Hi Fusion MFG community
As you may have noticed, the Mesh workspace just graduated from being a tech preview and is now a part of Fusion 360 user interface by default with the July 7, 2021 release. With this release you can:
The workflows we implemented contains several intelligent default settings and the ability to preview the outcome of each command before executing them, so that users don’t have be an expert in working with mesh data to benefit from this functionality.
This release also comes with several known limitations:
3mf related limits
Workspace related limits:
Face group related limits
With upcoming releases of Fusion 360, we will
We also would love to hear back from you.
So here are some of our questions:
No “Insert mesh” + “convert to solid” failed feedback please. You need to “generate face groups” before attempting to convert mesh bodies to solid bodies first 😊
There are no wrong answers here, we are just looking to start a conversation on how you plan to utilize the new mesh functionality in your workflow.
All questions, comments, and criticisms are welcome.
Thanks
Sualp Ozel, PE
Senior Product Manager
Additive Manufacturing
Would love to see MeshMixer's brush based "Sculpt" tool available directly in Fusion 360. Especially if the Fusion 360 implementation includes MeshMixer's auto mesh refinement when brush sculpting.
Our current workflow is:
Would love to keep this workflow entirely within Fusion 360. Currently Fusion 360 handles every step except 3.
What are you making?
There should be a organic mesh modeling update coming in the next months.
I've said previously:
MM should be 100% integrated with Fusion. Given the advancements 3D printing, the usage of Meshes, the prominence of 3D scanning and reverse engineering, and also specifically in my field the usage of CAD/CAM/Print for orthopedics & medicine, shouldn't that be the obvious end goal? Now even moreso with the official announcement of the abandonment of MM development & support. Surely cannot just let such a great swiss army knife die. I gladly pay for this.
Minature figurines, display props, and accessories. Some definitely have an organic look and MeshMixers Sculpt tool is indispensable for this.
Do you have a source for the "organic mesh modeling update"? A search did not yield anything. I would be interested to see what new features are to be expected.
https://youtu.be/MlbHz1rlppU?t=2933
And read the splash screen: https://www.meshmixer.com/
And I had a web meeting with some sales reps a few months ago, they said ya in the works for next year.
We the video is accurate. We are planning on releasing a 3rd mesh conversion option better suited for organic shapes in early 2022. In addition, we have focused our efforts to eliminate a lot of the shortcomings I recorded in my original post. So be on the lookout for a lot of improvements & bug fixes around Uploading 3MF files to Fusion 360, 3MF support with colors and textures, and improved support for mesh bodies in rendering. I hope to roll out these changes in the first half of 2022.
Thank you for the feedback. I agree, it would be very useful to have Mesh sculpting tools in Fusion 360 in the future. However we are not actively focused on those workflows right now.
Amazing. Will this have any of the organic mesh modelling tools of MM? Such as soft transform with a falloff. Or Smooth boundary of a selection?:
https://help.autodesk.com/view/MSHMXR/2019/ENU/?guid=GUID-AF536B4D-4BCF-41EF-AD66-38A3E161E467
https://help.autodesk.com/view/MSHMXR/2019/ENU/?guid=GUID-E668ED29-8943-4534-A2B4-45135297FF3C
One workflow that I have been struggling with is modifying face groups. Often I can use the Accurate method to initially generate face groups for a mesh body. This works out very close to what I want most the time, however, if the mesh groups cross a natural hard edge the resulting solid body either fails or you get messy geometry. To fix this you can select the face group(s) in question and re-generate just those with the Fast option and choose a small angle. BUT then you have to combine LOTS of triangles that were already grouped originally.
It would be a LOT faster if there was a command called Edit Face Group where I could choose to remove certain triangles from a face group. Even better if the command would allow me to select individual triangles from a neighboring face group and steal them for the current face group.
@JeffBradway I agree. The current workflow required to edit face groups is very clunky and inefficient.
I did find a decent work-around. First double click the mesh to enter the direct edit mode. Then there is the command to create a face group. When you use that command it steals triangles away from any other existing face groups. Then you can combine that new face group with the one you want, or leave it as its own. Working in the direct edit mode also has the advantage of being faster and not generating as much history that Fusion will have to run through.
Here is my feedback. Over the last month, accurate 3D scanners get much more affordable, so I guess a lot more people are going to use 3D scans in Fusion.
- Fusion slows down a lot if Meshes bigger > 100 Mb are imported. Even a small part is often bigger than that. So dealing with big data must be improved.
