Printing On The Whole Bed - Verse 2

[Bill Havins]

I have continued to do a number of very large and "flat" prints similar to the one included in this earlier thread (http://forum.seemecnc.com/viewtopic.php ... 9947#p9947). I thought I would share (repeat?) some of the things I am doing that help prevent warping when printing with ABS. (By "warping" I'm referring to corners of the print curling up and away from the print bed.)

First, when doing these big prints I am using a .70 mm nozzle. These prints take a long time, and they will just take too, too long if you use a smaller nozzle.

Second, if you have large thick horizontal "flats" to print the infill needs to be kept to 25% or less. The "skin" needs to be kept to .6mm or less. The horizontal top of the print in the earlier thread is 3mm thick. If the infill is set to more than 25%, or if the "skin" is set to thicker than .6mm, the "pull" that the cooling ABS will create will "pop" the outside corners up (and boy does it "pop" when it comes loose). "Skin" thickness and infill density seem to be two of the more important settings in avoiding "warp."

Third, adding "mouse ears" to outside corners seems to limit the tendency for a print to warp (see the photo in the earlier thread - you'll recognize what I mean by "mouse ears"). I have also found that a "raft," extending outside of the design by 7 or 8 mm also helps to prevent warping (when done in conjunction with these other recommendations). A raft alone will not prevent warping; you have to do all of these things together.

I continue to use KISSlicer (V1.1.1.10). When printing with black 1.75mm ABS (ordered from SeeMeCNC) I set the first layer temperature to 240 degrees; subsequent layers are printed at 235 degrees. I set my bed temperature to 80 degrees and leave it there. I am using a eighth-inch thick piece of non-tempered glass as a build surface. I spray it liberally with a cheap hair spray as the bed is heating. I thoroughly wash the glass in the kitchen sink after doing each large print. I set my bed "roughness" to .07 mm in KISSlicer.

There is a "trap" in KISSlicer that will destroy your prints (and your belts on your Rostock) if you don't know to look for it. Go to the "Printer" tab and then click on the "Speed" tab. The settings for "X,Y Travel Speed" and "XY Accel (mm/s^2)" are the killers. Mine are now set to 90 and 270, respectively. Why? My explanation is below.

When you are doing very large prints on a Rostock MAX, especially when you are using a raft, the extents of the print may approach the limits of the build diameter (i.e., 280 mm). If the carriage moves "diagonally" away from an axis to reach an outside corner of the print, the arms for that axis will approach a horizontal position. In that position there is a tremendous amount of "inertia" that the axis must overcome if the G-code should command it to move back toward the axis. If the "X,Y Travel Speed" or "XY Accel (mm/s^2)" speeds are set too high, the G-code command may spin the stepper motor too fast to overcome the inertia present on the end of the axis - the stepper motor will spin faster than the belt can move. The Rostock MAX will lose its position and, if this is allowed to repeat, the spinning pulley may grind the teeth off the belt.

When I first started using KISSlicer I used settings I found on this forum. I believe the "XY Accel (mm/s^2)" was set to 1500; when an axis' print arms are near horizontal, a move of 1500 mm/s "ain't gonna' happen." My current settings, 90 and 270, work well under all conditions. (EDIT: This discussion of "XY Accel" is in error; "X,Y Travel Speed" is the important variable. Please see ApacheXMD's post, and my response, below.)

Fourth, I choose to print in layer thicknesses that will allow me to complete prints as quickly as possible while still providing adequate quality. If the prints are more vertical (but include a large horizontal plane) I may print in .3 mm layers. I'm doing one of those as I type (25 mm tall and 250 mm wide). When design detail is important I may print in .2 mm layers.

Fifth, I don't use a fan to cool the ABS. When doing these huge prints I haven't found a spot to position a fan where it doesn't cool one corner more than the opposite corner. The cooler corner is going to warp. If I am doing a print with an overhang I will turn a fan on (using the control on the LCD) and manually point the fan at the overhang. Otherwise I have not found a fan to be helpful when printing on the "whole bed."

Finally, I don't print these huge prints using a fast print speed. The one I have going now is running at 15 mm/s. When you get out on the extremes of the build area fast print speeds and accelerated movements can cause the loss of position noted above. When you're 10 or more hours into a print you don't want to see (and hear) your Rostock MAX lose its position.

Just my two cents.

Bill

[Eaglezsoar]

Bill, thanks for such an informative writing. This is the type of stuff we all need to learn how to use our printers correctly.
Thanks for taking the time to write it.

Carl

[ApacheXMD]

I thought the acceleration setting in Kisslicer is used only to calculate estimated print times. Acceleration is set in firmware, no?

[Bill Havins]

In Process of Printing 25mm Tall

Almost Done and No Curl

ApachXMD,

Yes, you're right. "X,Y Travel Speed" is the setting that must be changed to limit travel speed in the G-code when not extruding. I found KISSlicer will crash when slicing if "XY Accel (mm/s^2)" is not adjusted down when "X,Y Travel Speed" is set to 90 (It has been several weeks since I made the changes).

One can make a number of changes in KISSlicer that will cause it to crash. It does not appear to have the error traps necessary to check math when settings are changed (or so it would appear).

Bill

[JohnStack]

Great post! Thanks!

[626Pilot]

I have X/Y Travel Speed at 500 and XY Accel at 1500. These settings work pretty well and I recommend putting them up fairly high. The reason is that they will always be constrained by the firmware accel and jerk settings. If those are up too high you can potentially make the pulleys jump teeth, and then you get air-spaghetti. It's easier to mess with the firmware settings (even in the middle of a print) so I just let the firmware be the limiting factor on speed.

The max jerk I use is 20mm/sec. I have had it all the way up to 60, and it can work okay there on things that don't involve lots of significant direction changes. (Smooth sculputres probably OK, "technical" parts with sharp edges not so much.) Acceleration is I think 800 or so by default, and I've had it as high as 3500, but now I settled on 1000 (printing) and 2000 (travel - to reduce stringing). That gives a decent tradeoff between speed, and not beating up the printer. Having the pulley jump over teeth is NOT good!

[artexmg]

I have continued to do a number of very large and "flat" prints similar to the one included in this earlier thread (http://forum.seemecnc.com/viewtopic.php ... 9947#p9947). I thought I would share (repeat?) some of the things I am doing that help prevent warping when printing with ABS. (By "warping" I'm referring to corners of the print curling up and away from the print bed.)

[...]

Finally, I don't print these huge prints using a fast print speed. The one I have going now is running at 15 mm/s. When you get out on the extremes of the build area fast print speeds and accelerated movements can cause the loss of position noted above. When you're 10 or more hours into a print you don't want to see (and hear) your Rostock MAX lose its position.

Just my two cents.

Bill

Bill, I just found out that, by mounting the hot end at the bottom side of the end effector (instead of its "default" top mounting), the extra height will limit the travel of the cheapskates when printing at the edges.

By using this simple method, I'd been able to print even a little bit outside of the print bed without having the cheapskates to drop down (and the belt being bitten!).

Also, it had helped me to increase the speed to 30mm/s and, some times, I'd gotten it up to 60mm/s (what a brave man I am sometimes!) .

Cheers!