Fractional Burgers Model (Fractional)¶
Quick Reference¶
Use when: Complex creep with glassy compliance, fractional retardation, and viscous flow
Parameters: 5 (\(J_g, J_k, \alpha, \tau_k, \eta_1\))
Key equation: \(J(t) = J_g + \frac{t^{\alpha}}{\eta_1\Gamma(1+\alpha)} + J_k[1 - E_{\alpha}(-(t/\tau_k)^{\alpha})]\)
Test modes: Relaxation, creep, oscillation
Material examples: Polymer composites, asphalt binders, bituminous materials, viscoelastic solids under load
Fractional Calculus Fundamentals
This model uses fractional calculus for power-law viscoelastic behavior. For mathematical foundations—SpringPot element, Mittag-Leffler functions, physical meaning of fractional order \(\alpha\), and derivation from molecular theory—see:
/user_guide/fractional_viscoelasticity_reference
Notation Guide¶
Symbol |
Units |
Description |
|---|---|---|
\(J_g\) |
1/Pa |
Glassy compliance (instantaneous elastic response) |
\(\eta_1\) |
Pa·s |
Viscosity of Maxwell dashpot (controls terminal flow) |
\(J_k\) |
1/Pa |
Kelvin compliance magnitude (retardation amplitude) |
\(\alpha\) |
dimensionless |
Fractional order (0 < \(\alpha\) < 1, controls power-law character) |
\(\tau_k\) |
s |
Retardation time (characteristic Kelvin timescale) |
\(E_{\alpha}(z)\) |
dimensionless |
One-parameter Mittag-Leffler function |
\(\Gamma(z)\) |
dimensionless |
Gamma function |
Overview¶
The Fractional Burgers Model combines a Maxwell element in series with a Fractional Kelvin-Voigt element, creating a five-parameter model that captures glassy compliance, viscous flow, and fractional retardation in a single compact framework. This model extends the classical four-element Burgers model by replacing the Kelvin-Voigt dashpot with a SpringPot, enabling power-law retardation instead of exponential relaxation.
The Fractional Burgers model is particularly effective for materials exhibiting complex creep behavior with both instantaneous elastic response, delayed fractional retardation, and long-term viscous flow. Common applications include polymer composites under load, asphalt binders, bituminous materials, and viscoelastic solids undergoing time-dependent deformation.
Mechanical Analogue:
[Maxwell Arm: Spring Gg + Dashpot η1] ---- series ---- [Fractional KV: Spring + SpringPot (Jk, α, τk)]
Physical Foundations¶
The Fractional Burgers model combines three distinct mechanical responses:
Instantaneous elastic response (glassy compliance \(J_g\))
Fractional retardation (SpringPot in Kelvin arm with time constant \(\tau_k\))
Long-term viscous flow (dashpot viscosity \(\eta_1\))
Microstructural Interpretation:
\(J_g\): Instantaneous bond stretching, glassy modulus
Fractional KV arm: Distributed retardation from hierarchical polymer network rearrangements
Maxwell dashpot: Irreversible chain flow, reptation, or permanent deformation
Governing Equations¶
Time Domain (Creep Compliance):
where \(E_{\alpha}(z)\) is the one-parameter Mittag-Leffler function:
Frequency Domain (Complex Compliance):
Complex Modulus:
Note: The inversion \(G^* = 1/J^*\) is exact for linear viscoelastic materials.
Parameters¶
Name |
Symbol |
Units |
Bounds |
Notes |
|---|---|---|---|---|
|
\(J_g\) |
1/Pa |
[1e-9, 1e3] |
Glassy compliance (instantaneous response) |
|
\(\eta_1\) |
Pa·s |
[1e-6, 1e12] |
Viscosity (Maxwell arm, controls terminal flow) |
|
\(J_k\) |
1/Pa |
[1e-9, 1e3] |
Kelvin compliance (retardation magnitude) |
|
\(\alpha\) |
dimensionless |
[0.05, 0.95] |
Fractional order (0.2-0.7 typical for polymers) |
|
\(\tau_k\) |
s |
[1e-6, 1e6] |
Retardation time (characteristic Kelvin timescale) |
Regimes and Behavior¶
Short Time (\(t \ll \tau_k\)):
Instantaneous glassy compliance plus early-time fractional flow from Maxwell arm.
Intermediate Time (\(t \sim \tau_k\)):
Fractional retardation dominated by Kelvin arm with power-law approach to equilibrium.
Long Time (\(t \gg \tau_k\)):
Unbounded creep (liquid-like) with constant compliance offset from glassy and Kelvin contributions.
