Warts and Seams

[jimc]

there were a couple small things there which are a good idea to change. things like random infill placement was turned on. not that this anything to do with the blobs. i made a few small changes that dont have anything to do with blobs and i made some in the ooze control settings as well. the black part in the pic looks like there are some blobs there but most of what it looks like is a gap from such a high coast. your extrusion width is set to your nozzle dia which isnt really good since the plastic doesnt extrude the width of the nozzle. i stuck that on auto which is .42mm. it should just be raised from that point. not too much lower. thats a whole different subject though. i took your file, made some changes and posted it back up here. do me a favor, without changing anything on it, just run that same print again with the same filament using the new profile. then post a pic. hardware wise it sound like all the usual stuff is checking out and the steps per mm is really close as it should be.

makergear_black.fff.zip

(2.96 KiB) Downloaded 344 times

[pyronaught]

It looks like Dale and Jim might have nailed it with the extrusion width recommendation of around .4 instead of the .35 I had it set to. Using auto sets it to .42, but then S3D leaves out the middle line of my 2mm wall box test. At .4mm it still leaves out the middle line, so I have to set extrusion width to .39 to have the 2mm wall print with exactly five lines. I'm still dialing in the extrusion multiplier with the 2mm box, but at least I'm getting the expected slots now where the excessive 7mm coast marks are instead of blobs. Once I switch back to my cylinder test model I bet I can get a wart free print and then dial back the coast to a reasonable value.

I'm a hairspray convert after doing all these tests. I didn't realize you could keep printing repeatedly over one coating of hair spray. Glue stick is only good for one print before you have to clean it off, and the platform must be below 50C before reapplying it or it gums up. So glue stick is really slow between prints due to the application time.

[pyronaught]

there were a couple small things there which are a good idea to change. things like random infill placement was turned on. not that this anything to do with the blobs. i made a few small changes that dont have anything to do with blobs and i made some in the ooze control settings as well. the black part in the pic looks like there are some blobs there but most of what it looks like is a gap from such a high coast. your extrusion width is set to your nozzle dia which isnt really good since the plastic doesnt extrude the width of the nozzle. i stuck that on auto which is .42mm. it should just be raised from that point. not too much lower. thats a whole different subject though. i took your file, made some changes and posted it back up here. do me a favor, without changing anything on it, just run that same print again with the same filament using the new profile. then post a pic. hardware wise it sound like all the usual stuff is checking out and the steps per mm is really close as it should be.

The attachment makergear_black.fff.zip is no longer available

The .42 extrusion width really seems to be the key here. I tried .39 on a cylinder test and still got big blobs (I think I might have mistyped the number, because they were huge). I had to redo my test model to have a wall thickness that is a multiple of .42 to keep from having an internal gap in the wall, so the image below has a 1.68 wall thickness. You can see the warts are considerably smaller now. The only thing I changed back was the random start points, since I like that better for checking the size of individual warts. I turned it back off and am doing a seam test now and can see the seam is much less noticeable than before. I think the ooze control settings will actually be responsive now, whereas before the .35 extrusion thickness was just blowing out any chance of the ooze controls having any effect. I'm using wipe on the seam test and not getting a double seam like before, so that is working now too.

So I guess the extrusion width is being used by the software to dictate how far apart to make the lines, and if it is too small then the lines start overlapping and you get an excess of goo.

Anyway thanks for all the help, this has got to be the most helpful and responsive forum I've ever been on!

wart_test.jpg (343.8 KiB) Viewed 9724 times

[pyronaught]

Here's the seam test with a 5mm wipe added. Coast was also increased to 2.8. The seam looks worse in the image than it actually is due to the light making it stand out.

seam_test.jpg (361.77 KiB) Viewed 9721 times

[insta]

That's looking much nicer. I know it's a bit of work, but I'd honestly suggest you go through my calibration process above, just to make sure your esteps are dialed in. It will make everything else much easier.

[pyronaught]

I did have to lower the extrusion width to .41 to get more side contact between lines. With .42, when I crush the cylinder in my hand I can see and hear the layers separating. I guess that's why they make this parameter a variable

EDIT: it was actually the extrusion multiplier that had to be increased in order to get the concentric perimeters to stick to each other better. Lowering the extrusion width had no effect on that for some reason.

