[Return to main site] 

Z-axis way off for me


#1

Hello everyone,
I am seeing a large problem with the Z axis. It appears to be off by quite a bit. I just printed a part that is supposed to be .260" vertically. It turned out only .158?? This is WAY off for such a small distance. I must be doing something wrong. I am printing with the orange ABS. MY settings were 20 second exposure 7mm lift at 15mm/s.

Is there something I should be looking for here?


#2

Can you post a screenshot of the Kudo3D host settings page and the Creation Workshop slicing profile? You should check to see that the Z layer height is exactly the same between the two.

Also, what XY resolution are you printing at?


#3

I don’t know if this helps, but i found re-calibrating my z-zero to not be as tight to the floor of the vat helped me a lot.


#4

Thanks for the responses. I just double checked the slice settings vs the kudo software settings for layer height and they are both set at 0.075mm. I will also redo my zero to see if that helps.

James, I read in a post when I was searching that you mention something called z-axis compression, but I couldn’t find what that meant. What is that, and could I be experiencing that?

PS I just wanted to take a second to thank you James for taking all the time you have to post your videos and help on this forum. Without all of that help before getting the printer I would have been lost. It is because of you that my first prints have been successful.


#5

Z-axis compression is just a short way of describing the symptom of the modeling being squished in the Z dimension somehow.

There can be several different causes.

What I recall seeing described thus far is:

  • the Z dimension being too large or too small from the model size due to a mismatch between CW slicing height and the host software's layer height
<li> the XY dimension being off (due to XY not being calibrated), resulting in the part looking like it's too short or too tall, but in reality it's due to the XY being too wide or too narrow.</li>

Measuring the part would differentiate these two cases.

I think there may have been one oddball case where there may have been Z rounding issues due to a layer height setting that exceeded the resolution of the Z-axis linear stage. The spec sheet for the Titan 1 indicates a maximum Z resolution of 5 microns, which is .005mm. Thus while .075 should be within the resolution limit, settings where the 3rd digit is not a multiple of 5 will exceed the limit. (E.g. .072 is bad, .070 is fine, .087 is bad, .085 is fine).


#6

I’m experiencing something very similar. I’m printing Marvin by 3DHubs as a reference model but it comes out with a strange elongation (is not linear, it seams that “accelerate” towards Z Up). I’ve made a lot of prints to test if the cause could be the orientation, hollow the model, zero level, the type of resin, projector focus or XY Calibration but without success. The CW’s parameters are the same as in Kudo SW.
In the pic attached you can see clearly the earmuffs that are ovalized, they should be exactly circular (9mm diameter) instead present a dimension of 8.8x9.6mm (YxZ).
I’ve used MKJ G+, ABS, Spot-HT and printed at 50micon X/Y and 0.05mm Z.

Someone have an Idea of a possible cause/solution? Thanks in adevance.


#7

Hmm… it’s hard to figure out exactly where the problem is occurring on the rotated part, especially if your feeling is that the elongation increases as the Z height increases.

Could you try printing a different part like:

And do so flat, w/o rotation?

Use slow lifts (like 10-25mm/min) and make a note of which overhang feature (the square in the middle of the part or the rectangle at the bottom) is X and which is Y.

Since this part is stepped, we can measure each step to see if it’s truly 5mm tall, and if there is any drift, we can see where that starts.


#8

Thanks James for your reply. I’ve downloaded the test object and I’ll post the result of the print with some misures. I’ll use the spot-ht resin with the same csv parameters I’ve set for the other prints to have a fix point for comparing.


#9

First attempt failed! The piece came off the plate but something is possible to see. I’ll do another print this evening.
The first layer of cube is instead a parallelepiped as you can see in the pics. The orientation on the picture of the object is real. I’ll remake the print hoping it will not drop again.
I haven’t with me the cvs but the parameters are, resin Spot-ht, 50micron X/Y, 0.05mm Z, layer 1-2 60sec expos 7mm lift 10mm/sec speed, layer 3-588 7.2sec expos 4mm lift 15mm/sec speed.

