Fractional Viscoelasticity: Mathematical Reference

Note

This is the definitive reference for fractional calculus concepts in RheoJAX. All fractional model pages link here instead of duplicating this content.

Overview

Fractional calculus generalizes differentiation and integration to non-integer orders, providing a powerful mathematical framework for describing complex viscoelastic behavior that cannot be captured by classical integer-order models. In rheology, fractional derivatives enable the modeling of power-law relaxation, broad relaxation spectra, and self-similar microstructures using fewer parameters than multi-mode classical models.

Why Fractional Calculus in Rheology?

Most real materials exhibit viscoelastic behavior that deviates from simple exponential relaxation:

Experimental observations:

  • Power-law relaxation \(G(t) \sim t^{-\alpha}\) over multiple time decades

  • Broad relaxation spectra arising from structural heterogeneity

  • Frequency-dependent moduli with parallel slopes in log-log plots

  • Non-exponential creep that cannot be fit with single relaxation times

Classical model limitations:

  • Single relaxation time \(\tau\) (Maxwell, Zener) insufficient for complex materials

  • Multi-mode models require many parameters (5-20+) with limited physical insight

  • Exponential functions cannot capture power-law dynamics

Fractional model advantages:

  • Capture power-law behavior naturally with 3-5 parameters

  • Fractional order \(\alpha\) has clear physical meaning (spectrum breadth, microstructure)

  • Fewer parameters than multi-mode models while maintaining accuracy

  • Interpolate smoothly between elastic (\(\alpha=0\)) and viscous (\(\alpha=1\)) extremes

SpringPot Element

The SpringPot (Scott-Blair element) is the fundamental building block of fractional rheology, generalizing both elastic springs and viscous dashpots into a single element.

Mathematical Definition

The SpringPot constitutive equation relates stress and strain through a fractional derivative:

\[\sigma(t) = E_0 \, D^\alpha \gamma(t)\]
where:

  • \(E_0\): quasi-property with units Pa·s \(^\alpha\)

  • \(D^\alpha\): fractional derivative of order \(\alpha \in [0, 1]\)

  • \(\gamma(t)\): strain as a function of time

  • \(\sigma(t)\): stress as a function of time

Limiting Cases

The SpringPot smoothly interpolates between classical elements:

SpringPot Limiting Behavior

\(\alpha\) Value

Element Type

Constitutive Equation

\(\alpha = 0\)

Pure elastic spring

\(\sigma = E_0 \gamma\) (Hooke’s law)

\(0 < \alpha < 1\)

Fractional viscoelastic

\(\sigma = E_0 D^\alpha \gamma\) (intermediate behavior)

\(\alpha = 1\)

Pure viscous dashpot

\(\sigma = E_0 \, d\gamma/dt\) (Newton’s law)

Frequency-Domain Representation

In oscillatory shear (frequency domain), the SpringPot impedance is:

\[Z(\omega) = E_0 (i\omega)^\alpha = E_0 \omega^\alpha \left[\cos\left(\frac{\alpha\pi}{2}\right) + i\sin\left(\frac{\alpha\pi}{2}\right)\right]\]

This reveals that the SpringPot simultaneously contributes to both storage and loss moduli with a constant phase angle:

\[\delta = \frac{\alpha\pi}{2}\]

where \(\delta\) is the loss angle (phase shift between stress and strain).

Physical interpretation:

  • \(\alpha = 0\): \(\delta = 0^\circ\) (purely elastic, no phase shift)

  • \(\alpha = 0.5\): \(\delta = 45^\circ\) (balanced viscoelasticity)

  • \(\alpha = 1\): \(\delta = 90^\circ\) (purely viscous, maximum phase shift)

The storage and loss moduli contributions scale as:

\[\begin{split}G'(\omega) &\sim \omega^\alpha \cos(\alpha\pi/2) \\ G''(\omega) &\sim \omega^\alpha \sin(\alpha\pi/2)\end{split}\]

Key insight: Both moduli have parallel slopes of \(\alpha\) in log-log plots, which is the hallmark signature of fractional viscoelasticity.

