Tools & Helpers#

This page contains tools and helpers for constitutive material formulations.

View Force-Stretch Curves on Elementary Deformations

ViewMaterial(umat[, ux, ps, bx, statevars])

Create views on normal force per undeformed area vs stretch curves for the elementary homogeneous deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

ViewMaterialIncompressible(umat[, ux, ps, ...])

Create views on normal force per undeformed area vs stretch curves for the elementary homogeneous incompressible deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

Special Constitutive Materials

CompositeMaterial(material, other_material)

A composite material with two constitutive materials merged.

Volumetric([bulk, parallel, volumetric])

Neo-Hookean material formulation with deactivated shear modulus, i.e. the volumetric part of the strain energy density function \(U(J)\) only.

VolumetricStrainEnergy(bulk[, kind])

The volumetric part of the strain energy density function as a function of the volume ratio, \(U(J)\), with its first and second derivatives w.r.t.

LineChange([parallel])

Line Change.

AreaChange([parallel])

Area Change.

VolumeChange([parallel])

Volume Change.

Other

constitution.lame_converter(E, nu)

Convert the pair of given material parameters Young's modulus \(E\) and Poisson ratio \(\nu\) to first and second Lamé - constants \(\lambda\) and \(\mu\).

constitution.lame_converter_orthotropic(E, nu, G)

Convert elastic orthotropic material parameters to Lamé constants.

Detailed API Reference

class felupe.ViewMaterial(umat, ux=array([0.7, 0.75, 0.8, 0.85, 0.9, 0.95, 1., 1.05, 1.1, 1.15, 1.2, 1.25, 1.3, 1.35, 1.4, 1.45, 1.5, 1.55, 1.6, 1.65, 1.7, 1.75, 1.8, 1.85, 1.9, 1.95, 2., 2.05, 2.1, 2.15, 2.2, 2.25, 2.3, 2.35, 2.4, 2.45, 2.5]), ps=array([1., 1.05, 1.1, 1.15, 1.2, 1.25, 1.3, 1.35, 1.4, 1.45, 1.5, 1.55, 1.6, 1.65, 1.7, 1.75, 1.8, 1.85, 1.9, 1.95, 2., 2.05, 2.1, 2.15, 2.2, 2.25, 2.3, 2.35, 2.4, 2.45, 2.5]), bx=array([1., 1.05, 1.1, 1.15, 1.2, 1.25, 1.3, 1.35, 1.4, 1.45, 1.5, 1.55, 1.6, 1.65, 1.7, 1.75]), statevars=None)[source]#

Create views on normal force per undeformed area vs stretch curves for the elementary homogeneous deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

Parameters:
  • umat (class) – A class with methods for the gradient and hessian of the strain energy density function w.r.t. the deformation gradient. See Material for further details.

  • ux (ndarray, optional) – Array with stretches for uniaxial tension/compression. Default is linsteps([0.7, 2.5], num=36). The (absolute) times are required as second row for a time-dependent material, e.g. [stretches, times].

  • ps (ndarray, optional) – Array with stretches for planar shear. Default is linsteps([1.0, 2.5], num=30). The (absolute) times are required as second row for a time-dependent material, e.g. [stretches, times].

  • bx (ndarray, optional) – Array with stretches for equi-biaxial tension. Default is linsteps([1.0, 1.75], num=15)`. The (absolute) times are required as second row for a time-dependent material, e.g. [stretches, times].

  • statevars (ndarray or None, optional) – Array with state variables (default is None). If None, the state variables are assumed to be initially zero.

Notes

The load cases of a time-dependent material, i.e. a material with an attribute time_dependent which is True, include the (absolute) times of the stretches, e.g. ux=[stretches, times]. The time increments are the differences of the times, with the initial time zero (like a TimeStep), and they are passed to the material for each increment of a load case. Each load case starts at the initial time with the initial state variables. Holding times are given by repeated stretches.

Examples

>>> import felupe as fem
>>>
>>> umat = fem.OgdenRoxburgh(fem.NeoHooke(mu=1, bulk=2), r=3, m=1, beta=0)
>>> view = fem.ViewMaterial(
...     umat,
...     ux=fem.math.linsteps([1, 1.5, 1, 2, 1, 2.5, 1], num=15),
...     ps=None,
...     bx=None,
... )
>>> ax = view.plot(show_title=True, show_kwargs=True)
../../_images/tools-1.png

The load case of a time-dependent material includes the times, e.g. a loading with a holding time at the maximum stretch and an unloading with a lower stretch rate.

>>> umat = fem.MaterialStrain(
...     material=fem.linear_elastic_viscoelastic,
...     λ=2.0,
...     μ=1.0,
...     G=[3.0, 2.0],
...     τ=[1.0, 10.0],
...     statevars=((2, 3, 3),),
... )
>>> stretches = fem.math.linsteps([1, 1.1, 1.1, 1], num=[10, 10, 10])
>>> times = fem.math.linsteps([0, 1, 11, 31], num=[10, 10, 10])
>>> view = fem.ViewMaterial(umat, ux=[stretches, times], ps=None, bx=None)
>>> ax = view.plot(show_title=True, show_kwargs=True)
../../_images/tools-2.png
biaxial(stretches=None, **kwargs)[source]#

Normal force per undeformed area vs stretch curve for a equi-biaxial deformation.

Parameters:

stretches (ndarray or None, optional) – Array with stretches at which the forces are evaluated (default is None). If None, the stretches from initialization are used.

Returns:

3-tuple with array of stretches and array of forces and the label.

Return type:

tuple of ndarray

evaluate()#

Evaluate normal force per undeformed area vs stretch curves for the elementary homogeneous incompressible deformations uniaxial tension/compression, planar shear and biaxial tension. A load case is not included if its array of stretches (attribute ux, ps or bx) is None.

Returns:

List with 3-tuple of stretch and force arrays and the label string for each load case.

Return type:

list of 3-tuple

planar(stretches=None, **kwargs)[source]#

Normal force per undeformed area vs stretch curve for a planar shear deformation.

Parameters:

stretches (ndarray or None, optional) – Array with stretches at which the forces are evaluated (default is None). If None, the stretches from initialization are used.

Returns:

3-tuple with array of stretches and array of forces and the label.

Return type:

tuple of ndarray

plot(ax=None, show_title=True, show_kwargs=True, **kwargs)#

Plot normal force per undeformed area vs stretch curves for the elementary homogeneous incompressible deformations uniaxial tension/compression, planar shear and biaxial tension.

uniaxial(stretches=None, **kwargs)[source]#

Normal force per undeformed area vs. stretch curve for a uniaxial deformation.

