DLP/LCD Test Print

[jt2156]

Good afternoon,

I am developing a part using SLA 3D printing and obtaining some fairly decent results but because my part is very small I am getting an issue with resolution. I have read that DLP or similarly LCD 3D printing work in a slightly different way. I have also read that with an anti-aliasing tool, people have had a lot of success using DLP printers. Hence I am reaching out to see if anyone would be willing to test a print for me? I have contacted Elegoo directly, as they state very good specifications of their printers but they couldn't help and suggested I use this forum.

Would anyone be willing to help me and test a print? I am based in Bath, U.k.

My part consists of 121 cone shaped structures of height 0.6 mm and width 0.3 mm sat on a coin of 10 mm and thickness 2 mm. I print my parts supported and not directly onto the build plate and normally uses less than 1 ml of resin and 0.025 layer thickness. I have attached an .STL file and haven't added any supports/anti-aliasing features as I am not sure how to the stl, except through ChiTuBox/PreForm when setting up the print. 

Any help will be greatly appreciated.

Thank you,
Joe

[cliffyk]

The file does not seem to be attached?

[jt2156]

The file does not seem to be attached

Sorry, I didn't realise I could - it should be attached now.

[cliffyk]

Hello,

I just sliced and printed your file, using Chitubox v1.8.1 (0.020 mm layers) and my plain 'ol Mars printer.

The result is not promising as I do not believe this class of printer (not just my Mars, but all hobbyist printers in this class) have the resolution necessary to do what you wish to do--in fact I do not know if the photo-initiated curing resin technology is capable that sort of resolution.

To back up a bit here are the specs for my Elegoo Mars printer (serial # M20Silver20200401)

LCD Shutter aperture (maximum print) size: 120 x 68 mm (X x Y);
LCD Shutter pixels: 2560 x 1440 (X x Y);
∴ LCD resolution = 0.0469 x  0.0472 mm (X x Y);
Z-Axis resolution is selectable @ 0.010 - 0.050 mm;

The Elegoo Mars Pro specs are:

LCD Shutter aperture size: 115 x 65 mm (X x Y);
LCD Shutter pixels: 2560 x 1440 (X x Y);
∴ LCD resolution = 0.0449 x  0.0451 mm (X x Y);
Z-Axis resolution is selectable @ 0.010 - 0.200 mm;

The Elegoo Mars 2 specs are:

LCD Shutter aperture size: 130.56 x 82.62 mm (X x Y);
LCD Shutter pixels: 2560 x 1620 (X x Y);
∴ LCD resolution = 0.0510 x  0.0510 mm (X x Y);
Z-Axis resolution is selectable @ 0.010 - 0.200 mm;

Note that while this printer is touted as having a "high resolution" LCD shutter due to its 2560 x 1620 pixels; it is actually lower resolution due to its larger aperture (print surface).

The Elegoo Mars 2 Pro resolution specs are the same as the Mars 2 as listed above;

The Elegoo Saturn specs are:

LCD Shutter aperture size: 192 x 120 mm (X x Y);
LCD Shutter pixels: 3840 x 2400 (X x Y);
∴ LCD resolution = 0.0500 x  0.0500 mm (X x Y);
Z-Axis resolution is selectable @ 0.010 - 0.150 mm;

I.e. the vaunted Saturn model, while having a significantly larger LCD aperture still has the same 0.0500 mm resolution, a bit "grainier" than my plain ol' Mars and the Mars Pro.

I believe the Mars "Saturn" is close to "state-of-the-art" for this class of printer--again, these are "happy-homeowner" hobbyist machines seemingly mostly used to print collectible "miniatures" (I.e. dolls, etc.).

I gave my printout a 40 minute post print curing in my DIY curing chamber--a converted  0.7 ft³ microwave with a 30 W 405 nm LED in place of the Magnetron--I'll run out to the barn and get it so we can see what it looks like.

Here we go, in no particular order--sorry 'bout the debris, my shop in the barn is not the world's cleanest environment:













I believe the output could be improved some by tweaking exposure times, lift speeds, etc; however the primary issue as I see it is that your tiny cones, by definition, terminate vertically at a zero dimension "point"--however with these machines the smallest a point can possibly be is 0.0449 x  0.0451 mm (if using the the Mars Pro)--throw in the inherent imprecision of the resin printing process (light scatter in the resin, unwanted excursion of the curing once initiated, etc.) and the practical resolution is far larger than that ultimate mechanical potential. I would be surprised to find 0.1 x 0.1 mm in a final product.

[cliffyk]

Thinking this over further while dealing with my insomnia--it occurred to me that pictures are = 1000 or more words--so...

Using FreeCAD I created a single cone with a 0.3 mm base diameter and 0.6 mm height, then exported it to an .stl file and sliced it with Chitubox with  20 micron layers.

Examining the sliced .ctb file with UVTools (which would not properly display the layers) I was able to retrieve a visualization of how it would print (keep in mind we are building this cone of mushy little 0.05 x 0.05 x 0.02 partially cured "bricks" of resin:

Here it is in FreeCAD:



in CHITUBOX:



and in the UVTools print simulation (not pretty, it appears to be only 12 layers as the "inline" 20 micron layers stack up):



In the actual printing process the uppermost layers I suspect each, comprised of just a half-dozen or so "mushy little bricks", cannot survive the stress of being torn away from the FEP and are just left behind as detritus in the resin vat.

I am not intimate with DLP resolution or specific process, perhaps it could do this better?

[cliffyk]

I thought of another way to visualize this. Here are Chitubox's previews of layers 1, 8, 15, 16, 22, and 23 (bottom to top) of our 0.3 x 0.6 mm (dia. x H) cone as presented in the "sliced" model window.

layer 1 - comprised of just 32 little 0.05 x 0.05 x 0.02 mm "mushy bricks":


layer 8 - just 16 "bricks":



layer 15 - just 5 "bricks":



layer 16 - down to only 1:



layer 22 - also just 1 "brick":



layers 23 to 31 - do not print, as they are smaller than a single "brick":



Some quick research into the DLP process seems to reveal that their resolution is no better and often not as good as MSLA (Masked Stereolithography Printing--I.e. "LCD"). Laser SLA may have some advantage, however the best I see in that arena is 10 micron (0.010 mm) XY resolution--with 25 micron being more typical. That twice as "high" as my lowly Mars, but still not really enough to meet the 0.3 x 0.6 mm "cone" test.

I also found tremendous levels of confusion regarding what these various resin based 3D printing systems are called--for example conventional wisdom almost exclusively limits the term SLA to MSLA (LCD) based systems--and I am as guilty of this as anyone.

This graphic comparison straightens out a lot of that confusion:

[cliffyk]

Formlabs offers a "free sample part" printing service--they are a VERY qualified 3D printing company and would very likely be able to tell you if what you seek is even possible...