Ideas on why these prints fail? or Why does it hate my sandwich?

I have had the Form1+ for about a week and I am very happy with it so far.

However, I have had a couple of print failures that I can’t rationalize with my understanding of how the machine works.

I have been printing several large, flat plate-like objects. When I print 1 or two on the same job there have been no problems. When have attempted to print more the inner pieces fail (if these were a sandwich, the bread prints fine but the meat fails).

I’ve adjusted the prints so that there is not a large surface area for the peel and spacing (the distance between them) does not seem to be the issue as I use the same spacing when printing 2 plates (which print fine). It should be noted that the failures always occur in the last 10-15% of the slices (which is annoying).

Below is a screenshot of the failed Preform model as well as a pic of the failure. The “unsupported” warning colors that that show up in Preform I think are a bug as these have never been a problem with 1 or 2 pieces.

Thanks for any ideas on why this might be happening.

Did you run low on resin?

What piece / pieces were left over in your resin tank?

Also: did you have a failed print before you printed this one? Make sure to scrape the resin tank after each print, and use a comb to search for debris in the resin. It is very easy to miss little pieces imbedded in the resin. (You can even go the extra mile and empty the resin tank and strain it through a paint strainer… you will be amazed at what ends up being caught in the strainer.)

Nope. Did not run low on resin. Following all failures there were patches of cured resin on the silicone under the failed area which I scraped off. I am very diligent about combing/cleaning the resin tank after every print (especially after a failed one). But again the failure has always been local to the center of the sandwich, never on the outside (bread) pieces. I have not had two failed prints in a row. I can print successfully in the same area as the failure occurred as long as I only print 2 plates max.

All that to say I don’t have any reason to believe this is due to a dirty resin tank which is why this is perplexing.

EDIT: I should mention that my 1 theory on this is that because it is a sandwich there is light bleed from the laser as it is curing the outer pieces. Eventually the light bleed causes a buildup of over-cured resin on the tank which causes the failure. Just a thought.

That’s an interesting theory, although I would think that would also happen around the perimeter of the other pieces if that were the case.

Normally when you find a patch adhered to the resin tank, it is because during a peel the model was pulled off of the supports and began to adhere to the resin tank.

I wonder what would happen if the three parts were not parallel to each other. It almost seems like the peel process is somehow altered by your geometry.

(Imagine the machine is starting the peel process, it builds up some tension then releases the tension as the 1st part peels… maybe as the peel reaches the inner part, it happens too fast (like a rubber band releasing its tension)? Just a wild theory…

EDIT: or, what if you rotate all three pieces to be roughly 45 degrees to the peel direction? Then the peel can take place as one event, instead of three discreet events.

P.S. more to my wild theory’s point… does the Z -height of the failed part correspond to the Z-height of your shortest part?

Wow I’m surprised that prints at all. I would have guessed the peel and return process would have blown out holes to equalize pressure as it printed all those closed cells. I’m guessing that is what happens. It’s possible Thomas_Alberti is right and that the more rapid peel rate as the tray snaps away results in the enough extra force as the pressure equalizes that it blows through the walls. But all in all closed cells are tricky and prone to blowouts.

@Thomas_Alberti …now that you mention it…there may be a correspondence to the shorter piece. I like your theory and will have to do some tests. I’ve been hesitant to rotate 45 because of issues that people have been reporting for tall pieces located far from the hinge side.

@RocusHalbasch there are no closed cells. It is an open diagonal grid on one side and I have had no issues that lead me to believe this has to do with blowout.

As a test, I just reprinted the failed piece in exactly the same location/position with exactly the same support structures but this time by itself rather than part of a sandwich. It printed flawlessly.

Sorry I thought there was a wall on the far side. Guess that explains why it didn’t fail earlier. :slight_smile: Oops. I still think Thomas_Alberti might be on to something though.

Ok so some updates here…and not exactly good ones.

I’ll admit I got a bit greedy in trying to print 4 pieces at once but it doesn’t seem to matter if the sandwich has 3 or 4 pieces in it…the failure always starts in the middle and works it’s way outward. I should also note that this print used a clean tray (I very carefully cleaned it out prior to this) and fresh resin that I received yesterday.

The first print I tried was the pieces at 45 degrees to test @Thomas_Alberti 's theory. Below is the print result.

As you can see the print failed again, in the middle and was creeping outward toward the outer pieces. I stopped the print about 80% of the way through because the failure tends to creep outward and would eventually ruin the other taller pieces. There was no correspondence between the height of the shorter pieces and the spot of the failure.

This is what was left sticking out of the remaining resin in the tray:

I removed this and another chunk that were stuck to the PDMS layer. Note the intense clouding in the area of the failure. This clouding was NOT there prior to this failure.

I decided to test one more idea which was to rotate the pieces perpendicular to the hinge side. I know this is not an ideal configuration but I thought that if it was indeed nonuniform peel stress that was causing this, a perpendicular configuration would at least change the location of the failure. This is what the configuration looked like (the hinge is to the right):

I set it to print and went to bed. In the morning I awoke to this:

Failure in the same location, plus this had somehow ripped the PDMS layer off the bottom of the tank. Self destruction. Fortunately it doesn’t look like any of the resin dripped into the base…but that’s about the only good news.

