I was quite excited about the new 3D print simulator. Until I tried it.
I read a couple of articles online talking about an incredible new service that apparently provided a virtual 3D printer that “behaves just like the real thing!”
Some publications suggested it could be used to test print jobs before running them to weed out problems. Others suggested it could be used to help educate everyone about 3D printers and how they operate.
With all that said, I couldn’t wait to try it out.
It turns out the tool is software that can run only on a Windows platform and requires a modern desktop with a current graphics card. I happen to have this available in the lab, so I downloaded the software and got it running.
The 3D Printer Simulator presents a virtual reality view of what appears to be an Ender-3 or similar desktop 3D printer on a table. You can spin around the view to see the action from all angles, hence the need for the graphics card.
The idea is that the “printer” will run a job and you can observe it (and the growing print) in real time. It’s also possible to slide the time forward or backward within the job with a slider. This is very much like the preview feature in popular 3D print slicing software, except that you see the job on what appears to be a lifelike printer.
The simulator provides several pre-sliced jobs, including a #3DBenchy. You can select from several pre-selected colours for the job, including a rainbow filament that slowly changes colour as it extrudes.
This sounds good so far, but then things went sideways for me.
When you start a print job, you experience not only the visual motion but also audio! The simulator presents astoundingly loud stepper motor noises as it moves to and fro.
That might have been realistic years ago, but there haven’t been any desktop 3D printers released in at least 3-4 years that produce those sounds. Virtually all machines now have silent stepper motors, so the simulator is simulating something from the past and therefore is less relevant. Imagine showing a newbie the simulator: they would expect 3D printers to make those noises, but they do not.
I saw no evidence that the simulator would present any of the common problems one encounters using desktop 3D printers, such as poor bed adhesion, stringing, erratic flow, blobs, colour changes, etc. The simulator simply runs through the sequence as if everything works.
That is not simulating reality, it’s simulating perfection. Desktop 3D printers, especially older ones that seem to be the subject of the simulation, fail constantly in many ways.
There is support for G-Code import. You can load in your own G-Code, sliced by you with your own slicer. You can then “run” the job on the simulator to see what happens.
I tried this several times with different G-Code files and in every case the simulation did not produce anything close to what was actually in the G-Code. Figurines with supported spindly arms came out as a massive square block, for example. I am also suspicious that there could be a height limit.
I gave up trying to make my own G-Code work, although the simulator does successfully complete its own preloaded jobs.
At this point I came to the conclusion that the simulator is most definitely not what was described on other sites. It’s not a tool for discovery of issues, and I don’t even think it’s particularly useful for showing people how a 3D printer works.
You cannot see any of the inner workings of the extruder during a job. You see only the rapid motion of the axes and a model appearing layer by layer.
For checking whether a print job will work, the simulator is effectively useless. It’s far better to use the built-in simulator found in the preview tab of your slicing software. That will show you exactly how the extrusions are made, revealing potential problems easily. You just cannot do that with this simulator.
If this sounds bad, I don’t want to denigrate the simulator. It is what it is: a fun simulation of a 3D printer. It is not something that can be used for any serious purpose as far as I can tell. The fault here likely lies in the erroneous reports that present the simulator as more than it really is.
Via 5MinLab
