# American Institute of Mathematical Sciences

December  2019, 11(4): 447-485. doi: 10.3934/jgm.2019023

## Networks of coadjoint orbits: From geometric to statistical mechanics

 Department of Mathematics, Imperial College, London SW7 2AZ, UK

* Corresponding author: So Takao

To Darryl Holm, for his 70th birthday

Received  April 2018 Revised  May 2019 Published  November 2019

Fund Project: AA acknowledges EPSRC funding through award EP/N014529/1 via the EPSRC Centre for Mathematics of Precision Healthcare. ST acknowledges funding through Schrödinger scholarship scheme.

A class of network models with symmetry group $G$ that evolve as a Lie-Poisson system is derived from the framework of geometric mechanics, which generalises the classical Heisenberg model studied in statistical mechanics. We considered two ways of coupling the spins: one via the momentum and the other via the position and studied in details the equilibrium solutions and their corresponding nonlinear stability properties using the energy-Casimir method. We then took the example $G = SO(3)$ and saw that the momentum-coupled system reduces to the classical Heisenberg model with massive spins and the position-coupled case reduces to a new system that has a broken symmetry group $SO(3)/SO(2)$ similar to the heavy top. In the latter system, we numerically observed an interesting synchronisation-like phenomenon for a certain class of initial conditions. Adding a type of noise and dissipation that preserves the coadjoint orbit of the network model, we found that the invariant measure is given by the Gibbs measure, from which the notion of temperature is defined. We then observed a surprising 'triple-humped' phase transition in the heavy top-like lattice model, where the spins switched from one equilibrium position to another before losing magnetisation as we increased the temperature. This work is only a first step towards connecting geometric mechanics with statistical mechanics and several interesting problems are open for further investigation.

Citation: Alexis Arnaudon, So Takao. Networks of coadjoint orbits: From geometric to statistical mechanics. Journal of Geometric Mechanics, 2019, 11 (4) : 447-485. doi: 10.3934/jgm.2019023
##### References:

