Experimental feature to enable pulsatile purging in the wipe tower, aiming to improve purging efficiency

[igiannakas]

Replaces PR #6258 following multi extruder changes

Description

This PR aims to explore the use of pulsatile flushing to improve the efficiency of the purge process when a wipe tower is used for single extruder, multi material setups (like the ERCF/Tradrack/Prusa MMU etc).

The theory behind it is that varying the flow and pressure in the nozzle while purging assists in detaching the adhered to filament in the hot end walls due to wall shear stress and shear rate. It is inspired by a similar technique used in the medical field, as outlined in papers like this: https://www.sciencedirect.com/science/article/abs/pii/S0268003320302217
https://link.springer.com/chapter/10.1007/978-3-030-03149-7_19

This PR implements a similar push-pause technique extended for the practical limitations of 3D printing.

Screenshots/Recordings/Graphs

A new set of options has been introduced as shown below. That section is only visible when pulsatile purging is enabled:

The way it works is like this:

1. Print a set of lines on the purge tower at high speed - these do the flush part (this is limited to the max of the volumetric flow rate of the filament)
2. Retract with a long-ish distance and medium speeds, to disrupt the continuous laminar flow in the nozzle
3. De-retract with a hard coded slow speed to pressurise the nozzle slowly
4. Purge one line with the slow speed
5. Repeat until the purge volume is reached

The idea is that each purge activity is broken into segments of fast purge, disrupted by a retraction and de-retraction operation and a slow purge to catch any material that may be stuck due to shear stress on the walls of the hotend.

The wipe tower looks like this with the default settings:

You can observe the fast purge and slow purge lines as well as the pause move (retraction and de-retraction).

Tests

I've commenced testing this to validate % improvement in purge volumes.

Initial tests have shown that it is effective in cleaning up the nozzle - I've observed significant amount of debris coming out from the nozzle when using pulsatile flushing compared to regular one speed flushing. The below were done in back to back prints with an admittedly a nozzle that could clearly do with a few cold pulls to clean it up. The top tower is the regular purging method and it looked immaculate. Pulsatile purge is the bottom one, where the efficacy of the pulse is causing carbon deposits to get detached from the nozzle and pushed out during purging.

Purging volume test

Top - Pulsatile - Bottom - Default

Test was Mihai's color transition test, where the bottom X layers are white only, then the middle Y layers are color swaps from white to red and red to white and the top Z layers are red only.

Default:
Slight pink on the top of the lines

Pulsatile
Same color throughout

There is every so slight colour bleed from red to white on the default profile with 700mm3 purge volume. Same purge volume on the pulsatile purge and it appears clear.

PS. you'll need a color accurate monitor to see the delta as it is very slight but also very obvious in person (as the camera + SRGB color space is really not sensitive to red colors as much as the naked eye :))

[igiannakas]

Converting to ready to review. I've been using this for a few months now and it has reduced my purge volume slightly to be meaningful. Also the same concept has now been introduced in blobifier purge code for ERCF where around 30% reduction of purge volume has been observed.