Hello everyone, and thanks in advance for the help.
I've been trying to print some dice that I designed, 8 and 10 sided. I wanted dice with larger numbers for a friend who has a hard time reading small numbers. While I started off with complex hollow designs and other cool designs, each revision has been more and more simple. I seem to have problems with the print where sides with a small overhang are distorted or malformed. (image links below.) I have a new M2, I've only used it a couple weeks.
In general, I've been using 0.1 layer height. I use Simplify3D and have been using the high quality default settings for the most part.
Things I've checked:
1) Bed leveling, I've tried with an extremely level bed, at least to the limit of my ability to feel the difference in drag on a piece of paper between the nozzle and bed.
2) I've calibrated the z stop to the same precision. (usual procedure I use is level once, repeat for fine adjustment, z stop, recheck level.)
3) the printer is located on a concrete floor, so no extra wobbling
4) I block extra air flow from the sides, though it is not enclosed, and the room is individually climate controlled at 76 degrees F.
5) I've used two different types of PLA filament, MakerGear at the recommended 215F and Hatchbox at 205F. (both worked fine for other prints, or other parts of the troubled prints, the MakerGear filament was fresh out of the packaging for one attempt.)
6) I have tried various support setups, with manual support, and changing the support resolution to 1mm, as well as changing max overhang to 30 degrees. The same issue happens even in supported areas, or areas with minimal overhang (10 to 20 degrees)
7) I printed spiral chess pieces which have complex curves and overhang successfully using 0.2mm layer height, if that helps narrow things down.
8) my print bed is lined with painter's tape, and has excellent adhesion, even after the print bed has cooled. (no issues with the prints being loose or coming unstuck.)
9) I've tried different infill amounts, 20 to 75%
I really wanted to make these dice for my friend, because he struggles with smaller numbers. The only idea I can come up with is that the basic shape of an 8 (or 10) sided die is oddly unfavorable to printing. If anyone could shed some light on this for me, I'd be really grateful.
Example Images Here: https://imgur.com/a/lxEpo
Unsure what is causing this print problem
Re: Unsure what is causing this print problem
u can divide it in 2 part and glue them back together ^_^ divide and conquer it's always the best way
i guess u are going to paint inside the number otherwise is pretty hard to read anyway
i guess u are going to paint inside the number otherwise is pretty hard to read anyway
Re: Unsure what is causing this print problem
The problems with overhangs are endemic to PLA filament. Due to the nature of the filament itself, it wants to warp, curl and (if it's being printed at too high a temp) drip.
Also, if you are printing over thin air (as in the deep recesses of the numbers) of course they are going to sag, there's nothing underneath them.
Some things you can do to improve it:
Here's the thing about PLA - it's a very counter-intuitive filament. You would think that you need to really heat it to get it to bond, but if you overdo it and it gets too liquid, it slags and drips and warps, and does not cool off quickly enough.
Particularly with overhangs, PLA needs to cool off as quickly as possible. You're actually making it much worse by blocking the air flow to the print and holding the heat in. If you have a small high speed desktop fan that you can point directly at the rear overhang areas, that will cool those off quickly enough to eliminate a lot of problems. (Generally, the overhang sides that do not face the bed fan directly show the worst issues with PLA.)
With any other shape, I would suggest that you turn the item so that the overhang faces the bed fan, but that one looks like it has overhangs on all sides, so no good there.
To increase the overall cooling exponentially, remove the fan guard from the bed fan. (And be careful not to stick filament into it when it's running.) But you will still need a separate fan for the rear, due to the shape of what you are trying to print.
Print at the lowest temperature that you can for PLA, and still get the filament to stick to the plate. For me, that was about 200° to 205°C for the extruder, and a bed temp of about 40° to 50°C. Printing at a lower temperature also helps the PLA to reach room temperature more quickly.
Print using the Outline Direction on the Layer tab going from Inside to Out. (Anytime you have overhangs.)
Sometimes you just have to throw in the towel and print with support. It's a last resort, since even with support, you aren't going to get a perfect result on those steep overhang areas. (Last picture - that one is going to be tough.) But it will be better than what you see now.
Good luck with it.....
Also, if you are printing over thin air (as in the deep recesses of the numbers) of course they are going to sag, there's nothing underneath them.
Some things you can do to improve it:
Here's the thing about PLA - it's a very counter-intuitive filament. You would think that you need to really heat it to get it to bond, but if you overdo it and it gets too liquid, it slags and drips and warps, and does not cool off quickly enough.
