Stitching Surfaces to make a complex solid

[rory.brittain]

Long story short I am trying to perform a computational fluid dynamic analysis on a complex shape (glass beads suspended in resin, I want to see how fluid will flow around the beads) but I need to convert the mesh into a solid object and am having trouble doing so. Here is my workflow. I should add I have no formal train in fusion so a lot of this is guess work and youtube

1. Import the .stl file of beads
2. Reduce the number of triangle/faces by using the "Make closed mesh tool" (the re-mesh and reduce mesh tools don't seem to work very efficiently) 
3. Convert the mesh to a surface
4. Use the validate tool to try and remove any issues
5. Click the stich button to try and create a solid
   1. This is where my issue begins, the tool states I have a 9 free edges. I am unsure if these need removing, and how to do it

[mango.freund · Accepted solution]

hi @rory.brittain  it is not possible to edit a photo, which should then be your body.

for this we need the mesh body, the stl. it is much easier to do this work with the app "meshmixer" and later the simulation in fusion.

----mango

[rory.brittain]

Thanks for getting back to me @mango.freund. 

I am a bit confused about what you mean with editing the photo, sorry! Do you mean you can't edit the photo I uploaded of the glass beads and would like a copy of the .stl? 

[mango.freund]

but yes, we need the work as you can offer it!

photos are helpful but not useful for the solution - don't you think so too? --mango

[jeff_strater]

Stitch reports free edges as an indicator that those edges are not going to be joined to other edges.  It is only a problem if you are trying to make a closed solid.  I can't tell from your image if that is the case or not.  Free edges are highlighted in red.  In this image, there are 5 planar faces (the top is open), and those 4 edges are free:

so, the answer to what you should do with those free edges depends on what you want the result to be,  If there are free edges, the result will be a surface body.  If that's OK, then you are finished.  If you are trying to get a resulting solid, then you need to give those edges something to connect to.

However, I feel I have to warn you - this is not a recommended workflow, to be honest.  Mesh -> a highly faceted BRep (each mesh facet is a separate face) will lead to very slow performance at best.  What are you trying to achieve here?

[rory.brittain]

Thanks for helping @jeff_strater. To give a bit more detail, the glass beads have been digitally recreated from an x-ray tomography scan, the software of which only outputs a surface mesh as far as I can tell. I have been using the meshing tools to reduce the number of faces, and will have access to a very powerful PC next week. I think mango has solved my solids problem by telling me to use meshmixer, but now its a case of trying to reduce the number of surface on the Brep model

I wanting to see how simulated flow around the beads compares to actual flow. 

[TheCADWhisperer]

---

@rory.brittain wrote:

I have been using the meshing tools to reduce...

I think mango has solved my solids problem

...now its a case of trying to reduce the number of surface on the Brep model

I wanting to see how simulated flow around the beads compares to actual flow. 

---

I don’t think this is the way to go.

C.an you Attach the original stl file before any attempt to reduce the number of faces?

[rushabhthakkar_910]

You can Export your File like this.. 

File> export> Type:- (.stl)> Location:- (Choose a Location where you want it to be saved on you Laptop/PC) and attach it here...

[rory.brittain]

I have tried to upload the .stl a few times now and it wont let me saying the file extension doesn't match. Not sure why as windows seems to think is an .stl 

[TrippyLighting]

ZIP .stl files and then attach.

In general such highly faceted geometry, when converted into a BRep has a relatively low chance of completing any form of FEM / Fluid Dynamics analysis. 

Usually BRep geometry will go through a meshing process in the FEM - very different from the .stl algorithm -  and simplifying BRep geomerty is a common practice.
The high facet count of .stl file (even when reduced) does not allow simplification and will possibly result in a very dense FEM mesh.

[rushabhthakkar_910]

Right Click on the File> Click on Add to Archive> Choose a Location you want it to be saved to> And share it here...

[mango.freund]

hallo @TrippyLighting es ist das gleiche wie mit dem "freund" vor einiger zeit -- eine "never ending story"

es muessen zutiefst innenliegende loecher/flaechen geflickt/gefuellt werden um die simulation gut werden zu lassen.

ich hab das bei dem mit fluidkoerper gefuellt -tagelang- und das hat immer noch nicht gereicht.

[TrippyLighting]

@mango.freund assuming that a simulation can be successfully completed on a highly faceted solid, you are correct. In that case the first task is to create that solid. I propose that Fusion 360 isn't the best software to do so, but it does have some tools in the Mesh environment to repair common mesh errors.

However, it cannot simply be taken for granted that a simulation on such a highly faceted BRep will be successful. The Algorithms that are employed to create the usually tetrahedral volumetric mesh structures required to fluid dynamics and other forms of FEM analyis assume relatively large (in relation to the overall size of the object) smooth BRep / NURBS surfaces. 

As such spending hours on repairing a mesh might not be the best use of time 😉

[rory.brittain]

Thanks you all for you help it really means a lot!

Bit of an update for you. I have managed to make the model a "solid" using mesh mixer so thank you @mango.freund for that tip, saved me loads of time! I have tried to reduce the mesh by cutting of the "sides" of the model (see pic below), this also has the added bonus of giving a nice planar surface for the CFD software. I have followed the advice regarding uploading the .stl and the website is still throwing the same error message. 

[TheCADWhisperer]

---

@rory.brittain wrote:

.I have followed the advice regarding uploading the .stl and the website is still throwing the same error message. 

---

What error are you getting when attempting to Attach your *.stl file to this forum?

I recommend that you STOP whatever other actions you are attempting as you are getting a lot of “questionable” advice in this thread

[rushabhthakkar_910]

I did not have any issues attaching a .stl file here...

[rory.brittain]

Hi Guys 

Bit of an update for you all, I got it to work (kind of) as most you said, the geometry was to complex for CFD. Hopefully access to the more expensive PCS will get it working

[TheCADWhisperer]

---

@rory.brittain wrote:

Hopefully access to the more expensive PCS will get it working

---

I recommend that you Attach the file here - I don’t think the computer used will have any relevance on the issue.

[rory.brittain]

Oh no, the stitching issue is resolved, as I mentioned using meshmixer to convert it to a solid was much easier than using fusion. I using Simscale to run the fluid Sim as well as CFD ultimate. Simscale is free and limits you on the the amount of time a sim will take, and the compelex geometry meant it would take a while. I am still learning how to use CFD ultimate, but fairly sure my old laptop will struggle. 

[TheCADWhisperer]

Attach your file here and end all doubt.

This is my final attempt.