Question What are the criteria for combing to work?

[bcsteeve 1]

I've got my printer pretty dialed in.  I use Cura 5.2.1.

 

I go to print a standing lithophane for my kid, and I get some pretty bad stringing.  Weird.  I figure it is the filament so I go to a known good roll... same thing.  Investigating, I look at the travel lines and I see a ton of them!  There shouldn't be any, or very few, as I have combing set to All and I've never really had stringing issues with that enabled.

 

I play around a bit, and I notice if I scale the model 300%, then almost all of the travel lines disappear (or, rather, they follow the route of the model as they're supposed to with combing).  But at 100% or even 200% I get a ton of straight across travels, which is very much counter-intuitive to that setting.

 

I gather it has something to do with the thin walls of the lithophane, but is there some requirement for combing that a wall be a certain thickness?  If so, what is that thickness?

 

I've uploaded my project which is after I've done my best to minimize the travels, and it is still really bad, but at least contained to the back where I can clean it up easily enough.  Without tweaking the z-seam alignment and increasing the minimum thickness of the model, it was like a spider's web!

 

I've done lithophanes before without an issue, so I'm not sure why this is a problem all the sudden. 

 

stringing.3mf

[GregValiant 1,318]

Hello @bcsteeve.  I took a look and yeah, there is a lot of travel for such a small print.

One of the problems I see is the geometry of the model itself.  The Inner Walls stop and go so there is going to be travel between the areas where there is Inner Wall to print.  The Infill only exists on the ends and Cura prints Infill separately so there is travel from one end to the other to complete the infill on each layer.

• You can try going to 4 walls and 0% infill.
• You have "Max comb distance with no retract" set to 30mm.  Drop that to 3mm to force more retractions.  That should allow you to turn off "Retract before outer wall".
• I see you have the retraction distance set to 2.0mm.  You have a direct drive on your printer?
• The setting "Avoid printed parts when traveling" tells Cura to compute the travel moves off the part.  I think you would be better off with as many combing moves as possible within the part.
• In Travel - set the Layer Start X and Layer Start Y to the same numbers as your Z seam location.

With your current settings there are about 2040 retractions in the gcode.  Making the changes above there are about 8000 retractions.  Almost all of those additional retractions are before travel moves and that should seriously cut down the amount of strings you are getting.

 

I see you have a lot of Flow settings at 85%.  Is there a reason you need to under-extrude by so much?  Single wall calibration cube perhaps?  This is not a single wall print.

 

 

 

[bcsteeve 1]

Hi @GregValiant

 

Firstly, thanks so much for taking the time to help.

 

Quote

• You can try going to 4 walls and 0% infill.

I actually did set it to 99 walls prior to printing... typical for lithophanes.  I guess I reverted that prior to saving the settings I uploaded.

 

Quote

• You have "Max comb distance with no retract" set to 30mm.  Drop that to 3mm to force more retractions.  That should allow you to turn off "Retract before outer wall".

Thanks.  I'll give that a shot

 

Quote

• I see you have the retraction distance set to 2.0mm.  You have a direct drive on your printer?

Correct

Quote

• The setting "Avoid printed parts when traveling" tells Cura to compute the travel moves off the part.  I think you would be better off with as many combing moves as possible within the part.

I wondered about that.  I tried with and without and it seemed to make no difference.

 

Quote

• In Travel - set the Layer Start X and Layer Start Y to the same numbers as your Z seam location.

That's something I hadn't thought of!  I'll give that a shot as well.

 

Quote

I see you have a lot of Flow settings at 85%.  Is there a reason you need to under-extrude by so much?  Single wall calibration cube perhaps?  This is not a single wall print.

Ok, so extrusion is something I've only recently bothered trying to calibrate, and yes I was going by a single wall calibration cube as per... well... just about all the tutorials out there!  Of course this is not a single wall print, but surely the calibration universally advised isn't just for single-wall printing, right?  I do have to say, I've had much more accurate printing results since changing to the lower figure.  85% is actually one of my higher ones.  Most of my rolls calibrated out to 82 to 83.5%.  With that said though, I have noticed some under-extruding issues, and this model was a prime example, in that there are little pin holes in the thinner areas that shouldn't be there.

 

Do you have a tip on best practices if you don't think I'm doing it right?

 

Thanks!

[bcsteeve 1]

I tried all the settings changes... this is still the preview, which has me thinking it'll still be a spiderweb mess 😞

 

That's pretty much what the preview looked like before those changes.

[GregValiant 1,318]

Because of the discontiguous areas of Inner Wall - the nozzle has to jump all over.  Making sure that there are retractions before any travel moves should eliminate a lot of the stringing.  With those changes you will get almost 4 times as many retractions and most of the additional retractions will be before the long travel moves and so the stringing should be a lot less.  Eliminating the infill will eliminate at least two combing moves per layer.

