Skip to first unread message
funBart
Dec 18, 2013, 6:00:57 PM12/18/13
to kisslicer-r...@googlegroups.com
Bertho, I had the same issue with using Kisslicer on the Ultimaker. I use the printed version of your extruder upgrade and have the "flow tweak" under the materials tab set to 1.15. It is 1 by default. You are right on about the calculations of wall thickness and extrusion width, number of loops, they have to match. I use 0.5 and 3 passes to make perfectly adhered hollow walls that range from 1.5-1.6mm when measured with calipers.
Bertho
- Newbie
- Posts: 9
OK, I will increase the max flow rate to see if it helps.
I will also run the test object with the different wall thicknesses.
Bertho
I will also run the test object with the different wall thicknesses.
Bertho
PenskeGuy
- Hero Member
- Posts: 1249
- Any Day I Learn Something New Is A Good Day
I'm thinking from your description that the 2nd and 3rd Loops may be starving. The head speed is significantly higher on the inner Loops than it is for the Perimeter. If the Max Flow Rate (mm^3/s) for the particular material is not high enough in relation to the speed set in the Printers > Speed Tab > Solid Infill and Support (or Sparse Infill--can't remember which right now--up all night... again) and/or the temperature is too low for that faster speed, it would cause this. Might try sneaking the Max Flow Rate up a bit and see if there is any change.
« Last Edit: March 11, 2013, 11:54:44 AM by PenskeGuy »
Bertho
- Newbie
- Posts: 9
I am extruding PLA from a 0.4mm nozzle 0.2mm layer and multiple samples of extruding in air gives me 0.42-0.44mm.
Printing 2mm walls with the perimeter set to 3 loops, I get individual free-standing vertical walls typically 0.35mm thick.
I do not know why it is thinner. The printer ought to smear out the filament to make it wider and stick together.
Testing extrusion setting a solid cube comes out OK.
For time being I have temporarily solved my problem in case of 2mm walls: pint single perimeter and then it will fill the in-between area.
Bertho
Printing 2mm walls with the perimeter set to 3 loops, I get individual free-standing vertical walls typically 0.35mm thick.
I do not know why it is thinner. The printer ought to smear out the filament to make it wider and stick together.
Testing extrusion setting a solid cube comes out OK.
For time being I have temporarily solved my problem in case of 2mm walls: pint single perimeter and then it will fill the in-between area.
Bertho
PenskeGuy
- Hero Member
- Posts: 1249
- Any Day I Learn Something New Is A Good Day
OK, I'll try to explain this better. I've gone through this a lot. I do have to say that I have not fully ruled Crowning out as a solution to this--getting stuff done is more important than testing for the sake of testing. Maybe when there is the luxury of time...
When using only Loops on a fully modeled thin wall part, and the distance between the Loop passes is right at a certain dimension, IOW, the inner pass of the Loop is close to the outer pass, nothing will go in there. Infill just looks at it and says: "Nothing for me to do here" and completely skips over it. Overlap won't do anything because Infill isn't there to influence. There isn't enough room for a second Loop, so there is a tiny gap. Drop the wall thickness and the two passes touch each other. Increase the wall thickness and more operations will take over and fill the gap; either with more Loops or Infill, but only after a certain threshold is passed.
I've had some walls that the inside and outside were intentionally not concentric. You could see them touch on one side, a gap opens part way around, Infill starts to do its thing diametrically opposite where the Loop passes were touching. Like I said, Crowning set at 0 may take care of this. I'll try to work this into my schedule, because I'd like to find a solution, too. At first, I was trying all manner of things to fix it. Then, I happened upon Dave's test and his results matched mine perfectly. Until I find a workaround, I avoid the 1.8mm-2.1mm thickness range if I want the walls to be extra strong.