[jimc]

ok so basically what is happening here is overextrusion. s3d will overlap lines a little when necessary to get them to fit in a space and not leave a gap. in a sense what your doing by changing the width is altering the overlap and changing the amt of overextrusion on the piece. printing a thin wall can be tricky because of this. really the best way to do thin walls is set your perimeters to 1 then set the infill to 90-100%. one other thing i changed on that fff profile was you had perimeter print order on inner then outer. while this is good on a model that has alot of steep angles, its not good on thin walls you are trying to print solid. inner then outer in the case of this model will end up giving you walls that are out of tolerance and make your outer wally sloppy because the inner is a little overstuffed due to overlapping a little. i really try not to print at 100% if its something that needs to be in tolerance or outer walls looking clean. even if you set the infill to 95% then that would be good and just as strong.

one other thing, if the edges of your extrusions are not touching each other then dont mess with the extrusion width. that should be changed by the extrusion multiplier to make it extrude more plastic. as you change the extrusion width s3d is also moving the lines closer or farther apart to match.

[ednisley]

So I guess the extrusion width is being used by the software to dictate how far apart to make the lines, and if it is too small then the lines start overlapping and you get an excess of goo.

It's even simpler than that. The slicer puts the lines exactly as far apart as the "thread width" setting, then drives the extruder to produce exactly enough plastic to create such a thread. That's what makes calibrating the actual thread width against what the slicer expects so critical: if the slicer expects a 0.40 mm thread, but the extruder actually produces 0.50 mm thread, you get too much plastic in the available space.

So using a thin-wall cube to make the actual width match the slicer's setting (by tuning the Extrusion Multiplier) isn't just a Good Idea, it's pretty much required before you can get Good Results... [grin]

However, your cylinder has another factor in play. Because the solid model in the STL file consists of many short straight segments, not a true curve, the distance between the straight walls will be slightly less than the true curved-wall thickness. A crude sketch shows the geometry:

Curved-wall approximation.jpg (111.5 KiB) Viewed 9704 times

The threads then fit inside those straight lines, so you can see that the distance between the threads isn't just the curved-wall thickness minus the threads; there's some empty space that the slicer can't fill.

if you know the curvature, you can figure the actual wall size or thread width required to make the answer come out right, but it's easier to do just as you've done: fiddle around while previewing the G-Code until everything fits. In general, if you're doing a thin-wall object, tuning the wall to fit a fixed thread size will work better, unless you absolutely must have a specific wall size.

Bear in mind that the thread width isn't a completely free variable: it must be larger than the nozzle inside diameter (the hole in the middle) and smaller than the nozzle outside diameter (the flat part on the end of the brass cone). For a 0.35 mm nozzle ID, that means the thread must be somewhere between 0.40 mm and 0.60 mm.

[pyronaught]

ok so basically what is happening here is overextrusion. s3d will overlap lines a little when necessary to get them to fit in a space and not leave a gap. in a sense what your doing by changing the width is altering the overlap and changing the amt of overextrusion on the piece. printing a thin wall can be tricky because of this. really the best way to do thin walls is set your perimeters to 1 then set the infill to 90-100%. one other thing i changed on that fff profile was you had perimeter print order on inner then outer. while this is good on a model that has alot of steep angles, its not good on thin walls you are trying to print solid. inner then outer in the case of this model will end up giving you walls that are out of tolerance and make your outer wally sloppy because the inner is a little overstuffed due to overlapping a little. i really try not to print at 100% if its something that needs to be in tolerance or outer walls looking clean. even if you set the infill to 95% then that would be good and just as strong.

one other thing, if the edges of your extrusions are not touching each other then dont mess with the extrusion width. that should be changed by the extrusion multiplier to make it extrude more plastic. as you change the extrusion width s3d is also moving the lines closer or farther apart to match.

A lot of times the thin wall tubes I'm making have threads on them though, so they have to be inside to outside in order to print them without requiring supports. Trying to dig support material out of threads just isn't worth the trouble. They aren't quite as thin walled as these test tubes I'm making though, they are usually closer to 4mm from the edge of the threads to the other side of the wall. I think that's thick enough to use regular infill though.

[pyronaught]

I think this is about as dialed in as I'm going to get it. I'm pretty happy with this though. Extrusion multiplier wound up being .95 and all dimensions exactly match the CAD model. Coast is still on the high side at 5.5mm with a 3mm wipe, but that's what got rid of the bumps. There might be a gap in some cases, which is why I use the random placement. Making a seam with settings that are right on the verge of leaving a gap would create a failure point in the model, but with them all distributed it is not a problem (unless it needed to hold air or water under pressure). It's outside to inside on this one, which I will use except when there are threads or some reason not to use it.

final.jpg (333.61 KiB) Viewed 9698 times