show all references

##### References:
In this figure, we plot the maximum real part of the eigenvalues $\lambda_S$ of the linearised system (58) corresponding to equilibrium solutions ${\boldsymbol \Pi}_e$ (eigenvectors of $\mathbb L$) with energy $\lambda_e$ (eigenvalue of $\mathbb L$ corresponding to eigenvector ${\boldsymbol \Pi}_e$). The blue crosses indicate the maximum real eigenvalues that are non-zero, corresponding to unstable solutions, and the red dots indicate those that are zero, corresponding to linearly stable solutions. The purple dots indicate the algebraic multiplicity $n$ of $\lambda_e$, whose values are to be read from the right axis. As expected, the equilibria corresponding to the highest and lowest energy are stable in both cases
In this figure, we display four non-trivial examples of linearly stable equilibrium solutions of the $20\times 20$ rigid body lattice. All the solutions are parallel to one of the axes (either $\Pi_1$ or $\Pi_3$) and have an 'argyle' pattern of aligned spins (2(a), 2(d)) or anti-aligned spins (2(b), 2(c)). The former pattern corresponds to configurations with low energy (negative values of $\lambda$), and the latter corresponds to configurations with high energy (positive values of $\lambda$). All four solutions given here correspond to the red dots in figure 1
This figure shows the possible $(\lambda_1, \lambda_2)$-pairs that one can have from solving equation (86). Fixing $\lambda_1$, the possible values of $\lambda_2$ that solve (86) are the eigenvalues of the graph Laplacian $\mathbb L(\lambda_1)$. We see that the eigenvalues $\lambda_2$ in the left panel 3(a) collapse as $\lambda_1 \rightarrow 0$, which does not occur in the right panel 3(b)
The two panels in this figure display the maximum, real part of the eigenvalues $\lambda_S$ of the linearised system (98) around the equilibrium configuration ${\boldsymbol \Gamma}_e$ (eigenvectors of $\mathbb L(\lambda_1)$) plotted against $\lambda_2$ (minus the eigenvalue of $\mathbb L(\lambda_1)$ corresponding to ${\boldsymbol \Gamma}_e$) for $\lambda_1 = 0.5$. Again, the blue crosses correspond to unstable equilibria, the red dots correspond to linearly stable equilibria and the purple dots correspond to the multiplicity of $\lambda_2$, whose values are to be read from the right axis. As expected, the configuration corresponding to the highest $\lambda_2$ (or, the lowest eigenvalue of $\mathbb L(\lambda_1)$) is stable in both cases
This figure shows the time evolution of the averaged spins (5(a), 5(c), 5(e)) and energy (5(b), 5(d), 5(f)) in a $20\times 20$ heavy top lattice, starting from different initial conditions, with dissipation (dashed lines) or without dissipation (bold lines). The blue, orange and green lines in the left panel correspond to the $\Gamma_1$, $\Gamma_2$ and $\Gamma_3$-components of the spins averaged over the lattice respectively and the same colours in the right panel correspond to the total, kinetic and potential energy. We observe metastability of the $\Gamma_1$ and $\Gamma_2$ axes in panels 5(c) and 5(a) in the dissipative case, and a partial synchronisation around the $\Gamma_3$-axis in the non-dissipative case in 5(a), 5(c) and 5(e)
The two panels in this figure show the temperature phase transition of the rigid body lattice for two different cases. The blue dots correspond to averaged magnetisation computed via direct simulation of the full stochastic equation (47) and the black line corresponds to the mean field approximation (103). Both cases exhibit a second order phase transition, characterised by a sudden loss of magnetisation (magnitude of the order parameter) at the critical temperature
The four panels in this figure show phase transitions in the heavy top lattice. In figure 7(a), the blue dots represent the $\Gamma_3$-components of the averaged position $\left\langle {\bf{\Gamma }} \right\rangle$ and in 7(b)-7(d), the blue, green and red dots represent the $\Gamma_1$, $\Gamma_2$ and $\Gamma_3$-components of the averaged position respectively. The black line in all four figures correspond to the mean field approximation (106). The top left panel 7(a) shows a phase transition similar to the rigid body phase transition, and the other panels 7(b)-7(d) display a 'triple-humped' phase transition. The corresponding critical temperatures are indicated by the vertical dashed lines. In 7(b)-7(d), we also varied the Casimirs (where $c_1 = {\boldsymbol \Pi} \cdot {\boldsymbol \Gamma}$ and $c_1 = \|{\boldsymbol \Gamma}\|^2$) to show that the three critical temperatures depend on their values, and may even disappear for large enough $c_1/c_2$. Due to the high dimension and non-compactness of the phase space, the mean field approximation requires intensive computation to be precise, explaining the visible errors, especially at low temperatures
