Can anybody recommend the best way to print cylindrical items that need good accuracy? Obviously everyone seems to suggest placing parts angled, but this just seems to ruin the ability of the printer to print cylindrical items that need to come out cylindrical.
Is it possible to get parts with flat bases to stick straight to the bed without needing support to improve the surface finish of the bottom face? I have a sample Formlabs Rook part here which does not have any support marks underneath… are they just stuck straight to the bed?
See the part I am trying to print below, any advice much appreciated!
The three big reasons you should print at an angle is to minimize peel force, prevent blowouts, and avoid pooling and undesired curing of resin on the top (bottom?) of a part during printing.
Just taking a glance at your part I would guess that you will only have issues with the last of those issues.
The FL rook sample I believe is printed on supports and sanded. (I would look at mine to confirm but my fiancee accidentally destroyed it.) I print a good amount of tight tolerance cylindrical parts for work and I get the best results from the orientation shown. I still print with supports on (barely adds any time and I know it works) so I cannot comment on how well printing direct on the platform would go. The face nearest the build plate is always pretty terrible looking but some quick sanding fixes that.
You might not even have issues with unwanted curing of the trapped resin using black, the darkest resin I use is “Tough”. I believe this is why black provides the best fine detail of the FL resins, limited bleed.
I am confident you will get a part you are happy with if you keep orientation as shown but add supports and sand the bottom face. If you decide to print directly on the platform be sure to let us know how it goes. A perfectly flat base IS appealing.
We design components and housings for motion picture lenses, so basically everything we do is a cylinder or circular. I’ve gotten the best finish with a few different grades of sand paper, a light bead blast (not possible for everyone, but useful), and a quick acetone wipe. The acetone, by itself, is pretty effective at hiding layer marks and light sanding, and it doesn’t seem to have a significant effect on the dimensional accuracy of the part. We have actually used parts from our Form1+ in trade shows, having only wiped them with acetone.
I get ±.005 - ±.001 depending on the location of the part on the build plate. This has been good enough for some jobs, not for others. When in doubt print a bit large and post process down to size (kinda defeats the point of a high precision printer but gets the job done).
Following Formlabs guide for finishing transparent parts should also take care of your sanding marks on black. Acetone does nothing for me, I seem to recall reading that mineral spirits clear parts up nicely. Most of my parts are for engineering use only so I check dimensions and use as is. I haven’t had much experience making “pretty” parts.
What I have had success with is painting resin over top of your sand marks and curing that extra layer. As a bonus you can print pockets for inserts and fully encapsulate them after installing. This method does add thickness so isn’t great when precision is required. (it works for me because most of my important features are internal and this just helps me visualize them. I have thought about cutting the resin with something to lower viscosity and allow for finer finishing but have never found the time to investigate. (Darn real work getting in the way!)
Post scaling and only in dead center of the plate. It’s not as romantic as it sounds. I generally print oversized and bring into spec in post processing. Really buried internal features are just used as starting points. But in general this isn’t an issue as we try not to design very difficult to manufacture parts.
±.001 is definitely best case, not normal. Sorry for any confusion.
I bought my form1 primarily for cylindrical parts (think of a round tv remote control), but have been sorely disappointed. I use it for gross testing, then send the files out to a third-party supplier for final fit and finish testing.
The biggest issue is the strong support point required in the peeling process. One side is always so ugly it is very difficult to sand clean; either the inside is nearly perfect and the outside is grossly pocked with supports (requiring HOURS of fixing to make it right), or the outside is nice, and the inside has unusable features.
…“professional” sla printed parts do NOT have this issue. Their support points are so tiny, I had to use a microscope to see them. One part came with some supports still attached, and I thought it was dust–they were that small.
Typically, on a 1.25" x 4" long part, my typical variations are .010" to .020". Good enough for many things, not good enough for final test.
My machine is the Form 1+, and I had the same results with the original Form 1.
Been there, done that. Even with the fewest number of the smallest possible supports, the cleanup is slow and tedious, if you are trying to get something cosmetically or dimensionally accurate. That is, IF the machine can even build the full part using those settings.
The reality is that the Form1 is (probably) fine for artistic items, but I do not make those types of part. My final builds (prototypes for an injection molded part) are done with a professional SLA printer due to the precision and accuracy…and lack thereof within my Form1+. My Form1+ (and previous Form1) are simply not good enough for all of my work.
I’m not saying I do not use it, and do not appreciate it. I’m just saying a $3k machine does not put out something good enough for my needs. Round holes are not printed round (requiring cleanup), supports leave huge blemishes (even at their smallest, they require very tedious cleanup), and the dimensions of the finished part are never spot-on. Even steps between layers at the finest resolution leave inconsistent blemishes that look like waves.
The worst part is that it is not fully predictable. I can’t just increase the scaling in one direction, because one from day to the next, it is not consistent. Maybe one day I built parts at 70F, and the next 76F, and the need for a heater is now demonstrated with the release of a Form2, but that recommendation was not in my instruction manual.
And no, I do not believe there is a “problem” with my Form1±-I believe these are merely limitations of the machine. Although I hoped it could put out top quality prints, I was never confident a $3k machine could match the quality of one 20 to 50 times more expensive.
I have all the same issues trying to make accurate engineering type parts on a Form1+, and generally I have a lot of success. One thing I do however is to go the opposite way with supports. Set the contact point to minimum 0.4mm, and then have a lot of them. This results in less sagging between points, and easier detachment of the support. Any remaining points are small enough to be knocked off with a scraper or even fingernail giving an overall better result.
I completely agree with this statement. As mentioned previously in this thread I can usually achieve utterly phenomenal results for a consumer printer. But when I do want to fine tune it seems to be an exercise in futility.
I love our Form1+ and it has paid for itself over and over based on simple ROI but it does have some gremlins that make it unacceptable for certain applications.