help needed on bad print - how do i fix?

#1
today i printed the white parts in the 2 photos attached. very similar grey parts were printed on an Ultimaker Original Plus with PLA a few weeks ago.

why are my parts so bad? i used most slic3r settings from the config file entered into cura and oriented it in an isometric type position (the grey parts were printed in an iso position, but i don't know if they were printed at same time or same exact orientation).

what am i doing wrong? waiting almost 3hrs for this print was not what i was hoping for...
 

Attachments

#2
Looks like retraction is your issue. And you may be over extruding. Its hard to tell with only 2 pictures. I don't use Cura either. Have you tried Slic3r with the configuration for the gMax?
 
#3
gCreate_Kyle said:
Looks like retraction is your issue. And you may be over extruding. Its hard to tell with only 2 pictures. I don't use Cura either. Have you tried Slic3r with the configuration for the gMax?
I did try the slic3r config files on another part that didn't turn out great. I'll try again, but I'm frustrated trying to print parts four times over 5 days changing settings and orientations.

Could the material be part of the issue? There are so many settings it's hard to know what to do next...
 
#4
Yes material greatly influences the performance of 3d printers. The gmax configuration files were calibrated to work with colorfabb PLA. Personally I use the cheapest PLA I can get with amazon prime from filament outlet. I absolutely have more problems with it but overall is much cheaper and works for my applications. My temperatures are about 20-30 degrees C higher than colorfabb temps set in the slic3r configuration. But this is the only difference in slic3r settings for me.

Loss performance I have noticed:
Bridges by about a factor of 4. I did a 150mm bridge with colorfabb and it sagged just a little. Filament outlet bridges. They fail at 30-40mm. Overhangs become sloppy at shallower angles but this is based on subjective experience. I have not measured a difference.

The maximun size of objects I can print is smaller because warping is much worse with cheap filament.

Layer and bed adhesion is weaker.

I don't have any problems with oozing like the pictures you have posted. Post pictures of prints when you used slic3r along with the g-code and I would be happy to look over your settings and help you out the best I can.
 

GORDON.LAPLANTE

Administrator
Staff member
#5
Hi Garrett,

Have you tried to use slic3r and load our default config settings? We printed your sample files using simplify3d, during our testing phase, but we did print some of your other files using our default slic3r config settings with success.

From what it looks like the blower fan may not be cooling properly. We tried using cura with mixed results and we've had the best results with slic3r.

If you email us your latest file we can slice it and send the gcode file to determine if this is the problem. We hope nothing became loose after your pickup which may be causing the issues.
 
#6
GORDON.LAPLANTE said:
Hi Garrett,

Have you tried to use slic3r and load our default config settings? We printed your sample files using simplify3d, during our testing phase, but we did print some of your other files using our default slic3r config settings with success.

From what it looks like the blower fan may not be cooling properly. We tried using cura with mixed results and we've had the best results with slic3r.

If you email us your latest file we can slice it and send the gcode file to determine if this is the problem. We hope nothing became loose after your pickup which may be causing the issues.
hi gordon - i did try the slic3r default settings after the first 2 print attempts, but i easily could have goofed something up. i'll email you the files and some other photos. i did have the y-axis pinion gear came loose so i had to figure how that was assembled to tighten it up, so i fixed that myself...

i'll email the files shortly...thanks!
 
#9
reprinted my "spike" part with standard config high resolution settings. here are the photos...

the internal details are surprisingly non-existent, the outer edges underneath and on the end are very poor, and the spikes on top have no point, just ball ends. i've included screengrabs of the cad file and the STL file too.

i'm still awaiting feedback or advice here. i'm starting to wonder if the gmax cannot print usable parts for basic mechanical assemblies for interacting/engaged parts.
 

Attachments

#13
I am printing this part as I write. I did change the model slightly though. I reoriented it vertically (upside down in reference to the STL file orientation) and boolean subtracted a feature where the overhang was horizontal. See the pictures. You may or may not be able to do this. I don't know the application for this but the "grain" of the layers may be optimal printing vertically.

The orientation that you choose to print greatly influences the quality and strength characteristics of the model. It is also good practice to avoid using so much support structure. Its important to keep this in mind while modeling a part for an FDM printer. I almost never use supports or a raft.

You could also tweak the support structure settings. If you want to keep the orientation and the model the same. Try increasing the overhang threshold and decreasing the pattern spacing. Be cautioned, this may make it very hard to remove support structure.

This part is very small and detailed. A fine nozzle and dissolvable supports would work best for this print. What units did you use?
 

Attachments

#15
gCreate_Kyle said:
I am assuming your not using mm. cm maybe?
hi kyle - i'm designing in solidworks in inches, then exporting to STL which i think the default is mm.

your part looks better than mine, but the channels look wavy on the spikes side. the feature/part you boolean subtracted is functionally needed. not sure how i can orient to get that back in there.

is the quality of the printed part more of a design issue or printer limitation on small part and feature accuracy capability?
 
#16
You can leave that overhang part of the model on there. You will just have to generate support material from the bed up. I was only attempting to minimize it. Use a g-code viewer to see where slic3r puts your supports. http://gcode.ws/

If I convert the model into inches it becomes 18.5 inches wide and 30 inches tall.
 
#17
gCreate_Kyle said:
You can leave that overhang part of the model on there. You will just have to generate support material from the bed up. I was only attempting to minimize it. Use a g-code viewer to see where slic3r puts your supports. http://gcode.ws/

If I convert the model into inches it becomes 18.5 inches wide and 30 inches tall.
yeah, solidworks exports it in MM while i design it in INCH. will that improve anything if i export the STL in same dimensions?

also, i'm printing another small part right now, and i noticed the gmax screen all jacked up? (see attached photo)

i'm surprised there isn't software out there that optimizes the part orientation for printing... there seems to be way more art and experience required to get a good print than i anticipated, and i still don't know what to look for or how to analyze the part to orient and prep it for slicing/gcode to get a good print.

:cry:
 

Attachments

#19
Yes, art and experience will be difficult to remove from FDM printing. You should watch our slic3r/model tutorial on advanced techniques. https://www.youtube.com/watch?v=gmFrffDxULE

I recommend the other videos in that series as well. This one goes over orientation and model features to optimize the print.

Often I cut models in half and either glue or bolt them together. The cut down the middle makes a nice flat surface for the bed and bolting them together makes them very strong.

Re: screen
eject the SD card and reinsert it. Or restart the printer. This is a grounding issue possibly caused by the long LCD screen ribbon. We are trying to figure out how to permanently fix this.
 

GORDON.LAPLANTE

Administrator
Staff member
#20
Hi Garrett,

We apologize for the email bouncing back. We have been switching our email system over and we had many problems last week. Everything should be working well now. If you send us the files again we can slice and test print it to see what is the problem.

Best,

Gordon