What has caused this error, and how to fix it?

[Rockel83]

I’m trying to print this poop chute now:

[ Jun.14 ] Poop Chute & Basket for X1C/P1S/P1P by TrocheHole

First printed the ramp with Jayo PETG, which almost came out perfect. Even without support while the slopes are only 30 degrees.

Then I started printing the basket today. I had canceled the first print early, because it had some deformation on a corner.

I thought it could be the case that the seam point was in that specific corner and that had something to do overhang/retraction/airflow. So I tried to rotate the print about 90 degrees on the buildplate.

The second print (5 hours) seemed to go pretty good. But ther were some irregularities in the same corner, while rotated 90 degrees. But I could see it verry well, because it was pretty out of sight now. So I decided to just let it finish after regular inspections.

Bit this print came out terrible. There were far more things going on in this corner then I had expected. And I don’t know what it is or what caused it.

I’m using the same filament as with the ramp on the same settings and that one came out perfect. The slope angle is the same on both, so not having used support would have been equal I guess.

And it’s just in 1 corner, where the other one looks fine on the same angle.

I still thought it would have been because of the seam point (retraction/starting point), because it was again on the verry same corner whil rotated 90 degrees.

But the seam point even isn’t on this specific corner. It’s like 2cm further.

Thos print actually had cost me a lot of filament and time. So I need to understand/ fix this before going to larger prints/projects.

It was printed on a (almost brand new) P1S with Jayo PETG on the standard “generic PETG profile” (only set the nozzle temp to 250 and the bed temp to 85, the rest of it’s settings are standard)

PS. The last 2 pictures are from the ramp, which came out pretty good, even on the corners and the same overhang angle

[R3d-Five]

Not sure I’m onto anything, but did you have “slow down for overhangs” checked. Looks a little like under extrusion.

[Rockel83]

Thanks John,
"Slow down for overhangs is activated indeed. It might be the pictures, but personally I don’t think it’s underextrusion. At least, there’re no gaps or thin/underextruded parts and the large bottomplate is just perfect. I think underextrusion would have been verry obvious there already I guess.
And it’s just in this specific corner, also when I rotated it on the Z-axis. So almost seems to be something in the slicing proces/g-code? Besides only this corner, the rest almost looks perfect, and it’s quite a big print.

[R3d-Five]

My bad. I’ve only had something similar with cooling, but that’s clearly not happening here.
Better minds than mine will be along.

[Rockel83]

No problem at all! I really appreciate your input and your point of view. And your effort for looking into it, so thanks again.

And maybe you’re near with your point of view or maybe it might be related. But just underextrusion would seem strange to me, because other parts are almost perfect.

Thanks again John
3D printing sometimes can be really confusing, since there might be tons of reasons why some problems occur sometimes haha

[user_3026326371]

So nice to see so many pics

You problem is caused by a combination of speed, temp and slicing faults.
Check the preview and you will see how your problem areas have either speed or flow rate changes.
Changing the wall generator can often fix this.
Check the Wiki in regards to the difference between Calssic and Arachne

If the wall generating changes did not help:
Consider to reduce some speeds.
Check the wall speeds in the preview.
You want to use the slowest speed you see there for the entire model.
Time consuming, I know but it does help a lot.

Another thing worth considering is to switch off the slowing down features.
Sounds counter-productive…
Well, it is because it means you will have to wait longer for your print to complete…
Do a test print with some corners of different types and angles as well as some overhangs.
Find the fastest (mainly wall loops) speed that still produces a great result.
Use this speed as the default for inner loops and a bit lower for outer loops.
To go even further you can try an infill that actually bonds to the walls.
Some are a bit like adding another loop along the wall while others only ever touch small sections.
The more shared line length there is the less distortion for overhangs.
If the preview does not show any major slowing downs or flow changes your print should come out just fine again.

[Rockel83]

Manny thanks for your help and your insights!

Sometimes it might be a bit confusing. The standard presets seems to work good in general. And I’m getting great results with adjusted Generic PETG preset.

