Next question…
Is this simple print enough to expose failures in printer control algorithms and hardware? The upper part is smooth and strong. The lower part shows a variety of artifacts.
This example is PLA printed at a nominal 100mm/s (fast) and 240C (hot). There are slicer (Cura) artifacts at the tips of the fins/nacelles. This is a “vase mode” (in Cura “spiralize”) print, so a single wall. Printed with a 0.4mm nozzle, 0.3mm layers, and 0.8mm line width.
The more complex paths of the lower parts mean the print head is changing speed quite a lot. This (should) mean changes in thermal input (and output) to the nozzle. The rate at which heat travels through the print head is much(!) slower than movement. Pretty sure the printer control algorithms cannot keep up.
Rate of travel should be gated to ensure enough heat for good bonding between layers. Heat pumped into the head should be sufficient to allow high rates of travel.
Clearly, this is not happening.
You have combined issues in this. try to make a shape that only exposes one flaw at a time, that way the problem with the software can be identified and solved one issue at a time. These things only get solved when the problem is repeatable at the same place, every time. If its an intermittent problem then the software people will say its not their fault.
@MidnightVisions Well, your interest is isolating causes is well-placed, but printers need to deal well with differential movement. This is a fairly simple shape, though the interaction is between movement and thermal input is decidedly not simple.
This is a case 3D FDM printers need to be able to routinely resolve, though perhaps not in the range of 8-bit controllers.
Try Marlin 2.x on a 32bit controller, with Linear Advance pressure algorithm enabled.
This is mechanical issue. Not rigid frame. When printing circle path, frame not loaded. When print complex shape, head suddenly change direction the frame twisted.
(See Occham’s razor law: Among the explanations given to things, the simplest is the true one)
Slicers really don’t have any way to handle hot end thermal dynamics effects. Too many complete unknowns (unknown to both the slicer and user.) Firmware would need to do that. Even there, it’s not trivial. It’s really up to the end-user to provide a combination of flow rate, tempset point, and material that is feasible.
@Preston_Bannister if you want a pathological print for firmware motion planners, look for the high-res scan of Michelangelo’s Pieta face of Mary on Thingiverse. High poly count plus organic curves with varying tortuosity. Acceleration schemes that trigger slowdowns on corner sharpness (IE almost all of them) can’t handle that kind of geometry. They don’t recognize a small-radius arc with high facet count as a corner. Plus a high segment rate will bog down a lot of controllers. (Slicers have different segment decimation algorithms built in though.)
If you want a pathological print for thermal management, just do a long, thin box shape (like a 100mm x 4mm wall) with fairly dense infill and a few perimeters. During the long perimeters, you get high extrusion rates. During the short infill runs, you get sustained acceleration slowdowns. The transitions from one to the other will show thermal control issues.
@Ryan_Carlyle Bust of Mary (from Pieta by Michelangelo), almost no support