Validity and Assumptions¶
Linear viscoelasticity: Strain amplitudes remain small (< 5-10% typically)
Isothermal conditions: Temperature constant throughout measurement
Time-invariant material: No aging, degradation, or structural evolution
Supported test modes: Creep (primary), oscillation
Fractional order bounds: 0.05 < \(\alpha\) < 0.95 for numerical stability
Liquid-like behavior: Unbounded creep at long times (\(\eta_1\) finite)
What You Can Learn¶
This section explains how to translate fitted Fractional Burgers parameters into material insights and actionable knowledge.
Parameter Interpretation¶
- Glassy Compliance ( \(J_g\) ):
The instantaneous elastic response upon stress application.
For graduate students: \(J_g\) reflects short-range bond stretching and angle deformation in the glassy state. For polymers, \(J_g \approx 1/G_\infty\) where \(G_\infty\) is the glassy modulus (~1 GPa for many polymers).
For practitioners: \(J_g\) sets the immediate strain upon loading. Critical for impact resistance and short-time deformation.
- Kelvin Compliance ( \(J_k\) ):
Controls the magnitude of delayed (retarded) elastic deformation.
Retardation magnitude: \(\Delta J = J_k\)
For polymers, relates to chain rearrangements in constrained environments
Typical values: \(10^{-6}\) to \(10^{-2}\) Pa\(^{-1}\)
- Fractional Order ( \(\alpha\) ):
Governs the breadth of the retardation spectrum and power-law character.
\(\alpha \approx 0.2\)–0.3: Very broad spectrum, highly heterogeneous (filled systems)
\(\alpha \approx 0.4\)–0.5: Moderate breadth, typical for polymer composites
\(\alpha \approx 0.6\)–0.7: Narrower spectrum, more uniform structure
\(\alpha \to 1\): Exponential retardation (classical Burgers)
Physical interpretation: Lower \(\alpha\) indicates greater polydispersity in relaxation times arising from structural heterogeneity, filler distribution, or molecular weight distribution.
- Viscosity ( \(\eta_1\) ):
Controls the rate of unbounded creep at long times.
Slope of J(t) at long times: dJ/dt = 1/\(\eta_1\)
For polymers, relates to molecular weight via \(\eta_1 \sim M_w^{3.4}\) (reptation)
Determines processability and long-term dimensional stability
- Retardation Time ( \(\tau_k\) ):
Characteristic timescale for the fractional Kelvin-Voigt relaxation.
Marks the transition from glassy to retardation-dominated regime
Temperature-dependent: follows WLF or Arrhenius behavior
Material Classification¶
Parameter Pattern |
Material Type |
Examples |
Key Characteristics |
|---|---|---|---|
High \(J_k/J_g\) (> 10) |
Soft viscoelastic solid |
Polymer composites, filled elastomers |
Large delayed compliance |
Low \(\alpha\) (< 0.3) |
Highly heterogeneous |
Asphalt, bitumen, nanocomposites |
Very broad spectrum |
High \(\eta_1 (> 10^6\) Pa·s) |
High MW polymer |
Melts, concentrated solutions |
Slow terminal flow |
Low \(\eta_1 (< 10^3\) Pa·s) |
Low MW or diluted |
Modified bitumen, soft materials |
Rapid creep |
Diagnostic Indicators¶
\(J_g \approx 0\) or poorly constrained: Insufficient early-time data; use faster sampling or estimate from high-frequency \(G'\)
Linear \(J(t)\) at all times: No retardation; use simple Maxwell liquid instead
\(\alpha\) near bounds (0.05 or 0.95): Data may not support fractional behavior; try classical Burgers (\(\alpha\) = 1)
Strong \(J_k\) - \(\tau_k\) correlation: Need better data coverage in intermediate regime
Application Examples¶
- Asphalt Pavement Design:
Use Burgers model to predict rutting under sustained traffic load. The terminal flow (\(\eta_1\)) determines permanent deformation rate, while \(J_k\) and \(\alpha\) control elastic recovery.
- Polymer Composite Selection:
Compare \(J_k\) values between formulations. Lower \(J_k\) means better dimensional stability under load. Monitor \(\alpha\) for filler dispersion quality.
- Food Texture Analysis:
Fit creep data from cheese or dough. High \(J_k\) indicates soft, easily deformable texture. Use \(\alpha\) to quantify structural heterogeneity.