But when downloading a design from Makerworld, you can still choose to use your own presets, or the preset included in the 3mf file.

Sometimes some settings in the 3mf file seems to work better. It might have better or faster results then my own preset.

Also now with this design, I was actually using the integrated preset. The same preset I had used for the previous printed ramp, which came out great and much better then expected without support.
It actually seems that Arachne is selected in this verry same preset. It actually saves a lot of time with its adaptive line width. It’s an hour shorter then on Clasic haha.

But I actually think the problem might be somewhere in the speed/acceleration.
While I recieved your message, i just had started a new print. It was already laying down it’s first layer. And again there began to occur some defects in that verry same corner.

Outer shell drawing actually seems to go verry smooth and not that fast either. It actually looks verry steady.

Unless it reaches that specific corner…
It just draws the outer shells verry persistent and then on that verry same sharp corner it accelerates as crazy to the bigger inner corner.
It’s only on this specific corner… That crazy acceleration…

I feel I need to change something in the slicing process here…

PS. I think this might be a big part of the problem here

[user_3026326371]

One fix would be to change this corner…
You have one short line than the long straight.
If you change this corner to have a small radius instead the slicer should adjust to a lower speed.
Of course you can also change the wall speeds and/or acceleration to match the speed for the other rounded parts.
With the current geometry the slicers sees this corner as a great way to speed up.
But the change from slow to fast on such a short distance messes with the pressure advance calculations.
It might help to double check the k-factor settings for this one as well.

Either make the corner round or lower the max speed for the walls, if that’s not enough lower the acceleration to something used in the high quality profiles for lower layer heights.

[Rockel83]

Thanks for your insights! appreciate it!

I already calibrated the flow rate for this (roll) type of filament twice yes. It seems to be just fine. Also on other parts of the (and other) prints, they come out just great.

It’s only this corner which messes up the whole print. Which is strange, because there seem to be little people complaining about these presets in this very same 3mf file, while it’s also a verry popular design on Makerworld. And seeing pictures of others, it seems to come out great on similar printers. While this design is specific for the X1 en P1 series, so probably most of those are printing on the verry same machines.

So maybe it might be something with the filament I’m using, where a lot of other people seem to have printed with PLA.

But at least for me the problem seems to be in that specific corner. I’m having another print running now. And other parts come out great. Actually never had such a clean first layers or layers this clean on such a big printed surface.

PS. I just stopped the print while writing this, because that verry same corner is getting to be messed up again…

I’m going to try my own settings now. Because I really don’t understand why I seem to be one of few people that seems to have problems with this specific print…

• While printer settings and hardware are stock, and should be the same with most others.
• Others might have been using the same settings and presets included in the 3mf file as wel.
• Seems that verry same corner is also just fine with most of them, without the need of changing the design.

Although I agree that it might have solved my problem when that corner was just a curve, instead of 3 different vectors that close to each other. Although when seeing the design, it actually doesn’t looks strange. It actually makes sense.

I actually was thinking going to print these parts in the far more expensive PETG-CF filament. I’m glad I didn’t. That would have been a waste of filament this way haha

[Rockel83]

This is the last print.

I’ve tried to put the seam line on that verry same edge and corner where it starts to accelerate again.
Hoping forcing to stop acceleration this way. My logic was that after slowing down on the slope/bridge edge, it would finish the layer at the end of it, retracting and start a new layer from there. Forcing it to stop accelerating while extruding…

This only seemed to have a minor effect. I think there is a slight improvement so far, but it still looks terrible.

While the rest of the print so far is just great and almost close to perfect so far:

I don’t see actually any artefacts or problems that might point extreme miscalibrations or extreme settings conflicts.
There seems to be a decent flow. All lines are even and bottom and top almost seem to be close to perfect without signs to over or under extrusion or other artifacts. At least to my eyes and understanding.

It’s only that verry same corner which is ruining the print all the time:

[kip]

OrcaSlicer has fixed this issue a while back, Bambu Studio is late on this.

Explanation details link below, if you use the OrcaSlicer latest version. Should improve the print quality.

