- Generated by
1.12.0
|
NEML2 2.0.0
|
Ctrl+F or Cmd+F to search the entire page.The following symbols are used throughout the documentation to denote different components of function definition.
Refer to System Documentation for detailed explanation about this system.
Calculate the advective stress, \( p_s \), taking the form of \( p_s = \frac{c}{3J} P_{ij}F_{ij} \). Here, \( J, P, F \) are the the deformation gradient Jacobian, the 1st Piola-Kirchhoff stress and the defomration gradient. \( c \) is the volume change coefficient.
advective_stress 🇴 The average advective stresscoefficient 🇵 Coefficient cdeformation_gradient 🇮 The deformation gradientjit Use JIT compilation for the forward operatorjs 🇮 The Jacobian of the deformation gradient associated with the swelling and phase changejt 🇮 The Jacobian of the deformation gradient associated with the thermal and volume expansionpk1_stress 🇮 1st Piola-Kirchhoff stressproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Define the variable as a function of temperature according to the Arrhenius law \( p = p_0 \exp \left( -\frac{Q}{RT} \right) \), where \( p_0 \) is the reference value, \( Q \) is the activation energy, \( R \) is the ideal gas constant, and \( T \) is the temperature.
activation_energy 🇵 Activation energyideal_gas_constant The ideal gas constantjit Use JIT compilation for the forward operatorparameter 🇴 The output parameterproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.reference_value 🇵 Reference valuetemperature 🇮 TemperatureDetailed documentation link
Map the flow rate (i.e., the consistency parameter in the KKT conditions) to the rate of internal variables. This object calculates the rate of equivalent plastic strain following associative flow rule, i.e. \( \dot{\bar{\varepsilon}}_p = - \dot{\gamma} \frac{\partial f}{\partial k} \), where \( \dot{\bar{\varepsilon}}_p \) is the equivalent plastic strain, \( \dot{\gamma} \) is the flow rate, \( f \) is the yield function, and \( k \) is the isotropic hardening.
equivalent_plastic_strain_rate 🇴 Rate of equivalent plastic strainflow_rate 🇮 Flow rateisotropic_hardening_direction 🇮 Direction of associative isotropic hardening which can be calculated using Normality.jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
The plastic flow direction assuming an associative J2 flow.
flow_direction Flow directionjit Use JIT compilation for the forward operatormandel_stress Mandel stressproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Map the flow rate (i.e., the consistency parameter in the KKT conditions) to the rate of internal variables. This object calculates the rate of kinematic plastic strain following associative flow rule, i.e. \( \dot{\boldsymbol{K}}_p = - \dot{\gamma} \frac{\partial f}{\partial \boldsymbol{X}} \), where \( \dot{\boldsymbol{K}}_p \) is the kinematic plastic strain, \( \dot{\gamma} \) is the flow rate, \( f \) is the yield function, and \( \boldsymbol{X} \) is the kinematic hardening.
flow_rate 🇮 Flow ratejit Use JIT compilation for the forward operatorkinematic_hardening_direction 🇮 Direction of associative kinematic hardening which can be calculated using Normality.kinematic_plastic_strain_rate 🇴 Rate of kinematic plastic strainproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Map the flow rate (i.e., the consistency parameter in the KKT conditions) to the rate of internal variables. This object calculates the rate of plastic strain following associative flow rule, i.e. \( \dot{\boldsymbol{\varepsilon}}_p = - \dot{\gamma} \frac{\partial f}{\partial \boldsymbol{M}} \), where \( \dot{\boldsymbol{\varepsilon}}_p \) is the plastic strain, \( \dot{\gamma} \) is the flow rate, \( f \) is the yield function, and \( \boldsymbol{M} \) is the Mandel stress.
flow_direction 🇮 Flow direction which can be calculated using Normalityflow_rate 🇮 Flow ratejit Use JIT compilation for the forward operatorplastic_strain_rate 🇴 Rate of plastic strainproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Avrami-Erofeev nucleation model, takes the form of \( f = k(1-a)(-ln(1-a))^n \), where \( k \) is the reaction coefficient, \( n \) is the reaction order, and \( a \) is the degree of conversion
conversion_degree 🇮 Degree of conversionjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.reaction_coef 🇵 Reaction coefficientreaction_order 🇵 Reaction orderreaction_rate 🇴 Reaction rateDetailed documentation link
Relate the porous flow capillary pressure to the effective saturation using the Brooks Corey correlation taking the form of \( P_c = P_t S_e^{-\frac{1}{p}} \). Here \( S_e \) is the effective saturation, \( P_t \) is the threshold pressure at zero saturation, and \( p \) is the shape parameter
capillary_pressure 🇴 Capillary pressure.effective_saturation 🇮 The effective saturationexponent 🇵 The shape parameter pjit Use JIT compilation for the forward operatorlog_extension Whether to apply logarithmic extensionproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.threshold_pressure 🇵 The threshold entry pressuretransition_saturation The transistion value of the effective saturation below which to apply the logarithmic extensionDetailed documentation link
Map the flow rate (i.e., the consistency parameter in the KKT conditions) to the rate of internal variables. This object defines the non-associative Fredrick-Armstrong kinematic hardening. In the model, back stress is directly treated as an internal variable. Rate of back stress is given as \( \dot{\boldsymbol{X}} = \left( \frac{2}{3} C \frac{\partial f}{\partial \boldsymbol{M}} - g \boldsymbol{X} \right) \dot{\gamma} \). \( \frac{\partial f}{\partial \boldsymbol{M}} \) is the flow direction, \( \dot{\gamma} \) is the flow rate, and \( C \) and \( g \) are material parameters. The complete Chaboche model adds static recovery terms \( - A \lVert \boldsymbol{X} \rVert^{a - 1} \boldsymbol{X} \), so the model includes kinematic hardening, dynamic recovery, and static recovery. \( A \) and \( a \) are additional material parameters.
A 🇵 Static recovery prefactorC 🇵 Kinematic hardening coefficienta 🇵 Static recovery exponentback_stress 🇮 Back stressback_stress_rate 🇴 Back stress rate, defaults to back_stress + _rateflow_direction 🇮 Flow directionflow_rate 🇮 Flow rateg 🇵 Dynamic recovery coefficientjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Compose multiple models together to form a single model. The composed model can then be treated as a new model and composed with others. The system documentation provides in-depth explanation on how the models are composed together.
additional_outputs Extra output variables to be extracted from the composed model in addition to the ones identified through dependency resolution.automatic_nonlinear_parameter Whether to automatically add dependent nonlinear parametersautomatic_scaling Whether to perform automatic scaling. See neml2::NonlinearSystem::init_scaling for implementation details.automatic_scaling_miter Maximum number of automatic scaling iterations. No error is produced upon reaching the maximum number of scaling iterations, and the scaling matrices obtained at the last iteration are used to scale the nonlinear system.automatic_scaling_tol Tolerance used in iteratively updating the scaling matrices.jit Use JIT compilation for the forward operatormodels Models being composed togetherpriority Priorities of models in decreasing order. A model with higher priority will be evaluated first. This is useful for breaking cyclic dependency.production Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
The contracting geometry model, often encountered in non-isothermal decomposition or solid-gas reactions, takes the form of \( f = k(1-a)^n \), where \( k \) is the reaction coefficient (often temperature-dependent), \( n \) is the reaction order, and \( a \) is the degree of conversion.
conversion_degree 🇮 Degree of conversionjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.reaction_coef 🇵 Reaction coefficient, kreaction_order 🇵 Reaction order, nreaction_rate 🇴 Reaction rateDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Copy the value from one variable to another.
