Unresolved Issues While Exporting DAMASK Multi-Phase Simulation Results to VTK

[Anmol-Bakshi]

I’m performing a 2-phase DAMASK simulation (Iron and Ti) on a 3D-RVE using DAMASK_grid v3.0.1 and encountered multiple export-related issues during post-processing with the python-damask API.
Outputs like sigma, F, P, epsilon_U^0.0(F), epsilon_V^0.0(F) were correctly computed using:
r.add_strain(F='F', t='U', m=0.0)
r.add_strain(F='F', t='V', m=0.0)
r.add_stress_Cauchy()

r.export_VTK(output='*', mode='cell', target_dir='output_vtk')
ERROR: ValueError: shape mismatch: value array of shape (202612,3) could not be broadcast to indexing result of shape (202612,1)

Difficulty in Adding Phase-Specific Fields Manually
Attempted to export custom fields like:

• Orientation component: Iron/O_q0
• Slip system resistance: Ti/xi_sl_0
• Slip shear: Iron/gamma_sl_0
  But: r.phase['Iron']['mechanical']['O_q0'] = r.phase['Iron']['mechanical']['O'][:, 0].reshape(-1, 1) and other variants (r.result[...], r.data[...], r['...']), all resulted in:
• AttributeError: 'Result' object has no attribute 'result'
• TypeError: 'Result' object is not subscriptable
• IndexError: only integers, slices (...) are valid indices

Even dir(r) doesn’t show any direct method to assign new fields into the internal structure in a safe/documented way.

How to add orientations, gamma_sl, xi_sl correctly to export_VTK(...)?

[eisenlohr]

When you say "2-phase simulation", are you referring to a grid that has a single component at each grid point that is either iron or titanium, or are you simulating a situation where each grid point contains both phases (in certain proportion)?

[Anmol-Bakshi]

It’s the first case. I’m using fullfield with N_constituents: 1, so each grid cell contains a single phase only. The microstructure is biphasic at the cell level: a cell is either Iron or Ti, not a mixture.

Concretely: my .vti has a cell-data array named material with values 0/1; these map to two entries in material: in the YAML:

material[0] → Iron (v: 1.0)

material[1] → Ti (v: 1.0)

There are no multi-constituent cells (no local volume fractions other than 1.0).

Also, my material.yaml file looks like this:
homogenization:
fullfield:
N_constituents: 1
mechanical:
type: pass
output: ['F', 'P', 'sigma', 'material']
phase:
Iron:
references: ['https://en.wikipedia.org/wiki/Iron']
lattice: cI
rho: 7874.0
mechanical:
output : ['O', 'xi_sl', 'gamma_sl']
elastic:
type: Hooke
references: ['W. F. Hosford. The mechanics of crystals and textured polycrystals. Oxford University Press, New York-Oxford 1993. https://doi.org/10.1002/crat.2170290414']
C_11: 233500000000.0
C_12: 135500000000.0
C_44: 118000000000.0
plastic:
type: phenopowerlaw
output: ['xi_sl', 'gamma_sl']
N_sl: [1]
dot_gamma_0_sl: [0.0017]
n_sl: [6]
a_sl: [2.3] # Strain rate sensitivity
xi_0_sl: [121.5e6] # Changes yield stress
xi_inf_sl: [366.e6] # Controls strain hardening
h_0_sl-sl: [3.8e+8] # Modifies work-hardening rate
h_sl-sl: [1.0] # Slightly reduced interaction
damage:
type: none
Ti:
references: ['https://www.totalmateria.com/page.aspx?ID=CheckArticle&site=ktn&NM=221']
lattice: hP
c/a: 1.640
rho: 4506.0
mechanical:
output : ['O', 'xi_sl', 'gamma_sl']
elastic:
type: Hooke
references: ['W. F. Hosford. The mechanics of crystals and textured polycrystals. Oxford University Press, New York-Oxford 1993. https://doi.org/10.1002/crat.2170290414']
C_11: 228100000000.0
#C_11,T: -59.15
#C_11,T^2: 0.01156
C_12: 140360000000.0
#C_12,T: 21.92
#C_12,T^2: -0.01621
C_44: 39190000000.0
#C_44,T: -18.57
#C_44,T^2: -0.003745
C_33: 294280000000.0
#C_33,T: -41.10
#C_33,T^2: 0.00733
C_13: 107330000000.0
#C_13,T: 2.965
#C_13,T^2: -0.005767
#T_ref: 293.15
plastic:
type: phenopowerlaw
output: ['xi_sl', 'gamma_sl']
N_sl: [3, 0, 0, 0] # basal, prism, -, 1. pyr<c+a>
dot_gamma_0_sl: [0.0017, 0.0, 0.0, 0.0]
n_sl: [20, 0, 0, 0]
a_sl: [2.0, 0.0, 0.0, 0.0]
# C. Zambaldi et al.:
xi_0_sl: [349.e6, 0.0, 0.0, 0.0]
xi_inf_sl: [568.e6, 0.0, 0.0, 0.0]
# L. Wang et al. :
# xi_0_sl: [127.e+6, 96.e+6, 0.0, 240.e+6]
h_0_sl-sl: [200.e6, 0.0, 0.0, 0.0]
h_sl-sl: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
damage:
type: none
material:

• constituents:
  • phase: Iron
    O: [0.820456968582669, -0.339844403730458, 0.175932915225707, -0.424739630005737]
    v: 1.0
    homogenization: fullfield
• constituents:
  • phase: Ti
    O: [1.0, 0.0, 0.0, 0.0]
    v: 1.0
    homogenization: fullfield

[eetorres]

Did you try r.export_VTK() without parameters?

