---
title: Homo-Energetic Solutions in Kinetic Theory
url: https://www.emergentmind.com/topics/homo-energetic-solutions
type: topic
---

# Homo-Energetic Solutions in Kinetic Theory

Searching arXiv for the specified paper and closely related work on homoenergetic solutions and the Rayleigh–Boltzmann equation.
Homoenergetic solutions are a particular class of solutions to the linear Boltzmann–Rayleigh equation in which the spatial dependence is absorbed into a velocity shift induced by a prescribed deformation field. In the formulation studied in “Homoenergetic solutions for the Rayleigh-Boltzmann equation: existence of a stationary non-equilibrium solution” [2501.14454], they are used to describe the dynamics of Boltzmann gases under mechanical deformations, with emphasis on simple shear. Their significance lies in the fact that their long-time behavior cannot be described by Maxwellian distributions and depends strongly on the homogeneity of the collision kernel. For cut-off collision kernels with homogeneity $\gamma \geq 0$, including Maxwell molecules ($\gamma=0$) and hard potentials with $0 \leq \gamma < 1$, the analysis establishes both well-posedness in the space of non-negative Radon measures and the existence of a stationary non-equilibrium solution under the non-equilibrium condition induced by shear [2501.14454].

## 1. Kinematic ansatz and reduction to a homogeneous problem

The starting point is the linear Boltzmann–Rayleigh equation for a one-particle distribution
$$
g=g(t,x,w)\ge 0
$$
on $\mathbb{R}_+\times \mathbb{R}^3_x \times \mathbb{R}^3_w$,
$$
\partial_t g + w\cdot \nabla_x g = \mathcal{L}g(w).
$$
A homoenergetic solution is introduced through the ansatz
$$
g(t,x,w)=f(t,v), \qquad v:=w-\Sigma(t,x),
$$
with deformation field
$$
\Sigma(t,x)=L(t)x,
$$
where the matrix $L(t)$ solves
$$
\dot L + L^2 = 0,\qquad L(0)=A\in M_3(\mathbb{R}).
$$
Substitution yields a spatially homogeneous equation for $f$,
$$
\partial_t f(t,v)+\nabla_v\cdot(L(t)\,v\,f(t,v))=\mathcal{L}f(t,v).
$$
For the classical choice of simple shear,
$$
A=\begin{pmatrix}0&K&0\\0&0&0\\0&0&0\end{pmatrix},
$$
one has $A^2=0$ and therefore $L(t)=A$ for all $t$. In this case the reduced equation becomes
$$
\partial_t f - K\,v_2\,\partial_{v_1}f = \mathcal{L}f.
$$

This reduction isolates the competition between transport induced by shear and collisions encoded in $\mathcal{L}$. The resulting equation is spatially homogeneous but not equilibrium-preserving in the Maxwellian sense. The paper identifies this as the mechanism behind the non-equilibrium stationary states studied later [2501.14454].

## 2. Collision operator and homogeneity of the kernel

The collision operator $\mathcal{L}$ is obtained by linearizing the classical Boltzmann binary-collision operator around a Maxwellian background $M$. With post- and pre-collisional velocities denoted by $(v',v_*')$ and $(v,v_*)$, and with
$$
n=\frac{v-v_*}{|v-v_*|},
$$
the operator is written as
$$
\mathcal{L}f(v)
=\int_{\mathbb{R}^3}\int_{S^2} B\bigl(n\!\cdot\!\omega,\;|v-v_*|\bigr)
\bigl\{M(v_*')\,f(v') - M(v_*)\,f(v)\bigr\}\,d\omega\,dv_*,
$$
where the collision rule is
$$
v' = v - \bigl((v-v_*)\!\cdot\!\omega\bigr)\,\omega, \qquad
v_*' = v_* + \bigl((v-v_*)\!\cdot\!\omega\bigr)\,\omega,
$$
and the normalized Maxwellian is
$$
M(v)=(2\pi)^{-3/2}\exp\!\bigl(-|v|^2/2\bigr).
$$

