Power-Law Viscosity Models

• Power-law viscosity models are constitutive frameworks that define non-Newtonian fluids by relating shear rate to local apparent viscosity using algebraic exponents.
• They are applied in calculating shear-thinning and shear-thickening behaviors in systems ranging from polymer solutions and biological tissues to geological flows and plasmas.
• The formulation, based on the Ostwald–de Waele law and tensor generalizations, supports experimental, analytic, and computational studies by addressing stability and regime transitions.

A power-law viscosity model is a constitutive framework for describing non-Newtonian fluids in which the local (apparent) viscosity depends algebraically on the magnitude of the deformation-rate tensor or shear rate. This nonlinearity underlies the behavior of both shear-thinning (pseudoplastic) and shear-thickening (dilatant) fluids and is central to the rheology of many soft matter systems, polymer solutions, dense suspensions, geological flows, biological tissues, and plasma transport. The classical Ostwald–de Waele law and its tensorial generalizations are widely used as the baseline for capturing this dependence in the momentum equations, enabling analytic, experimental, and computational analyses in diverse geometries and flow regimes.

1. Mathematical Formulation and Variants

The standard form of the power-law viscosity constitutive relation (Ostwald–de Waele law) can be written as

$$\boldsymbol T = K\,|\boldsymbol{D}(u)|^{n-1}\,\boldsymbol{D}(u)$$

where $\boldsymbol{T}$ is the extra stress tensor, $$\boldsymbol{D}(u) = \nabla u + (\nabla u)^\mathsf{T}$$ is the rate-of-deformation tensor, $K$ is the consistency index (units of Pa·s$^n$), $n$ is the flow-behavior index, and $|\boldsymbol{D}(u)|$ is the Frobenius norm (second invariant) of $\boldsymbol{D}(u)$. For scalar (simple shear) flows, the relation reduces to

$$\tau = K\,\dot\gamma^{\,n} \qquad \mu_\text{app}(\dot\gamma) = K\,\dot\gamma^{\,n-1}$$

where $\tau$ is the shear stress and $\boldsymbol{T}$0 is the shear rate.

• $\boldsymbol{T}$1 describes shear-thinning (e.g., polymer solutions, mud, ice)
• $\boldsymbol{T}$2 describes shear-thickening (e.g., certain colloidal suspensions, some polymer melts)
• $\boldsymbol{T}$3 recovers the Newtonian case

In tensor form, the invertibility of the relation rests on $\boldsymbol{T}$4. For $\boldsymbol{T}$5, one may employ the Glen–Nye formulation by inverting the stress–strain constitutive law, avoiding the divergence at vanishing strain rate (Noble et al., 2012).

Intrinsic limitations of the classical, unregularized model arise for $\boldsymbol{T}$6 where well-posedness is lost in free-surface or boundary-layer flows, necessitating regularization when modeling singular or near-static regions (Noble et al., 2012, Helanow et al., 2023).

2. Physical Regimes and Non-Newtonian Behavior

Power-law viscosity models capture the ubiquity of shear-rate-dependent viscous responses in a wide range of materials:

• Shear-thinning (pseudoplasticity): fluids whose apparent viscosity decreases as the local shear rate increases, typical for aqueous polymer solutions, blood, gels, and engineered slurries (Sochi, 2014, Daripa et al., 2023). The microstructural origin often involves coil compression (not extension) in polymers (Dunstan, 2018), non-affine network rearrangements in gels (Aime et al., 2018), or the presence of power-law velocity distributions in plasmas (Wang et al., 2018).
• Shear-thickening: fluids with an increasing viscosity with increased shear rate, found in dense colloidal suspensions or certain polymer melts (Noble et al., 2012, Jadhao et al., 2019).
• Transition regimes: Real fluids deviate from pure power-law scaling at very low or high shear rates, where plateaus are observed (described by generalized Carreau or Carreau–Yasuda models) (Shahsavari et al., 2015, Gao et al., 2023, Jadhao et al., 2019). For elastohydrodynamic lubrication, the viscosity scaling crosses from Carreau (power-law) at low pressure/temperature to Eyring (thermally activated) at high stress (Jadhao et al., 2019).