from 🇮 Variable to copy value fromjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 Variable to copy value toDetailed documentation link
Crack geometric function associated with the AT-1 functional, \( \alpha = d \)
crack 🇴 Value of the crack geometric functionjit Use JIT compilation for the forward operatorphase 🇮 Phase-field variableproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Crack geometric function associated with the AT-2 functional, \( \alpha = d^2 \)
crack 🇴 Value of the crack geometric functionjit Use JIT compilation for the forward operatorphase 🇮 Phase-field variableproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
This class defines a cubic anisotropic elasticity tensor using three parameters. Various options are available for which three parameters to provide.
coefficient_as_parameter Whether to treat the coefficients as (trainable) parameters. Default is true. Setting this option to false will treat the coefficients as buffers.coefficient_types Types for each parameter, options are: INVALID, P_WAVE_MODULUS, POISSONS_RATIO, YOUNGS_MODULUS, SHEAR_MODULUS, BULK_MODULUS, LAME_LAMBDAcoefficients 🇵 Coefficients used to define the elasticity tensorjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Calculate the dimensionless inner and outer radii of the reaction product
inner_radius 🇴 Dimensionless inner radius of the product phasejit Use JIT compilation for the forward operatorouter_radius 🇴 Dimensionless outer radius of the product phaseproduct_fraction 🇮 Volume fraction of the product phaseproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.solid_fraction 🇮 Volume fraction of the solid phaseDetailed documentation link
Calculate the void fraction rate of change
diffusion_coefficient 🇵 Diffusion coefficient of the rate-limiting species in the product phasejit Use JIT compilation for the forward operatorliquid_reactivity 🇮 Reactivity of the liquid phase, between 0 and 1molar_volume Molar volume of the rate-limiting (liquid) speciesproduct_dummy_thickness Minimum product thickness to avoid division by 0product_inner_radius 🇮 Inner radius of the product phaseproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.reaction_rate 🇴 Product phase substance (mol/V) rate of changesolid_inner_radius 🇮 Inner raidus of the solid phasesolid_reactivity 🇮 Reactivity of the solid phase, between 0 and 1Detailed documentation link
Calculate the effective saturation, taking the form of \( S = \frac{\frac{\phi}{\phi_\mathrm{max}} - S_r}{1-S_r} \) where \( \phi \) is the volume fraction of the flowing fluid, \( \phi_\mathrm{max} \) is the maximum allowable volume fraction and \( S_r \) is the residual saturation.
effective_saturation 🇴 Effective saturationfluid_fraction 🇮 Volume fraction of the fluidjit Use JIT compilation for the forward operatormax_fraction 🇵 Maximum allowable volume fraction of the fluidproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.residual_saturation 🇵 Liquid's residual volume fractionDetailed documentation link
Calculate the total volume of a control mass. The volume has the form of \( V = \dfrac{M}{1-\phi_{o}} \sum_i \frac{\omega_i}{\rho_i} \), where \( \omega_i \) and \( \rho_i \) are respectively the mass fraction and the density of each component; \( \phi_{o} \) is the volume fraction accounting for leakage from the control mass; \( M \) is the reference mass of the composite.
composite_volume 🇴 Volume of the compositedensities 🇵 Densities of the components in the compositejit Use JIT compilation for the forward operatormass_fractions Mass fractions of the components in the compositeopen_volume_fraction 🇮 Open volume fraction accounting for leakageproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.reference_mass 🇵 Reference mass of the compositeDetailed documentation link
Calculates the elastic strain rate as \(\dot{\varepsilon} = d - d^p - \varepsilon w + w \varepsilon \) where \( d \) is the deformation rate, \( d^p \) is the plastic deformation rate, \( w \) is the vorticity, and \( \varepsilon \) is the elastic strain.
deformation_rate 🇮 Name of the deformation rateelastic_strain 🇮 Name of the elastic strainelastic_strain_rate 🇴 Name of the elastic strain ratejit Use JIT compilation for the forward operatorplastic_deformation_rate 🇮 Name of the plastic deformation rateproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.vorticity 🇮 Name of the vorticityDetailed documentation link
Define the relationship between non-dimensionalized porosity and permeability. The exponential porosity-permeability relation takes the form of \( K_0 \exp \left[ a(\varphi_o-\varphi) \right] \) where \( a \) is the scaling parameter; \( \varphi_0 \) and \( K_0 \) are the reference porosity and permeability respectively.
jit Use JIT compilation for the forward operatorpermeability 🇴 Porous flow permeabilityporosity 🇮 porosityproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.reference_permeability 🇵 the reference permeabilityreference_porosity 🇵 the reference porosityscale 🇵 Scaling constant in the exponential lawDetailed documentation link
By default, if \( a \ge 0, b \ge 0, ab = 0 \) then the Fischer Burmeister (FB) condition is: \(a+b-\sqrt(a^2+b^2)\), where a, b is the first_var and second_var respectively and first_inequality = second_inequality = 'GE'. One could set first_inequality = 'LE' (i.e. \( a \le 0, b \ge 0, ab = 0 \), FB conditions is \(-a+b-\sqrt(a^2+b^2) \)). Same goes for second_inequality = 'LE'.
first_inequality Type of inequality for the first variable.Default: GE. Options are LE, GEfirst_var 🇮 First conditionfischer_burmeister 🇴 Fischer Burmeister conditionjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.second_inequality Type of inequality for the second variable.Default: GE. Options are LE, GEsecond_var 🇮 Second conditionDetailed documentation link
Checks the value of the modified Rodrigues parameter by checking if \( \left\lVert r \right\rVert > t \), with \( t \) a threshold value set to 1.0 by default and replacing all the orientations that exceed this limit with their shadow parameters values.
input_orientation 🇮 Name of input tensor of orientations to operate on.jit Use JIT compilation for the forward operatoroutput_orientation 🇴 Name of output tensorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.threshold Threshold value for translating to the shadow parametersDetailed documentation link
Map the flow rate (i.e., the consistency parameter in the KKT conditions) to the rate of internal variables. This object defines the non-associative Fredrick-Armstrong kinematic hardening. In the model, back stress is directly treated as an internal variable. Rate of back stress is given as \( \dot{\boldsymbol{X}} = \left( \frac{2}{3} C \frac{\partial f}{\partial \boldsymbol{M}} - g \boldsymbol{X} \right) \dot{\gamma} \). \( \frac{\partial f}{\partial \boldsymbol{M}} \) is the flow direction, \( \dot{\gamma} \) is the flow rate, and \( C \) and \( g \) are material parameters.
C 🇵 Kinematic hardening coefficientback_stress 🇮 Back stressback_stress_rate 🇴 Back stress rate, defaults to back_stress + _rateflow_direction 🇮 Flow directionflow_rate 🇮 Flow rateg 🇵 Dynamic recovery coefficientjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Gurson-Tvergaard-Needleman yield function for poroplasticity. The yield function is defined as \( f = \left( \frac{\bar{\sigma}}{\sigma_y + k} \right)^2 + 2 q_1 \phi \cosh \left( \frac{1}{2} q_2 \frac{3\sigma_h-\sigma_s}{\sigma_y + k} \right) - \left( q_3 \phi^2 + 1 \right) \), where \( \bar{\sigma} \) is the von Mises stress, \( \sigma_y \) is the yield stress, \( k \) is isotropic hardening, \( \phi \) is the porosity, \( \sigma_h \) is the hydrostatic stress, and \( \sigma_s \) is the void growth back stress (sintering stress). \( q_1 \), \( q_2 \), and \( q_3 \) are parameters controlling the yield mechanisms.
