Combatting Edge Curl

[JohnStack]

I can't seem to get a 90 mm x 120 mm x 5 mm ABS print working without edge curl.

Some particulars:

I have a well-calibrated printer.

0.50 nozzle
235 head temp
65 bed temp, glass with painters tape
0.25 layer thicknesses - but I've tried lower
I've tried skirts and rafts in Kiss - with inflation of 8, and inflate support of 2

Kiss seems to ignore raft settings and quite frankly, Repetier is not suited for this type of printing - I've tried all sorts of infill with it - different speeds, fill densities, etc. It simply gives me too much or not enough plastic - with the same curl.

What's the secret to Kiss w/rt external support? Should I include a 0.25 thick, 1" raft wide in the drawing and just cut it off after I'm done printing? Seems like that's not what we're doing here...

Another question while I'm at it: Why does edge curl occur at opposite corners when the part is so large?

[Polygonhell]

Can you show a picture of the part, large prints can be tricky in ABS.
Sharp outside corners area the usual areas you see warping, the quick fix is to add a circle 2 layers high centered on the corners that cause issues, often referred to as Micky mouse ears. Less perimeters, less solid layers and less dense infill will also help.
The curl is just the plastic shrinking as it cools, you print layer 1 by the time your done, it's cooled and shrunk, you print layer 2 on top of it and as it shrinks it causes the layers below to curl upwards, you need to have enough adhesion to the bed to keep the part flat during the print.
I like hairspray on glass for ABS, but as parts get large ABS starts to be difficult.
You can also try and keep the temperature of the entire part elevated, the idea being if the entire part cools together it won't warp,I've seen people use heat lamps to do this.
PLA is pretty trivial to print at that size, but it comes with it's own issues.

[bvandiepenbos]

I also prefer hairspray on glass at 80-90c
Sticks really really well.
Though I still have issues like you do John with anything very large curling up in ABS.
Often with larger parts the hairspray sticks so well to glass that the shrinking part pulls up chunks of glass!
Have yet to find a good method.
The "Mouse Ears" sounds like a good idea.

~*Brian

[jetpad]

I think in general less infill will help. The best I've been able to do with battling ABS warping was to print directy on glass cleaned with Acetone, no fans, and the print bed surrounded with the wrap-around foam board (that is suppose to go on the bottom of the Rostock) to hold the heat in.

[JohnStack]

Thanks for replying!

(apologies for the photos in advance. My Gallaxy SIII is not great at closeups.)

The design is a phone holder for an outlet. I made it in Autodesk Inventor Fusion, it has design issues but I was eager to print it.

Kisslicer - good vertical (80mm) but the edge kept curling. I had medium infill set. I even held the bad edge down when the head wasn't in the area. The best print of the bunch but the edge curl killed it. I don't think the print will need supports for the circle.

Can't really see the edge curl here but it's about 3mm up on each corner.

Played with infill.

Reduced infill and speed but changed infill to "straight" - that definitely didn't work - I should have set the extrusion down a bit.

Less infill and mouse ears! I'll try those. I'll also try keeping the heat in. I think there's room for innovation there. The round tube or perhaps just printing something...

[cambo3d]

have you tried added circular disc's at the corners (anti-warping discs). this seems to work for some.

[JohnStack]

Does speed have anything to do with curling too? I didn't notice an appreciable difference.

I do wonder what it would be like if we got Repetier's rafting capabilities with Kisslicer's general flexibility....

BTW, when I exported from Inventor Fusion, the print came out sideways - like the base was to be printed on its side. I rotated it in Netfab Basic.

That would make four software packages to do a print:
Inventor Fusion
Netfab Studio
Kisslicer
Repetier for control and management purposes....

[JohnStack]

have you tried added circular disc's at the corners (anti-warping discs). this seems to work for some.

Definitely going to try that. I did try a raft...but repetier had so many other problems....

Mouse ears it is...

[geneb]

John, I think you're confusing Repetier-Host and Slic3r.

You might want to make up a little cardboard shield that you can use to prevent drafts from reaching the printer - that will cause ABS to pull up like that as well.

g.

[JohnStack]

@Geneb, I'm always confused. Aphasia perhaps.

I use Kisslicer to slice, then load the G-code in Repetier to print. Might seem like an extra step but it means that I get control over the printer with software.

And yes, to me, Slic3r and Repetier are almost one in the same...except for the Repetier controls portion. (Geesh. Smack forehead.)

(Launched from one interface in Mac. Ima thinking Repetier should disassociate themselves from Slic3r soon if the software doesn't get any better.)

Anyway, here is the drawing with the mousey ears that is going to the Wobbly Weeble next.

[JohnStack]

A very slow print but a decent one. For some reason, the areas around the balls didn't get filled in - but I blame that on geometry and translation.

The mouse ears were only 0.25mm think - I was amazed how they held the corners. I increased the print temps to 70 and reduced infill. I kept the tape since I wanted to keep things a bit consistent. I forgot to add insulation as suggested and then decided to see it through.

This print is 90 x 110 x 75mm high.

Things I didn't like:

The top surface (would love Kiss to have top thickness)
The vertical part is weak. I can fix the design but it's weak because of infill on bottom of the vertical.
I also found a dead spot on my bed by doing this. I had to pause the print and pull off unattached stuff in dead corner.

This was a very slow print - like 3.5 hours. I'll adjust the drawing and print it a bit faster, perhaps with thicker layers.

[geneb]

You can reduce the size of those ears by half and get the same result I suspect.

