I want to print a C-130 Hercules for remote control flight

[schafe 3]

I am currently building a C-130 out of balsa. Its slow and gluing balsa together seems way too 80's. I have been playing around with the idea of printing one. I know that there have been at least 3 planes that have been printed and at least 1 has flown. Search "First flight of 3D printed plane" on Youtube. I don't think anyone has tried a scale model. Someone needs to. I'm going to give it my best. The neat part is you can simply reprint badly damaged sections.

Here is where I am at so far:

- There are 3D files available on Turbosquid that can be downloaded as an .obj file. I have been using Blender to modify the exterior mesh of a sample plane for the required interior spaces. I plan on designing "tubes" into the fuselage and wings to slide carbon fiber shafts through for strength. Once the plane is redesigned I will divide it up into sections that will fit in a printer.

- It seems that ABS is the preferred material for its strength vs PLA.

- I'm just about ready to order a UM2. Just haven't been brave enough to hit "Buy" yet.

- I don't think stringers and strange bumps from the printing process will matter much as I can simply sand them out.

- I'm not going to print the control surfaces attached like in the video.

Here are the unanswered issues and questions I can currently think of:

- I will have to test PLA and ABS vs balsa, plywood and hardwood spars typical of RC plane construction. In the above video you notice that the interior of the fuse is empty compared to a built-up RC plane. I think I'll take advantage of "Print this for me" and see if someone will make me some samples. I understand that the plane in the video was printed with professional equipment.

- A heated print chamber is recommended for ABS prints. Do you actively heat the chamber or simply insulate it so the heat from the bed and print head doesn't escape?

Thoughts, ideas and advice welcome. Does this even seem possible?

Thanks

 

[dree 0]

Hi,

It seems very possible to me.

I recommend learning some 3D designing and building your own plane so you have more control over strength and weight of your plane parts.

Maybe you can kick off with some sketchup C-130 models http://sketchup.google.com/3dwarehouse/search?q=C-130&styp=m&scoring=t&btnG=Zoeken but I don’t think these models are made for real flight.

Don’t underestimate the strength of PLA, the only thing I worry about is the engine… does it get hot? In that case you’re right about using ABS.

Heated chamber is indeed not actively heating itself.

Dree

 

[schafe 3]

Thanks for the link. I'll experiment with the file later. I like free. Free is good.

I had not considered the operating temp of the electric system. Good point. The batteries get warm, but not enough that you can't hold them. The speed controls get warm, but they are usually placed in a recession on the outside of the fuse or wing for maximum air flow. The motors also get warm, but typically there is ventilation designed into the engine nacelles.

 

[jonnybischof 60]

In terms of weight, a printed PLA part is pretty similar to balsa wood if you print the parts hollow with some 25% infill. I guess this won't be a problem anyways as you can also build (bigger) RC planes from stronger and heavier plywood and still get them to fly.

You can use PLA, it won't decompose while your plane is in the air

ABS might be stronger (haven't printed ABS myself yet) but the UM2 can do both materials anyways.

Just don't forget that your maximum print size for a single part is about 20x20x20 cm.

 

[woofy 12]

Hopefully, you mean 25.5x25.5x20.5 cm for the UM2.

 

[gr5 2,265]

Not double checking but from memory I'm pretty sure it's 235X225X205mm.

I would go ABS. PLA even just in hot water out of the sink faucet will soften much too much. ABS is much more frustrating but I would recommend you just start right out with ABS and not even bother learning about PLA. The heated chamber is only if the shrinking is a problem. For example hollow fuselage parts printed edge up and down will probably print fine without heat. But some parts will be a problem.

Or consider buying PLA90 (google it).

 

[schafe 3]

The web site says 230X225X205. Which of these would be the height?

My plan is to print the fuse and wings with a 1mm skin to start. I will lower that until the exterior starts to feel flimsy. I will use the carbon fiber shafts for strength. This should eliminate the balsa ribs, plywood bulkheads, and hard wood stringers hopefully making the entire plane lighter. I will have to bulk up the area where the wing, rudder and elevator attaches.

I sure would like to order a UM2, but there seems to be a lot of confusion right now. I guess I'll wait until the reviews start showing up from people who get the first units.

I'm working with the 3D file mentioned above. Looks good so far.

 

[umuser 0]

I've built and flown traditional RC planes and heli's. It would be interesting to see your parts. Balsa is very rigid for it's weight, and resistant to elongation, and that's why it's still used. The main thing I'd worry about with printing is elongation from temperature. If ABS or PLA spars, leading edges, or trailing edges change much in length due to temp or humidity, it would warp your wing.

 

[dree 0]

This is the way the infill is printed, pretty srtrong I think, this is 20% infill.

