Surface defects looking like seams

I have printed six identical parts on the same build plate. All parts have the same surface defects at the same position. Material is Bambu PLA Basic, printed on textured PEI plate with chamber temperature at around 35°C. This defects only occur on the inner surface of the bore.

Slicer settings are bambu “0.20mm Strength @BBL X1C” with the following changes:

  • top shell layers 10 (instead of 5)
  • bottom shell layers 10 (instead of 3)
  • order of wall: outer first

printer firmware version is 01.08.00.00 (latest)
bambu studio software version is 1.9.2.57 (latest)

Position 1:

Position 2 and 3:

Position 4:

As you can see, these defects do not occur on the top and bottom shell layers (first ten and upper ten layers are not affected), which is weird since all layers there are “outer wall”.

Any ideas?

edit corrected messed up photos

Reposition the seams?perhaps the seams are better implemented by Bambu studio elsewhere.
Test changes on one part rather than 6?

Deleted to not be misleading…

No to the two of you, that would be to easy :wink:

These are NOT seams. The seams are manually positioned. I just edited my entry-post, there you can see the position of the seams.
These are surface defects that LOOK like seams. It is also very interesting that these are arranged symmetrical

Here another picture: some seams are in the small horizontal bores, some on the selected edge. But none are on the big bore.

That is indeed a mysterious phenomena.

If I may suggest the following troubleshooting steps.

  1. Cut the model in the slicer but leave it on the same spot in the plate. What we are looking for is if the model itself has the defect or if the travel of the neighboring models are influencing the other models.
  • If the pattern disappears, then try changing the print sequence from the default by-layer to by-object. You will have to provide much, much greater spacing as the slicer will now print one object and then move to the next, in order to prevent collision, it will require a lot of space as shown by the circles that will appear when you select this mode.

  1. Try taking one model and rotate it 45 degrees. If the print defect moves with the model, it is an indication that the model is defective or has some mathematical artifact induced by the slicer. If that is the case then you can try the following:
  • Reduce the number of facets by right-clicking the model and select simplify. This will be at the cost of resolution so use it sparingly. However, if there is a mathematical anomaly in your model, this could break the coincidental interference caused the the slicer being confused by certain layers.
  • Alternatively, if you have access to the CAD file, you can export it as a STEP file rather than an STL and see if that too changes the behavior. If it isn’t available as a CAD or STEP file, try this online conversion tool and convert the STL there. This is the one I used, they all suck but this one sucks the least. STL to STEP Converter Online - AnyConv
  1. Have you tried scaling the model vertically and does that change the location or density of the artifacts?
  2. Have you tried different layer heights? If not, use extremes, 0.28 and 0.08 as an example and see if that changes the behavior of the model.

There are probably more troubleshooting steps you might take but these are the ones that first come to mind.

2 Likes

Thank you for your suggestions @Olias

Before I start with testing I need some clarification on some points:

  1. what do you mean by “cut the model in the slicer”? I constructed one part and copied it six times in the slicer, so there are six parts. Do you mean printing just one single part?

  2. The file already is a STEP File. Mathematical artefacts can almost be ruled out since I constructed this in CatiaV5, but I will make a test print in 45° orientation!
    Do you think simplifying the model the way you described still makes sense with my specific file?

  3. Not yet, I will make a test print and scale it 50% in Z

  4. Not yet, I will make test prints

What a curious issue…

Since there’s already some really good advice on further troubleshooting, just a few additional thoughts:

  • Are the issues in exactly (!) the same locations for all parts printed or are there slight variations? The first would point towards model artefacts and simplifaction or repair which should at least give changes. The latter would point towards settings and minor machine maintenance (cleaning rods, re-tightening belts).
  • Can you spot particular features at these locations in the slicer? For example, do these defects occur in areas where the infill connects to the walls?
  • Are these layers printed with a particularly high flow or are there cooling changes in the layer? I have had prints with “slow down for overhangs” being affected by fan latency leading to completely different areas of the model receiving an extra cooling blast where it was no longer needed.
  • Do the defective layers relate to model features such as the holes? Do defects occur after a set travel path from the seams? The stable distances between defects seem to point in that direction but defects appear to start at too low a layer height…
  • Do you have the chance to closely observe while it is printing those defect prone layers?
  • And the obvious: Do the defects change/move when you print at half speed?

:crossed_fingers:

Yes, I am referring to isolating one part on the plate. Then using the (c)ut tool to split the model into sections thus isolating the part of the model that is causing you grief. Here’s how that might work. This example uses a Torus and a Cube primitive combined into an Assembly.

I’m using Orca 2.1 in this example which looks a little different than Bambu Studio but it works the same.

Step 1 Select the a “single” part and use the cut tool, Shortcut 'c".

Step 2 Select the Arrow handles, they will change color and activate when you over over them. A middle solid ball indicates changing height. This can sometimes be tricky but patient.

You’ll want to uncheck the part you want to discard, in this example it’s called the lower part in red. Also uncheck the “Place cut” box as if you leave that on it will flip the part so that side is in contact with the plate.

Repeat this process but use the X vector arrow. Note in this case I flip which part I want and this time it turned out to be the lower part in pink. And of course, you execute the cut by clicking on Perform Cut.

Just remember that CTRL-Z works here to undo any step if you make a mistake. Multiple CTRL-Z will back step to the beging.

When you’re done, you should have isolated that section of the model that cause problems.

Just remember to make sure that during your cut, you experiment with printing the model during each cut. This will help understand if the Slicer is being confused by render error during the slicing process.

Actually, that’s a common misconception. Mathematical artifacts are rendering errors caused by the slicer algorithm and can occur in any format during conversion to mesh, including STEP, OBJ, STL, and 3MF files. DXF doesn’t apply because it isn’t 3D.

