gCreate Spaceship- Fullsize print with questions

#1
Here are some pix of the gCreate Spaceship I printed yesterday and the settings I used for each piece. This was the largest print I've completed so far. Overall, I'm pleased since the intent is to finish the model off with paint giving it a distressed aluminum look ; detailing the rivets, portholes and rocket boosters.

The print quality does spawn some questions though.

It looks like there may be layer shifting around the portal windows but not ABOVE the portals. The problem seems to have corrected itself (see pix). I'm not sure if that's a printer issue or a Slic3r issue. Another forum member had a similar problem and posted the same issue on another thread. He indicated he switched to Simplify3D but the problem didn't go away. This was several months ago and both Simplify3D and the gMax have been updated . I am unsure of the cause and how to fix it. The Simplify3D website says they support gMax printers. This might be a recent addition to their site since it lists the XT as a supported printer - Another member mentioned Simplify3D to Gordon and urged him to get the gMax config to them. I mention all of this since there is no trial for Simplify3D and I would gladly spend the money if I knew it was a slicing engine problem. At this point, I am trying to understand why this is happening.

Seam alignment was random so I am unsure what blobs are layer changes and what can be fixed by additional tweaks.

There is also artifacting to the right of the entire column of rivets in one of the photos. This only occurred on one face. Another face of the rocket the rivets didn't have the issue.

Last issue was the upper portion of the antennae (see pic). Not sure how to address that. (hotend, bed, slic3r) The "knob" and portion below is not the same quality as the rest of the piece.

I was not going for super high quality. Everything was based on medium quality and tweaked from there. Again, the intent was to finish it off with paint. Using the latest Marlin firmware for XT and gMax Slic3r config bundle as of this post, the only slicer settings that were tweaked were perimeters, infill, layer height and perimeter speed. (see below).

All parts scaled @ 100%

Spaceship Main
Perimeters 3 (2 is recommended - the rocket has some heft to it)
Infill: 20% (could probably go a little lower - unless you plan to fly in it).
Layer height .25 (small tweak)
First layer height .28mm
Perimeter speed: 45% <-------- 50% is ok too (and faster)
No support, brim or tall skirt.
Print time: 9 hours 16 minutes
Notes: This took some time to remove from the bed - It was really stuck. I was concerned about breaking one of the little rocket boosters with the putty knife so I chipped away at each one lightly until I could get a small corner of the knife under the booster. This took a couple of minutes on each one. (Be patient - 2 minutes is way better than having to print another one for 9 hours+. Then with light pressure, I was able to pry them loose). The main booster released from the bed by grabbing the body of the spaceship and lightly torquing it off. Areas of the main spaceship printed better than others. - vertical faces on the fin arches connecting to the body (really nice). Main body contour (some areas not as good).

Spaceship Top:
Perimeters: 2 (I printed the spaceship first and after feeling how hefty it was I settled on 2 for the rest of the parts)
Infill: 20%
Layer height .25 (small tweak)
First layer height .28mm
Perimeter speed: 45%
No support, brim or tall skirt.
Print time: Approx 1 hour 30 minutes
Notes: You'll need to sand down the "peg" of the Spaceship top so it can seat fully onto the Spaceship. Be careful when handling. The antennae is fragile and I broke it off while sanding. No worries, I just printed another. :)

Flame
Perimeters: 2
Infill: 20%
Layer height .25 (small tweak)
First layer height .28mm
Perimeter speed: 45%
Print time: Approx 1 hour 40 minutes

Smoke
Perimeters: 2
Infill: 18% (this would have been really heavy with a higher number. It feels very strong at 18% not to mention a faster print & less filament - could probably go less for the main smoke body but I am not sure how strong the "stem" part would be. After I snapped the antennae , I was a little more cautious)
Layer height .30mm (small tweak)
First layer height .28mm
Perimeter speed: 70% (ramped this up for increased turbulence look and shorter print time)
Print time: Approx 11 hours 35 minutes
Notes: I increased the layer height for a couple of reasons. 1) Time and filament 2) I thought the added roughness would increase the turbulent look of the smoke. It was definitely faster, but you can tell the difference in layer height when looking close. I'm happy with the way it printed but I wanted to point that out. From a few feet away you can't tell. The whole smoke piece released from the bed very easily. A couple of pokes from the the putty knife and a it pried off with minimal pressure. No warping at all along the edges on the large, flat bottom.

I still have a few tweaks and a ton to learn but I feel the printer is now ready to start printing prototypical products and provide solutions to our clients - It's intended purpose. - There is no doubt this printer will pay for itself in short order :) So that just leaves the learning curve - any insight on the print issues is appreciated.