Groet Dree

 

[woofy 12]

Not double checking but from memory I'm pretty sure it's 235X225X205mm.

 

The UM2 page still says 230 x 225 x 205 mm.

But go to News->Blog.

There Sander mentions, then confirms in a subsequent post, the larger build size:

' X25,5 * Y25,5 * Z20,5 cm' this is the correct build size.

 

[gr5 2,265]

My plan is to print the fuse and wings with a 1mm skin to start.

 

The UM1 and UM2 have a .4mm diameter hole in the nozzle. This means it prefers to print everything (in X,Y directions) with a multiple of .4mm. So skin ("shell thickness" in cura) should be set to either .8mm or 1.2mm. If you set to 1mm it will make 2 passes. One or both of which will overextrude. It will probably work okay but .8mm is probably better.

You can print for example a section of fuselage by printing a vertical cylinder (conical cylinder) and telling Cura to not print top or bottom solid infill in the expert settings. But you will probably have much more complicated shapes with holes and such. Which means you will have to model the thickness of the walls of the fuselage. This will have to be done in CAD and will be annoying and maybe difficult but most CAD software can do this.

 

When you do that you must have walls at least .8mm thick or Cura will simply not print them. You can print thicker than .8mm to any width e.g. 1mm and if you select solid infill, Cura will underextrude the exact right amount to fill in the last .2mm properly although this is kind of buggy - or at least it won't be the ideal pattern of movement. But it will work. Whereas .8mm or 1.2mm walls will work as desired (2 or 3 passes of the nozzle).

 

[Daid 306]

I've build and flown glider-model-planes, made from balsa.

And, I know my way around the 3D printer.

Now, most things can fly with a big enough engine attached to it. So depending on what you want exactly weight might not even be a big issue. Strength however is. Most planes have strong inlays to keep the wings from flexing. 3D printing would allow for stronger internal parts where you want, and even lighter stronger planes in theory.

It would be awesome to see a project like this, and I would help in the best ways possible. It has been on my mind like forever, but the people that I know that build model-planes do not know their way around computers enough to get this done. And I... I'm always short on time...

 

[schafe 3]

Thanks for the advice everyone. I'm still plugging away at this. I'm 100% new to 3D CAD and printing but I have a pretty good technical background and have been flying RC for 25 years or so. I'm having good luck with Blender as my CAD program. It seems that I had better design the plane in mm. I had been working in inches. Once I get a section finished I will post it somewhere for analysis. That brings up the point:

Where is a good place to post a .stl file for people to access?

 

[gr5 2,265]

Where is a good place to post a .stl file for people to access?

 

Youmagine.com

 

[dree 0]

Where is a good place to post a .stl file for people to access?

You can use Youmagine or Dropbox, put your file in the public map and put a link here.

Dtree

 

[Daid 306]

Youmagine is being funded by Ultimaker right now. We hope to make big projects like the C-130 better maintainable then thingiverse, but it's not really as big as thingiverse right now. While thingiverse is run by Makerbot and flooded with crap.

The UM2 has a build envelope of 230x225x205 (the 255x255 from Sander is wrong. No idea where he got that from)

We also have experiment UM+ printers here at UM, which have a build height of 550 instead of the normal 200. Would be perfect for wings. But no idea if there ever will be plans to produce it.

 

[woofy 12]

The UM2 has a build envelope of 230x225x205 (the 255x255 from Sander is wrong. No idea where he got that from)

 

Ah well: It be what it be.

 

[SandervG 1,521]

yes yes yes.. i made a mistake ! :oops:

I already changed it in the news blog, and i will also post it here:

The correct size is: x23 y22.5 z20.5

I got that other information from someone, don't remember who.

Just an unfortunate miscommunication I think. Sorry for any inconvenience caused.

 

[LePaul 354]

Youmagine is being funded by Ultimaker right now. We hope to make big projects like the C-130 better maintainable then thingiverse, but it's not really as big as thingiverse right now. While thingiverse is run by Makerbot and flooded with crap.

 

I agree completely, Thingiverse was very appealing for me at first but the sheer amount of very old files and , well, landfill, is pretty obvious.

I'm also finding GrabCAD.com is a neat place to find drawings for projects I would love to print someday. To that end, the drawing to done and all I have to do is tweak and figure out how to print

 

[schafe 3]

Would it be a good idea for me to download Cura and experiment with the settings there? I'm just not challenged enough trying to learn Blender. . .

 

[gr5 2,265]

Cura is free and easy to use to see how things will slice. YOu can use the xray view to see problems (for cura) in cad files and you can slice it and look at layer view to see if it is printing solid where you expected and openings where you expected.

