3D Printed 64mm MiG 21

[telnar1236]

I've had a project on-again/off-again for decades mainly because of career. Now retired, I need a large workshop which I hope to finish this spring. Which has given me the itch again. I was just using JavaFoil, but the author clearly says the 2D, closed-form methods he uses drastically fail once in stall/turbulent conditions. These two tools you mention... do they have their own CAD or do they import from something like DFX? Although lift/drag are important, I'm looking for after stall conditions and interference problems. I'd also like to be able to export the pressures being applied to the model, so I can suck them back into my FEM to do the structural analysis. Do these analysis programs give pressure data as a function of time? For instance vortex shedding. I think the dynamic fluctuations that might excite natural frequencies in the structure would be far more critical than the ultimate static loads that JavaFoil gives me. aka...

Tacoma Narrows Bridge collapses - History Channel

What is the project? Simflow is what you would want for what you're describing. You can import STL and other 3d files directly into the workspace. You can do all of what you're describing in terms of finding pressure and interference issues. I haven't tried to export the pressures, but I would be shocked if you couldn't. Transient analysis is also possible which will give you the values over time. The big concern would be that the free version has a 200,000 node limit on mesh size and the built in mesher uses hex cells which aren't great for rotational flows. I have had trouble getting the coarse mesh to converge for highly rotational and turbulent flows like you would see with vortex shedding. You might look into hand calculations to check if turbulence will cause resonance. The frequency primarily depends on flow velocity and the width of the object so you should be able to calculate if it's at a risky frequency at the least.

Kármán vortex street - Wikipedia

XFLR5 stops converging in stall conditions and is much more airplane wing oriented.

 

[Inq]

Sorry... getting a battery of questions as I start downloading these programs. The first thing that hit me on the xflr5 page was, "XFLR5 is an analysis tool for airfoils, wings and planes operating at low Reynolds Numbers." Do you know if it is invalid for higher Reynolds Numbers? I'm looking at Reynold's Number of 200K to 2.5M with the normal operating range around 1.5M.

 

[telnar1236]

I recall using some of LLNL (Lawrence Livermore National Laboratory) programs years ago. If you have a need to go over the 200,000 nodes, have you ever looked into either:

LLNL or NASA - https://software.nasa.gov/

I had not looked into that. Some of those programs look extremely useful.

 

[telnar1236]

Sorry... getting a battery of questions as I start downloading these programs. The first thing that hit me on the xflr5 page was, "XFLR5 is an analysis tool for airfoils, wings and planes operating at low Reynolds Numbers." Do you know if it is invalid for higher Reynolds Numbers? I'm looking at Reynold's Number of 200K to 2.5M with the normal operating range around 1.5M.

I'm not really sure. Re = 200,000 is definitely fine. I've used it for analysis at Re = 10,000,000 but I haven't ever compared the results to experimental data. Generally, I would expect it to underpredict drag and lift at high Reynolds numbers. Really, XFLR5 is for preliminary analysis and sizing. The lift numbers should be pretty good since inviscid approximations are pretty good for high aspect ratio wings, but the drag numbers are pretty questionable even if viscous analysis is being used.

 

[Inq]

I always used Nastran for FEA professionally. It is required for all government, DOD, DOE, etc. I found the likes of Dyna3D and Calculix for my own goofing off, but it was brutal to use, requiring Linux and no real GUI interface except the old kind of type in rotations, et. Brutal to use. I finally found Mecway when still in Beta and got a permanent license for cheap. Don't know how friendly the Nasa/LLNL based CFD programs are, but normally, the government stuff is not!

 

[telnar1236]

I always used Nastran for FEA professionally. It is required for all government, DOD, DOE, etc. I found the likes of Dyna3D and Calculix for my own goofing off, but it was brutal to use, requiring Linux and no real GUI interface except the old kind of type in rotations, et. Brutal to use. I finally found Mecway when still in Beta and got a permanent license for cheap. Don't know how friendly the Nasa/LLNL based CFD programs are, but normally, the government stuff is not!

I have some experience running stuff from the command line. A large part of XFLR5 is actually just a nice GUI for an older tool called XFOIL which is almost completely operated without even having a graphical display (there is one for results and one to show the effects of geometry changes, but it's still all controlled by the command line), and another tool I still use pretty frequently is AVL which was written by the same team and has similar levels of graphics. Simflow mostly meets my needs for hobby grade CFD as it is, but it's always neat to play around with new tools.

