3D Printed WWII B-17G

[DeepEndofPool]

Image reflects current state of the design.

Length: 1179 mm (~46 in)
Wingspan: 1633 mm (~64 in)
Weight goal: <2 kg
Designed in: Fusion 360
Sliced in: Cura 3.4.1
Printed on: undecided but probably Creality Ender 3
Motors: not yet decided
Servos: ?9Gs?

Design origin: This is a translation of a balsa/plywood model plan found on the web. There is no name or info block on the plan so I don't know the real origins or how accurate it is. Comparing to historical photos it doesn't look too bad.

Lots left to do yet and lots of little issues with what's been done. The main thing missing is all the internal structure. My design process is iterative. I cobble something together, then add refinements as I go. I have attempts at gun turrets and other details, but I'm not happy with them yet. I was blocked on how to use Fusion 360 to create the support structures, but LocalFiend, here on this form, provided a great explanation of how to do it. Thanks!

My interest is in combining 3D printing and RC flying, but there is a foam board B-17 plan on this forum. Check out https://forum.flitetest.com/index.php?threads/my-first-swing-at-designing-a-plane-b-17.39601/ and https://forum.flitetest.com/index.php?threads/b-17-build-log.39765/

This is my first plane design as well. Also first time using Fusion 360 or 3D printing. Pretty steep hill to climb. I'll be going slow.

Thanks for reading.
Chip

 

[b-29er]

I would strongly advise you get in touch with @localfiend, since he's done successful 3d printed aircraft and the structural infills required. 3d printed aircraft can very easily get overweight, and ABS is not the easiest material to print with...

 

[Fluburtur]

I just made a 3d printed plane but it has foam wings, also no internal structure because I was too lazy to add it.
It works very well but im on the third nose so far. It took me forever to design some things such as the canopy and weird lofted parts like the wings and tail.

 

[localfiend]

Cool man, looking pretty good so far. Got a few orders of operations things that could help you out.

Before you even touch making the internal structure, make sure your timeline is clean (no errors), and make sure that you can make all the tools you'll need to cut the spars. So offsets, shells, whichever method you chose, make sure that you can in fact make them all before drawing spars. That may seem like extra work, but it'l save you a lot in the end.

Part of all of that is making sure you have good clean joints between things like wings and the fuselage, nacelles etc... If you have ugly joints or excessive fillets, making the tools you need will be almost impossible.

Also, get the control surface hinges all worked out before hand, as well as applicable battery hatches, landing gear setup, servo slots and wire runs, servo horns, firewall and motor mounting.

Finally, export chunks of your model and slice them to get an idea of the total weight you're looking at. Expect to add about 35-40% more weight when all the internal structure is in place (maybe more if you need to go tongue and groove on wing parts for example). Add that all up with the expected electronics weight, and do the math for wing loading. It really is critical for 3D printed planes as you have very little wiggle room. They're a lot heavier than foamboard, which will really increase tail weight. You're going to need a lot of ballast in the form of battery weight to balance a 3D printed plane, so keep that excess in mind.

Once you get all of the above done, I can go over what making the internal structure is going to involve. For something this size, you're really going to want to split your cad file up into multiple pieces before adding spars. The internal structure is incredibly CPU intensive, so breaking stuff up when it's ready for internals will save you from massive amounts of hair pulling and headaches.

I've also got quite a few work arounds for making spar cutting tools. So if you get to that point, and are having trouble creating offsets or shells, let me know.

 

[DeepEndofPool]

make sure your timeline is clean (no errors),

It's clean. I had to start over four times to get it that way, but it's currently clean. That said, it is "lumpy". Since I'm learning 360 as I go, parts are made with different tools with a mish-mash of multi-bodies, components and imports from separate files. It all hangs together so far.

making sure you have good clean joints

I don't know what you mean by clean joints. Do you mean things like gaps and surfaces that don't match?

get the control surface hinges all worked out before hand, as well as...

Yeah, you're right. I've got a few problems there. Hinges for ailerons and flaps are there but I made a mistake and they bind up (interfere with the wing). That's one I know I have to fix. I did cutouts for 9G servos on a previous iteration, but haven't got back that far yet. Do you have any idea what the throw on control surfaces should be? The plan I started from has no recommendation so I assumed +-45 degrees for ailerons and more for flaps. I think I read on one of the FT Simple planes that their throw is more like +-15 degrees but control surfaces look to be proportionally bigger. I'm still noodling on motor mounts and gear mechanics.

export chunks of your model and slice them to get an idea of the total weight

I've done a few pieces, but encouraging me to do more is the right thing. Wing loading calc is something I will have to learn. It's got to be common so I'll google it, but if you know of a good "how to" reference, I'd like to have it.