Cheers,

Larry

The forum only allows 3 attachments per post. Pix of print problems to follow.....
 

Attachments

#4
Cool! I forgot how awesome that things looks. I am going to have to print one.

It looks to me like one belt is taut. While the other is loose. This can cause some rippleing after certain features that require the motor to change direction or accelerate. Like bungee jumping on a really small scale. Your rivets and windows could cause this. Which explains why some are good (acceleration on the taut axis). While others are bad (acceleration on the loose axis). You can always slow down a bit and see how that helps.

The antenna sag. This is caused by the printer laying down a layer while the layer beneath it is still cooling. If you can, raise your fan speed. Watch that the extruder can keep a stable temp. Try that first on a small print.
You could lower extruder temps. This is a good technique for knowing your temp ranges and a nice g-code lesson.https://www.youtube.com/watch?v=FSOPsRiiOZk. Open g-code in notepad. edit>find>z10. If your not using a g-code editor.(watch the video)
And finally under filament settings>cooling>cooling thresholds. There are settings for print speeds for layers that print quickly because they have a small area.
My dads fix to layers sagging... Print 2 models at the same time. That's not always an option. But some useful oldman wisdom nonetheless.

Make one change at a time and note taking is very helpful. Let me know what you come up with.
 
#7
It's nice to see they have the GMAX on the simplify3d site. I was the one telling Gordon to get in contact with them since it worked so well out of the gate for me.

In any case, if you are curious about simplify3d and if using it could address some of these issues I can always slice up the model for you for so you can see how it works before potentially purchasing it.
 
#8
gCreate_Kyle said:
It looks to me like one belt is taut. While the other is loose. This can cause some rippleing after certain features that require the motor to change direction or accelerate. Like bungee jumping on a really small scale. Your rivets and windows could cause this. Which explains why some are good (acceleration on the taut axis). While others are bad (acceleration on the loose axis). You can always slow down a bit and see how that helps................
Make one change at a time and note taking is very helpful. Let me know what you come up with.
Kyle, thanks! I am pretty sure that is the issue. My x-axis is nice and taut but there is slop in the Y Axis belt. The spring does not move at all on the x axis but the Y axis, pulls the spring and then accelerates. The fact that you identified this as an issue and my Y belt is loose makes me think, technically it's an easy fix. Now how do I pull this off physically?

During the build, I documented the entire process and any issues I ran into. One of the issues I had was the belts. I am new to all of this. A) I had no idea how tight they should be (nothing to reference) and had a difficult time getting the Y-belt as tight as the X axis belt. I sent an email to Gordon suggesting a 10 second video or a little documentation showing the best way to install the belts. Is there some tip or technique that makes initially tightening the belt easier?

Now that LIRlene is built what is the best way to tighten the Y axis belt? Gordon had mentioned that to tighten the belt on an ongoing basis you turn the printer over and just adjust the two screws on the clamp. Turn the printer over? I don't think I follow. What about the gantry? Does that act as a prop to triangulate the frame for support while I work on the exposed carriage/belt?. That is just way too scary of a visual in my mind. Can you clarify the best way to tighten the Y belt on an operational printer? I think I would rather remove the bed. Please advise.

And I totally agree with you - Only introduce one change at a time. Otherwise you're trying to hit a moving target.

Thanks again,

Larry
 
#9
grumble said:
It's nice to see they have the GMAX on the simplify3d site. I was the one telling Gordon to get in contact with them since it worked so well out of the gate for me.

In any case, if you are curious about simplify3d and if using it could address some of these issues I can always slice up the model for you for so you can see how it works before potentially purchasing it.
Grumble,

Yep, it was you I was referring to and thanks for the offer to slice. I think Kyle identified the issue. So first, I am going to try and tighten up the Y axis belt and print another spaceship (only way to know for sure)

Question regarding Simplify3D. - Does it allow you to specify which extruder you can use specifically for Skirt & Brim? While experimenting/playing with Slic3R, I set the brim and skirt to 0mm to see if I could "cheat" and let the software create a different color trim, around the lower edges using the second extruder. It worked. I figured this might be a neat and easy way to "brand' a product line with a distinct color scheme along the bottom edge where the whole object is one color but the complementary trim or material at the bottom is another. More importantly it will let the software do natively what I would otherwise have to do with two stl models (one for each material/color). An example being the Spaceship. If I had black filament in extruder one, and silver in extruder two, I would like to be able to specify to print the spaceship in black (extruder one) but print the brim or skirt using extruder two. I would set the skirt height/brim width to how ever many layers I wanted the trim color to be and the software would take care of the rest. Another example if rubber was in the second extruder, I could create the raft, BRIM and Skirt with the extruder of my choice. give the object a nice rubber bottom that I could create POST STL creation. I don't see why I can't manipulate the skirt brim and raft to achieve some pretty unique results.

