Decentralized Dissipative Force Model

• Decentralized dissipative force model is a framework where local interactions and coarse-graining yield effective friction without imposing an external global damping law.
• It encompasses diverse formulations—from atomistic DPD mapping and continuum viscoelastic impacts to CTP-based coarse-graining—that extract dissipation from microscopic dynamics.
• The approach enhances simulation fidelity and clarifies debates by showing how effective dissipation arises intrinsically, resolving misconceptions about equilibrium contributions.

The decentralized dissipative force model denotes a class of descriptions in which dissipation is not inserted as a single global friction law, but emerges from locally defined interactions, hidden degrees of freedom, or distributed continuum stresses. In the cited literature, this idea appears in several technically distinct settings: atomistically grounded dissipative particle dynamics (DPD), reduced effective dynamics in the Closed Time Path formalism, viscoelastic contact mechanics, pairwise DPD thermostats, active Brownian swarms with DPD-like interactions, and open graph networks with bulk and boundary driving. Across these formulations, the common structural feature is that the dissipative term is derived from local exchange, local pairwise coupling, or coarse-graining over unresolved modes rather than imposed from an external constitutive prescription (Sokhan et al., 2018, Polonyi, 2015, Goldobin et al., 2015, Altamirano et al., 2016, Renger et al., 2022, Lobaskin et al., 2013).

1. Core definition and conceptual scope

Across the cited works, “decentralized” refers to a shared structural property rather than a single standardized formalism. The dissipative force is generated by the system’s own local structure: by atoms entering and leaving an invariant volume element in a host fluid, by integrating out hidden coordinates in a closed system, by local viscous stresses in the bulk of a viscoelastic material, by short-ranged pairwise DPD interactions, or by bulk and boundary fluxes on a finite graph. In each case, there is no need for a single privileged damping agent or an externally inserted global friction kernel (Sokhan et al., 2018, Polonyi, 2015).

This notion is clearest when contrasted with phenomenological damping laws such as $F=-\gamma \dot x$. In the atomistic-to-DPD construction, the drag coefficient is obtained as a direct statistical average over microscopic force–velocity correlations. In the CTP treatment, nonconservative forces arise from couplings between doubled histories $(x^+,x^-)$. In viscoelastic impact, the dissipative force is the contact integral of a locally solved stress field. In standard DPD, dissipation is embedded directly in short-ranged pairwise interactions rather than in a system-wide thermostat. In open networks, the force field splits into a dissipative symmetric part and a non-dissipative antisymmetric part, with bulk and boundary contributions treated on the same footing (Sokhan et al., 2018, Polonyi, 2015, Goldobin et al., 2015, Altamirano et al., 2016, Renger et al., 2022).

A recurring implication is that dissipation need not be associated with explicit microscopic friction. One line of work shows that even a closed, non-interacting system can generate dissipative effective forces once one focuses on a reduced set of variables, provided infinitely many soft normal modes accumulate near $\omega=0$. Another line shows that an ideal gas can exhibit dissipative effective dynamics for collective or composite observables. A further line demonstrates that in DPD the random and dissipative terms are dynamically essential for temperature control while contributing negligibly to long-time equilibrium pressure (Polonyi, 2015, Altamirano et al., 2016).

2. Atomistic derivation in dissipative particle dynamics

A central molecular realization of the decentralized dissipative force model is given by the mapping of DPD into classical molecular dynamics through the Brownian quasiparticle (BQ) (Sokhan et al., 2018). The BQ is defined as “all atoms contained at a particular moment of time in an invariant volume element of the fluid” with fixed size $a$. Because atoms enter and leave the volume, the BQ is an open system and its mass fluctuates. Its velocity is the centroid velocity of the atoms currently inside it. This replaces the picture of an abstract mesoscopic bead immersed in a solvent by a coarse-grained object built from the solvent itself.

