Flow rate vs extrusion speed (eSteps)

Related to the discussion in the “Cant get prints to correct size” thread, I throught it might be intresting for some people to see the influence of extrusion speed to the actualy extruded filemant length.
This small test is inspired by the very good article here: Exploring Extrusion: Variability and Limits. But this article is valid for the extruder of an Ultimaker (which seems to be quite bad in 2013 compared to the TAZ), and I did the simplification to measure the length of filament when extruding 100mm at different speeds instead of measureing the weight of the extruded material.

First, what settings are used in my test:
Filament: eSun PLA (Diameter: 2.89mm)
Nozzle temperature: 205°
Nozzle diameter: 0.4mm instead of the original 0.5mm
Distance between the washers of the extruder lever: 5mm
My eSteps/mm I’m comparing to: 801

What I did was to prime the nozzle, then make a mark at the filament like you do for esteps calibration, extrude 100mm at different speeds and measure how much filament was driven through the extruder.

The table:
In the first column you can see the extrusion speed. Remeber, 100mm/min is the recommendation from Lulzbot for the esteps calibration. Next one is the flow rate, then the length of filament requestet (always 100mm) followed by the measured length.
The esteps/mm is the value that would be necessary to achive 100% (100mm), next one is the measured flow deviation because I fixed my esteps to 801.
The last column might be interesting for some people because you might be more used to think in print speed than in volume per second or extrusion speed. So you can get a feeling what x mm/min extrusion speed is in reality.

Here is the same data in a graph:

Yes, I’m a little bit under-extruding according to this test. But 801steps/mm gives me a perfect surface quality, so thats the deviation between theory and reality :wink:
Like in the linked article: Underextrusion is everywhere. But it’s not nearly as bad as for the old Ultimaker. But remember to use a extrusion speed for you calibration that matches your average print speed. As you can see in the table, depending on the choosen speed you may end up with something between 817 and 880 steps/mm, a quite hughe range!
To think one step further, it might be nice for some extruders (or if somebody is working with high speed differences in one print) to have an option in the slicer to auto-adapt the flow rate according to the speed…