flow_invariant 🇮 Effective stress driving plastic flowisotropic_hardening 🇮 Isotropic hardeningjit Use JIT compilation for the forward operatorporo_invariant 🇮 Effective stress driving porous flowproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.q1 🇵 Parameter controlling the balance/competition between plastic flow and void evolution.q2 🇵 Void evolution rateq3 🇵 Pore pressurevoid_fraction 🇮 Void fraction (porosity)yield_function 🇴 Yield functionyield_stress 🇵 Yield stressDetailed documentation link
Relates elastic strain to stress with some non-isotropic tensor. This verion implements a general relation using the elasticity tensor, expressed as an SSR4 object
compliance Whether the model defines the compliance relationship, i.e., mapping from stress to elastic strain. When set to false (default), the model maps elastic strain to stress.elastic_stiffness_tensor 🇵 Elastic stiffness tensorjit Use JIT compilation for the forward operatororientation 🇮 Active convention orientation from reference to currentproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate_form Whether the model defines the stress-strain relationship in rate form. When set to true, the model maps elastic strain rate to stress rate.strain 🇮 Elastic strainstress 🇴 StressDetailed documentation link
Green-Lagrange strain, \( E = \frac{1}{2} (C - I) \), where \( C = F^T F \) is the right Cauchy-Green tensor and \( I \) is the identity tensor.
deformation_gradient 🇮 The deformation gradientjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.strain 🇴 The Green-Lagrange strainDetailed documentation link
Local mass balance used in conjunction with the GTNYieldFunction, \( \dot{\phi} = (1-\phi) \dot{\varepsilon}_p \).
jit Use JIT compilation for the forward operatorplastic_strain_rate 🇮 Plastic strain rateproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.void_fraction 🇮 Void fraction (porosity)void_fraction_rate 🇴 Rate of void evolutionDetailed documentation link
The smooth step function defined by Hermite polynomials
argument 🇮 Argument of the smooth step functioncomplement Whether takes 1 to subtract the function.jit Use JIT compilation for the forward operatorlower_bound 🇧 Lower bound of the argumentproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.upper_bound 🇧 Upper bound of the argumentvalue 🇴 Value of the smooth step functionDetailed documentation link
Update an implicit model by solving the underlying implicit system of equations.
implicit_model The implicit model defining the implicit system of equations to be solvedproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.solver Solver used to solve the implicit systemDetailed documentation link
This class defines an isotropic elasticity tensor using two parameters. Various options are available for which two parameters to provide.
coefficient_as_parameter Whether to treat the coefficients as (trainable) parameters. Default is true. Setting this option to false will treat the coefficients as buffers.coefficient_types Types for each parameter, options are: INVALID, P_WAVE_MODULUS, POISSONS_RATIO, YOUNGS_MODULUS, SHEAR_MODULUS, BULK_MODULUS, LAME_LAMBDAcoefficients 🇵 Coefficients used to define the elasticity tensorjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Map Cauchy stress to Mandel stress For isotropic material under small deformation, the Mandel stress and the Cauchy stress coincide.
cauchy_stress 🇮 Cauchy stressjit Use JIT compilation for the forward operatormandel_stress 🇴 Mandel stressproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Calculates the Kocks-Mecking normalized activation as \(g = \frac{kT}{\mu b^3} \log \frac{\dot{\varepsilon}_0}{\dot{\varepsilon}} \) with \( \mu \) the shear modulus, \( k \) the Boltzmann constant, \( T \) the absolute temperature, \( b \) the Burgers vector length, \( \dot{\varepsilon}_0 \) a reference strain rate, and \( \dot{\varepsilon} \) the current strain rate.
activation_energy 🇴 Output name of the activation energyb Magnitude of the Burgers vectoreps0 Reference strain ratejit Use JIT compilation for the forward operatork The Boltzmann constantproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.shear_modulus 🇵 The shear modulusstrain_rate 🇮 Name of the effective strain ratetemperature 🇮 Absolute temperatureDetailed documentation link
Switches between rate independent and rate dependent flow rules based on the value of the Kocks-Mecking normalized activation energy. For activation energies less than the threshold use the rate independent flow rule, for values greater than the threshold use the rate dependent flow rule. This version uses a soft switch between the models, based on a tanh sigmoid function.
activation_energy 🇮 The input name of the activation energyflow_rate 🇴 Output name for the mixed flow rateg0 🇵 Critical value of activation energyjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate_dependent_flow_rate 🇮 Input name of the rate dependent flow raterate_independent_flow_rate 🇮 Input name of the rate independent flow ratesharpness A steepness parameter that controls the tanh mixing of the models. Higher values gives a sharper transition.Detailed documentation link
Calculates the temperature-dependent flow viscosity for a Perzyna-type model using the Kocks-Mecking model. The value is \( \eta = \exp{B} \mu \dot{\varepsilon}_0^\frac{-k T A}{\mu b^3} \) with \( \mu \) the shear modulus, \( \dot{\varepsilon}_0 \) a reference strain rate, \( b \) the Burgers vector, \( k\) the Boltzmann constant, \( T \) absolute temperature, \( A \) the Kocks-Mecking slope parameter, and \( B \) the Kocks-Mecking intercept parameter.
A 🇵 The Kocks-Mecking slope parameterB 🇵 The Kocks-Mecking intercept parameterb The Burgers vectoreps0 The reference strain ratejit Use JIT compilation for the forward operatork Boltzmann constantproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.shear_modulus 🇵 The shear modulustemperature 🇮 Absolute temperatureDetailed documentation link
The critical value of the normalized activation energy given by \( g_0 \frac{C-B}{A} \)
A 🇵 The Kocks-Mecking slopeB 🇵 The Kocks-Mecking interceptC 🇵 The Kocks-Mecking horizontal valueintercept 🇴 The interceptjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Calculates the temperature-dependent rate sensitivity for a Perzyna-type model using the Kocks-Mecking model. The value is \( n = \frac{\mu b^3}{k T A} \) with \( \mu \) the shear modulus, \( b \) the Burgers vector, \( k\) the Boltzmann constant, \( T \) absolute temperature, and \( A \) the Kocks-Mecking slope parameter.
A 🇵 The Kocks-Mecking slope parameterb The Burgers vectorjit Use JIT compilation for the forward operatork Boltzmann constantproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.shear_modulus 🇵 The shear modulustemperature 🇮 Absolute temperatureDetailed documentation link
The yield stress given by the Kocks-Mecking model. \( \sigma_y = \exp{C} \mu \) with \( \mu \) the shear modulus and \( C \) the horizontal intercept from the Kocks-Mecking diagram.
C 🇵 The Kocks-Mecking horizontal interceptjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.shear_modulus 🇵 The shear modulusDetailed documentation link
Define the relationship between non-dimensionalized porosity and permeability. The Kozeny-Carman porosity-permeability relation takes the form of \( K = K_0 \frac{\varphi^n (1-\varphi_0^m)}{\varphi_0^m (1-\varphi)^n} \) where \( n \) and \( m \) are shape parameters. \( varphi_0 \) and \( K_0 \) are the reference porosity and permeability respectively.
jit Use JIT compilation for the forward operatorm 🇵 Shape parameter mn 🇵 Shape parameter npermeability 🇴 Porous flow permeabilityporosity 🇮 porosityproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.reference_permeability 🇵 the reference permeabilityreference_porosity 🇵 the reference porosityDetailed documentation link
Update the trial stress under the assumptions of J2 plasticity and isotropic linear elasticity
coefficient_as_parameter Whether to treat the coefficients as (trainable) parameters. Default is true. Setting this option to false will treat the coefficients as buffers.coefficient_types Types for each parameter, options are: INVALID, P_WAVE_MODULUS, POISSONS_RATIO, YOUNGS_MODULUS, SHEAR_MODULUS, BULK_MODULUS, LAME_LAMBDAcoefficients 🇵 Coefficients used to define the elasticity tensorelastic_trial_stress 🇮 Initial trial stress assuming a purely elastic stepequivalent_plastic_strain 🇮 Current guess for the equivalent plastic strainjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.updated_trial_stress 🇴 Trial stress corrected for the current increment of plastic deformationDetailed documentation link
Relate elastic strain to stress for linear isotropic material. \( \boldsymbol{\sigma} = K \tr \boldsymbol{\varepsilon}_e + 2 G \text{dev} \boldsymbol{\varepsilon}_e \), where \( K \) and \( G \) are bulk and shear moduli, respectively. Other pairs of Lame parameters are also supported.
