My first print (after the formlabs butterfly) was an obelisk, download from thingiverse, 1:1 print, and it came out (nearly) great!
(note: ALL of my prints are made with the suggested standard settings clear 0.5)
The exeption is the bottom: of course I used the support standard settings.
In the middle of the print, the print base, the supports, and the object all clumped together. This means there is an area of about one cm that has become “part of the model” and has to be removed the “hard way”.
I then modifed the model with some tinkercad playaround, an printed the object directly to the building platform. On this there was success, very good looking bottom of the object, and it stuck to the platform perfect.
The print itself failed, I have no idea why, as before, the same object, printed out well - and this upper part od the obelist was not modified by me.
Here some behaviour also can be seen, I think it is called “delamination”? which cant be seen at the print before…
at next, as one of my goals is to print mechanical parts, I tried this theme.
based on:
http://support.formlabs.com/entries/24479916-Miscalibration-circles-turn-into-ellipses
Marks experience, I printed the same parts:
While they -physically- look good, I experienced the same as mark:
nolt really round!
While Mark stated he even could not assemble the parts, I did a second print, and could assemble the object.
That said, it was extremely hard to assemble. The assembling was only possible because of the flexibility of the cured resin. The object moves, but you would not discribe this as moving at the moment. You need extremely force, at every move the part “disassembles” itself, and a lot of “sand” (= abrasive, ist this the correct word?) comes out of the object.
conclusion: at this stage of printing calibration, the Form1 is definitely not suitable for printing mechanical parts or parts that should have given dimensions.
I investigated more, and did a 5min session in tinkercad - building an object which has 10x10x10mm dimension.
The result is constant, but constant unsymmetrical!
To have comparable results with others, based on
http://support.formlabs.com/entries/24462731-Form1-Linearity-Calibration-Procedure-Available-
Damien B.s suggestions:
http://exploreideasdaily.wordpress.com/2013/07/12/episode-2-form1-calibration-strikes-back/
I printed the same parts, took that 100 measures, and filled out the excel form:
NOTE: I am not an expert in caliper measuring. The objects on this things are so small, depending on the “pessure” I gave to the object, etc., I received different values each time. So in my opinion, the chance of measurement errors on this test is relatively high. I then tried to do that 100 measuring points “the same way” - leveling out errors.
That said, how I interpret the excel file, my printer would have a 0.76% error in X-axis and 0.65% error in Y-axis, which sounds partly better, and partly worse than Damiens values. To be sure, maybe Damien can analyse the excel?
But it does not reach out for real items that should have a certain value.
Cylindre5x_measure_0-8-1_squarefish.xlsx
The following item should have 75mm at the whole lengh:
As the part was too big to place it horizontal on the buliding platform, I had to place it angular.
"The frame for the “window” did not really come out symmetrical, but I am happy it printed.
After some printing time, there began to appear some “vertical lines” which were not at the beginning. Well, as the surface has to be sanded, it might not be that problem.
At the last two photos, this was a fast download from tinkercad, adding company logos. Unfortunately, during print I thought that I MIGHT run out of resin, and to be sure, I paused the printer, refilled resin, and continued. The pause was no longer than 45 seconds and all was done carefully without applying force anywhere.
After resuming, at the end of the print, I realised a “visible line” through the object, at that level where the print was paused.
In general, I think the FORM 1 is a great printer. Easy to use (exept the Form Finish kit, removing the object from the build platform is the most stressful part of the whole print job), and with partly great results. Alhough, I think it is necessary that Formlabs provides some easy “self doing calibration procedure”.
When you construct a part and need it to be 75mm, and then it is 74mm or less, the part itself looks nice but can not be used for original purpose, only “tinkered together”. Hope this can be solved some time.
So is not one of the printers calibrated correctly? I haven’t seen anyone post up a perfectly round gear or object that has proper measurements when using calipers. Is Formlabs working to fix this problem. Also is this something that will be fixed with the software or is everyone who have received a printer going to need replacement parts?
Mine was too far off, and in the end I had to get it replaced. The replacement seems to print much more accurately, and when I print that gear set now it works better.
In my experience, you need to give the mechanical parts some time to cure fully, right after printing and cleaning they still have some stickyness/softness. After a while they harden fully and take on a slightly yellow color. At that point, the gears seem to work pretty well for me.
I have some other issues with the replacement machine that I am currently looking at together with Formlabs’ support team. That experience has been really great so far, I have to say, they are very helpful and friendly.
Martin, thanks for the detailed photos and measurement results. I’ll have to check my machine to see how it compares!
In defense of the Form1, these results are fairly typical of what I have seen from other (even much more expensive) 3D printer options. I have had SLS prints made (on a $300K+ machine) that were 4 to 5 inches in overall length which varied by several hundredths of an inch in certain areas. Even printers like the objet and projet series can be significantly off when you pull out the calipers. Large “flat” surfaces are notoriously difficult to create with any of the technologies that I have tried. Print a completely flat 5 inch square that is 1/8" thick and put it on a truly flat piece of granite or glass… If it does not “rock” at all then please report back and let us know what printer it was! These are hurdles that all 3D printer manufacturers are working on.
Thank you for the detailed photos, this is very helpful!