Particularly with overhangs, PLA needs to cool off as quickly as possible. You're actually making it much worse by blocking the air flow to the print and holding the heat in. If you have a small high speed desktop fan that you can point directly at the rear overhang areas, that will cool those off quickly enough to eliminate a lot of problems. (Generally, the overhang sides that do not face the bed fan directly show the worst issues with PLA.)
With any other shape, I would suggest that you turn the item so that the overhang faces the bed fan, but that one looks like it has overhangs on all sides, so no good there.
To increase the overall cooling exponentially, remove the fan guard from the bed fan. (And be careful not to stick filament into it when it's running.) But you will still need a separate fan for the rear, due to the shape of what you are trying to print.
Print at the lowest temperature that you can for PLA, and still get the filament to stick to the plate. For me, that was about 200° to 205°C for the extruder, and a bed temp of about 40° to 50°C. Printing at a lower temperature also helps the PLA to reach room temperature more quickly.
Print using the Outline Direction on the Layer tab going from Inside to Out. (Anytime you have overhangs.)
Sometimes you just have to throw in the towel and print with support. It's a last resort, since even with support, you aren't going to get a perfect result on those steep overhang areas. (Last picture - that one is going to be tough.) But it will be better than what you see now.
Good luck with it.....
Re: Unsure what is causing this print problem
Just got another look at the size of those dice - they're pretty tiny. If you are printing them one at a time, the heat soak effect is probably also coming into play. (Happens when the printing time for a particular layer is too short to give the previous layer time to harden up.)
Solution - print multiples of the same item on the plate to give the previous layer time to harden before the next layer gets laid down, or print several of the different dies at once.
Okay - now I'm done.
Solution - print multiples of the same item on the plate to give the previous layer time to harden before the next layer gets laid down, or print several of the different dies at once.
Okay - now I'm done.
Re: Unsure what is causing this print problem
Pekish: that may be the way to go if I just can't get it. The d8 would be easy to divide along the equator.
Jules: that is extremely helpful information! It also helps to explain why the problem areas are on the opposite side of the fan. I will definitely work on the temperature. I have been hesitant to print many pieces at once, both because I don't want multiple failures, and because it seems that is when I have the most issues with bed adhesion.
I hadn't considered the temperature so much because I had been printing with ABS, and I figured that the MakerGear starting settings would be the best way to go before trying to experiment. I will attempt your recommendations, then try ABS if the temperature issues are still causing problems.
The dice themselves are quite large for polyhedral dice, about twice the 'standard' size, but still small compared to other items. I have a lot of hobby painting experience, and experience in 3D modeling, so printing lets me combine the two. once I get a good print, I will fill and smooth the surface, prime it a few times with the air brush, and then apply the color. A little brush work for the numbers, and clear coat, and they should look pretty nice.
Jules: that is extremely helpful information! It also helps to explain why the problem areas are on the opposite side of the fan. I will definitely work on the temperature. I have been hesitant to print many pieces at once, both because I don't want multiple failures, and because it seems that is when I have the most issues with bed adhesion.
I hadn't considered the temperature so much because I had been printing with ABS, and I figured that the MakerGear starting settings would be the best way to go before trying to experiment. I will attempt your recommendations, then try ABS if the temperature issues are still causing problems.
The dice themselves are quite large for polyhedral dice, about twice the 'standard' size, but still small compared to other items. I have a lot of hobby painting experience, and experience in 3D modeling, so printing lets me combine the two.
once I get a good print, I will fill and smooth the surface, prime it a few times with the air brush, and then apply the color. A little brush work for the numbers, and clear coat, and they should look pretty nice.Re: Unsure what is causing this print problem
Well, take some pics before you gift them, so we can see the finished product....
Re: Unsure what is causing this print problem
I will be sure to post some in this thread when I get something worth finishing. I will have to do some experimentation on smoothing the print texture, I'm used to working with other plastics, resin, epoxy, etc. I'm torn between sanding, which is precise but time consuming, and using a filler coating. The filler coating isn't as controllable, but being able to print several and try different mediums and specific mixtures should simplify the process. I'm leaning towards an acrylic resin suspension at the moment, assuming I can get good bonding without a primer.
I tried cooling the extrusion temperature and bed temperature somewhat, and placed a desk fan so that it blew across the build platform. (I turned it on after the first layer was done.) I also printed 6 dice at once, 3 d8s and 3 d10s. I had some adhesion problems, I lost 1 die, and there was some lifting/curling from the edges. (While I've read that bed adhesion and warping is supposed to be a bigger problem with ABS, I seem to have more issues in this regard with PLA than ABS.) Switching from kapkon tape to masking tape on the bed for PLA did help a lot, at least with the print completely separating. It is kind of annoying to have to switch tape and re-level the bed when I want to change filament type though.
