I'm trying a solid with 2% gyroid infill for the wing for the first modular prop plane and the results seem promising. Being able to model the wings as solids would save a lot of time and make it much easier for others to design custom wings. The internal ribs are at 45 degrees which removes the need for a dedicated spar except for where the wing joins the fuselage. The servo pockets are small enough that they can printed without supports even though they have a 90-degree overhang. For the modular prop plane, there are 3 sections to each wing which are joined using CA glue (inboard, outboard, and wingtip) but the F-104 for has a small enough span that each wing is short enough to be printed in one go.
telnar1236
Elite member
The good news is that the F-104 flies. The bad news is that the recording stopped a few seconds into the flight - my RunCam is pretty beat up and the battery sometimes shakes loose. The other bad news is that I ended up crashing again. This time, the problem was two-fold. First, while I added a second spring to the pitch axis on the transmitter, I didn't add one to the roll axis which made the plane very sensitive to roll inputs (and my rates were probably a bit high). Second, I overcompensated for the pitch-up problem from last time and this time the plane was extremely nose-heavy. I was able to get it mostly stable (you can kind of see this at the very end of the video) and fly a bit longer after the video cuts off, but it wasn't a long flight.
The good news is that it has enough power on 4s, which I was a bit worried about and held together up until the crash. On to version 3! The only changes this time will be to reduce the nose-weight a bit and changes to my transmitter rates. Hopefully nothing else goes wrong!
Realistically, the prop design will actually fly before the jet now (which would maybe have been a good idea to start with).
The good news is that it has enough power on 4s, which I was a bit worried about and held together up until the crash. On to version 3! The only changes this time will be to reduce the nose-weight a bit and changes to my transmitter rates. Hopefully nothing else goes wrong!
Realistically, the prop design will actually fly before the jet now (which would maybe have been a good idea to start with).
telnar1236
Elite member
Precisely. Pretty hard to control roll when you have the elevator pinned full back just to achieve level flight.A perfect example of PIO (pilot induced oscillation)
I've started learning openSCAD and Prusa Slicer. Anxious to download your stl files for the prop version.
telnar1236
Elite member
The STLs should be available once I've had a successful flight with the prop version. Hopefully next weekend, but with how badly I've been flying lately, maybe a couple weeks after thatA perfect example of PIO (pilot induced oscillation)
I've started learning openSCAD and Prusa Slicer. Anxious to download your stl files for the prop version.
telnar1236
Elite member
I'm using normal PLA. LW-PLA isn't strong enough to make the connection hardware that makes a modular design possible, so I would need to print some PLA parts and some LW-PLA parts if I went for an LW-PLA design which would almost double the part count. Normal PLA is also cheaper and easier to print with, so I've been trying to switch over to it wherever possible. I used LW-PLA for a few previous projects, and it's great for reducing weight with thicker walls, but a lot of that weight can just be designed out of a PLA design.
telnar1236
Elite member
The modular prop plane is mostly printed now. The only major fix I need to make is a reduction in the thickness of the PETG springs in the gear since they are currently far too stiff.
telnar1236
Elite member
Thanks! It has a 36" wingspan and will weigh in between 750 and 900 grams with between a 3s 1500 and a 4s 2200 battery. I designed it to use a locking mechanism only where necessary for modularity. Otherwise, the parts are connected using CA.
telnar1236
Elite member
Unfortunately, the weather this weekend did not cooperate. I woke up both days to thunderstorms and 20 mph winds. The good news is it gave the chance to experiment with different filaments on the F-104. I tested both ABS and carbon fiber PLA. Both worked. The PLA gave a very stiff print, as expected, but no real weight savings. On the other hand, the ABS produced extremely neat and strong prints, and was 17% lighter than the PLA versions I've printed.
telnar1236
Elite member
The modular prop plane flew today and had the first flight that didn't end in a crash. Unfortunately, I don't have video. There are two major problems with the design that need to be addressed before the STLs are ready to be released. The first is that the plane was extremely tail heavy. I added an extra section of fuselage to make the nose longer and still needed to add 3.5 oz of weight to the nose and fly with a 4s 2200 mAh pack all the way forward as opposed to the designed for 3s 1500 mAh pack. I didn't want to add the weight of a RunCam to the already heavy tail to try and get footage. The second is that while the gyroid infill structure in the wings is strong enough to support the aircraft in flight, even at well above its designed weight, it is not strong enough to support the point loads of the landing gear. The plane ended up needing to be hand launched due to the weak gear mounts, instead of performing an ROG takeoff. The good news is that the aerodynamics worked great. It was very stable and easy to fly, although the lack of rudder was pretty noticeable. Because it was underpowered due to the high weight, I decided to land after only a few circles around the field. Foolishly, I didn't remove the gear despite them being only partially attached, and one of them punched through the wing preventing any further flights.
To fix the issue with the plane being tail heavy, I plan to simply increase its size (make it longer and increase the wingspan). The tail is far stronger than it needs to be for the current size of the plane, so increasing the overall size will let me make it relatively lighter. I also plan to use a larger and heavier motor which will give me more power instead of just adding a ton of weight to the nose. To fix the problem with the weak gear mounts, I will fully model the inside of wing, and, while I'm at it, the tail. I was hopeful that simply using infill would be strong enough and make the task of modeling the plane simpler, and in some cases (gear not mounted in the wing), it would be, but overall it was a failed experiment for applications with the majority of designs. It might be possible to partially model the inside of the wing (e.g. model the structure needed to mount the gear and nothing else), but I can also cut down on weight by fully modeling the structure, so I'm giving up on that part of the idea.
To fix the issue with the plane being tail heavy, I plan to simply increase its size (make it longer and increase the wingspan). The tail is far stronger than it needs to be for the current size of the plane, so increasing the overall size will let me make it relatively lighter. I also plan to use a larger and heavier motor which will give me more power instead of just adding a ton of weight to the nose. To fix the problem with the weak gear mounts, I will fully model the inside of wing, and, while I'm at it, the tail. I was hopeful that simply using infill would be strong enough and make the task of modeling the plane simpler, and in some cases (gear not mounted in the wing), it would be, but overall it was a failed experiment for applications with the majority of designs. It might be possible to partially model the inside of the wing (e.g. model the structure needed to mount the gear and nothing else), but I can also cut down on weight by fully modeling the structure, so I'm giving up on that part of the idea.
telnar1236
Elite member
I've been a bit busy and unable to work on the modular plane as much as I would have liked these past couple of weeks, but the third version of the jet is coming together, this time in a peppermint flavor.
The different colors show the different sections of the plane that bolt together using the modular connection. This most recent version includes further improvements to reduce weight and increase structural rigidity in a couple of key locations. I'm hesitant to give a timeframe for when it will fly, but I hope to test it sometime soon.
The different colors show the different sections of the plane that bolt together using the modular connection. This most recent version includes further improvements to reduce weight and increase structural rigidity in a couple of key locations. I'm hesitant to give a timeframe for when it will fly, but I hope to test it sometime soon.
