3

When choosing a layer height, I know that often you go as fine as your printer will do for better precision, but sometimes you go a little thicker, for speed, for example.

I also see 0.1 mm and 0.2 mm as common thicknesses.

What are my options here? When I'm working on a part where I want to print a draft piece, and the quality matters less, can I set it to 0.15 mm? 0.11 mm? The Ultimaker Cura slicer I normally use will let me put in almost anything, but what can it really do? If I can use values in between simple 0.1 mm increments, are there reasons I might want to do so?

For reference, I have a Monoprice Maker Select Plus with a 0.4 mm nozzle and, again, Ultimaker Cura as the slicer. But more general answers for other printer types and slicers are also encouraged. I want to know about this generally, and not just for one printer.

0scar
  • 32,029
  • 10
  • 59
  • 135
Joel Coehoorn
  • 2,115
  • 3
  • 11
  • 30

4 Answers4

4

Layer hight can be theoretically anything as long as it fits into these ranges:

  • it needs to be at least one step for the Z-Axis motor to be physically possible
  • it should be at max 3/4th of your nozzle diameter to create an adhesion surface

In praxis, the lowest setting for layer height due to physical limitations of the extrusion systems is around 0.05 mm.

Also, not any number is possible, it is dependant on the Z-axis system: since one or one-half step is the smallest rotation that the stepper can achieve, the raise that belongs to this partial rotation is how much one and the next layer height can be together. The step limitation usually is of no concern though, unless one has a very steep lead screw.

As a rule of thumb: doubling layer thickness results in almost half the print time. 0.1, 0.2 and 0.3 mm are common because they allow easy checking the printed accuracy in Z-axis with calipers.

Trish
  • 20,169
  • 10
  • 43
  • 92
  • @JashJacob No, stepper motors only allow full or half steps, where a step depends on the number and arrangement of coils - and the range is given in degrees. 5 microns are a length, not a rotation. 5 microns is what is the *result* of the turn via the leadscrew. – Trish Jan 07 '19 at 14:49
3

You decide which layer height you want based on the quality you desire, but never go over about 75 % of your nozzle diameter, so with your 0.4 mm nozzle never choose layer heights larger than 0.3 mm. The rationale of this rule of thumb is that the filament leaves the nozzle as a tube and needs to be flattened to make it adhere to the previous layer.

0scar
  • 32,029
  • 10
  • 59
  • 135
3
  1. The thicker the layer, the smaller the number of interlayer contact zones (this is a weak point), the higher the strength of the part.
  2. For good interlayer adhesion, the thickness of the layer should not exceed half the diameter of the nozzle (this is pure geometry). The recommended value is 0.8 * 0.5 * D.
  3. There is an opinion that the thickness of the layer should be a multiple of the height of the Z climb to one full engine step (in order to avoid accumulation of error).

Thus, for an engine with a step of 1.8 degrees and a screw with a step of 8 mm per revolution (most common configuration of Z-axis), the layer thickness must be a multiple of 0.04 mm. For a nozzle with a diameter of 0.4 mm, the recommended layer thickness is 0.16 mm, maximum valid is 0.2 mm.

  • No, sorry, I meant exactly **count** of interlayer contact areas. The smaller the number of such areas, the higher the strength. – Dmitry Kasatsky Jan 07 '19 at 13:05
  • Ahhh, yea, I got that twist... but count and area are proportional. you might mean "number of extrusions/contact area"? that is antiproportional to the strength, as you gain in-layer-inter-extrusion boundries. – Trish Jan 07 '19 at 13:32
  • Translation difficulties... I meant "the number of interlayer contact zones". Finally fixed. – Dmitry Kasatsky Jan 07 '19 at 13:38
  • I would like to debate the first point, many of my failed prints have failed through layers, not only alongside the layers. Could you explain where the formula comes from? – 0scar Jan 07 '19 at 13:49
  • Sorry, @0scar, I did not understand. The first point does not contain formulas, only general considerations. The strength of the interlayer adhesion zone is less than the layer itself, this is obvious. The strength of the interlayer adhesion zone varies from layer to layer and is distributed between the minimum and maximum values. The more layers, the greater the likelihood of a zone with a strength close to the minimum. The ideal case is a molded part (one layer, no interlayer adhesion zones, maximum strength). – Dmitry Kasatsky Jan 07 '19 at 18:18
  • @DmitryKasatsky The second point does contain a formula, comments cannot contain an enter (new line), therefore it looks at though I meant it to be for the first point while I meant the second, please explain. Nice explanation about the strength of layers, but I seriously doubt that considering the failures I've seen. It all depends on the loading of the part. – 0scar Jan 07 '19 at 18:39
  • @0scar, the effect of the number of layers on strength is small, probably it can be neglected. I'll look for the source of the formula from the second point. – Dmitry Kasatsky Jan 07 '19 at 20:23
  • @0scar, sorry, but the source in Russian: https://3dtoday.ru/blogs/filamentarno/the-intricacies-of-3d-printing-part-2-theory-and-practice/ This is an article by a famous russian filament manufacturer. Try to use Google translator if you want. – Dmitry Kasatsky Jan 08 '19 at 10:21
  • If I understand this right: 1.8 degrees is 200 steps to get a full revolution (360 degrees ÷ 1.8 = 200). Therefore the 8 mm per-revolution for the threads is also 8 mm per 200 steps, such that 8 / 200 results in the .04 mm per step. Thus, rather than .1 mm layers (the advertised machine minimum), I might do much better at .12 mm, which is an exact multiple of .04 mm we move at each step. Does that sound right? – Joel Coehoorn Apr 30 '19 at 17:43
  • Doing more research, and the rods used in a given printer can vary widely, so the 8 mm revolution distance is just one example. The main thing is the firmware in the printer knows what it needs to do. Also, stepper motors can take partial steps, so we can get a lot more granular. – Joel Coehoorn Apr 30 '19 at 19:17
2

