Gaps between adjacent linear infill

[ttaiwo 1]

Hello,

 

Firstly the parts I am printing are for conducting tensile testing. Therefore the printing choices may seem unusual for standard 3D printing, but they are essential for my tests. If I indicate that a proposed solution is not feasible, it could be due to these specific requirements.

 

In the images provided, you can observe gaps in the magenta section of my specimens, which consist entirely of infill with no walls. The infill setting is on 'lines' at 100%, not connected, and aligned with the arrow's direction. I'm using an Ultimaker S3 with dual extrusion, where Printhead 1 uses Ultimaker Magenta PLA and Printhead 2 uses Ultimaker Blue PLA. Notably, Printhead 2 does not exhibit this defect.

 

Specimens with significant gaps tend to crack across the specimen, perpendicular to the direction of the arrow, during loading. If a crack intersects a gap, it then travels along the gap before continuing across the specimen, resulting in a stair-like break, as shown in the final image. This issue could be impacting my test results.

Efforts to resolve this include cleaning and adjusting the tension in Feeder 1, swapping print cores and materials, and increasing the line width from 0.4 to 0.45. Unfortunately, these steps have not eliminated the gaps. The frequency of this issue is causing me to discard every other print.

 

I have yet to try printing at a higher temperature, which could affect my previous results. Calibrating the e-steps of Feeder 1/Extruder 1 on the Ultimaker S Series is another consideration, though I'm not sure it's possible. The specimen's total width (20mm) is a multiple of the line width (0.4mm),  negating the possibility of mismatched dimensions being the cause too.

 

I welcome any questions or suggestions from the community as I explore further solutions.

 

 

 

 

 

Edited December 11, 2023 by ttaiwo

[Dustin 129]

The gaps look like one of a few things.

Printing too cold on the extruder (or too fast)
Feeder tension not set correctly (guide: https://support.ultimaker.com/s/article/000002935)
Print core needs cleaning (guide: https://support.ultimaker.com/s/article/1667337928445)

[ttaiwo 1]

Hi Dustin,

 

Thank you for your suggestions. Currently, I'm using CURA 5.5, which features enhanced speed profiles for the Ultimaker S Series printers. These profiles reduce print times and if I am not mistaken one way they do that is by speeding up infills. However, the issue I'm facing was also present in CURA 5.4, which I believe predates these faster profiles. However if nothing else works I will revert to an older CURA version or reduce the print speed manually. 

 

I have previously adjusted the feeder tension to both 3/4 and full, starting from the assumed default halfway point. Unfortunately, this adjustment hasn't had any impact.

 

Today, I updated the printer's firmware. While I'm certain it will not resolve the issue, it seemed worth a try. Additionally, I've cleaned both print cores and replaced the bowden tubes for print core 1. I've left the feeder tension at 3/4 for now and I'll conduct three test prints at the temperatures of 200°C, 210°C, and 220°C and will update you on the outcomes.

[gr5 2,173]

There are many possible csauses of underextrusion.  Most of them are in the print core (but not all!). 

 

One quick fix is to just print 3X slower. 

 

Also make all the print speeds the same (typically you set the "speed" and this affects about 7 other speeds like infill and if you are doing strength tests and want solid infill then you should really print all at the same speeds - I think if you do engineering profiles it also does everything at the same speed - not certain).

 

However the speeds in the profiles were chosen carefully and it should be able to handle the default profile speeds as long as you don't mess with layer height and temperature in a profile.

 

So another quick fix is to swap out the print core.  They are considered "cheap" as they are cheaper than filament (if you for example get a new one every 5 spools of filament).  Every S3 should come with two AA 0.4 cores.  Did you try swapping?  Did you try duncan's suggestion of cleaning the core?  There is a process explained in the maintenance menu (cold pull).

 

The filament tends to caramelize into a brown or black crud on the inside of the nozzle.  It tends to be a pretty thin layer that reduces the nozzle diameter.  Or you can get chunks that partially clog the nozzle.  Cold pulls fix the chunks.  metal needles fix the layer blockage.

 

You can check in the menu how many meters of filament and how many hours have passed through a given print core.  Can you post that number here?

 

One quick test is to use the MOVE menu (start with middle menu on left and then top left icon and then "..." icon).  play with how many times you hit the down arrow per second and you should be able to get the filament from the nozzle to the bed in a little under 10 seconds.  Let us know how long that takes for you.  This can save the time of doing a print to know if you've fixed the issue or not.

[ttaiwo 1]

I gr5, 

 

Today I cleaned the print core using the hot pull and cold pull as instructed on the Ultimaker. I have swapped them around before but we have another Ultimaker S3 that has some bed levelling issues so I will swap a print core from there and order some brand new ones in the meantime. As I have recently swapped the print cores around the amount of filament and hours in use for both are:

 

Print core 1 (swapped): Time hot = 38d 17h, Material amount = 975.92m

Print core 2 (original):  113d 8h, Material amount = 2119.24m

 

I tried using the move menu and it seemed responsive. I do not think the under-extrusion is significant enough to notice by my human standards. I have been using the stock speed profiles with little modification so I thought the machines could handle it too, I am just about to throw a print on at 220 degrees C so I will try the 3X slower print after. 

[Dustin 129]

For clarification.. 
If your using the newer "Speed" profiles.. there really isnt much headroom in those profiles.. they are already pushing the print cores near their limits as far as thermals go.

If your increasing speeds of other profiles.. thats another story.

[ttaiwo 1]

Thanks for the suggestions I have been doing test prints since my last post using some of the suggestions and other things to try and solve the issues. Apart from reducing walls, top and bottom to zero and adjusting the infill direction I am pretty much using the stock normal 0.15mm layer print profile for the Ultimaker S3. In summary, what did not work:

 

- Printing at a temperature of 220 degrees C

- Print at 30mm/s

- Print at 25mm/s 

 

What has seemed to help:

- Drying the PLA in a dry box 

- Printing from a dry box

 

The best result so far is:

- Printing at 10mm/s (hence the delay in replying as these take 16hr +) 

 

So this is current where I am at the moment. So I have ordered some fresh Magenta, Blue and White Ultimaker PLA and will see if a fresh spool has this issue to see if it's a moisture issue. Printing directly from the dry box is not ideal because it restricts the filament path into the feeder, but drying it every couple of days is possible. 

 

Apart from trying a fresh filament spool, I am not sure what else to try. I found a forum mentioning a setting called Cura extrusion multiplier which sounds like this would help me with my issue but I cannot find this option in the printer settings I have found Flow Rate Compensation Factor and Flow Rate Compensation Max Extrusion under experimental however I am not sure what these are doing and how best to use them, any insight into these settings? Finally, I might try reducing the power of the part cooling fans as I am not doing overhangs etc. 

 

Thanks in advance

Edited December 18, 2023 by ttaiwo