Printing hollow framelike/boxlike objects

I’m at a bit of a loss here…
I’ve been trying to print some casings for a client and, while i had some luck with two of them…

… i can’t figure out a way to successfully print the third one no matter which orientation i try.
And i’ve tried 4 till now. By not successful i don’t mean it’s failing to print - it prints just fine, it’s that at least one of the walls (and usually at least two) have horrible surface finish and or end up deformed. Here’s the latest example:

That one was printed like this…

… and it had the best overall dimensionality and least amount of deformation/ripped layers, but the worst surface quality of any of the previous attempts by far:

Now, i know flat surfaces like these are a bad match for the Form1+ (and most SLA printers in general), but these particular objects are an even worse match for FDM printers (i know, i’ve tried - there were lots of issues, as simple as they look) so i’d really like to nail printing them successfully. In part because i’m pretty sure there’ll be similar requests at some point.

Besides, i think i’ve got the ones with the solid (or at least perforated) top cover down fine, it’s the hollow, frame-like shapes like this last one that are giving me a headache.

If anyone has a better idea (Formlabs folks?) on how to orient and support this object, here’s the .form file (note that the top surface has to remain free of support pockmarks as that’s where the printed/laminated mask with legend normally goes):

https://dl.dropboxusercontent.com/u/1702513/t0657_orig2.form

It looks like there are no supports in that area. The Pre-Form files is corrupt so I can’t open it and look. Maybe just pop a few extra supports in that window area.

Try opening the .form from within Preform (instead of double clicking). It’s a bug i’ve reported a few days ago.

Re: supports, there are supports all along those edges (at equal distances) and, if you look at the slices, you can see they’re pretty well supported by the previous layers (i.e. they’re not just thin, bridging slivers of material). It’s hard to tell from the photos, but even the parts that were directly supported on those photos have a pretty crappy finish.

A version of the print i did earlier, with denser supports (manually positioned) ended up worse - the edge that was supported at every 5mm or so ended up distorted by the supports (concave between them, instead of straight).

Bizzarely, the smallest casing, the one with the vents, ended up cleanest, despite being printed with the least amount of supports (basically, just along the two shortest edges).

BTW, is clear v2 supposed to be that elastic and bendy? These boxes are easily deformable, even when cured, and the Goldberg polyhedron i printed earlier actually bounces off the table.

I didn’t expect that.

That is strange I could only open like you said.
Orientation looks good to me, I agree with Josh that it just needs more support bellow those areas.

The part is soft during the print process, and the whole area is one big overhang until that one support finally connects. Glide through the slices and you will see how it takes forever to connect because it is on he far outside. I am in a bit of a hurry, but I reworked your supports.
Ante.form (372.6 KB)

3 Likes

Thanks for taking the time to build this setup; going to print it in clear02 for the sake of comparison to A_E’s original results. I haven’t flipped through the layers on yours vs the original yet, but looking forward to doing so when I have a moment. This is really helpful stuff for those of us just learning.

Yeah, realized it’s pretty soft even after the print process… I expected that side to be relatively stiff and for it to support itself along the Z axis by the point it’s built up (instead of every layer having to have to be supported laterally). Obviously, that’s not enough at that length and without a top side closed off.

Thanks for taking the time to tweak the supports, printing it now!

PS: I was originally avoiding adding supports to the sides, so they remain unblemished (the casings go into a bigger assembly as subassemblies and are stacked close without a lot of room, from what i’ve understood, so i didn’t want the sides to have warts i’d have to sand down). But considering the whole thing is very flexible even when cured, that’s not an issue anymore :smile:

Well that was interesting…
Additional supports did help with the surface quality immensely, but it also created a sort of “tenting” effect on the faces.

Which is when i’ve noticed that none of the sides are actually flat and the top left corner was distorted (off by a full mm, and not just bent out of shape - the peel actually distorted it during the build).

I think i need to completely rethink the orientation, this is slowly turning into more support than object, lol.

(and yes, i realize this is a problematic shape for this tech and resin now - it’s very “non-self-supporting” (that’s a technical term) across all of its faces… pretty much the only thing worse would be to print just a thin frame)

1 Like

Yea, but don’t get too obsessed with the perfect print. I used to do that because it is so easy to tell yourself the next one will be perfect. But results are a compromise. I just made the decision to sand all my customers models for them and clear-coat them. It was a pain, but they loved it.

Solid advice, but right now, i’m more concerned about it being dimensionally off (i’m mostly bugged by that “bent” edge, that’s just not gonna fly).

But you’re right, should’ve stopped at the first few attempts (that were pockmarked, but at least straight) :smiley:

Yea a bent edge wouldn’t be good for something that assembles. Are you going to add more supports and print again, or sand the first one smooth?

I’m gonna try one more orientation i had in mind (chalking this up as a learning experience) and if that ends up worse, going to sand the first one.

Also don’t forget you can sometimes achieve better results by breaking it into pieces and assembling them after the fact. I would have tried to be of more assistance but I currently don’t have a printer.

Yeah, i kinda ran away from that on my FDM machine. That’s how i originally printed them, but then gave up because it was too much hassle (dimensions went out of whack, among other issues).

I’ve printed the latest attempt with a different orientation and got fairly nice surfaces, but deformed corners (two this time). I guess it’s a tradeoff.

Funny thing - i tried one in black and the result was worse. It seems that the black resin is way more floppier than clear before fully cured / during the print. I had to support it from the front to get a straight print, which resulted in a horrid top surface (looks almost molten). Will take photos later.

In the meantime, printing the helical wall test at 50 um in black, i’m interested in what the result will be.

Okay, here’s the black one:

This one was printed almost vertically and on its short side. The face that looks “molten” or “wet” is the one that was facing the bottom (i.e. the build platform) the tiniest bit. The one that was facing the top (i.e. the vat) turned out perfectly (that’s the matte looking bottom or inside one in the second shot). The key thing is - the difference between the two angles was minimal. Really unexpected result, i expected that to be pretty much the ideal position, as far as the surface quality is concerned.

In brighter news, the previous prints turned out to fit the assembly perfectly, with no manual calibration required. And apparently, the finish was more than satisfactory (as far as the client is concerned).

Thanks for sharing. Not sure what to say about the results though. More supports can help a little. But mechanical stuff tends to print like that.

Well, for what it’s worth, more supports in this particular (last) print wouldn’t help with the top/bottom surface as they’re a single (thin) feature.

If one face’s texture is fine, lack of support can’t influence the other face of the same thin, flat sheet, right?

Rotating around Z (to change the peel direction) might, slightly.

I’m surprised by the differences between clear and black, though. Black is even softer than clear (and takes longer to post cure). The thinnest features on that casing are so flexible and soft they’re almost like rubber.

For a moment, i thought i’ve accidentally printed on the wrong resin preset and went back to check, but nope.

Also surprised by the surface finish. Difference in bottom vs. top facing sides’ finish usually screams “laser overshoot” (overcure in depth) to me, but black should be the least prone to it and clear the most. From what i’m seeing, it’s the opposite.

Yea if one side of the same feature is good then it’s probably not supports. But too few supports can be like it’s sitting on springs. Especially if you are having under-curing.

Is this the normal state of things for black Formlabs resin?

Cause i’m finding the fact i can tie the supports in a knot a slight bit ridiculous. It also seems the supports in black are a little bit thinner than in clear on top of the crazy flexibility/softness.

And before someone asks - yes, this is cured and after being left to air for a day. And even if it weren’t, it’d still be ridiculous.