Voxel Smoothing.

[UOMO]

Hello there,
Do you know if in the current TexGen version is possible to implement any smoothing algorithm for the voxel meshing trough python scripting? I know that in this work it has been done successfully "Potter, Pinho. Mesh generation and geometrical modelling of 3D woven composites with variable tow cross-sections. Comput Mater Sci 2012.51:103-111"...
Thank you in advance!
/UOMO.

[louisepb]

Hi Uomo,

There is nothing incorporated into TexGen at the moment although I know that people have implemented the algorithm that you refer to. A colleague has developed a method of smoothing the voxel mesh using Hypermesh and I am currently working with him to implement this in TexGen. It's a 'watch this space' at the moment!
Best regards,
Louise

[UOMO]

Hi Uomo,

There is nothing incorporated into TexGen at the moment although I know that people have implemented the algorithm that you refer to. A colleague has developed a method of smoothing the voxel mesh using Hypermesh and I am currently working with him to implement this in TexGen. It's a 'watch this space' at the moment!
Best regards,
Louise

Thank you Louise!
Is that the only option you know so far? Have you guys tried to "filter it" somehow in Matlab for example?
How complicated is the smoothing procedure you have in HyperMesh?
I'm really interested and need something like that, since I'm dealing with stress concentrations, and having voxels is not a good idea...
Thanks!!
/UOMO.

[louisepb]

Hi Uomo,
Sorry for the delayed reply - I've been out of the office. I'll make some enquiries and get back to you.

Best regards,
Louise

[UOMO]

Hi Uomo,
Sorry for the delayed reply - I've been out of the office. I'll make some enquiries and get back to you.

Best regards,
Louise

Thank you Louise, looking forward to it.
Best,
/UOMO.

[UOMO]

Hi Uomo,
Sorry for the delayed reply - I've been out of the office. I'll make some enquiries and get back to you.

Best regards,
Louise

Thank you Louise, looking forward to it. BTW, the algorithm you have is that one used in this conference paper?: http://www.escm.eu.org/eccm16/assets/0850.pdf (for going from Fig.5a to 5b).
I still interested, thank you!
/UOMO.

[louisepb]

Hi Uomo,

Yes, the method is the one described in the conference paper. The theory is described in Qing Pan's thesis. You can find it here: https://www.dropbox.com/s/vxwt9u0u3tqpn ... 5.pdf?dl=0

I am trying to find out whether we can have access to the code to use with TexGen. I'll have to get back to you on that one.

Hope that helps,
Louise

[UOMO]

Hi Uomo,

Yes, the method is the one described in the conference paper. The theory is described in Qing Pan's thesis. You can find it here: https://www.dropbox.com/s/vxwt9u0u3tqpn ... 5.pdf?dl=0

I am trying to find out whether we can have access to the code to use with TexGen. I'll have to get back to you on that one.

Hope that helps,
Louise

Thank you very much for the useful information. In fact, I have kind of implemented such algorithm in Matlab. I used the suggested lambda=0.631 and mu=-0.673. However, I don't get the nice smoothness that is shown in the thesis (Figure attached after 500 iterations). Also, I just selected a different group (outer surface of the domain of the yarn, attached) and got the result in the last figure attached. Since it is not specified how the groups are smoothed (same weights, etc.)... I have some questions...

- Smoothness is only performer only in the Z- direction, isn't it? i=3.
- Smoothness is performed in all the groups? i.e., in the 10 of then?
-How the neighbours are defined? I mean, belonging to a group or taking into account neighbours from other groups?
- What does the group of interface and i-direction (with i=X,Y,Z) Edge nodes means? Is it the cross section of the yarn intersecting with the surfaces?... .
- The new "mesh" will seems not be periodic anymore...have you deal with this?


Thank you again Louise!
Best,
/UOMO.

[louisepb]

Hi Uomo,

Unfortunately I only have as much information about this as you do. The researcher has finished his PhD and is no longer in the office but I have contacted him to see if he can shed any light on this. I'll get back to you if I hear any more.

Sorry not to be more help,
Louise

[UOMO]

Hi Uomo,

Unfortunately I only have as much information about this as you do. The researcher has finished his PhD and is no longer in the office but I have contacted him to see if he can shed any light on this. I'll get back to you if I hear any more.

Sorry not to be more help,
Louise

Thank you Louise! I improved a bit such a code, attach image... however still looking forward to hear from you back since still, it's not as shown in the thesis...
Cheers!
/UOMO.

[George_Bal]

Hello,

is the smoothing script from the thesis mentioned above available after all (Taubin smoothing technique for Texgen voxel mesh)?
I think it would help a lot of people, especially when it comes to strength prediction.

Thank you,
George.

[louisepb]

Hi George,

I'm afraid that the script from the thesis isn't available. Another researcher in our group is currently working on a voxel smoothing method which will, hopefully, be integrated into TexGen in due course. I'm not sure what the timescale will be on that though.

Best regards,
Louise

[George_Bal]

Thank you for your answer Louise.

Best,
George.