1 Foundational principles

1.1 First law requirements (energy conservation)

Thermodynamic consistency requires that a model respect conservation of energy. At the macroscopic level, this means the bookkeeping of internal energy, mechanical work, and heat transfer must satisfy a first-law energy balance. In differential form, the rate of change of energy is tied to fluxes of energy and to power supplied by stresses and other generalized forces. For constitutive modeling, this manifests as constitutive relations that do not create or destroy energy unaccountably; for example, the model’s definitions of heat flux and stress power must be compatible with the chosen energy variable.

1.2 Second law requirements (entropy inequality)

The second law constrains the allowed direction of processes by enforcing nondecreasing total entropy for an isolated system, or equivalently a nonnegative entropy production rate. In modeling terms, one constructs an entropy balance in which the divergence of an entropy flux plus an entropy production term yields the net entropy change. Thermodynamic consistency typically means the model guarantees nonnegative entropy production for all physically admissible states and process directions, not only for solutions that satisfy equilibrium.

1.3 Equilibrium thermodynamics and state functions

A consistent equilibrium framework treats key quantities—such as internal energy, entropy, and thermodynamic potentials—as state functions. This ensures that equilibrium relations are integrable and path independent. If the model introduces state variables and constitutive laws inconsistently, it can yield contradictions like cyclic path-dependent “equilibrium” work or entropy changes. In practice, consistency involves defining thermodynamic variables so that equilibrium conditions follow from extremum principles or from derivative relations among potentials.

1.4 Compatibility with conservation laws

Beyond the first and second laws, consistency must also align with other conservation requirements relevant to the modeled system. For continuum theories, this includes conservation of mass, momentum, and possibly species. Constitutive and closure relations must be chosen so that the overall system of equations is compatible: for instance, constitutive choices for stress, heat flux, and diffusion must not violate momentum or species balances, and they must combine with energy/entropy balances to yield a mathematically closed, physically meaningful set of governing equations.

2 Thermodynamic potentials and state variables

2.1 Choice of thermodynamic potential

Thermodynamic potentials reorganize information so that constitutive equations and equilibrium conditions take convenient forms. A model’s consistency depends on using potentials that match the variables one treats as independent, as well as on maintaining the correct derivative relationships.

2.1.1 Internal energy representation

In an internal-energy formulation, internal energy is treated as a function of entropy and extensive variables such as volume (or density) and composition. The model then expresses other thermodynamic quantities through partial derivatives, for example temperature and pressure derived from gradients of internal energy with respect to entropy and volume. Consistency requires that these derivative definitions reproduce the correct equilibrium behavior and that constitutive laws respect the associated entropy and energy balances.

2.1.2 Helmholtz free energy representation

The Helmholtz free energy is useful for systems held at fixed volume and temperature. With entropy eliminated in favor of temperature via a Legendre transform, the constitutive structure often becomes simpler when temperature is prescribed. Thermodynamic consistency requires that the free energy and its derivatives correctly connect stress measures, heat-related relations, and composition-dependent effects, while still ensuring a valid entropy production statement for irreversible processes.

2.1.3 Gibbs free energy representation

The Gibbs free energy is commonly chosen for fixed pressure and temperature. By replacing volume with pressure through a Legendre transform, it provides direct access to chemical potentials and equilibrium composition relations. For multi-component systems, consistency is especially sensitive: chemical potentials defined from the Gibbs potential must enter mass-action-like closures and diffusion laws in a way that preserves entropy inequality and yields correct phase and chemical equilibrium limits.

2.2 Fundamental thermodynamic relations

2.2.1 Maxwell relations and integrability

Thermodynamic potentials imply cross-derivative identities that, when satisfied, guarantee integrability of state-variable relations. These identities—commonly expressed as Maxwell relations—follow from the existence of a potential whose mixed derivatives commute. Inconsistent constitutive modeling can break these conditions, leading to conflicting predictions for responses derived along different thermodynamic paths. A practical way to enforce this is to derive constitutive equations from a potential rather than prescribing independent differential relationships.