 

I have a thing about using single wall calibration cubes.  It boils down to this...when the "Volume of Filament In" is equal to the "Volume of Extrusion Out" then by definition the "flow" is 100%.  You calibrate your e-steps to achieve that.

Everyone who tries to calibrate using a "single wall cube" ends up at around 85% flow.  So where is that 15% shortage of plastic coming from?  It can't be from the length of the extrusion which is fixed, it can't be from the layer height which is fixed, so it can only come from the width of the extrusion.  If your line width is 0.5mm then the index distance between adjacent extrusions is 0.5mm.  You need 100% flow to fill the void below the nozzle.  Calling for 85% flow means your extrusion is only .425 wide.  A .425 wide strip of plastic will not fill a 0.5 wide gap.  That (by definition) is under-extrusion.

The best way to judge "Flow" is a skin.  Scale a flat topped calibration cube to 100 x 100 x 1 and set all your flows to 100%.  Make your top/bottom line directions [0,90] and print the file.  About 1/2 way through the top layer tune the flow with the LCD to 85%.  When it's done look at the top layer through a magnifying glass (or better, a microscope).  I'll betcha the 100% flow side looks better than the 85% side.   Under a microscope you might even be able to see the layer below through the 0.07 gaps between extrusions in the top layer.

Now, if you ever have a need to print a model that is single walled and it is absolutely necessary to have the walls exactly .4mm (or whatever) wide then by all means calibrate for that specific condition.  For everything else, calibrate the e-steps and you are done.  Some materials may need a bit of tweaking but 15%?  Probably not.

[bcsteeve 1]

That sounds perfectly logical.  

 

Often times, the 3D printing community is the blind leading the blind, and sometimes certain things that appear to work (or make some logical sense) becomes pervasive, and when you see the same guide/advice repeated through the Internet, I suppose it is sensible to assume there's some truth.  But that doesn't mean it necessarily is!

 

I admit I didn't put a lot of thought into it.  I was printing some guy's puzzle box and everything was jammed up.  He helpfully had a test print and all signs were pointing to over-extrusion.  I found one of many guides and calibrated e-steps then did the single-wall cube test and ended up with my ~85% settings... and the puzzle box worked.  Does that mean my settings are right?  Not necessarily.  Perhaps he subscribed to the same flawed logic and has his set to 85% (or whatever) and tweaked his designs for that result.  Who knows?

 

But I am happy with my current settings other than this stringing, which I'm also now happy to report I've resolved.   I had my retract distance at 2.0mm as you noted, and I often saw advice to start at 2mm and adjust by 0.1mm as needed, but nothing really said which direction.  I suppose I thought if zero meant bad stringing and 2mm was a good starting point, that the proper direction would be to go higher.  I tried 3mm, 4mm... nothing helped.

 

Finally I printed a tower just now with 2.0, 1.6, 1.2, 0.8, 0.4 and 0... clearly there's a winner!

 

So really, my problem is solved.  It wasn't the question I asked, but I'm good.  And I very much appreciate your assistance!

 

 

[GregValiant 1,318]

"...the 3D printing community is the blind leading the blind,"  I'm retired and I bought my Ender to give me something to do while stuck in the house during Covid.  Searching for advice I first came across the Reddit group and yeah, it was a bunch of folks guessing.  Finally I came across this forum and the people actually helped and gave advice that was sound and wonder-of-wonders it actually worked.

When printing articulating devices, or print-in-place models, or pieces that must fit together, much has to do with the tolerances built into the model.  In the case of that puzzle box making a change to the flow made sense and worked, but was it a fix "for all things" or a patch to get past a glitch in the tolerances designed into that particular model.

Cura calculates the extrusion area as a simple rectangle and so the volume of an extrusion is just the area times the length.  If you have accurately measured your material diameter (my Matter Hacker filament is consistently 1.72mm diameter) and entered it into the "Diameter" setting within the Printer Settings plugin, and your E-steps/mm is calibrated, then you should be good.

On a single wall print the extrusion is un-confined on both sides and so it isn't a rectangle but rather a bulged out flattened oval.  In that particular case then yes, you will be measuring the bulge and yes, it is wider than your set "Line Width" BUT the volume is right, the cross section of the extrusion is just out of shape.

 

Also consider that the Line Width has nothing to do with the width of the actual extrusion but is rather the Index Distance between adjacent extrusions.  Now the result of that Index Distance is indeed a rectangle of particular dimensions.  It's the "Flow" that fills that rectangle.  I'm just saying that at 85% Flow (and a lot of people end up at 85% after using a single wall calibration cube) the rectangle isn't full.

"I bought a 12oz beer - how come there is only 10.2oz in my glass?"

If all of that sounded like a well practiced rant, it's because it is.

 

Edited December 14, 2022 by GregValiant

[bcsteeve 1]

Yep, what you say makes perfect sense.  I appreciate your taking the time.