Re: [Getfem-users] Theory behind mdbrick_Coulomb_friction.set_augmented_matrix

[Konstantinos Poulios]

Hi Yves,

thanks for the response. After your hint that there are two different
types of augmentation I understand both coulomb_num.pdf and
2007_HR_contact.pdf better and my previous question can be considered
as answered. On the question if I really need the Hansbo augmentation
I can't give any answer yet. In my case I have contact between two
elastic bodies with non-conforming meshes. For the contact condition I
search for contact nodes on both sides (interchanging master and
slave) so I get a number of multiplier variables which is more "rich"
than the grid on each separate contact surface. I am not sure if in
this case the inf-sup condition still holds. Anyway I am going to
experiment with this in the next days.

Regarding your remark about the new brick system, I would definitely
prefer to work with it instead of the old one. After your mail I
focused my intents on converting my model to the new system. I have
tried to create a very simplified (frictionless) version of the
Coulomb_friction brick for the new brick system. As soon as I make it
fundamentally work, I will publish it here for further discussion and
development. For the moment, I miss any example of a non-linear
problem in the new brick system. Are there such examples that I 've
overlooked?

For the future I would also be very interested in the development of
contact problems in combination with large displacements/deformations.

Regards

Kostas