Friction With Bristle Dynamics (FrBD) Framework

• Friction-With-Bristle-Dynamics (FrBD) is a physically motivated framework uniting nonlinear friction laws with internal bristle rheologies to connect microscopic deformations and macroscopic friction.
• It employs advanced viscoelastic models, including Generalized Maxwell and Kelvin-Voigt elements, ensuring properties like boundedness and passivity for robust control applications.
• The framework supports both lumped and distributed formulations, validated against classical models, and is applicable to scenarios such as rolling contacts and tire dynamics.

The Friction-With-Bristle-Dynamics (FrBD) framework constitutes a comprehensive class of physically motivated, control-oriented friction models that integrate nonlinear rate-and-state-dependent friction laws with internal bristle-like rheologies. Originally motivated by the need to bridge microscopic frictional mechanisms with macroscopic observations, FrBD encompasses a broad variety of lumped and distributed descriptions, ranging from classical LuGre-type ordinary differential equations (ODEs) to advanced hyperbolic and parabolic partial differential equation (PDE) systems modeling spatially distributed friction phenomena, including rolling contact, directionality, viscoelastic relaxation, and control passivity. The framework enables the embedding of highly general viscoelastic microconstitutive laws (e.g., generalized Maxwell and Kelvin-Voigt branches) and nonlinear velocity-dependent friction coefficients, yielding models that are provably bounded, passive, and well-suited for applications in robotics, tire dynamics, and tribological systems.

1. Microscopic-Macroscopic Coupling and Internal States

At the core of FrBD is the hypothesis that macroscopic frictional forces emerge from the collective deformation and relaxation of microscopic “bristles”—idealized asperities or hairs mediating contact between solid surfaces. Let $v$ denote the rigid-body sliding velocity and $z$ the representative bristle deflection; the bristle tip velocity is then $v_s = v - \dot{z}$, where $\dot{z}$ is the rate of bristle deformation. The macroscopic friction force is generated by the product $F_f(t) = p\,f(v_s(t))$, where $p$ is the normal load and $f(v_s)$ is a nonlinear friction function, often reflecting a Stribeck-type or more complex empirical law. The bristle force $f$ can include contributions from viscoelastic elements governed by internal ODEs or PDEs, with additional internal variables representing branch forces or deformations in multi-element rheologies. This formulation enables direct coupling between bristle microstates and the nonlinear frictional response (Romano et al., 20 Jan 2026, Romano et al., 10 Feb 2026).

2. Rheological Representations: GM and GKV Elements

The FrBD framework supports highly general viscoelastic bristle models. Two canonical forms are the Generalized Maxwell (GM) and Generalized Kelvin-Voigt (GKV) elements:

• GM (n+1-branch):

$$\begin{cases} f(t) = \bar{k}_0\,z(t) + \sum_{i=1}^n f_i(t), \ \dot{f}_i(t) = -\tau_i^{-1} f_i(t) + \bar{k}_i\,\dot{z}(t), \quad i=1,\ldots,n. \end{cases}$$

• GKV (n+1-branch):

$$\begin{cases} f(t) = \bar{k}_0\,\Big[z(t) - \sum_{i=1}^n z_i(t)\Big], \ f(t) = \bar{k}_i\,z_i(t) + \bar{c}_i\,\dot{z}_i(t), \quad i=1,\ldots,n. \end{cases}$$

The nonlinear bristle evolution is then given by a closed-loop system such as:

$z$0

with $z$1 defined by the above viscoelastic equations. This systematically connects the micro- and macro-dynamics and subsumes first-order LuGre-type models as special cases (Romano et al., 20 Jan 2026, Romano et al., 10 Feb 2026).

3. Well-Posedness, Boundedness, and Passivity

FrBD models are rigorously analyzed in terms of key mathematical properties:

• Boundedness: The adoption of quadratic storage functions for the internal states ensures that, for any bounded input velocity $z$2, all internal variables and macroscopic friction forces remain uniformly bounded.
• Passivity: The models are provably passive, i.e., the total energy dissipated satisfies an integral inequality of the form:

$z$3

where $z$4 is an explicit storage function (e.g., a quadratic form in $z$5). Passivity holds for all physically meaningful parametrizations (e.g., $z$6; $z$7; $z$8), without requiring ad hoc damping or restrictive matrix conditions (Romano et al., 20 Jan 2026, Romano, 20 Jan 2026).

4. Distributed and Multi-Dimensional FrBD Models

FrBD encompasses distributed and multidimensional frictional settings indispensable for modeling spatially-extended contacts such as tire-road interaction and rolling bearings:

• Distributed FrBD: The internal states become spatial fields $z$9 over the contact region $v_s = v - \dot{z}$0, and the evolution is governed by hyperbolic (and in some cases parabolic) PDEs:

  $v_s = v - \dot{z}$1

  where $v_s = v - \dot{z}$2 is a vector of distributed bristle states, $v_s = v - \dot{z}$3 is the transport velocity, and $v_s = v - \dot{z}$4 encodes (possibly spin-dependent) rigid-relative velocities. Models can be classified as linear or semilinear depending on the coupling between spin and internal states (Romano, 11 Jan 2026, Romano, 20 Jan 2026, Romano, 3 Mar 2026).

