Unable to print a single box with even walls

[NVNDO]

Hey everyone,

all boxes I print have a protruding irregular outer wall where the solid base is printed compared to the rest of the walls.

I’m trying to fix this since over a week and I’m unable to do so on both my printers A1M and P1S. I tried countless settings, layer-time, fan-time, speed, temps, less acceleration and what-not.
Nothing seems to work. I feel defeated and need help …please…

On the left box with the green arrow I had slown down all walls to 50%, acceleration, changed layer-time and fan-speed. it seemed a step in the right direction, but decreasing speeds and altering those settings further did not help to fully mitigate this and have since wasted an entire spool…

Edit: I assume the answer lies somewhere in between those settings, but I can actually not understand whats the issue here. The wall consists of 4 walls printed side by side throughout the entire box including the solid base. If the entire wall is basically printed the same - why are there irregularities?

[Chris1974]

After slicing you can take a look in the preview tab. Standard it shows “Line Type” but in the pulldown menu you can select other thing to view such as speed, layer time, fan speed, flow, temperature etc.

Looking at those previews might give you a clue what is going on at that specific place.

But based on how it looks I suspect it has something to do with shrinkage…

[NVNDO]

Thx Chris, that was my main window for the last week.
I set everything i could to be even on all layers, but it did not help.
Here an image of the layer-time (only screenshot I made)

One can clearly see how the solid surface takes much more time (which is obvious) but I unfortunately can not see the independent wall-speed (which would be interesting) as I think, that the wall-speed of that solid part is different from the upper walls speed. By changing the layer-time I can unfortunately only change the overall time (and that did not help either)

The fan-speed also increased to its maximum at that solid layer which I also adjusted to be even with the rest. I of course see differences in all boxes, some better, some worse, but they’re so minimal that it could also just be basic print variances. I’m truly lost.

How can such a basic box / cube / thing be so challenging to print…
This is my overall problem in 3D printing as I generally only print geometrical, flat, even, things on which even the tiniest error shows.

[Chris1974]

When a solid base is connecting to a thin wall you often see these lines due to difference in shrikage… I think that is what is going on, but I can’t see what the geomtry inside the box is like…

[ElektroQuark]

Try printing the outer perimeter first.

[lkraus]

This seems to be a pretty common problem. I would first try setting Order of walls to outer/inner. The theory is that the outer wall can be aligned on the previous layer before the large flat floor of the box can begin to shrink.

Another common suggestion is to change the design, adding an interior fillet where the walls meet the floor of the box. I’ve not seen a good explanation for how this helps.

[user_3026326371]

It is a problem I never had until I got a Bambu printer LOL

Things that make these weird wall ‘features’ happen, or at least affect the severity:
Layer time…
It seems that whenever the layer time goes down walls start to struggle.
This becomes more prominent the longer the wall path is before it makes a more or less sharp turn again.
Flow rate…
Similar to the layer time, if it goes up fast the print struggles.
The new discontinuity feature in 1.9.5 tries to mitigate this a bit, especially for overhangs by changing the ramping speeds.
Does not fix the problem though.
Pressure changes…
See under flow rate - this can be a pain with lots of retractions or models having many short lines on the perimeters.

Why is it so annoying?
Because it should NOT happen!
It is not even speed related as it happens no matter what the average print speed is set to.
Even in vase mode it can ruin prints by suddenly speeding up when the wall starts.
The first layer of the vase mode wall is printed much slower and no option to address this - as it should not happen.

Especially on long, flat walls you can often observe how the issue is less prominent on one end of the wall.
Checking the slicer preview shows the clear relationship between acceleration, PA calculations and ramping.
Depending towards which end of the flow rate calibration you sit the ‘wrong’ line’ is either perfect at the start or at the end.
But going for the golden middle of the flow rate settings still won’t fix the problem - apart from these flow rates being useless for normal print situations.

