Having recently switched from CrealityV3+Cura to Bambulab A1 + Studio I am very content in general. However, there is one area where my experience with Bambulab Studio is worse than it was with Cura, namely in the ability to produce very thin and easily removable support. I have lost a number of prints because the support simply did not get off, and this is not because I haven’t experimented with lots of settings in Studio.This is definitely an area where improvements are needed. Unless of course I am missing something… What do you guys and gals do to ensure that the support structures can be easily removed?
I find it often difficult to remove the support structure from a print. Bambu Lab Studio should give more options to print very light and thin support, like the 5% support density in Ultimaker Cura.
Would you kindly post the screengrab of the model as it sits om your your build in the preview mode as well as the slice mode? It’s hard to visualize where you’re experiencing problems with just a verbal description. Better yet, include the 3MF file here.
If your supports are sticking too well, it’s usually because you’re over extruding a bit. Try knocking back the flow parameter for the filament a couple 100ths and see if that helps.
Also, while I’m something of a soap-boxer on this, filament that isn’t dry will over extrude and make supports stickier (or too sticky). If your filament hasn’t been through a dryer, put it through a dryer.
See if this helps:
https://www.reddit.com/r/BambuLab/comments/1coiy6m/guide_how_to_produce_easy_to_remove_supports/
If you want to give OrcaSlicer a try (It is a version of BambuSlcer with added features), there a are a few helpful features.
Orca Slicer has a nice setting under “Cooling” in the filament settings called “support interface fan speed.” Set that to 100%
Also set “fan speed for overhangs” to 100% (this is available in BS as well)
This will use aggressive cooling for the top layer of the support, and the bottom layer of the model. Doing this weakens the bond between the support and the model.
If you would kindly share a screenshot of your support settings, we can weigh in with some specific suggestions - also a screenshot of your model with support from the preview tab. There are so many different ways to go about doing support, many of which involve model specific choices.
In te picture you can see how the support sticks to the model (this of course was the underside when printing) and is almost impossible to remove. I also include trhe 3mf file.
schommelpont (Meshed).3mf (86,7 KB)
see my response to Olias with photo and 3mf file. Thanks for any suggestions.
Put it through a dryer? But for how long would that help? My filaments (middle of the road PLA) are not kept in a special atmospheric conditioned closet so it is exposed to normal in-house conditions, I cannot imagine that being a problem because this is what all hobbyist 3D printers do.
You may not be able to imagine it, but that doesn’t make it less true.
If you want clean supported surfaces, you must calibrate flow, and you can’t accurately calibrate flow if you don’t have dry filament.
Hobbyists may not care enough about the quality of their prints, or they may not recognize that they can make them better. But anyone who’s a serious 3D printer knows dry filament is critical to achieving the best possible results. It’s why there are so many filament drying solutions available on the market.
Thanks for that link. I have the impression that setting support line width to .25 mm instead of the default .45 mm may be a (the?) crucial setting because on my last print (10 mins ago) the support got off very easily with those recommended settings.
Getting your .3mf was quite helpful. Thank you!
Before support, the first thing I would consider is printing the model in parts and assembling it. I think you could break it into three/four parts:
-The base and uprights
-The track
-The swing ferry piece that hangs from the track (might be two pieces)
If you do want to print this as one piece, here is how I would do the supports. First, I opened your file in OrcaSlicer, where there are a few more options. You will likely want to download OrcaSlicer.
Under “Support”
-Changed to “Snug.” Not too important, just uses less filament.
-We should use interface layers. It is counterintuitive, but we need more contact so we aren’t bridging and sagging constantly.
-Lets make the interface spacing .7, wider than default.
-We don’t need bottom interface really
-Increase XY just a tad to prevent sticking to walls
Under “Cooling” in the filament settings.
We are going to make some changes that will make the top interface layer, and bottom model layer less apt to want to adhere to each other.
-Set “Support interface fan speed” to 100%. This will make the interface less sticky.
-Make sure “Fan speed for overhangs” is set on 100%
-Make sure “Force cooling for overhangs and bridges” is turned on
Under Quality
Lets make the support interface line width as narrow as a .4 nozzle can handle. This will make the supports a bit more delicate, with a smaller contact patch.
A Third idea you could try is to print the model at a 45degree angle. That will reduce your supported area by a lot. You’d need to add a “sacrificial kickstand” to your model so that your bed adhesion will be good until the support kicks in. I didn’t add the kickstand, but here is the model at 45deg.
Thanks for posting the 3MF, this takes the guesswork out of your question.
To embellish on @movingimage excellent post and offer up some visuals of tips that I found work. You do NOT want turn off top interface layers as you have in your example.
Instead, we rely on those interface layers to be “weakened” contact points that can be broke off more easily. Also, you can increase the base patter to be wider(3mm in this example) which will make the contact points even more brittle and easier to snap off as shown in this revised settings.
On the subject of orientation. @movingimage is spot on with the change in orientation. However I might even say go even bolder. By that I mean, since you’re forced to use supports, print the entire model in the vertical. You will get greater symmetry in the results. Yes, this will take 3X the amount of time to print but you are obviously pursuing appearance over function so spend the time.
Look, you have to resign yourself that you’re going to need supports no matter what. So making those supports more uniform and spread out will at least make any post-processing much easier. If you do this correctly, a screw driver should allow the supports to just pop off. If 3mm spacing doesn’t work then try two other extremes, 0.80 spacing to match 2X of the nozzle and also try 5mm spacing just to see the results.
