Print distorted in one axis... due to varying thermal expansion pressure?

[jkflowers 1]

Hi everyone,

I am working with an Ultimaker S5 in an academic context. This week, I hav e been trying to print a holder for some fiber optic probes, shown below

 

 

 

The part is coming out distorted, as shown up-close below

 

 

 

You can see that the distortion occurs right above the point where support ends. I am guessing what is happening is that the part is pushed outwards above that point due to much greater thermal expansion pressure from the PLA than the breakaway. Is that correct? And if so, do you have any tricks to mitigate it? I am not too concerned by the distortion on this side of the part, but it also distorts one of the tubes on the other side of the part (more difficult to show). I think I will try some different orientations and will get it right eventually, but this problem has been more difficult to google than any of the issues I've had previously, so I thought it might be worthwhile to ask directly.

[gr5 2,168]

Could you please post your project file (or at least your STL file)?  It will contain your model and all your machine settings and your profile and the settings override for that.

 

Mostly I want to see how the part is *supposed* to be.

 

In general all distortions are usually explained by contraction.  Not expansion.  So maybe the PVA is contracting and pulling inwards at the ends?  If you have a micrometer maybe you can measure if one end is shorter than it should be or if the other end is longer than it should be.

 

It looks like you printed a raft.  Perhaps the raft is pulling things inwards.  Perhaps it would do better without a raft.  If you are having problems with things sticking to the glass bed then use the glue stick and then use a wet tissue to spread the glue around evenly and to also thin out the glue (reduce the total amount) both of which will help parts stick to the glass.

[gr5 2,168]

Also make sure you are using engineering profiles.  the default profiles are for parts that look good.  The engineering profiles are for parts that are dimensionally accurate.  I don't think changing to an engineering profile will make any difference in this specific case.

[jkflowers 1]

Hey gr5! I loved your video on adhesion.


I am attaching the project file. You are right that contraction, not expansion, is the issue. That makes far more sense when I think about it, and is confirmed by my calipers. The dimension i measured (from the front of the thinner tube to the back) should be 133.6 mm. It is 133.4 mm above the fault and 132.7 mm below it.

The raft was not the issue. This image is actually my 3rd time trying to print the part, and the raft was a new thing I was trying to maybe improve support of the bending tube. It obviously did not work.

I have never heard of engineering profiles, but I will figure out what those are and probably start using them on my next print.

And potentially of interest... here is an image of a very similar part I printed which did not have this problem. The main (though not only) difference between these two pieces is that the thinner tube extends much lower below the base in the one that failed, meaning that there is more support, including beneath the back end.

 

 

 

 

UMS5_ftir-holder-3.3mf

[gr5 2,168]

[gr5 2,168]

I'm not sure what the issue is but we are assuming the white breakaway is pulling on the bottom tips of the cylinders?  Towards the other end of the part?

 

If so it appears to be doing this by only pulling on the tip as you can see a gap - here there is a gap between the support and the part between the red circle and the teal support.

 

If this is the issue then at least getting rid of the raft will have one of the 2 cones touching the print bed.

 

[gr5 2,168]

I hate to say this but the easiest solution would be to measure the angle of tilt that you want to correct and correct for it in cad.  It's an ugly solution but will give you very good results.

 

How is the inner diameter of those probe tubes?  A more common issue in 3d printing is that vertical holes are smaller than the cad model.  Typically by around 0.4mm.

[jkflowers 1]

earlier i said i thought the pla bridge was pushing that part of the print outwards. i guess that it obviously wrong, but it seems possible that the opposite could be happening -- that maybe the parts below the bridge are pulled in by the contraction because there is nothing to support them? I don't know if that makes sense. As you say, it can't really be the breakaway pulling on things; it is not even in contact with the bent tube at all.

We do sometimes have problems getting the tolerance right on holes. I am not terribly systematic on how I deal with that, and could probably improve, but in this case, since I wanted them to be particularly snug, I actually prototyped them in a few sizes to see what fits best.

Measuring the bend and correcting for it (in Fusion 360) sounds like a pain. I will do it if nothing else works, but I think I will default to my slightly easier backup plan first, and see if it is possible to move the tubes down so that I can print this upside-down with no support at all. I was hoping this was an effect people had seen before and could easily diagnose, but I appreciate your responsiveness and help, and am a little bit star-struck anyway.
 

[gr5 2,168]

Oh don't embarrass me!  🤭

 

Thanks about the video comments.  My videos the last 2 years have improved greatly but none of the recent ones are about 3d printing unfortunately.

 

Anyway if you figure it out some day let us know.

[jkflowers 1]

On the off-chance anyone comes across this, I never ended up figuring out the cause of this, but it is clear that the bridge on this piece causes some kind of deforming force to be exerted on surrounding elements. What I ended up doing is moving things around so that the tubes are level with the bridge, as shown in the attached image, and the whole thing could be printed upside-down without support. When I printed this, it actually still came out with problems, as the bridge had somehow caused both of the tubes to shrink -- both were about .25 mm smaller than they should be, which implies that the support on my previous part was not the cause of the problem. I got around this on yet another print by setting horizontal hole expansion to 0.125.

Edited December 13, 2022 by jkflowers
adding more detail

[gr5 2,168]

In general, vertical holes (and inner diameter of vertical cylinders) always print about .4mm too small.

 

PLA is like snot when melted - it sticks to itself strongly.  Like a liquid rubber band.

 

And it is also cooling and shrinking in the first milliseconds that it leaves the nozzle.  So it is like a liquid rubber band.  While printing the circle of the inside of those cylinders it is being pulled inwards while still in the liquid "mucus" state.  The outside of cylinders and corners are also pulled inwards but not nearly so much because they are supported by existing solid filament further inwards.  The force is not very strong so it doesn't take much to hold it in place.

 

In contrast the warping caused by long sections (like when printing a pencil flat on the bed) cooling/shrinking can be very strong.