Parameters:

stretches (ndarray or None, optional) – Array with stretches at which the forces are evaluated (default is None). If None, the stretches from initialization are used.

Returns:

3-tuple with array of stretches and array of forces and the label.

Return type:

tuple of ndarray

class felupe.ViewMaterialIncompressible(umat, ux=array([0.7, 0.75, 0.8, 0.85, 0.9, 0.95, 1., 1.05, 1.1, 1.15, 1.2, 1.25, 1.3, 1.35, 1.4, 1.45, 1.5, 1.55, 1.6, 1.65, 1.7, 1.75, 1.8, 1.85, 1.9, 1.95, 2., 2.05, 2.1, 2.15, 2.2, 2.25, 2.3, 2.35, 2.4, 2.45, 2.5]), ps=array([1., 1.05, 1.1, 1.15, 1.2, 1.25, 1.3, 1.35, 1.4, 1.45, 1.5, 1.55, 1.6, 1.65, 1.7, 1.75, 1.8, 1.85, 1.9, 1.95, 2., 2.05, 2.1, 2.15, 2.2, 2.25, 2.3, 2.35, 2.4, 2.45, 2.5]), bx=array([1., 1.05, 1.1, 1.15, 1.2, 1.25, 1.3, 1.35, 1.4, 1.45, 1.5, 1.55, 1.6, 1.65, 1.7, 1.75]), statevars=None)[source]#

Create views on normal force per undeformed area vs stretch curves for the elementary homogeneous incompressible deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

Parameters:
  • umat (class) – A class with methods for the gradient and hessian of the strain energy density function w.r.t. the deformation gradient. See Material for further details.

  • ux (ndarray, optional) – Array with stretches for incompressible uniaxial tension/compression. Default is linsteps([0.7, 2.5], num=36). The (absolute) times are required as second row for a time-dependent material, e.g. [stretches, times].

  • ps (ndarray, optional) – Array with stretches for incompressible planar shear. Default is linsteps([1, 2.5], num=30). The (absolute) times are required as second row for a time-dependent material, e.g. [stretches, times].

  • bx (ndarray, optional) – Array with stretches for incompressible equi-biaxial tension. Default is linsteps([1, 1.75], num=15). The (absolute) times are required as second row for a time-dependent material, e.g. [stretches, times].

  • statevars (ndarray or None, optional) – Array with state variables (default is None). If None, the state variables are assumed to be initially zero.

Notes

The load cases of a time-dependent material, i.e. a material with an attribute time_dependent which is True, include the (absolute) times of the stretches, e.g. ux=[stretches, times]. The time increments are the differences of the times, with the initial time zero (like a TimeStep), and they are passed to the material for each increment of a load case. Each load case starts at the initial time with the initial state variables. Holding times are given by repeated stretches.

Examples

>>> import felupe as fem
>>>
>>> umat = fem.Hyperelastic(fem.extended_tube, Gc=0.2, Ge=0.2, beta=0.2, delta=0.1)
>>> preview = fem.ViewMaterialIncompressible(umat)
>>> ax = preview.plot(show_title=True, show_kwargs=True)
../../_images/tools-3.png
>>> umat = fem.OgdenRoxburgh(fem.NeoHooke(mu=1), r=3, m=1, beta=0)
>>> view = fem.ViewMaterialIncompressible(
...     umat,
...     ux=fem.math.linsteps([1, 1.5, 1, 2, 1, 2.5, 1], num=15),
...     ps=None,
...     bx=None,
... )
>>> ax = view.plot(show_title=True, show_kwargs=True)
../../_images/tools-4.png

The load case of a time-dependent material includes the times, e.g. cycles with increasing maximum stretches at a constant stretch rate.

>>> umat = fem.Hyperelastic(
...     fem.finite_strain_viscoelastic, mu=1.0, eta=1.0, nstatevars=6
... )
>>> stretches = fem.math.linsteps([1, 1.5, 1, 2, 1], num=[5, 5, 10, 10])
>>> times = fem.math.linsteps([0, 0.5, 1, 2, 3], num=[5, 5, 10, 10])
>>> view = fem.ViewMaterialIncompressible(
...     umat, ux=[stretches, times], ps=None, bx=None
... )
>>> ax = view.plot(show_title=True, show_kwargs=True)
../../_images/tools-5.png
biaxial(stretches=None)[source]#

Normal force per undeformed area vs stretch curve for a equi-biaxial incompressible deformation.

Parameters:

stretches (ndarray or None, optional) – Array with stretches at which the forces are evaluated (default is None). If None, the stretches from initialization are used.

Returns:

3-tuple with array of stretches and array of forces and the label.

Return type:

tuple of ndarray

evaluate()#

Evaluate normal force per undeformed area vs stretch curves for the elementary homogeneous incompressible deformations uniaxial tension/compression, planar shear and biaxial tension. A load case is not included if its array of stretches (attribute ux, ps or bx) is None.

Returns:

List with 3-tuple of stretch and force arrays and the label string for each load case.

Return type:

list of 3-tuple

planar(stretches=None)[source]#

Normal force per undeformed area vs stretch curve for a planar shear incompressible deformation.

Parameters:

stretches (ndarray or None, optional) – Array with stretches at which the forces are evaluated (default is None). If None, the stretches from initialization are used.

Returns:

3-tuple with array of stretches and array of forces and the label.

Return type:

tuple of ndarray

plot(ax=None, show_title=True, show_kwargs=True, **kwargs)#

Plot normal force per undeformed area vs stretch curves for the elementary homogeneous incompressible deformations uniaxial tension/compression, planar shear and biaxial tension.

uniaxial(stretches=None)[source]#

Normal force per undeformed area vs stretch curve for a uniaxial incompressible deformation.

Parameters:

stretches (ndarray or None, optional) – Array with stretches at which the forces are evaluated (default is None). If None, the stretches from initialization are used.

Returns:

3-tuple with array of stretches and array of forces and the label.

Return type:

tuple of ndarray

class felupe.CompositeMaterial(material, other_material)[source]#

A composite material with two constitutive materials merged. Each material keeps its own state variables.

Parameters:

Notes

The state variables of the composite material are the flattened state variables of the first material, followed by the flattened state variables of the second material, i.e. the size of the state variables is the sum of the sizes of the state variables of both materials. The state variables are split for the evaluation of the materials and the updated state variables of both materials are concatenated again. Hence, the order of the materials doesn’t matter, e.g. nh & viscoelastic and viscoelastic & nh are equal, and several materials with state variables are supported, also nested composite materials like a & b & c. If the state variables are not given (None) or if the input list contains only one array, the input list is passed to both materials and the state variables of the first material are returned.