The local character of the construction is geometric. When BQ volumes overlap, atoms are reassigned by a Voronoi tessellation so that each atom belongs to the nearest sphere. This enforces single membership and momentum conservation. The bead–bead interaction is then not postulated in advance; it emerges from the partitioning of the host atomistic system. The total bead force is obtained by summing atomic pair forces across the two BQ volumes,

${\bf f}_{ij}(r_{ij}) = \sum_{\substack{n\in \Omega_i,\m\in \Omega_j}{\bf f}_{nm}(r_{nm})},$

and the bead velocity is written as

$${\bf u}_{i}(t) = \sum_{n\in \Omega_i}{\bf v}_{n}(t).$$

The resulting DPD force is decomposed into conservative, dissipative, and stochastic parts,

$${\bf f}_{ij}(r_{ij}) = {\bf f}_{ij}^{({\rm c})}(r_{ij}) + {\bf f}_{ij}^{({\rm d})}(r_{ij}) + {\bf f}_{ij}^{({\rm s})}(r_{ij}),$$

with

$${\bf f}_{ij}^{({\rm c})}(r_{ij}) = \alpha(r_{ij}){\bf e}_{ij},$$

$${\bf f}_{ij}^{({\rm d})}(r_{ij}) = -\gamma(r_{ij})u_{ij}{\bf e}_{ij},$$

$${\bf f}_{ij}^{({\rm s})}(r_{ij}) = \sqrt{2k_{\rm B}T\gamma(r_{ij})/\Delta t}\,\theta_{ij}{\bf e}_{ij}.$$

For the conservative interaction, the potential is obtained by Boltzmann inversion of the bead radial distribution function,

$(x^+,x^-)$0

and in the appendix the fit is

$(x^+,x^-)$1

leading to

$(x^+,x^-)$2

The dissipative term is extracted by time-reversal symmetry. Since the drag term is the only one odd under velocity reversal,

$(x^+,x^-)$3

so that

$(x^+,x^-)$4

This is the most explicit atomistic statement of the decentralized force model: $(x^+,x^-)$5 is a local conditional average over pair forces and relative velocities, not a parameter fitted from macroscopic transport. The fitted form later used is

$(x^+,x^-)$6

The method is validated on a supercritical Lennard-Jones fluid. The extracted conservative interactions reproduce the radial distribution functions of the BQ subsystem, and the DPD model built from the fitted $(x^+,x^-)$7 and $(x^+,x^-)$8 gives good agreement in structure. For dynamics, the BQ velocity autocorrelation function has a zero slope at the origin and a hydrodynamic long-time tail, whereas standard DPD decays exponentially. The reported agreement in radial distribution functions, temperatures, and qualitative transport trends supports the DPD parameterization while also showing the limits of the Markovian DPD approximation for long-time hydrodynamics (Sokhan et al., 2018).

3. Emergent dissipation from coarse-graining and hidden modes

A more formal realization is given by the CTP treatment of irreversibility and decoherence in an ideal gas (Polonyi, 2015). The central result is that dissipative effective forces can arise internally in a closed, non-interacting system when only a reduced set of variables is observed. The physical mechanism is the mixing of infinitely many soft normal modes. Each microscopic mode is reversible, but when the spectrum contains infinitely many low-frequency modes accumulating near $(x^+,x^-)$9, finite-time observation cannot resolve them all, and the observed subsystem experiences a gradual leakage of energy into unresolved modes.

In the harmonic bath model, the bath is characterized by the spectral density

$\omega=0$0

and the effective force on the observed coordinate is generated by the bath self-energy. The low-frequency expansion produces a frequency shift, a viscous force proportional to $\omega=0$1, and higher-derivative corrections. The Newtonian friction term is

$\omega=0$2

so low-frequency spectral weight directly controls dissipation.