coefficient_as_parameter Whether to treat the coefficients as (trainable) parameters. Default is true. Setting this option to false will treat the coefficients as buffers.coefficient_types Types for each parameter, options are: INVALID, P_WAVE_MODULUS, POISSONS_RATIO, YOUNGS_MODULUS, SHEAR_MODULUS, BULK_MODULUS, LAME_LAMBDAcoefficients 🇵 Coefficients used to define the elasticity tensorcompliance Whether the model defines the compliance relationship, i.e., mapping from stress to elastic strain. When set to false (default), the model maps elastic strain to stress.jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate_form Whether the model defines the stress-strain relationship in rate form. When set to true, the model maps elastic strain rate to stress rate.strain 🇮 Elastic strainstress 🇴 StressDetailed documentation link
Map equivalent plastic strain to isotropic hardening following a linear relationship, i.e., \( h = K \bar{\varepsilon}_p \) where \( K \) is the hardening modulus.
equivalent_plastic_strain 🇮 Equivalent plastic strainhardening_modulus 🇵 Hardening modulusisotropic_hardening 🇴 Isotropic hardeningjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Calculates elastic strain energy density based on linear elastic isotropic response
coefficient_as_parameter Whether to treat the coefficients as (trainable) parameters. Default is true. Setting this option to false will treat the coefficients as buffers.coefficient_types Types for each parameter, options are: INVALID, P_WAVE_MODULUS, POISSONS_RATIO, YOUNGS_MODULUS, SHEAR_MODULUS, BULK_MODULUS, LAME_LAMBDAcoefficients 🇵 Coefficients used to define the elasticity tensordecomposition Strain energy density decomposition types, options are: VOLDEV, SPECTRAL, NONEjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.strain 🇮 Elastic strainstrain_energy_density_active 🇴 Active part of the strain energy densitystrain_energy_density_inactive 🇴 Inactive part of the strain energy densityDetailed documentation link
Map kinematic plastic strain to back stress following a linear relationship, i.e., \( \boldsymbol{X} = H \boldsymbol{K}_p \) where \( H \) is the hardening modulus.
back_stress 🇴 Back stresshardening_modulus 🇵 Hardening modulusjit Use JIT compilation for the forward operatorkinematic_plastic_strain 🇮 Kinematic plastic strainproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Simple linear slip system hardening defined by \( \dot{\tau} = \theta \sum_{i=1}^{n_{slip}} \left| \dot{\gamma}_i \right| \) where \( \theta \) is the hardening slope.
hardening_slope 🇵 Hardening ratejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.slip_hardening 🇮 Name of current values of slip hardeningslip_hardening_rate 🇴 Name of tensor to output the slip system hardening rates intosum_slip_rates 🇮 Name of tensor containing the sum of the slip ratesDetailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
Object to setup a model for mixed control. Copies the values of the fixed_values and the mixed_state (the conjugate) into the input variables used by the model.
above_variable 🇴 The prescribed variable when the control signal is greater than the thresholdbelow_variable 🇴 The prescribed variable when the control signal is less than the thresholdcontrol 🇮 The control signal.fixed_values 🇮 The name of the fixed values, i.e. the actual values being imposed on the modeljit Use JIT compilation for the forward operatormixed_state 🇮 The name of the mixed state tensor. This holds the conjugate values to those being controlledproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.threshold The threshold to switch between the two controlDetailed documentation link
Store the first derivatives of a scalar-valued function in given variables, i.e. \( u_i = \dfrac{f(\boldsymbol{v})}{v_i} \).
from Function arguments to take derivatives w.r.t.function Function to take derivativejit Use JIT compilation for the forward operatormodel The model which evaluates the scalar-valued functionproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to Variables to store the first derivativesDetailed documentation link
Define the Olevsky-Skorohod sintering stress to be used in conjunction with poroplasticity yield functions such as the GTNYieldFunction. The sintering stress is defined as \( \sigma_s = 3 \dfrac{\gamma}{r} \phi^2 \), where \( \gamma \) is the surface tension, \( r \) is the size of the particles/powders, and \( \phi \) is the void fraction.
jit Use JIT compilation for the forward operatorparticle_radius 🇵 Particle radiusproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.sintering_stress 🇴 Sintering stresssurface_tension 🇵 Surface tensionvoid_fraction 🇮 Void fractionDetailed documentation link
Defines the rate of the crystal orientations as a spin given by \( \Omega^e = w - w^p - \varepsilon d^p + d^p \varepsilon \) where \( \Omega^e = \dot{Q} Q^T \), \( Q \) is the orientation, \( w \) is the vorticity, \( w^p \) is the plastic vorticity, \( d^p \) is the plastic deformation rate, and \( \varepsilon \) is the elastic stretch.
elastic_strain 🇮 The name of the elastic strain tensorjit Use JIT compilation for the forward operatororientation_rate 🇴 The name of the orientation rate (spin)plastic_deformation_rate 🇮 The name of the plastic deformation rateplastic_vorticity 🇮 The name of the plastic vorticityproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.vorticity 🇮 The name of the voriticty tensorDetailed documentation link
Perzyna's viscous approximation of the consistent yield envelope (with a power law), i.e. \( \dot{\gamma} = \left( \frac{\left< f \right>}{\eta} \right)^n \), where \( f \) is the yield function, \( \eta \) is the reference stress, and \( n \) is the power-law exponent.
exponent 🇵 Power-law exponentflow_rate 🇴 Flow ratejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.reference_stress 🇵 Reference stressyield_function 🇮 Yield functionDetailed documentation link
Define the (cummulative, as opposed to instantaneous) linear isotropic phase transformation (from phase A to phase B) eigenstrain, i.e. \( \boldsymbol{\varepsilon}_\mathrm{PT} = \Delta V f \boldsymbol{I} \), where \( \Delta V \) is the volume fraction change when going from phase A to B, \( f \) is the phase fraction (0 to 1, A to B).
eigenstrain 🇴 Eigenstrainjit Use JIT compilation for the forward operatorphase_fraction 🇮 Phase fractionproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.volume_fraction_change 🇮 Change in volume fraction going from phase A to phase BDetailed documentation link
Caclulates the plastic deformation rate as \( d^p = \sum_{i=1}^{n_{slip}} \dot{\gamma}_i Q \operatorname{sym}{\left(d_i \otimes n_i \right)} Q^T \) with \( d^p \) the plastic deformation rate, \( \dot{\gamma}_i \) the slip rate on the ith slip system, \(Q \) the orientation, \( d_i \) the slip system direction, and \( n_i \) the slip system normal.
crystal_geometry The name of the Data object containing the crystallographic information for the materialjit Use JIT compilation for the forward operatororientation 🇮 The name of the orientation matrix tensorplastic_deformation_rate 🇴 The name of the plastic deformation rate tensorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.slip_rates 🇮 The name of the tensor containg the current slip ratesDetailed documentation link
Caclulates the plastic spatial velocity gradient as \( l^p = \sum_{i=1}^{n_{slip}} \dot{\gamma}_i Q \left(d_i \otimes n_i \right) Q^T \) with \( l^p \) the plastic spatial velocity gradient, \( \dot{\gamma}_i \) the slip rate on the ith slip system, \(Q \) the orientation, \( d_i \) the slip system direction, and \( n_i \) the slip system normal.