The gradients and the hessians are evaluated in chunks of cells, by a pool of threads if any of the materials is parallel.

The composite material is time-dependent if any of the materials is time-dependent (attribute time_dependent). The time increment dtime is only passed to the time-dependent materials, e.g. viscoelastic & fem.Volumetric(bulk=5000).

Warning

Do not merge two constitutive materials with the same keys of material parameters. In this case, the values of these material parameters are taken from the first constitutive material.

Examples

>>> import felupe as fem
>>>
>>> nh = fem.NeoHooke(mu=1.0)
>>> vol = fem.Volumetric(bulk=2.0)
>>> umat = nh & vol
>>> ax = umat.plot()
../../_images/tools-6.png
chunksize = 8192#
copy()#

Return a deep-copy of the constitutive material.

gradient(x, *args, out=None, **kwargs)#

Return the evaluated gradient(s) of the strain energy density function, evaluated by _gradient(x, *args, **kwargs) in chunks of cells.

Parameters:
  • x (list of ndarray) – The list with input arguments. These contain the extracted fields of a FieldContainer along with the old vector of state variables, [*field.extract(), statevars_old].

  • *args (tuple, optional) – Optional arguments for _gradient().

  • out (ndarray or None, optional) – A location into which the first gradient is stored (default is None).

  • **kwargs (dict, optional) – Optional keyword-arguments for _gradient(), e.g. the time increment dtime of a time-dependent material.

Returns:

A list with the evaluated gradient(s) of the strain energy density function and the updated vector of state variables.

Return type:

list of ndarray

Raises:
  • ValueError – If the material is time-dependent and the time increment dtime is missing (or None).

  • TypeError – If the material is not time-dependent and a time increment dtime is given.

hessian(x, *args, out=None, **kwargs)#

Return the evaluated upper-triangle components of the hessian(s) of the strain energy density function, evaluated by _hessian(x, *args, **kwargs) in chunks of cells.

Parameters:
  • x (list of ndarray) – The list with input arguments. These contain the extracted fields of a FieldContainer along with the old vector of state variables, [*field.extract(), statevars_old].

  • *args (tuple, optional) – Optional arguments for _hessian().

  • out (ndarray or None, optional) – A location into which the first hessian is stored (default is None).

  • **kwargs (dict, optional) – Optional keyword-arguments for _hessian(), e.g. the time increment dtime of a time-dependent material.

Returns:

A list with the evaluated upper-triangle components of the hessian(s) of the strain energy density function.

Return type:

list of ndarray

Raises:
  • ValueError – If the material is time-dependent and the time increment dtime is missing (or None).

  • TypeError – If the material is not time-dependent and a time increment dtime is given.

is_stable(x, hessian=None)#

Return a boolean mask for stability of isotropic material model formulations.

At a given deformation gradient, a normal force is applied on each principal stretch direction. If the resulting incremental stretches are positive, the material model formulation is considered to be stable at the given deformation gradient.

Parameters:
  • x (list of ndarray) – The list with input arguments. These contain the extracted fields of a FieldContainer.

  • hessian (ndarray or None, optional) – Second partial derivative of the strain energy density function w.r.t. the deformation gradient. Default is None.

Returns:

Boolean mask of stability.

Return type:

ndarray

Notes

Warning

This stability check will lead to a singular matrix for isotropic (hyperelastic) material model formulations without a volumetric part.

Examples

First, let’s check the stability of the Neo-Hookean material model formulation. The stability is evaluated on (valid) principal stretches of a biaxial deformation. All deformations are stable.

>>> import numpy as np
>>> import felupe as fem
>>>
>>> umat = fem.NeoHooke(mu=1.0, bulk=2.0)
>>> view = umat.view()
>>> λ = view.biaxial()[0]
>>>
>>> F = np.zeros((3, 3, 1, λ[0].size))
>>> for a in range(3):
...     F[a, a] = λ[a]
>>>
>>> umat.is_stable([F])
array([[ True,  True,  True,  True,  True,  True,  True,  True,  True,
         True,  True,  True,  True,  True,  True,  True]])

Now, let’s check the stability of the Mooney-Rivlin material model formulation. The stability is evaluated on (valid) principal stretches of a biaxial deformation. Biaxial deformations are only stable up to a longitudinal stretch of 1.35.

>>> import numpy as np
>>> import felupe as fem
>>> import felupe.constitution.tensortrax as mat
>>>
>>> umat = fem.Hyperelastic(
...     mat.models.hyperelastic.mooney_rivlin,
...     C10=0.25,
...     C01=0.25,
... ) & fem.Volumetric(bulk=5000)
>>> view = umat.view()
>>> λ = view.biaxial()[0]
>>>
>>> F = np.zeros((3, 3, 1, λ[0].size))
>>> for a in range(3):
...     F[a, a] = λ[a]
>>>
>>> umat.is_stable([F])
array([[ True,  True,  True,  True,  True,  True,  True,  True, False,
        False, False, False, False, False, False, False]])
optimize(ux=None, ps=None, bx=None, incompressible=False, relative=False, **kwargs)#

Optimize the material parameters by a least-squares fit on experimental stretch-stress data.

Parameters:
  • ux (array of shape (2, ...) or (3, ...) or None, optional) – Experimental uniaxial stretch and force-per-undeformed-area data (default is None). The (absolute) times are required as third row for a time-dependent material.

  • ps (array of shape (2, ...) or (3, ...) or None, optional) – Experimental planar-shear stretch and force-per-undeformed-area data (default is None). The (absolute) times are required as third row for a time-dependent material.

  • bx (array of shape (2, ...) or (3, ...) or None, optional) – Experimental biaxial stretch and force-per-undeformed-area data (default is None). The (absolute) times are required as third row for a time-dependent material.

  • incompressible (bool, optional) – A flag to enforce incompressible deformations (default is False).

  • relative (bool, optional) – A flag to optimize relative instead of absolute residuals, i.e. (predicted - observed) / observed instead of predicted - observed (default is False).

  • **kwargs (dict, optional) – Optional keyword arguments are passed to scipy.optimize.least_squares().

Returns:

  • ConstitutiveMaterial – A copy of the constitutive material with the optimized material parameters.

  • scipy.optimize.OptimizeResult – Represents the optimization result.

Notes

Warning

At least one load case, i.e. one of the arguments ux, ps or bx must not be None.

Note

For JAX-based materials, double-precision is required to optimize material parameters.

