Large printed parts are not flat even though they don't unstick from bed
Hi,
I have a strange issue. I am printing some large parts in PLA, like this one:
And I finally managed to tweak the settings so that the part doesn't lift from the bed - even after it fully cools down at the end - however, when I finally remove it from the bed (I do it by bending the spring steel sheet very very slightly, and it pops right off), I see that the part is not flat. The surface that lays on the bed is concave.
Again, I checked thoroughly after the print, and saw that the middle did not warp and the part was perfectly flat prior to removing it from the bed. I make sure to wait at least 1 hour after the end of the print before even removing the sheet from the magnetic bed, but it seems to make no difference. When removing the print from the steel sheet, I tried bending the sheet both inwards and outwards - but that makes no difference either, the part ends up concave either way.
This seems to be some kind of internal stress in PLA that gets released after the part is no longer held to the bed, and causes that warp. If that is true, I have no idea how to prevent that stress, or how to release it while keeping the part flat. Annealing would probably fix it, but it would also make the whole part shrink, throwing off the dimensions. Or it might even make it worse.
I tried different print temperatures (in the end I had to stick to higher ones because otherwise I get warping and lifting during print), but even at lower temps, there seems to be no difference in this regard. I tried different PLA filaments (including Prusament PLA), as well as drying them in the dehumidifier overnight prior to printing. Doesn't help either.
The bed itself on my printer is dead flat - I checked it with a straightedge.
How can I fix this issue? For some parts, that initial surface being flat is absolutely essential, as for some assemblies it might be the reference surface, meaning that the alignment and accuracy of the assembly of other parts depend on it being as flat as the bed of the printer itself.
I'm attaching the .3mf for reference. Please note that the GCODE takes a while to generate.
RE: Large printed parts are not flat even though they don't unstick from bed
For some reason I can't attach that 3mf... And forum won't let me edit my post. Here is the link to download it: https://ufile.io/czoxgcz8
RE: Large printed parts are not flat even though they don't unstick from bed
The forum allowed file list is very limited. To attach files you need to zip them up into an archive.
RE: Large printed parts are not flat even though they don't unstick from bed
That's one of the limit of FDM printing: tolerances are quite large. You can see many measuring or control tools on 3D prints libraries, but prints are often not accurate enough for real usage. Your example is a good one: large flat part with sharp corners is prone to deformation during heating and cooling. You should control, calibrate and cut the plastic (back to substractive manufacturing) to obtain an accurate tool.
RE: Large printed parts are not flat even though they don't unstick from bed
That's one of the limit of FDM printing: tolerances are quite large. You can see many measuring or control tools on 3D prints libraries, but prints are often not accurate enough for real usage. Your example is a good one: large flat part with sharp corners is prone to deformation during heating and cooling. You should control, calibrate and cut the plastic (back to substractive manufacturing) to obtain an accurate tool.
Yeah, I'm aware there are limits to 3D printing tolerances, but what I'm getting here is really, really bad. I have seen people printing much larger and more complex parts in PLA without getting these concave surfaces, so there must be something that can be done to fix this. I don't have any subtractive manufacturing machinery available to grind down that surface to flat, and doing it by hand (sander block) won't achieve sufficient flatness either.
RE: Large printed parts are not flat even though they don't unstick from bed
One thing that would be interesting to know, after you remove the print surface from the bed but before you remove the part is the bed bowed up in the middle ?
I've had large parts (nearly full bed size) slabs that have pulled the print sheet away from the magnets but were fully adhered to the bed. That was round the edges though and I eventually added extra temporary binder clips to hold the edges down to fix that but I am wondering if due to your design its actually pulling the bed away in the middle ?
People have modified their designs to include cut outs that relieve stress along certain axis but that dont impact part strength and usability. Dont know if anything like that is possible with your though.
RE: Large printed parts are not flat even though they don't unstick from bed
One thing that would be interesting to know, after you remove the print surface from the bed but before you remove the part is the bed bowed up in the middle ?