When using only Loops on a fully modeled thin wall part, and the distance between the Loop passes is right at a certain dimension, IOW, the inner pass of the Loop is close to the outer pass, nothing will go in there. Infill just looks at it and says: "Nothing for me to do here" and completely skips over it. Overlap won't do anything because Infill isn't there to influence. There isn't enough room for a second Loop, so there is a tiny gap. Drop the wall thickness and the two passes touch each other. Increase the wall thickness and more operations will take over and fill the gap; either with more Loops or Infill, but only after a certain threshold is passed.
I've had some walls that the inside and outside were intentionally not concentric. You could see them touch on one side, a gap opens part way around, Infill starts to do its thing diametrically opposite where the Loop passes were touching. Like I said, Crowning set at 0 may take care of this. I'll try to work this into my schedule, because I'd like to find a solution, too. At first, I was trying all manner of things to fix it. Then, I happened upon Dave's test and his results matched mine perfectly. Until I find a workaround, I avoid the 1.8mm-2.1mm thickness range if I want the walls to be extra strong.
« Last Edit: March 03, 2013, 04:40:47 PM by PenskeGuy »
funbart
- Sr. Member
- Posts: 307
Curious. Most of the time when not having set the filament diameter correctly, it's resulting in the layers not attaching, not the walls.
As rule the nozzle diameter is also a good starting point for the path width.
It could be that with this model, Kisslicer is spacing out the loops, resulting in a just to large distance? But Crwoning threshold should solve that (set to 1mm).
Also, it can be only the loops have to much distance to the infill/solid? You can make them more overlap under
the Printer settings, subtab Hardware. Adjust the slider "loop / infill overlap" to the right, resulting in more attaching of infill and loops.
But with a photo, probably more to tell.
Bart
As rule the nozzle diameter is also a good starting point for the path width.
It could be that with this model, Kisslicer is spacing out the loops, resulting in a just to large distance? But Crwoning threshold should solve that (set to 1mm).
Also, it can be only the loops have to much distance to the infill/solid? You can make them more overlap under
the Printer settings, subtab Hardware. Adjust the slider "loop / infill overlap" to the right, resulting in more attaching of infill and loops.
But with a photo, probably more to tell.
Bart
Makerbot Replicator 1 dual
Minimalistic Mk8 replacement by Emmett
Sailfish Firmware
Kisslicer Pro 1.1.0 Alpha
Minimalistic Mk8 replacement by Emmett
Sailfish Firmware
Kisslicer Pro 1.1.0 Alpha
max72
- Full Member
- Posts: 148
It could be the extrusion width is less than expected?
Maybe the filament is not the nominal diameter but slightly less, or your material - print combination is behaving differently. For instance orbi pla90 has the opposite problem: if the temperature is too high it "boils" increasing the dimensions.
Massimo
Maybe the filament is not the nominal diameter but slightly less, or your material - print combination is behaving differently. For instance orbi pla90 has the opposite problem: if the temperature is too high it "boils" increasing the dimensions.
Massimo
Bertho
- Newbie
- Posts: 9
Thanks Massimo:
I was experimenter last night and realized part of the problem: I have 2mm walls and a 0.4mm nozzle and extrusion set to 0.4mm
I also specified 3 perimeter loops. Basically I got 5 loops of 0.4mm to get the wall thickness 2mm.
They were beautifully vertical walls like paper side by side but in most areas they were not sticking to the adjacent wall. In other words, layer to layer works great but side to sides are not adhering.
After realizing the 3 loops issue I did a test with a single loop and specified 100% fill and that worked well and as expected. That turned out to be what you suggested.
That still does not explain why the vertical walls are not sticking together. I have no picture since to I needed the part and I used super-glue to wick between the layers and lock them together.
Bertho
I was experimenter last night and realized part of the problem: I have 2mm walls and a 0.4mm nozzle and extrusion set to 0.4mm
I also specified 3 perimeter loops. Basically I got 5 loops of 0.4mm to get the wall thickness 2mm.
They were beautifully vertical walls like paper side by side but in most areas they were not sticking to the adjacent wall. In other words, layer to layer works great but side to sides are not adhering.
After realizing the 3 loops issue I did a test with a single loop and specified 100% fill and that worked well and as expected. That turned out to be what you suggested.