As said, none of my non-Bambu printers behaves like this.
What’s even worse is for some prints it seems as if the extrusion path changes.
On a perfectly flat wall a difference of close to 1mm is just not possible with extrusion related issues alone to blame.
Why do I keep blaming Bambu ?
Well… :

Try to print a simple storage box, like 10 by 15cm with 5mm walls and bottom.
Plain and simple.
Assuming your filament if half way calibrated the print SHOULD come out just fine but it won’t.
Once the bottom is done there WILL be a line around the outside of the box.
Now please do try the print outer wall thing first as it IS logical it would fix the issue…
Nope, still no luck and despite the outer wall supposed to be printed onto the previous - the printer won’t do it!

So far the only way to fix it an acceptable level is to go to snail speeds for the printing :
Use the SLOWEST speed in the print settings as the default value.
That usually is the speed for overhangs.
Get all other print speeds within 20% of that speed, preferably make them all even.
This WON’T fix the layer time issue but it does wonders to the pressure in the melting chamber if there is not too many retractions involved.
If you can’t do without retraction then try to tune the settings to the shortest pull/push and slowest retraction speed that still provides good results.

In case you have a high flow hotend as the default option to print things:
Some filaments allow to really push the speed and flow rate limits.
Sadly that also means that for areas where we still have to slow down to normal speeds, like overhangs, the pressure changes can be quite severe.
As the majority of models out there won’t really allow the max speeds those hotends are capable of it makes sense to use them more ‘wisely’.
The printer, with a max flow rate set to, let’s say, 35 won’t have to slow down as much and might even go above the set default print speed.
Tweaking the speed settings to get more uniform flow rates can really help to minimise those pressure changes in the melting chamber.
Funny enough, for most prints, there is no meaningful increase in print time by lowering some print moves a bit.
Like outer walls for a complex bust or such - while 200mm/s may be shown in the slicer preview, this speed is for the MAX the head might reach during this loop.
The average speed for this outer wall loop though might be as low as 100mm/s.

I like to have two base profiles for my high flow hotend.
The first aimed to max out all speeds for models that allow for it.
The second aimed to provide the best possible print quality, mainly in terms of surface quality.
As this wall problem exists basically since day one without Bambu having provided a WORKING solution I guess we can only send out more support or feature requests to fix this one day.

[NVNDO]

@Chris1974

but I can’t see what the geomtry inside the box is like…

They’re just boxes with a plane bottom and a 2mm fillet

Here a picture of all the boxes I printed so far (with different settings) of which not one is even close to good enough.

@ElektroQuark

Try printing the outer perimeter first.

& @lkraus

try setting Order of walls to outer/inner.Here’s the result.

Another common suggestion is… …adding an interior fillet

That’s what I did here, but the result is even worse. The lines might be less pronounced, but other unwanted artifacts start to show.

@user_3026326371

I’m glad you said that XD:

It is a problem I never had until I got a Bambu printer

because usually its me saying this and its usually not a great start into a conversation (apparently) Even though its a fact sometimes.

Your post is eyeopening and I am currently testing speeds I would otherwise have never though of testing. Currently printing @50mm/s (Speeds a standard Ender 3 easily surpasses with excellent quality) and it seems like this is the solution to this exact problem.

Printed with standard settings, but all speeds @50mm/s

There’re still some minor lines, but nothing compared to what I had been experiencing even at slow 50% speed settings.

My second test (still running) includes a reduction in Acceleration speed for Outer walls @3500 (instead of the 5000) and this seems to do the trick, at least from what I can already see.

Let’s wait for the print to finish and see.

@BrotherC

Try to print a simple storage box, like 10 by 15cm with 5mm walls and bottom.
Plain and simple.
Assuming your filament if half way calibrated the print SHOULD come out just fine but it won’t.
Once the bottom is done there WILL be a line around the outside of the box.
Now please do try the print outer wall thing first as it IS logical it would fix the issue…
Nope, still no luck and despite the outer wall supposed to be printed onto the previous - the printer won’t do it!

As this wall problem exists basically since day one without Bambu having provided a WORKING solution I guess we can only send out more support or feature requests to fix this one day.

Is there already a request on Github or somewhere else I, we can join?!

Thank you so much for your post. It was a blast to read and very refreshing to see someone understanding my frustration.