Obviously, this is not supposed to happen. After clean up, I noticed that there appears to be some clouding near the corner of the mirror as well:

So in summary:

  • It doesn’t appear to be the nonuniform peel force discussed earlier.
  • The failure always happens in the middle of the sandwich and spreads like a cone shape to the exterior pieces
  • The failure is not because of bad geometry of my print pieces. Pieces consistently print fine when there are only 1 or 2 (same piece, same location, same support structures) - it is only when they are part of a sandwich that I see this failure.
  • The failure is not due to a dirty print bed or contaminated resin.
  • The failures cause clouding on the PDMS (known issue)

Needless to say, I will be opening a support ticket. Disappointed.

Holy crap! I’ve never seen the PDMS pulled out that way!

Here is the response from FL this morning as well as my response. I hope I’m not off base in my understanding of what they are getting at.

Hi Ben,

My name is Luke and I work on the support team here at Formlabs.
Thanks for reaching out; I’m sorry to hear that you are having printing trouble. I have just a few questions and things to check so we can diagnose your issue properly.

I’m passing your post along to our engineering team to get their feedback, but the direct cause of your PDMS layer ripping out is the forces adhering the resin to the PDMS were greater than the forces adhering the PDMS to the acrylic of the resin tank. The strong adherence between the part and the PDMS could be caused by a number of factors, including tank age and the height of the models, as well as where the models were positioned in the tank. We actually recommend that when printing tall narrow models that you orient the long axis of the part perpendicular to the hinge, as this provides more stability during the peel.

Printing tall thin models is inherently difficult to do, as the taller a part is, the less stable it is, and the more partially cured resin build up around the part. With such a dense packing of parts as you’ve got here, resin can’t flow away from the part during the peel as effectively, resulting ultimately in bits of resin curing to the base of the tank. These small pieces grow in size as the print progresses, ultimately causing failure when they engulf the layer that’s printing.

If you like, you can send me your file and I can pass it along to our print technicians to get their feedback. I’ll let you know once our engineering team weighs in, but I hope this information is helpful in the meantime! Let me know if you have any other questions.

Best,

Luke

And my response:

Hi Luke,

I understand that printing tall thin models is difficult. I hope that you’ve had a chance to look at the thread I started to try to understand this issues involved here.

A couple points:
The parts are not densely packed. As you can read in the thread, the failures only occur with a sandwich, not with 2 pieces in exactly the same configuration. The pieces have nearly a 1/2" between them which should be more than enough for the resin to flow back in given the foot print. Remember, I am not printing solid objects but a 1mil shell and a .5mil lattice. That is not very much resin so I have trouble accepting the resin-flow explanation. Or maybe I’m misunderstanding the “can’t flow away from the part” rationale?
As for the perpendicular pieces ripping off the PDMS layer, the tear occurred very late in the build (approx 80%) with a relatively small cross-section of the model pieces. This means that the resistance to the moment forces created by the peel were actually much less than they were earlier in the print. True, perhaps the larger cross-sections helped to loosen the layer, but still…
In the meantime, I’d like to try to avoid my printer being out of commission for as short a time as possible. What shall I do about the mirror? Try to clean it? Are you sending me a new tank to replace the one that was destroyed? Action here is critical. We can work out issues with my file later…as you can read in the thread, I know that I can consistently print 2 pieces at a time without issues and would like to continue my work.

Obviously the tearing of the PDMS layer is the biggest issue, but please do not ignore the other failure problems I was having.

File is attached.

Thanks,

Ben

Comments from the peanut gallery are welcome.

How old was that tank? In months and prints?

I really hope formlabs is researching alternatives to the PDMS and are serious about replacing it.

I’ve been doing a lot of research in this arena and have found a potential candidate for it.

The key here is not to replace the PDMS for the sake of replacing it with something that doesn’t cloud. No. The key is to modify the tank in a way to make those peel forces weaker and to improve the print quality of large objects, surfaces and to lessen the chance of failures.

I’m working on a prototype and will let you know if it works well.

But as they said, printing large items with SLA is tough. That’s why most printers that are used to print large objects are the ones where the projector (or laser) is shining into a tub of resin from the top and the part is slowly raised upwards. There is no peeling necessary. Drawbacks are usually lower resolution, slower prints, and large amount of resin required.

I was telling @RocusHalbasch on another thread how I had to use a Vaseline spray to rejuvenate my PDMS once. Maybe this is something we should all do right away when the tank is new and clean to create a non-stick barrier. Thoughts Monger or Rocus?

Josh,

From what I know about PDMS, it’s a porous material. If you put any sort of vaseline on it, it will plug those tiny pores up and actually worsen it’s ability to inhibit curing by oxygenation.

I wouldn’t use anything on the PDMS.

Have you seen curing by oxygenation? That doesn’t make sense to me because I’ve never seen a skin form on the top of the resin in the tank. It’s my belief that when users scrape the PDMS good and hard what they are really doing is squeegeeing out the ingredients that have settled into the pores. I think plugging the pores with Vaseline is the real deal.

PS - I’m not sure how Discourse handles my edits, so… @RocusHalbasch again too.

I didn’t say curing by oxygenation. It’s the opposite.

What happens is when the resin is touching the PDMS, a very thin layer of the resin is not being cured because of the oxygen in the pdms. That’s what allows part to be peeled away.

I really do apologize for disagreeing twice, but that doesn’t make sense either. I don’t think the squeegee action actually dries the PDMS well enough for gasses like oxygen or nitrogen to bond. And from my experience using silicone molds, a dry mold will adhere terrible. Mold-releases like the Vaseline (petroleum jelly) are a must for releasing from silicone molds after you use them more than a time or two.

Josh,

If it works for you, I say go for it. I’m just telling you what I know and you don’t have to agree with it :wink:

@Ben_C, Did you “let the PDMS breathe” before the print that pulled your PDMS off? And how old was it?