That attempt did seem to be better regarding the gooey/warped overhang edges, despite the adhesion issue. Makes it seem like I'm going in the right direction. I'm currently at work, but have another print going. I positioned the desk fan in a way that seems to be better for even air flow, though I don't have a setup at the moment that I'd consider ideal. (I'm an atmospheric scientist by day, so I can geek out on fluid dynamics.) I readjusted the extrusion temperature upward slightly because it seemed too low, and changed the print bed temperature up to 70 degrees to helpfully alleviate the issue with adhesion, I figured the lower temperature combined with the extra cooling might be responsible. The warmer extruder should also hopefully help, I noticed the layer 1 texture was smoother as it didn't seem to flow into the tape texture as well. Finally, I changed the speed settings in the cooling tab, allowing greater speed reduction with a larger time threshold per layer. I haven't figured out how to adjust the overall print speed, the relevant settings in the software seem to be about hardware configuration rather than print speed. the print time is now estimated at 5 hours by Simplify3D, compared to 3 hours for the last attempt.
I know doing all this at once really ruins the possibility of narrowing things down to one specific cause... but being at work I can't do multiple attempts and to retry with a single change as they fail. 3D printers are a little annoying in this regard. I can handle long print times, and I expect to be more successful as I learn more, but the length of each "trial and error" cycle is so long that it can be frustrating. What is your opinion on cooling fan mounts/cowls for better airflow? It seems to be popular on 3d file sites like thingiverse, but I can't help but think that if it was so beneficial printers would come that way by default.
Thanks again for the assistance. I probably can't contribute towards printing-specific stuff, but I'm going to look around the forums. I should be able to help out with finishing/painting and other subjects. One thing I really want to explore is using my printer to produce molds for resin and epoxy casting. If I can figure out a good way to to finish printed surfaces, I might be able to produce molds for quicker production of smooth items or specialty material like carbon fiber/etc.
I tried cooling the extrusion temperature and bed temperature somewhat, and placed a desk fan so that it blew across the build platform. (I turned it on after the first layer was done.) I also printed 6 dice at once, 3 d8s and 3 d10s. I had some adhesion problems, I lost 1 die, and there was some lifting/curling from the edges. (While I've read that bed adhesion and warping is supposed to be a bigger problem with ABS, I seem to have more issues in this regard with PLA than ABS.) Switching from kapkon tape to masking tape on the bed for PLA did help a lot, at least with the print completely separating. It is kind of annoying to have to switch tape and re-level the bed when I want to change filament type though.
That attempt did seem to be better regarding the gooey/warped overhang edges, despite the adhesion issue. Makes it seem like I'm going in the right direction. I'm currently at work, but have another print going. I positioned the desk fan in a way that seems to be better for even air flow, though I don't have a setup at the moment that I'd consider ideal. (I'm an atmospheric scientist by day, so I can geek out on fluid dynamics.) I readjusted the extrusion temperature upward slightly because it seemed too low, and changed the print bed temperature up to 70 degrees to helpfully alleviate the issue with adhesion, I figured the lower temperature combined with the extra cooling might be responsible. The warmer extruder should also hopefully help, I noticed the layer 1 texture was smoother as it didn't seem to flow into the tape texture as well. Finally, I changed the speed settings in the cooling tab, allowing greater speed reduction with a larger time threshold per layer. I haven't figured out how to adjust the overall print speed, the relevant settings in the software seem to be about hardware configuration rather than print speed. the print time is now estimated at 5 hours by Simplify3D, compared to 3 hours for the last attempt.
I know doing all this at once really ruins the possibility of narrowing things down to one specific cause... but being at work I can't do multiple attempts and to retry with a single change as they fail. 3D printers are a little annoying in this regard. I can handle long print times, and I expect to be more successful as I learn more, but the length of each "trial and error" cycle is so long that it can be frustrating. What is your opinion on cooling fan mounts/cowls for better airflow? It seems to be popular on 3d file sites like thingiverse, but I can't help but think that if it was so beneficial printers would come that way by default.
Thanks again for the assistance. I probably can't contribute towards printing-specific stuff, but I'm going to look around the forums. I should be able to help out with finishing/painting and other subjects. One thing I really want to explore is using my printer to produce molds for resin and epoxy casting. If I can figure out a good way to to finish printed surfaces, I might be able to produce molds for quicker production of smooth items or specialty material like carbon fiber/etc.