2.3 Thermodynamic derivatives and response functions

Response functions (heat capacity, compressibility, susceptibilities) depend on second derivatives of potentials. Thermodynamic consistency therefore implies not only correct first derivatives but also appropriate properties of second derivatives. For example, positive heat capacities and stability against small perturbations often correspond to convexity/concavity conditions on potentials. When models are inconsistent, derived response functions may become unphysical (e.g., negative dissipation-related coefficients or unstable behavior in regimes where stability should hold).

3 Consistent constitutive modeling

3.1 Constitutive relations and closure

A thermodynamic model becomes predictive only when constitutive relations close the governing balances. Consistency means that closure choices are compatible with the chosen thermodynamic framework and do not contradict the energy and entropy balances. This typically involves selecting constitutive variables (such as stress, heat flux, and diffusion fluxes) and expressing them in forms consistent with derivative definitions from potentials or with empirically motivated yet thermodynamically constrained structures.

3.2 Entropy production as a constraint

Nonnegative entropy production can be used as a direct constraint on admissible constitutive forms. The standard approach identifies products of “fluxes” and “forces” within the entropy production rate. By constructing constitutive laws that ensure each contribution does not lead to negative total entropy generation, one obtains inequality-safe models. This constraint can also guide the functional dependence of transport coefficients and coupling terms, yielding restrictions such as sign requirements and bounds that protect the model from violating the second law.

3.3 Onsager reciprocity and near-equilibrium behavior

Near equilibrium, linear irreversible thermodynamics often yields a structure where entropy production is quadratic in thermodynamic forces. Under standard assumptions (microscopic reversibility and appropriate parity under time reversal), the coupling coefficients satisfy Onsager reciprocity relations. Thermodynamic consistency can thus impose symmetry conditions on the phenomenological transport matrix. In practice, models intended for near-equilibrium regimes incorporate these symmetries so that cross-effects (e.g., heat-driven diffusion or stress-driven heat flow) align with the entropy production structure.

3.4 Practical criteria for consistency checks

Consistency checks typically include: (i) verifying that the model’s balance laws reduce to the correct equilibrium relations; (ii) deriving entropy production from the governing equations and confirming it is nonnegative under admissible states; (iii) confirming integrability through potential-based definitions; and (iv) checking that linearized response and transport coefficients satisfy expected stability or positivity conditions. When a model fails a check, the failure often localizes to a particular constitutive assumption—such as an ad hoc flux expression not derived from the same potential structure or a sign choice that can make entropy production negative in some parameter range.

4 Irreversible thermodynamics

4.1 Local equilibrium assumption

Irreversible thermodynamic formulations frequently rely on local equilibrium: each small material element is assumed to be characterized by state variables (like temperature, chemical potentials, and strain) at which equilibrium thermodynamics applies. Under this assumption, entropy is treated as a state function of local fields, enabling entropy balances and constitutive constraints. If local equilibrium is invalid—such as in extremely fast processes or in systems with strong non-equilibrium internal variables—one must enlarge the state space and introduce additional evolution equations while preserving entropy inequality.

4.2 Flux–force frameworks

A common representation expresses entropy production as a sum of products between generalized fluxes and conjugate thermodynamic forces. For example, heat conduction couples heat flux to a gradient of inverse temperature, and diffusion couples mass flux to gradients of chemical potential over temperature. This structure clarifies which constitutive dependencies are needed for dissipation and how coupling terms enter. Thermodynamic consistency is often enforced by ensuring the constitutive relations make these flux–force products yield a nonnegative total entropy production.

4.3 Nonnegative entropy production

Ensuring nonnegative entropy production requires more than choosing positive coefficients in isolation. The full expression may contain coupling between different processes, and the sign of the overall quadratic form must remain nonnegative. A consistent approach expresses entropy production in a form where coefficients form a positive semidefinite operator or matrix. In nonlinear regimes, one likewise verifies that the entropy production remains nonnegative for all admissible rates or gradients.

