Hi Ronghai,

you can constrain the displacements using the periodic boundary conditions - then the strain on the boundaries would be the same as well. Or you need something else?

r.

On 03/15/2016 03:26 PM, Ronghai Wu wrote:

Hi sfepy developers and users,

I am wondering is it possible to have periodic strain boundary condition in sfepy? For example, in a 2D uniaxial tension along y direction, how to constrain the strain at left and right boundary to be same when solving mechanical equilibrium equation?

Regards Ronghai

Hi Ronghai,

So you mean something like enforcing a symmetric deformation of the neck? That should be automatic, provided the mesh is good enough and the loads/BCs are symmetric.

As for constraining the strain, maybe you would need some kind of a mixed formulation, where strain is an independent variable, but I am not sure I understand what you want to achieve.

r.

On 03/16/2016 12:00 PM, Ronghai Wu wrote:

Hi Robert,

Thanks for your reply. Here I do not mean periodic displacement boundary condition, but periodic strain boundary condition. The reason is, still taking 2D uniaxial tension along y direction as example. If I fix the "u.x" both at top and bottom boundary. There will be necking at left and right boundary as tension goes on (but I do not want to have free surface boundary condition). However, if using periodic displacement boundary condition, there would be no displacement at left and right boundary, and the elastic stress-strain would be larger than free surface case. I am thinking if there is periodic strain boundary condition, then the necking is retained and stress-strain would be between free surface and periodic displacement cases. So, I am wondering is it possible to have periodic strain boundary condition in sfepy, sounds kind of odd though.

Regards Ronghai

在 2016年3月15日星期二 UTC+1下午10:28:54，Robert Cimrman写道：

...

Hi Ronghai,

you can constrain the displacements using the periodic boundary conditions

then the strain on the boundaries would be the same as well. Or you need something else?

r.

On 03/15/2016 03:26 PM, Ronghai Wu wrote:

...

Hi sfepy developers and users,

I am wondering is it possible to have periodic strain boundary condition in sfepy? For example, in a 2D uniaxial tension along y direction, how to constrain the strain at left and right boundary to be same when solving mechanical equilibrium equation?

Regards Ronghai

OK, good luck with that!

r.

On 03/18/2016 11:32 AM, Ronghai Wu wrote:

Hi Robert,

I thought it over, and I noticed what I wanted to do just could not help that much in scientific point of view but complicate solving procedure. So, I will use periodic displacement boundary condition. Anyway, thanks always for your help.

Regards Ronghai

在 2016年3月18日星期五 UTC+1上午11:16:37，Robert Cimrman写道：

...

Hi Ronghai,

So you mean something like enforcing a symmetric deformation of the neck? That should be automatic, provided the mesh is good enough and the loads/BCs are symmetric.

As for constraining the strain, maybe you would need some kind of a mixed formulation, where strain is an independent variable, but I am not sure I understand what you want to achieve.

r.

On 03/16/2016 12:00 PM, Ronghai Wu wrote:

...

Hi Robert,

Thanks for your reply. Here I do not mean periodic displacement boundary condition, but periodic strain boundary condition. The reason is, still taking 2D uniaxial tension along y direction as example. If I fix the "u.x" both at top and bottom boundary. There will be necking at left and right boundary as tension goes on (but I do not want to have free surface boundary condition). However, if using periodic displacement boundary condition, there would be no displacement at left and right boundary, and the elastic stress-strain would be larger than free surface case. I am thinking if there is periodic strain boundary condition, then the necking is retained and stress-strain would be between free surface and periodic displacement cases. So, I am wondering is it possible to have periodic strain boundary condition in sfepy, sounds kind of odd though.

Regards Ronghai

在 2016年3月15日星期二 UTC+1下午10:28:54，Robert Cimrman写道：

...

Hi Ronghai,

you can constrain the displacements using the periodic boundary

conditions

...

• then the strain on the boundaries would be the same as well. Or you need something else?

r.

On 03/15/2016 03:26 PM, Ronghai Wu wrote:

...

Hi sfepy developers and users,

I am wondering is it possible to have periodic strain boundary condition in sfepy? For example, in a 2D uniaxial tension along y direction, how to constrain the strain at left and right boundary to be same when solving mechanical equilibrium equation?

Regards Ronghai