-
willnewton
- Posts: 479
- Joined: Wed Apr 09, 2014 8:32 pm
Re: Unsure what is causing this print problem
I run hatchbox at 205º on layer one and drop to 200º on layer 2, but you can run lower to 200-195º with no issues on small stuff. That might help.
Maybe slow your print speed 20mm/sec and try again. Also check the Enable Speed Reduction box for layers lower than 15 sec to 10%.
As Jules said, print multiples. Try 4 of them arranged in a 100mm square on the bed.
Zipped up below is rsilver's FFF profile from http://www.thingiverse.com/thing:806946, it a good starting place for nice settings and may improve your results.
Possibly choose a font that has bolder detailing within the letter. It looks like you used Impact, which makes a bold punchout, but leaves tiny negative space that is inherently difficult to print. Helvetica Bold or Verdana may do better. Also, reducing the depth of the letters may help.
Maybe slow your print speed 20mm/sec and try again. Also check the Enable Speed Reduction box for layers lower than 15 sec to 10%.
As Jules said, print multiples. Try 4 of them arranged in a 100mm square on the bed.
Zipped up below is rsilver's FFF profile from http://www.thingiverse.com/thing:806946, it a good starting place for nice settings and may improve your results.
Possibly choose a font that has bolder detailing within the letter. It looks like you used Impact, which makes a bold punchout, but leaves tiny negative space that is inherently difficult to print. Helvetica Bold or Verdana may do better. Also, reducing the depth of the letters may help.
- Attachments
-
- M2-0.35mm.fff.zip
- (3.14 KiB) Downloaded 310 times
I'm finally back to where I started two days ago!
A thread with some stuff in it I update every once in a while. viewtopic.php?f=8&t=9
See some of my stuff http://www.thingiverse.com/willnewton/favorites
A thread with some stuff in it I update every once in a while. viewtopic.php?f=8&t=9
See some of my stuff http://www.thingiverse.com/willnewton/favorites
Re: Unsure what is causing this print problem
Just something to aspire to, down the road......Did you happen to see the two-tone ones that Tim printed with the dual? (He did a freaking awesome job of it - no painting required.)
viewtopic.php?f=6&t=2848
viewtopic.php?f=6&t=2848
Re: Unsure what is causing this print problem
Just a note, I hope it doesn't seem like I'm neglecting this thread. I'm still really new, so my posts have to be approved by a moderator. Understandably for the season, that takes a while.
Willnewton, thanks for the link to the settings, I'll try it out when I'm home. I've been running hatchbox at 200-205, but hadn't tried changing extrusion temperature between layers yet. More recently, I've been using the MakerGear filament. As mentioned in my prior post (that hasn't been approved yet) I have done the speed reduction for layers lower than a certain time. I'm not sure how to globally reduce print speed yet, something else to look up on the weekend when I get a chance. Christmas weekend and school winter vacation, so I've been playing with my kids a lot. I'm posting the results of my last trial with the settings in that previous post below. I was already looking at the font, I just haven't had time yet to make the change. The negative space isn't too bad so far, so I'll probably try helvetica. right now, my working meshes are 300mm tall, and I use 80mm lettering with 10mm depth. the negative space portions are raised by 8mm instead of 10, so that they are also recessed. It ends up being about a 1/10th scale for printing. It's a lot shallower than the examples I looked at on thingiverse, but I think further reduction would help so I'll try that too.
Jules, I hadn't seen that yet, but it looks pretty darned good. I haven't sprung for a dual extruder, I figured I should figure out a single one first. Right now the dice are serving multiple purposes, figuring out print quality, and later serving as a test bed for finishing. (and of course, for use in our D&D campaign) For those dual-color dice, a quick sanding to smooth, and a nice clear coat to give a shine and further smooth, and it'd be pretty awesome, better looking than store bought dice for sure.
This latest attempt with the changed settings has some noticeable improvement. I am going to try uploading pictures directly to this post. There is still some distortion on the sides opposite from the air flow, but nowhere near as bad as previously. I had adhesion problems with the last print, but only for 1 die on the corner of the build plate. I arranged 6 dice in a row along the X axis, so that all the dice were getting airflow from the desk fan roughly evenly without blocking each other. I positioned them on the back of the build plate, so that the incoming air wouldn't be too cool, and the HBP thermister would be more representative of the temperature under the print. I figure that the bed needs to be leveled again, that might have caused the issue with that last die in the corner.
