Thicker supports should help unless they are under exposed.
Are some resins worse than others?
I am really thinking it’s the profile for the resin.
I have printed thin wall objects like this with no issues (clear 02).
Homemade stud finder with 1.7mm walls, 135mm long
[quote=“KenCitron, post:41, topic:19857”]
Thicker supports should help unless they are under exposed.
[/quote]Not necessarity, since some of the warping is taking place in parts of the model where supports may not be doing the… supporting. And it’s only in the range 1-2" above the build plates. If if were only the supports, if would continue to warp all the warp to the top.
[quote=“KenCitron, post:41, topic:19857”]
Are some resins worse than others?
[/quote] I have only tried Tough and Standard Grey. Tough seems to be words, but I have definite warping in grey as well.
[quote=“KenCitron, post:41, topic:19857”]
I am really thinking it’s the profile for the resin.
[/quote] It could be a temperature issue (resin heater not working correctly), layers being pressed (needs z-adjustment) or a host of other things.
And I’m not sure at this point if the thickness of he parts has any effect on the warping because it’s occurring in both thick and thin areas.
Have you tried a clear resin or gray, one of the basic resins?
Yes, I stated so in my last reply (and there are photos showing warped prints in Standard Grey at the start of the thread).
I see. The unevenness bottom would be from the supports where partially cured resin collects. Over the course of the print, low angles on the model were supports connect will have some distortion. That is the nature of the beast with SLA printers. The denser the supports and heavier they are the worse it can be. This problem is less evident on brand new tanks and fresh resin but will be worse as the PDMS wears.
Typically many print with least detail side for supports where cleanup is the easiest.
I stopped using IPA for part cleanup because it degrades the parts drastically. It warps them and makes them brittle. What bothers me is if the part is warped prior to cleanup as you showed on the gray box shapes. If that was the uncleaned part then there are some serious issues going on. I am not that familiar with the F2 peel mechanism but never experienced that with my F1+.
Looking at the prints you have you shouldn’t need that many supports. A print like yours I would have half the number as you show. Is it possible to have too many? would that cause issues other than use up a liter of resin and make cleanup a 12 day job?
[quote=“KenCitron, post:45, topic:19857”]
I stopped using IPA for part cleanup because it degrades the parts drastically.
[/quote] IPA has not degraded the parts I’ve printed at all. The warping is occurring during the print, possibly due to the resin being too hot, therefore staying soft and easily affected by the peeling force.
1 Like
I don’t know what is your cleaning method with IPA but I have never encountered any issues that I could trace back to the IPA bath. It does make the parts more flexible if you leave them in for 40+ mins but I have never had the need to leave them in for more than 20min.
As for too many supports I think it’s a good habit to get into when troubleshooting in order to rule out the lack of supports, which is probably what @JoatrashFX did. I don’t see how too many supports would be detrimental to the print, except for the marks left on the model.
That is interesting! I always assumed the Z offset value was only effective for the raft, but the fact that they are recommending to change the value would indicate that is has an effect on the whole print ? This would definitely affect the force applied to the print during the horizontal part of the peel operation.
If it is happening during the print process then the resin is under cured especially the supports.
Running the exposure at the minimum will extend the live of the tank but too low you will have problems while running the exposure too high will have a strong part at print but will degrade the PDMS faster and if too high the part will be difficult to remove from the build platform. As I mentioned before I think the exposure is too low.
Was there a recent update to PreForm?
Did you have any problems with prior versions of PreForm?
If the IPA is clean, 15 minutes has been more than enough for me.
Formlabs themselves wanted to add more supports as well. I usually stick to the autogen, then add them around strategic points.
[quote=“JohnHue, post:47, topic:19857”]
the fact that they are recommending to change the value would indicate that is has an effect on the whole print ? This would definitely affect the force applied to the print during the horizontal part of the peel operation.
[/quote] The thinking was that the pressure from the plate is pushing the print slightly with each layer. I don’t know if raising by 0.1mm did much though, as I’m still seeing warping. Heightening the print several cm above the plate did more. (Right now I’m running another test where I’ve raised the model even more.)
[quote=“KenCitron, post:48, topic:19857”]
Running the exposure at the minimum will extend the live of the tank but too low you will have problems
[/quote] These things are handled automatically by the Form 2. (I’m guessing F1/F1+ were less automated?)