crystal_geometry The name of the Data object containing the crystallographic information for the materialjit Use JIT compilation for the forward operatororientation 🇮 The name of the orientation matrix tensorplastic_spatial_velocity_gradient 🇴 The name of the plastic spatial velocity gradientproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.slip_rates 🇮 The name of the tensor containg the current slip ratesDetailed documentation link
Caclulates the plastic vorcitity as \( w^p = \sum_{i=1}^{n_{slip}} \dot{\gamma}_i Q \operatorname{skew}{\left(d_i \otimes n_i \right)} Q^T \) with \( d^p \) the plastic deformation rate, \( \dot{\gamma}_i \) the slip rate on the ith slip system, \(Q \) the orientation, \( d_i \) the slip system direction, and \( n_i \) the slip system normal.
crystal_geometry The name of the Data object containing the crystallographic information for the materialjit Use JIT compilation for the forward operatororientation 🇮 The name of the orientation matrix tensorplastic_vorticity 🇴 The name of the plastic vorticity tensorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.slip_rates 🇮 The name of the tensor containg the current slip ratesDetailed documentation link
Power degradation function to degrade the elastic strain energy density, \( g = \left( 1-d \right)^p (1-\eta) + \eta \)
degradation 🇴 Value of the dedgradation functioneta Residual degradation when d = 1jit Use JIT compilation for the forward operatorphase 🇮 Phase-field variablepower Power of the degradation functionproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
This particular model implements a power law recovery of the type \( \dot{k} = -\left(\frac{\lVert k \rVert}{\tau}\right)^{n-1} \frac{k}{\tau} \)
isotropic_hardening 🇮 Isotropic hardening variableisotropic_hardening_rate 🇮 Rate of isotropic hardening, defaults to isotropic_hardening + _recovery_ratejit Use JIT compilation for the forward operatorn 🇵 Recovery exponentproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.tau 🇵 Recovery rateDetailed documentation link
This object defines kinematic hardening static recovery on a backstress term. This particular model uses a power law for recovery \( \dot{X} = - \left(\frac{\lVert X \rVert}{\tau}\right)^{n-1} \frac{X}{\tau} \)where \( n \) is the power law recovery exponent and \(\tau\) is the recovery rate.
back_stress 🇮 Back stressback_stress_rate 🇴 Back stress rate, defaults to back_stress + _recovery_ratejit Use JIT compilation for the forward operatorn 🇵 Static recovery exponentproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.tau 🇵 Static recovery rateDetailed documentation link
Define the relationship between non-dimensionalized porosity and permeability. The power law porosity-permeability relation takes the form of \( K_0 \left( \frac{\varphi}{\varphi_0} \right)^p \). \( varphi_0 \) and \( K_0 \) are the reference porosity and permeability respectively.
exponent 🇵 Exponent in the power lawjit Use JIT compilation for the forward operatorpermeability 🇴 Porous flow permeabilityporosity 🇮 porosityproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.reference_permeability 🇵 the reference permeabilityreference_porosity 🇵 the reference porosityDetailed documentation link
Power law slip rule defined as \( \dot{\gamma}_i = \dot{\gamma}_0 \left| \frac{\tau_i}{\hat{\tau}_i} \right|^{n-1} \frac{\tau_i}{\hat{\tau}_i} \) with \( \dot{\gamma}_i \) the slip rate on system \( i \), \( \tau_i \) the resolved shear, \( \hat{\tau}_i \) the slip system strength, \( n \) the rate senstivity, and \( \dot{\gamma}_0 \) a reference slip rate.
gamma0 🇵 Reference slip ratejit Use JIT compilation for the forward operatorn 🇵 Rate sensitivity exponentproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.resolved_shears 🇮 Name of the resolved shear tensorslip_rates 🇴 Name of the slip rate tensorslip_strengths 🇮 Name of the tensor containing the slip system strengthsDetailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
Define the backward Euler time integration residual \( r = s - s_n - (t - t_n) \dot{s} \), where \(s\) is the variable being integrated, \(\dot{s}\) is the variable rate, and \(t\) is time. Subscripts \(n\) denote quantities from the previous time step.
automatic_scaling Whether to perform automatic scaling. See neml2::NonlinearSystem::init_scaling for implementation details.automatic_scaling_miter Maximum number of automatic scaling iterations. No error is produced upon reaching the maximum number of scaling iterations, and the scaling matrices obtained at the last iteration are used to scale the nonlinear system.automatic_scaling_tol Tolerance used in iteratively updating the scaling matrices.jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇮 Variable ratetime 🇮 Timevariable 🇮 Variable being integratedDetailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Calculate the Jacobian of a second order tensor.
determinant 🇴 The determinant of the input tensorinput 🇮 The second order tensor to calculate the determinant ofjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Calculate the first order discrete time derivative of a variable as \( \dot{f} = \frac{\Delta f}{t-t_n} \), where \( \Delta f \) is the variable with the increment, and \( t \) is time.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇴 The variable's rate of changetime 🇮 Timevariable 🇮 The incremental valueDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Calculate linear combination of multiple R2 tensors as \( u = c_i v_i + s \) (Einstein summation assumed), where \( c_i \) are the coefficients, and \( v_i \) are the variables to be summed. \( s \) is a constant offset.
coefficient_as_parameter By default, the coefficients are declared as buffers. Set this option to true to declare them as (trainable) parameters. This option takes a list of booleans, one for each coefficient. When the length of this list is 1, the boolean applies to all coefficients.coefficients 🇵 Weights associated with each variable. This option takes a list of weights, one for each coefficient. When the length of this list is 1, the same weight applies to all coefficients.constant_coefficient 🇵 The constant coefficient added to the final summationconstant_coefficient_as_parameter By default, the constant_coefficient are declared as buffers. Set this option to true to declare them as (trainable) parameters.from_var R2 tensors to be summedjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to_var 🇴 The sumDetailed documentation link
Multiplication of form \( A B \), where \( A \) and \( B \) are second order tensors. A and B can be inverted and/or transposed per request.
A Variable AB Variable Binvert_A Whether to invert Ainvert_B Whether to invert Bjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The result of the multiplicationtranspose_A Whether to transpose Atranspose_B Whether to transpose BDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
Extract the symmetric part of a R2 tensor
input 🇮 Rank two tensor to splitjit Use JIT compilation for the forward operatoroutput 🇴 Output symmetric rank two tensorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Extract the skew symmetric part of a R2 tensor
input 🇮 Rank two tensor to splitjit Use JIT compilation for the forward operatoroutput 🇴 Output skew symmetric rank two tensorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
Solve the consistent plasticity yield envelope by solving the equivalent complementarity condition
\[ r = \dot{\gamma} - f^p - \sqrt{{\dot{\gamma}}^2 + {f^p}^2} \]
flow_rate 🇮 Flow ratejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.yield_function 🇮 Yield functionDetailed documentation link
Power degradation function to degrade the elastic strain energy density, \( g = \frac{\left( 1-d \right)^p}{\left( 1-d \right)^p + Q\left(d \right)} \) where, \( Q\left(d \right) = b_{1}d\left( 1+b_{2}d+b_{2}b_{3}d^2 \right)\)
degradation 🇴 Value of the dedgradation functioneta Residual degradation when d = 1fitting_param_1 Material dependent fitting parameter 1fitting_param_2 Material dependent fitting parameter 2fitting_param_3 Material dependent fitting parameter 3jit Use JIT compilation for the forward operatorphase 🇮 Phase-field variablepower Power of the degradation functionproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Calculates the resolved shears as \( \tau_i = \sigma : Q \operatorname{sym}\left(d_i \otimes n_i \right) Q^T \) where \( \tau_i \) is the resolved shear on slip system i, \( \sigma \) is the Cauchy stress \( Q \) is the orientation matrix, \( d_i \) is the slip direction, and \( n_i \) is the slip system normal.