Edit: Testprint is done!
It looks flawless - @50mm/s and 3500 Acceleration for Outer walls.

I think we can all agree that this (slow printing) should not be the solution to a printer announced of running at 500mm/s

I will now proceed to print the entire cube and see what result (I) we get.

[NVNDO]

Seems like I went far too early…

hm…

[user_3026326371]

Nice !
You came a long way in such a short mount of time.
Took me a very long time to work those things out and get good enough results.
Glad it wasn’t wasted time and now helps others.

It is VERY hard to compensate for the many factors causing this issue.
Especially hence flow rate calibrations and such are linked to the print speed.
You are already closer than many came after many days of messing with the settings.

I know there were a few support request and bug reports in relation to this in the past.
But the ‘fixes’ if you want to call them were all based on the existing, flawed, calculations.
More and more users pushing the speed limits while more and more (mainly new customers) face issues they can’t explain or fix.
So what’s the solution? Of course you fix things !
This happened on a large scale with the 1.9.5 Studio version.
Rather than increasing or sticking with default speeds Bambu reduced a lot of them.
Like the default outer wall speed dropped from 200 to 160mm/s.
Combined with the magic we can’t see it all helps to mitigate this issue and some other to some degree.
What it does not though is addressing the actual causes.

You now worked out that not all is golden in the Bambu world and that in order to fix certain certain things you have to go back to more traditional print speeds.
So how is a proper support request supposed to work considering Bambu clearly IS aware of those things and the factors causing it?
A fix would have been not to decrease default speeds and to render working user profiles or 3MF containing them useless.
A fix would have started by least admitting that it exists and that it takes an awful lot of work to correct the relevant calculations.
I other slicers and for other high speed printers we see this wall issue at times but also WORKING and easy to apply solutions to get the correct result.

Since your cubes are useful, easy to print and all that, why not abuse them for a new listing combined with a support request/bug fix ?
Users can print them, compare their results and also compare the print profiles of other users.
Bambu should have little issue using this date to hone on the few culprits throwing those PA and pathway calculations out of the window…
Anyway…
What does the slicer preview say about the speeds and flow rates for the transition from overhang to straight wall ?
Judging by your image and based on my results I would say the transition is not too sudden.
Is the inside bottom below the vertical part of the wall?
If so try to reduce the acceleration by another 10 to 15% to see if this improves the result.
Did you do a fresh temp test for those new speeds you now use ?
Am asking as a lower print speed usually means you can also lower the extrusion temp a bit.
Sometimes being a bit on the high side is a good thing, like for combined infill.
Ideal is when bridging, overhangs and solid infills come out with good layer bond and not any higher then what is required to get the right level of gloss on the surface.
High/soft means the filament can be pushed into voids easier but also that it can be pushed outwards easier.
Low/firmer means the extruded filament might not creep into the smallest of cavities but in return comes with a more consistent outer surface.
If you pint a cylinder with many straight faces rather than being based on a true circle you can observe the effect quite easy, even in vase mode.
At higher speeds the head not only overshoots a bit but the mechanics can also resonate.
Like corrugations on the dirt road you get a ripple like effects from those corner moves.
A bit like waling the tight rope.
You have to keep the body, the filament flow well balanced.
Whenever you make a move this balance changes and you need to compensate.
Same for the actual pathway, resonant movements and all that.
The problem we have is like having changed the big umbrella or long balance stick with a set of heavy balls hanging on ropes.
Whenever we try to compensate we have to somehow predict how those heavy balls will react and try to throw us off the rope.
Until Bambu fixes the relevant calculations and provides more reliable hotends we can only do what we would do on the rope = going slow, steady and without and sudden moves

But: Isn’t it great to learn so much, try so much only to try to solve problems we would not have without a 3D printer ? LOL

[Chris1974]

Is it possible for you to change the geometry to add fillets or chamfers so there is a more gradual transition from the bottom to the walls? That might make it a lot better.

[NVNDO]

Glad it wasn’t wasted time and now helps others.