 [1] Yuncherl Choi, Taeyoung Ha, Jongmin Han, Sewoong Kim, Doo Seok Lee. Turing instability and dynamic phase transition for the Brusselator model with multiple critical eigenvalues. Discrete & Continuous Dynamical Systems - A, 2021  doi: 10.3934/dcds.2021035 [2] Peter Benner, Jens Saak, M. Monir Uddin. Balancing based model reduction for structured index-2 unstable descriptor systems with application to flow control. Numerical Algebra, Control & Optimization, 2016, 6 (1) : 1-20. doi: 10.3934/naco.2016.6.1 [3] Braxton Osting, Jérôme Darbon, Stanley Osher. Statistical ranking using the $l^{1}$-norm on graphs. Inverse Problems & Imaging, 2013, 7 (3) : 907-926. doi: 10.3934/ipi.2013.7.907 [4] F.J. Herranz, J. de Lucas, C. Sardón. Jacobi--Lie systems: Fundamentals and low-dimensional classification. Conference Publications, 2015, 2015 (special) : 605-614. doi: 10.3934/proc.2015.0605 [5] Wolf-Jüergen Beyn, Janosch Rieger. The implicit Euler scheme for one-sided Lipschitz differential inclusions. Discrete & Continuous Dynamical Systems - B, 2010, 14 (2) : 409-428. doi: 10.3934/dcdsb.2010.14.409 [6] Jean-François Biasse. Improvements in the computation of ideal class groups of imaginary quadratic number fields. Advances in Mathematics of Communications, 2010, 4 (2) : 141-154. doi: 10.3934/amc.2010.4.141 [7] Michael Grinfeld, Amy Novick-Cohen. Some remarks on stability for a phase field model with memory. Discrete & Continuous Dynamical Systems - A, 2006, 15 (4) : 1089-1117. doi: 10.3934/dcds.2006.15.1089 [8] Mikhail Gilman, Semyon Tsynkov. Statistical characterization of scattering delay in synthetic aperture radar imaging. Inverse Problems & Imaging, 2020, 14 (3) : 511-533. doi: 10.3934/ipi.2020024 [9] Jan Prüss, Laurent Pujo-Menjouet, G.F. Webb, Rico Zacher. Analysis of a model for the dynamics of prions. Discrete & Continuous Dynamical Systems - B, 2006, 6 (1) : 225-235. doi: 10.3934/dcdsb.2006.6.225 [10] Simone Calogero, Juan Calvo, Óscar Sánchez, Juan Soler. Dispersive behavior in galactic dynamics. Discrete & Continuous Dynamical Systems - B, 2010, 14 (1) : 1-16. doi: 10.3934/dcdsb.2010.14.1 [11] Lucas C. F. Ferreira, Jhean E. Pérez-López, Élder J. Villamizar-Roa. On the product in Besov-Lorentz-Morrey spaces and existence of solutions for the stationary Boussinesq equations. Communications on Pure & Applied Analysis, 2018, 17 (6) : 2423-2439. doi: 10.3934/cpaa.2018115 [12] Alexey Yulin, Alan Champneys. Snake-to-isola transition and moving solitons via symmetry-breaking in discrete optical cavities. Discrete & Continuous Dynamical Systems - S, 2011, 4 (5) : 1341-1357. doi: 10.3934/dcdss.2011.4.1341 [13] Juan Manuel Pastor, Javier García-Algarra, Javier Galeano, José María Iriondo, José J. Ramasco. A simple and bounded model of population dynamics for mutualistic networks. Networks & Heterogeneous Media, 2015, 10 (1) : 53-70. doi: 10.3934/nhm.2015.10.53 [14] Tomáš Roubíček. An energy-conserving time-discretisation scheme for poroelastic media with phase-field fracture emitting waves and heat. Discrete & Continuous Dynamical Systems - S, 2017, 10 (4) : 867-893. doi: 10.3934/dcdss.2017044 [15] Paula A. González-Parra, Sunmi Lee, Leticia Velázquez, Carlos Castillo-Chavez. A note on the use of optimal control on a discrete time model of influenza dynamics. Mathematical Biosciences & Engineering, 2011, 8 (1) : 183-197. doi: 10.3934/mbe.2011.8.183 [16] Guirong Jiang, Qishao Lu. The dynamics of a Prey-Predator model with impulsive state feedback control. Discrete & Continuous Dynamical Systems - B, 2006, 6 (6) : 1301-1320. doi: 10.3934/dcdsb.2006.6.1301 [17] Mingxin Wang, Qianying Zhang. Dynamics for the diffusive Leslie-Gower model with double free boundaries. Discrete & Continuous Dynamical Systems - A, 2018, 38 (5) : 2591-2607. doi: 10.3934/dcds.2018109 [18] Wen-Bin Yang, Yan-Ling Li, Jianhua Wu, Hai-Xia Li. Dynamics of a food chain model with ratio-dependent and modified Leslie-Gower functional responses. Discrete & Continuous Dynamical Systems - B, 2015, 20 (7) : 2269-2290. doi: 10.3934/dcdsb.2015.20.2269 [19] Mats Gyllenberg, Jifa Jiang, Lei Niu, Ping Yan. On the classification of generalized competitive Atkinson-Allen models via the dynamics on the boundary of the carrying simplex. Discrete & Continuous Dynamical Systems - A, 2018, 38 (2) : 615-650. doi: 10.3934/dcds.2018027

2019 Impact Factor: 0.649