This model has a small handle at the top, which I cannot print satisfactorily.
There seems to be open space inside the handle (in particular right inside the outline) and thereby lack of support even though I selected 100% fill and Minimum Infill Length 0.5 mm. (Shorter Minimum Infill Length made, the filament feed stop, so that the upper part of the handle was "printed" with air.)
I tried many settings, but obviously need help to hit the right parameters.
I include the factory file for the exact object on the picture.
I very much appreciate any hint.
Best regards,
Niels
Flashforge Dreamer
Detail won't be good
- Attachments
-
Re: Detail won't be good
Hi Niels,
to me this looks like massive overheating due to very small features and printing only this single part/instance.
Try to print two of them a few cm (maybe 8-10cm) apart and make sure your part cooler is running when reaching the top. It may also be helpful to have a look at "Print islands sequentially without optimization" on the "Layer" tab of your process. Just slowing down the print speed won't help as you need the head to move away to give the small features some time to cool down, especially when printing PLA. 20% - 30% infill should be enough here.
https://www.simplify3d.com/support/prin ... verheating
bye,
Daniel
to me this looks like massive overheating due to very small features and printing only this single part/instance.
Try to print two of them a few cm (maybe 8-10cm) apart and make sure your part cooler is running when reaching the top. It may also be helpful to have a look at "Print islands sequentially without optimization" on the "Layer" tab of your process. Just slowing down the print speed won't help as you need the head to move away to give the small features some time to cool down, especially when printing PLA. 20% - 30% infill should be enough here.
https://www.simplify3d.com/support/prin ... verheating
bye,
Daniel
Re: Detail won't be good
Hi Daniel
Thank you very much for your reply and very relevant points.
I tried to print two models with distance 8 cm. Fan was running. I tried with and without "Print islands sequentially without optimization". Print temperature was rather low 185 C. Nothing helped. It - still - looks like the fill is not sufficient even I set it to 100%.
Then I got the idea to slice it with Flashprint *attempting* to set the same parameters. And got a perfect result! Even with hole in the handle that I planned from the beginning. So, it cannot be an overheating issue.
I have to study further where there are differences between the S3D and Flashprint.
I believe that Flashprint had a higher print speed than S3D (4800 versus 3600 mm/min), but need to be sure that there are no %-reductions that I am not aware of.
Any hints are still very, very welcome.
Regards,
Niels
Thank you very much for your reply and very relevant points.
I tried to print two models with distance 8 cm. Fan was running. I tried with and without "Print islands sequentially without optimization". Print temperature was rather low 185 C. Nothing helped. It - still - looks like the fill is not sufficient even I set it to 100%.
Then I got the idea to slice it with Flashprint *attempting* to set the same parameters. And got a perfect result! Even with hole in the handle that I planned from the beginning. So, it cannot be an overheating issue.
I have to study further where there are differences between the S3D and Flashprint.
I believe that Flashprint had a higher print speed than S3D (4800 versus 3600 mm/min), but need to be sure that there are no %-reductions that I am not aware of.
Any hints are still very, very welcome.
Regards,
Niels
Re: Detail won't be good
Hi Niels,
good to hear you made some progress printing that part.
Just to make sure, we're talking about this feature on the top?
That shouldn't require any infill at all. Looks like a perfect 45° and with proper cooling/settings this should come out quite nice, regardless of the slicer.
1. For testing I'd move the parts along Z until there's only the interesting features with a base for adhesion left:
(I've used -12.1 inside your factory file to get 1st layer, one solid layer and "the interesting part")
2. Slice and try to print without infill on both slicers
3. Take a 60-90 second video while each slicers code is printing
Compare single layer times and look closely where the difference is.
Just guessing, but I think you might find a difference in how the slicers handle possible slowdowns in movement here.
On the "Speeds" Tab there's an option to slow down the print if the layer would be printed in less than x seconds.
Depending on part arrangement, part size and other values this can be a tricky option. While slowing down all movements
can help (more time for objects to cool down while the head is travelling to the next one) this could also be exactly what's
ruining the print - if print times for walls and infill get longer to a point where the head is long enough right above
the object to heat up previous layers too much (and moves/rips walls apart when moving away).
In case of the model as included in your factory file there's another way to avoid those problems:
Just print them this way and don't use any support / small features at all. But orientation is also dependent on where you
want your parts strong and multiple other variables. So rotating an object can be a solution to some problems.