4.4 Linear vs. nonlinear regimes

In the linear regime, entropy production is typically quadratic and the second law imposes matrix positivity and symmetry (when applicable). In nonlinear regimes, the entropy production structure still constrains allowable functional forms, but reciprocity may not hold in the same way and coefficients may depend on state and magnitude of driving forces. Thermodynamic consistency in nonlinear modeling thus focuses on inequality preservation for arbitrary process directions rather than only on linearized coefficients.

5 Multi-component and multi-phase systems

5.1 Chemical potentials and mixture thermodynamics

For mixtures, chemical potentials play a central role in equilibrium and in non-equilibrium transport. Consistency requires that the chemical potentials used in diffusion laws and phase relations are consistent with the mixture free energy (or Gibbs potential) chosen as the thermodynamic basis. Additionally, constraints such as the Gibbs–Duhem relation link chemical potentials, temperature, and pressure, reducing independent degrees of freedom. A model that treats these relations inconsistently may overcount variables or generate contradictions in predicted compositions.

5.2 Phase equilibria consistency

Phase equilibrium conditions—such as equality of temperature, pressure, and chemical potentials across interfaces between phases—must follow from the model’s thermodynamic potentials. Consistency is tested by ensuring that, in the equilibrium limit, the model’s constitutive equations and constraint equations yield the standard coexistence criteria. For example, if a phase-field or sharp-interface approach is used, its thermodynamic driving forces must be derived so that the predicted equilibrium interface conditions do not conflict with potential-based equilibrium thermodynamics.

5.3 Interfacial thermodynamics and jump conditions

Multi-phase models often involve interfaces where quantities can jump. Interfacial thermodynamics introduces surface free energy and defines interfacial conjugate variables, leading to jump conditions that relate stresses, chemical potential differences, and curvature effects (depending on the framework). Thermodynamic consistency requires that these jump relations are compatible with the overall energy and entropy balances, ensuring that interfacial motion or diffusion across an interface dissipates energy in a manner consistent with the second law.

5.4 Stability and convexity conditions

Stability of mixtures and phases is tied to curvature properties of thermodynamic potentials. Consistent models ensure that the relevant potentials are convex (or concave) in the appropriate variables, yielding positive definiteness of response functions and preventing unphysical spinodal behavior in regimes where stability should persist. In multi-component systems, the stability conditions become matrix inequalities involving Hessians of potentials, and maintaining them across parameter ranges is a key practical aspect of thermodynamic consistency.

6 Continuum mechanics coupling

6.1 Stress–strain and thermodynamic forces

Coupling thermodynamics to mechanics requires identifying which stress measure is work-conjugate to which kinematic variable and ensuring that the mechanical power input appears correctly in the energy balance. Thermodynamically consistent models define stresses as derivatives of an elastic or free-energy density with respect to strain-like variables, or otherwise specify constitutive relations that align with the same energy/entropy structure. This ensures that the model’s mechanical response and dissipative behavior are consistent with the underlying thermodynamic framework.

6.2 Viscosity, heat conduction, and dissipation

Viscous dissipation and heat conduction contribute to entropy production. Consistency requires that constitutive laws for viscosity and thermal conductivity lead to nonnegative dissipation, typically by imposing positivity and, in multi-dimensional settings, tensor positive semidefiniteness. When there are cross-couplings between mechanical and thermal effects (e.g., thermo-viscous interactions), the combined dissipation must still satisfy the entropy inequality, meaning the set of coupled transport coefficients must form a structure compatible with a nonnegative entropy production rate.

6.3 Thermoelastic and thermo-viscoplastic models

Thermoelastic and thermo-viscoplastic theories introduce temperature-dependent elastic energy and internal variables for plasticity or viscoplastic flow. Thermodynamic consistency in such models often requires: (i) a free-energy or energy-storage function for reversible parts; (ii) evolution laws for internal variables that produce dissipation matching entropy production; and (iii) coupling terms so that plastic work is appropriately partitioned into stored energy and heat/dissipation consistent with the first and second laws. This prevents, for instance, constitutive choices where the model could predict negative dissipation during plastic flow.