EDIT: The pics are too big, I’ll repost them later.


#10

Here are the pics of the first print … but I’ve realized I made some mistake when sliced the object. The piece has X/Y dimension of 12.5mmx12.5mm instead of 25mmx25mm exactly the half … I don’t understand where I’ve modified the CW parameters since I’ve used the same as Marvin Object, mah!
I’ll remake the slicing and the printing.


#11

Hi All, here are two prints of the calibration cube, one with MKJ G+ red and one with Spot-HT (without added pigment).
The object are not “calibrated”, the dimensions/misures are not constant (5mmx5mmx5mm).
Here are some pics …


#12

… and here are the csv parameters I’ve used (X/Y 50micron, Z 50micron) and the two tables rapresenting the misures.
The axis x,y are respectively: the side with the rectangle hole, the side with square hole. Z is the height.

There is a difference caused by the resin, but the overall result is that the dimensions are not exact …

Any idea to solve it? Thanks in advance.


#13

The dimensions are close enough that the problem is most likely something that you need to deal with via exposure and pigment settings rather than there being something drastically off with your Z layer height.

The ideal values would be 5, 10, 15, 20, 25mm for each step in each dimension, and you seem to be between -0.6 and +0.6mm in XY, and consistently small in Z.

This is what I would expect with the combination of resin shrinkage and light bleed.

Light bleed for this part would primarily affect XY, but not Z since the part has no overhangs. This behavior seems to be reflected in your results where we see that in Z, every dimension is smaller than nominal, never bigger.

I have not printed with Spot-HT, but G+ has some peculiar shrink characteristics and also exhibits more light bleed (if you don’t add pigments) which can make it difficult for you to get exact dimensions.

There’s been a thread on this subject that would be worth reading at:

http://www.kudo3d.com/forums/topic/outside-dimensions-perfect-inside-dimensions-are-off/#post-15648

From the different resins I’ve tried, 3DM-ABS’s dimensional accuracy is noticeably better than anything else. PR48 is hopeful, but I haven’t done extensive testing. Both MJ G+ and FTD IB have their quirks (which are quite different from each other), and I’ve not been able to do better than around +/- 0.3mm with those (using extra pigments and exposure changes).


#14

Thanks a lot James, I was worried about was something in the printer since I’ve used 3 kind of resin and a lot of different exposure parameters.
As a double check, since I’ve also the 3DM ABS, I’ll print the calibration cube with that resin and verify if the dimensions will be more accurate.

Do you think that the CW Build Size Adjust parameters could be of some help?

I’ve read the thread you repost, it’s not quite clear how is made the donut/disk calibration object but it’s clear that to print exact dimensions you need to make a recursive process of calibration. I’ll study better the thread and make some test.


#15

The build size adjustment parameters will adjust your XY sizing. However, you need to be careful that since resin shrinks, if you dial this parameter such that it incorporates the shrinkage (as opposed to measuring the optical projection of the grid), then your slice size settings will become specific to a particular combination of resin and pigment.


#16

I just wanted to follow up to my original post to let people know what my problem was in case it can help someone else. After printing some larger models, I found that my overall z height was spot on. However, certain features we way off. For example, I have a recessed cut in one of my part walls that is .300" tall (z axis). On the print I was getting around .270". Being .030" off was horrible. After testing print settings to get my wall thicknesses correct, and realizing that my exposure time was way too high (I went from 12 seconds to 1.2 seconds using 3dm-ABS) I finally figured out what was happening. The over-exposing was causing extra layers to cure until there was enough distance from the bottom to stop curing unwanted areas.

In a nut shell, when you are printing an overhang area, if you over expose you will drastically mess up your z measurements for that area of the feature.


#17

Hi, I got similar kind of issue. print part appeared as stretched height (Z-axis). I use 50um resolution for XY and layer thickness as 0.02mm. See the print picture.

Can any one help me by suggesting correct procedure of setting-up parameters in CW before slicing if that is an issue?