Jeff Hiller
COMSOL Employee
Please login with a confirmed email address before reporting spam
Posted:
3 years ago
28 juin 2022, 09:30 UTC−4
Hello Chui,
See Henrik's answer here and mine here.
Best,
Jeff
-------------------
Jeff Hiller
Hello Chui,
See Henrik's answer [here](https://www.comsol.com/forum/thread/307921/can-i-change-material-property-for-each-element-if-yes-then-how?last=2022-05-17T13:50:18Z) and mine [here](https://www.comsol.com/forum/thread/145781/gradually-change-material-parameter-s-as-a-function-of-the-domain-position?last=2017-06-12T12:38:29Z).
Best,
Jeff
Please login with a confirmed email address before reporting spam
Posted:
3 years ago
28 juin 2022, 21:26 UTC−4
Hello Chui,
See Henrik's answer here and mine here.
Best,
Jeff
Thanks for your answering.
Unfortunately, values of elastic modulus and poisson ratio can not be expressed by a specific function. They are statistically related to the element location. I need to assign element mechanical properties after meshing.
For better understanding and simplicity, assuming that I want to change the elastic modulus of a specific element, and now I have retrieved the global stiffness matrix, as well as node and mesh information, in livelink for matlab. What should I do next?
Thanks, looking forward to your reply.
>Hello Chui,
>
>See Henrik's answer [here](https://www.comsol.com/forum/thread/307921/can-i-change-material-property-for-each-element-if-yes-then-how?last=2022-05-17T13:50:18Z) and mine [here](https://www.comsol.com/forum/thread/145781/gradually-change-material-parameter-s-as-a-function-of-the-domain-position?last=2017-06-12T12:38:29Z).
>
>Best,
>
>Jeff
Thanks for your answering.
Unfortunately, values of elastic modulus and poisson ratio can not be expressed by a specific function. They are statistically related to the element location. I need to assign element mechanical properties after meshing.
For better understanding and simplicity, assuming that I want to change the elastic modulus of a specific element, and now I have retrieved the global stiffness matrix, as well as node and mesh information, in livelink for matlab. What should I do next?
Thanks, looking forward to your reply.
Henrik Sönnerlind
COMSOL Employee
Please login with a confirmed email address before reporting spam
Posted:
3 years ago
29 juin 2022, 02:13 UTC−4
Updated:
3 years ago
29 juin 2022, 02:14 UTC−4
Even if the material data are statistically dependent on the location, it seems easier to do it inside COMSOL multiphysics than trying to work outside.
Here are two blog posts about generating random material data:
https://www.comsol.com/blogs/how-to-generate-randomized-inhomogeneous-material-data/
https://www.comsol.com/blogs/paper-mechanics-and-the-benefits-of-modeling-paperboard-formation/
-------------------
Henrik Sönnerlind
COMSOL
Even if the material data are statistically dependent on the location, it seems easier to do it inside COMSOL multiphysics than trying to work outside.
Here are two blog posts about generating random material data:
Please login with a confirmed email address before reporting spam
Posted:
3 years ago
29 juin 2022, 03:56 UTC−4
Henrik
Hi Henrik, thanks for your reply. The model I want to build looks like as the following picture .
The gray part is material A, and the black part is material B. For all elements belong to material A, elastic moduli are spatially random distributed in the domain of material A, but their values follow some kind of statistical distribution. The same with material B.
I don't think it can be fullfilled without programming, and the links you give me may not suitable for my problem.
I prefer to calculate and assemble element stiffness matrice in matlab, and then replace the original matrice given by comsol.
Could you give me any suggestions?
>Henrik
Hi Henrik, thanks for your reply. The model I want to build looks like as the following picture ![](https://ars.els-cdn.com/content/image/1-s2.0-S0892687516300395-gr4.jpg).
The gray part is material A, and the black part is material B. For all elements belong to material A, elastic moduli are spatially random distributed in the domain of material A, but their values follow some kind of statistical distribution. The same with material B.
I don't think it can be fullfilled without programming, and the links you give me may not suitable for my problem.
I prefer to calculate and assemble element stiffness matrice in matlab, and then replace the original matrice given by comsol.
Could you give me any suggestions?