This is achievable in Slic3r but they combine the brim and skirt setting with the support material setting in terms of which extruder to use. I'm curious if Simplify3D is a little more granular and separates them.

I do realize that by adding skirt and brim it will increase the dimensions at the bottom of the print since I am adding thickness at the base for the different colored trim.

Thanks,

Larry
 
#10
Kyle,

I was able to tighten the Y belt quite a bit. There is still a very small amount of play in the spring tensioner but it is a lot tighter than what it is before. I removed the bed and leaned the printer up against the wall so I could access the two inside screws on the bottom of the Y belt clamp. I used a needle nose pliers to grip the loose end of the belt in my left hand as I loosened the clamp screws with my right. This was to prevent the loose end of the belt from coming out of the clamp completely and also to give me the ability to pull the belt end tight towards the rear of the printer as I loosened the screws. Once the belt was taut, I tightened the clamp screws. It is still not as tight as the X-axis belt, but it is much tighter with a VERY small amount of flex in the spring. Before it was quite a bit.

I replaced the bed and re-leveled it. Unfortunately, it did not solve the problem. The issue occurred in the exact same areas and seems to be more pronounced. See comparison pix below. After seeing this for the second time, it occurred to me that if it is a mechanical issue due to the belt, I should be able to rotate the model and reproduce the issue on the face where it did NOT occur previously. I rotated the model in Slic3r 90% on the platen and I'm currently running another test print now. That should be enough rotation to see if its a mechanical issue. I also thought to do a few z cuts in Slic3r and just print the main body part as a separate .stl file so I can isolate the printing of the windows and not have to wait for the rocket boosters to print. But that would not give us an apples to apples comparison to isolate the problem. It would be interesting to see though if the problem went away by printing it as a separate .stl. since it would be lower on the print bed (whereas now in the full print it is elevated by the boosters) Another option would be to do multi spaceship print with the spaceship in various rotations and placement on the platen but that is an option of last resort - Unless of course, I need a fleet of Spaceships for the next Flash Gordon movie. - Ain't troubleshooting fun :shock: ).

Larry
 

Attachments

#12
Kyle,

I downloaded your test file from thingiverse - thanks. The third spaceship is several hours into the print so I am going to let it finish. May as well see how the rotated one comes out. I will print out the test file per your instructions and will post my results for the third spaceship and your test model when they're done.

I appreciate the help.

Larry
 
#13
shaqFoo said:
Grumble,

Yep, it was you I was referring to and thanks for the offer to slice. I think Kyle identified the issue. So first, I am going to try and tighten up the Y axis belt and print another spaceship (only way to know for sure)

Question regarding Simplify3D. - Does it allow you to specify which extruder you can use specifically for Skirt & Brim? While experimenting/playing with Slic3R, I set the brim and skirt to 0mm to see if I could "cheat" and let the software create a different color trim, around the lower edges using the second extruder. It worked. I figured this might be a neat and easy way to "brand' a product line with a distinct color scheme along the bottom edge where the whole object is one color but the complementary trim or material at the bottom is another. More importantly it will let the software do natively what I would otherwise have to do with two stl models (one for each material/color). An example being the Spaceship. If I had black filament in extruder one, and silver in extruder two, I would like to be able to specify to print the spaceship in black (extruder one) but print the brim or skirt using extruder two. I would set the skirt height/brim width to how ever many layers I wanted the trim color to be and the software would take care of the rest. Another example if rubber was in the second extruder, I could create the raft, BRIM and Skirt with the extruder of my choice. give the object a nice rubber bottom that I could create POST STL creation. I don't see why I can't manipulate the skirt brim and raft to achieve some pretty unique results.

This is achievable in Slic3r but they combine the brim and skirt setting with the support material setting in terms of which extruder to use. I'm curious if Simplify3D is a little more granular and separates them.

I do realize that by adding skirt and brim it will increase the dimensions at the bottom of the print since I am adding thickness at the base for the different colored trim.

Thanks,

Larry

You can specify which extruder you want for the skirt and brim in Simplify3D. Here is a screenshot of how that works.

As far as tightening the y axis, my neighbor helped me and using 2 people makes it much easier as it's hard to keep tension on what you are also tightening by yourself.
 

Attachments

#14
Larry, try turning down your small perimeter speed a tad, and maybe dropping your perimeter acceleration a bit. I can't be certain, but coming off the rivets, it looks like your getting a little ringing from the sharp change of direction.
 
