Can DAMASK Simulate Texture Evolution Under Stress

[CGH20171006]

Hello, everyone!

I’ve been working with DAMASK and trying to simulate the behavior of polycrystalline materials under tensile stress. I’ve run a few basic cases, but I am unable to see any grain orientation-related information in the hdf5 results. Specifically, I want to track the evolution of the grain texture during deformation.

So I think that DAMASK might not account for grain orientation transitions in its default model.

Is this correct? or do I need to add something to the model or define additional parameters to track the texture evolution? Any guidance would be greatly appreciated!

[eisenlohr]

The relevant output quantity you need to include in the HDF5 file is called "O", which is the lattice orientation in quaternion representation. Please add to your material.yaml file in the phase: NameOfPhase: mechanical: output list an entry for O.

[CGH20171006]

The relevant output quantity you need to include in the HDF5 file is called "O", which is the lattice orientation in quaternion representation. Please add to your material.yaml file in the phase: NameOfPhase: mechanical: output list an entry for O.

OK, with your answer, I have solved this problem. But I didn't get a significant crystal rotation phenomena when I ran the DAMASK code. I guess the low strain in my code causes this. However, the damask will post this error when I try to get more strain.

 Piola--Kirchhoff stress       / MPa =
        0.3356       -45.9545      -234.4250
      -47.2996         0.3001      -165.9533
     -622.0288      -424.0315    177533.0421

 ... doing gamma convolution ...............................................

 ... reporting .............................................................

 error divergence =         0.96 (8.49E+07 / m, tol =  8.88E+07)
 error stress BC  =         0.00 (3.36E+05 Pa,  tol =  1.78E+08)

 ===========================================================================

 increment 67513 converged

 ###########################################################################
 Time  6.75140E+02s: Increment 67514/100000-1/1 of load case 1/1
 Increment 67514/100000-1/1 @ Iteration 1≤0≤100

 deformation gradient aim       =
   0.6279416    0.0000000    0.0000000
   0.0000000    0.6586008    0.0000000
   0.0000000    0.0000000    1.6751400

 ... evaluating constitutive response ......................................

 ┌─────────────────────────────────────────────────────────────────────┐
 │                        error 402                                    │
 ├─────────────────────────────────────────────────────────────────────┤
 │ invalid orientation specified                                       │
 │ fromMatrix                                                          │
 │                                                                     │
 └─────────────────────────────────────────────────────────────────────┘

 DAMASK terminated on:
 Date:               03/12/2024
 Time:               04:29:51
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_UNDERFLOW_FLAG IEEE_DENORMAL IEEE_INEXACT_FLAG
STOP 1

Load.yaml is here:

solver: {mechanical: spectral_basic}

loadstep:
  - boundary_conditions:
      mechanical:
        dot_F:
          - [x,   0.0,  0.0]
          - [0.0,   x,  0.0]
          - [0.0,   0.0,  0.001]
        P:
          - [0.0,   x,  x]
          - [x,   0.0,  x]
          - [x,   x,  x]
    discretization: {t: 1000.0, N: 100000}
    f_out: 1000

material.yaml is here:

homogenization:
  direct:
    N_constituents: 1
    mechanical: {type: pass}
phase:
  '1':
    lattice: hP
    c/a: 1.587
    mechanical:
      output: [F, P, F_e, F_p, L_p, O]
      elastic: {C_11: 162400000000.0, C_33: 180700000000.0, C_44: 49700000000.0, C_12: 92000000000.0, C_13: 69000000000.0, type: Hooke}
      type: phenopowerlaw
      references: ['C. Zambaldi et al., Journal of Materials Research 27(1):356-367, 2021, https://doi.org/10.1557/jmr.2011.334', 'L. Wang et al., Acta Materialia 132:598-610, 2017, https://doi.org/10.1016/j.actamat.2017.05.015']
      N_sl: [3, 3, 0, 12]
      dot_gamma_0_sl: [0.001, 0.001, 0.0, 0.001]
      n_sl: [20, 20, 1.0, 20]
      a_sl: [2.0, 2.0, 1.0, 2.0]
      xi_0_sl: [349000000.0, 150000000.0, 0.0, 1107000000.0]
      xi_inf_sl: [568000000.0, 1500000000.0, 0.0, 3420000000.0]
      h_0_sl-sl: [200000000.0, 200000000.0, 0.0, 200000000.0]
      h_sl-sl: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, -1.0, -1.0, -1.0, -1.0, -1.0, -1.0, -1.0, -1.0, -1.0, -1.0, -1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]

at the last increment of this job (67514), the "u" is 8.3e-6. Did this index mean stain or displacement? But I this what ever is it, maybe this value is too low for the crystal rotation.

Thanks for your assistant!!!

[eisenlohr]

A few comments may be helpful.

• The total amount of strain (extension to F_zz = 2) is probably too much and will result in excessively distorted cells that will not allow the solution to converge.
• The discretization (N) is quite fine and can be reduced by one or even two orders of magnitude. Usually, a strain increment on the order of 1e-3 should converge without much issue in a purely mechanical problem.
• I am surprised by the strength levels (xi) that are listed in the material.yaml file. All those are on the order of GPa... Are you certain that this is intended?
• Please also make sure that the -1 entries in the h_sl-sl matrix correspond to interactions with the (switched off) third slip system. Otherwise, the associated softening due to the negative sign could lead to instabilities and non-convergence.