The material parameters are the keyword-arguments of the material. They are set on a copy of the material, i.e. its keyword-arguments and its attributes with the same names, as well as the material parameters of its materials, e.g. of a composite material. The time increment is not a material parameter, i.e. a constant time increment dtime of the keyword-arguments is not optimized. The experimental data of a time-dependent material includes the (absolute) times of the data points, e.g. ux=[stretches, forces, times], which are taken for the time increments of the load case, see ViewMaterial.

import jax

jax.config.update("jax_enable_x64", True)

The vector of residuals is given in Eq. (3) in case of absolute residuals

(1)#\[ \begin{align}\begin{aligned}\begin{split}\boldsymbol{r} &= \begin{bmatrix} \boldsymbol{r}_\text{ux} \\ \boldsymbol{r}_\text{ps} \\ \boldsymbol{r}_\text{bx} \end{bmatrix}\end{split}\\r_\text{ux}(\lambda_i) &= P_\text{ux}(\lambda_i) - P_\text{ux, observed}(\lambda_i)\\r_\text{ps}(\lambda_i) &= P_\text{ps}(\lambda_i) - P_\text{ps, observed}(\lambda_i)\\r_\text{bx}(\lambda_i) &= P_\text{bx}(\lambda_i) - P_\text{bx, observed}(\lambda_i)\end{aligned}\end{align} \]

and in Eq. (4) in case of relative residuals.

(2)#\[ \begin{align}\begin{aligned}\begin{split}\boldsymbol{r} &= \begin{bmatrix} \boldsymbol{r}_\text{ux} \\ \boldsymbol{r}_\text{ps} \\ \boldsymbol{r}_\text{bx} \end{bmatrix}\end{split}\\r_\text{ux}(\lambda_i) &= \frac{ P_\text{ux}(\lambda_i) - P_\text{ux, observed}(\lambda_i)}{ P_\text{ux, observed}(\lambda_i) }\\r_\text{ps}(\lambda_i) &= \frac{ P_\text{ps}(\lambda_i) - P_\text{ps, observed}(\lambda_i)}{ P_\text{ps, observed}(\lambda_i) }\\r_\text{bx}(\lambda_i) &= \frac{ P_\text{bx}(\lambda_i) - P_\text{bx, observed}(\lambda_i)}{ P_\text{bx, observed}(\lambda_i) }\end{aligned}\end{align} \]

Examples

The Anssari-Benam Bucchi material model formulation is best-fitted on Treloar’s uniaxial and biaxial tension data [1]_.

>>> import numpy as np
>>> import felupe as fem
>>>
>>> λ, P = np.array(
...     [
...         [1.000, 0.00],
...         [1.240, 2.30],
...         [1.585, 4.16],
...         [2.180, 6.00],
...         [3.020, 8.80],
...         [4.030, 12.5],
...         [4.760, 16.2],
...         [5.750, 23.6],
...         [6.850, 38.5],
...         [7.250, 49.6],
...         [7.600, 64.4],
...     ]
... ).T * np.array([[1.0], [0.0980665]])
>>>
>>> umat = fem.Hyperelastic(fem.anssari_benam_bucchi)
>>> umat_new, res = umat.optimize(
...     ux=[λ, P], incompressible=True, relative=True
... )
>>>
>>> ux = np.linspace(λ.min(), λ.max(), num=50)
>>> ax = umat_new.plot(incompressible=True, ux=ux, bx=None, ps=None)
>>> lines = ax.plot(λ, P, "C0x")
../../_images/tools-9.png

See also

scipy.optimize.least_squares

Solve a nonlinear least-squares problem with bounds on the variables.

References

parallel = False#
plot(incompressible=False, **kwargs)#

Return a plot with normal force per undeformed area vs. stretch curves for the elementary homogeneous deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

Parameters:
Return type:

matplotlib.axes.Axes

See also

felupe.ViewMaterial

Create views on normal force per undeformed area vs. stretch curves for the elementary homogeneous deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

felupe.ViewMaterialIncompressible

Create views on normal force per undeformed area vs. stretch curves for the elementary homogeneous incompressible deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

screenshot(filename='umat.png', incompressible=False, **kwargs)#

Save a screenshot with normal force per undeformed area vs. stretch curves for the elementary homogeneous deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

Parameters:
  • filename (str, optional) – The filename of the screenshot (default is “umat.png”).

  • incompressible (bool, optional) – A flag to enforce views on incompressible deformations (default is False).

  • **kwargs (dict, optional) – Optional keyword-arguments for ViewMaterial or ViewMaterialIncompressible.

Return type:

matplotlib.axes.Axes

See also

felupe.ViewMaterial

Create views on normal force per undeformed area vs. stretch curves for the elementary homogeneous deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

felupe.ViewMaterialIncompressible

Create views on normal force per undeformed area vs. stretch curves for the elementary homogeneous incompressible deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

property time_dependent#

A flag which is True if any of the materials is time-dependent. Then, the time increment dtime is passed to the time-dependent materials.

view(incompressible=False, **kwargs)#

Create views on normal force per undeformed area vs. stretch curves for the elementary homogeneous deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

Parameters:
  • incompressible (bool, optional) – A flag to enforce views on incompressible deformations (default is False).

  • **kwargs (dict, optional) – Optional keyword-arguments for ViewMaterial or ViewMaterialIncompressible, e.g. the load cases. The load cases of a time-dependent material include the (absolute) times, e.g. ux=[stretches, times].

Return type:

felupe.ViewMaterial or felupe.ViewMaterialIncompressible

See also

felupe.ViewMaterial

Create views on normal force per undeformed area vs. stretch curves for the elementary homogeneous deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

felupe.ViewMaterialIncompressible

Create views on normal force per undeformed area vs. stretch curves for the elementary homogeneous incompressible deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

class felupe.Volumetric(bulk=None, parallel=False, volumetric='volume-ratio')[source]#

Neo-Hookean material formulation with deactivated shear modulus, i.e. the volumetric part of the strain energy density function \(U(J)\) only.

Parameters:
  • bulk (float or None, optional) – Bulk modulus of the volumetric part of the strain energy density function. It is required if volumetric is a kind and it must be None if volumetric is an instance, which holds its own parameters (default is None).

  • parallel (bool, optional) – A flag to evaluate the gradient and the hessian in chunks of cells by a pool of threads (default is False).

  • volumetric (str or VolumetricStrainEnergy, optional) – The volumetric part of the strain energy density function \(U(J)\), given by a kind with the bulk modulus, "volume-ratio", "stretch" or "log", or by an instance of VolumetricStrainEnergy, e.g. of a custom subclass. Default is "volume-ratio".

Examples

The volumetric part is combined with a material formulation for the distortional part of the strain energy density function.

