Question on Minimum Feature Size

I've recently become interested in 3D printing, and I'm researching the printers before I buy my first one. I've noticed nearly all of the printers list the layer thickness, nozzle diameter, and filament diameter, but nowhere can I find a minimum feature size that can be reliably printed by the tool. Is there a rule of thumb for estimating minimum feature size? Or is it more a matter of how well you can calibrate your printer so the minimum feature size = Nozzle Diameter x Layer Thickness?


I'm mostly intrested in finding out the minimum size of a gear tooth I can make with a 3D printer. What is the smallest sized gear tooth I can reliably expect to make with a Prusa?

In the horizontal plane (X and Y), minimum feature width is determined by your nozzle diameter. Around 0.5mm wide is a good ballpark for minimum width. Minimum length is probably about the same.

Minimum height (z) can be as low as 0.01mm on a very well calibrated printer. A more typical minimum height would be 0.05mm.

Thanks! What would I have to do to get it calibrated well enough to do a 0.01mm layer? Most tools usually list a 0.1mm layer thickness.

That is because below 0.1mm, printing reliably on a wide variety of objects with the same settings becomes very difficult. In most cases there is no improvement in surface finish, and surface finish may actually become worse.

Put another way, layer heights of 0.1mm and above are pretty "plug and play". Below that you have to know what you are doing and be prepared to do some tweaking.

Thanks again!

So, just to be clear, going below 0.1mm in a layer thickness requires software tweaks specific to my printing job. Are there any hardware changes that would also have to be made? A better extruder, etc.?

Can you give me some examples of what sort of tweaks I would have to do for certain conditions? Or, are there any resources online that I can go to that would show me how to do this?

You have to remember that with a 3d printer, feature size is sort of a continuum. While you might be able to achieve a certain layer height and mechanical resolution, in the end you are squirting plastic out of a nozzle. For example, you are never really going to get a perfectly sharp corner.

It's also important to consider that when you print with a lower layer size, you are creating a wider width of bead. For instance, if you're using a nozzle with a diameter of .35 mm you will have a bead size with a cross-sectional area of a .35 mm circle. If you print with a layer height of .1 mm you are squashing that bead down in height, and it will increase in width. The cross-sectional area will remain essentially the same, so the width must increase if the height decreases.

Based on my calculations it would go like this for a .35 mm orifice:

Layer Height       Bead Width
   0.3               0.385
   0.2               0.524
   0.1               0.984

These are theoretical values, of course, and your mileage may vary, but it gives some idea of what changing the layer height will do in terms of the bead laid down.

DGC PartWorks Wrote:
> Based on my calculations it would go like this for
> a .35 mm orifice:
>
>
> Layer Height Bead Width
> 0.3 0.385
> 0.2 0.524
> 0.1 0.984
>
>
> These are theoretical values, of course, and your
> mileage may vary, but it gives some idea of what
> changing the layer height will do in terms of the
> bead laid down.


This is not true, because it assumes a fixed flow of plastic at the nozzle. In reality, the slicer adjusts the flow to suit the requested width (or width over thickness). So, printing very thin layers actually means extruding very little plastic per mm of length, while keeping the extrudate width constant.

Edited 3 time(s). Last edit at 11/02/2012 02:29PM by dzach.

> These are theoretical values, of course, and your
> mileage may vary, but it gives some idea of what
> changing the layer height will do in terms of the
> bead laid down.

As the previous poster says this is not true, the bead is stretched as it exist the nozzle, in general you want a w/h ratio of about 1.5, that ratio needs to get bigger as you reduce the layer height, because it also needs to be close to the nozzle width.

The problem with printing at very thin layers and width is that you end up stretching the extruded filament to the point of breaking so that sparse infills and overhangs fail. In this case, you need to print at 100% infill with no overhangs to have any success. Also, when your nozzle size is much larger than the width that you're trying to print, you don't really end up with the desired x-y resolution.

Thanks, everyone for answering my questions.

So, to summarize the comments:

The minimum x-y feature size is determined by the nozzle diameter. The feature height in the z direction is determined by the layer thickness, which is variable, but has limitations if you go too thin, such as poor sparse infills and overhangs. A good rule of thumb for the z axis is a layer thickness that is 2/3 the nozzle diameter.

Is this essentially correct? Are there any other details that need to be added?

There is nothing that limits the layer thickness to 2/3 of the nozzle diameter, but you have to increase the width as you decrease the layer thickness as DGC Pathworks pointed out. In other words, you have to sacrifice x-y resolution if you want to increase z resolution.