A common mistake is selecting too high a STEP version, believing more features are better. Higher STEP versions include irrelevant data like material and manufacturing information, which can cause the slicer to fail. Always use version AP203(the lowest ver) if available to avoid this issue.

Here’s what that looks like in OnShape.

This is one of the many things that had me defect from Fusion 360 to OnShape. Fusion 360 doesn’t provide this level of export options and neither does FreeCAD.

That is unknown, which is why I suggested it as a good diagnostic step. Using multiple models, you can clone the model and simplify the second one. This isolates one variable at a time by reducing the number of facets on one model while leaving the original intact. If you’re using Windows, you can use 3D Builder directly, the same algorithm used by Orca and Bambu Studio for simplification.



Note that the Windows 3D builder utility only works on mesh formats such as STL and OBJ but you can always export the model from the slicer to an STL file if need be, or do it from inside CAD.

Final Note:

This is just one remedy path. The goal in the steps above is to isolate if the model is the problem. It’s important that I note that I am fully aware that your artifacts may not be model related but slicer settings related. I only bring up model diagnostics assuming that finding a slicer setting remedy isn’t producing a remedy just yet. So just be warned that this could wind up being a waste, but it is a valid diagnostic path that has served me well when making complex CAD models that often fail during slicing.

If any step yields a different result, modify the model to find a remedy. You don’t need to complete all steps; just verify if the model’s behavior changes. If it does, skip other troubleshooting steps and investigate conversion options. This could involve re-exporting in a different format (e.g., STL or OBJ) or using another app like FreeCAD to see if another export algorithm alters the model vertices. I do this often, and it has become second nature.

Good Luck :four_leaf_clover:

3 Likes

Since my last print, I fixed my X1Cs skew issue (skew deviation from 0,15% to almost 0) by adjusting the hardware (belts and position of the gantry).

Yet, the surface issues are the same.

@Olias first Post (#5)

  1. Printing only one part (the middle one in Y+), but with different wall order: default setting: first inner, outer second

    Result: surface defects are the same. So we can rule out the wall printing order and the number of parts on the plate.
    From now on, I will only print one part for all following tests, since the surface defects are still there.

  2. Printing the part with the same slicer profile from my initial post, but a 45° rotation

    Aaaaannndd it’s gone!
    Bild2

    No more surface defects with a 45° rotation!

  3. Scaling the model veritically
    – yet to be tested

  4. different layer heights
    – yet to be tested

@EnoTheThracian (#6)

  • Are the issues in exactly (!) the same locations for all parts printed or are there slight variations? The first would point towards model artefacts and simplifaction or repair which should at least give changes. The latter would point towards settings and minor machine maintenance (cleaning rods, re-tightening belts).

    Issues are in exactly the same locations. Machines belts can be ruled out → look first sentence of my post

  • Can you spot particular features at these locations in the slicer?

    Where it comes to the slicer, it gets interesting. I first thought I’m on to something. For example, some of the inner walls have some “stops” in G-code. This means, the circles are not executed as a whole circle, but in circle segments. Yet, these stops are not occuring at all the defect positions and the outer wall also is not affected. Furthermore, I realizied that even straight lines are executed as several line segments. I have no idea why the slicer is working this way.

  • For example, do these defects occur in areas where the infill connects to the walls?

    That’s hard to say, the infill is connected to the walls in many places. I don’t know how to measure the real position exactly enough and then transfer the results somehow into the slicer to see a possible dependency.

  • Are these layers printed with a particularly high flow or are there cooling changes in the layer? I have had prints with “slow down for overhangs” being affected by fan latency leading to completely different areas of the model receiving an extra cooling blast where it was no longer needed.

    The slice doesn’t show any of this. Actually all layers seem to be the same, since they are geometrical identical. As already mentioned, the lower and upper ten layers just work fine.

  • Do the defective layers relate to model features such as the holes? Do defects occur after a set travel path from the seams? The stable distances between defects seem to point in that direction but defects appear to start at too low a layer height…

    I didn’t find any of that correlations, you mind taking a look into my file?

  • Do you have the chance to closely observe while it is printing those defect prone layers?

    I can hardly see anything in there, mainly because of the black material

  • And the obvious: Do the defects change/move when you print at half speed?

    Yet to be tested!

@Olias second post (#7)

I’ll just continue your numbering:

  1. Cutting and isolating the part of the model with surface defects

Yet to be tested

  1. Saving model as STEP version AP203 → checked this, CatiaV5 R19 saves default AP203

Here you see an image where I imported the regarding STEP file into catia, with view option „displays geometry views in shading mode with edges and hidden edges“

Bild1


Summary so far:

I haven’t tested every suggestion, mainly because there are no surface issues when printing with 45° rotation. I’d first like to discuss my knew findings before printing eventually unnecessary models.

Which move would you suggest to be the next?

To make it easier for everyone, especially regarding possible slicer issues, here are my bambu studio 3mf file and the STEP part.

@EnoTheThracian

  • And the obvious: Do the defects change/move when you print at half speed?

I just printed the part V3 with 50% speed (silent mode) with the same configuration as point 1 in my previous post. Surface defects are still there however they are not that raised as with 100% speed.

Any idea?

Ok so the thing I notice in your problem areas is that there are travel moves over your zones 1 and 4 from the infill print which, if you have some nozzle oozing, could deposit material on the inner loop and that in turn can result in those “hick-ups”.


So for one try with “Avoid crossing walls” checked.
second eliminate/reduce the oozing from the nozzle

@Panamon_Creel you’re right, that seems reasonable!

I just printed another part V4 with “Avoid crossing walls”.
Unfortunately: no improvement :frowning:

Please show me your Line type preview screen with all Line types checked