>>> import felupe as fem
>>>
>>> umat = fem.NeoHooke(mu=1.0) & fem.Volumetric(bulk=5000, volumetric="stretch")

See also

felupe.NeoHooke

Nearly-incompressible isotropic hyperelastic Neo-Hookean material formulation.

felupe.VolumetricStrainEnergy

The volumetric part of the strain energy density function as a function of the volume ratio.

chunksize = 8192#
copy()#

Return a deep-copy of the constitutive material.

function(x)#

Strain energy density function per unit undeformed volume of the Neo-Hookean material formulation.

Parameters:

x (list of ndarray) – List with the Deformation gradient F (3x3) as first item

gradient(x, *args, out=None, **kwargs)#

Return the evaluated gradient(s) of the strain energy density function, evaluated by _gradient(x, *args, **kwargs) in chunks of cells.

Parameters:
  • x (list of ndarray) – The list with input arguments. These contain the extracted fields of a FieldContainer along with the old vector of state variables, [*field.extract(), statevars_old].

  • *args (tuple, optional) – Optional arguments for _gradient().

  • out (ndarray or None, optional) – A location into which the first gradient is stored (default is None).

  • **kwargs (dict, optional) – Optional keyword-arguments for _gradient(), e.g. the time increment dtime of a time-dependent material.

Returns:

A list with the evaluated gradient(s) of the strain energy density function and the updated vector of state variables.

Return type:

list of ndarray

Raises:
  • ValueError – If the material is time-dependent and the time increment dtime is missing (or None).

  • TypeError – If the material is not time-dependent and a time increment dtime is given.

hessian(x, *args, out=None, **kwargs)#

Return the evaluated upper-triangle components of the hessian(s) of the strain energy density function, evaluated by _hessian(x, *args, **kwargs) in chunks of cells.

Parameters:
  • x (list of ndarray) – The list with input arguments. These contain the extracted fields of a FieldContainer along with the old vector of state variables, [*field.extract(), statevars_old].

  • *args (tuple, optional) – Optional arguments for _hessian().

  • out (ndarray or None, optional) – A location into which the first hessian is stored (default is None).

  • **kwargs (dict, optional) – Optional keyword-arguments for _hessian(), e.g. the time increment dtime of a time-dependent material.

Returns:

A list with the evaluated upper-triangle components of the hessian(s) of the strain energy density function.

Return type:

list of ndarray

Raises:
  • ValueError – If the material is time-dependent and the time increment dtime is missing (or None).

  • TypeError – If the material is not time-dependent and a time increment dtime is given.

is_stable(x, hessian=None)#

Return a boolean mask for stability of isotropic material model formulations.

At a given deformation gradient, a normal force is applied on each principal stretch direction. If the resulting incremental stretches are positive, the material model formulation is considered to be stable at the given deformation gradient.

Parameters:
  • x (list of ndarray) – The list with input arguments. These contain the extracted fields of a FieldContainer.

  • hessian (ndarray or None, optional) – Second partial derivative of the strain energy density function w.r.t. the deformation gradient. Default is None.

Returns:

Boolean mask of stability.

Return type:

ndarray

Notes

Warning

This stability check will lead to a singular matrix for isotropic (hyperelastic) material model formulations without a volumetric part.

Examples

First, let’s check the stability of the Neo-Hookean material model formulation. The stability is evaluated on (valid) principal stretches of a biaxial deformation. All deformations are stable.

>>> import numpy as np
>>> import felupe as fem
>>>
>>> umat = fem.NeoHooke(mu=1.0, bulk=2.0)
>>> view = umat.view()
>>> λ = view.biaxial()[0]
>>>
>>> F = np.zeros((3, 3, 1, λ[0].size))
>>> for a in range(3):
...     F[a, a] = λ[a]
>>>
>>> umat.is_stable([F])
array([[ True,  True,  True,  True,  True,  True,  True,  True,  True,
         True,  True,  True,  True,  True,  True,  True]])

Now, let’s check the stability of the Mooney-Rivlin material model formulation. The stability is evaluated on (valid) principal stretches of a biaxial deformation. Biaxial deformations are only stable up to a longitudinal stretch of 1.35.

>>> import numpy as np
>>> import felupe as fem
>>> import felupe.constitution.tensortrax as mat
>>>
>>> umat = fem.Hyperelastic(
...     mat.models.hyperelastic.mooney_rivlin,
...     C10=0.25,
...     C01=0.25,
... ) & fem.Volumetric(bulk=5000)
>>> view = umat.view()
>>> λ = view.biaxial()[0]
>>>
>>> F = np.zeros((3, 3, 1, λ[0].size))
>>> for a in range(3):
...     F[a, a] = λ[a]
>>>
>>> umat.is_stable([F])
array([[ True,  True,  True,  True,  True,  True,  True,  True, False,
        False, False, False, False, False, False, False]])
optimize(ux=None, ps=None, bx=None, incompressible=False, relative=False, **kwargs)#

Optimize the material parameters by a least-squares fit on experimental stretch-stress data.

Parameters:
  • ux (array of shape (2, ...) or (3, ...) or None, optional) – Experimental uniaxial stretch and force-per-undeformed-area data (default is None). The (absolute) times are required as third row for a time-dependent material.

  • ps (array of shape (2, ...) or (3, ...) or None, optional) – Experimental planar-shear stretch and force-per-undeformed-area data (default is None). The (absolute) times are required as third row for a time-dependent material.

  • bx (array of shape (2, ...) or (3, ...) or None, optional) – Experimental biaxial stretch and force-per-undeformed-area data (default is None). The (absolute) times are required as third row for a time-dependent material.

  • incompressible (bool, optional) – A flag to enforce incompressible deformations (default is False).

  • relative (bool, optional) – A flag to optimize relative instead of absolute residuals, i.e. (predicted - observed) / observed instead of predicted - observed (default is False).

  • **kwargs (dict, optional) – Optional keyword arguments are passed to scipy.optimize.least_squares().

Returns:

  • ConstitutiveMaterial – A copy of the constitutive material with the optimized material parameters.

  • scipy.optimize.OptimizeResult – Represents the optimization result.

Notes

Warning

At least one load case, i.e. one of the arguments ux, ps or bx must not be None.

Note

For JAX-based materials, double-precision is required to optimize material parameters.