crystal_geometry The name of the data object with the crystallographic informationjit Use JIT compilation for the forward operatororientation 🇮 The name of the orientation matrixproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.resolved_shears 🇴 The name of the resolved shearsstress 🇮 The name of the Cauchy stress tensorDetailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
Convert a Rot (rotation represented in Rodrigues format) to R2 (a full rotation matrix).
from 🇮 Rot to convertjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 R2 to store the resulting rotation matrixDetailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
Define the backward Euler time integration residual \( r = s - s_n - (t - t_n) \dot{s} \), where \(s\) is the variable being integrated, \(\dot{s}\) is the variable rate, and \(t\) is time. Subscripts \(n\) denote quantities from the previous time step.
automatic_scaling Whether to perform automatic scaling. See neml2::NonlinearSystem::init_scaling for implementation details.automatic_scaling_miter Maximum number of automatic scaling iterations. No error is produced upon reaching the maximum number of scaling iterations, and the scaling matrices obtained at the last iteration are used to scale the nonlinear system.automatic_scaling_tol Tolerance used in iteratively updating the scaling matrices.jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇮 Variable ratetime 🇮 Timevariable 🇮 Variable being integratedDetailed documentation link
Interpolate a SR2 as a function of the given argument. See neml2::Interpolation for rules on shapes of the interpolant and the argument. This object performs a bilinear interpolation.
abscissa1 Scalar defining the abscissa values of the first interpolation axisabscissa2 Scalar defining the abscissa values of the second interpolation axisargument1 🇮 First argument used to query the interpolant along the first axisargument2 🇮 Second argument used to query the interpolant along the second axisdim Intermediate dimension along which to interpolatejit Use JIT compilation for the forward operatorordinate SR2 defining the ordinate values of the interpolantoutput 🇴 SR2 output of the interpolantproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Calculate the Jacobian of a second order tensor.
determinant 🇴 The determinant of the input tensorinput 🇮 The second order tensor to calculate the determinant ofjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Perform forward Euler time integration defined as \( s = s_n + (t - t_n) \dot{s} \), where \(s\) is the variable being integrated, \(\dot{s}\) is the variable rate, and \(t\) is time. Subscripts \(n\) denote quantities from the previous time step.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇮 Variable rate of changetime 🇮 Timevariable 🇴 Variable being integratedDetailed documentation link
Calculate the first order discrete time derivative of a variable as \( \dot{f} = \frac{\Delta f}{t-t_n} \), where \( \Delta f \) is the variable with the increment, and \( t \) is time.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇴 The variable's rate of changetime 🇮 Timevariable 🇮 The incremental valueDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Calculate the invariant of a symmetric second order tensor (of type SR2).
invariant 🇴 Invariantinvariant_type Type of invariant. Options are: INVALID, EFFECTIVE_STRAIN, VONMISES, I2, I1jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.tensor 🇮 SR2 which is used to calculate the invariant ofDetailed documentation link
Calculate linear combination of multiple SR2 tensors as \( u = c_i v_i + s \) (Einstein summation assumed), where \( c_i \) are the coefficients, and \( v_i \) are the variables to be summed. \( s \) is a constant offset.
coefficient_as_parameter By default, the coefficients are declared as buffers. Set this option to true to declare them as (trainable) parameters. This option takes a list of booleans, one for each coefficient. When the length of this list is 1, the boolean applies to all coefficients.coefficients 🇵 Weights associated with each variable. This option takes a list of weights, one for each coefficient. When the length of this list is 1, the same weight applies to all coefficients.constant_coefficient 🇵 The constant coefficient added to the final summationconstant_coefficient_as_parameter By default, the constant_coefficient are declared as buffers. Set this option to true to declare them as (trainable) parameters.from_var SR2 tensors to be summedjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to_var 🇴 The sumDetailed documentation link
Interpolate a SR2 as a function of the given argument. See neml2::Interpolation for rules on shapes of the interpolant and the argument. This object performs a linear interpolation.
abscissa Scalar defining the abscissa values of the interpolantargument 🇮 Argument used to query the interpolantdim Intermediate dimension along which to interpolatejit Use JIT compilation for the forward operatorordinate SR2 defining the ordinate values of the interpolantoutput 🇴 SR2 output of the interpolantproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
Thermal annealing recovery for a hardening variable of type SR2.For temperatures below \( T_1 \) the model keeps the base model hardenign rate.For temperatures above \(T_1\) but below \(T_2 \) the model zeros the hardening rate.For temperatures above \(T_2\) the model replaces the hardening rate with \( \dot{h} = \frac{-h}{\tau} \) where \( \tau \) is the rate of recovery.
T1 🇵 First stage annealing temperatureT2 🇵 Second stage annealing temperaturebase 🇮 Underlying base hardening variablebase_rate 🇮 Base hardening ratejit Use JIT compilation for the forward operatormodified_rate 🇴 Output for the modified hardening rate.production Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.tau 🇵 Recovery rate for second stage annealing.temperature 🇮 TemperatureDetailed documentation link
Calculate the first order discrete time derivative of a variable as \( \dot{f} = \frac{f-f_n}{t-t_n} \), where \( f \) is the force variable, and \( t \) is time.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇴 The variable's rate of changetime 🇮 Timevariable 🇮 The variable to take time derivative withDetailed documentation link
Convert a symmetric rank two tensor to a full tensor
input 🇮 Symmetric tensor to convertjit Use JIT compilation for the forward operatoroutput 🇴 Output full rank two tensorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
Define the backward Euler time integration residual \( r = s - s_n - (t - t_n) \dot{s} \), where \(s\) is the variable being integrated, \(\dot{s}\) is the variable rate, and \(t\) is time. Subscripts \(n\) denote quantities from the previous time step.
automatic_scaling Whether to perform automatic scaling. See neml2::NonlinearSystem::init_scaling for implementation details.automatic_scaling_miter Maximum number of automatic scaling iterations. No error is produced upon reaching the maximum number of scaling iterations, and the scaling matrices obtained at the last iteration are used to scale the nonlinear system.automatic_scaling_tol Tolerance used in iteratively updating the scaling matrices.jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇮 Variable ratetime 🇮 Timevariable 🇮 Variable being integratedDetailed documentation link
Interpolate a Scalar as a function of the given argument. See neml2::Interpolation for rules on shapes of the interpolant and the argument. This object performs a bilinear interpolation.
abscissa1 Scalar defining the abscissa values of the first interpolation axisabscissa2 Scalar defining the abscissa values of the second interpolation axisargument1 🇮 First argument used to query the interpolant along the first axisargument2 🇮 Second argument used to query the interpolant along the second axisdim Intermediate dimension along which to interpolatejit Use JIT compilation for the forward operatorordinate Scalar defining the ordinate values of the interpolantoutput 🇴 Scalar output of the interpolantproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Perform forward Euler time integration defined as \( s = s_n + (t - t_n) \dot{s} \), where \(s\) is the variable being integrated, \(\dot{s}\) is the variable rate, and \(t\) is time. Subscripts \(n\) denote quantities from the previous time step.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇮 Variable rate of changetime 🇮 Timevariable 🇴 Variable being integratedDetailed documentation link
Calculate the first order discrete time derivative of a variable as \( \dot{f} = \frac{\Delta f}{t-t_n} \), where \( \Delta f \) is the variable with the increment, and \( t \) is time.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇴 The variable's rate of changetime 🇮 Timevariable 🇮 The incremental valueDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Calculate linear combination of multiple Scalar tensors as \( u = c_i v_i + s \) (Einstein summation assumed), where \( c_i \) are the coefficients, and \( v_i \) are the variables to be summed. \( s \) is a constant offset.