The formalism uses doubled variables $\omega=0$3 and the action

$\omega=0$4

with the final-time condition

$\omega=0$5

The decomposition

$\omega=0$6

separates the physical coordinate from the difference coordinate. After integrating out the environment, the effective action becomes

$\omega=0$7

or equivalently

$\omega=0$8

where $\omega=0$9 generates the conservative part, $a$0 contains the coupling between the two CTP branches and generates dissipation, and the imaginary part of $a$1 gives decoherence. The effective equation of motion is

$a$2

The paper’s structural conclusion is that dissipative effective forces can be represented as semiholonomic forces in the CTP formalism. This is expressed through

$a$3

The same framework also governs decoherence in quantum theory, where the reduced density matrix acquires a suppression factor controlled by $a$4. The broader significance is that classical friction, decoherence, and mass or frequency renormalization are presented as different manifestations of the same reduced-description mechanism (Polonyi, 2015).

The ideal-gas extension sharpens the decentralized interpretation. Even without an external probe, composite observables such as currents or bilocal fields can display dissipative effective dynamics because they couple to infinitely many normal modes. For the current $a$5, the effective action is

$a$6

with linear response

$a$7

This shows that a free gas can generate effective dissipation internally once the observable is composite and the unresolved sector is sufficiently large (Polonyi, 2015).

4. Local continuum mechanics in viscoelastic collisions

A distinct but closely related version of decentralized dissipation is obtained in the rigorous derivation of dissipative contact forces between colliding viscoelastic bodies (Brilliantov et al., 2014, Goldobin et al., 2015). The physical setting is a collision between two smooth convex bodies of possibly different materials, with small impact velocity, no plastic deformation, no fragmentation, and elastic recovery after separation. The impact is assumed slow enough that the deformation rate is small compared to the speed of sound and the material relaxation time is much shorter than the collision duration.

The continuum mechanics equation is

$a$8

with elastic stress

$a$9

and viscous stress

${\bf f}_{ij}(r_{ij}) = \sum_{\substack{n\in \Omega_i,\m\in \Omega_j}{\bf f}_{nm}(r_{nm})},$0

Here

${\bf f}_{ij}(r_{ij}) = \sum_{\substack{n\in \Omega_i,\m\in \Omega_j}{\bf f}_{nm}(r_{nm})},$1

with ${\bf f}_{ij}(r_{ij}) = \sum_{\substack{n\in \Omega_i,\m\in \Omega_j}{\bf f}_{nm}(r_{nm})},$2 and ${\bf f}_{ij}(r_{ij}) = \sum_{\substack{n\in \Omega_i,\m\in \Omega_j}{\bf f}_{nm}(r_{nm})},$3. For the two bodies there are separate displacement fields ${\bf f}_{ij}(r_{ij}) = \sum_{\substack{n\in \Omega_i,\m\in \Omega_j}{\bf f}_{nm}(r_{nm})},$4 and ${\bf f}_{ij}(r_{ij}) = \sum_{\substack{n\in \Omega_i,\m\in \Omega_j}{\bf f}_{nm}(r_{nm})},$5, and at the contact plane ${\bf f}_{ij}(r_{ij}) = \sum_{\substack{n\in \Omega_i,\m\in \Omega_j}{\bf f}_{nm}(r_{nm})},$6,

${\bf f}_{ij}(r_{ij}) = \sum_{\substack{n\in \Omega_i,\m\in \Omega_j}{\bf f}_{nm}(r_{nm})},$7

The perturbation scheme is organized by the small parameter

${\bf f}_{ij}(r_{ij}) = \sum_{\substack{n\in \Omega_i,\m\in \Omega_j}{\bf f}_{nm}(r_{nm})},$8

with ${\bf f}_{ij}(r_{ij}) = \sum_{\substack{n\in \Omega_i,\m\in \Omega_j}{\bf f}_{nm}(r_{nm})},$9. Zeroth order recovers static Hertz contact, while first order incorporates viscous effects through

$${\bf u}_{i}(t) = \sum_{n\in \Omega_i}{\bf v}_{n}(t).$$0

This is the key improvement over the quasi-static approximation: the dissipative correction is not taken as a viscous stress alone, but as the consistent first-order solution of the boundary-value problem including the elastic correction induced by dissipation.

