So, if I want to use it commercially, the best option is always the Single Mesh one.
Thanks, Manuel! :)
I think that´s the way to go
This was the first question I had when I got my up plus 2 and I did some tests to figure it out.
Single meshes are most reliable. if you model has thickness it's technically 2 meshes combined with opposite facing normals. The dicer uses the opposite facing normals to create the instructions to lay out the filament.
It does not have to be one mesh but it cannot have overlapping in the mesh. Technically everything in the .stl gets grouped under one shape / mesh node.
Say you had two intersecting spheres: you would need to perform a boolean operation to remove the faces from the inside where the spheres intersect. The problem with having faces overlapping is the printer will try to print those faces and by doing so it will melt holes and seams into the area because its trying to make something inside of something else.
And its better to break multi-part models into seperate .stl files so people can print them separately if they want.
Here are some other guild-lines I follow.
If you've seen a 3d printer in action and can think about how surfaces are made, that helps. All surfaces must be a minimum of 1 mm thick in order to work on most home printers (unless you are doing a single surface shell print). Details can be smaller than 1 mm but the actual surface thickness must be 1 mm. 2-3 mm works great for functional objects like vases that you actually want to hold water. A single layer shell print will not hold water unless you treat it in various ways.
If you can, avoid creating overhangs that reach out from the main surface at hard angles near 90 degrees as this will require support material and the quality of the model in that section will be rougher.
I test every model that I make and sometimes have to go through 4-5 revisions before I can confirm that the design will work. Anyone without a 3d printer making designs for sale should consider the complexities encountered when someone buys your file and it fails. Not only does it waste their material but then they want to return their model. My fortune cookie model works great on my printer but my first buyer of it wanted to return it because they said the two halves didn't click together. I sent them 2 alternative files and they still left a bad review. And this is a file that I made 7 revisions on before publishin. I am going to predict that the return rate on 3d print models is going to be near that of womans clothing.
3d Printing for the masses is a design challenge for sure.
Anyone thinking that you can just load up a model and click print is wrong. Sure . . .some printers can deal with anything, but most home printers cannot. What is the optimal layer-height? What kind of infill is recommended. Are there any special instructions like 'do not print it from this angle'. Does it require support material to print successfully? How long will it take and how much material will it use. Some materials like ABS have a tendency to warp if there are large flat sections touching the print bed. Nylon like taulman 618 is much more forgiving, but not everyone is going to have the desire or capabilities to print in nylon.
These are all questions that a potential buyer may have.
Most models that I 3d print takes between 4-8 hours to print on medium quality and use about 1/15 of my filament reel. That is an investment that I would rather make then have my buyer get mad for wasting their time and money. At $30-40 per reel you want the design to work, but it never comes out quite how you imagined the first time and then its time to revise.
I'm not trying to troll but just to establish the realities involved, that 3d printing any given model is still an engineering challenge that should be worked through by the designer of the model. Hope this helps anyone with these questions!~
