VLB — What You Can Learn¶

This guide explains how to extract physical insights from VLB model parameters, how to interpret fitting results across protocols, and how to use the model for material design and quality control.

For notation and governing equations, see VLB Transient Network Models.

Parameter Interpretation¶

Network Modulus \(G_0\)¶

The network modulus encodes the density of elastically active chains:

\[G_0 = c k_B T\]

where \(c\) is the number density of active chains and \(k_B T\) is the thermal energy (\(\approx 4.1 \times 10^{-21}\) J at room temperature).

Calculating chain density:

\[c = \frac{G_0}{k_B T} \approx \frac{G_0}{4.1 \times 10^{-21}} \text{ chains/m}^3\]

Typical \(G_0\) by Material¶
Material	\(G_0\) (Pa)	Physical Interpretation
Dilute hydrogels	10 - 100	Low cross-link density; soft, swollen networks
PVA-borax gels	100 - 1000	Moderate density, dynamic boronates
Telechelic polymers	1000 - 10⁴	Concentrated end-functionalized chains
Vitrimers	10⁵ - 10⁶	High cross-link density, stiff network
Elastomers (comparison)	10⁵ - 10⁷	Permanent networks (VLB limit \(k_d \to 0\))

Dissociation Rate \(k_d\)¶

The dissociation rate controls the network lifetime and relaxation:

Kinetic Regimes¶
\(k_d\) (1/s)	Bond Lifetime	Regime
\(< 10^{-4}\)	> hours	Quasi-permanent. Effectively elastic on experimental timescales. Vitrimer-like.
\(10^{-4}\) — \(10^{-1}\)	seconds to hours	Slow exchange. Self-healing with long recovery times. Thermo-activated exchange.
\(10^{-1}\) — \(10^1\)	0.1 - 10 s	Dynamic gel. Active bond exchange. Typical associating polymers.
\(10^1\) — \(10^3\)	ms	Fast exchange. Liquid-like terminal behavior. Concentrated micelles.
\(> 10^3\)	< ms	Ultra-fast exchange. Near-Newtonian. Dilute associating solutions.

Derived Quantities¶

Zero-shear viscosity \(\eta_0 = G_0/k_d\):

Sensitive to both parameters. An increase in \(G_0\) or decrease in \(k_d\) raises viscosity. Compare with rotational viscometry for consistency.

Crossover frequency \(\omega_c = k_d\):

The SAOS crossover gives a direct read of \(k_d\) without fitting — simply identify the frequency where \(G' = G''\).

Weissenberg number \(\text{Wi} = \dot{\gamma}/k_d\):

Quantifies relative importance of deformation vs relaxation. \(\text{Wi} > 1\) means the material is being deformed faster than it can relax.

Pressure (normal force data):

In incompressible materials the pressure \(p\) is a Lagrange multiplier determined by boundary conditions. For confined geometries (parallel plates, cone-and-plate), the thrust force \(F_N\) is related to the first normal stress difference:

\[F_N = \frac{\pi R^2}{2} N_1 = \frac{\pi R^2}{2} G_0(\mu_{xx} - \mu_{yy})\]

where \(R\) is the plate radius. If normal force data is available, this provides an independent check on \(G_0\) and the steady-state Weissenberg number.

Diagnostic Signatures¶

Single-Exponential Relaxation¶

Expected: \(\ln G(t)\) is linear with slope \(-k_d\).

If deviation:

Upward curvature (slower-than-exponential at long times): residual permanent network (\(G_e > 0\)), use VLBMultiNetwork
Downward curvature (faster-than-exponential): multiple relaxation times, use VLBMultiNetwork with M > 1
Power-law tail: not Maxwell-like; consider fractional models (FMG, FZSS)

Newtonian Flow Curve¶

Expected: \(\sigma \propto \dot{\gamma}\) (constant viscosity).

If deviation:

Shear thinning: \(k_d\) increases with stress → need force-dependent \(k_d\) (Bell model, see VLB Advanced Theory & Numerical Methods)
Shear thickening: formation-enhanced kinetics (stretch-creation model, TNT family)
Yield stress: permanent network component or DMT/Fluidity model needed

SAOS Semicircle¶

Expected: Cole-Cole plot (\(G'\) vs \(G''\)) is a semicircle.

If deviation:

Depressed semicircle: broadened relaxation spectrum, use VLBMultiNetwork
Multiple arcs: well-separated relaxation modes
High-frequency uptick: solvent contribution (\(\eta_s > 0\))

Monotonic Startup¶

Expected: \(\sigma_{12}(t)\) monotonically increases to steady state.

