I also opened a ticket with BL on Z-height issues while printing tall models in regards to the Z-axis. The issue I have is collisions with the print head to the model. I’ve adjusted what I can, but seems to me z-axis calibration missing from BL Studio is a big issue that needs to be resolved.
4 Likes
Have you analyzed in the video recordings what was the reason for the collisions?
Is it scraping over the part, or is it having hard collisions with the tools head?
Simply stating that the issue is caused by the deviations that the OP is having does not really make any sense.
Those deviations might cause a part that is too high or too low, but your issue looks more like there is backlash on the Z-axis, or possibly over-extrusion.
Would be good to give more details on your use case.
1 Like
Gerrit,
I was present when it started and took it’s final toll on the model. Started with pronounced scraping. It’s not the first model this has happened to. I was using BL filament, so the settings should have been perfect for extrusion. I’ve printed a dozen models higher than 1/2 of the available height, and all of them were scraping against the model at the top most layers. Most of them on Bambu filament. So, no matter the cause, there should be an adjustment. And, I’ve tried all the available ones presented to me so far.
Just consider that if its scraping the model, it would move at the same height as when it is extruding.
This is most likely an over-extrusion issue. The fact that you use Bambu Filament is no guarantee that the flow is perfect.
You could consider a Flow calibration test to see the flow ratio is correct.
Also some crossing infill patterns might cause scraping of the tool head of the infill because the intersections are slightly higher than the rest of the infill.
Moisture can also cause over-extrusion.
But as mentioned, there could be a mechanical issue (backlash) that is causing this issue. But in that case it still has nothing to do with the 1.5% height deviation that the OP is experiencing.
Garrit,
Thanks for the input. Always below 20% humidity (according to AMS 2.0 Pro) and always do dynamic flow calibration. Can you suggest a good article on backlash?
The dynamic flow calibration is something totally different than flow rate calibration.
Flow rate calibration is a manual calibration that requires a test print. The calibration menu can be found in Bambu Studio after you activated “Engineering Mode “ under preferences.
Dynamic Flow Calibration refers to the calibration of the so called “Pressure Advance” parameter which is to optimize flow during acceleration and deceleration.
But slowly we are hijacking this thread. It would be good to create your own thread on this.
@moderators can you make this off topic part a separate thread?
1 Like
Please post a picture of your print on the print bed, you might also be experiencing with warping.
Gerrit,
You’ve been very helpful. And, I think the flow of conversation as of yet is not a hijack of the conversation but rather a progression to solutions on something that I and others have experienced. I do appreciate you providing them.
Now to get completely off topic… I am continuing to get a clue that the “just print and don’t worry” mentality that Bambu tries to portray is not all it’s cracked up to be. I will add “yet” to that statement. That said, there are obviously many dials that need to be turned in any print project, and to me it seems Bambu might be just a tad bit too protective in their software to handle all situations gracefully. Your suggestions seem extremely useful, I feel to have to turn on “engineering mode” for a hobbyist like me starts to stretch the expectations for what Bambu states they can easily provide. While I like most hobby only 3d printer fans, don’t mind diving into the small details and turning on special modes, to hide them (in those special modes) and or not provide them at all (i.e. the Z-axis calibration which can be handled easily with G-code) just shows the issue that Bambu needs to find a happy medium for and has not yet. I started 3D printing in the early 2000s after the patents protecting the intellectual property started to expire and many companies dove into providing hobby level printers. It’s been a fun ride, and still maturing even today. The frustration of 3D printing of past is slowly melting away (pun sort of intended) I just hate to see Bambu somewhat inadvertantly put in some blocks to moving past those frustrations. Thank you Gerrit, for being a help to me and others. We’ve lots to learn together.
Thanks dotnico, I did think of that, fortunately there was zero warping. I measured the print, and found no variance on any corner and the print was pretty close to perfect in size after the print, just slightly smaller due to shrinkage as expected.
I think Bambu Labs is trying to get to this “print and forget” mode, but 3D printing has a lot of variables.
Although BBL filament should run optimally, they might come from different sources, but also from the same source, production tolerances influence the diameter and composition of the filament.
Then tolerances in the printer might influence the result also.
As a third factor, you can see the environment the printer is working in. Temperature and humidity indirectly influence the result also.
To come back to the clean plate, I also meant small particles of filament that might raise the plate at some locations. If those are too large, a bed leveling procedure might not be able to compensate for that. But I presume that is not the case for you, but just wanted it to be mentioned.
Gonna go out on a limb here, and ask if you are using grid for sparse infill. If you are, read the following.
Grid infill is sometimes OK on small models, but on larger models (long layer times), it should be avoided.
The problem is that on every layer, the two directions are printed. On a small model, when an infill line is crossed, the plastic is still molten, so it happily continues, BUT it tends to raise little zits above the current layer, so the next layer interacts with them.
On a large model, by the time an infill line is crossed, the line being crossed may have had time to solidify and the tip of the nozzle hits it.
If you like that pattern of infill, rectilinear or crosshatch will work, because the infill is laid down in only one direction on a layer, and the next layer is laid down on the other direction, so the nozzle does not hit the crossed line.
Most other infill patterns are acceptable.
When the problem happens, it is hard on the nozzle, and can actually bend it.
The following infills also cross previous lines on the same layer, despite what Bambu Lab says about them:
Cubic
Triangles
Tri-Hexagon
Adaptive Cubic
Support Cubic
Here are some timed tests of sparse infills.
Times in hours:minutes, all done on the same model.
Using Generic PLA profile with 15% infill:
3:22 Lightning - use this if you don’t need strength
3:37 Concentric
3:45 Archimedean Curve
4:01 Line
4:07 Aligned Rectilinear
4:08 Rectilinear
4:26 Octagram Spiral
4:31 CrossHatch
4:46 3D Honeycomb
4:49 Gyroid
4:56 Hilbert Curve
5:33 Honeycomb
To see 10 of the 17 infills, see Fast infill and low consumption test - Free 3D Print Model - MakerWorld
The following video explains a lot about some of the problems we face.
The unseen world of 3d printing at 1000fps!
For definitive proof of Grid infill being bad, check it out at time 14:01.
Despite the ads, it it well worth watching the whole thing, and you can jump over the PCBWay stuff.
1 Like
Hi @Narg ,
If you could show some pic’s it would really help finding the root cause.
- Infill choice usually shows up as shredded infill tops
- Flow issues are usually associated with sand dune type waves on surfaces and walls although they can also show as low quality infill lines
- Warping is often seen as a local flow issue on top surfaces, in particular around corners and edges but absent in the centre
- Curling is particularly tricky to spot and likes to happen on overhangs, edges and supports
So the usual chain is:
- Dry filament, clean plate (if nothing else at least to rule them out)
- Use infills like rectangular, gyroid or honeycomb
- Consider 2 walls for support trees
- Keep the print environment warm to reduce uneven cooling (<~35°C for PLA)!
If errors presist, the real troubleshooting starts:
- Use adaptive layer height to counter curling, failing supports. For warping only, a global layer height reduction is helpful, as are mouse ears or a brim to some extend.
- Slow down
These two are particularly effective as they have cumulative cubic effects on warping/curling. They do of course increase print time though.
& 
1 Like