For the static solution, the contact pressure over the elliptical contact area $${\bf u}_{i}(t) = \sum_{n\in \Omega_i}{\bf v}_{n}(t).$$1 is

$${\bf u}_{i}(t) = \sum_{n\in \Omega_i}{\bf v}_{n}(t).$$2

and the Hertz-type law is

$${\bf u}_{i}(t) = \sum_{n\in \Omega_i}{\bf v}_{n}(t).$$3

For spheres of equal material, this reduces to

$${\bf u}_{i}(t) = \sum_{n\in \Omega_i}{\bf v}_{n}(t).$$4

At first order, the contact-plane stress satisfies

$${\bf u}_{i}(t) = \sum_{n\in \Omega_i}{\bf v}_{n}(t).$$5

with

$${\bf u}_{i}(t) = \sum_{n\in \Omega_i}{\bf v}_{n}(t).$$6

Integrating over the contact area yields

$${\bf u}_{i}(t) = \sum_{n\in \Omega_i}{\bf v}_{n}(t).$$7

and therefore

$${\bf u}_{i}(t) = \sum_{n\in \Omega_i}{\bf v}_{n}(t).$$8

The total normal force becomes

$${\bf u}_{i}(t) = \sum_{n\in \Omega_i}{\bf v}_{n}(t).$$9

The significance of this result lies in what it corrects. The quasi-static approximation can produce inconsistencies for different materials and can vanish unphysically when the shear modulus is very small or for nearly incompressible materials. The rigorous first-order treatment does not suffer from these inconsistencies because it includes both the viscous stress and the first-order elastic deformation caused by that stress (Goldobin et al., 2015, Brilliantov et al., 2014). In this setting, dissipation is decentralized because the force derives from distributed viscous losses in the material interior and from the local contact stress field, not from an abstract dashpot at the interface.

5. Pairwise dissipative interactions in DPD and active matter

In standard DPD, the decentralized dissipative force model takes the form of a local pairwise thermostat (Altamirano et al., 2016). Each coarse-grained bead interacts through three pairwise forces,

$${\bf f}_{ij}(r_{ij}) = {\bf f}_{ij}^{({\rm c})}(r_{ij}) + {\bf f}_{ij}^{({\rm d})}(r_{ij}) + {\bf f}_{ij}^{({\rm s})}(r_{ij}),$$0

where the dissipative force is

$${\bf f}_{ij}(r_{ij}) = {\bf f}_{ij}^{({\rm c})}(r_{ij}) + {\bf f}_{ij}^{({\rm d})}(r_{ij}) + {\bf f}_{ij}^{({\rm s})}(r_{ij}),$$1

the random force is

$${\bf f}_{ij}(r_{ij}) = {\bf f}_{ij}^{({\rm c})}(r_{ij}) + {\bf f}_{ij}^{({\rm d})}(r_{ij}) + {\bf f}_{ij}^{({\rm s})}(r_{ij}),$$2

and the conservative force is

$${\bf f}_{ij}(r_{ij}) = {\bf f}_{ij}^{({\rm c})}(r_{ij}) + {\bf f}_{ij}^{({\rm d})}(r_{ij}) + {\bf f}_{ij}^{({\rm s})}(r_{ij}),$$3

The weight functions satisfy

$${\bf f}_{ij}(r_{ij}) = {\bf f}_{ij}^{({\rm c})}(r_{ij}) + {\bf f}_{ij}^{({\rm d})}(r_{ij}) + {\bf f}_{ij}^{({\rm s})}(r_{ij}),$$4

with fluctuation-dissipation constraint

$${\bf f}_{ij}(r_{ij}) = {\bf f}_{ij}^{({\rm c})}(r_{ij}) + {\bf f}_{ij}^{({\rm d})}(r_{ij}) + {\bf f}_{ij}^{({\rm s})}(r_{ij}),$$5

This architecture makes DPD decentralized in an immediately operational sense. The dissipative force depends only on pairwise relative motion; there is no global friction field and no separate collective thermostat. Each pair exchanges momentum locally, and the random force compensates the dissipation on average so that the target temperature is maintained without an external thermostat (Altamirano et al., 2016).