Thanks for the really solid advice. I'll do my best to follow it. I appreciate the help!
Chip

 

[wilmracer]

OOOOHHHH... will be watching this one closely!

 

[DeepEndofPool]

Data in this post is pretty soft, but I got to thinking "How heavy is this thing going to be?" So I did a quick split on the parts to fit print volume and let Cura (slicer) estimate the amount of filament to be used for each. There is no internal structure yet so I estimated 50% increase for that. Lots of other SWAGs regarding electronics as well. The data is below, but the bottom line is that all up including electronics and battery will be somewhere in the neighborhood of 2600 grams or about five and three quarters pounds. This compares to the 3DLabs P-38 as follows.

	P-38	B-17
Wingspan	1410	1633 mm
Weight	1980	2600 g
Power	700	1400 W
Wing Load	65.9	~78 g/dm2

It's a bomber not a fighter, but will it get off the ground? We'll see how these numbers hold up over time.

----------------------------

Cura settings
CR-10 build volume (Ender 3 build volume will require more pieces)
No top or bottom layers
0.4 mm nozzle
0.25 mm layer height
wall line count = 1 (0.4 mm wall thickness)
no supports
infill 0%

WING including Nacelles
Aileron = 12 g
Flap = 19 g
Wing1 = 53 g
Wing2 = 47 g
Wing3 = 37 g
Wing4 = 36 g
Wing5 = 50 g
Wingtip = 5 g
__________________
Total = 259 g for one wing.

FUSELAGE including Vertical Stabilizer & Rudder
Nose = 4 g
Fuselage1 = 39 g
Fuselage2 = 42 g
Fuselage3 = 46 g
Fuselage4 = 51 g
Fuselage5 = 63 g
Fuselage6 = 12 g
Rudder = 13 g
___________________
Total = 270 g

HORIZONTAL Stabilizer including Elevator, one side
Elevator = 19 g
Stabilizer = 23 g
StabTip = 2 g
___________________
Total = 44 g for one side

Total printed parts
2x Wing = 518 g
2x H-Stab = 88 g
Fuselage = 270 g
___________________
Total = 876 g

Add 50% for internals = 1314 g

Most of the following from Hobby King website
Motors: Something like Quanum MT Series 4108 370KV 355W @ 108 g = 432 g
ESC options: 4x 20A @ 30 g = 120 g
2x 40A @ 43 g = 86 g <-- looks like the right trade off
1x 80A @ 95 g = 95 g
Receiver: Turnigy iA10B Receiver 10CH 2.4G = 20 g
Battery: 6S 5200mAh = 643 g
Servos: 7x HXT900 @ 9.8 g = 69 g

Total electronics = 1250 g

Grand Total = 1314 + 1250 = 2564 g

Whew!
--------------------------------------------

 

[DeepEndofPool]

No, I haven't given up. I'm slow.

Here is the result of a complete redesign of the wing. Same size. Same shape (mostly). But this one actually works (at least in Fusion simulation). The ailerons and flaps don't visibly bind against the wing. The servo pockets are there. The servo horn linkage has been planned. Pipes for wire to the servos and motors have been added.


Lot's more to do yet, but I thought I'd share this bit of progress. Is it right yet? I don't know. I know I've learned a lot, but won't really know if it works until it is printed.

I still haven't done the internal struts. I'm taking localfiend's advice and making sure everything else is in place first.

Still to be done on the wing:
- firewalls
- motor mounts
- terminating motor wire pipes to the right place
- trim flap and aileron ends for clearance (how much?)
- redesign the wingtip so it matches the new wing
- fillets where needed
- struts & motor bracing
- some solution for landing gear. thinking about fixed gear in the first version

Of course, what I did to the wing needs to be done to the stabilizer and rudder as well. ...

 

[wilmracer]

Nice! I don't know anything about designing planes for 3D printing but I'd strongly suggest doing what you're doing and following Localfiend's lead. Fixed gear for the prototype is probably a good idea. The tailwheel is dead simple but the mains may be a bit tricky to do with off the shelf units, particularly if you want to keep a scale-ish look. The struts actually angle forward and have linkages that run to the front of the bay that support the strut and retract the unit. I don't have the total rotation angle handy, but it is less than 90* so finding a unit that works would be a challenge.

Keep it up!

 

[DeepEndofPool]

Thanks for the encouragement. Yeah, that main gear is tricky. It's going to be interesting to find (or build) something that looks right and will support the weight.

And the plan doesn't provide much in the way of details.