The material parameters are the keyword-arguments of the material. They are set on a copy of the material, i.e. its keyword-arguments and its attributes with the same names, as well as the material parameters of its materials, e.g. of a composite material. The time increment is not a material parameter, i.e. a constant time increment dtime of the keyword-arguments is not optimized. The experimental data of a time-dependent material includes the (absolute) times of the data points, e.g. ux=[stretches, forces, times], which are taken for the time increments of the load case, see ViewMaterial.

import jax

jax.config.update("jax_enable_x64", True)

The vector of residuals is given in Eq. (3) in case of absolute residuals

(3)#\[ \begin{align}\begin{aligned}\begin{split}\boldsymbol{r} &= \begin{bmatrix} \boldsymbol{r}_\text{ux} \\ \boldsymbol{r}_\text{ps} \\ \boldsymbol{r}_\text{bx} \end{bmatrix}\end{split}\\r_\text{ux}(\lambda_i) &= P_\text{ux}(\lambda_i) - P_\text{ux, observed}(\lambda_i)\\r_\text{ps}(\lambda_i) &= P_\text{ps}(\lambda_i) - P_\text{ps, observed}(\lambda_i)\\r_\text{bx}(\lambda_i) &= P_\text{bx}(\lambda_i) - P_\text{bx, observed}(\lambda_i)\end{aligned}\end{align} \]

and in Eq. (4) in case of relative residuals.

(4)#\[ \begin{align}\begin{aligned}\begin{split}\boldsymbol{r} &= \begin{bmatrix} \boldsymbol{r}_\text{ux} \\ \boldsymbol{r}_\text{ps} \\ \boldsymbol{r}_\text{bx} \end{bmatrix}\end{split}\\r_\text{ux}(\lambda_i) &= \frac{ P_\text{ux}(\lambda_i) - P_\text{ux, observed}(\lambda_i)}{ P_\text{ux, observed}(\lambda_i) }\\r_\text{ps}(\lambda_i) &= \frac{ P_\text{ps}(\lambda_i) - P_\text{ps, observed}(\lambda_i)}{ P_\text{ps, observed}(\lambda_i) }\\r_\text{bx}(\lambda_i) &= \frac{ P_\text{bx}(\lambda_i) - P_\text{bx, observed}(\lambda_i)}{ P_\text{bx, observed}(\lambda_i) }\end{aligned}\end{align} \]

Examples

The Anssari-Benam Bucchi material model formulation is best-fitted on Treloar’s uniaxial and biaxial tension data [1]_.

>>> import numpy as np
>>> import felupe as fem
>>>
>>> λ, P = np.array(
...     [
...         [1.000, 0.00],
...         [1.240, 2.30],
...         [1.585, 4.16],
...         [2.180, 6.00],
...         [3.020, 8.80],
...         [4.030, 12.5],
...         [4.760, 16.2],
...         [5.750, 23.6],
...         [6.850, 38.5],
...         [7.250, 49.6],
...         [7.600, 64.4],
...     ]
... ).T * np.array([[1.0], [0.0980665]])
>>>
>>> umat = fem.Hyperelastic(fem.anssari_benam_bucchi)
>>> umat_new, res = umat.optimize(
...     ux=[λ, P], incompressible=True, relative=True
... )
>>>
>>> ux = np.linspace(λ.min(), λ.max(), num=50)
>>> ax = umat_new.plot(incompressible=True, ux=ux, bx=None, ps=None)
>>> lines = ax.plot(λ, P, "C0x")
../../_images/tools-12.png

See also

scipy.optimize.least_squares

Solve a nonlinear least-squares problem with bounds on the variables.

References

parallel = False#
plot(incompressible=False, **kwargs)#

Return a plot with normal force per undeformed area vs. stretch curves for the elementary homogeneous deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

Parameters:
Return type:

matplotlib.axes.Axes

See also

felupe.ViewMaterial

Create views on normal force per undeformed area vs. stretch curves for the elementary homogeneous deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

felupe.ViewMaterialIncompressible

Create views on normal force per undeformed area vs. stretch curves for the elementary homogeneous incompressible deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

screenshot(filename='umat.png', incompressible=False, **kwargs)#

Save a screenshot with normal force per undeformed area vs. stretch curves for the elementary homogeneous deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

Parameters:
  • filename (str, optional) – The filename of the screenshot (default is “umat.png”).

  • incompressible (bool, optional) – A flag to enforce views on incompressible deformations (default is False).

  • **kwargs (dict, optional) – Optional keyword-arguments for ViewMaterial or ViewMaterialIncompressible.

Return type:

matplotlib.axes.Axes

See also

felupe.ViewMaterial

Create views on normal force per undeformed area vs. stretch curves for the elementary homogeneous deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

felupe.ViewMaterialIncompressible

Create views on normal force per undeformed area vs. stretch curves for the elementary homogeneous incompressible deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

time_dependent = False#
view(incompressible=False, **kwargs)#

Create views on normal force per undeformed area vs. stretch curves for the elementary homogeneous deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

Parameters:
  • incompressible (bool, optional) – A flag to enforce views on incompressible deformations (default is False).

  • **kwargs (dict, optional) – Optional keyword-arguments for ViewMaterial or ViewMaterialIncompressible, e.g. the load cases. The load cases of a time-dependent material include the (absolute) times, e.g. ux=[stretches, times].

Return type:

felupe.ViewMaterial or felupe.ViewMaterialIncompressible

See also

felupe.ViewMaterial

Create views on normal force per undeformed area vs. stretch curves for the elementary homogeneous deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

felupe.ViewMaterialIncompressible

Create views on normal force per undeformed area vs. stretch curves for the elementary homogeneous incompressible deformations uniaxial tension/compression, planar shear and biaxial tension of a given isotropic material formulation.

property volumetric#

The volumetric part of the strain energy density function, see VolumetricStrainEnergy. For a kind, it is created with the current bulk modulus and it is None without a bulk modulus.

class felupe.VolumetricStrainEnergy(bulk, kind='volume-ratio')[source]#

The volumetric part of the strain energy density function as a function of the volume ratio, \(U(J)\), with its first and second derivatives w.r.t. the volume ratio.

Parameters:
  • bulk (float) – The bulk modulus \(K\).

  • kind (str, optional) – The kind of the volumetric part of the strain energy density function: "volume-ratio" for \(U(J)=K(J-1)^2/2\), "stretch" for \(U(J)=9K(J^{1/3}-1)^2/2\) or "log" for \(U(J)=K\ln(J)^2/2\), see Eq. (5). Default is "volume-ratio".

Notes

Unlike a ConstitutiveMaterial, which is evaluated by the deformation gradient, it is evaluated by the volume ratio \(J=\det(\boldsymbol{F})\). An instance U is callable, i.e. U(J) returns \(U(J)\), U.dUdJ(J) returns \(U'(J)\) and U.d2UdJdJ(J) returns \(U''(J)\). The derivatives w.r.t. the deformation gradient are obtained by the chain rule with the derivatives of the volume ratio, see VolumeChange.

