Thanks for the insight both of you! I have a lot of orders to finish with the 0.4 mm nozzle currently installed, but after that I'm definately switching to 0.8 to do some testing with larger objects and faster printing.
- 2 weeks later...
"There is a setting called "jerk" which is not jerk but instead the max change in instantaneous velocity at a vertex. In other words take the vel vector before and after a vertex and subtract them. "
I can't blame GR5 for this error, as the term jerk is rarely used in mathematics. The change in velocity is called acceleration, and the change in acceleration is called the jerk. While almost everyone knows what acceleration is, jerk is usually reserved for son-in-laws. It is the third derivative of a curve. It is also called jolt surge or lurch. It is the only time I have seen this derivative used for consumer products Being able to specify this change in acceleration demonstrates the precision needed in a good 3D device.
For those of a masochistic academic nature:
fourth derivative -- jounce or snap -- change in the jerk
fifth -- crackle -- change in the jounce (and as follows logically)
sixth -- pop
seventh -- lock
eight -- drop
I have never seen (or used) a high order derivative that jounce.
I hope someone is still awake should they have read the post.
I put jerk in quotes because it's not the right word but it's the word that Marlin uses in the code and on the settings in the menu system. So when Marlin or Ultimaker printers talk about "jerk" they are talking about an infinite acceleration - an instant change in velocity at a vertex (there's a lot of vertexes in a print - one for each gcode approximately). The same word: "jerk" is used unfortunately in the firmware for: Marlin, Repetier, Sprinter, Redeem, and more.
The jerk setting on UM by default is 20mm/sec. So if you are going into a 90 degree corner at 100mm/sec it will slow down to actually 14.14mm/sec at the corner (it's complicated).Edited by Guest
Marlin does this. There is a setting called "jerk" which is not jerk but instead the max change in instantaneous velocity at a vertex. In other words take the vel vector before and after a vertex and subtract them. The magnitude of the difference vector is your speed change. That is kept to 20mm. So for 90 degrees the speed goes down to 14mm/sec. For sharper angles (say 180 degrees) that goes down to 10mm/sec). For very very slight angles the limiting vertex speed is quite high. Maybe as high as 1000mm/sec (which is faster than the max speed (300mm/sec) of the printer so it won't limit at all.
Link to post
Share on other sites
One thing I noticed over the weekend regarding nozzle sizes is this:
For a given layer height, switching to a larger nozzle size caused more filament to be used. At least that is what Cura 2.1.2 estimates.
For example, if I say I want 0.25 layer height and I use the 0.4mm nozzle, I get, say, 17m of filament estimated will be used. Keeping all else the same, and changing to the 0.8 nozzle, I get around 25m of filament that will be used.
I suspect this is because the infill line width will be that much wider, using more plastic. It probably would contribute to a stronger part too, but that is just a guess.
If you want to keep the amount of filament used down (at the cost of speed), then the smaller nozzle might help. ... Assuming Cura's estimates are reasonably accurate.
I did not think of upping the layer height beyond 0.25, however. So, if 0.4mm layer height is feasible with the 0.8mm nozzle, that might help offset things. Again, just a guess.
Anyway, this is just an observation I noticed. Hope it helps!
Link to post
Share on other sites