Power-law exponents, consistency indices, and critical transition rates must often be empirically determined or tabulated, possibly as functions of composition, temperature, pressure, or concentration (Daripa et al., 2023, Shahsavari et al., 2015, Dunstan, 2018).

3. Theoretical and Scaling Foundations

A range of first-principles and phenomenological theories underpin power-law viscosity models:

• Microscale statistical origins: In colloidal gels and soft matter, fractional Maxwell models—constructed from fractional derivatives ("springpot")—capture power-law viscoelastic response from a broad spectrum of relaxation times (Bonfanti et al., 2020, Aime et al., 2018).
• Polymeric solutions: Scaling arguments reveal that in concentrated solutions, the viscosity $\boldsymbol{T}$7 scales as the 9th power of the coil radius $\boldsymbol{T}$8 ($\boldsymbol{T}$9), and thus as a power-law in shear rate: $$\boldsymbol{D}(u) = \nabla u + (\nabla u)^\mathsf{T}$$0, $$\boldsymbol{D}(u) = \nabla u + (\nabla u)^\mathsf{T}$$1 (Dunstan, 2018).
• Free-volume transport: In dense hard-sphere fluids, the macroscopic viscosity obeys a power-law in geometric or thermodynamic free-volume: $$\boldsymbol{D}(u) = \nabla u + (\nabla u)^\mathsf{T}$$2, revealing entropy scaling and resolving discrepancies in empirical exponential ("Cohen–Turnbull–Doolittle") or mode-coupling forms (Liu, 2020).
• Kinetic models for gases and plasmas: Chapman–Enskog expansion yields $$\boldsymbol{D}(u) = \nabla u + (\nabla u)^\mathsf{T}$$3 for inverse-power-law molecular interactions, with $$\boldsymbol{D}(u) = \nabla u + (\nabla u)^\mathsf{T}$$4 set by the intermolecular potential exponent; non-Maxwellian velocity-statistics (e.g., Tsallis-q) in plasmas lead to transport coefficients that are inverse power-laws of the statistical parameter (Hu et al., 2020, Wang et al., 2018).

Detailed analysis of instability mechanisms, such as stick–slip in molecular sliding friction (Prandtl model), directly generates power-law-like detachment and stress–rate relations (Gao et al., 2023).

4. Flows, Modeling Methodologies, and Regime Transitions

Power-law models have been applied in numerous flow situations and model hierarchies:

• Thin-film and shallow-water flows: Derivation of asymptotically consistent Benney–shallow-water systems for thin power-law films down inclines shows that the streamwise momentum diffusion and non-Newtonian terms are critical to correctly predict "roll wave" formation and the linear stability threshold. For $$\boldsymbol{D}(u) = \nabla u + (\nabla u)^\mathsf{T}$$5, the classical formulations become ill-posed unless singularity removal strategies or regularization are implemented (Noble et al., 2012).
• Porous and fibrous media: Effective Darcy-like mobility functions and pore-scale network models for power-law fluids have been developed, demonstrating that at high local shear rates (low porosity or high velocity), the macroscopic mobility becomes independent of the Carreau number and exhibits pure power-law scaling (Shahsavari et al., 2015, Sochi, 2014).
• Boundary-layer analysis: Generalizations of the Blasius problem for 2D boundary layers with power-law viscosity reveal that wall-shear and boundary-layer thickness vary systematically with $$\boldsymbol{D}(u) = \nabla u + (\nabla u)^\mathsf{T}$$6, and Töpfer-style non-iterative transformation algorithms can be employed for similarity solutions (Fazio, 2020).

Transition criteria between Newtonian and power-law regimes are often codified via dimensionless groups, e.g., the Carreau number or rescaled Carreau number incorporating the effective pore-scale shear rate and geometric parameters (Shahsavari et al., 2015), or an Ohnesorge number in the context of extensional vs. shear rheology (Matsumoto et al., 2024).