Let's start with the stepper motors used to move the z-axis. Most printers use a Nema-17 motor, and most Nema-17 motors turn 1.8° per step. For the motors that don't use 1.8°, almost all of them are .9°, such that assuming 1.8° is pretty safe for most us (if you have a .9° motor and wrongly assume 1.8°, the motor will just take two steps instead of one).

But these motors don't move the print head or bed directly. Rather, they will turn a threaded rod to lift or lower the necessary part. We need to figure out the linear distance traveled along the rod per step.

There are two ways to determine this. First, we can look at the threads on the rod, and think about the linear distance traveled if the rod turns one complete revolution. Use a sharpie marker to trace a thread, and then measure the distance.

Alternatively, if we check the specs for the rod in terms of threads and pitch. Most 3D printers use 8 mm rods, but looking at my own printer these aren't standard metric threads. Indeed, the pitch can vary quite a bit from printer to printer. Both 2 mm and 3 mm pitch is common, with the number of threads on the rod varying from 1 to 4. To find the linear distance of a revolution, you multiply the threads by the pitch. A 3 mm pitch with 1 thread has a linear distance of 3 mm. A 2 mm pitch with 4 threads has a linear distance of 8 mm.

With either method, we know the linear distance for one complete revolution (360°). But a single step only moves a small portion of that: 1.8°. Thankfully, this works out to an even number: 360/1.8 = 200. There are 200 steps per revolution.

Now we know enough to figure out the step distance. The math looks like this:

StepDistance = (RodPitch*RodThreads) / 200 

As an example 2 mm pitch and 4 threads produces this expression:

(2 * 4) / 200

and the final result is:

.04 mm per step

Those values are common, but not universal, so you need to know what numbers to plug in for your printer.

You should also be aware that many stepper motors can take half steps or even divide a single step up to 256 times. If you can watch your printer smoothly move from very low to very high, or vice versa, without jerking for each step, this is what it's doing. That means your printer may be able to make adjustments as fine as .00015625mm. At this point, we're down in nano-meter territory. I think it's probably best to assume your printer wants to take at least a half step when moving from layer to layer.

Now lets look at applying this .04 mm per step value to a real printer. Please remember: this is just an example, where the .04 mm number is based off a specific rod type. You need to know the values for your specific z-axis rods.

Let's say you have a printer that advertises a .1 mm minimum layer height and a .4 mm nozzle. Given the .04 mm steps, you might do much better actually using .12 mm layers, which is an exact multiple of .04 mm. But then, if we consider half steps, .1 mm could be just fine. On the other end of the range, we don't really want to go above 75% of our nozzle width. That's .3 mm. Given a .04 mm step height, a better max is really .28 mm... but, again, considering half steps, we could in theory choose to do the exact .3 mm height.

The main thing I want to take away here is for when you're looking for an in-between layer height: .1 mm is too slow, but .2 mm is too course. It's probably best to bump up the layer in exact half-steps. For this example, that's .02 mm increments. So after .1 mm we try .12 mm, then .14 mm, and so on. Finer adjustments may be possible, but there's a power of two law at play here, so keep dividing things in half, rather than by 10ths. If you don't get good results this way, try assuming full steps, and start at .12 mm for the layer height, then go to .16 mm, and so on.

Joel Coehoorn
  • 2,115
  • 3
  • 11
  • 30
  • 1
    Hi Joel, see also [Full steps mode for Z axis stepper controller](https://3dprinting.stackexchange.com/questions/5796/full-steps-mode-for-z-axis-stepper-controller/5974#5974) – 0scar Apr 30 '19 at 21:22