Speaking of which, I've noticed that the build platform seems to lose its level rather quicker than I expected. I've noticed that after a few hours of print time if I check the level, it is no longer true. I don't really know what the acceptable amount of unevenness is yet, but the bed is noticeably not level after 10 hours or so of print time. I suppose because we're dealing with small fractional turns of the leveling screws, like a 1/8th or even 1/16th of a turn, it's not that surprising given all the moving the platform does. I do kind of wish the screws had a flatter thread though, the final adjustments I find are 1/32nd to 1/64th of a turn, so fine that I have to make several attempts to get it right.
As you can see on the pictures, overall quality is much better. I still get that protruding-looking area near the print bed, like that face has extra material, which ruins the die. The early overhang isn't very neat either. I was using the default 4 solid top and bottom layers, but I think reducing them to 2 would help, a smaller mass of solid hot plastic keeping things warm. I used 70 or 75% infill on this print, because I want the dice to have a more substantial weight. (should improve randomness in the dice, and has a nicer feel.) I think I should try reducing that as well for heat reasons, though it does increase print time. For some reason, Simplify3D likes to start a new layer on the same object that it finished the last layer on, but varies which object it finishes with. I should probably figure out how to make it start a new layer on one of the dice that are cooler. I'll put that on the list for when I'm trying to figure out slowing the overall print speed.
Thanks again everyone for the help
Willnewton, thanks for the link to the settings, I'll try it out when I'm home. I've been running hatchbox at 200-205, but hadn't tried changing extrusion temperature between layers yet. More recently, I've been using the MakerGear filament. As mentioned in my prior post (that hasn't been approved yet) I have done the speed reduction for layers lower than a certain time. I'm not sure how to globally reduce print speed yet, something else to look up on the weekend when I get a chance. Christmas weekend and school winter vacation, so I've been playing with my kids a lot.
I'm posting the results of my last trial with the settings in that previous post below. I was already looking at the font, I just haven't had time yet to make the change. The negative space isn't too bad so far, so I'll probably try helvetica. right now, my working meshes are 300mm tall, and I use 80mm lettering with 10mm depth. the negative space portions are raised by 8mm instead of 10, so that they are also recessed. It ends up being about a 1/10th scale for printing. It's a lot shallower than the examples I looked at on thingiverse, but I think further reduction would help so I'll try that too. Jules, I hadn't seen that yet, but it looks pretty darned good. I haven't sprung for a dual extruder, I figured I should figure out a single one first. Right now the dice are serving multiple purposes, figuring out print quality, and later serving as a test bed for finishing. (and of course, for use in our D&D campaign) For those dual-color dice, a quick sanding to smooth, and a nice clear coat to give a shine and further smooth, and it'd be pretty awesome, better looking than store bought dice for sure.
This latest attempt with the changed settings has some noticeable improvement. I am going to try uploading pictures directly to this post. There is still some distortion on the sides opposite from the air flow, but nowhere near as bad as previously. I had adhesion problems with the last print, but only for 1 die on the corner of the build plate. I arranged 6 dice in a row along the X axis, so that all the dice were getting airflow from the desk fan roughly evenly without blocking each other. I positioned them on the back of the build plate, so that the incoming air wouldn't be too cool, and the HBP thermister would be more representative of the temperature under the print. I figure that the bed needs to be leveled again, that might have caused the issue with that last die in the corner.
Speaking of which, I've noticed that the build platform seems to lose its level rather quicker than I expected. I've noticed that after a few hours of print time if I check the level, it is no longer true. I don't really know what the acceptable amount of unevenness is yet, but the bed is noticeably not level after 10 hours or so of print time. I suppose because we're dealing with small fractional turns of the leveling screws, like a 1/8th or even 1/16th of a turn, it's not that surprising given all the moving the platform does. I do kind of wish the screws had a flatter thread though, the final adjustments I find are 1/32nd to 1/64th of a turn, so fine that I have to make several attempts to get it right.
As you can see on the pictures, overall quality is much better. I still get that protruding-looking area near the print bed, like that face has extra material, which ruins the die. The early overhang isn't very neat either. I was using the default 4 solid top and bottom layers, but I think reducing them to 2 would help, a smaller mass of solid hot plastic keeping things warm. I used 70 or 75% infill on this print, because I want the dice to have a more substantial weight. (should improve randomness in the dice, and has a nicer feel.) I think I should try reducing that as well for heat reasons, though it does increase print time. For some reason, Simplify3D likes to start a new layer on the same object that it finished the last layer on, but varies which object it finishes with. I should probably figure out how to make it start a new layer on one of the dice that are cooler. I'll put that on the list for when I'm trying to figure out slowing the overall print speed.
Thanks again everyone for the help
- Picture with a penny for scale
- With supports still attached