Questions encountered in Adaptively Refining Mesh

[Weidong Pian]

Hello,Deal.ii mates
I’m studying works contributed to deal.II conmunity recently,that is,tutorial step-44 and “Quasi-Static Finite-Strain Quasi-incompressible Visco-elasticity”.When I try to extend your program of the latter with automatic refinement,I encountered an interesting trouble,stucking me for two days.I found that if I transfer all the three component of solutions from coarse mesh to refining mesh directly,then the interpolated solution U seems correct,but the interpolated solution p and interpolated solution J seems totally chaotic(as pictures show).This directly results in a NAN in RES_U,RES_P and RES_J.I can’t understand how this chaos happens,so I would sincerely appreciate your help and opinions.
The figure on the top shows solution befor adaptive refinement and solutiontransfer,that on the bottom shows solution after refinement and solutiontransfer.I would really appreciate it for any suggestions

Best regards,

WeiDong Pian

[Wolfgang Bangerth]

On 11/7/25 20:17, 'Weidong Pian' via deal.II User Group wrote:
>
> I’m studying works contributed to deal.II conmunity recently,that is,tutorial
> step-44 and “Quasi-Static Finite-Strain Quasi-incompressible Visco-
> elasticity”.When I try to extend your program of the latter with automatic

> refinement,I encountered an interesting trouble,stucking me for two days.*I

> found that if I transfer all the three component of solutions from coarse mesh
> to refining mesh directly,then the interpolated solution U seems correct,but
> the interpolated solution p and interpolated solution J seems totally
> chaotic(as pictures show).This directly results in a NAN in RES_U,RES_P and

> RES_J.*I can’t understand how this chaos happens,so I would sincerely

> appreciate your help and opinions.
> The figure on the top shows solution befor adaptive refinement and
> solutiontransfer,that on the bottom shows solution after refinement and
> solutiontransfer.I would really appreciate it for any suggestions

Wei-Dong,
I think you need to show us how you transfer the solution. Are you calling one
of the DoFRenumbering functions *after* you interpolate the solution onto the
mesh with SolutionTransfer?

Best
W.

--
------------------------------------------------------------------------
Wolfgang Bangerth email: bang...@colostate.edu
www: http://www.math.colostate.edu/~bangerth/