At equilibrium, however, the role of the dissipative and random terms is primarily thermostatic rather than thermodynamic. The pressure is written in virial form as

$${\bf f}_{ij}(r_{ij}) = {\bf f}_{ij}^{({\rm c})}(r_{ij}) + {\bf f}_{ij}^{({\rm d})}(r_{ij}) + {\bf f}_{ij}^{({\rm s})}(r_{ij}),$$6

and the numerical results show that the random force contribution becomes negligible very quickly, the dissipative force contribution also vanishes after sufficient simulation time, and the conservative force is the only term that matters for long-time equilibrium pressure. This remains true even when the dissipation strength is changed. A common misconception is therefore that the dissipative term contributes substantially to equilibrium pressure simply because it is present in the equations of motion; the reported result is the opposite for converged long-time averages (Altamirano et al., 2016).

The same local pairwise principle appears in active matter in the ABP-DPD hybrid model of collective motion (Lobaskin et al., 2013). There, active Brownian particles with an internal energy depot obey

$${\bf f}_{ij}(r_{ij}) = {\bf f}_{ij}^{({\rm c})}(r_{ij}) + {\bf f}_{ij}^{({\rm d})}(r_{ij}) + {\bf f}_{ij}^{({\rm s})}(r_{ij}),$$7

with thrust

$${\bf f}_{ij}(r_{ij}) = {\bf f}_{ij}^{({\rm c})}(r_{ij}) + {\bf f}_{ij}^{({\rm d})}(r_{ij}) + {\bf f}_{ij}^{({\rm s})}(r_{ij}),$$8

and stationary speed

$${\bf f}_{ij}(r_{ij}) = {\bf f}_{ij}^{({\rm c})}(r_{ij}) + {\bf f}_{ij}^{({\rm d})}(r_{ij}) + {\bf f}_{ij}^{({\rm s})}(r_{ij}),$$9

valid for $${\bf f}_{ij}^{({\rm c})}(r_{ij}) = \alpha(r_{ij}){\bf e}_{ij},$$0. The interparticle force is

$${\bf f}_{ij}^{({\rm c})}(r_{ij}) = \alpha(r_{ij}){\bf e}_{ij},$$1

with pairwise dissipative term

$${\bf f}_{ij}^{({\rm c})}(r_{ij}) = \alpha(r_{ij}){\bf e}_{ij},$$2

Here the alignment mechanism is not a Vicsek-style rule but frictional suppression of relative motion between neighbors. The order-disorder transition is continuous according to Binder cumulant analysis, and the critical influx rate scales as

$${\bf f}_{ij}^{({\rm c})}(r_{ij}) = \alpha(r_{ij}){\bf e}_{ij},$$3

The model shows that collective motion can emerge from self-propulsion combined with decentralized dissipative collisions, without explicit velocity-copying or a centralized alignment field (Lobaskin et al., 2013).

6. Bulk–boundary decomposition on open graphs

A network-theoretic version of the decentralized dissipative force model is developed for an open linear network on a finite graph (Renger et al., 2022). The system consists of an irreducible finite graph $${\bf f}_{ij}^{({\rm c})}(r_{ij}) = \alpha(r_{ij}){\bf e}_{ij},$$4, with internal hopping rates $${\bf f}_{ij}^{({\rm c})}(r_{ij}) = \alpha(r_{ij}){\bf e}_{ij},$$5 and boundary injection and removal rates $${\bf f}_{ij}^{({\rm c})}(r_{ij}) = \alpha(r_{ij}){\bf e}_{ij},$$6, $${\bf f}_{ij}^{({\rm c})}(r_{ij}) = \alpha(r_{ij}){\bf e}_{ij},$$7. The macroscopic density evolves as