 

[Inq]

@telnar1236,
Am I remembering correctly that you use standard PLA on your planes?
As you might remember... I strictly use ABS. My printer is even modified to really only use ABS for structural (not cosmetic) prints. I HAD to buy two rolls of PLA to do some things for my daughter's wedding. Cosmetic things that required bling... thus I found something called "silk" PLA. It's on the way.

I had some old PLA from when I first got the printer and tried to use it. Maybe its old... but I get nothing but strings and no sticking to the bed and really just a big CF. (I hope that is PG-13 enough for the forum. Surely every 13 year old has seen "Heartbreak Ridge"). I don't like PLA. Anyway... I'm expecting to have lots of this left over after the wedding. Is it usable for anything useful? 😂

 

[telnar1236]

@telnar1236,
Am I remembering correctly that you use standard PLA on your planes?
As you might remember... I strictly use ABS. My printer is even modified to really only use ABS for structural (not cosmetic) prints. I HAD to buy two rolls of PLA to do some things for my daughter's wedding. Cosmetic things that required bling... thus I found something called "silk" PLA. It's on the way.

I had some old PLA from when I first got the printer and tried to use it. Maybe its old... but I get nothing but strings and no sticking to the bed and really just a big CF. (I hope that is PG-13 enough for the forum. Surely every 13 year old has seen "Heartbreak Ridge"). I don't like PLA. Anyway... I'm expecting to have lots of this left over after the wedding. Is it usable for anything useful? 😂

I have mostly used LW-PLA in the past, but I am trying to switch over to primarily designing for normal PLA. As you can see with this project, I have had some trouble keeping the weight down using LW PLA techniques, but I'm getting better at it. I think I've even commented before that I've had exactly the opposite experience. PLA works great for me, while ABS won't stick to the bed or warps so badly it's unusable. It sounds like the old PLA might have gotten too much water in it. If PLA is left to sit, it absorbs water and starts to behave like that. I live somewhere humid, and it's never happened to me, but someone I know, who lives pretty close to where I am, has that problem constantly, so who knows. Have you tried ASA? I've heard some good things about it as an ABS substitute but haven't seen much done with it.

Normal PLA is great for rc planes. It has better layer adhesion than ABS and is a bit stronger, but at the cost of being a lot denser. Silk PLA is more brittle and weaker than normal PLA. I've seen some posts where people have successfully printed a plane with it, but it's not something I've tried.

 

[telnar1236]

The partially assembled fuselage for the MiG 21 fell off a table and broke pretty badly while I was working on it. I'm temporarily putting the design of a 3D printed MiG 21 on hold. I'm not giving up completely however, so it will eventually be redesigned to work with my modular airplane system which should both resolve all the TWR and inlet geometry issues and make repairs easier. https://forum.flitetest.com/index.php?threads/a-modular-plane-design-my-next-big-project.74036/

 

[GH05T]

Not to be a bother but could you happen to release the stl files, there's not a single other person who's released any stl files for a mig 21 anywhere on the internet and I want to start a project for one on my own but would like a place to start

 

[telnar1236]

Not to be a bother but could you happen to release the stl files, there's not a single other person who's released any stl files for a mig 21 anywhere on the internet and I want to start a project for one on my own but would like a place to start

Here they are. Had to split them into a couple of zip files because of limits on upload size. Honestly, you would probably be better off starting off from scratch if you're doing a new design. I couldn't get it to fly, but it still might be capable of ROG takeoff with a decent power system (I never really was able to try that properly). But in general, the MiG 21 actually has pretty simple geometry and can be modeled somewhat quickly in Fusion 360, so, for anyone looking at this, a design from scratch is the easiest way to go. Also, as mentioned in the post where I abandoned this project, my now mostly functional modular plane design will eventually have a MiG 21 variant, so if you don't mind waiting a month or a few, that is coming.

 

[GH05T]

I kind of want to try using the ground effect holes that the real mig21 uses for cooling, instead I want to add them in a certain angle to make ground effect inlets, it reduces the pressure under the plane though so it may not help with lift but the edf will get air

 

[telnar1236]

I kind of want to try using the ground effect holes that the real mig21 uses for cooling, instead I want to add them in a certain angle to make ground effect inlets, it reduces the pressure under the plane though so it may not help with lift but the edf will get air

A lot of edfs use inlets on the bottom so shouldn't be a problem