Under Grad’s cut-off assumption,
$$
\sup_{v\in\mathbb{R}^3}\;\int_{S^2}B\bigl(n\!\cdot\!\omega,|v-v_*|\bigr)\,d\omega<\infty.
$$
The full homoenergetic equation under simple shear is then
$$
\partial_t f - K\,v_2\,\partial_{v_1}f
=\int_{\mathbb{R}^3}\int_{S^2}B(n\!\cdot\!\omega,|v-v_*|)
\bigl\{M(v_*')f(v')-M(v_*)f(v)\bigr\}\,d\omega\,dv_*.
$$

The collision kernel is assumed homogeneous of degree $\gamma$ in the relative speed,
$$
B\bigl(n\!\cdot\!\omega,\,|v-v_*|\bigr)
=b\bigl(n\!\cdot\!\omega\bigr)\;|v-v_*|^\gamma,\qquad \gamma\in\mathbb{R},
$$
so that for any $\lambda>0$,
$$
B\bigl(n\!\cdot\!\omega,\lambda\,|v-v_*|\bigr)
=\lambda^\gamma\,B\bigl(n\!\cdot\!\omega,|v-v_*|\bigr).
$$
The summary distinguishes three regimes: $\gamma=0$ for Maxwell molecules, $0<\gamma<1$ for moderately hard potentials, and $\gamma<0$ for a hyperbolic-dominated regime. In the cases actually treated, namely $\gamma\ge 0$, the qualitative balance between shear transport and collisions is central. The text states that $\gamma=0$ gives a perfect balance at large velocities, while $\gamma>0$ yields a collision-dominated regime [2501.14454].

## 3. Weak formulation and well-posedness in non-negative Radon measures

The well-posedness theory is developed for
$$
f(t,\cdot)\in C([0,\infty),\mathcal{M}^+(\mathbb{R}^3)),
$$
where $\mathcal{M}^+(\mathbb{R}^3)$ denotes the non-negative finite Radon measures. Weak solutions are defined by testing against $\phi\in C_c^1([0,\infty)\times\mathbb{R}^3)$:
$$
\int_{\mathbb{R}^3}\phi(T,v)\,f(T,dv)
-\int_{\mathbb{R}^3}\phi(0,v)\,f(0,dv)
=\int_0^T\!\int_{\mathbb{R}^3}\Bigl[
\partial_t\phi +K\,v_2\partial_{v_1}\phi +\mathcal{L}^*\phi\Bigr]\,f(t,dv)\,dt,
$$
with $\mathcal{L}^*$ the adjoint collision operator.

The main argument proceeds through the backward adjoint problem
$$
-\partial_t\psi + K\,v_2\,\partial_{v_1}\psi = \mathcal{L}^*\psi,\qquad \psi(T,v)=\psi_T(v),
$$
which is then rewritten forward in time. The generator
$$
A\psi := -K\,v_2\,\partial_{v_1}\psi + \mathcal{L}^*\psi
$$
is considered on the Banach space $C_0(\mathbb{R}^3)$ with a domain imposing sufficient decay or growth control. The paper shows that $(A,D(A))$ is a Markov generator, via the Hille–Yosida or Lumer–Phillips theorem, and therefore generates a positive, strongly continuous semigroup $S(t)$ on $C_0(\mathbb{R}^3)$ [2501.14454].

Solutions are then defined by duality:
$$
\int_{\mathbb{R}^3}\phi(v)\,f(t,dv)
:=\int_{\mathbb{R}^3}[S(t)\phi](v)\,f(0,dv).
$$
This construction automatically satisfies the weak formulation. Uniqueness follows by testing the difference of two solutions against the dual semigroup.