Although slowing down speeds would have been easy, who would (now in 2024) print at 50mm/s? Probably no one, therefore no wonder I didn’t go that slow. 70mm/s was my limit at which I though, “Damn! This is not working… what on earth?!” and didn’t even think of slowing down even more.

It is VERY hard to compensate for the many factors causing this issue.

And not only this one, almost all custom configs for a specific part have sometimes taken me days to figure out. When going Bambu I thought, finally just printing, no more tinkering, no more unnecessary waste by printing several prototypes and quality-checks. I faced the reality pretty quickly and went back to applying knowledge. But some cases, like this one, are driving me nuts… This part would have printing flawlessly on all my previous printers beyond 50mm/s.

This happened on a large scale with the 1.9.5 Studio version… …like the default outer wall speed dropped from 200 to 160mm/s.

Did they? Updated the moment the release came out and it still says 200…

THIS

Since your cubes are useful, easy to print and all that, why not abuse them for a new listing combined with a support request/bug fix ?
Users can print them, compare their results and also compare the print profiles of other users.

is a brilliant idea!!! Some people have more time than others. For some it would be great challenge and fun to solve, for others great to have people solve it for them.

Judging by your image and based on my results I would say the transition is not too sudden.
Is the inside bottom below the vertical part of the wall?

The inside bottom is slightly above the edge where the straight wall begins - The bottom is exactly where the defect begins and it lasts until the fillet ends.

I’m starting to wonder if using a fillet is the culprit here
(when my current print ends - I’ll make a test without)

Did you do a fresh temp test for those new speeds you now use ?

Nope! ^^ And I will not do so. There has to be a way with proper speeds… and now as I’m thinking. I don’t have to wait for anything (I have 3 printers here) dough!! XD I always forget Hahaha… Let me try that without fillet first. OKAY. Print is on its way.

High/soft means the filament can be pushed into voids easier but also that it can be pushed outwards easier.

One could also try to just lower the temps… Argh… there is just so much one can try.
See… exactly what you said. If I wasn’t alone with testing, imagine 20 people and each trying a different setting. Bam!" In one hour we would be as far as me on my (or anyone else) on their own in 2 days.

At higher speeds the head not only overshoots a bit but the mechanics can also resonate.
Like corrugations on the dirt road you get a ripple like effects from those corner moves.
A bit like waling the tight rope.

You mean the typical Bambu walls? Those which have the exact same pattern as the belts? XD
I know what you mean… sadly…

Isn’t it great to learn so much, try so much only to try to solve problems we would not have without a 3D printer ? LOL

You just made my day! ROFL

[NVNDO]

Hey @Chris1974 I’m currently running a test-print without fillet.
The design is mine, therefore I can change everything I want, yes.

Next step: A cube and each wall with a different chamfer or fillet to see what that does.
It will either totally mess up everything, or give some interesting insights.

[DeVerm]

I recommend you experiment with using a circular pattern for solid infill and a different sparse infill like gyroid.

Also, try to disable the “slow for overhang” setting and try a different wall generator.

If possible, test with a different material, like a glass or carbon fiber filled filament.

[user_3026326371]

You again confirmed how the overall layer time can make a a mess of outer walls for no obvious reason - thanks !

I see your point(s) here - really do, especially since I also see how my other printers don’t have those struggles.
There is no sense to print a seemingly simple thing at crawling speeds.
Not if almost identical designs are able to max out the speed of machine while giving flawless results.

As it happened I needed to print something in glow PLA.
Meant I had to swap my high flow for the 0.6mm hardened steel hotend.
Imagine my surprise when the print came out without those wall features I get from the same model printed with the 0.4mm hotened.
Never made this comparison, never paid attention either…
And that really got me thinking…

Bambu is optimised mainly for the 0.4mm nozzle.
And optimised here does not mean optimised for best results in ALL cases.
Like for cars we have different models and modes to choose from, like 0.2mm, fine, draft and so on.
All tuned for a specific purpose - the lower the layer height and the higher the quality level the better the print.
At this stage it hit me that we are going all wrong here…

Bambu does all this intentionally!
The 0.24mm is for draft use, 0.2 for general use and below that things start to get interesting.
It was never intended to provide the best possible quality for all layer heights!
And for other nozzle sizes Bambu is way behind in terms of providing great profiles.
No surprise then that for a 0.6mm nozzle the flaws in the calculations are less severe.
Those much higher flow rates mean things have to slow down accordingly in order to work during flow rate changes.

