Help Cura slicer speed inconsistency

[AlexHall304 0]

Since last week I've been analyzing prints with the color schemes that are integrated within the cura slicer.

However, I noticed some strange inconsistencies within the speed of cylinders. I have added cylinders and a flow test for purposes of testing my Ender 3.

The strange thing is when I look at the speed, the speed of the cylinders all of the sudden drop a significant amount when the flow test is done printing.

 

(Another strange thing that I noticed is that the walls do not have any inconsistencies in speed when there is a draft shield enabled)

 

My question is, does anyone know why this is happening and does anyone know how to solve this? I'd be really grateful :)

 

(I included screenshots with regular printing speed, absurdly high printing speeds and a print with draft shield enabled).

 

[GregValiant 982]

The draft shield takes time to print.  At the bottom and the top of the model - the skins take time to print.  So what is going on is that you are bumping up against your "Minimum Layer Time" and Cura is calculating the speeds for the intermediate layers so as to maintain the Minimum Layer Time (and with a nod towards the "Minimum Speed").

If you want to maintain the speed across all the features then you might want to add either the draft wall, or a sacrificial calibration cube to insure that all layers take more than the Minimum Layer Time.  If the layer time is too short then the plastic doesn't have time to harden and the print can turn into mush.

 

To calibrate "Flow":

1. Calibrate your E-Steps/mm and make adjustments in your printer as necessary.
2. Measure the exact diameter of the filament and enter it into Cura.
3. Look at the prints and let them talk to you.

Any other "flow calibration" is at best confusing and at worst just plain wrong as very few real-life models re-create the exact specific conditions of a flow model.  If an extrusion requires 5mm³ of plastic then the printer should deliver 5mm³ of plastic and not 80% or whatever number some "Flow Model" told you it should be.