[Paper Review] Finite time blow-up and condensation for the bosonic Nordheim equation

Miguel Escobedo, Juan J. L. Velázquez|arXiv (Cornell University)|Jun 23, 2012

Gas Dynamics and Kinetic Theory21 references4 citations

TL;DR

This paper establishes the existence of classical solutions to the bosonic Nordheim equation that blow up in finite time and demonstrates finite-time Bose-Einstein condensation for weak solutions. Using energy and moment estimates, it proves that initial bounded, finite-energy distributions can develop singularities at zero momentum and accumulate a positive fraction of particles at the origin within finite time, confirming physical conjectures via rigorous analysis of the kinetic equation with quantum statistics.

ABSTRACT

The homogeneous bosonic Nordheim equation is a kinetic equation describing the dynamics of the distribution of particles in the space of moments for a homogeneous, weakly interacting, quantum gas of bosons. We show the existence of classical solutions of the homogeneous bosonic Nordheim equation that blow up in finite time. We also prove finite time condensation for a class of weak solutions of the kinetic equation.

Motivation & Objective

• To rigorously establish the existence of finite-time blow-up in classical solutions of the homogeneous bosonic Nordheim equation.
• To prove the occurrence of finite-time Bose-Einstein condensation in weak solutions, defined as the formation of a positive Dirac mass at zero energy within finite time.
• To resolve a long-standing physical conjecture by showing that initial bounded, finite-energy distributions can develop singularities and condense at the origin.
• To analyze the role of quantum statistics in the Nordheim equation, particularly the cubic collision terms $ q_3(F) $, which distinguish it from classical Boltzmann dynamics.

Proposed method

• The authors analyze the homogeneous bosonic Nordheim equation in the form of a nonlinear integro-differential kinetic equation with delta-function conservation laws for momentum and energy.
• They use a transformation to radial symmetry in energy space, reducing the equation to a form involving the distribution $ g(t,\epsilon) $, where $ \epsilon = |p|^2/2 $.
• Classical solutions are constructed via a mild formulation in $ L^\infty $, and blow-up is proven by contradiction using energy and moment estimates.
• For weak solutions, the existence of a Dirac mass at $ \epsilon = 0 $ is shown using a limiting argument and continuity in the weak topology of measures.
• The proof relies on key lemmas (e.g., Lemma 8.1, 8.7, 8.10) that control the growth of $ \|f\|_{L^\infty} $ and the absence of mass at zero initially.
• The construction of global weak solutions combines bounded mild solutions on $ [0, T_*] $ with a continuation method from [19] beyond $ T_* $, ensuring weak continuity in time.

Experimental results

Research questions

• RQ1 Can classical solutions of the bosonic Nordheim equation blow up in finite time despite initial boundedness and finite energy?
• RQ2 Does the presence of quantum statistics in the Nordheim equation lead to the formation of a Bose-Einstein condensate in finite time?
• RQ3 Is the time of blow-up equivalent to the time of condensation formation, as conjectured in the physical literature?
• RQ4 Can weak solutions of the equation develop a Dirac measure at zero energy in finite time, even when initial data are smooth and bounded?

Key findings

• Finite-time blow-up occurs for classical solutions of the bosonic Nordheim equation when initial data are bounded and decay sufficiently fast at infinity, with blow-up time $ T_{\text{max}} \leq \frac{K_3}{(1 - 2^{-\beta})} \rho^\beta $.
• Finite-time condensation is proven: there exists a finite $ t^{**} > 0 $ such that $ \int_{\{0\}} g(t^{**}, \epsilon) \, d\epsilon > 0 $, even when initial data satisfy $ f_0 \in L^\infty(\mathbb{R}^+; (1+\epsilon)^\gamma) $ with $ \gamma > 3 $.
• The blow-up time $ T_{\text{max}} $ is less than or equal to the condensation time $ T_{\text{cond}} $, but equality $ T_{\text{max}} = T_{\text{cond}} $ is not proven and remains an open question.
• The onset of a Dirac mass at $ \epsilon = 0 $ is not considered condensation if it occurs at $ \epsilon > 0 $, as only supercritical Bose-Einstein distributions with mass at zero are stationary solutions with macroscopic occupation.
• The analysis confirms that the cubic collision term $ q_3(F) $, arising from Bose statistics, is responsible for the blow-up and condensation, distinguishing the Nordheim equation from classical Boltzmann dynamics.
• The results validate numerical and physical conjectures that singularity formation at $ p=0 $ corresponds to the onset of Bose-Einstein condensation in weakly interacting quantum gases.