I've had large parts (nearly full bed size) slabs that have pulled the print sheet away from the magnets but were fully adhered to the bed. That was round the edges though and I eventually added extra temporary binder clips to hold the edges down to fix that but I am wondering if due to your design its actually pulling the bed away in the middle ?
People have modified their designs to include cut outs that relieve stress along certain axis but that dont impact part strength and usability. Dont know if anything like that is possible with your though.
To answer that first question - yes, when I remove the sheet from the bed, the part bends the sheet. However, that bending force is not strong enough to overcome the magnets, so the part is kept flat while it's cooling. Doesn't help, though...
I would be interested to know how to design these stress relief cuts. Without knowing how they are made, I can't say for sure if they are feasible in my design. After all, not only does that surface have to be flat, but it also needs to have sufficient strength in remaining flat when forces are applied. If you have any more info on this, it would be great.
RE: Large printed parts are not flat even though they don't unstick from bed
This is some inverse warping which is just something inherent, even to PLA with large but relatively flat pieces. This hill shaped warping (surface is bent downwards towards the edge) is the result of having a solid bottom layer and not enough vertical structure. The major factor here is printbed temperature.
I had a similar issue with very flat, bed filling PLA pieces. One thing one can do is minimize the bottom surface thickness to the absolute minimum. The best results however are achieved by reducing bed temperature as much as possible, maybe even choose not to heat the bed at all. Use glue stick or specialized adhesion agents to secure bed adhesion at lower bed temperatures, use a brim if necessary.
For me it was already sufficient to reduce bed temperature from 60°C to 40°C and use Kores stick for adhesion to almost eliminate the issue.
Mk3s MMU2s, Voron 0.1, Voron 2.4
RE:
The cause of this warping is the vertical temperature gradient when you use a heated bed. Everything looks perfectly flat as long as the bed is heated. When you cool the bed and remove the part however, the bottom is cooling down more than the upper layers. Therefore it also shrinks more. This is happening after the part is removed from the bed. That bending post cooling might also take a while. That means the warping might be slightly worse the day after removal than directly after removal. This depends also on the stiffness and geometry of the party.
Mk3s MMU2s, Voron 0.1, Voron 2.4
RE: Large printed parts are not flat even though they don't unstick from bed
Large flat surfaces, the bane of FDM printing... Well, one of them.
I found this to be more of a problem the more Z height there is to the model. Just a flat slab usually prints fine so kind of hard to test if it's going to be an issue or not.
I have never done it myself but I like @thejiral's idea of printing with no heat on the bed. Certainly worth a try.
Formerly known on this forum as @fuchsr -- until all hell broke loose with the forum software...
RE: Large printed parts are not flat even though they don't unstick from bed
This is some inverse warping which is just something inherent, even to PLA with large but relatively flat pieces. This hill shaped warping (surface is bent downwards towards the edge) is the result of having a solid bottom layer and not enough vertical structure. The major factor here is printbed temperature.
I had a similar issue with very flat, bed filling PLA pieces. One thing one can do is minimize the bottom surface thickness to the absolute minimum. The best results however are achieved by reducing bed temperature as much as possible, maybe even choose not to heat the bed at all. Use glue stick or specialized adhesion agents to secure bed adhesion at lower bed temperatures, use a brim if necessary.
For me it was already sufficient to reduce bed temperature from 60°C to 40°C and use Kores stick for adhesion to almost eliminate the issue.
Thank you! That sounds very promising. I hope I can get enough grip on that surface with glue stick alone. I am not a fan of the glue stick, considering how long it takes to clean it up, even with soaking in isopropyl... But if it works, it will be amazing. I will try this long print again overnight, and will let you all know how it goes.
RE: Large printed parts are not flat even though they don't unstick from bed
To clean it up, just take your sheet to the sink. A bit of hot water and dish washing soap, and it's off in seconds. IPA will just smear it around.
Formerly known on this forum as @fuchsr -- until all hell broke loose with the forum software...
