We are looking into it. We will let you know when we have an update to share. This will most likely take a few days. Thank you for you patience.
Thank you Jeff, I tested the files you attached above, and the “index, dimension 1, above upper bounds” type error as follows is persistent. I want to know whether the above attached project modified by you can successfully run in your computers. Thank you.
And another question referring to the following quote is that if the reason for the quoted problem is unclean .stl file, why the “index, dimension 1, above upper bounds” type error still appeared when I replaced the origin .stl with the clean one uploaded. Thank you.
And could you please recommend me a software to remove the duplicate entities including facets, vortexs and lines? Since I can not find the tools and commands in solidworks to do the above job and flip their norm (if you know, could you please tell me)? It would be better if it is open-sourced. Thank you.
Hey Jeff, I hope I am not bothering you too frequently, I am willing to know whether you have any updates for the problem I have posted. Now that you have posted the gas velocity distribution across the chamber, could you please tell me how long you have simulated in that case and can you give me the evolution of bulk density, volume fraction and outlet mass flowrate as well? Thank you for your consideration. I need your help.
We have a potential fix for the meshing issue on Windows that we will make available in the next few days (21.3.1 point release). Even if you can mesh, it will take too long to run on a PC, you will need access to a cluster. I only ran for 0.1s to troubleshoot and I did not analyze any data.
Thanks for your reply, but when I encountered the “index, dimension1, above upper bounds” problem when I mesh the cleaned files provided by you.
Why my PC can not even build that? due to RAM, Algo Lib, stochastic effect, version difference from you, windows operation systems?
I really want to know the reason. Thank you.
Hi Jeff this info is very important for me. Could you please answer me? Thank you.
Ju Wu - please understand that we all have full-time jobs and MFiX support is provided on a best-effort basis. We are all working on our own research and development of MFiX. Please be patient when asking questions on the forum and do not repeat the same questions in different threads. Thanks.
I am personally using Blender right now, but I have used meshlab in the past. I don’t know about solidswork, but I imagine they have some clean up tools.
Regarding the out of bound error while meshing on Windows, please try the 21.3.2 point release. I hope this fixes your issue.
Thank you Jeff, could you please explain why this type of problem happened only under WINDOWS? Now that I found out that the the installation and create-environment command under WINDOWS and LINUX are the same, so I assume they create and install the same software under Anaconda. Thank you.
Thank you Jeff, how to demonstrate two slices at the same time as follows?
Here each slice is saved in a different vtu file, and the two files are loaded in Paraview. If your vtk region is set as a plane, it will save a slice (on cell thick).
You can also do the slicing in Paraview if you have the full 3D domain in the vtu file.
Thank you Jeff, as you said you only simulate 0.1s, can you tell me your settings for computing cluster such as the node number, the cores, the task-per-core and so on? And I wonder whether the project you attached finally converged or not since I found our defects/holes in the generated mesh by your .mfx and .stl files. Thank you in advance for your answers.
I used 4 nodes, 40 cores per node, so a total of 160 cores. I can’t say if this would eventually run to completion . I know you had some pretty high velocities so it is possible it would fail at some point.
Hi Jeff, I have imported the same .stl file to the blender 2.9, there are two functions one is “merge vortex” and the other is “clean up” under “modeling” bar “mesh” subbar. But it shows that 0 vortex removed. So could you please tell me which method I can use in Blender to detect the double faces, vortex automatically and remove them then? Thank you in advance.
I used an older version (2.79) that has a “remove double” option. Maybe the equivalent for 2.9 is Mesh>cleanup>merge by distance. You may need to adjust the merge distance and make sure all faces/vertices are selected.
Thank you Jeff, I can modify the merge distance, I have the following questions.
-
so what is the recommended “merge distance” you have configured?
-
And whether the “optimal” procedure is to generate the extra fine stl with as small deviation and angle tolerance as possible and then set a relatively larger “merge distance” in the Blender to remove surplus entities while avoid bad representation? How to trade off that?
-
As you said, we have to remove duplicate entities including vortex, faces, edges, so how do you remove duplicate faces, edges in Blender? with the same function “merge by distance” under cleanup in mesh bar?
Thank you in advance.
- There is no universal value. It depends on the scale and unit of the stl file. If the STL file coordinates are expressed in meters and you want to merge points or remove details less than 0.1 mm , then the distance will be 0.0001. If the unit is in millimeters, the the distance would be 0.1. Start with small values and you can gradually increase it. If it is too large you will start losing details in the geometry.
- To me the ideal procedure is to generate the cleanest STL file possible with few elongated triangles and do the clean up to remove duplicates and the few bad triangles.
- In 2.79, “remove double” removes duplicate vertices, edges and faces. Not sure about 2.9, you should see the outcome of the merge telling you exactly what was done.