Does anyone here know if the tail wheel is steerable on a full-sized B-17?

 

[varg]

Neat. I'm in the process of designing my first 3d printed RC plane now but it's not scale, so it's an exercise in nailing down design basics. This is an ambitious project. Every additional part, like the 4 engine nacelles, adds to the weight and complexity. Despite being decently experienced with 3d modeling I can't offer any input on fusion since I use solidworks.

	P-38	B-17
Wingspan	1410	1633 mm
Weight	1980	2600 g
Power	700	1400 W
Wing Load	65.9	~78 g/dm2

It's a bomber not a fighter, but will it get off the ground? We'll see how these numbers hold up over time.

I'm hoping it turns out lighter for you to get that wing loading down. And hopefully the proportions of the plans you're working from aren't 100% scale and have been altered for better flying characteristics as a small model.

Does anyone here know if the tail wheel is steerable on a full-sized B-17?

It freely casters and is lockable, but you'll want yours to steer.

 

[DeepEndofPool]

I'm hoping it turns out lighter as well. I made fairly conservative estimates, so I can hope.

I don't even really know if the model plan I started from flew all that well. And I'm experiencing the complexity you mentioned. I'm learning just how truly naïve I am. So far, it's still fun. I'm slogging through trying to balance making it printable, have at least minimally decent flight characteristics, fidelity to the original plane, and fidelity to the plan, all at the same time.

Even just staying true to the plan is impossible. It's a pencil drawn plan. I've found mistakes where the formers don't match the 3-view. Which is right? And the center lines down the 3-views either aren't in the center or in some cases aren't straight (side view "sags" in the middle). On such a plan the width of a pencil line was plenty accurate when cutting wood (sand & fill), but when the printer will lay down a fraction of a millimeter of material that pencil line isn't really accurate enough.

So it's all a compromise and I'm "designing" more than I thought I would. When I find discrepancies in the plan or something like ailerons don't work, I go back to historical photos to see what was, then I fiddle with Fusion to see what I can implement. Hopefully, that will lead me to a flying plane.

Thanks for the info on the tail wheel. I agree that steering is one where functionality trumps accuracy.

 

[wilmracer]

Does anyone here know if the tail wheel is steerable on a full-sized B-17?

Varg is right on the tailwheel. You definitely want yours to steer, and you might consider differential thrust in the outboard motors to help with low speed steering. Our models don't have brakes on the mains to help with steering so do whatever you have to to get the steering you need.

Attached are some of the reference shots I have for the B-17 gear. Good luck with your build!

 

[localfiend]

What airfoil are you using? Looks symmetrical, if it is, you might want to consider something with more lift for this type of plane.

I'd probably stick in some washout at the tips well.

 

[foamtest]

Does anyone here know if the tail wheel is steerable on a full-sized B-17?

I would make the tail wheel swivel and just use differential thrust to save weight and unnecessary complexity with the first model.

 

[DeepEndofPool]

Thanks for the feedback.

The airfoil is lifting at the root and

symmetrical at the tip. When I redesigned the wing, I eliminated all but these two formers. I will need to go back and check the intermediate stations to see how close they are to the plan formers and how quickly the curve transitions.

The plan does not explicitly call out washout, but lining up the spar cutouts in the formers there appears to be a bit. Something else to check.

 

[DeepEndofPool]

W-15 (start of ailerons) former matches up pretty well in terms of airfoil shape, but I was wrong about the washout. Same AoA at tip as at root. How many degrees would you suggest?

 

[DeepEndofPool]

make the tail wheel swivel and just use differential thrust

I like saving weight and fixed gear in the first prototype is very likely (but final decision is a ways off yet). I'm not sure about the differential thrust (mostly from my lack of skill standpoint). How would controls work on that?

 

[wilmracer]

The 17 is supposedly a pretty gentle flyer, particularly if you are going to do counter rotating props. 1-2 degrees of washout should be plenty (I would think)

For the thrust steering the setup would depend on your radio, but basically you would have the outboard motors slaved to the rudder channel (perhaps on a single mix that can be turned on/off). When you give right rudder the left outboard motor speeds up and the right outboard slows down, and vice versa. Your rudder should be perfectly adequate at speed, but for low speed ground taxiing the differential thrust will greatly cut down on your turn radius. Think about some of the FT Sea Ducks you may have seen and how they can basically spin like a top in one spot. You don't need your 17 to do that, but you get the idea.

 

[DeepEndofPool]

Thrust Steering: Okay, that sounds like more of a setup configuration (something I can do) versus a flying skill (something I would have to learn). Cool.