[Feature] Introduced option to control amount of internal bridging, fixing internal bridge missing for some sloped surfaces

[KanneKaffe]

The project is set to PLA. If you switch to PETG, you will probably also have to adjust the speeds. If I change the print profile to Generic PETG, it still results in speeds well over 200mm/s. If it were my project, I would set the outer wall speed to 80mm/s to be on the safe side, the inner walls as well, it also affects the infill and the solid infill (all at 80mm/s). I would set the overhang speeds to at least 40mm/s, no slower, but perhaps a little more cooling.

For example, if the speeds for the infill are too high, there may be problems with the overlying layers if the infill has not been printed properly (the filament must be able to melt sufficiently in the time it takes to extrude it, so give the machine more time for this).

Many paths lead to the goal, you will certainly not try them all.

PS: Because it’s obligatory: you have dried the filament sufficiently

[Rockel83]

I was thinking taking a step back, and doing some teat prints first again. All came put great.

Also did a overhang test print. Although small, I was still impressed that it showed verry good results until 70, which represents 20 degrees in bambu studio if I’m correct. The angle in the print is 30 degrees (60 on the test print).

This was made on the the “generic PTFE” profile with unchanged speed settings, 200/300mm/s for outer and inner walls.

Al the previous/failed prints were made on the 3mf or the same “generic PTFE” settings. With speeds on both of 100/150 mm/s for the outer and inner walls.

Since the overhang test was printed on unchanged “generic PTFE” speed and looked pretty good, I decided to give a try on higher speeds on a smaller basket.

And it came out significantly better.

Almost with clean corners on a 30 degrees slope without support.
Maybe not to call it perfect. But the differences are quite big.

Maybe not the same comparison, because the failed print is a slightly different design. And it will be sliced differently after that failed corner.

I’m going to try to “isolate” that part on the actually print now and going to print it on stock speeds of the generic PTFE preset to see if I might get the same results without slowing down to 100/150 mm/s

[Rockel83]

Just did a test on a new model, but this time with unchanged speed settings as on the small basket.

Just had stopped the print midway, while it was already clear that this part was way better already on the standard “generic PETG” profile with unchanged speeds of 300/200 mm/s for the walls.

Although it seems to be sliced differently after cutting a big part of it, for not wasting too much filament.

Seems the seam line now is exactly on the corner where on previous prints the errors occured.

[user_3026326371]

Great to see you make good progress here!

[KanneKaffe]

The overhang is not necessary. Why do you have to attach such a stupid stiletto to it? It’s standing on the ground! If it bothered me, I would probably add a cube in Bambu Studio that exactly covers the area of the overhang and otherwise disappears into the printing plate. Then the additional object gets an extra infill that requires little filament and would only be printed with 2 perimeters. This would be sufficient as a support. The “designer” could have simply designed the overhang differently so that it is easier to print.

I have just looked at the model again. It is simply incomprehensible. There was every freedom to model this angle at 40° to 45°. You can recognise this when you create the model.

[Rockel83]

The “stiletto” fits in the hole of the ramp, to hold everything in place and prevent the basket from being separated. Which is actually a good design I think. Although it looks a bit strange when separated from the ramp with this stiletto and it should not need to reach the ground. But from the other hand, than it would have needed support. It’s actually just a notch to keep everything firmly in place.

Although I don’t know why the designer choose to go to the max printable angle of that slope (without support). I guess around 35 to 40 degrees would have been a bit better yes.

But it’s unclear to me what causes this problem. The smallest basket just printed fine, with just a little imperfections compared to all the other baskets.
But it’s a difference of day and night, while the slope angle and the specific corners and edges are exactly the same.

I’m printing now a huge AMS riser, which has similar overhangs, but on a much longer surface. Also without support, with the same filament and about the same settings. But besides a few reasonable imperfections on the slope part, it comes out almost perfect.

So I’m actually surprised why such a small and relatively easy basket causes such problems.