coefficient_as_parameter By default, the coefficients are declared as buffers. Set this option to true to declare them as (trainable) parameters. This option takes a list of booleans, one for each coefficient. When the length of this list is 1, the boolean applies to all coefficients.coefficients 🇵 Weights associated with each variable. This option takes a list of weights, one for each coefficient. When the length of this list is 1, the same weight applies to all coefficients.constant_coefficient 🇵 The constant coefficient added to the final summationconstant_coefficient_as_parameter By default, the constant_coefficient are declared as buffers. Set this option to true to declare them as (trainable) parameters.from_var Scalar tensors to be summedjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to_var 🇴 The sumDetailed documentation link
Interpolate a Scalar as a function of the given argument. See neml2::Interpolation for rules on shapes of the interpolant and the argument. This object performs a linear interpolation.
abscissa Scalar defining the abscissa values of the interpolantargument 🇮 Argument used to query the interpolantdim Intermediate dimension along which to interpolatejit Use JIT compilation for the forward operatorordinate Scalar defining the ordinate values of the interpolantoutput 🇴 Scalar output of the interpolantproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Calculate the multiplication (product) of multiple Scalar variable with a constant coefficient. Using reciprocal, one can have the reciprocity of variable 'a', aka. '1/a'
coefficient 🇵 The coefficient multiply to the final productfrom_var Scalar variables to be multipliedjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.reciprocal List of boolens, one for each variable, in which the reciprocity of the corresponding variable is taken. When the length of this list is 1, the same reciprocal condition applies to all variables.to_var 🇴 The multiplicative productDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
Create a diagonal symmetric rank 2 tensor with values filled by a scalar
input 🇮 Symmetric tensor to convertjit Use JIT compilation for the forward operatoroutput 🇴 Output full rank two tensorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Thermal annealing recovery for a hardening variable of type Scalar.For temperatures below \( T_1 \) the model keeps the base model hardenign rate.For temperatures above \(T_1\) but below \(T_2 \) the model zeros the hardening rate.For temperatures above \(T_2\) the model replaces the hardening rate with \( \dot{h} = \frac{-h}{\tau} \) where \( \tau \) is the rate of recovery.
T1 🇵 First stage annealing temperatureT2 🇵 Second stage annealing temperaturebase 🇮 Underlying base hardening variablebase_rate 🇮 Base hardening ratejit Use JIT compilation for the forward operatormodified_rate 🇴 Output for the modified hardening rate.production Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.tau 🇵 Recovery rate for second stage annealing.temperature 🇮 TemperatureDetailed documentation link
Calculate the first order discrete time derivative of a variable as \( \dot{f} = \frac{f-f_n}{t-t_n} \), where \( f \) is the force variable, and \( t \) is time.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇴 The variable's rate of changetime 🇮 Timevariable 🇮 The variable to take time derivative withDetailed documentation link
Calculates the slip system strength for all slip systems as \( \hat{\tau}_i = \bar{\tau} + \tau_0 \) where \( \hat{\tau}_i \) is the strength for slip system i, \( \bar{\tau} \) is an evolving slip system strength (one value of all systems), defined by another object, and \( \tau_0 \) is a constant strength.
constant_strength 🇵 The constant slip system strengthjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.slip_hardening 🇮 The name of the evolving, scalar strengthslip_strengths 🇴 Name of the slip system strengthsDetailed documentation link
SlopeSaturationVoce isotropic hardening model, \( \dot{h} = \theta_0 \left(1 - \frac{h}{R} \right) \varepsilon_p \), where \( R \) is the isotropic hardening upon saturation, and \( \theta_0 \) is the initial hardening rate. In addition to the reparameterization, this model differences from the VoceIsotropicHardening model in that it defines the hardening rate in a non-assocative manner. This is sometimes handy, for example in supplementing the model with static recovery.
flow_rate 🇮 Flow rateinitial_hardening_rate 🇵 Initial hardening rateisotropic_hardening 🇮 Isotropic hardening variableisotropic_hardening_rate 🇮 Rate of isotropic hardening, defaults to isotropic_hardening + _ratejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.saturated_hardening 🇵 Saturated isotropic hardeningDetailed documentation link
Calculates the sum of the absolute value of all the slip rates as \( \sum_{i=1}^{n_{slip}} \left| \dot{\gamma}_i \right| \).
dim The intermediate dimension over which to sum the slip rates.jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.slip_rates 🇮 The name of individual slip ratessum_slip_rates 🇴 The output name for the scalar sum of the slip ratesDetailed documentation link
Define the linear isotropic phase change deformation Jacobian for a freezing liquid or a melting solid, i.e. \( J = \left( 1 + \alpha c \phi^f + (1-c) \phi^f \Delta \Omega \right) \), where \( \alpha \) is the coefficient of swelling, \( \Delta \Omega \) is relative difference of the reference volume between the two phases, \( \phi^f \) is the fluid fraction associated with swelling, and \( c \) is the phase fraction.
fluid_fraction 🇮 Volume fraction of the fluid phase.jacobian Phase change deformation Jacobianjit Use JIT compilation for the forward operatorphase_fraction 🇵 Phase fraction during the phase change. 0 means all solid, 1 means all liquid.production Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.reference_volume_difference 🇵 Relative difference between the reference volumes of the two phases.swelling_coefficient 🇵 Coefficient of phase expansionDetailed documentation link
Define the symmetric Hermite interpolation function, taking the form of \( \dfrac{1}{x_h-x_l}(24c^2-32c^3) \) for \( 0 le c le 0.5 \); \( \dfrac{1}{x_h-x_l} (24(1-c)^2 - 32(1-c)^3) \) for \( 0.5 le c le 1 \), and 0.0 otherwise. Here, \( c = \frac{x-x_l}{x_h-x_l} \) where \(x_l\) and \(x_h\) are the lower and upper bound for rescaling the input argument.
argument 🇮 Argument of the smooth step functionjit Use JIT compilation for the forward operatorlower_bound 🇧 Lower bound of the argumentproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.upper_bound 🇧 Upper bound of the argumentvalue 🇴 Value of the smooth step functionDetailed documentation link
Define the linear isotropic thermal deformation Jacobian, i.e. \( J = 1 + \alpha (T - T_0) \), where \( \alpha \) is the coefficient of thermal expansion (CTE), \( T \) is the temperature, and \( T_0 \) is the reference (stress-free) temperature.
CTE 🇵 Coefficient of thermal expansionjacobian Thermal deformation Jacobianjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.reference_temperature 🇧 Reference (stress-free) temperaturetemperature 🇮 TemperatureDetailed documentation link
Define the (cummulative, as opposed to instantaneous) linear isotropic thermal eigenstrain, i.e. \( \boldsymbol{\varepsilon}_T = \alpha (T - T_0) \boldsymbol{I} \), where \( \alpha \) is the coefficient of thermal expansion (CTE), \( T \) is the temperature, and \( T_0 \) is the reference (stress-free) temperature.