RE:
Give it a try with the glue stick. Wash it the way fuchsr described for removal. It's a bit of additional work but really not that much. Removing the glue in water is not difficult at all.
The more expensive specialized adhesion agents from Magigoo and co are even easier to remove and might be even stronger, if simple glue stick is not sufficient to keep the part on the plate. But first I would try if simple glue stick is good enough.
PS: I would try 40°C bed temperature first and test if you get the part to stick to the very end. If everything works fine but some warping remains you can try even unheated bed next.
Mk3s MMU2s, Voron 0.1, Voron 2.4
RE: Large printed parts are not flat even though they don't unstick from bed
I've also used Layerneer bed weld with good success.
Formerly known on this forum as @fuchsr -- until all hell broke loose with the forum software...
RE: Large printed parts are not flat even though they don't unstick from bed
I tried at 40°C and glue stick, and while perimeters go down nicely, small details like circles, supports, don't seem to stick at all, and keep dragging with the nozzle. Is it okay to increase the temperature only for that first layer and then cool down to 40°C, or will that make the part unstick mid print?
RE: Large printed parts are not flat even though they don't unstick from bed
I don't think it is too much of an issue if the first layer is printed at a higher temperature as long as you go to the lower temperature right after the first layer. It might solve slight adhesion issues.
But if you are printing with Prusament PLA, you shouldn't have such adhesion issues with glue stick and 40°C. I would recommend not to use Prusaslicer support foundation though. You can use the slicer supports but I would add 100%infill bases for the first layer where supports touch the bed (either design a custom object with the height of the first layer or add some cubes, cylinders etc and bring them to the right shape in slicer). To be honest I am not a big fan of how Slicer does supports and one thing is because of how it handles their first layers.
Is your z-offset spot on? I would recommend you to do a first layer calibration with 40°C and glue stick and possibly fine tune the offset until things look good. If you can't get it working, maybe you should give the build plate a proper water + dish soap wash. And when you applay the glue stick, do so to the cold plate and don't overdo it. It should cover the relevant space but shouldn't be too thick.
Mk3s MMU2s, Voron 0.1, Voron 2.4
RE:
PS: Lower printbed temperatures are of course less forgiving than 60°C where results can still be just fine even with a slightly off z-offset for example. There is a reason after all, why Prusa chooses that temperature for it's default settings. I spent quite some time on this "inverse warping" issue though and found that the bed temperature is the most powerful and still easiest fix to the problem in many cases.
While I am not sure how challenging it is to get PLA to stick reliably at unheated conditions I can confirm that 40°C still worked decently fine, as long as one uses glue stick. Use brims to be on the safe side.
I like this Magigoo manual. The later pages are of course about their products but in the beginning they have a fairly good introduction to questions of warping, bed adhesion and factors one can optimize to improve results.
https://cdn-3d.niceshops.com/upload/file/Magigoo_Manual.pdf
Mk3s MMU2s, Voron 0.1, Voron 2.4
RE:
Sorry, guys... It's just not working. I cleaned the spring steel sheet with hot water and soap, didn't touch it after, dried thoroughly, then applied single coat of PVA glue, heated up to 60°C, and tried doing the Z height calibration. The filament simply doesn't stick to the bed at any nozzle height. I tried them all. Without glue it is fine, with glue it doesn't work. I tried this over 10 times now, all possible temperature (from unheated to 80°C) and Z height combinations, but I'm getting nowhere. PLA just won't stick at all if the glue is present. I guess I just got lucky on that first attempt... It's almost as if the glue is made to prevent PLA from sticking on purpose. I went through the Magigoo manual, but didn't find an answer to this...
RE: Large printed parts are not flat even though they don't unstick from bed
I just tried again, bumping the temperature to 100°C. Without glue, PLA sticks like crazy. With glue, it doesn't stick at all, same as with any other temperature.
RE: Large printed parts are not flat even though they don't unstick from bed
I just tried again, bumping the temperature to 100°C. Without glue, PLA sticks like crazy. With glue, it doesn't stick at all, same as with any other temperature.
How did 100°C affect warping?