#16
crussell said:
Larry, did you ever get this to come out clean? I'm curious to what the final resolution was.
Chris, I was away for a few weeks and haven't had the opportunity to post. I've printed about seven different cubes with varying results but nothing solved the problem.

In a nutshell, I went through the entire printer and found several issues that made me wonder how I was able to print the rocket in the first place

1) The Y-belt wasn't loose it was fine. The tensioner spring was flexing on the y belt due to the Y carriage rollers being pulled too tight against the rails. When I built the printer, I had no idea how tight it should be or how easy the bed should move. The bed moved and the motors didn't complain (it never sounded like the motors were missing steps/clicking). I loosened the rail slightly and now there is no movement in the tensioner spring at all and the bed rolls easier. NOTE- due to the v-slot wheels being pulled too tight against the rail, some of the wheel got shaved off resulting in a "loose" wheel which caused the bed to move ever so slightly. There was an accumulation of white shavings on the right wheels of the Y carriage. This is how I noticed the rail was too tight. The movement was solved by tightening the wheel bolts to close the gap between the spacer and the wheel. Now it is a snug fit & rolls smooth.

2) Extruder #1 had subtle movement to it during printing. I noticed this when it crossed over a layer that had already printed. It wasn't suppose to touch it at all but it wasn't level so one side of the extruder hit the previous level and physically moved. It was subtle but noticable. The dual extruder came fully assembled and I only had to mount it on the x -carriage rail. I solved the issue by removing both extruders, reseating them and tightening down their belt clamps. Extruder #1 had a loose belt clamp and this is why it was moving. Now both hot ends are sitting firmly in the carriage.

3) The X-carriage was not tight enough against the X-rail and this allowed the carriage to "rotate" slightly as if it were on a "bbq spit". I tightened down the x-carriage and the movement is much less. It's not as rigid as I would like but it's much better. I don't think there is a way to mitigate the movement completely.

4) I adjusted all of the steppers and the printer runs much quieter now and there is no visible/audible clues of missing steps.

5) Per your suggestion, I slowed down the perimeter speeds and I also lowered the JERK settings per another user's suggestion.

6) Releveled the bed.

7) Crossed my fingers.

Attached are the latest two cubes that just came off the printer. As you can see, they are far from perfect. They were printed one right after the other. The red one came from extruder #1 and the black extruder #2. The ringing issue is most apparent on the right & left sides of the cube (pulling along the y axis) and there is now horizontal striping between the rivets along the x-axis.

I can print large and small scale items without any problems. There's only an issue when there is a print with surface detail. I attached a dual color print that was much larger and printed without problems. There is no surface detail but you can see the layer quality throughout. Also, the printer is printing accurately dimension wise. I've printed a 50mm and 150mm calibration angle at 5mm height and they're accurate to within .1 mm. I am assuming this is an acceptable tolerance.

Any further help would be appreciated.

Cheers,

Larry
 

Attachments

#17
Additional images for my previous post (board limit per post is 3)

BTW, - the quality of the translucent red PLA is not as good as the black PLA. I received filament from different vendors and there is a noticeable difference in filament quality between the two. The red, seems to "pit' easier and doesn't adhere to the bed like the black.

(Kyle just addressed filament quality in this post:
viewtopic.php?f=12&t=423)

The red, PLA is not what is causing the issues (i printed the cube in other colors with the same results) but I thought I would mention this since the images show the red quality being much poorer than the black.

This is apparent in several of the layers of the cone where there are gaps/splits on the red but not the black. What is the best way to address this?

As always, any help is greatly appreciated.

Cheers,

Larry
 

Attachments

#18
LIRLENE!
She looks great.

Ok, the right and left side rippleing is actually an x-axis issue. The x-axis makes the movements for those rivet features and the g window on the right and left sides. Probably similar to your y-axis carriage problems either loose belt, or v-slot wheel friction. Although could be a number of things from a loose bolt on the frame to a well timed earthquake. I put my printer on carpet pads to dampen vibrations.

Your y-axis rippling is quite small. It is challenging the eliminate this effect. I am thinking I might print this model at a 45 degree angle to see the difference. Maybe if the force is distributed on both axes than the ripple will be even smaller.

That weird layer shift is the extruder offset shifting. The extruder for your red filament could be loosening from hitting previously printed layers.
Those wheel tension screws you mentioned for tightening the carriage to the open beam are actually meant for leveling the z height of the extruders. Before you take my word for it I will have check with Gordon on this statement. I say this because it is what I did.
 
#20
By banding I am a assumeing you are talking about the rippling on the perimeter. This is caused by recoil after the extruder moves to "scribe" a feature. Too much acceleration, a loose belt, or too much friction will cause this.