With that in mind it makes sense that certain problems we keep having disappear if we go for a lower layer height - all related calculations change.
What causes bad artefacts at 0.2mm won’t happen at 0.12 layers as all speeds and acceleration values changed downwards.
Think about it - despite the flow rate going down with the layer height the speed also goes down not up

Another thing that can now be explained is why a quality high flow hotend performs so much better.
All those vital calculations also include the pressures and the melting chamber and the PA values.
While the Bambu hotends are too restrictive in terms of flow rate and come with excessive pressures in the melting chamber - a high flow hotend is almost like free flow at low flow rates.
Pressure changes are far less severe as well, so these calculations provide a totally different print result just by using a different hotend.
Now for the big questions:
What would be the perfect hotend for the machine?
Why did Bambu not optimise things for their own hotends ???

[NVNDO]

Hey DeVerm,

thx for commenting. As there is only infill in the bottom part and the problem occurs several layers later I discard changing infill as a solution. (same box without any infill behaves exactly the same btw)

I could of course try different filaments but I don’t think this will bring me any closer to finding a solution. It’s about solutions, settings, not avoiding the problem.

Using GEEETECH PETG since several years without any issues and I’ll probably stick with this brand although I sometimes thought about Overture, which also seems to be pretty reliable.

PD: I can of course print in several other filaments if you’re curious.
PD2: Never used fibrous filaments and will probably never do in the future after seeing this video here: https://www.youtube.com/watch?v=RLt9l6YxvHk Scary sh*t this is!!

[NVNDO]

I’m pretty amazed by how deep you dig.

From what you’re are saying I filtered out (3) things.

1. Lower layer height
2. Bigger Nozzle
3. High-flow Nozzle

Glad I have everything I need.
I have additional 0.2 & 0.6 Bambu Hotends (not yet used)
and a High-flow Hotend with Stainless and Hardened Steel CHT Nozzle. Highflow but not Hight quality as it was around 35€ (but as everything from certain brands is so extremely overpriced I’d guess that this is actually a reasonable price for a hotend)
It looks very promising. I have nothing to loose, so I’ll give that a try… AFTER printing at 0.16mm

So far, non of my tests resolved the problem. Printing at 50mm/s and 3500 Wall acceleration did almost give a flawless result (which depends hugely from which angle you look) but I’m not going to print at those speeds. Simply “NO!” Why?: Because “NO!” XD

[DeVerm]

The reason for the circular solid infill pattern and different sparse infill patterns is to prevent warping later on in the print. Even a tiny warp that is near impossible to see will result in surface imperfections.

You have a square part which works very well for a Archimedian or concentric solid infill pattern. Also, long straight lines in sparse infill create warp so a gyroid pattern would prevent that.

I’m not sure if a fiber filled filament is what works for the look and feel of this print but it is a very good bet for high accuracy prints.

[NVNDO]

The cube is 60x60mm, no chance of warping here, but I see your point. On bigger objects and a 120x120 cube it could of course make sense. I usually use the new Cross hatch which works great but for little Infill I still use Grid to increase equal strength.

Fiber filament will not enter my home “ever”! Hahaha… I have a huge problem with harmful stuff I can not see and therefore control.

[NVNDO]

I think I got it!

Well… @user_3026326371 got it!

I just used 0.16mm Standard settings. The imperfection is NOT GONE, but gone.
Depending on how you tilt the cube to a lightsource you can still see something. But that something must be in qm because its basically invisible from what ever other angle you look at it until you get that 0.5º at which its visible. I call this “Good enough”.

Here are some picture of the Test-section.

I will now print the full cube (cause you never know) and celebrate then (when things work out)
Wish me luck!!