For $\gamma=0$, an alternative approach is also described. One introduces the characteristic function
$$
\varphi(t,k)=\int f(t,v)e^{-ik\cdot v}\,dv,
$$
which satisfies a transport-plus-integral-operator equation. According to the summary, this equation is amenable to a Banach fixed-point argument in the space of characteristic functions. This suggests that the Maxwell case permits an analytically distinct treatment because the homogeneity of the kernel simplifies the Fourier-side structure [2501.14454].

## 4. Stationary non-equilibrium states under shear

The central stationary problem is the steady equation
$$
-K\,v_2\,\partial_{v_1}f_\infty(v)=\mathcal{L}f_\infty(v),\qquad \int f_\infty(dv)=1.
$$
In weak form, for every $\phi\in C_c^2(\mathbb{R}^3)$,
$$
\int_{\mathbb{R}^3}\bigl[-K\,v_2\partial_{v_1}\phi +\mathcal{L}^*\phi\bigr](v)\;f_\infty(dv)=0.
$$
The solution $f_\infty$ is therefore stationary in the co-moving shear frame but remains non-equilibrium because the shear deformation produces a persistent balance between transport and collision effects rather than relaxation to the background Maxwellian.

The construction differs in the two regimes considered in the paper.

| Regime | Shear condition | Main compactness set |
|---|---|---|
| $\gamma=0$ | Small shear $K\in[0,K_0)$ | $V=\{\mu\in\mathcal{M}^+(\mathbb{R}^3): \mu(\mathbb{R}^3)=1,\ \int |v|^2\mu(dv)=\mathrm{tr}\,M_\infty\}$ |
| $0<\gamma<1$ | Arbitrary $K>0$ | $W=\{\mu\in\mathcal{M}^+:\mu(\mathbb{R}^3)=1,\ \int|v|^2\mu(dv)\le C_0\}$ |

In the Maxwell case, the argument begins with closed ODEs for the second-order moment matrix
$$
M(t)_{jk}=\int v_j\,v_k\,f(t,dv).
$$
These satisfy
$$
\dot M + K\bigl(e_1\otimes M_{2\cdot}+M_{2\cdot}\otimes e_1\bigr)=A(M)+C\,I,
$$
where $A$ is a linear operator and $CI$ comes from loss-gain balance. For $K<K_0$, the spectrum of $A$ lies in the left half-plane, so there exists a unique stable stationary matrix
$$
M_\infty=-A^{-1}(C\,I).
$$
The set $V$ defined by mass one and fixed second moment is weak* compact in $\mathcal{M}^+$. The dual semigroup $S^*(t)$ leaves $V$ invariant and is weak* continuous. Schauder’s fixed-point theorem then yields a nontrivial measure $f_\infty\in V$ with $S^*(t)f_\infty=f_\infty$ for all $t$, hence a steady solution [2501.14454].

For hard potentials with $0<\gamma<1$, the key input is a uniform-in-time bound on the second moment
$$
M_2(t)=\int |v|^2 f(t,dv).
$$
Using a Povzner-type estimate, the paper proves that for any $K>0$,
$$
\frac{d}{dt}M_2(t)\le -\alpha\,M_2(t)+C,
$$
with $\alpha>0$ for a suitable choice of constants. Therefore $M_2(t)$ remains bounded. The weak* compact set $W$ is invariant under the semigroup, and Schauder’s theorem again yields a stationary solution.

In strong form, the stationary equation is
$$
-K\,v_2\partial_{v_1}f_\infty(v)
=\int_{\mathbb{R}^3}\int_{S^2}b(n\!\cdot\!\omega)\,|v-v_*|^\gamma
\bigl\{M(v_*')\,f_\infty(v') - M(v_*)\,f_\infty(v)\bigr\}\,d\omega\,dv_*.
$$
A plausible implication is that the existence mechanism depends less on explicit solvability than on the combination of invariant convex sets, moment control, and weak* compactness.

## 5. Maxwell molecules: small-shear stationarity and large-shear growth

A distinctive feature of the Maxwell case $\gamma=0$ is the existence of two qualitatively different behaviors depending on the size of the shear parameter $K$. The paper states that there is a threshold $K_0$ separating small- and large-shear regimes [2501.14454].