The kinds are quadratic in the Seth-Hill strain \(E_m = (\lambda^m - 1) / m\) of order \(m\) of the volumetric stretch \(\lambda = J^{1/3}\), see Eq. (5), with \(m=3\) for "volume-ratio", \(m=1\) for "stretch" and the limit \(m \rightarrow 0\) for "log".

(5)#\[U(J) = \frac{9K}{2} E_m^2 = \frac{9K}{2 m^2} \left( J^{m/3} - 1 \right)^2\]

For all kinds, the second derivative at the undeformed configuration is the bulk modulus, \(U''(1)=K\). The volumetric part and its first and second derivatives are given in Eq. (6) for kind="volume-ratio" (default),

(6)#\[ \begin{align}\begin{aligned}U(J) &= \frac{K}{2} \left( J - 1 \right)^2\\U'(J) &= K \left( J - 1 \right)\\U''(J) &= K\end{aligned}\end{align} \]

in Eq. (7) for kind="stretch"

(7)#\[ \begin{align}\begin{aligned}U(J) &= \frac{9K}{2} \left( J^{1/3} - 1 \right)^2\\U'(J) &= 3K\ \frac{J^{1/3} - 1}{J^{2/3}}\\U''(J) &= K\ \frac{2 - J^{1/3}}{J^{5/3}}\end{aligned}\end{align} \]

and in Eq. (8) for kind="log".

(8)#\[ \begin{align}\begin{aligned}U(J) &= \frac{K}{2} \ln \left( J \right)^2\\U'(J) &= K\ \frac{\ln \left( J \right)}{J}\\U''(J) &= K\ \frac{1 - \ln \left( J \right)}{J^2}\end{aligned}\end{align} \]

The kinds are evaluated by these expressions of the volume ratio, i.e. not by the general form, which is singular for \(m=0\). The cubic root of the kind "stretch" is evaluated by numpy.cbrt(), which is real-valued also for negative (linearized) volume ratios. The smaller the order \(m\), the stiffer is the volumetric response in compression and the softer in tension. The first derivative \(U'(J)\), i.e. the hydrostatic part of the Cauchy stress, is unbounded for \(J \rightarrow 0\) for the kinds "stretch" and "log". In tension, it is bounded by \(3K/4\) at \(J=8\) and by \(K/e\) at \(J=e\) for these kinds, where their second derivatives change their sign.

A custom volumetric part of the strain energy density function is a subclass which implements the methods __call__(J), dUdJ(J) and d2UdJdJ(J). The undeformed configuration is free of stress only if \(U'(1)=0\), and a positive second derivative \(U''(J)>0\) is required for a stable volumetric response.

The volumetric part of the strain energy density function is passed by the argument volumetric to NeoHooke, Volumetric, NearlyIncompressible and SolidBodyNearlyIncompressible, either as a kind together with the bulk modulus or as an instance (then, without the bulk modulus).

Examples

>>> import numpy as np
>>> import felupe as fem
>>>
>>> U = fem.VolumetricStrainEnergy(bulk=5000, kind="stretch")
>>> J = np.array([0.9, 1.0, 1.1])
>>>
>>> U(J)
array([26.79710794,  0.        , 23.44513171])
>>> U.dUdJ(J)
array([-555.32721366,    0.        ,  454.39256332])
>>> U.d2UdJdJ(J)
array([6165.48776188, 5000.        , 4127.92559565])

A custom volumetric part of the strain energy density function is defined by a subclass.

>>> class MyVolumetricStrainEnergy(fem.VolumetricStrainEnergy):
...     def __init__(self, bulk):
...         self.bulk = bulk
...
...     def __call__(self, J):
...         return self.bulk / 4 * (J**2 - 1 - 2 * np.log(J))
...
...     def dUdJ(self, J):
...         return self.bulk / 2 * (J - 1 / J)
...
...     def d2UdJdJ(self, J):
...         return self.bulk / 2 * (1 + 1 / J**2)
>>>
>>> U = MyVolumetricStrainEnergy(bulk=5000)
>>> U.d2UdJdJ(1.0)
5000.0

The kinds or an instance are passed to the materials and the solid bodies.

>>> umat = fem.NeoHooke(mu=1, bulk=5000, volumetric="stretch")
>>> umat = fem.NeoHooke(mu=1, volumetric=U)
>>>
>>> mesh = fem.Cube(n=3)
>>> field = fem.FieldContainer([fem.Field(fem.RegionHexahedron(mesh), dim=3)])
>>> umat = fem.NeoHooke(mu=1)
>>> solid = fem.SolidBodyNearlyIncompressible(umat, field, volumetric=U)

See also

felupe.VolumeChange

Volume Change.

felupe.NeoHooke

Nearly-incompressible isotropic hyperelastic Neo-Hookean material formulation.

felupe.Volumetric

Neo-Hookean material formulation with deactivated shear modulus.

felupe.NearlyIncompressible

A nearly-incompressible material formulation to augment the distortional part of the strain energy function by a volumetric part and a constraint equation.

felupe.SolidBodyNearlyIncompressible

A (nearly) incompressible solid body with methods for the assembly of sparse vectors/matrices.

__call__(J)[source]#

Return the volumetric part of the strain energy density function \(U(J)\).

Parameters:

J (float or ndarray) – The volume ratio.

Returns:

The volumetric part of the strain energy density function.

Return type:

float or ndarray

d2UdJdJ(J)[source]#

Return the second derivative of the volumetric part of the strain energy density function w.r.t. the volume ratio \(U''(J)\).

Parameters:

J (float or ndarray) – The volume ratio.

Returns:

The second derivative w.r.t. the volume ratio. For kind="volume-ratio", it is the bulk modulus (a scalar).

Return type:

float or ndarray

dUdJ(J)[source]#

Return the first derivative of the volumetric part of the strain energy density function w.r.t. the volume ratio \(U'(J)\).

Parameters:

J (float or ndarray) – The volume ratio.

Returns:

The first derivative w.r.t. the volume ratio.

Return type:

float or ndarray

kinds = ('volume-ratio', 'stretch', 'log')#
class felupe.LineChange(parallel=False)[source]#

Line Change.

\[d\boldsymbol{x} = \boldsymbol{F} d\boldsymbol{X}\]

Gradient:

\[\frac{\partial \boldsymbol{F}}{\partial \boldsymbol{F}} = \boldsymbol{I} \overset{ik}{\otimes} \boldsymbol{I}\]
function(extract)[source]#

Line change.

Parameters:

extract (list of ndarray) – List of extracted field values with Deformation gradient as first item.

Returns:

F – Deformation gradient

Return type:

ndarray

gradient(extract, parallel=None)[source]#

Gradient of line change.