6.4 Energy balance in moving/deforming media

In moving and deforming continua, the energy balance includes convective transport and work performed by stresses on moving boundaries. Consistency requires that the form of the energy balance aligns with the choice of control volume and with the kinematic description used in the constitutive theory. The entropy balance must then follow from the same underlying derivation. If different pieces—stress power terms, heat flux definitions, or internal energy update rules—are used with inconsistent conventions, the resulting coupled system can violate thermodynamic constraints even if each sub-model appears reasonable in isolation.

7 Mathematical formulation and constraints

7.1 Integrability and potential-based formulations

Mathematically consistent thermodynamics often relies on expressing constitutive relations through potentials. Potential-based formulations make integrability explicit: gradients of a potential yield conjugate variables, and mixed derivatives commute. This provides a systematic route to ensuring Maxwell relations and other thermodynamic identities automatically. When a model instead specifies constitutive relations as independent differential rules, integrability must be checked separately and can easily be violated.

7.2 Convexity/concavity and well-posedness

Convexity or concavity properties of thermodynamic potentials have implications for both physical stability and mathematical well-posedness of the governing equations. For equilibrium thermodynamics, correct curvature ensures positive response functions and prevents instabilities. In dynamic or coupled PDE systems, these curvature properties can influence hyperbolicity, monotonicity, and the sign of certain energy norms. While mathematical criteria vary by theory class, thermodynamic convexity typically supports existence and stability arguments by providing Lyapunov-like structures tied to energy or entropy.

7.3 Invariance and coordinate transformations

Thermodynamic consistency is preserved under proper transformations of variables and coordinates, such as changes from one set of generalized coordinates to another, or transformations between reference and current configurations in continuum mechanics. A consistent model defines state variables and fluxes so that the governing laws maintain their form under these changes, using tensorial transformations where needed. Failure to respect invariance can introduce spurious terms that break energy or entropy balances.

7.4 Boundary and initial condition compatibility

Even a thermodynamically well-constructed constitutive model can yield inconsistent behavior if boundary and initial conditions are incompatible with the entropy inequality or the energy balance. Consistency checks therefore include ensuring that specified boundary data (temperature, chemical potentials, tractions, velocities, or fluxes) do not force the model into states where entropy production becomes negative. Initial conditions must also lie within the thermodynamically admissible set—often the set where potentials are defined and stability inequalities hold.

8 Validation and example workflows

8.1 Deriving governing equations from potentials

A common workflow begins by choosing thermodynamic potentials and identifying state variables. One then computes conjugate variables as derivatives of the potential and formulates balance laws. Constitutive relations are selected to match the reversible parts with potential derivatives and to represent irreversible parts through entropy-consistent dissipation terms. This derivation-first approach makes consistency easier to verify because most constraints are enforced at the level of definitions.

8.2 Verifying entropy production in a model

After assembling the governing equations, one computes the entropy balance and isolates the entropy production term. The remaining step is to show that the entropy production is nonnegative for admissible processes. In practice, this can involve algebraic manipulation to rewrite entropy production as a sum of quadratic forms, checking sign properties of coefficients, and confirming that coupling terms do not lead to indefinite forms. If the inequality cannot be established, the model’s constitutive assumptions or variable choices typically need revision.

8.3 Benchmarking against equilibrium limits

A consistency-oriented validation includes checking equilibrium and near-equilibrium limits. The model should reproduce known equilibrium relations—such as coexistence conditions, chemical potential equalities, and correct response behavior as dissipation vanishes. For irreversible frameworks, one also checks that, as driving forces diminish, the model’s predictions converge to linearized thermodynamic predictions and respect expected reciprocity or transport symmetries where applicable.

8.4 Common pitfalls and how to avoid them

Frequent failure modes include: (i) mixing variable conventions (e.g., using incompatible stress or energy measures); (ii) specifying flux laws that are not consistent with the derived entropy production form; (iii) using ad hoc couplings without verifying positivity of the resulting dissipation; and (iv) violating integrability by prescribing inconsistent derivative relations. Avoidance strategies include deriving as much as possible from potentials, explicitly computing entropy production, enforcing tensor positivity in coupled transport, and performing limit checks against equilibrium thermodynamics and known linear response results.