• Steady-State and Transient Behavior: The distributed formulation admits direct computation of steady-state force–slip (“action”) surfaces and captures nontrivial transient phenomena such as spatial relaxation lengths, overshoots, and hysteresis, even under time-varying normal loads or patch geometries.
• Well-posedness and Stability: For bounded spatial domains and mild regularity on pressure and velocity profiles, existence and uniqueness of (mild/classical) solutions, input-to-state stability (ISS), and input-to-output stability (IOS) are established. Passivity is generalized to the $v_s = v - \dot{z}$5 setting, with appropriate quadratic storage functionals (Romano, 20 Jan 2026, Romano, 3 Mar 2026).

5. Connection to Rate-and-State, Directionality, and Emergent Friction Phenomena

FrBD generalizes and unifies several previously distinct classes of friction models:

• LuGre and Rate-and-State Laws: Classical LuGre models arise as the special case of an elastic bristle (single spring, no micro-damping) with closed-form inversion of the friction law. FrBD extends this by systematically embedding arbitrary viscoelasticity and nonlinear friction characteristics (Romano et al., 10 Feb 2026, Romano et al., 20 Jan 2026).
• Directional Friction and Nap Effects: Asymmetry and directionality in friction, including with-the-nap/against-the-nap effects, emerge from geometric and energetic factors in bristle–substrate interaction, as rigorously proved in homogenization limits of extended Prandtl-Tomlinson models. Limiting Coulomb friction coefficients factor as $v_s = v - \dot{z}$6, where $v_s = v - \dot{z}$7 is the normal load and $v_s = v - \dot{z}$8 is a geometric function of bristle angle and substrate profile; bristle angle and surface asymmetry both contribute to directional dry friction (Gidoni et al., 2016). Motion-direction inversion in vibration-driven bristle-bots is likewise quantitatively predicted by FrBD asymptotics in terms of geometric and inertial parameters (Cicconofri et al., 2017, Cicconofri et al., 2014).
• Sliding vs Rolling Contact: FrBD provides a unified framework for both sliding and rolling contacts; the governing PDE structure and nonlinear bristle laws are the same modulo the transport velocities of each body through the contact patch. Both processes manifest multi-scale viscoelastic relaxation captured within the same system class (Romano, 8 Jun 2026).

6. Applications in Control Design and Engineering Systems

FrBD models are directly suited for observer and controller synthesis in robotic and automotive systems:

• Friction-Observer Design: Bristle deflection and internal viscoelastic forces are estimated via reduced-order observers based on the same ODE or PDE structure as the plant friction dynamics. Passivity guarantees stability of the observation error dynamics (Romano et al., 20 Jan 2026).
• Sliding-Mode and Passivity-Based Control: Joint mechanical systems, tire models, and soft-contact robotic manipulators have been successfully controlled using FrBD-based passive feedback, exploiting built-in dissipation and strict output-passivity to guarantee closed-loop asymptotic stability (Romano et al., 20 Jan 2026, Romano et al., 11 Jan 2026).
• Advanced Simulation: FrBD-based tire and rolling-contact models (including string-tyre and flexible-carcass variants) reproduce memory effects, micro-shimmy, and transient frictional behavior under real steering maneuvers; these models admit fast numerical solution exploiting sparse discretizations and implicit integration schemes (Romano, 3 Mar 2026, Romano et al., 11 Jan 2026).

7. Experimental and Numerical Validation

Quantitative validation of FrBD includes:

• Benchmarking against Classical Models: Lumped and distributed FrBD models have been rigorously compared against LuGre/Dahl and classical brush/tyre models on standard tribological tests, including pre-sliding displacement, frictional lag, and stick–slip cycles. FrBD is found to match or slightly outperform traditional models, with enhanced ability to tune micro-damping and relaxation (Romano et al., 10 Feb 2026, Romano et al., 11 Jan 2026).
• Experimental Realizations: Bristle-bot locomotion, rolling bearing tests, and diaphragm valve actuation have been modeled and validated with FrBD, showing that theoretical predictions (e.g., inversion frequency, average speed) match measured outcomes within experimental tolerances (Cicconofri et al., 2017, Cicconofri et al., 2014, Romano et al., 10 Feb 2026).

---

References

• (Romano et al., 20 Jan 2026): Linear viscoelastic rheological FrBD models
• (Romano, 11 Jan 2026): Two-dimensional FrBD friction models for rolling contact
• (Romano et al., 10 Feb 2026): First-order friction models with bristle dynamics: lumped and distributed formulations
• (Romano, 20 Jan 2026): Two-dimensional FrBD friction models for rolling contact: extension to linear viscoelasticity
• (Romano, 3 Mar 2026): Bare and stretched string tyre models with distributed FrBD dynamics
• (Romano, 8 Jun 2026): Dynamic sliding and rolling friction models for linear viscoelastic contact pairs
• (Romano et al., 11 Jan 2026): Semilinear single-track vehicle models with distributed tyre friction dynamics
• (Cicconofri et al., 2017, Cicconofri et al., 2014): Bristle-bot analytical and experimental analyses
• (Gidoni et al., 2016): Genesis of directional friction through bristle-like mediating elements