Although I’m printing this riser on a standard 0.20 preset, the basket was printed on a 0.16 preset.
The standard 0.20 preset already has the outer wall speeds set down to 60mm/s. So I’m not sure why the standard speed of the 0.16 profile is set to 200mm/s.

Maybe I should try to print it in that verry same 0.20 profile after finishing my riser…

[KanneKaffe]

@Rockel83

Although I don’t know why the designer choose to go to the max printable angle of that slope (without support). I guess around 35 to 40 degrees would have been a bit better yes.

That’s the crazy thing, the angle doesn’t matter at all. If it had been steeper, it would have been easier to print. The ramp underneath where the stick hangs would then have been shorter. The upper ramp on which the filament slides down could have remained at the same angle. This would not have changed the function.

Perhaps he realised afterwards when it was already “too late”. Changing the design again could be a bit of work. If you have created things for the 3D printer often enough, you automatically pay attention to things like this, because you want others to be able to print it properly. With technical models, I always make sure that angles where overhangs occur are between 40° and 45°.

Well, what the heck. He wrote that you should use a support structure.

[Rockel83]

Well yes… I founded that comment about just adding a support structure a bit strange too.
Support for a slope like this size will neither come out clean, and will result in even more filament. So that does not seems to be a solution no.

I agree that it probably would have been easier to print a slope of a bigger angle and yes, 5 to 10 degrees might have been a big difference. This is around 30 degrees now. so a bit on the limitation of FDM-printing.

But even this would have been doable…
I believe it’s also a on how it’s sliced. But it seems I don’t have much of an influence on that, maybe in conjunction with PETG, which behaves a bit differently then PLA.

Whatever I try, I get the same results:

• drastically slowing down
• increasing the layer height
• increasing temperature
• decreasing temperature
• Calssic vs Arachne

And always in that same corner, while the start of the slope actually looks nice.

Only had good results with te smallest basket, which seems to be the verry same slope as all the biggerbaskets.
It also came out pretty clean so far when cutting off the rest of the basket. So only printing the slope part.
But then after trying to print a bigger basket in it’s whole, I just have the verry same issues in that verry same corner again.

After slicing that smallest basket, which was pretty OK, it seemed the seamline was all over the straight part of the slope. Which looks logical. because of this it just stops and retracts at the end and does not need to accelerate to go around the corner. This might be a solution I guess. So I even tried to draw the seam line on that verry same corner, but again without results after printing.

I also seem to be the only one with this specific problem so far when looking on the design page. But I believe all other examples are printed with PLA.
I also find it strange that I first printed the big ramp from the verry same design/3MF file, which has a slope around the same angle, and this came out almost perfect.

I’ve tested my printer settings after with some other tests and so far my printer seems to be configurated pretty decent. Test prints come out good, with correct dimensions. There are no serious signs of over- or underextrusion. And even printed a complete AMS riser after. And all come out pretty decent and close to great/perfect.

After tried to print a basket again, but just the same problem again…

Maybe I should just look for an other design poopshute. But I actually liked this design because of the possibility of magnet inserts in combination with the removable baskets, in diferent sizes.

But again, it’s just that strange that the ramp came out great, but that I’m facing these kind of problems and frustration on a simple, small basket with a slope…

btw, when looking at the first published foto’s. It’s just a total mess on that right corner on the bottom. But let’s say 1/3 of the top part actually came out pretty clean, which is still on the same slope angle.

[KanneKaffe]

Have you tried adding supports? I would choose a tree structure and space the supports about 1cm apart at the top.

Because you mentioned that you have increased the layer height. You would have to reduce it for the overhangs. Try variable layer height. See picture:

You also need to cool the overhanging walls well. Speed and temperature must go together with PETG so that the results are correct. It may be helpful to further reduce the speed of the overhangs, possibly down to 40mm/s, while at the same time cooling 100% with the component cooler. It might also help to increase the number of walls (although I wouldn’t recognise any advantage here when trying it out in the slicer).

All just mind games. I thought about printing the part myself, but I no longer have any PETG in stock. Either PLA or ABS. I would have to try ABS first because I never print it, but I have a few rolls here.