Anyway, if you have read the thread you will have seen several people mention resin temperatures during printing rising to 45 degrees Celsius. The Tough resin (which is the one taking up most of the discussion) will begin to go soft at a little over tha over timet, even after curing, so the heat from the resin building up over several hours seems to be allowing the peel process to deform the model, cured or not.
[quote=“KenCitron, post:48, topic:19857”]
Did you have any problems with prior versions of PreForm?
[/quote] Yes.
Another small update.
Here’s the result of a height test I did yesterday (0.1mm Z shift). This model was printed in the orientation shown in the photograph, but I added some trash geometry (cut off in the photo) so that I could raise the model almost to the limit of the build volume. The warping is still there- you can see a definite “banana bend” where it sags in the middle. So the warping seems to occur over the entire build volume, not just near the build plate. It IS however, a lot stronger near the build plate.
I am currently printing a test using 0.2mm Z-adjustment as per Formlabs recommendations. The same model at a slightly steeper angle, but closer to the build plate. It also has 30-40% more supports overall. Over the last few prints, it looks like the temperature has stayed under 40 degrees C.
1 Like
that level of warping is not unusual given the geometry of the part, and the flexibility of Tough resin.
The thing about shrinkage is the difference between masses of volume.
Imagine a dumbell. Two spheres connected by a narrow bar. If its cast or printed solid it has three competing stresses acting on its shape. There is the overall shrinkage along the length of the barbell entire that wants to pull the two ends together.
But then there is the masses of the spheres on either end that are trying to shrink toward their own centers of mass. These two additional stresses are PULLING on either end of the narrow neck, even as the narrow neck is trying to pull the twp spheres toward each other. The result, in, say metal casting, would be a tear somewhere along the narrow bar- almost every time confined to the juncture of bar and ball where the mass suddenly changes.
even the most uniform part is going to warp somewhat in the SLA process. And what I see in this part looks perfectly within what I expect from ANY SLA printer- even the quarter million dollar ones.
However- the pronounced warping you are seeing at a specific height is excessive and more importantly, obviously hardware related.
If there was shrinkage then printing a solid cube the middles would sink in. I don’t remember seeing that with a 3d printed part. That is typically an issue when heat is involved such as injection molding, metal castings etc. Also known as suckback.
Think John Hue might be spot on with the temperature setting on this causing warping. Possible to turn the heater off for that resin?
Perhaps, but for a material that costs almost $300 per liter (here in Europe), that is marketed at engineers needing “functional parts and assemblies”, I expect a lot more in terms of precision.
I have verified that the warping occurs over the entire volume- it’s just (of course) much more pronounced closer to the build plate.
Sadly, no. It might be possible in “open mode” but that throws out a lot of other functions, so isn’t an option.
The irony is that with one particular model of those I’m testing (the “tube”), my Ultimaker prints it virtually flawlessly, with a surface finish that is actually just as good as, if not better, than the Form 2 when taking into account the support material on the latter. And it took maybe 2-3 test prints to get there, at a fraction of the material cost. To be fair, I did have to split it into three (seamless once assembled) parts on the Ultimaker, but the two machines are not supposed to be on even footing on a model like this. Today I actually did a test and printed the split setup on the F2. There was still warping, though not a lot, but all the supports meant quite a bit of cleanup. Still the best result so far on the Form 2 with the model, which is frustrating. The Form 2 did however win time-wise. (The UM2 took about 2/3 longer.)
1 Like
Your expectations may be unrealistic.
I don’t know what you are comparing their prints to… but as SLA printing technology goes- the Form 2 compares very favorably to machines running a quarter million dollars.
I have made parts by every means imaginable. Short of milling stable solid objects to spec directly, there is NO means of making repeated parts that does not entail some level of warping or distortion that must be compensated for in design. And when I have something like nylon milled, I still have to compensate for the frictional heat of cutting the nylon causing distortions from spec.
Themoplastics are easier to get to higher tolerances than any printing technology because of the precision of the tooling, and the fact that the material is fully homogenous- all liquid at once and all cooling together, so that the distortions are more predictable and so the tooling can be adjusted to compensate reliably.
From long experience with SLA I am pretty good at compensating for the distortion in the From 2 output in design, even though I only go to that effort when the the part is dimension critical.
