There are a lot, I mean a great many, parameters that impact the quality of a 3D print. This is what makes calibrating a printer so much fun. Two of the parameters that provide the most visible impact are layer height and printer speed. These two parameters are also the ones that most dramatically impact print time. The higher the resolution, the more layers, the longer the print. Obviously the faster the printer goes, the shorter the print!
When I got the order for the truck I was in the process of setting my prices on 3D Hubs so I decided to do some tests using one of the tank models that I downloaded. Obviously the time that a model takes to print has to be the largest factor in how much it is going to cost via my hub. Equally obviously the customer should be able to expect a high quality print!
The printer that I used for this test was the Wanhao Duplicator 4S printing ABS. I printed seven copies of the same model as follows:
Model |
Resolution |
Speed |
Time |
0 |
0.1 |
|
74 |
1 |
0.15 |
Fast |
48 |
2 |
0.15 |
Slow |
66 |
3 |
0.2 |
Fast |
36 |
4 |
0.2 |
Slow |
58 |
5 |
0.25 |
Fast |
30 |
6 |
0.25 |
Slow |
47 |
The "Fast" speed was 43mm/sec and the "Slow" speed was 25mm/sec. I did not do two speeds for the 0.1 resolution print as I am not pricing that resolution on the Duplicator 4S. The time is minutes.
Above is an image of all the test prints arranged with the fast prints in the front (0, 1, 3, 5) and the slow prints in the back (2, 4, 6). The differences in quality that you get via the layer height are patently obvious!
Not so much the difference in quality from the speed. This surprised me as I know there can be a huge difference but it obviously depends on the model.
One of the issues that I have seen that can be caused by speed are
whiskers. As I write this I am doing a print that I know is running
faster than it should and there are whiskers!
|
Speed Whiskers |
There are several parameters in most slicers that deal with this kind of issue (oozing). In Simplify3D you can adjust retraction, coasting, and wiping. In the case above I have it set to to a retraction and to coast (at the end of a loop of plastic). I did not have it set to wipe as I tend to see this as a negative more often than a positive. I also know from experience with this print that slowing it down would result in a clean print. I chose not to do that as the manual cleanup was worth the fast time to print.
In any case, no matter how hard I look I am not seeing a difference between the slow and fast prints when looking at my test cases.
|
Left=Fast, Right=Slow |
|
Left=Fast, Right=Slow |
Above are images of the test prints. You might be able to claim that the layers look more uniform but I am not seeing something jump out at me as to a big difference.
I am going to do another print at an even higher speed and see what evidence that produces. My suspicion is that we will not see much of a difference.
Here are the test prints that more clearly show the differences in resolution (layer height).
|
0.1mm |
|
0.15mm |
|
0.2mm |
|
0.25mm |
I printed two more copies of the test model, this time with a larger speed difference of 25mm/sec and 50mm/sec. There don't appear to be any fatal flaws showing up but the stitching of the filament loops is definitely tighter on the slower prints.
|
Faster - Slower |
|
Slower - Faster |
I don't think that this experiment yielded the result that I expected but I think that I know the reason why. Namely that these models are simply too small. I doubt that the extruder assembly is ever up to speed and there are no large distances for acceleration and deceleration to produce the kind of jerking that would cause an issue.
Given that this test did not give me a result my experience has been that bad things can start to happen above 50mm/sec and sometimes even a little below that threshold. Based on my experience I decided to use the speeds of 33, 37, and 42 for the three levels of quality that I offer on my two printers on 3D Hubs. Obviously I will adjust on the fly for individual prints!