Stress analysis

Section author: Florian Zill (Helmholtz Centre for Environmental Research GmbH - UFZ)

The following example from the ogs benchmark collection is used for the stress analysis:

<https://www.opengeosys.org/docs/benchmarks/thermo-mechanics/creepafterexcavation/>

from ogstools.meshplotlib import examples, plot, setup
from ogstools.meshplotlib.plot_features import plot_streamlines
from ogstools.propertylib import mesh_dependent, presets

setup.reset()
setup.length.output_unit = "km"
mesh = examples.mesh_mechanics
fig = plot(mesh, presets.displacement)

Tensor components

We can inspect the stress (or strain) tensor components by indexing.

fig = plot(mesh, presets.stress["xx"])
fig = plot(mesh, presets.stress["xy"])

• 
• 

Principal stresses

Let’s plot the the principal stress components and also overlay the direction of the corresponding eigenvector in the plot. Note: the eigenvalues are sorted by increasing order, i.e. eigenvalue[0] is the most negative / largest compressive principal stress.

eigvecs = presets.stress.eigenvectors
fig = plot(mesh, mesh_property=presets.stress.eigenvalues[0])
plot_streamlines(
    ax=fig.axes[0], mesh=mesh, mesh_property=eigvecs[0], plot_type="lines"
)

fig = plot(mesh, mesh_property=presets.stress.eigenvalues[1])
plot_streamlines(
    ax=fig.axes[0], mesh=mesh, mesh_property=eigvecs[1], plot_type="lines"
)

fig = plot(mesh, mesh_property=presets.stress.eigenvalues[2])
plot_streamlines(
    ax=fig.axes[0], mesh=mesh, mesh_property=eigvecs[2], plot_type="lines"
)

We can also plot the mean of the principal stress, i.e. the magnitude of the hydrostatic component of the stress tensor. see: ogstools.propertylib.tensor_math.mean()

fig = plot(mesh, presets.stress.mean)

Von Mises stress

see: ogstools.propertylib.tensor_math.von_mises()

fig = plot(mesh, presets.stress.von_Mises)

octahedral shear stress

see: ogstools.propertylib.tensor_math.octahedral_shear()

fig = plot(mesh, presets.stress.octahedral_shear)

Integrity criteria

Evaluating models regarding their integrity is often dependent on the geometry, e.g. for a hypothetical water column proportional to the depth. Presets which fall under this category make use of ogstools.propertylib.mesh_dependent.

The hypothetical water column used in the integrity criteria would initially use existing “pressure” data in the mesh, otherwise it is automatically calculated as the following:

mesh["pressure"] = mesh_dependent.p_fluid(mesh)
fig = plot(mesh, presets.pressure)

/builds/ogs/tools/ogstools/.venv-devcontainer/lib/python3.10/site-packages/pyvista/core/dataset.py:1975: UnitStrippedWarning: The unit of the quantity is stripped when downcasting to ndarray.
  scalars = np.asanyarray(scalars)

But since this assumes that the top of the model is equal to the ground surface, the resulting pressure is underestimated. In this case we have to correct the depth manually. Then the pressure is calculated correctly:

mesh["depth"] = mesh_dependent.depth(mesh, use_coords=True)
fig = plot(mesh, "depth")
mesh["pressure"] = mesh_dependent.p_fluid(mesh)
fig = plot(mesh, presets.pressure)

• 
• 

Dilantancy criterion

see: ogstools.propertylib.mesh_dependent.dilatancy_critescu()

fig = plot(mesh, presets.dilatancy_critescu_tot)
fig = plot(mesh, presets.dilatancy_critescu_eff)

• 
• 

Fluid pressure criterion

see: ogstools.propertylib.mesh_dependent.fluid_pressure_criterion()

fig = plot(mesh, presets.fluid_pressure_crit)