Extremely inefficient slicing

[ooper 1]

I have a model that is somewhat like a maze about 4" x 4". The model gets sliced by Cura in such a way that it is moving around WAY more than it should. It will print a little bit, then move all around the maze (following the shape of the maze) to another part, print a little, then move to another part on the opposite side of the maze and so forth. It just keeps moving back and forth across the maze printing a small amount before deciding to move again. Any ideas on how to remedy this? Why can't it do all printing it needs to do in one area and then move on to another area?

[Labern 775]

What cura version are you using?

[nallath 1,124]

Slice it on a different operating system (yea, really. We use hash maps and those tend to give different results on different OSes)

[ultiarjan 1,223]

Really? no idea what a hash map is but Cura creates a different path per OS? that doesn't sound very desirable.

[neotko 1,417]

 

!

[nallath 1,124]

Only how it travels from one island to the other. There is no real alternative to this (save spending enormous amounts of time / effort in re-inventing the wheel)

[ultiarjan 1,223]

Raises an interesting question..... whats the most efficient OS ?

[nallath 1,124]

That's pretty much impossible to say. It's pretty much random (not true random, but close enough).

[ooper 1]

I am running Windows 10, Cura 15.04.

[gr5 2,230]

Consider unchecking "combing". This will at least move directly to the next area instead of following the inside of the maze walls and will save you lots of time. It might reduce the quality slightly but hopefully you won't notice any difference.

combing is among the retraction settings.

[ooper 1]

I will try unchecking combing, although it seems like w/o combing I had trouble with the feeder grinding down the filament due to lots of retraction. Thanks for the suggestion though. I will let you know how it goes.

Of course, a better solution would be more intelligent slicing to avoid moving across the print so frequently to do just a tiny bit of printing between each move. I don't understand the algorithms used to determine slicing, but there appears to be some room for improvement there.

[gr5 2,230]

Picking the most efficient order of infill is very very complicated. What seems simple to a human is incredibly hard to do in an algorithm. This algorithm was just explained in another set of posts just over the last few weeks - there is this tree structure or grid structure where the slice is grided into 4 squares and then those into 4 subsquares and repeated on down to a small size and every vertex is placed in a square and when it finishes with a line of infill it uses this to find the closest line to do next.

And there's other stuff I forgot already.

Anyway it tends to pick the closest line to do next ignoring the fact that if combing is on it may have to travel a LONG way around.

I think a better algorithm might be to try to do the lines from Y=0 to Y=200 or something like that (top to bottom basically). At least anytime the next line is more than 1cm away.

[ooper 1]

Thanks for your help here gr5. Usually in situations like this, there is either not enough "pain" being experienced by the community to warrant the time needed to fix/improve the algorithm or not enough pain being experienced by the people who have the power to make the change. I'm not bad-mouthing anyone. I am a software developer myself, so I know that the available resources are usually applied to where it makes the most business sense.

It's just unfortunate for me that this issue and a handful of others have caused me to have to look elsewhere for my 3D printing needs. Don't get me wrong, I love the UM2 and will continue using mine, but it has been difficult keeping my three going. I suppose they are not really meant to be running 24x7x365

[ooper 1]

Oh and by the way, I turned off combing and I was able to print successfully! There was a little extra cleanup needed on the finished print with a small file, but the feeder did not grind the filament to a point of stopping the feed.

[gr5 2,230]

There's an inexpensive slicer called Simplify3D that almost everyone who tries it loves.

[ooper 1]

Yes I have it I use it too and will most likely be using it exclusively very soon.