CTE 🇵 Coefficient of thermal expansioneigenstrain 🇴 Eigenstrainjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.reference_temperature 🇧 Reference (stress-free) temperaturetemperature 🇮 TemperatureDetailed documentation link
Relate the porous flow capillary pressure to the effective saturation using the van Genuchten correlation for capillary pressure, taking the form of \( a \left( S_e^{-\frac{1}{m}} - 1 \right)^{1-m} \). Here \( S_e \) is the effective saturation, \( a \) and \( m \) are shape parameters
a 🇵 Shape parameter acapillary_pressure 🇴 Capillary pressure.effective_saturation 🇮 The effective saturationjit Use JIT compilation for the forward operatorlog_extension Whether to apply logarithmic extensionm 🇵 Shape parameter mproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.transition_saturation The transistion value of the effective saturation below which to apply the logarithmic extensionDetailed documentation link
Define the backward Euler time integration residual \( r = s - s_n - (t - t_n) \dot{s} \), where \(s\) is the variable being integrated, \(\dot{s}\) is the variable rate, and \(t\) is time. Subscripts \(n\) denote quantities from the previous time step.
automatic_scaling Whether to perform automatic scaling. See neml2::NonlinearSystem::init_scaling for implementation details.automatic_scaling_miter Maximum number of automatic scaling iterations. No error is produced upon reaching the maximum number of scaling iterations, and the scaling matrices obtained at the last iteration are used to scale the nonlinear system.automatic_scaling_tol Tolerance used in iteratively updating the scaling matrices.jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇮 Variable ratetime 🇮 Timevariable 🇮 Variable being integratedDetailed documentation link
Interpolate a Vec as a function of the given argument. See neml2::Interpolation for rules on shapes of the interpolant and the argument. This object performs a bilinear interpolation.
abscissa1 Scalar defining the abscissa values of the first interpolation axisabscissa2 Scalar defining the abscissa values of the second interpolation axisargument1 🇮 First argument used to query the interpolant along the first axisargument2 🇮 Second argument used to query the interpolant along the second axisdim Intermediate dimension along which to interpolatejit Use JIT compilation for the forward operatorordinate Vec defining the ordinate values of the interpolantoutput 🇴 Vec output of the interpolantproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Perform forward Euler time integration defined as \( s = s_n + (t - t_n) \dot{s} \), where \(s\) is the variable being integrated, \(\dot{s}\) is the variable rate, and \(t\) is time. Subscripts \(n\) denote quantities from the previous time step.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇮 Variable rate of changetime 🇮 Timevariable 🇴 Variable being integratedDetailed documentation link
Calculate the first order discrete time derivative of a variable as \( \dot{f} = \frac{\Delta f}{t-t_n} \), where \( \Delta f \) is the variable with the increment, and \( t \) is time.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇴 The variable's rate of changetime 🇮 Timevariable 🇮 The incremental valueDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Calculate linear combination of multiple Vec tensors as \( u = c_i v_i + s \) (Einstein summation assumed), where \( c_i \) are the coefficients, and \( v_i \) are the variables to be summed. \( s \) is a constant offset.
coefficient_as_parameter By default, the coefficients are declared as buffers. Set this option to true to declare them as (trainable) parameters. This option takes a list of booleans, one for each coefficient. When the length of this list is 1, the boolean applies to all coefficients.coefficients 🇵 Weights associated with each variable. This option takes a list of weights, one for each coefficient. When the length of this list is 1, the same weight applies to all coefficients.constant_coefficient 🇵 The constant coefficient added to the final summationconstant_coefficient_as_parameter By default, the constant_coefficient are declared as buffers. Set this option to true to declare them as (trainable) parameters.from_var Vec tensors to be summedjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to_var 🇴 The sumDetailed documentation link
Interpolate a Vec as a function of the given argument. See neml2::Interpolation for rules on shapes of the interpolant and the argument. This object performs a linear interpolation.
abscissa Scalar defining the abscissa values of the interpolantargument 🇮 Argument used to query the interpolantdim Intermediate dimension along which to interpolatejit Use JIT compilation for the forward operatorordinate Vec defining the ordinate values of the interpolantoutput 🇴 Vec output of the interpolantproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
Calculate the first order discrete time derivative of a variable as \( \dot{f} = \frac{f-f_n}{t-t_n} \), where \( f \) is the force variable, and \( t \) is time.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇴 The variable's rate of changetime 🇮 Timevariable 🇮 The variable to take time derivative withDetailed documentation link
Voce isotropic hardening model, \( h = R \left[ 1 - \exp(-d \bar{\varepsilon}_p) \right] \), where \( R \) is the isotropic hardening upon saturation, and \( d \) is the hardening rate.
equivalent_plastic_strain 🇮 Equivalent plastic strainisotropic_hardening 🇴 Isotropic hardeningjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.saturated_hardening 🇵 Saturated isotropic hardeningsaturation_rate 🇵 Hardening saturation rateDetailed documentation link
Voce hardening for a SingleSlipStrength type model defined by \( \dot{\tau} = \theta_0 \left( 1 - \frac{\tau}{\tau_f} \right) \sum_{i=1}^{n_{slip}} \left| \dot{\gamma}_i \right| \) where \( \theta_0 \) is the initial rate of work hardening, \( \tau_f \) is the saturated, maximum value of the slip system strength, and \( \dot{\gamma}_i \) is the slip rate on each system.
initial_slope 🇵 The initial rate of hardeningjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.saturated_hardening 🇵 The final, saturated value of the slip system strengthslip_hardening 🇮 Name of current values of slip hardeningslip_hardening_rate 🇴 Name of tensor to output the slip system hardening rates intosum_slip_rates 🇮 Name of tensor containing the sum of the slip ratesDetailed documentation link
Calculate the volume-adjusted deformation gradient, i.e. \( F_e = J^{-\frac{1}/{3}} F \), where \( F \) is the pre-adjusted deformation gradient and \( J \) is the total jacobian of the volumetric deformation gradients to be removed.
input 🇮 Input deformation gradientjacobian 🇮 The jacobian that controls the volume adjustment of the input deformation gradientjit Use JIT compilation for the forward operatoroutput 🇴 Output adjusted deformation gradientproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.Detailed documentation link
Define the (cummulative, as opposed to instantaneous) linear isotropic volume expansion eigenstrain, i.e. \( \boldsymbol{\varepsilon}_V = (\frac{V}{V0})^(1/3)-1 \boldsymbol{I} \), where \( V \) is the current volume, and \( V0 \) is the reference (initial) volume.
eigenstrain 🇴 Eigenstrainjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.reference_volume 🇵 Reference (initial) volumevolume 🇮 VolumeDetailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Perform explicit discrete exponential time integration of a rotation. The update can be written as \( s = \exp\left[ (t-t_n)\dot{s}\right] \circ s_n \), where \( \circ \) denotes the rotation operator.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇮 Variable rate of changetime 🇮 Timevariable 🇴 Variable being integratedDetailed documentation link
Define the implicit discrete exponential time integration residual of a rotation variable. The residual can be written as \( r = s - \exp\left[ (t-t_n)\dot{s}\right] \circ s_n \), where \( \circ \) denotes the rotation operator.
automatic_scaling Whether to perform automatic scaling. See neml2::NonlinearSystem::init_scaling for implementation details.automatic_scaling_miter Maximum number of automatic scaling iterations. No error is produced upon reaching the maximum number of scaling iterations, and the scaling matrices obtained at the last iteration are used to scale the nonlinear system.automatic_scaling_tol Tolerance used in iteratively updating the scaling matrices.jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.rate 🇮 Variable ratetime 🇮 Timevariable 🇮 Variable being integratedDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
A parameter that is just a constant value, generally used to refer to a parameter in more than one downstream object.
jit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.value 🇵 The constant value of the parameterDetailed documentation link
Average a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum a dynamic dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
A parameter that is defined through an input variable. This object is not intended to be used directly in the input file.
from 🇮 The input variable that defines this parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the parameter, default to 'parameters/object_name'Detailed documentation link
Average an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Sum an intermediate dimension
dim The dimension over which to perform the reduction.from 🇮 Variable to reducejit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The reduced variableDetailed documentation link
Convert the parameter to state variable.
from 🇵 The input parameterjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.to 🇴 The name of the variables, default to 'state/object_name'Detailed documentation link
Classical macroscale plasticity yield function, \( f = \bar{\sigma} - \sigma_y - h \), where \( \bar{\sigma} \) is the effective stress, \( \sigma_y \) is the yield stress, and \( h \) is the isotropic hardening.
effective_stress 🇮 Effective stressisotropic_hardening 🇮 Isotropic hardeningjit Use JIT compilation for the forward operatorproduction Production mode. This option is used to disable features like function graph tracking and tensor version tracking which are useful for training (i.e., calibrating model parameters) but are not necessary in production runs.yield_function 🇴 Yield functionyield_stress 🇵 Yield stressDetailed documentation link
1.12.0