Cross test finished last night and to me the results look more than acceptable. There are some minor visible surface ripples, but to the touch the parts feel perfectly smooth. White made it hard to photo, but I did my best. Note that there were some little blobs on the surface that washed away in the IPA bath.
Though yours do look better than mine, I still find them a little rough… What layer thickness did you prrint them? It should be smooth all over. I know mine has printed similar objects without any visible surface ripples before. Still, your quaility is much better than mine right now, so if it’s acceptable to you, then that’s great
I’m guessing from the presence of the ripples those where printed at 25um. Am I right? Anyway if that is the case they should come out great at 50um or 100um. However if that wasn’t at 25um that is not so great.
As for the ripples on the hinge side. I’ve had minor ripples at 25um on all but one of my old Form1’s. As sad and unnecessary as I think they are, I don’t think Formlabs is trying to get the printer good enough to never have minor ripples, though I believe it’s not far out of their reach. Also they can be masked or eliminated with orientation.
However that being said, all the little bumps and the severity of the ripples are worse than I would expect out of an old Form1, not a lot but noticeably.
@Alex_Vermeer, I do agree that ideally prints would come out perfect. Also keep in mind these were with minimal cleaning… I’m still getting my technique down. I don’t have the flaking and excessive gel-ing that I’ve seen some with on here, so that’s what I was basing my acceptabliity on. Naturally a perfect print would be preferred
@RocusHalbasch, Yes this was at 25um. I agree with you all that there really shouldn’t be ripples and Formlabs should be trying to fix this. If orientation can eliminate the effects that is great, but I’m sure at some point I will have a large enough part that doesn’t have as much wiggle room. I am printing the black helix at 25um now!
@KevinHolmes, Unfortunately I cannot post pictures of my design, as my customer is still awaiting her patent pending status. Like I mentioned, the shape is essentially a +/- 3" OD hemisphere with 0.125" walls. I am going to try it next with the flat side facing the platform… Preform kept orienting it with the sphere facing the platform so I think this may be part of the problem. I am also experimenting with techniques, so I may have made the support density a bit too lean…
@Annino is this hemisphere open, like a bowl, or closed like a solid sphere cut in half? If it’s a bowl do both the outside and inside of the bowl need to be a clean surface? If not does either the the outside or inside need to be an especially clean surface? Can it be done in multiple parts and assembled or must it be one part?
That is generally a bad idea. This is why I suggest people learn to orient their own parts, Preform has a tendency to make stupid mistakes. You are lucky the walls a so thick otherwise you would probably have gotten holes in the part from blowouts. Also the lowest point of the sphere when printed in this orientation will suffer from much of the same problems as a plane parallel to the print surface, that is to say, in a nutshell, it will require a lot of supports and come out awful. The reason for all of this has to do with the force of removing the cured layers and how strongly they adhere to the PDMS, and also the equalization of pressure during the peel and return when there are closed off sections at some point in the print.
Bowl like hemispheres are a tricky problem for the Form1+ and how to go about producing it depends on what sacrifices you are willing to make, and what features you value most. There are many ways to do it. On a well calibrated printer I would prefer the following method.
Orient the hemisphere so the flat side is down. Next split it in half and separate the halves by several millimeters. Now print it just like that with no further orientation and no supports. When this is done printing the parts will have no marks from supports, but you might want to sand the edge where the two halves will mate to make it extra flat for cleanly mating. The last step would be to assemble the two halves and “solder” them together with some extra resin and a 405nm laser pointer. I like this method as it gives a print with no surface problems from supports, however it does have a seam. Anyway if you would like other suggestions let me know.
Whoa if it didn’t stick it will probably not hold this time either. First lower your platform by increasing the platform height value a little (0.1mm or 0.2mm). The greater the number the lower the platform will be when the print starts.
It is open like a bowl. There are some features protruding from it so it cannot lie flat on the build platform. I do like the idea of splitting it in half though. I could split and orient the surfaces that will become the seam against the build platform. Ideally the entire part should be free of defects. I do not mind cleaning up support nubs tho, as long as the surface underneath them look acceptable. I haven’t attempted fusing parts yet, but I did get a 405nm laser pointer for this purpose. My logic tells me that when you do this method you only create an outside layer that is fused, correct? Would you have to work a bevel into your part edge and build up the seam like a weld to get something strong?
I think I will try one at 50% scale in the new orientation to see if that helps. If not then I may have to try splitting it in two. I wish I could just post my file, that would make things a lot easier!!
That is correct, and that is a good strategy. One other problem you might find with printing with no supports is sometimes Preform over-cures the part of you object closest to the platform resulting in a slight ledge, if you try it you will see.
So attempt #2 was indeed successful. I was a bit worried because for the first ~100 layers the peel motor sounded like it was working really hard… almost stalling like an impact wrench under load… then it started sounding normal but I figured that was because I lost the print.
I’ll let the pictures speak for themselves
Black @ 25um, centered in the build platform, no tweaks to the .form file.
@Annino oh my word. That is way better than I expected, there is no waviness at all. That is way better than when Formlabs tried it. That is exactly what I want my printer to do, well minus the unnecessary 5mm of over-cure. I can not easily express my awe and jealousy. @Sam_Jacoby and @Stephen that’s all I wanted my printer to do right there. Ok, that is enough of the theatrics, now back to socking gift horses in the mouth. Do you have a pair of calipers and if so could you take a bunch of quick measurements of the outer wall thickness around the top edge to see if the wall is thicker in the back or the front, and also the a bunch of the diameter to see if it is close to round or slightly wider from front to back? In case that seems like a joke from the rest of my comment I am actually serious. I have had problems getting round objects to come out round, meawhile that print came out so nice I think it might be pretty close. If it is I want to know because I would like to know what a well calibrated printer is actually capable of.
@RocusHalbasch, yeah this one really caught me by surprise, especially after the failure. It really is just shy of being perfect.
Your request for measurements doesn’t seem like a joke at all. For a test piece it actually makes perfectly good sense! What is curious is that the “ledge” of overcure is actually to spec… 21+/- 0.05mm. The rest of the body all the way to the end ranged from 20.85-20.95 circumferentially with a few outliers a few hundredths greater or lower.
@EvanFoss, Yeah I’m happy with this one All these laser threads got me nervous but now I’m feeling a bit better!
@Annino Those measurements sound good to me, though a little odd that only the overcured part is the right size. That gives me hope they can get my printer up to good standards. Thank you so much for doing these tests, and sharing. We really needed some info on when things go right to balance out the view of the whole picture.
Did the 50% scale test with the flat parallel to the platform and it came out PERFECT, albeit with visible layers since I did it at 100um. I ran one half at full scale and it also came out very very nice. Since all the supports touch that flat surface, there is a bit of cleanup and filing required, but that is a small sacrifice I am willing to make to get the rest of the part so perfect.
One thing I did notice is that the support base takes up nearly all the build platform… and thus the force required for the peel is very high. So high, that you can visibly see the build platform deflecting down at the front right corner with the resin tank during a peel, and the peel motor working under significant stress. Everything goes back to normal after the base is complete, but I was surprised to see so much deflection.