$${\bf f}_{ij}^{({\rm c})}(r_{ij}) = \alpha(r_{ij}){\bf e}_{ij},$$8

In flux form,

$${\bf f}_{ij}^{({\rm c})}(r_{ij}) = \alpha(r_{ij}){\bf e}_{ij},$$9

with continuity equation

$${\bf f}_{ij}^{({\rm d})}(r_{ij}) = -\gamma(r_{ij})u_{ij}{\bf e}_{ij},$$0

The large-deviation Lagrangian $${\bf f}_{ij}^{({\rm d})}(r_{ij}) = -\gamma(r_{ij})u_{ij}{\bf e}_{ij},$$1 has zero-cost flux $${\bf f}_{ij}^{({\rm d})}(r_{ij}) = -\gamma(r_{ij})u_{ij}{\bf e}_{ij},$$2, and the force is defined as the derivative of the cost at zero flux,

$${\bf f}_{ij}^{({\rm d})}(r_{ij}) = -\gamma(r_{ij})u_{ij}{\bf e}_{ij},$$3

$${\bf f}_{ij}^{({\rm d})}(r_{ij}) = -\gamma(r_{ij})u_{ij}{\bf e}_{ij},$$4

This force is split into symmetric and antisymmetric components: $${\bf f}_{ij}^{({\rm d})}(r_{ij}) = -\gamma(r_{ij})u_{ij}{\bf e}_{ij},$$5

$${\bf f}_{ij}^{({\rm d})}(r_{ij}) = -\gamma(r_{ij})u_{ij}{\bf e}_{ij},$$6

The symmetric part is the dissipative force and has the form

$${\bf f}_{ij}^{({\rm d})}(r_{ij}) = -\gamma(r_{ij})u_{ij}{\bf e}_{ij},$$7

where the quasipotential is

$${\bf f}_{ij}^{({\rm d})}(r_{ij}) = -\gamma(r_{ij})u_{ij}{\bf e}_{ij},$$8

The steady state $${\bf f}_{ij}^{({\rm d})}(r_{ij}) = -\gamma(r_{ij})u_{ij}{\bf e}_{ij},$$9 solves

$${\bf f}_{ij}^{({\rm s})}(r_{ij}) = \sqrt{2k_{\rm B}T\gamma(r_{ij})/\Delta t}\,\theta_{ij}{\bf e}_{ij}.$$0

and is coordinate-wise positive and unique. The antisymmetric part is non-dissipative, independent of $${\bf f}_{ij}^{({\rm s})}(r_{ij}) = \sqrt{2k_{\rm B}T\gamma(r_{ij})/\Delta t}\,\theta_{ij}{\bf e}_{ij}.$$1, and encodes departures from detailed balance.

The crucial structural statement is that the symmetric and antisymmetric forces are orthogonal “in a certain sense.” Because the cost is non-quadratic, this is not ordinary Euclidean orthogonality. Instead, a generalized pairing is introduced such that

$${\bf f}_{ij}^{({\rm s})}(r_{ij}) = \sqrt{2k_{\rm B}T\gamma(r_{ij})/\Delta t}\,\theta_{ij}{\bf e}_{ij}.$$2

This vanishing allows the large-deviation cost to decompose cleanly into dissipative and non-dissipative parts. The antisymmetric force leaves the quasipotential unchanged and generates motion tangent to its level sets, while the symmetric force drives relaxation toward the steady state (Renger et al., 2022).

When the dissipative part is removed, the zero-cost dynamics associated with the antisymmetric sector becomes Hamiltonian: $${\bf f}_{ij}^{({\rm s})}(r_{ij}) = \sqrt{2k_{\rm B}T\gamma(r_{ij})/\Delta t}\,\theta_{ij}{\bf e}_{ij}.$$3 with conserved energy

$${\bf f}_{ij}^{({\rm s})}(r_{ij}) = \sqrt{2k_{\rm B}T\gamma(r_{ij})/\Delta t}\,\theta_{ij}{\bf e}_{ij}.$$4

This provides a sharp separation between relaxation and circulation. The numerical examples show that full and symmetric dynamics converge to $${\bf f}_{ij}^{({\rm s})}(r_{ij}) = \sqrt{2k_{\rm B}T\gamma(r_{ij})/\Delta t}\,\theta_{ij}{\bf e}_{ij}.$$5, whereas antisymmetric dynamics orbits around it; the same pattern holds both for pure bulk forcing and for combined bulk-boundary forcing (Renger et al., 2022).