The other thing to note. Go bold with the X-Y gap. Not just more than the default of 0.35. Think more like 1-2mm. It will ensure that the filament will stay far away from vertical walls and make removal far easier. It will allow as an example, the ability to use a pry tool to get in between the walls and the supports.
I’m a big believer in large X-Y gaps. My minimum gap is 0.5mm and I will consistently go to 1mm if the surface is not angled by a lot (or simply, flat surfaces take big numbers very well, angled surfaces need to be closer to the bottom of the scale).
I also like thinning out the support line width as much as I can. As stated prior, small contact patch = easier to remove.
I tend to use snug supports but that’s more of a preference, and I mostly stay away from tree supports as they have a spotty rep for consistency.
I’m hoping for two hot ends in the same head in the new machines so multi-material will work for supports without bumping print time through the roof.
I agree with all you say.
Tree supports are good when the supported areas are points hanging in the air. So, statues/minis/other organic stuff. In that case, trees work great to minimize time and filament. The other place I use them is with mechanical stuff where I don’t want the base of the support to be on my model, and trees are needed to reach the bed. In all other cases, regular/snug.
The only reason I have an AMS is to do dual material with 0-z-gap support, and yet I still mostly avoid it because it is so slow. So, I agree, a dual-extrusion head would take care of this perfectly. I’m not sure that is the use case for most people though. I am curious how much faster the dual head will print multicolor – I think that is what the majority of people want. Looking at the last Bambulab patent, as best I can read it, maybe it shaves off a third of the time on a 4 color print? I may be totally misreading the patent though
Hi, thank you, movingimage, and Olias and Just4memike and Rocketsled for your extensive comment om this important issue.
Yes printing in parts is what I aften do, especially with technical stuff where parts have to exactly fit other parts (I am creating moving robots with Arduinos). This model was a sidestep, it is a bold engineering idea of a friend that I offered to make a print of. Printing in one piece is a nice demonstration of what can be done - the swing can actually move (along the rails) once you remove the support between the rails.
I have experimented with both horizontal and vertical printing positions. Disadvantage of vertical is (next to print duration) that the bottom side of the rails will be hard to get smooth due to sticking support so the movement becomes difficult.
I did consider 45 degrees but indeed, then you have to manually add smart extra bed adhesion stuff that can be broken of easily (perhaps something with a 0.2 mm gap with the model, like you use when printing moving things?). But this is certainly worth trying.
What do you think of the suggestion of Rocketsled to slightly diminish the flow in the filament settings (e.g. from .95 to .92 or so)?
As for your other advice and suggestions, I am going to do some directed experiments regarding support and hope to report on my results here.
Regards, Michiel
You see… that is an important thing to note. I’ll explain in a moment.
Now that we understand your background as a parts maker and that this is a scale model for a functional part, the design objective becomes clear.
It seems your aim was to create a print-in-place movable model. While this approach works well for “toys”, it may limit the effectiveness of your engineering proof-of-concept unless your goal is to produce an “educational tool” where avoiding assembly or reducing errors is a priority.
If my assessment of your goals is incorrect, feel free to clarify. In the meantime, I’d like to suggest an alternative approach.
Instead of designing the model as a single piece, consider breaking it down into an assembly. As a fellow parts maker, I follow a cardinal rule: avoid printing over air whenever the build plate can provide support. This approach enables you to take advantage of smooth build plates and ironing settings to create surfaces that slide better, which, given your description of the model, would be a desirable feature for the rails and carriage. When combined with techniques like 100% wall loops, this can result in dimensionally accurate parts with properties similar to molded plastic—a “holy grail” for parts builders.
Using the slicer’s paint tool for illustration purposes, I’ve shown one possible way the model could be broken into assemblies. Each color represents a separate part, though you could combine certain sections for ease of assembly. Adapting your original CAD design in this way offers several benefits:
- Optimized settings for each component: For example, the base may not require high infill, while arches might need greater stiffness.
- Support-free printing: Breaking parts down allows them to lie flat, simplifying the process.
- Faster print times: Flat-laying parts print more efficiently.
- Improved iterative design: Modifying one section, such as support towers, becomes easier without reprinting the entire model.
For assembly, you could use any cyanoacrylate glue, but my preference is acrylic cement, as it welds parts rather than merely gluing them, resulting in an extremely permanent assembly that, if done correctly, would be indistinguishable from a single printed part.
schommelpont (Meshed)-Colored.3mf (92.1 KB)
Hi again. My support results are much better now that I have changed a lot of settings according to the suggestions on this (and the Reddit) forum. However, one should still be careful… The support linewidth of 0.25 (instead of the standard 0.45) does make removal much easier but may lead to broken support trees when they have long branches. I have experimented with a beautiful but truly complex design (not by me alas), a 3D version of Eschers Relativity drawing.
Relativity.3mf (180,8 KB)
I’m glad we were able to help!
Your point about the .25mm supports is well taken. With tree supports they can be thin enough that they don’t hold together well or are unstable.
I have made a request on the OrcaSlicer GitHub to separate the line thickness setting for “support” from “support interface.” That way we could have thick, strong supports, with thin delicate interfaces. I hope they do it!
I agree, used Cura with my ender printers for years and i cannot figure out how to reduce the amount of supports in Bambu vs what Cura would create, attached is the same part in Bambu and in cura, and I like the reduced number of supports in cura, trying to find out how to do this in Bambu.