Fitting Guidance¶
Recommended Data Collection:
Creep test (primary): 4-5 decades in time (e.g., 0.1 s - \(10^4\) s)
Sampling: Log-spaced, minimum 50 points per decade
Stress level: Within LVR, verify with amplitude sweep
Temperature control: ±0.1°C for polymers, ±0.5°C for bitumen
Initialization Strategy:
# From creep data J(t)
Jg_init = J(t_min) # Instantaneous compliance
eta1_init = t / (J(t) - J(t_min)) at long time # Terminal slope
Jk_init = (J(t_mid) - Jg_init) * 0.5 # Mid-range magnitude
tau_k_init = t where retardation is 50% complete
alpha_init = 0.5 # Default starting point
Optimization Tips:
Fit in log(compliance) space for better conditioning
Use weighted least squares with log-spaced weights
Constrain \(J_g < J_k\) (glassy stiffer than Kelvin)
Verify residuals are random, not systematic
Common Pitfalls:
Overfitting: Don’t fit Burgers if classical 4-element model suffices
Underfitting: If residuals show curvature, may need additional Kelvin element
Wrong regime: Ensure data captures all three regimes (glassy, retardation, flow)
Usage¶
from rheojax.models import FractionalBurgersModel
from rheojax.core.data import RheoData
import numpy as np
# Create model
model = FractionalBurgersModel()
# Fit to experimental creep data
t_exp = np.logspace(-1, 4, 100) # 0.1 s to 10,000 s
J_exp = load_creep_data() # Load your data
# Automatic fit
model.fit(t_exp, J_exp, test_mode='creep')
# Inspect fitted parameters
print(f"Jg = {model.parameters.get_value('Jg'):.2e} Pa⁻¹")
print(f"Jk = {model.parameters.get_value('Jk'):.2e} Pa⁻¹")
print(f"α = {model.parameters.get_value('alpha'):.3f}")
print(f"τk = {model.parameters.get_value('tau_k'):.2e} s")
print(f"η₁ = {model.parameters.get_value('eta1'):.2e} Pa·s")
# Predict creep at new times
t_new = np.logspace(-2, 5, 200)
data = RheoData(x=t_new, y=np.zeros_like(t_new), domain='time')
data.metadata['test_mode'] = 'creep'
J_pred = model.predict(data)
# Bayesian uncertainty quantification
result = model.fit_bayesian(
t_exp, J_exp,
num_warmup=1000,
num_samples=2000,
test_mode='creep'
)
intervals = model.get_credible_intervals(result.posterior_samples, credibility=0.95)
See Also¶
Generalized Fractional Maxwell (Two-Order) — generalized two-SpringPot formulation
Fractional Kelvin-Voigt (Fractional) — Kelvin arm used inside Burgers
Mastercurve (Time-Temperature Superposition) — build broadband spectra for better fitting
FFTAnalysis — convert relaxation to frequency domain
../../examples/advanced/04-fractional-models-deep-dive — notebook comparing Burgers family
Material Examples¶
Polymer Composites (\(J_g \approx 10^{-6}-10^{-5}\) 1/Pa, \(\alpha \approx 0.3-0.5\)):
Filled elastomers (carbon black, silica fillers)
Fiber-reinforced polymers under sustained load
Polymer nanocomposites (clay, CNT fillers)
Asphalt and Bitumen (\(\eta_1 \approx 10^4-10^7\) Pa·s, \(\alpha \approx 0.4-0.6\)):
Asphalt concrete (temperature-dependent)
Bituminous binders for road pavements
Roofing materials
Food Materials (\(J_k \approx 10^{-4}-10^{-2}\), \(\alpha \approx 0.2-0.5\)):
Cheese (long-term creep)
Dough (wheat flour, viscoelastic retardation)
Semi-solid fats (margarine, butter)
Biological Tissues (\(\alpha \approx 0.2-0.4\)):
Ligaments and tendons under sustained stress
Intervertebral discs (viscoelastic creep)
Experimental Design¶
Creep Test (Primary Application):
Step stress: Apply constant stress \(\sigma_0\) within LVR
Time span: Cover 4-5 decades (e.g., 0.1 s - \(10^4\) s)
Sampling: Log-spaced to capture all three regimes
Analysis: Fit \(J(t)\) to identify \(J_g\) (instantaneous), \(J_k\) (retardation), \(\eta_1\) (slope at long time)
Frequency Sweep (Oscillatory):
Frequency range: 0.001-100 rad/s (wide span critical)
Strain amplitude: Within LVR (0.5-5%)
Analysis: Fit \(G'(\omega)\), \(G''(\omega)\) simultaneously
Verification: Check terminal flow region (\(G'' \sim \omega\), \(G' \sim \omega^2\))
Fitting Strategies¶
Initialization from Creep Data:
\(J_g\): Extrapolate \(J(t \to 0)\) (instantaneous compliance)