If stress overshoot observed:

Constant-\(k_d\) VLB cannot produce overshoot
Indicates force-dependent breakage (Bell model)
Or structure-dependent kinetics (DMT model)

Multi-Network Spectrum Analysis¶

When fitting VLBMultiNetwork, the relaxation spectrum \(\{(G_I, k_{d,I})\}\) reveals network structure:

Spectrum Interpretation¶

Spectrum Feature	Physical Interpretation
Single dominant mode	Narrow bond lifetime distribution; well-defined network
Two well-separated modes	Two distinct bond types (e.g., hydrogen bonds + ionic cross-links)
Broad spectrum (many modes)	Polydisperse bond lifetimes; heterogeneous network
\(G_e > 0\)	Permanent cross-links alongside reversible ones
Large \(\eta_s\)	Significant un-networked polymer contribution

Mode Assignment Strategy¶

Sort modes by \(k_d\): fastest \(k_d\) = most labile bonds
Compare \(G_I\): largest \(G_I\) = most abundant bond population
Cross-reference with chemistry: match timescales to known bond kinetics (e.g., boronate esters: \(k_d \sim 0.1\) s^-1; H-bonds: \(k_d \sim 10^3\) s^-1)

Application Examples¶

Hydrogel Design¶

Goal: Tune self-healing rate and modulus.

Approach:

Fit SAOS data with VLBLocal to extract \(G_0\) and \(k_d\)
\(G_0\) controls stiffness → adjust cross-linker concentration
\(k_d\) controls healing time → modify cross-linker chemistry (e.g., catechol-metal: slow, boronate: moderate, host-guest: fast)

Quality metric: Self-healing efficiency \(\propto k_d \cdot t_{heal}\) → higher \(k_d\) means faster healing but lower toughness (trade-off).

Vitrimer Characterization¶

Goal: Determine exchange kinetics from rheology.

Approach:

Perform stress relaxation at multiple temperatures
Fit VLBLocal at each \(T\) to extract \(k_d(T)\)
Plot \(\ln k_d\) vs \(1/T\) → Arrhenius activation energy

\[k_d(T) = k_d^0 \exp\!\left(-\frac{E_a}{k_B T}\right)\]

The activation energy \(E_a\) characterizes the bond exchange mechanism.

Telechelic Network Diagnostics¶

Goal: Distinguish loop fraction from bridge fraction.

Approach:

Fit SAOS with VLBMultiNetwork (2 modes)
Faster mode → loop relaxation (non-load-bearing)
Slower mode → bridge relaxation (load-bearing)
\(G_{bridge}/G_{total}\) estimates the bridge fraction

This is complementary to TNTLoopBridge, which models loop-bridge kinetics explicitly.

Batch Quality Control¶

Goal: Detect batch-to-batch variations in cross-link density.

Approach:

Establish baseline \(G_0^{ref}, k_d^{ref}\) from a reference batch
Fit each new batch with VLBLocal
Flag deviations:
- \(G_0/G_0^{ref} < 0.9\): under-crosslinked
- \(G_0/G_0^{ref} > 1.1\): over-crosslinked
- \(k_d/k_d^{ref} > 1.5\): accelerated degradation
- \(k_d/k_d^{ref} < 0.5\): kinetic trapping

Cross-Protocol Validation Workflow¶

A robust characterization uses multiple protocols to validate the model:

Step 1: SAOS (primary)

Extract \(G_0, k_d\) from crossover
Verify Cole-Cole semicircle

Step 2: Stress relaxation (validation)

Verify \(G(0) = G_0\) from SAOS
Verify exponential decay with slope \(-k_d\)

Step 3: Startup shear (validation)

Verify \(\sigma^{ss} = \eta_0 \dot{\gamma}\) matches flow curve
No stress overshoot (constant \(k_d\))

Step 4: Creep (optional)

Verify \(J(0) = 1/G_0\)
Verify \(dJ/dt = 1/\eta_0\)

Consistency check:

\[\eta_{SAOS} = \lim_{\omega \to 0} \frac{G''}{\omega} \stackrel{?}{=} \eta_{flow} = \frac{\sigma}{\dot{\gamma}} \stackrel{?}{=} \eta_{creep} = \frac{1}{G_0 \cdot dJ/dt}\]

If all three agree, the constant-\(k_d\) VLB model is appropriate. Discrepancies indicate rate-dependent structure or non-Maxwell behavior.

When VLB Is Not Enough¶

The constant-\(k_d\) VLB model has clear limitations. Here is how to recognize them and which extension to consider:

Observation	VLB Prediction	Reality	Next Step
Shear thinning	Newtonian	\(\eta \propto \dot{\gamma}^{n-1}\)	Force-dependent \(k_d\) (Bell)
Stress overshoot	Monotonic	Overshoot at high Wi	Bell \(k_d\) or DMT
LAOS harmonics in \(\sigma_{12}\)	\(I_3/I_1 = 0\)	\(I_3/I_1 > 0\)	Nonlinear \(k_d\) or FENE
Extensional hardening	Singularity at \(\dot{\varepsilon} = k_d/2\)	Bounded growth	Langevin finite extensibility
Aging	Time-independent	Properties change at rest	DMT or Fluidity-Saramito
Power-law relaxation	Single exponential	\(G(t) \propto t^{-\alpha}\)	Fractional models (FMG, FZSS)
Shear banding	Homogeneous	Banded profiles	`VLBNonlocal`
Vitrimer BER kinetics	N/A (no evolving natural state)	Associative exchange, topology rearrangement	HVM (Hybrid Vitrimer Model) (HVM)
NP-filled vitrimer	N/A (no filler effects)	Payne effect, dual freezing temperatures	HVNM (Hybrid Vitrimer Nanocomposite Model) (HVNM)