Happy hunting,
Daniel
good to hear you made some progress printing that part.
Just to make sure, we're talking about this feature on the top?
That shouldn't require any infill at all. Looks like a perfect 45° and with proper cooling/settings this should come out quite nice, regardless of the slicer.
Still looks like one to me... but that doesn't mean im right. Hard to tell without seeing the parts getting printed.nfol wrote: Then I got the idea to slice it with Flashprint *attempting* to set the same parameters. And got a perfect result! Even with hole in the handle that I planned from the beginning. So, it cannot be an overheating issue.
It would be nice to know what the slicers have done differently. Just a few ideas, maybe it'll help somehow somewhere... ^^I have to study further where there are differences between the S3D and Flashprint.
I believe that Flashprint had a higher print speed than S3D (4800 versus 3600 mm/min), but need to be sure that there are no %-reductions that I am not aware of.
Any hints are still very, very welcome.
1. For testing I'd move the parts along Z until there's only the interesting features with a base for adhesion left:
(I've used -12.1 inside your factory file to get 1st layer, one solid layer and "the interesting part")
2. Slice and try to print without infill on both slicers
3. Take a 60-90 second video while each slicers code is printing
Compare single layer times and look closely where the difference is.
Just guessing, but I think you might find a difference in how the slicers handle possible slowdowns in movement here.
On the "Speeds" Tab there's an option to slow down the print if the layer would be printed in less than x seconds.
Depending on part arrangement, part size and other values this can be a tricky option. While slowing down all movements
can help (more time for objects to cool down while the head is travelling to the next one) this could also be exactly what's
ruining the print - if print times for walls and infill get longer to a point where the head is long enough right above
the object to heat up previous layers too much (and moves/rips walls apart when moving away).
In case of the model as included in your factory file there's another way to avoid those problems:
Just print them this way and don't use any support / small features at all. But orientation is also dependent on where you
want your parts strong and multiple other variables. So rotating an object can be a solution to some problems.
Happy hunting,
Daniel
Last edited by parallyze on Thu May 02, 2019 12:45 pm, edited 1 time in total.
Re: Detail won't be good
I've just taken notice of something that may cause trouble inside your factory file...
Infill is set to 100% using a width of 200%. Using an Extrusion Width of 0.4mm that's 0.8mm infill. I'm not sure the average 0.4 nozzle can handle .8mm properly. Was this set on purpose?
If I were given the factory file without any further information I'd probably set it up something like this:
- Change Infill width to 100%, lower Infill to somewhere in the range of 20% - 30%
- Untick "Merge all outlines into a single solid model" on the "Advanced" Tab
- Remove support structures
- Rotate part
- Duplicate part
- Align parts
- Increase top/bottom layers because of the thin 0.1mm layers
- Align start points (layer tab) so you don't get retracts on the overhang
Of course this all depends on the wanted result. You could come away without any infill in this case by setting the amount of perimeters/extrusion width.
If the walls have a thickness of 3mm you could get away with 3 shells @ 0.5mm extrusion width without any infill:
(Just an illustration, haven't measured anything of the model to get a precise extrusion width, it's more of an example)
Note: I do not recommend this as an alternative to printing it the way you intended to do and skip the parts about finding the
difference between your slicers. You will have that problem again one day if you just re-arrange stuff to avoid it now.
bye,
Daniel
Infill is set to 100% using a width of 200%. Using an Extrusion Width of 0.4mm that's 0.8mm infill. I'm not sure the average 0.4 nozzle can handle .8mm properly. Was this set on purpose?
If I were given the factory file without any further information I'd probably set it up something like this:
- Change Infill width to 100%, lower Infill to somewhere in the range of 20% - 30%
- Untick "Merge all outlines into a single solid model" on the "Advanced" Tab
- Remove support structures
- Rotate part
- Duplicate part
- Align parts
- Increase top/bottom layers because of the thin 0.1mm layers
- Align start points (layer tab) so you don't get retracts on the overhang
Of course this all depends on the wanted result. You could come away without any infill in this case by setting the amount of perimeters/extrusion width.
If the walls have a thickness of 3mm you could get away with 3 shells @ 0.5mm extrusion width without any infill:
(Just an illustration, haven't measured anything of the model to get a precise extrusion width, it's more of an example)
Note: I do not recommend this as an alternative to printing it the way you intended to do and skip the parts about finding the
difference between your slicers. You will have that problem again one day if you just re-arrange stuff to avoid it now.
bye,
Daniel
Re: Detail won't be good
Hi Daniel,
Thank you so much for using time on this!
