Best way to edit STL file?

Bambu Lab Software Bambu Studio
1

I wanted to print something, but it wasn’t good enough and I wanted to make some changes to it. I will go through a few methods I tried and hopefully somebody has a better suggestion for me to use in the future. In my attempts, a combination of Fusion 360 and Bambu/Orca was the best.

I first started with Blender and the way you have to select every triangle face and line is a pain, and sometimes the triangle faces extend past the section I want to select. I tried the bisect command, but it didn’t work. And in general, it seems like a pain to use. Maybe I will take some time to learn it one day.

I then looked into converting it to a parametric object in Fusion 360. Since I have the free version, there is a new conversion method that I couldn’t use. The method available to me first does a basic conversion, then you have to delete extra triangle faces and fusion will fill it in to get a smooth body. While fillets have to be deleted and then you use the loft tool to add them back. This also seemed like a pain.

I really like the way you can split up an STL object in Bambu Studio, I wish other STL editors were so easy to use and worked as well as Bambu/Orca, and/or I wish Bambu/Orca had better CAD like drawing tools.

I only had to adjust one section of the object, so I brought the STL into Fusion and used the convert mesh tool. Then I traced a sketch of the section I wanted to replace, extruded it and made all my other adjustments. I then recorded the Z position of the bottom of the section object. I brought the original STL into Orca and split it 0.1 mm higher then the section object Z location. I kept the bottom part of the split and discarded the top. I then brought in the section object and assembled it with the original STL I split. I positioned the section object at the correct Z height, then zoomed in and positioned it correctly in the X and Y axis.

My final method with Fusion and Bambu was the best/easiest of the three, but I can’t believe there is not a better way. I want something that works like Bambu Studio but has more and better design tools.

2

I’ve been on a similar search for the “perfect tool” for mesh editing. This came as an outgrowth of searching for the one tool that can convert raster to vector formats such as taking a pixel image of a font and converting it to a vector format. The science is the same but just in two dimensions and nobody has been able to solve that problem either. This challenge for me goes back all the way to the late 80’s when I was trying to convert CAD drawings. Although the tools are better, they are still only approximations and fail when one scales the object.

This effort for me goes back much further than my 3D printing experience but rather when I was dabbling in VR game development and I was taking OBJ models and trying to convert them into a dynamically changing scene. You learn a lot when you have to write code to manipulate something inside of a 3D construct. If anyone remembers POVRAY, that’s where I cut my teeth on 3D technology. Back in the eighties I had delusions of being the next ILM. :wink:I did immerse myself into blender all the way back to 2.6 which was over 10 years ago. Although I keep a current copy loaded on my system I’ve found that the trajectory of the tool has become more complicated to use than I have a taste for, albeit far more powerful too. But Blender is not my first go to for mesh editing anymore.

You could try Autodesk Meshmixer which many of the features found there have been migrated into Fusion360 but it has a more simplified interface. Even though development ceased in 2018, it can still be downloaded. https://meshmixer.com/download.html Consider it Fusion360-lite for mesh editing and you get the idea of the audience it’s aimed at but for slicing and dicing, it may be that goldilocks zone of what your looking for.

My one takeaway regarding my effort for the holy grail of mesh to vector conversion tool, is that even the very, very expensive professional mesh to vector conversion tools are only marginally better than the free ones included in things like Fusion360. There’s also a complete absence of an open source effort for mesh to vector conversion. So my conclusion is that this is one science project that has eluded modern mathematics.

What I might suggest is this online converter.

While not perfect, of all the conversion tools I’ve tried to use, it sucks the least. Mind you that anything above a certain vertices count and it pukes like the rest. Of course like any mesh to vector conversion, it really only works well on geometric shapes and completely pukes on organic shapes.

Using that tool, once you have the STL converted into a STEP file then you can start editing using CAD parametric tools.

One decent video on the topic which gave me some additional insights was from the Teaching Tech Channel. His video on “How to modify STLs” gave me some additional ideas using OnShape which I found handy and find myself using quite handy especially when I am trying to borrow a component of an STL that I found online that is just a piece of an apparatus I want to design. One example of this is the articulating modular camera mount that was designed for the PI.

I’ve used just about every tool in my search for the perfect mesh editing tool. Some have some cool features but none are a complete suite of tools. So if you find one that solves everything, please share it.

Here’s a random list of things that I’ve been searching for that I just can’t find in any tool.

  1. Conversion of a circle into an actual arc-based mathematical construct. All current tools balk and just create a vertices-based object, it other words, they just say ■■■■-it and take the short cut.
  2. True parametric conversion. I want something to allow me to take a mesh and once converted into a STEP file, also assign length and width to the X, Y and Z axis. You would think that this would be easy, but it’s absent from the industry and nobody seems to want to conquer this challenge.
  3. Provide for a closed mesh upon import. I mean really? We can send a man to the moon but we can’t close a manifold? I realize first hand how difficult a mathematical problem this is, but nobody has been able to solve it.
  4. Break away from the triangle-only vertices mesh route and incorporate other more flexible polygons. I’ve read some white papers on the topic and the math is definitely over my head but I got the gist and it is possible. However, nobody wants to tackle solving the more than 3 point solution that all vertices-based mesh models rely on.
2 Likes
3

Thanks, I will try it some of those things.