Continuing the discussion from [Roman] Dodecahedrons:
When I wrote in the above mentioned thread that I was thinking of making Roman type dodecahedrons with the lost-wax casting method the first two answers mentioned 3D printing:
I think this is a great idea (and thanks to both for the suggestion), but I have no idea where to start. Does anyone have an idea what I need or wants to share experiences?
I have zero experience of 3D printing, so I can’t help with any of the detail, except to say that ready-made files for printing Roman Dodecahedra exist on thingiverse, so at least you shouldn’t have to model it from scratch.
But I would need a program, I guess, to handle those files from thingiverse, copy them to a USB stick and carry them to a 3D printing shop? Or should I buy my own 3D printer? Is thingiverse a metaphor or does a site called thingiverse exist?
You see when I write absolute beginner I am not exagerating, but I have been musing about 3D printing for some time and it seems dodecahedrons are a good start: symmetric, not too big, not too small, not too simple, not overly complex.
So I have a computer: what next?
Thingiverse is a website where people upload and share models they have created so that other people can just download and print them (‘just’ being an oversimplification); the alternative to using a ready made model is to design it yourself from scratch using a CAD program, which has its very own learning curve - As I say, I can’t really help you with any of the detail, but just wanted to let you know that the learning curve for CAD is something you will be able to choose to skip, either at first, or for as long as you like.
If you have a local 3d printing shop (or perhaps your library has 3d printing), I would suggest starting with them first to find out what you need.
And yes, https://www.thingiverse.com/ is a real site. The files you download are called “stl” which is a 3d model. What you do with that model really depends on the 3d printer and the software for that printer. But basically you load that file into the software, edit various printing settings, and then you’re off to the races.
To print a couple of models, you do not need to invest in a (cheap or expensive) 3-D printer: find a nearby shop/place that has one.
It is possible to download files from a site like Thingiverse, but to tweak them you may want to perform some basic operations using CAD software. FreeCAD has its flaws but it is free software, or there are products like Autodesk Fusion 360 and Solidworks.
Thanks to all! That is a starting point. 
Personally, I do all of my designing in TinkerCad. It’s a simple tool, and I probably should use something more complicated, except that I’ve already figured out how to do all of the complicated things in it. I, personally, don’t get anything off of Thingiverse, because to me, the designing is the fun part.
Then, once I have the design, I take it to the public library, and get it printed there. This process basically consists of mailing them the STL file, telling them my library card number, telling them what color I want, and then about a week later (there’s a longish queue), they email me back and tell me I can come pick it up. And pay for it; it’s not free, but it’s very cheap. Very few of the things I’ve made have been over a dollar.
Or at least, that was my process. Nowadays, I’m full-time employed at a school, so the next thing I need to print will presumably be at the school’s makerspace. Which will mean that I’ll have to set the settings on the machine myself, but that’s something I should get more practice at, anyway.
3D printers (the models I’m familar with anyway) don’t accept STL files directly. After you have your design in an STL file, you have to run your STL file through a program called a “slicer” to generate a “g-code” file. With most printers, you then copy the g-code file onto a USB stick, plug the stick into the printer, and run the print.
I’ve done prints by downloading STL files from thingiverse and printing them unchanged, by downloading and tweaking them in a CAD program (I’ve used SketchUp and Fusion 360), or by designing from scratch in a CAD program. There is indeed a steep learning curve for the latter approach. Fusion 360 has more functionality than SketchUp but is much harder to use. With any design, creating proper “supports” can be tricky. You can’t print arbitrary shapes because shapes that are disconnected in early layers but will be connected in later layers will fall apart before you get them connected. So you have to add temporary scaffolding that holds the parts in place during the print, and them remove the scaffolding manually after the print is done, with knives and sandpaper and such.