Mittag-Leffler Functions

Mittag-Leffler functions play the same role in fractional viscoelasticity as exponential functions do in classical models. They provide the exact analytical solutions for fractional differential equations governing viscoelastic constitutive relations.

One-Parameter Mittag-Leffler Function

The one-parameter Mittag-Leffler function is defined by the infinite series:

\[E_\alpha(z) = \sum_{k=0}^{\infty} \frac{z^k}{\Gamma(\alpha k + 1)}\]

where \(\Gamma\) is the gamma function (generalization of factorial to real numbers).

Key Properties:

  1. Recovers exponential: \(E_1(z) = \exp(z)\) (classical limit)

  2. Initial value: \(E_\alpha(0) = 1\) for all \(\alpha > 0\)

  3. Asymptotic behavior:

    • Short times: \(E_\alpha(-t^\alpha) \approx 1 - t^\alpha/\Gamma(\alpha+1)\)

    • Intermediate times: \(E_\alpha(-t^\alpha) \sim t^{-\alpha}\) (power-law decay)

    • Long times: \(E_\alpha(-t^\alpha) \sim \exp(-A \, t^{\alpha/(1-\alpha)})\) (stretched exponential)

  4. Interpolation: Smoothly interpolates between exponential (\(\alpha=1\)) and power-law (\(0<\alpha<1\))

Physical Meaning in Relaxation:

The relaxation modulus for fractional models typically has the form:

\[G(t) = G_0 \, E_\alpha\left(-\left(\frac{t}{\tau_\alpha}\right)^\alpha\right)\]

This captures:

  • Initial plateau: \(G(0) = G_0\) (elastic response)

  • Power-law relaxation: \(G(t) \sim G_0 (t/\tau_\alpha)^{-\alpha}\) at intermediate times

  • Broad relaxation spectrum: Continuous distribution of relaxation times

  • Characteristic time \(\tau_\alpha\): Time scale for onset of power-law decay

Two-Parameter Mittag-Leffler Function

The two-parameter generalization adds a second parameter \(\beta\):

\[E_{\alpha,\beta}(z) = \sum_{k=0}^{\infty} \frac{z^k}{\Gamma(\alpha k + \beta)}\]

Key Properties:

  1. Reduces to one-parameter: \(E_{\alpha,1}(z) = E_\alpha(z)\)

  2. Initial value: \(E_{\alpha,\beta}(0) = 1/\Gamma(\beta)\)

  3. More flexible asymptotics: Controls short-time behavior via \(\beta\)

Applications in Fractional Models:

  • Creep compliance: \(J(t)\) often involves \(E_{\alpha,1+\alpha}(-t^\alpha)\)

  • Complex constitutive equations: Fractional Maxwell Liquid uses \(E_{1-\alpha,1-\alpha}\)

  • General viscoelasticity: Provides exact solutions for arbitrary fractional orders

Computational Note

RheoJAX computes Mittag-Leffler functions using the mittag_leffler module (rheojax.utils.mittag_leffler), which implements:

  • One-parameter: E_alpha(z, alpha) via series expansion + asymptotic approximations

  • Two-parameter: E_alpha_beta(z, alpha, beta) via series expansion

These functions are JAX-compatible and GPU-accelerated for fast evaluation in optimization and Bayesian inference.

Physical Meaning of Fractional Order α

The fractional order \(\alpha\) is not an arbitrary fitting parameter – it has deep physical significance related to material microstructure and relaxation dynamics.

1. Relaxation Spectrum Width

The fractional order \(\alpha\) quantifies the breadth of the relaxation time distribution:

Spectrum Breadth Interpretation

\(\alpha\) Value

Spectrum Type

Physical Meaning

\(\alpha = 1\)

Narrow (Dirac delta)

Single relaxation time (classical exponential)

\(0.7 < \alpha < 1\)

Moderate breadth

Few dominant relaxation processes

\(0.3 < \alpha < 0.7\)

Broad distribution

Continuous spectrum over many decades

\(\alpha \to 0\)

Very broad (power-law)

Hierarchical or fractal structure, no characteristic time

Mathematical connection:

For fractional models, the relaxation time spectrum \(H(\tau)\) is approximately:

\[H(\tau) \sim \tau^{-(1-\alpha)} \quad \text{for } \tau_{\text{min}} < \tau < \tau_{\text{max}}\]
where:

  • Narrow spectrum (\(\alpha \to 1\)): \(H(\tau) \to \delta(\tau - \tau_0)\) (Dirac delta)

  • Broad spectrum (\(\alpha \approx 0.5\)): \(H(\tau) \sim \tau^{-0.5}\) (power-law distribution)

2. Microstructural Heterogeneity

Lower \(\alpha\) values indicate greater structural heterogeneity at the molecular/microscopic level:

For cross-linked networks (e.g., elastomers, hydrogels):

  • \(\alpha < 0.5\): Hierarchical structure with multiple length scales

    • Broad cross-link density distribution

    • Polydisperse mesh sizes

    • Fractal or self-similar network architecture

  • \(\alpha \approx 0.5\): Critical gel-like behavior

    • Sol-gel transition point

    • Percolation threshold

    • Maximum structural disorder

  • \(\alpha > 0.5\): More homogeneous networks

    • Narrow cross-link density distribution

    • Approaching regular lattice structure

For polymer melts:

  • \(\alpha < 0.5\): Broad molecular weight distribution (polydispersity)

    • Significant chain length heterogeneity

    • Branched or star polymers

    • Complex intermolecular interactions

  • \(\alpha \approx 0.7\text{--}0.9\): Relatively monodisperse linear polymers

    • Narrow molecular weight distribution

    • Simple chain dynamics (reptation)

3. Material Character (Solid vs. Liquid vs. Gel)

The fractional order \(\alpha\) influences the dominant viscoelastic character:

Material Character Classification

\(\alpha\) Range

Dominant Character

Typical Materials

\(\alpha < 0.3\)

Strong solid-like

Stiff gels, covalently cross-linked elastomers, biological tissues

\(0.3 < \alpha < 0.5\)

Solid-like viscoelastic

Soft gels, filled polymers, weak networks

\(\alpha \approx 0.5\)

Critical gel (balanced)

Gel point, percolation threshold, \(G' \approx G''\) across all \(\omega\)

\(0.5 < \alpha < 0.7\)

Liquid-like viscoelastic

Concentrated polymer solutions, weak gels

\(\alpha > 0.7\)

Strong liquid-like

Polymer melts, dilute solutions, approaching classical Maxwell

Oscillatory shear signature:

  • \(\alpha < 0.5\): \(G'(\omega) > G''(\omega)\) at low frequencies (elastic dominance)

  • \(\alpha \approx 0.5\): \(G'(\omega) \approx G''(\omega)\) across all frequencies (critical gel)

  • \(\alpha > 0.5\): \(G''(\omega) > G'(\omega)\) at low frequencies (viscous dominance)

4. Typical α Ranges by Material Class

Extensive experimental studies have established typical fractional order ranges for common materials:

Fractional Order Ranges by Material

Material Class

Typical \(\alpha\)

Notes

Cross-linked polymer networks

0.3 - 0.6

Natural rubber, synthetic elastomers, cured epoxies

Filled elastomers

0.2 - 0.5

Carbon black or silica-filled rubber; lower \(\alpha\) due to filler-polymer interactions

Hydrogels (chemical)

0.4 - 0.7

Covalently cross-linked PVA, alginate, PAA

Hydrogels (physical)

0.3 - 0.5

Non-covalent cross-links, weaker structure

Biological tissues (soft)

0.1 - 0.4

Skin, tendons, cartilage; very broad spectra from hierarchical collagen/elastin

Biological tissues (stiff)

0.3 - 0.5

Bone, dentin, cornea

Semi-crystalline polymers

0.3 - 0.5

Polyethylene, polypropylene; crystalline vs amorphous phase relaxation

Polymer melts (linear)

0.7 - 0.9

Linear homopolymers; approaching classical Maxwell behavior

Polymer melts (branched)

0.5 - 0.7

Long-chain branched polymers, star polymers

Concentrated polymer solutions

0.5 - 0.8

Above overlap concentration \(c^*\)

