How Is Extrusion Width Calculated?

[inventabuild 5]

There is no explicit field for me to enter the extrusion width, so I'm wondering how Cura figures it out? Also, is the formula for extrusion width different for the first layer vs the other layers?

 

[Dim3nsioneer 558]

What about the nozzle size?

 

[aviphysics 9]

I am pretty sure in Cura, that the extrusion width is the "nozzle size". If you make that number bigger, it will put out more filament and you will get a wider line width. Obviously, making it much narrower than your nozzle width is not going to yield great results (think about what under extrusion looks like. Making it wider than the flat surrounding the nozzle is also not that great.

IIRC, the material feed rate is going to be the ("nozzle size")*("layer thickness")*("print speed"), meaning that Cura calculates the path as if it were a long rectangle (ignoring the fact that the ends of the lines are more like semicircles.) I think one of the purposes of the "infill overlap" setting is to help correct for that.

 

[inventabuild 5]

My nozzle size is 0.35 mm, but the Cura's gcode is giving me an extrusion width of 0.44 mm. My layer height is 0.22 mm and my first layer height is 0.3 mm.

Any ideas why my extrusion width is coming out to 0.44 mm with a 0.35 mm nozzle?

 

[inventabuild 5]

by the way, it's the first layer that is giving me a 0.44 mm extrusion width.

 

[aviphysics 9]

From the nozzle diameter, it sounds like you are not using a UM or UMO. What printer are you using?

In the meantime...several things things.

1) First layer line width is easily messed up by the relative height of the platform, so it could be that. Does the fill on the first layer look like you are putting out too much plastic? Try printing something with a large flat solid bottom and take a look. The plastic will get squished into little ridges if there is not enough room between the hot end and platform (this seems like the most likely circumstance, if your problem is just with the first layer.)

2) There is a separate setting to adjust first layer line width ("initial layer line width" (%)".) You should make sure that is set correctly. If this is the case, then the bottom layer would probably look somewhat smooth, unlike the first case, but with wider lines than the other layers.

3) If the filament diameter settings is wrong, your printer will put out more or less filament than it thinks it ought to. This would affect all layers, and the fill in any layer with 100% fill will look messed up, like described in the first case.

4) If your "extruder steps/mm" on your printer's firmware is wrong, it will put out more or less filament that it ought to. This would affect all layers, and the fill in any layer with 100% fill will look messed up, like described in the first case.

5) If Cura was configured in a way that made it think you are using an UM2, it will send the printer the volume of material to use instead of distance to move the filament. If you are not using a printer that expects this, it will make your printer use the wrong amount of filament. This would affect all layers, and the fill in any layer with 100% fill will look messed up, like described in the first case.

 

[inventabuild 5]

Thank you for the suggestions. So I did not see the extrusion width variable in Cura's output gcode so the 0.44 mm first layer width was a calculated value by looking at the gcode finding two parallel lines and adjacent loop segments, comparing the coordinates, etc to make the calculation. The 0.44 mm first layer width is not a measurement by eye, sorry for not making this clear in my postings.

My layer height is 0.22 mm and first layer height is 0.3 mm are data values I entered into Cura's respective fields.

I am using a TAZ 4 with Cura 14.09 which is Lulzbot's Cura edition: https://www.lulzbot.com/cura

 

[inventabuild 5]

...and I have nozzle size set to 0.35 mm.

 

[aviphysics 9]

If it is just in the first layer, I would consider that there could be something wrong with the first layer line width setting. If it isn't that, then it is pretty hard for my to guess.

You could try generating a gcode file from a vanilla cura version, and see if it has the same line width problem. The problem could potentially be isolated to the lulz bot edition.

 

[gr5 2,270]

Cura uses a combination of two things to determine the width of the "traces" or "extrusion width" as you call it. I'll abbreviate to "ext-wid".

nozzle and shell thickness.

If shell width is a multiple of nozzle width then the ext-wid is the nozzle width. So .35mm nozzle, .7mm shell gets .35mm ext-wid. If the values aren't multiples then it's more complicated. Cura tries to make an integral passes to get the correct shell width. So if shell width is say .88mm and nozzle is .35mm Cura might choose either two .44 passes or 3 .293 passes. Cura chooses the two .44 passes. It will go all the way up to 150% of nozzle size for ext-wid before switching to 3 passes. So a 1.06mm wall will require 3 passes of .35333mm each.

 

[gr5 2,270]

I had to leave suddenly - anyway cura always picks a larger ext-wid than the nozzle - up to 2X. Never smaller. So if you ask for a shell of .69 with nozzle of .35 it does it in one pass with equivalent flow of almost 200%.

 

[Daid 306]

https://github.com/daid/Cura/blob/SteamEngine/Cura/util/profile.py#L1117

 

[aviphysics 9]

https://github.com/daid/Cura/blob/SteamEngine/Cura/util/profile.py#L1117

 

What is the "-0.0001" supposed to do? It would seem like in almost any situation where it would make a difference, lineWidth would already be greater than 1.5*nozzleSize, ensuring that linecount would yield the same result with or without that subtraction. Maybe I am missing something.

 

lineCount = int(wallThickness / (nozzleSize - 0.0001))

[Daid 306]

You're missing rounding errors :-)