I’d recommend doing something like using a public library printer initially as Chronos suggests, until you get the hang of things and decide that you’re going to print enough to make it worth buying your own printer. I’ve owned two printers, using two very different technologies. My first printer was from LulzBot. It is a traditional filament printer: it sucks in a spool of plastic filament, runs it to a “print head” that melts it and moves around to deposit the melted filament where needed. My second printer is a Elegoo Mars. It uses photosensitive liquid resin rather than filament, and prints a whole layer at a time. It is of course faster to print a layer at a time rather than a voxel at a time. It also has better resolution. I think my LulzBot had a resolution of about .3 mm while the Elegoo has a resolution of .05 mm. Generally you can’t even detect the layers on a print from the Elegoo, while the layers were quite noticable on the LulzBot prints. I’ve also had more prints fail with the filament printer than the liquid resin printer; there’s just more to go wrong with all the mechanical movement of the head. However the LulzBot could print larger prints; something like 12"x12"x12". The Mars is limited to about 120 mm x 70 mm x 155 mm. (You can sometimes print larger things in separate pieces and glue them together.) Speed is an issue with any type of printer. A print can take anywhere from a couple of hours to a day or more depending on the size. (Running a 24 hour print and having it fail in the last hour is frustrating, to say the least.)
Just to not overcomplicate things, of you are going to a shop/library to print, you probably won’t have to do this. They’ll do it for you.
If there is a random printer random people have access to, I do not see how you can avoid checking all the settings and calibration yourself before you print; who knows what they did to it and whether it is even in working order?
I get things printed from
If you have a good shop near you, where you can do it yourself, that may be more fun and probably cheaper, but i bet there is someone with a “shop” at shapeways already selling the things, and you could just buy one.
I wouldn’t start with Roman dodecahedrons. They are a somewhat challenging 3D printing problem since they require supports to be printed for all of the pointy corners and hollow bits.
They shouldn’t be crazy difficult to do, but I wouldn’t recommend that as project number one.
How serious do you want to get into this? Is it a means of printing a few Roman artifacts, or is this something you plan on expanding into a full-on hobby?
Or you can buy one someone else designed on shapeways. 
Go to their marketplace
And search for “Roman dodecahedron”. There are a lot available.
(No, i have no connection with shapeways behind having bought a couple dozen items from them.)
Since you replied to me, I assume you thought I was referring to the complexity of designing them–no, I expect someone already has made them and the OP has access to a nice pre-made STL.
My point was that even with an existing STL file, with a traditional filament-based printer these are going to be tricky to print, so it might be best to cut one’s teeth on a few simpler models first.
It depends on how sophisticated and how deep you want to go. We have 3D printers in our STEM lab. They are limited in size to creating an object with dimensions, roughly, of 12x12x12. They have a menu of programmed objects that they can produce but, if you want to get into programming custom made objects, you can create the programming by inputting the proper information into the machine. These cost us about $5,000 a piece, so a good one is pricey.
Oh, one more thing that our administration didn’t think of before hand. Because of all the precise movement, they tend to need a lot of servicing. If you get one, you should get a maintenance contract with it.
I would recommend you find your nearest FabLab, and go speak to a flesh person there.
OPEN LAB
Jeden Mittwoch (ausgenommen Feiertage): Open Lab mit einer kostenlosen Führung um 19 Uhr, bei der wir alle Maschinen vorstellen und deine Fragen beantworten!
Alternativ dazu kannst du nach Terminvereinbarung zu einer kostenlosen Besichtigung (max. 5 Personen aus einem Haushalt) vorbeikommen und dich bei uns im Lab umschauen.
And they can assist you if der blinkenlighten is kaput.
I have been thinking of building a small airship (1.5 - 3 m3) with a concrete design idea I have for which 3D printing seems to me to be the only sensible way to proceed. My design idea is quite filigree, for lack of a better word, but I believe it would give me better maneuvrability in certain conditions than the traditional quadcopter drones. With a width of up to 1.5 m and a legth of maybe 2.5 m this is not something to 3D-print at home, I guess. So I would start with the dodecahedron or something like that and see if I get the knack of it and whether I like it.
It would also be a great exercise in procrastination for the other projects I have, while those other projects could always be useful to procrastinate on that new one. Procrastination is something I am both good and creative at.
Sounds more than sensible, thanks!
They have Blinkenlightens! That settles it: I must try this. 
That is an understatement. A home printer with a 50 cm x 50 cm base is considered large.