Here's the test I've developed for the Anycubic Photon LCD printer.

https://github.com/altLab/photon-resin-calibration

This is one of the steps to find the best resin settings. if you want to integrate something alike in the SLS/DLP section feel free to pull ideas from there I'm also open to collaborate with you guys if you want.

Best Regards

The underside is not flat on the bed - is that supposed to be testing something, or is it just to annoy us ? (We can't add supports because then the overhang test won't work, ...)

original is here ; https://www.thingiverse.com/thing:2656594

fun that you pretend 'autodesk' now made this

The tower OpenScad configurator code was based on Temp Calibration Tower by eibwen

The process was defined by miself, it is released under CC.BY feel free to reference it modify or add to the protocol if you like.

https://www.thingiverse.com/thing:3183883

probably in one or all of the README files

The team at formlabs who created the Form2 SLA machine have a jacad generated STL to test machines:

https://pastebin.com/XsYbhss9

The resulting testform object is very good at quantifying SLA performance down to high resolution features.

From the code you should see that it generates:

• Positive features (wires, walls, etc).
• Negative features (holes, slots, etc.).
• Structural features (bridges, overhangs, etc.).
• Dimensional accuracy (small dimensions, larger dimensions).

More explicitly:

• bridges on 2 axis
• internal thread
• hole geometry
• column geometry
• unsupported walls on 2 axis
• slots on 2 axis
• overhang angles on 2 axis
• text rendering
• hex stud
• external thread
• base definition test (print directly to build surface)

Surely it is a good candidate for comparing such machines - where each must be tested individually.

One issue we have is horizontal holes experiencing sagging during cure/cooling process.
Any chance of adding a horizontal hole feature (maybe to the pillar?)

PrusaSlicer-2.2.0+linux-x64-202003211856
Loading of a model file failed.

How about adding a spherical shape?

I have noticed spheres tend to expose all sorts of smoothness issues.

In presented example, X error and Y error values are presented as positive or negative values, while they should be presented as absolute values. For example, let's assume we've got following X measurements: 26mm, 19mm, 15mm, 9.5mm, 4.5mm, then, using current method X Error (X Error=Measuerd X - Target) will be: 1mm, -1mm, 0mm, -0.5mm, 0.5mm, and then average error will be 0, while real error is huge. Method should be changed to use absolute values as X Error and Y Error values (X,Y Error=abs(Measured X,Y - Target)

There should be a unified place where users can submit results. Something very simple as this board would do, currently its hard to hunt down the results.

Specifically regarding the FFF (FDM) comparison, if the goal is to get an apples-to-apples comparison between printers, then we should also specify the nozzle diameter and layer height to be used.

Likely there will be updates to this test, please add some version/revision number.

Relating to issues: #7 #8 #10 #27

It's...fairly difficult to gain a baseline level of understanding of the 3D printing world, and adding information like what is being asked for in these issues to a benchmark project like this would go a long way to helping create that easily accessible reference of basic knowledge that often permeates other hobbies.

If you'll pardon me to plead my case:

Having just setup my printer, done a few of the pre-compiled models that had prebuilt gcode bundled on an SD card, I decided I wanted to try this kickstarter test shape and compare to the known score for the model printer I have so I can see how well I've set the thing up as compared to the factory-assembled version. Unfortunately no one seems willing or able to provide the standard settings that were used to slice the model when creating that benchmark.

In a comment for closing #10, @andreasbastian wrote "Standard settings are the settings that a printer/slicer/materials manufacturer/provider prescribe for using their product." but this doesn't actually answer any of the questions being asked or help understand how to use the model. I've consulted every manual I can find for my printer and slicer, and nowhere does anything suggest an "optimal", "default", or "standard" layer height, infill %, infill pattern, or value for any parameter that might impact the quality or time taken to print.

I use PrusaSlicer and this is further complicated by several fairly equally recommended quality profiles ranging from "Ultradetail" to "Draft". These profiles vary estimated print times between 2 hours and 22 hours. The layer height settings of these profiles varies between 0.05mm and 0.3mm. Whether supports are printed is also up in the air but doesn't vary by profile.

I think they should be printed without support material and that adding supports would defeat the point of the overhang and bridging tests, but no one will write anything down to confirm or deny, officially, how I'm supposed to print this model. The only real evidence I have is the DrVax youtube videos comparing Ender 5 and Prusa MK3 where he shows removing the print from the sheet. In his video he doesn't have any support material on these, which are implied to be fresh from the printer.

But presumably layer height resolution can also impact how a lot of model features score like fine flow control, negative space, maybe even bridging and overhangs but I'm just guessing. So when the readme says something like "Prusa i3 MK3 scores 22.5"...what settings were used to get that score? At very least what basic profile or layer height? If you included how long it took to print I could at least pick the quality profile that's nearest to that length but as it stands I am completely in the dark here.

Obviously there is almost infinite tweakability in printer and slicer settings but just by picking a different built-in profile and not knowing anything I can make a Prusa take over a day to print this model, but no one will say whether that's the right way to do it. And measuring print time like suggested in #8 is almost useless without this baseline assumption of the resolution the model is being printed at since even small changes to that layer height value change the print time dramatically.

tl;dr: Please for Pete's sake, just include the most basic print parameters for most slicers in the readme, it'll help out the newbies immeasurably and really increase consistency in benchmark data.

FDM Assessment missing photo in section 3, Fine Negative Features.

If we are looking to get a true apples-to-apples comparison between printers, the amount of time spent printing the objects should also be taken into comparison.

Just my two cents, but I would suggest that additional points be allocated towards print speed with a faster time to completion being awarded more points than slower printing. (and this should be the actual print time - not the theoretical time calculated by the slicer)

Please consider changing the model image to something other than a White Model on a White Background. Please see the difference in the images below.

3MF is lightweight, portable, and a heck of a lot better than STL ;) @andreasbastian, let me know if I can help generating the files (if you're open to including them)!

Can you provide stp format so we can edit?
we can't use the availabe format for editing

You should disclaim, or better have a config json of this so called standart!

I suspect that the answer is "of course they should be printed without support material". However, in the interests of getting the most consistent results across users, it would be good to specify this.

Looks like a score might be missing in the FDM Protocol, under "Overhangs". 1, 3, 4, and 5 are listed, but not 2. The photo above shows a print with a score of 2.

in order to test how it spread the heat away when in a small spot print area for each layer ?

Looks like the rightmost 2 negative features plugs are not properly labelled.
They read 0.3 and 0.5 instead of 0.5 and 0.6 (or 0.7 may be?) respectively.

We ran the test print on an Ultimaker 2 which has been upgraded with Ultimaker 2+ parts set. Part was placed and prepared using Cura 3.5.1

File: ksr_fdmtest_v4.stl

Material: PLA

Nozzle: 0.4mm

Layer Height: 0.1mm

Infill: 20%

Generate supports: No

Print time: 9h 31min