> Just to make sure, we're talking about this feature on the top?
Yes.
> Infill is set to 100% using a width of 200%. Using an Extrusion Width of 0.4mm that's 0.8mm infill. I'm not sure the
> average 0.4 nozzle can handle .8mm properly. Was this set on purpose?
Not on purpose. I will look more into the significance of that. And also keep your other suggestions in mind in the future.
However, I think I found the actual problem:
In the factory file attached, the "Retraction Distance" is 1.0 mm and "Extra Restart Distance" is -0.4 mm. This file gives the bad print. The "handle" is almost halfened (infill sat to 40%).
If I change "Extra Restart Distance" to 0, the model comes out correctly.
I sat the restarts distance to -0.4 because i had a seam problem on earlier, cylindrical models.
Again: Thank you very much for your help.
Best regards,
Niels
Thank you so much for using time on this!
> Just to make sure, we're talking about this feature on the top?
Yes.
> Infill is set to 100% using a width of 200%. Using an Extrusion Width of 0.4mm that's 0.8mm infill. I'm not sure the
> average 0.4 nozzle can handle .8mm properly. Was this set on purpose?
Not on purpose. I will look more into the significance of that. And also keep your other suggestions in mind in the future.
However, I think I found the actual problem:
In the factory file attached, the "Retraction Distance" is 1.0 mm and "Extra Restart Distance" is -0.4 mm. This file gives the bad print. The "handle" is almost halfened (infill sat to 40%).
If I change "Extra Restart Distance" to 0, the model comes out correctly.
I sat the restarts distance to -0.4 because i had a seam problem on earlier, cylindrical models.
Again: Thank you very much for your help.
Best regards,
Niels
- Attachments
-
- Detail - S3D test3.factory
- (13.7 KiB) Downloaded 211 times
Re: Detail won't be good
Hi Niels,
nozzle geometry. Personally I'm using 0.4mm nozzles at 0.4mm up to 0.65mm. 0.8mm can work but you may
have artifacts, that may be misleading when looking for errors. Try to think of that 200% infill when printing
some materials like PETG, which sometimes can react a bit tricky to sudden volume changes like this because of
pressure changes inside the nozzle.
Walls get printed at 0.4mm and 30mm/s according to the first factory file. While infill prints at 60mm/s
and 0.8mm - so your extruder has to output double the amount of plastic. At 0.1mm that usually is not a problem
but from 0.25mm upwards this can be hard for some extruders
I have to admit I didn't even check that one, just had a look at the extrusion width. But maybe one of the other things might
help someday somewhere.
But missing up to -0.4mm of material after retracts could be a good cause for problems on such a small feature...
bye,
Daniel
Using a bigger extrusion width than nozzle diameter usually works fine. How far you can go often depends onnfol wrote: > Infill is set to 100% using a width of 200%. Using an Extrusion Width of 0.4mm that's 0.8mm infill. I'm not sure the
> average 0.4 nozzle can handle .8mm properly. Was this set on purpose?
Not on purpose. I will look more into the significance of that. And also keep your other suggestions in mind in the future.
nozzle geometry. Personally I'm using 0.4mm nozzles at 0.4mm up to 0.65mm. 0.8mm can work but you may
have artifacts, that may be misleading when looking for errors. Try to think of that 200% infill when printing
some materials like PETG, which sometimes can react a bit tricky to sudden volume changes like this because of
pressure changes inside the nozzle.
Walls get printed at 0.4mm and 30mm/s according to the first factory file. While infill prints at 60mm/s
and 0.8mm - so your extruder has to output double the amount of plastic. At 0.1mm that usually is not a problem
but from 0.25mm upwards this can be hard for some extruders
However, I think I found the actual problem:
In the factory file attached, the "Retraction Distance" is 1.0 mm and "Extra Restart Distance" is -0.4 mm. This file gives the bad print. The "handle" is almost halfened (infill sat to 40%).
I have to admit I didn't even check that one, just had a look at the extrusion width. But maybe one of the other things might
help someday somewhere.
But missing up to -0.4mm of material after retracts could be a good cause for problems on such a small feature...
bye,
Daniel