\(\eta_1\): Slope of \(J(t)\) at long time \(\to 1/\eta_1\)
\(J_k\): Mid-time plateau height minus \(J_g\)
\(\tau_k\): Time where retardation is half-complete
\(\alpha\): Curvature of retardation region in log-log plot
Optimization:
Use weighted least squares (log-spaced weights)
Constrain \(J_g < J_k\) (glassy stiffer than Kelvin)
Fit in compliance space for creep, modulus space for oscillation
Verify residuals random across all time/frequency decades
Usage Example¶
from rheojax.models import FractionalBurgersModel
import numpy as np
# Create model
model = FractionalBurgersModel()
# Set typical parameters for polymer composite
model.parameters.set_value('Jg', 1e-6) # 1/Pa
model.parameters.set_value('eta1', 1e5) # Pa·s
model.parameters.set_value('Jk', 5e-6) # 1/Pa
model.parameters.set_value('alpha', 0.4) # dimensionless
model.parameters.set_value('tau_k', 10.0) # s
# Predict creep compliance
t = np.logspace(-1, 4, 100)
J_t = model.predict(t, test_mode='creep')
# Fit to experimental creep data
# t_exp, J_exp = load_creep_data()
# model.fit(t_exp, J_exp, test_mode='creep')
Limiting Behavior¶
\(\alpha \to 1\): Classical Burgers with exponential Kelvin retardation
\(J_k \to 0\): Maxwell + fractional flow only (no retardation)
\(\eta_1 \to \infty\): Fractional Kelvin-Voigt (bounded creep, no flow)
\(\tau_k \to 0\): Instantaneous Kelvin response, \(J(t) = J_g + J_k + t/\eta_1\)
\(\tau_k \to \infty\): Kelvin arm inactive, simple Maxwell
Model Comparison¶
Burgers vs Fractional Burgers:
Classical Burgers: Exponential retardation (\(\alpha = 1\))
Fractional Burgers: Power-law retardation (0 < \(\alpha\) < 1)
Use Fractional when creep shows curved transition in log-log plots
Burgers vs Fractional Maxwell Gel:
Burgers: 5 parameters, includes delayed elasticity (Kelvin arm)
FMG: 3 parameters, single relaxation mode
Use Burgers for complex creep with multiple timescales
Troubleshooting¶
Issue: Cannot identify \(J_g\) from data
Cause: Insufficient early-time resolution
Solution: Use faster sampling or estimate from high-frequency modulus
Issue: Oscillatory fit poor at low frequencies
Cause: Terminal flow region not captured
Solution: Extend frequency sweep to lower \(\omega\) (< 0.01 rad/s)
Issue: Parameter correlation (\(J_k\) and \(\tau_k\))
Cause: Insufficient data in retardation regime
Solution: Focus measurements on intermediate timescale (\(t \sim \tau_k\))
Tips & Best Practices¶
Fit creep first: Compliance space more natural for Burgers model
Verify terminal flow: Confirm linear \(J(t)\) vs \(t\) at long time
Check bounds: Ensure \(J_g < J_k\) (physically meaningful)
Use transforms: Apply FFTAnalysis to convert creep → oscillation
Log-log plots: Visualize all three regimes clearly
References¶
See Also¶
Metzler, R., Schick, W., Kilian, H.-G., & Nonnenmacher, T. F. “Relaxation in filled polymers: A fractional calculus approach.” Journal of Chemical Physics, 103, 7180-7186 (1995). https://doi.org/10.1063/1.470346
Friedrich, C. “Relaxation and retardation functions of the Maxwell model with fractional derivatives.” Rheologica Acta, 30, 151-158 (1991). https://doi.org/10.1007/BF01134604
Heymans, N. & Bauwens, J. C. “Fractal rheological models and fractional differential equations for viscoelastic behavior.” Rheologica Acta, 33, 210-219 (1994). https://doi.org/10.1007/BF00437306
Nonnenmacher, T. F. & Glöckle, W. G. “A fractional model for mechanical stress relaxation.” Philosophical Magazine Letters, 64, 89-93 (1991). https://doi.org/10.1080/09500839108214672
Podlubny, I. Fractional Differential Equations. Academic Press (1999). ISBN: 978-0125588409
Generalized Fractional Maxwell (Two-Order) — generalized two-SpringPot formulation
Fractional Kelvin-Voigt (Fractional) — Kelvin arm used inside Burgers
Mastercurve (Time-Temperature Superposition) — build broadband spectra for better fitting
FFTAnalysis — convert relaxation to frequency domain
../../examples/advanced/04-fractional-models-deep-dive — notebook comparing Burgers family