7. Comparative structure, limits, and recurrent misconceptions

The cited formulations are technically heterogeneous, but they share a recognizable architecture.

Setting	Local source of dissipation	Representative result
Atomistic DPD mapping	Atom exchange in invariant volumes; local force–velocity averages	$${\bf f}_{ij}^{({\rm s})}(r_{ij}) = \sqrt{2k_{\rm B}T\gamma(r_{ij})/\Delta t}\,\theta_{ij}{\bf e}_{ij}.$$6
CTP reduced dynamics	Mixing of infinitely many soft normal modes	$${\bf f}_{ij}^{({\rm s})}(r_{ij}) = \sqrt{2k_{\rm B}T\gamma(r_{ij})/\Delta t}\,\theta_{ij}{\bf e}_{ij}.$$7
Viscoelastic impact	Bulk viscous stress and local contact stress	$${\bf f}_{ij}^{({\rm s})}(r_{ij}) = \sqrt{2k_{\rm B}T\gamma(r_{ij})/\Delta t}\,\theta_{ij}{\bf e}_{ij}.$$8
Standard DPD	Short-ranged pairwise drag and matched noise	$${\bf f}_{ij}^{({\rm s})}(r_{ij}) = \sqrt{2k_{\rm B}T\gamma(r_{ij})/\Delta t}\,\theta_{ij}{\bf e}_{ij}.$$9
Open graph network	Dissipative symmetric force plus non-dissipative antisymmetric force	$(x^+,x^-)$00
ABP-DPD swarm	Pairwise frictional suppression of relative motion	$(x^+,x^-)$01

A first recurrent misconception is that dissipative models of this type must rely on an externally inserted damping law. The atomistic DPD, CTP, and viscoelastic contact results all argue otherwise: dissipation can be derived from local atomistic exchange, from integrating out hidden coordinates, or from a perturbative continuum solution that resolves bulk viscous losses (Sokhan et al., 2018, Polonyi, 2015, Goldobin et al., 2015).

A second misconception is that dissipation necessarily requires explicit microscopic interactions. The ideal-gas analysis shows that effective dissipation can arise from coarse-graining over infinitely many soft normal modes even in a non-interacting system, and that composite observables may display dissipative dynamics although the underlying gas is free (Polonyi, 2015).

A third misconception concerns DPD equilibrium observables. Because the dissipative and random forces are essential to the dynamics, they are sometimes presumed to contribute materially to long-time equilibrium pressure. The numerical analysis of DPD virials shows instead that their contributions vanish after sufficient simulation time, while the conservative term determines the equilibrium pressure (Altamirano et al., 2016).

A fourth issue is the status of quasi-static contact laws. The rigorous viscoelastic derivations identify a genuine limitation of the quasi-static approximation: it neglects the first-order elastic correction induced by viscous stress, even though that correction is of the same perturbative order. This is why the corrected coefficient is not just the quasi-static one, and why the rigorous result restores physical consistency for different materials and avoids unphysical vanishing in limiting cases (Brilliantov et al., 2014, Goldobin et al., 2015).

Taken together, these works suggest a coherent but plural picture. A decentralized dissipative force model is not a single equation class; it is a modeling principle in which dissipation is generated by local pairwise exchange, local continuum fields, or reduced-description coupling to unresolved modes. The practical consequence is a route from microscopic or mesoscopic structure to effective irreversible dynamics without postulating a global friction term in advance (Sokhan et al., 2018, Renger et al., 2022).