Kiss has a thickness field that is used for the walls as well as the top and the bottom. If you want 5 layers on the top & bottom, just multiply your layer thickness by 5 and stick the result in that thickness field. I don't have KS handy at the moment or I'd give you a screenshot.

g.

[JohnStack]

Thanks. In general, I haven't been that happy with Kiss' top surfaces. Sides come out great though. I'll do another print at a higher speed and change the rest - and then move on.

Kinda fun to work all this out; however, it's a bit of a time sink - given the fact that all objects to be printed have their own specific characteristics. Those will hopefully be resolved in software - as time goes on.

[Polygonhell]

FWIW I think you are under extruding looking at the top surface.
I'm surprised none of the slicers have made it easy to add mickey mouse ears yet, Slic3rs brim is close, but there is a significant advantage to having the interior corners between the circle and the part.

[Eaglezsoar]

John, what software did you use to create the mouse ears?
Is Inventor Fusion that you mentioned one of those programs that can take years to learn all the nuances and
I read somewhere that you have to have Autocad before Fusion will work, is that true?

[barnett]

question on mouse ears:

I'd like to add them to my part. Is it just 2 layers high, touching the edges of the part ... trim them off with an xacto knife? Do you leave any kind of gap or perforation?

thanks!

[626Pilot]

Hairspray on kapton on glass. It will hold onto your first layer like a vise.

[Flateric]

Cake that hairspray on too, don't be afraid to use it. My glass looks like really clear crystals of ice in opacity. Not able to see through it, but you can see shapes sorta of idea. I also find I do not need to recoat every time and second and third uses seem sticker if anything. If I see a smooth and shiny patch where the HS (hair spray) peeled up then I give it a refresher shot.

I also think that having the heated bed temp just right is important and unfortunately this can be different from 1 filament to another even from the same guys. Color alone being the only difference. But you will be in the ball park.

I also like the circular disks, but if you want to absolutly ensure no curl or lift at all, follow the steps above and use a nice wide skirt, or a raft. Skirt is very reliable, raft is an almost 100% fix for me everytime. But then this can affect the flat finish of the side facing downwards.

These are all reasons why polycarbonate is my personal favorite print medium of the moment. Yes, it can curl like a mad B*tch, BUT, if you pre-coat the clean spotless glass with UHU-Glue stick, wait till is dries nice and clear (only a few minutes) and print with no bed heat at all. It will almost never curl at all. If you do get slight curl and skirt or raft will do the trick every single time garrenteed. Extra perks of the polycarb is it nis really something to look at, very crystalline in appearance.

Ooops, almost forgot, you do HAVE to use a all metal hot end with the polycarb (lexan) due to it's very high print temps requirments. (270c-310c depending on speed of print and desired crystal transparancy.

It's really really tough plastic also.

[JohnStack]

I can't wait to try it! Just about ready for PLA, then Nylon...of course a few repraps in between! Thanks for all the feedback!

[626Pilot]

I can't wait to try it! Just about ready for PLA, then Nylon...of course a few repraps in between! Thanks for all the feedback!

PLA is very well behaved on the print surface. With heated glass you don't need any tape or hair spray. To get a piece off you can just torque it (requires less torque than ABS), or wait for the plate to cool a few tens of degrees, and it will just lift off with little or no force. It never curls up - once it sticks to the glass, it's just there. I'm printing ABS at the moment, but only beause the stock hotend sucks for PLA. As soon as I get the Budaschnozzle mount stuff printed, I'll be using that, and probably go back to PLA.

[Capt Jack]

I use straight Kapton tape to hold my ABS down. Works great!
FYI, it seems your bed temp of 70c is a bit low. The bed temps usually recommended for printing ABS is 100-120c. I have mine set to 120 for the first layer and then 115 for the remainder. Solved my curling issues.
Of course, 120c on my system is closer to 110c when checked with my multimeter.

[geneb]

Funny thing is, I've printed well over 100 hours of PLA with the stock hot end and I've never had a single issue. A PEEK fan is _required_ to keep it from jamming (like all other hot ends).

g.

[Flateric]

I have no real problems with the stock hotend either. But in comparison to a good jhead it's just no contest.

Compared to a stock hotend, the jhead drools far less, needs far less retract, can "instant" on when starting to print and takes less time to heat up. It also is less prone to failure and simplier in design and weighs less. Not as likely to have issues with shorting of thermistors or the heating resistors. Heater cartridges also last longer. and far easier to deal with as well. That's alot of pro's towards a jhead in compared to the stock hotend. But yes, the stocker does do a fair job too.

[626Pilot]

Funny thing is, I've printed well over 100 hours of PLA with the stock hot end and I've never had a single issue. A PEEK fan is _required_ to keep it from jamming (like all other hot ends).

g.

Maybe you have better PLA than I do. I have tried one and two 10mm fans blowing directly on the PEEK section, and I have tried temperatures from 170 to 220. It ALWAYS jams up sooner or later.

[626Pilot]

I have no real problems with the stock hotend either. But in comparison to a good jhead it's just no contest.

Compared to a stock hotend, the jhead drools far less, needs far less retract, can "instant" on when starting to print and takes less time to heat up. It also is less prone to failure and simplier in design and weighs less. Not as likely to have issues with shorting of thermistors or the heating resistors. Heater cartridges also last longer. and far easier to deal with as well. That's alot of pro's towards a jhead in compared to the stock hotend. But yes, the stocker does do a fair job too.

SeeMe should just dump the stock hotend and ship Rostocks with J-heads. They have more than enough skill to produce really good ones and they could do a lot of business selling these since the hotends.com guy is always sold out.