Cura 5.5 and massive pillowing

[LePaul 321]

I have prints that keep experiencing pillowing even though they have 5 or 6 top layers 

On this one, with 20% infill and 6 top layers, this is the result.  0.25mm layer height with a 0.8 core/nozzle

 

Here's the results and my settings...thanks

 

[Gero 54]

have also found this. Was printing done with the new Balanced profile? I think with the higher speeds there are still slight problems here. In this case, the material at the 0.8mm nozzle does not have enough time to completely close the layers

 

[LePaul 321]

Yes, I had tried the "new" balanced "fast" printing setting that clearly the Cura team didn't test adequately.

 

Very frustrating as we have a media event today that is coming to see our fleet of Ultimaker's in action.

 

I'm now re-slicing in 5.3 and re-printing since these newest builds are totally unreliable and untested.

 

 

[Dustin 130]

Can you both please share what materials your printing with? and your project files please?
I will share these with the Dev Team.

[Gero 54]

4 minutes ago, Dustin said:

Can you both please share what materials your printing with? and your project files please?
I will share these with the Dev Team.

Printed with Tough PLA Black. 

Here I have the project file.

UMS5_BigSpoolHolder_HalterV1_TPLA_test1.3mf

[Gero 54]

52 minutes ago, LePaul said:

Yes, I had tried the "new" balanced "fast" printing setting that clearly the Cura team didn't test adequately.

 

Very frustrating as we have a media event today that is coming to see our fleet of Ultimaker's in action.

 

I'm now re-slicing in 5.3 and re-printing since these newest builds are totally unreliable and untested.

 

 

i agree, the new profiles with higher speeds are not quite "ready" yet. if we look at the layer adhesion at the higher speeds or how well several outer walls adhere to each other, our first impression is that the material flow and nozzle temperature still need to be adjusted. otherwise, a strong decrease in the strength of the components is to be expected so far. 

however, the engineering profiles are working very well so far. simply because printing is still slower here. since the profiles only print a little faster here, there are no problems in this respect. 
due to the quality of life improvements in 5.5, i'm clearly on 5.5 right now, but i'm only using the engineering profiles for the time being. it runs quite reliably with them. 

[LePaul 321]

49 minutes ago, Dustin said:

Can you both please share what materials your printing with? and your project files please?
I will share these with the Dev Team.

 

I can't share the part publicly in this forum but I can send you the project file via PM, if that is ok?

 

These prints were done on the Ultimaker S5, Core 0.8 and Tough PLA

 

[Cuq 198]

1 hour ago, Gero said:

Printed with Tough PLA Black. 

Here I have the project file.

UMS5_BigSpoolHolder_HalterV1_TPLA_test1.3mf 316.33 kB · 2 downloads

 

Yes certainly something wrong , in 5.4 for the same profil the Print Speed was 45 mm/s and now in 5.5 it's 100 mm/s  For sure you've almost divided the time by 2.  

 

Cura 5.4 

 

Cura 5.5

 

 

So the pillowing effect is certainly due to this high top/bottom Speed.

 

Edited November 8, 2023 by Cuq

[LePaul 321]

Faster isn't always better!

 

[Slashee_the_Cow 271]

9 hours ago, LePaul said:

Faster isn't always better!

Ambulances excluded. But it took me about zero time to realise that just because my new Ender-3 V3 SE could print at 250mm/s and accelerate at 4000mm/s² doesn't mean it should. I didn't dial down the acceleration enough though and it started ripping out supports on the fast travel moves away from them, so I wrote a post to severely limit the acceleration while printing supports and until it starts printing something that isn't support.

[PaulKuiper 11]

We have observed pillowing in the PETG AA0.8 0.2mm layer height modes (0.3mm and 0.4mm are OK), which will be fixed in Cura 5.6. We fixed this by increasing the  'Bridge wall flow' and the 'Bridge skin flow'. Try setting these to 200% to verify that it has the desired effect. Note: the top layers start with a single bridging layer.

We tested pillowing for PLA, T-PLA AA0.8 modes, but never observed it. We even lowered the layer count from 6 to 5 because of this. However, pillowing can be sensitive to flow changes. The moment bridging threads break, pillowing will be inevitable (even 6 top layers will not close these gaps).

We will investigate if we should also increase the bridge flows for the PLA, T-PLA and ABS for the AA0.8, 0.2mm layer heights (to much flow can cause sagging in normal bridging areas).

[PaulKuiper 11]

I just noticed that you also changed the layer height to 0.25, which means a 25% flow rate increase. The high speed modes have been designed to be at the edge of the flow rate limits of the cores. This change will thus cause severe under extrusion.

It is better to choose the 0.3mm layer height mode in this case and reduce the layer height to 0.25mm.
We have a pull request open to warn the user if the flow rates of the cores are exceeded (e.g. when changing speed, layer height, flow settings). So hopefully we will guide the user better in the future.
 

[FredericM 1]

We were indeed already aware of this happening with PETG after releasing Cura 5.5, which has already been fixed for 5.6. I've been running some tests, and I observe it now too with Tough PLA, and managed to fix it in the same way as with PETG. I'm now also going to test PLA and ABS for completeness, and it should be fixed in time for the 5.6 release.

 

Thank you for bringing this to our attention!

[LePaul 321]

5 hours ago, PaulKuiper said:

I just noticed that you also changed the layer height to 0.25, which means a 25% flow rate increase. The high speed modes have been designed to be at the edge of the flow rate limits of the cores. This change will thus cause severe under extrusion.

It is better to choose the 0.3mm layer height mode in this case and reduce the layer height to 0.25mm.
We have a pull request open to warn the user if the flow rates of the cores are exceeded (e.g. when changing speed, layer height, flow settings). So hopefully we will guide the user better in the future.
 

 

Hi

 

I've been using 0.8 nozzles for a long time and find 0.25 layer height gives great detail and reduced layer lines, especially on large prints.  That means less post processing for me - think sanding, filler, sand, prime.  Rinse and repeat.

 

That said, the "Balanced" profiles should be labeled as experimental, which they are.  

 

I was hopeful that owning an Ultimaker S5 (well, 7 of them!) would mean that the latest version of Cura would be thoroughly tested on your native machines.  

 

Your last few releases have been plagued with bugs that are painfully obvious to your customers and users.  Please slow down.  Test thoroughly.  Gather input.  Don't be pressed to release until it's truly ready.  (I come from a software dev background, so I know what you're probably up against.)

 

Usually I wouldn't jump in and try a newest version but I did this time.  And it set me back a few days and plenty of wasted material.

 

Thanks for listening to my feedback.