5. Experimental Validation and Multiaxial Rheology

Experimental investigations confirm and refine the power-law framework across deformation modalities:

• Shear vs. extensional viscosity: Recent experimental studies using capillary breakup (CaBER-DoS) and shear rheometry demonstrate that for low-viscosity shear-thinning fluids with Ohnesorge number $$\boldsymbol{D}(u) = \nabla u + (\nabla u)^\mathsf{T}$$7, extensional and shear viscosity follow power laws with identical exponents, aligning with Carreau predictions. This enables inference of extensional rheology from shear data in jetting, printing, and coatings (Matsumoto et al., 2024).
• Parameter identification: Exponents $$\boldsymbol{D}(u) = \nabla u + (\nabla u)^\mathsf{T}$$8 in power-law models extracted from direct rheometry, filament-thinning, or oscillatory datasets have typical values in the range $$\boldsymbol{D}(u) = \nabla u + (\nabla u)^\mathsf{T}$$9 for shear-thinning industrial fluids, while fractional exponents $K$0 in viscoelastic (FM) models are often $K$1 for gels and cells (Bonfanti et al., 2020, Aime et al., 2018).
• Limitations: At very low velocities or high concentrations, regularization or alternative constitutive forms (Carreau, Cross, Eyring) become necessary as the simple power-law expression breaks down.

6. Computational Implementation and Numerical Stability

Power-law viscosity models pose unique challenges and opportunities for numerical modeling:

• Nonlinear solvers and preconditioners: For steady or unsteady flows, especially when $K$2, the viscosity may diverge at zero strain rates, causing extreme ill-conditioning in discretized systems. Regularization (e.g., via $K$3) is typically applied to avoid singularities (Helanow et al., 2023). Viscosity-scaled mass-matrix Schur preconditioners exhibit robust eigenvalue clustering and mesh-independence, essential for Newton–Krylov solvers in, e.g., glaciological flow simulations (Helanow et al., 2023).
• Wall-modeled turbulence: LES models explicitly incorporating power-law viscosity in wall-stress closures are necessary to capture drag reduction and mean velocity profiles in turbulent flows of non-Newtonian fluids, as conventional Newtonian wall-stress models fail in the presence of strong shear-thinning (Taghvaei et al., 11 Nov 2025).
• Hybrid schemes: Domain-decomposition and discontinuous finite element methods, combined with pointwise updates of empirical $K$4 and $K$5, enable robust simulation of spatially and temporally evolving rheology, as in polymer flood EOR processes (Daripa et al., 2023).

Grid resolution, regularization parameter choice, and the implementation of accurate subgrid models are vital for ensuring numerical stability and fidelity to the true power-law viscous response, particularly in complex multi-phase or multiphysics flows.

7. Relation to Generalized and Fractional Models

Power-law viscosity models are subsumed or complemented by generalized viscoelastic and fractional frameworks:

• Carreau, Cross, and Eyring models: These provide interpolating forms for viscosity as a function of shear rate, seamlessly capturing both the Newtonian plateaus and power-law thinning regimes. Carreau–Yasuda models with an additional cross-over exponent $K$6 allow highly accurate fits to molecular-dynamics data and are justified microscopically by the distribution of instability thresholds (Gao et al., 2023, Jadhao et al., 2019).
• Fractional viscoelasticity: Replacement of integer-order constitutive elements with fractional derivatives ("springpots") generates both the observed power-law rheology and the broad relaxation spectra in gels, biological tissues, and amorphous solids (Bonfanti et al., 2020, Aime et al., 2018).
• Microscopic instability models: The Prandtl–Tomlinson sliding instability delivers power-law frictional and viscous scaling directly, with the exponent set by the specifics of the underlying mechanical instability (Gao et al., 2023).

The scope of power-law viscosity models thus extends from strictly viscous to complex viscoelastic and microstructurally disordered materials, with broad applicability underpinned by unified mathematical structures and physical principles.