Emulsions

0.4 - 0.7

Droplet size polydispersity and interfacial dynamics

Colloidal gels

0.2 - 0.4

Particle network with weak attractive interactions

Critical gels

0.45 - 0.55

Sol-gel transition, gelation point

Physical Interpretation Summary

Key takeaway: The fractional order \(\alpha\) is a structural fingerprint that encodes:

  1. How broad the relaxation spectrum is (spectrum width)

  2. How heterogeneous the microstructure is (structural disorder)

  3. Whether the material is solid-like or liquid-like (material character)

  4. What physical processes dominate relaxation (molecular vs network dynamics)

Lower \(\alpha\) values indicate:

  • Broader relaxation spectra

  • More heterogeneous microstructure

  • More solid-like character

  • Hierarchical or fractal organization

Higher \(\alpha\) values indicate:

  • Narrower relaxation spectra

  • More homogeneous microstructure

  • More liquid-like character

  • Approaching classical exponential behavior

Fractional Models in RheoJAX

RheoJAX implements 11 fractional models organized into families based on their mechanical analogs:

Fractional Maxwell Family (4 models):

Fractional Zener Family (4 models):

Advanced Fractional Models (3 models):

See Models Handbook for detailed model documentation.

Key References

Foundational Theory:

  • Mainardi, F. (2010). Fractional Calculus and Waves in Linear Viscoelasticity. Imperial College Press. ISBN: 978-1-84816-329-4

    The definitive reference on fractional calculus in viscoelasticity.

  • Schiessel, H., Metzler, R., Blumen, A., Nonnenmacher, T.F. (1995). “Generalized viscoelastic models: their fractional equations with solutions.” J. Phys. A 28, 6567–6584. https://doi.org/10.1088/0305-4470/28/23/012

    Original derivation of fractional viscoelastic models.

  • Gorenflo, R., Kilbas, A.A., Mainardi, F., Rogosin, S.V. (2014). Mittag-Leffler Functions, Related Topics and Applications. Springer. https://doi.org/10.1007/978-3-662-43930-2

    Comprehensive treatment of Mittag-Leffler functions.

Physical Interpretation:

  • Mainardi, F., Spada, G. (2011). “Creep, Relaxation and Viscosity Properties for Basic Fractional Models in Rheology.” European Physical Journal Special Topics, 193, 133-160. https://doi.org/10.1140/epjst/e2011-01387-1

    Physical meaning of fractional parameters in rheology.

  • Friedrich, C., Braun, H. (1992). “Generalized Cole-Cole Behavior and its Rheological Relevance.” Rheologica Acta, 31, 309-322. https://doi.org/10.1007/BF00418328

    Connection between fractional order and relaxation spectrum width.

Applications:

  • Koeller, R.C. (1984). “Applications of fractional calculus to the theory of viscoelasticity.” J. Appl. Mech. 51, 299–307. https://doi.org/10.1115/1.3167616

    Early application of fractional calculus to viscoelasticity.

  • Metzler, R., Klafter, J. (2000). “The Random Walk’s Guide to Anomalous Diffusion: A Fractional Dynamics Approach.” Physics Reports, 339(1), 1-77. https://doi.org/10.1016/S0370-1573(00)00070-3

    Broader context: fractional dynamics in physics.

Further Reading

Within RheoJAX Documentation:

  • /user_guide/model_selection — Decision flowcharts for choosing fractional vs classical models

  • Architecture Overview — Template Method pattern for smart initialization

  • /examples/advanced/04-fractional-models-deep-dive — Jupyter notebook with case studies

External Resources:

  • Podlubny, I. (1999). Fractional Differential Equations. Academic Press. ISBN: 978-0-12-558840-9

  • Hilfer, R. (Ed.) (2000). Applications of Fractional Calculus in Physics. World Scientific. ISBN: 978-981-02-3457-7

  • Tarasov, V.E. (2010). Fractional Dynamics: Applications of Fractional Calculus to Dynamics of Particles, Fields and Media. Springer. https://doi.org/10.1007/978-3-642-14003-7

See Also

  • Models Handbook — Complete model catalog with governing equations

  • /user_guide/core_concepts — RheoData, parameters, and test modes

  • /user_guide/modular_api — Direct model API usage

  • /user_guide/bayesian_inference — Bayesian inference for fractional models