Parameters:

extract (list of ndarray) – List of extracted field values with Deformation gradient as first item.

Returns:

Gradient of line change

Return type:

ndarray

class felupe.AreaChange(parallel=False)[source]#

Area Change.

\[d\boldsymbol{a} = J \boldsymbol{F}^{-T} d\boldsymbol{A}\]

Gradient:

\[\frac{\partial J \boldsymbol{F}^{-T}}{\partial \boldsymbol{F}} = J \left( \boldsymbol{F}^{-T} \otimes \boldsymbol{F}^{-T} - \boldsymbol{F}^{-T} \overset{il}{\otimes} \boldsymbol{F}^{-T} \right)\]
function(extract, N=None, parallel=None)[source]#

Area change.

Parameters:
  • extract (list of ndarray) – List of extracted field values with Deformation gradient as first item.

  • N (ndarray or None, optional) – Area normal vector (default is None)

Returns:

Cofactor matrix of the deformation gradient

Return type:

ndarray

gradient(extract, N=None, parallel=None)[source]#

Gradient of area change.

Parameters:
  • extract (list of ndarray) – List of extracted field values with Deformation gradient as first item.

  • N (ndarray or None, optional) – Area normal vector (default is None)

Returns:

Gradient of cofactor matrix of the deformation gradient

Return type:

ndarray

class felupe.VolumeChange(parallel=False)[source]#

Volume Change.

\[d\boldsymbol{v} = \text{det}(\boldsymbol{F}) d\boldsymbol{V}\]

Gradient and hessian (equivalent to gradient of area change):

\[ \begin{align}\begin{aligned}\frac{\partial J}{\partial \boldsymbol{F}} &= J \boldsymbol{F}^{-T}\\\frac{\partial^2 J}{\partial \boldsymbol{F}\ \partial \boldsymbol{F}} &= J \left( \boldsymbol{F}^{-T} \otimes \boldsymbol{F}^{-T} - \boldsymbol{F}^{-T} \overset{il}{\otimes} \boldsymbol{F}^{-T} \right)\end{aligned}\end{align} \]
function(extract)[source]#

Gradient of volume change.

Parameters:

extract (list of ndarray) – List of extracted field values with Deformation gradient as first item.

Returns:

J – Determinant of the deformation gradient

Return type:

ndarray

gradient(extract)[source]#

Gradient of volume change.

Parameters:

F (ndarray) – Deformation gradient

Returns:

Gradient of the determinant of the deformation gradient

Return type:

ndarray

hessian(extract, parallel=None)[source]#

Hessian of volume change.

Parameters:

extract (list of ndarray) – List of extracted field values with Deformation gradient as first item.

Returns:

Hessian of the determinant of the deformation gradient

Return type:

ndarray

felupe.constitution.lame_converter(E, nu)[source]#

Convert the pair of given material parameters Young’s modulus \(E\) and Poisson ratio \(\nu\) to first and second Lamé - constants \(\lambda\) and \(\mu\).

Parameters:
  • E (float) – Young’s modulus.

  • nu (float) – Poisson ratio.

Returns:

  • lmbda (float) – First Lamé - constant.

  • mu (float) – Second Lamé - constant (shear modulus).

Notes

\[ \begin{align}\begin{aligned}\lambda &= \frac{E \nu}{(1 + \nu) (1 - 2 \nu)}\\\mu &= \frac{E}{2 (1 + \nu)}\end{aligned}\end{align} \]
felupe.constitution.lame_converter_orthotropic(E, nu, G)[source]#

Convert elastic orthotropic material parameters to Lamé constants.

Parameters:
  • E (list of float) – List of the three elastic moduli \(E_1, E_2, E_3\).

  • nu (list of float) – List of three poisson ratios \(\nu_{12}, \nu_{23}, \nu_{31}\).

  • G (list of float) – List of three shear moduli \(G_{12}, G_{23}, G_{31}\).

Returns:

  • lmbda (list of float) – List of six (upper triangle) first Lamé parameters \(\lambda_{11}, \lambda_{12}, \lambda_{13}, \lambda_{22}, \lambda_{23}, \lambda_{33}\).

  • mu (list of float) – List of the three second Lamé parameters \(\mu_1,\mu_2, \mu_3\).

Notes

The orthotropic material parameters are converted to orthotropic Lamé constants.

The compliance matrix as the inverse of the stiffness matrix with the parameters \(E_i\), \(\nu_{ij}\) and \(G_{ij}\) is given in Eq. (10).

(9)#\[\begin{split}\boldsymbol{C}^{-1} = \begin{bmatrix} \frac{1}{E_1} & -\frac{\nu_{21}}{E_2} & -\frac{\nu_{31}}{E_3} & 0 & 0 & 0 \\ -\frac{\nu_{12}}{E_1} & \frac{1}{E_2} & -\frac{\nu_{32}}{E_3} & 0 & 0 & 0 \\ -\frac{\nu_{13}}{E_1} & -\frac{\nu_{23}}{E_2} & \frac{1}{E_3} & 0 & 0 & 0 \\ 0 & 0 & 0 & \frac{1}{G_{12}} & 0 & 0 \\ 0 & 0 & 0 & 0 & \frac{1}{G_{23}} & 0 \\ 0 & 0 & 0 & 0 & 0 & \frac{1}{G_{31}} \end{bmatrix}\end{split}\]

The stiffness matrix with the Lamé constants is denoted in Eq. ortho-matrix.

(10)#\[\begin{split}\boldsymbol{C} = \begin{bmatrix} \lambda_{11} + 2 \mu_1 & \lambda_{12} & \lambda_{13} & 0 & 0 & 0 \\ \lambda_{11} & \lambda_{12} + 2 \mu_2 & \lambda_{13} & 0 & 0 & 0 \\ \lambda_{11} & \lambda_{12} & \lambda_{13} + 2 \mu_3 & 0 & 0 & 0 \\ 0 & 0 & 0 & \frac{\mu_1 + \mu_2}{2} & 0 & 0 \\ 0 & 0 & 0 & 0 & \frac{\mu_2 + \mu_3}{2} & 0 \\ 0 & 0 & 0 & 0 & 0 & \frac{\mu_3 + \mu_1}{2} \end{bmatrix}\end{split}\]

Eq. (10) is evaluated and inverted numerically to extract the Lamé constants.

See also

felupe.LinearElasticOrthotropic

Orthotropic linear-elastic material formulation.

felupe.constitution.tensortrax.models.hyperelastic.saint_venant_kirchhoff_orthotropic

Strain energy function of the orthotropic hyperelastic Saint-Venant Kirchhoff material formulation.