That still does not explain why the vertical walls are not sticking together. I have no picture since to I needed the part and I used super-glue to wick between the layers and lock them together.
Bertho
PenskeGuy
- Hero Member
- Posts: 1249
- Any Day I Learn Something New Is A Good Day
Dave White, over on the BfB forum, performed a test a couple of years ago using Axon1 as the slicer:
This was on a RapMan 3.1, so results would vary using newer hardware and slicer.
I've attached his test files, so you can run one on your machine.
Quote
Over the last few weeks I?ve been playing with all sorts of settings and software to try to really get to grips with what they really mean. I got frustrated by doing large prints so I designed a couple of objects along the way that allowed me to see results quickly. So I?ve attached them here so others can use them.
The square and diamond are simple 25x25 parts that allowed me to check on solid surfaces (shells and surface layers at 0, 45, and 90 degrees.
But what was really starting to frustrate me was printing objects with thin walls (containers). Walls that were really thin were not printed, quite thin walls ok, and thick walls were no problem? it was the ones in between.
So I designed a ?Walltester? with walls from 0.1 to 2.0mm in 0.1 increments and also 2.5 and 3.0mm for good measure. This allowed me to see..
The smallest wall size that is processed
Wall that print as a single line
Walls that print as 2 lines side by side
Walls that print as 2 lines with a gap
Walls that print as multiple lines
Walls that print as multiple lines with zig zag infill
So for example with BfB Axon default ABS setting (0.25mm layers) set to 4 solid surface layers ,3 shells and 40% fill density I get the following?.
The smallest wall size that is processed =0.6 (0.1 ? 0.5 missing)
Walls that print as a single line = 0.6
Walls that print as 2 lines side by side 0.7 to 1.3
Walls that print as 2 lines with a gap 1.4 to 1.8
Walls that print as multiple lines 1.9 and 2.0
Walls that print as multiple lines with zig zag infill 2.5 and 3.0
This was on a RapMan 3.1, so results would vary using newer hardware and slicer.
I've attached his test files, so you can run one on your machine.
thin-wall-test-objects.zip (58.34 kB - downloaded 16 times.)max72
- Full Member
- Posts: 148
Do you have a picture?
If you have an "external" cluster and an "internal" one, with a single space in between this is due to the fact tha the inside space is less than extrusion distance.
In this case either use 1 in the number of loops settings (and in between use high infill), or play with crowning.
What do you have as infill settings?
Massimo
If you have an "external" cluster and an "internal" one, with a single space in between this is due to the fact tha the inside space is less than extrusion distance.
In this case either use 1 in the number of loops settings (and in between use high infill), or play with crowning.
What do you have as infill settings?
Massimo
Bertho
- Newbie
- Posts: 9
I am using an Ultimaker and I am new to KISSlicer but I have seen this problem with other slicers:
For example, when printing the twisted salt shaker the inside wall is not attached to the outside wall.
Yesterday I printed a big box, 170 x 100mm and 6mm tall with 6 vertical dividers. The wall thickness is 2mm.
In most places the walls consisted of side by sides thin vertical layers, beautifully uniform but not attached to each other sideways.
I got a 0.4 mm nozzle and printed 0.4 wide.
The box looks perfect otherwise.
As a test I just printed the the solid 20 x 20 x 10mm cube it it looks normal so my filament and settings are presumably OK.
Bertho
For example, when printing the twisted salt shaker the inside wall is not attached to the outside wall.
Yesterday I printed a big box, 170 x 100mm and 6mm tall with 6 vertical dividers. The wall thickness is 2mm.
In most places the walls consisted of side by sides thin vertical layers, beautifully uniform but not attached to each other sideways.
I got a 0.4 mm nozzle and printed 0.4 wide.
The box looks perfect otherwise.
As a test I just printed the the solid 20 x 20 x 10mm cube it it looks normal so my filament and settings are presumably OK.
Bertho
Reply all
Reply to author
Forward
This conversation is locked
You cannot reply and perform actions on locked conversations.
0 new messages