Problems printing objects that extend towards the edges of the screen

[chri2]

I've tried printing planes, because I wanted to learn how to get a case for some PCB out of the printer. The case is much larger than the printer allows to print, so I cut it into pieces and tried different ways printing it. The parts that do not have overhangs are best printed flat on the platform.

The other parts didn't come out very well, yet. I always have problems towards the edges of the screen, like here:

For this object:

This has been printed in Anycubic Water-Wash Resin+ using the following parameters:

Layer Height: 50
Initial Expos.: 40
Exposure Time: 7.00
UV POWER: 255
Bottom Layers: 5
Trans. Laysers: 0
B. Lift Height: 8.0
Lift Height: 5.0
Lift Speed: 0.50
Retr. Speed: 2.90
B.Lift Speed: 0.20
Rest Time: 0

I'd think those are quite slow and conservative, but I'm a beginner. There are other errors in the print above than the edges for sure. The supports end before they bow over to the vertical plain I printed. So there must be other factors to improve than the one I'll look at now.

I was wondering what could be the differences between printing in the middle of the screen and towards the edges and came up with two possible factors:

• peel force is probably different towards the edge of the release film
• light distribution might be different

I tried to mitigate the difference in peel force using slow lifting speeds which improved the result, but didn't stop the problems toward the edges.

To look at the light distribution I put a camera on the head and took pictures of the screen:

I took a picture while running the CLEANING RESIN VAT option. The picture is strongly under-exposured to offer a view towards the light source. The two faulty areas on the right edge are a problem with my screen that shouldn't have effected the print, because I avoided them when placing the object in the slicer:

The LEDs on the bottom of the tube can be clearly identified and there are some strong reflections from the mirror funnel.

I then put the vat on the printer filled with clear Water-Washable Resin+ and took another picture while running CLEANING RESIN VAT. This time the light has to filter trough the release film and the very low level of resin in the vat and the light distribution can be seen clearly:

The object I printed successfully and used to repair a remote control looks like this:

It is centered on the middle of the screen where the strongest light is emitted if my picture from above can be taken as some measure on how much light is emitted in which part of the screen.

Conclusion

I don't know about the effects of over-exposure when printing SLA, but for the edges of the screen the only solution seems to be to increase the exposure time even more.

Idea for improvement

It could be possible to compensate uneven light distribution if this would make sense at all:

• make a brightness map of the screen (e.g. by taking a picture)
• load the brightness map into the firmware
• calculate an exposure factor for each pixel depending on the amount of light in its area
• while printing one layer
  • start with all pixels for the png transparent
  • switch pixels in brighter areas to black after the given exposure time
  • leave pixels in darker areas of the screen longer transparent according to their calculated exposure factor

[chri2]

To illustrate the issue with another example: I used the service https://3dp.rocks/lithophane/ recommended by Alfredo Bonini to make a relief from this picture of a city wall in Campeche, México

and sliced it to be printed flat like this lying flat on the platform:

The result shows on the right between D and C edge the layers above the base layers (exposure 40 sec, same settings as above) missing. The missing piece below the D is the problem with my screen that has nothing to do with the suspected matter of the uneven exposure at the edges.

BTW: Thanks a lot for the hint to the page with the tool to convert pictures to reliefs! It is really great fun and I'm looking forward to learn how to do that in freecad!

[chri2]

If uneven light distribution is an issue - maybe it could also be solved optically by using a diffuser inside the funnel. If I can get my hands on one that is good for UV light I'll try...

[chri2]

I replaced my broken screen to get rid of the little cut-out around the 'D' and tried using the following parameters:

Layer Height: 50
Initial Expos.: 30
Exposure Time: 20.00
UV POWER: 255
Bottom Layers: 5
Trans. Laysers: 0
B. Lift Height: 8.0
Lift Height: 5.0
Lift Speed: 0.20
Retr. Speed: 2.90
B.Lift Speed: 0.20
Rest Time: 0

The platform tilted during printing and after solving that the print came out quite well: