Output in logfile

[jguilie]

Hello,

First of all, let me thank you for all the great work. Your code is very easy to read and contain a lot of original materials.

Actually, I'm working with the EDF uncoupled type and I know that i can record in logfile the cauchy stress tensor for each element and each time step which is great.

But, I would like more. I would like to record the cauchy stress tensor separately in each direction of integration on the sphere for each element and each time step. I know that this information is obviously in the source code during the numerical integration but i can't find a way to add a new record in logfile.

I don't mind if I have to do a little coding for that but please tell me how to do it.


Best regards,

Joachim Guilie

[jguilie]

I think I found the beginning of a solution in the case of an internal variable (damage model). I've added some lines in DataStore.h and DataStore.cpp and I succeeded in printing (in logfile) the damage for each element and each converged time step. Am I near the solution for cauchy on the sphere ?

Best,

[ateshian]

Hi Joachim,

Saving the stress in the fibers for each fiber direction within FEBio can be a challenge. The stress is calculated inside a loop, which you can find in FESphericalFiberDistribution.cpp (or FEEllipsoidalFiberDistribution.cpp, or the corresponding routines for an uncoupled strain energy formulation), in the routine mat3ds FESphericalFiberDistribution::Stress(FEMaterialPoi nt& mp). The fiber directions represent fiber bundles with associated cross-sectional areas on the unit sphere (representing the fractional fiber content for that direction). The stresses are added together within the loop to provide the overall stress tensor contributed by all the fibers. There is no simple way to extract and save the stress contribution for each direction.

However, it is possible to reproduce this calculation in a custom post-processing routine: From FEBio, you can save the nine components of the deformation gradient to the log file, for every element and every time step. Then, in a standalone post-processing routine, you use this deformation gradient to calculate the fiber strain for each fiber direction, then the stress based on the chosen constitutive relation. In other words, your reproduce the code in FESphericalFiberDistribution::Stress(FEMaterialPoi nt& mp) and you add a couple of lines to save the stress for each fiber direction.

Best,

Gerard