[zy-worker]

Did you solve this? I meet the same problrm

[dmentock]

@zy-worker can you provide your setup files? (The python script to generate the input files and the input files themselves)

[zy-worker]

Thank you for your answer. I want to simulate HCP and FCC dual phases, but the number of slip systems is different - HCP has 18 and FCC has 12. How can I achieve separate output for each phase as well as combined output for both phases?

homogenization:
homo:
N_constituents: 1
mechanical: {type: pass}
damage:
type: pass
output:

• T
• phi
• f_phi
  phase:
  A:
  lattice: hP
  c/a: 1.6333
  mechanical:
  output: [F, P, F_p, L_p, O]
  elastic: {type: Hooke, C_11: 106e+9, C_12: 60e+9, C_44: 28e+9, C_13: 28e+9, C_33: 28e+9}
  plastic:
  type: dislotwin
  output:
• rho_dip
• rho_mob
• gamma_sl
• Lambda_sl
• tau_pass
  D: 1.525e-05
  tau_0: [55e+6, 55e+6, 55e+6, 55e+6]
  N_sl: [3, 3, 0, 6]
  b_sl: [2.546e-10, 2.546e-10, 2.546e-10, 2.546e-10]
  rho_dip_0: [1, 1, 1, 1]
  rho_mob_0: [1e12, 1e12, 1e12, 1e12]
  v_0: [10e-5, 10e-5, 10e-5, 10e-5]
  Q_sl: [8.361e-20, 8.361e-20, 8.361e-20, 8.361e-20]
  Q_cl: 1.876e-19
  p_sl: [0.8, 0.8, 0.8, 0.8]
  q_sl: [1.5, 1.5, 1.5, 1.5]
  i_sl: [45, 45, 45, 45]
  V_at: 6.6e-29
  D_a: 7
  D_0: 0.000623
  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]
  damage:
  type: anisobrittle
  output: [Psi_D, f_phi, r_W]
  N_cl: [3]
  g_crit: [0.5e+7]
  s_crit: [0.006666]
  dot_o_0: 1.e-3
  p: 20
  l_c: 4.0e-6
  mu: 0.01
  N_sl: [12]
  G_0: 10.0
  onoff_damage: true
  onoff_plastic: true
  gamma_plastic_lock: 0.2
  B:
  lattice: cF
  mechanical:
  output: [F, P, F_p, L_p, O, xi_sl, gamma_sl]
  elastic: {type: Hooke, C_11: 106e+9, C_12: 60e+9, C_44: 28e+9}
  plastic:
  type: dislotwin
  output: [rho_dip, rho_mob]
  D: 1.525e-05
  tau_0: [55e6]
  N_sl: [12]
  b_sl: [2.546e-10]
  rho_dip_0: [1]
  rho_mob_0: [1e12]
  v_0: [10e-5]
  Q_sl: [8.361e-20]
  Q_cl: 1.876e-19
  p_sl: [0.8]
  q_sl: [1.5]
  i_sl: [45]
  V_at: 6.6e-29
  D_a: 7
  D_0: 0.000623
  h_sl-sl: [1, 1, 5.123, 0.574, 1.123, 1.123,1 ]
  damage:
  type: anisobrittle
  output: [Psi_D, f_phi, r_W]
  N_cl: [3]
  g_crit: [0.5e+7]
  s_crit: [0.006666]
  dot_o_0: 1.e-3
  p: 20
  l_c: 4.0e-6
  mu: 0.01
  N_sl: [12]
  G_0: 10.0
  onoff_damage: true
  onoff_plastic: true
  gamma_plastic_lock: 0.2

import damask
res = damask.Result('fn')
res.add_stress_Cauchy()
res.add_strain()
res.add_equivalent_Mises('sigma')
res.add_equivalent_Mises('epsilon_V^0.0(F)')
res.export_VTK()

[eisenlohr]

Hi @zy-worker and @Anmol-Bakshi, before answering the question, I would like to suggest that, for future reference, YAML snippets should be fenced within

```YAML
put the snippet here
```

so that they are readable and easy to copy and paste.

The solution to your original issue would be to select a view that includes only one of the two phases:

result.view(phases='A').export_VTK()
result.view(phases='B').export_VTK()

If you want, you can ask for specific output, such as export_VTK(output='xi_sl').

---

As a side remark, I believe that a few of your above-mentioned configuration keys are problematic:

• onoff_damage and onoff_plastic: I am not sure that DAMASK is honoring these keywords
• B > mechanical > output: xi_sl and gamma_sl are both constitutive quantities, not continuum mechanical ones... They are invalid as outputs at that level but need to go within plastic > output.