Re: [mor4ansys] Is Mor4Ansys right for me?

[Evgenii Rudnyi]

On 31.07.2012 11:59 ansibin said the following:
> Hello,
>
> I have read many about mor4ansys but i'm not sure if it is the right
> product for me or if there are functions in Ansys witch do the same.
>
> I have a big structure with about 1 million knodes. On this Structure i
> glue a piezo. Then i want to extract the state space form to use the model
> in Matlab Simulink.
> I tried this an my Problems are:
>
> Export the state space matrices with spmwrite -> only mechanical DOFs
> Normal Substrutring and HBMAT -> results are not good

You may want to look at

Automatic Generation of Compact Electro-Thermal Models for Semiconductor
Devices
http://modelreduction.com/Applications/Thermal.html

where you find a comparison of substructuring and MOR for a thermal
problem. Note that for a mechanical problem, the difference between
substructuring and MOR is smaller, the comparison is here

mor4ansys: Generating Compact Models Directly from ANSYS Models
http://modelreduction.com/Applications/Structural.html


> CMS Substrutring -> Only mechanical DOFs are allowed.
>
> Is there any way to export state space Matrices from Ansys with electrical
> and mechanical DOFs in matrix size that can handle by Matlab?

You probably have seen examples here

http://modelreduction.com/Applications/Piezoelectric.html

Some problem is that if you look at equations, for example in the ANSYS
Theory Reference, a piezo-electric problem is not directly compatible
with MATLAB. That is, formally you cannot convert it to the state space
form, as the M matrix is degenerated. Do not forget that the K matrix is
indefinite that also adds some problems.

Some solution will be presented soon at MME 2021

NEW MODELLING APPROACH FOR MICRO ENERGY HARVESTING SYS-TEMS BASED ON
MODEL ORDER REDUCTION ENABLING TRULY SYSTEM-LEVEL SIMULATION

You may want to contact Tamara Bechtold for a preprint.

Evgenii