- To not slow down Fusion too much, the 3d scanned mesh must be reduced so much, that the geometry is getting a lot less accurate. Consequently, reverse engineering gets less accurate. So the super-accurate 3D scans are not worth anything if they can not be used in full resolution in Fusion.
- I used to make face groups for critical areas (bores, mounting points etc.) and then invert the selection, to only reduce the mesh on non-critical areas. This works usually for one operation. If I want to reduce the mesh a second time, Fusion brings up a failure message and can not do the operation. I'm not sure if this is a bug or if there are limitations between the border of high and low-density mesh.
- I'm really missing proper geometry detection tools like those found in GOM inspect. This is really important and speeds up reverse engineering and makes it more exact. ATMO in Fusion, a simple plane can only be defined by 3 points. The proper definition should be done with a Gaussian algorithm to best fit a plane to the scanned surface. Tools are needed to quickly define and measure regular geometry, like inside-, outside diameters, rectangular surfaces.
Here is a good video. All he does in Gom inspect before he imports the model into Fusion, should be possible directly in Fusion https://www.youtube.com/watch?v=V3xAwpENbU4
- Lastly, I would like to import point clouds directly into fusion, clean them up and finally mesh them adaptively. So critical areas should have high-density mesh, non-critical rough mesh. This could help a lot to bring the file size down and speed up Fusion.
I know you guys are working on most of the issues I have with the Mesh workflow, but I was curious if there is any talk about having "repair mesh" be an embedded option in the 3D printing export dialogue. It's not a big deal, but some slicers out there are very bad at repairing meshes themselves, and even Preform apparently uses a Fusion utility when it repairs meshes.
So, when I need to fix an exporting STL before it will print, I have to open a separate project, re-import the mesh Fusion just made, repair it, and then re-export it. It doesn't take that long and it usually works great, but it feels a little silly.
I've returned to trying out the Mesh Environment in F360 during the Trial Extension period to try out the mesh to solid conversion tools. So far, nothing I have works with the converter (and I suspect this is my inexperience) despite hours of working with re-meshers, blender, Insta-Meshes and others to prep and simplify mesh files.
In doing all this a number of points struck me that I WISH were in Fusion:
1. A mesh simplifier that included a change to Quad Meshes for manual work towards a form. Maybe a plugin for insta-meshes (flawed) could do this, but I'd like to see something reliable developed.
2. A Mesh Scaler tool that would allow a mesh to be precisely scaled between two (or 3) vertices for exact linear scaling, and exact X/Y/Z transformation.
3. (Most Useful) A way to 3D sketch on a Mesh using Vertices as Points (so you don't have to use the create mesh section tool to create surfaces or sectionally recreate a mesh object as a solid.
e.g. below is a Miata Dash Tombstone in both basic mesh form and an imperfect 8K quad re-mesh of the OBJ file. Neither auto-converts to a solid, but heck with a very small selected set of vertices points I could EASILY create a 3D sketch with lines and arcs between points & then use surface lofts between them + thicken to recreate this as a solid. Far easier than trying with section sketches. (All this if I were dead certain the scale of the mesh were correct). Perhaps this suggestion would be better parsed to the Sketch Interface Team?
Also better support for 3MF because as I understand it, that format can ALSO do Quad Meshes as WELL as contain measurement data & color data etc.
~Paul K.
Thanks for the suggestion. Yes doing an edit on face group to reduce mesh count does help to an extent, and allows a partial body to be generated. It's a start. But it does bring up another suggestion.
If one could 3D project (include 3D geometry of the outline) of a face group, THAT would be a useful thing, instead of just being able to project a mesh body cross section geometry on whatever XY plane a sketch is attached (similar to create mesh section sketch I suppose).
I'm sure much of this comes down to my own inexperience in this work flow, but these were a few things that struck me as features (if available) that could make the process (from whatever way one attacks it) easier.
These were all just practice items. The Tombstone seemed like one of more *simple* STL/OBJ files I have that would be nice as a solid instead of a mesh. I'll probably try the organic conversion before time runs out.
Anyway, maybe these suggestions will inspire some updates.
~Paul K.
converting triangle-based meshes to quads is a much more complex ordeal than you might think. The only software I know of with a truely robust tool for this is the full version of Zbrush.
At the moment fusion simply can’t do much with triangle-heavy meshes. They’re just too complex to transform into CAD elements.
In the Autodesk Portfolio there is also "Recap Photo" which is a good candidate as a Quad remesher...
you have an export setting dedicated for Fusion (Quad in medium quality) if I remember well.
It should be a good idea to include those algorithms in Fusion directly...
Can't find what you're looking for? Ask the community or share your knowledge.