But I do think there is something physically wrong with your machine, because neither I nor any other poster here has seen similar exaggerated warping across a narrow height band, across the entire platform, regardless of part orientation.
Whatever is causing that might also be responsible for a significant amount of the less severe warping you are seeing.
For example… the warping I get can be identified to be related to part geometry ( as in the dumbel example above ) In which the warping is the same regardless of part orientation, Or it is due to how the part is oriented and supported- for example I printed two prints the other day that both had a flaw where a small section of the print was thicker and formed an angled facet on the print… examining it, I could tell that the portion that was off had been initially laid down as a very small, Isolated minima that was supported by only one or two supports… this allowed that supported minima to be shifted on peel, and the next layer then filled in and got similarly shifted- as the part built the isolated area got larger, with more supports resisting the drag of peel and so the isolated spot moved less and less, and by the time the minima was connected to the rest of the print body, it was robustly enough supported to match up with the correct contours on the rest of the print.
that flaw was just an oversight on my part because I know better than to orient a part in that way.
experiment with design features and orientations that will minimize what warpage you can… but that strong curl at 1-2" Z is something else.
You may want to just send the file out to 3d Hubs and have someone bang it out on a polyjet rather than pull your hair out.
Seems ridiculous that there are these kinds of issues with the F2 that are not solved. They market the machine for prototyping but what is the point if the parts come out like that?
I don’t remember having those kinds of problems with my F1+
Since Formlabs are specifically marketing the F2, and the Tough resin, as “the ideal choice for prototyping strong, functional parts and assemblies” I am least expecting the parts to be straight when they are supported, do not contain any extreme geometry, such as long, thin rods with no connections.
I have been using prints as a complement to scratch-built and machined movie props and model parts since around 2007, mainly through vendors, with the big Stratasys and 3D Systems machines. I can’t recall the exact materials I tried on those machines (it’s been a few years) but there were a couple of different wax-supported plastics that I had varying results with. Some of them were simply useless, others were a little more stable. (I think Shapeways still offers one of the better ones in their “ultra detail” options. It’s stable and accurate, but brittle as hell.) I’ve also done a lot things in SLS nylon, which I eventually grey to hate because it’s almost unworkable. Eventually (around 2014) I got a Ultimaker and completely stopped using services except for when I needed extreme detail parts, or metal. Overall, FDM has provided the best results, though the post-processing can be a royal pain.
Certainly, and that even applies to old-school methods like resin-casting as well. But it’s hard to compensate for a problem like this, unless I somehow devise a method of accurately bending the entire model in the opposite direction before exporting to STL!
That’s what what worries me, because Formlabs are beginning to lean towards it being “machine limitations”.
Sure, and I get the same printing in FDM, designing for overhangs, no supports and so on. You learn to see it- “That part of the model isn’t going to print right, so I need to change something”. But compensating for “the machine is going to warp the entire part no matter what I do” is something else.
I used to do that, but you have to wait a long time and it gets very expensive after a while. I also like being self-sustaining. For the last few years, I have had very little need for outside services. The Ultimaker has been a solid work-horse and there have been VERY few times when I have been limited by it. (Usually, when I showed folks what I had done with it, they were very surprised.) The Form 2 was supposed to make things easier.
One last thought, did you ever try printing with the cup side down?
I would try rotating the model over so enough you don’t need internal supports 45degrees or so and long edge to the tank.
I would manually place the supports and put in just enough to eliminate the red warnings in PreForm. Also shorter supports would be more stable.
Give that a try before jobbing it out.
If you have to send it out then a polyjet is your best bet.
I used to use various companies to do my 3d printing and majority of the time they would be late shipping and F@#$ up the print. It is nice to be independent.
I’ve printed it in just about every possible orientation. Ironically, the one that has yielded the best results is one that goes against Formlabs’ recommendations! Printing it “standing” straight up in an 90 degree angle yielded the least amount of overall (visible) warping, probably due to the smaller footprint. (The little “ribs” on the top did however, not print entirely straight.)
Oh, I wouldn’t send this out anyway. It already prints perfectly on my old Ultimaker 2 . (The cleanup is actually easier on the UM2 since I can print with less supports.) But I wanted the added durability and weight of the Tough resin, compared to PLA. (At any rate, I need the machine to be able to print this kind of part, or it will be useless in the future.)
2 Likes
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.