Negative G1

[tuanhie...@gmail.com]

Hello,

I conduct a simulation to calculate G with the presence of an initial stress. I obtain negative G1 for all the point on the crack front. How can I understand the results?

mpirun --allow-run-as-root -np 2 ./crack_propagation -my_file BM_A3.h5m -my_order 2 -mofem_mg_verbose 1 -mofem_mg_coarse_order 1 -mofem_mg_levels 2 -my_ref 0 -my_geom_order 1 -my_ref_order 3 -my_residual_stress_block 1 -my_mwls_approx_file BM_A3_Ini.h5m -my_internal_stress_name SIEFELGA -mwls_dm 1 -mwls_number_of_base_functions 4 -se_number_of_refinment_levels 1 2>&1 | tee log_BM_A3

You can find in the attached files the med file, rmed file for initial stress field and the log file.

Best regards,

Pham

[Lukasz Kaczmraczyk]

Hello,

I am unable to reproduce your results, you need to send me bc configuration file. 

The negative values of g1 cam happen in three cases, 

1)

The approximation is very poor, for that case is an easy fox, increase approximation order to -me_ref_order

2) Approximation of internal stress is poor, you can try to use this parameter:

-mwls_dm 2  # Use more points 

-mwls_number_of_base_functions 10 # Use then base functions (2nd order polynomial)

3)

 It is a strong mix mode.   It will mean that crack is strongly rotating while it is propagating.