However Cura is only a slicer. You can't change the model (other than scale and rotation).

For that you need CAD. I love google sketchup which is free and find it easy to use but it just doesn't do all kinds of things you might need to do with a complicated shape like a wing. It's fine for editing existing CAD models (like a wing). And for breaking a model up into multiple pieces and making simple changes.

But you might want to learn blender instead. All of these CAD programs have tons of free youtube tutorials or web page tutorials and such. All of them are quite a bit of time before you are really good at them.

OpenSCAD is another free cad program which is TOTALLY different. There is no mouse clicking - you don't draw anything - you type in basic shapes like cube(10,20,30) which is a 10 by 20 by 30 mm cube. And then transform and twist and add and subtract shapes and you can use for loops to make gears and logic and parameters to make generic gear builders and there are libraries - tons of libraries to make these things easier (such as making 3d text on your model).

And there are many more! Freescad is another one. I love this one but it crashes on me soooo much. And seems buggy sometimes (sometimes it refuses to extrude or revolve and can be frustrating).

Solidworks seems to be the standard but it is NOT free. I've been meaning to try tinkerCAD as I heard good things about that one.

You should ask around, talk to mechanical engineers maybe. Play with a few free cad's or do their trial version.

 

[jonnybischof 60]

Imho, Google Sketchup (I decided to stay with V.8 instead of the new 2013) is perfect for a beginner to get to know his printer. I've started making my own models last week. By now I've created a lot of stuff without much trouble and with great results. However, if you use Blender, you can probably ignore Sketchup...

I definetly recommend using Cura for creating the G-Code. The X-Ray view is a perfect tool to get good results: If there's anything red in X-Ray view, it needs to be fixed. For example, if you put two cubes side to side to form a single block, then you have to get rid of the surfaces which are between the two cubes -> surfaces inside of a solid object are going to get the slicer into trouble...

If you need a real CAD tool instead of just a 3D-drawing program, you should take a look at PTC Creo Elements / Direct Modeling.

It's free to use, and it's based on Wildfire / ProE which is one of the industry leaders for professional CAD tools.

Personally, I hate it because it's just NOT intuitive. I don't ever want to touch it again, but as soon as I have to create some serious models, I'll probably have to...

 

[schafe 3]

Here is what would be the 2nd section back from the nose.

https://www.youmagine.com/designs/c-130-fuse-section

I have not quite figured out an easy way to join the tubes to the fuse exterior. I would prefer more of a solid fillet between the fuse and the tubes. However, Cura seems to think everything is ok. (I just realized that the tubes start above the build table - oops.) Blender identifies several problems for 3D printing but once again Cura seems happy. For future design attempts I will likely print the tubes in 2 or 3 sections to save weight. I'll have to taper the bottom of the tubes to allow for proper printing. I also realized that the offset fuse section that would slide into its mating section is not necessary. I could eliminate that and use carbon fiber material or something similar to glue the sections together and fill the seam with balsa filler.

I don't understand how some of the print settings in Cura affect print time. Why does changing shell thinkness result in a print time change if this is a solid object? Same goes for bottom/top thickness and fill density.

 

[gr5 2,265]

I tried to slice this with cura at 1X scale (wow - this is going to be a big plane!!).

I had to remove the skirt in expert settings because it was outside the build volume.

There were some problem spots on the bottom due to the cylinder not touching the build plate in a few places. There are two arcs that don't touch the build plate. Do you need those? If so you can just flip it over.

Also you have these cylinders which start about 1cm in from either end. Those are a problem as they are floating in the air. For those you might have to add "support".

Also the 4 small cylinders seem to have a wall between them and the main cylinder. This is where you should be adding the fillet. To make a good strong connection. Right now it barely touches and those 4 cylinders could probably just pull right off the main cylinder.

Looks like a good start though! This looks like a huge project.

 

[gr5 2,265]

Okay re-reading - I see you already came to the same conclusions.

 

and fill the seam with balsa filler.

 

Well - wouldn't it be better like this design though? Where one section fits inside the next? Easier to get perfect seams. If they mate flush then it's harder to get it aligned perfectly in X and Y.

 

Why does changing shell thinkness result in a print time change if this is a solid object?

 

If you want solid infill, it's probably faster to just set shell to 20mm (solid) which will give you basically infill parallel to the walls of your cylinder. If you do normal infill it does diagonal to the Ultimaker bed which takes longer to print.

The print times are always a low estimate. Pretty close for an object like this but not for small objects. This is because it doesn't include the acceleration - it assumes full speed for every line segment but in reality, Marlin decelerates and accelerates the print head at every junction of two line segments. But your object will print close to "at speed" because it is large and doesn't have thousands of tiny corners to zig zag around.