For small shear, $K\in[0,K_0)$, the operator governing the second-moment dynamics has strictly negative spectrum. As a consequence, all moments converge exponentially to those of the unique stationary matrix $M_\infty$. The summary further reports that numerically one can expand in powers of $K$ to obtain
$$
M_\infty = I + O(K),
$$
so that $f_\infty$ is a small non-equilibrium perturbation of the Maxwellian. This does not mean the stationary state is Maxwellian; rather, it remains shear-driven and only approaches the Maxwellian configuration in a perturbative sense as $K\to 0$.

For large shear, $K>K_0$, the moment ODE matrix acquires a real eigenvalue
$$
p(K)>0,
$$
and the second moments grow like
$$
e^{p(K)t}.
$$
The summary explicitly states that no bounded steady solution can exist in this regime. Instead, a self-similar asymptotic behavior is conjectured:
$$
f(t,v)\sim e^{-3\beta t}\;F\Bigl(v\,e^{-\beta t}\Bigr),
$$
for suitable $\beta=\beta(K)$, where $F$ solves
$$
-\,\beta\,\nabla_v\cdot(vF)
\;-\;K\,v_2\partial_{v_1}F
\;=\;\mathcal{L}F,\qquad \int F=1.
$$

This dichotomy is one of the paper’s main conceptual outcomes. It indicates that, for Maxwell molecules, shear can be either weak enough to permit a stationary non-equilibrium balance or strong enough to drive unbounded moment growth. The summary describes this as a bifurcation at $K=K_0$ between steady-state relaxation and unbounded, self-similar growth. This suggests that the precise balance implied by $\gamma=0$ is structurally marginal: neither unequivocally collision-dominated nor purely transport-dominated.

## 6. Conceptual significance and relation to non-equilibrium kinetic theory

The results place homoenergetic solutions in a framework where mechanical deformation is built directly into the ansatz, rather than treated as a perturbation of an equilibrium evolution. In this setting, the long-time behavior is governed by the interaction between the deformation field and the homogeneity of the collision kernel. The paper’s conclusions may be summarized as follows: well-posedness holds in the space of non-negative Radon measures; a stationary non-equilibrium state exists when collisions dominate sufficiently, namely for $\gamma>0$ and arbitrary $K>0$, or for $\gamma=0$ with small shear $K<K_0$; and the Maxwell case exhibits a transition to a non-stationary self-similar regime when $K>K_0$ [2501.14454].

A common misconception would be to identify stationary solutions of the reduced equation with equilibrium Maxwellians. The analysis does not support that interpretation. The stationary objects constructed here satisfy a balance law in the sheared frame and arise specifically under a non-equilibrium condition induced by the presence of the shear deformation. Even in the perturbative Maxwell case, where $M_\infty=I+O(K)$ numerically, the stationary state is described as a non-equilibrium perturbation rather than equilibrium itself.

Another potential misconception is that the cut-off assumption or the linearization around a Maxwellian background trivializes the dynamics. The summary instead shows that the combination of transport, collision homogeneity, and moment propagation leads to distinct regimes requiring different tools: Markov semigroup theory, Fourier-transform methods, moment ODEs, Povzner-type estimates, and Schauder fixed-point arguments. This suggests that even in the linear Rayleigh–Boltzmann setting, homoenergetic solutions retain a rich non-equilibrium structure.

Within the scope described in the paper, homoenergetic solutions provide a mathematically controlled setting for studying non-Maxwellian steady states and deformation-driven asymptotics in kinetic theory. The analysis is limited to simple shear and cut-off collision kernels with $\gamma\ge 0$, particularly Maxwell molecules and hard potentials with $0\le \gamma<1$. A plausible implication is that extensions to other deformation fields or to different homogeneity classes would require new mechanisms for moment control and compactness beyond those established here.

Source: https://www.emergentmind.com/topics/homo-energetic-solutions