Vector decomposition of ev_cauchy_strain

[Radko Bankras]

Hi,

The following code is used in example linear_elasticity/material_nonlinearity.py to evaluate the Cauchy strain during post processing:

strain = problem.evaluate('ev_cauchy_strain.i.Omega(u)', mode='el_avg')
out['cauchy_strain'] = Struct(name='output_data', mode='cell', data=strain, dofs=None)

Using ParaView to visualize the result on a tetrahedral mesh from VTK output, I find the strain to have 8 decomposition components with arbitrary names 1 through 8.
Is it possible to obtain the strain as a vector in Cartesian components?

Regards,
Radko

[Radko Bankras]

Correction: 9 decomposition components with arbitrary names 0 through 8.

[Robert Cimrman]

Hi Radko,

what do you mean by "as a vector"? The strain is a tensor (let's say e) with 9
components in 3D. The components you see in paraview are e_11, e_12, e_13,
e_21, e_22, e_23, e_31, e_32, e_33, starting from 0.

r.

> Regards,
> Radko
>

[Radko Bankras]

Hi Robert,

Ah ... that makes sense. I guess my confusion started by trying to grasp everything in sfepy at once and then looked at how ParaView visualizes tensor components on mesh cell surfaces. I have now added some line probes to my script, based on examples/linear_elasticity/linear_elastic_probes.py. That helps. I should play more with this and look for that Dummies book though.

Regards,
Radko