DIMENSION PROBLEMS- Help

[KenW]

Hi All-
I am printing on a Saturn with Siraya fast and directly on a magnetic build plate.
My prints are all over the place in dimensions.  As much as 2mm smaller in the X axis and at least 3mm 
in the Y axis.  
My settings are 0.1 layer ht   bottom count 5  Exposure 2.5 and  bottom exposure 40s.
The part sticks to the plate no problem, and bending the magnetic plate, they pop off 
but when I measure, it is way off.
Any help would be appreciated.
Ken

[cliffyk]

What you are experiencing seems to be a commonly reported issue with magnetic build plates. How did the printer perform before you installed it? Have you tried printing without it?

[KenW]

What you are experiencing seems to be a commonly reported issue with magnetic build plates. How did the printer perform before you installed it? Have you tried printing without it?

I never printed directly on the build plate, but don't recall any dimension issues before.

[cliffyk]

The obvious thing to do is to return the printer to the it's "as built" specifications (I.e. remove the aftermarket build plate and any other modifications) and see how it performs...

[KenW]

The obvious thing to do is to return the printer to the it's "as built" specifications (I.e. remove the aftermarket build plate and any other modifications) and see how it performs...

Hi Cliff, I just ordered another plain build plate, as I don't want to rip off the magnet yet.
I also did an experiment by printing 2" squares on the plate and varying some of the settings, like exposure time etc. I found that there were some interesting changes happening.  The squares were supposed to be 50.8x50.8mm  and I was getting some at 51.19, in the first tests, but the last one was 50.8x50.8.   So I now have to duplicate those settings on a larger, or more complicated shape and see what happens.

[cliffyk]

Exposure time can have great impact on dimensions as what we are doing here is producing partially cured (hence the need for final post-printing curing) stacks of tiny mushy blocks, with exposure time being the most controllable factor in how mushy they are. and how well they stick together..

With your observed error of just +0.77% (51.19/50.8) it won't take much to alter the final dimension.

The 10 micron layer height also add a "twist" as it means there are 5X as many layers (as compared to the 50 micron default for most SLA printing)--thus more opportunity for cumulative error. 10 microns is 0.00039", do you really need that precision?

50 micron = 0.00197"

[KenW]

Exposure time can have great impact on dimensions as what we are doing here is producing partially cured (hence the need for final post-printing curing) stacks of tiny mushy blocks, with exposure time being the most controllable factor in how mushy they are. and how well they stick together..

With your observed error of just +0.77% (51.19/50.8) it won't take much to alter the final dimension.

The 10 micron layer height also add a "twist" as it means there are 5X as many layers (as compared to the 50 micron default for most SLA printing)--thus more opportunity for cumulative error. 10 microns is 0.00039", do you really need that precision?

50 micron = 0.00197"

Am I confused then as the usual layer height is .05 and I used .1   so isn't the .05 smaller than .1?  I'm not using .01.

---

So to clarify:  Chitubox layers in mm are usually .05 (.00196")  And I chose to use .1 (.00393")

[cliffyk]

Am I confused then as the usual layer height is .05 and I used .1   so isn't the .05 smaller than .1?  I'm not using .01.

---

So to clarify:  Chitubox layers in mm are usually .05 (.00196")  And I chose to use .1 (.00393")

I apologize for the confusion--I misread your first post as stating you were using .01 (10 micron) layers, when in fact you wrote "0.1 mm" (100 micron); 2X the default 0.05 mm default. Layer height will also affect the final dimensions of our
"mushy little blocks".

[KenW]

So, in my experiments, I took a small part measuring 50.8 x 15.3 x 8.36 high. The print came out 48.34 x 14.82 x 8.09 high (all mm).
Then I calculated the percentage the part is smaller - 4.84% x 3.137% x 3.229%.
I then upscaled the model by those percentages and printed again. I got 50.57 x 15.3 x 8.47. Much closer now.
So is there an average I need to use, or does each axis need its own percentage adjustment to get to my actual size?

[cliffyk]

That is exactly how I do it...

However the needed correction will be affected by overall volume and relative X, Y, Z dimensions as the "tiny mushy bricks" pile up--any "average" correction will be just a good place to start based on experience and what your gut tells you.

Keep in mind that while the Z-axis resolution for your Saturn is variable in 0.00125 mm (1.25 micron) increments via layer height, X & Y maximum resolution is fixed by the LCD pixel size (0.05 mm, 50 microns).

Actual print dimensions (even if 100% accurate) will always be multiples of those maximum resolutions.

Keep in touch and let us know how things are going...