---
title: 'Nagle–Kardar Model: Mixed-Range Ising System'
url: https://www.emergentmind.com/topics/nagle-kardar-model
type: topic
---

# Nagle–Kardar Model: Mixed-Range Ising System

Searching arXiv for recent papers on the Nagle–Kardar model and related usage of the term.
The **Nagle–Kardar model** denotes a one-dimensional Ising-like spin system with competing short-range and long-range interactions. In its standard classical form, it combines a nearest-neighbor coupling and a mean-field coupling on a ring of spin-\(\tfrac12\) variables, and it is notable for exhibiting a line of continuous phase transitions, a first-order transition line, and a tricritical point despite its one-dimensional geometry [2511.07207]. Recent work has also used the same name for a quantum extension with a transverse field [2604.25837]. In a separate strand of the literature, the label “Nagle–Kardar” has occasionally been used for the continuous Hwa–Kardar model of anisotropic self-organized criticality, but that usage refers to a different field-theoretic object and should be distinguished from the Ising-like model discussed here [1508.00236].

## 1. Definition and Hamiltonian

The classical Nagle–Kardar model is defined on a one-dimensional ring of \(N\) Ising spins,
\[
s_i=\pm 1,\qquad i=1,\dots,N,
\]
with periodic boundary conditions \(s_{N+1}=s_1\) [2511.07207]. Its Hamiltonian is
\[
H = -J \sum_{i=1}^N s_i s_{i+1}
    - \frac{K}{2N} \left(\sum_{i=1}^N s_i\right)^2
    - h \sum_{i=1}^N s_i, \tag{1}
\]
where \(J\) is the nearest-neighbor coupling, \(K\) is the mean-field coupling, and \(h\) is an external field [2511.07207].

This form makes explicit the coexistence of two interaction ranges. The nearest-neighbor term is sensitive to local domain structure: \(J>0\) favors ferromagnetic alignment, while \(J<0\) favors antiferromagnetic order. The mean-field term couples every spin to every other spin equally and favors global magnetization when \(K>0\) [2511.07207]. The model is therefore a paradigmatic mixed-range Ising system. A useful reformulation introduces the total magnetization
\[
M=\sum_{i=1}^N s_i,\qquad m=\frac{M}{N},
\]
and the number of defects \(2S\), where defects are bonds with \(s_i s_{i+1}=-1\),
\[
2S=\sum_{i=1}^N \frac{1-s_i s_{i+1}}{2}. 
\]
In terms of \(M\) and \(S\), the Hamiltonian becomes
\[
H=-J(N-4S)-\frac{K}{2N}M^2-hM. \tag{2}
\]
This representation is central for both equilibrium analysis and dynamical coarse-graining [2511.07207, 2606.25919].

The same mixed-range structure is used in finite-size thermodynamic studies, often with couplings denoted \(J_s\) and \(J_l\), or in reduced form \(K_s=\beta J_s\) and \(K_l=\beta J_l\) [2404.18324, 2508.15430]. In that notation, the model is a one-dimensional Ising chain with nearest-neighbor interaction and equivalent-neighbor ferromagnetic interaction of strength \(J_l/N\) under periodic boundary conditions [2404.18324].

## 2. Equilibrium structure and tricriticality

The static properties are exactly accessible through transfer-matrix and mean-field methods [2511.07207]. Using a Hubbard–Stratonovich transformation, one obtains an effective free-energy density whose minimum determines the equilibrium magnetization \(m^*\). The corresponding self-consistency equation is
\[
m^{*} = \frac{\sinh\big( (h + K m^*)/T \big)}
{\sqrt{\sinh^2\big( (h + K m^*)/T \big) + e^{-4 J/T}}}. \tag{3}
\]
The equilibrium defect density \(s^*\) is then determined by \(m^*\) [2511.07207].

At zero field, \(h=0\), the phase diagram in the \((J/T,K/T)\) plane contains a continuous transition line,
\[
\frac{K}{T}=e^{-2J/T}, \tag{4}
\]
a first-order transition line for sufficiently negative \(J/T\) and sufficiently large \(K/T\), and a tricritical point where the two meet,
\[
\frac{J}{T}=-\frac{\log 3}{4},\qquad \frac{K}{T}=\sqrt{3}. \tag{5}
\]
These results recur in later finite-size and dynamical studies [2511.07207, 2606.25919].

A Landau expansion of the free energy around \(m=0\),
\[
\mathcal{F}(m)=\mathcal{F}_0+\frac{a_2}{2}m^2+\frac{a_4}{4}m^4+\frac{a_6}{6}m^6+\dots,
\]
shows that along the second-order line \(a_2=0\) but \(a_4>0\), whereas at the tricritical point both \(a_2=0\) and \(a_4=0\), leaving \(a_6>0\) as the leading stabilizing term [2511.07207]. This yields mean-field Ising static exponents on the critical line,
\[
\beta=\frac12,\qquad \nu=\frac12,\qquad d_u=4,
\]
and tricritical exponents
\[
\beta=\frac14,\qquad \nu=\frac12,\qquad d_u=3. 
\]
The change in upper critical dimension from \(4\) to \(3\) reflects the shift from an effective \(\phi^4\) to an effective \(\phi^6\) description at tricriticality [2511.07207].

A notable feature is that the one-dimensional geometry does not preclude bulk criticality because the infinite-range term changes the effective thermodynamics. This suggests that the model should not be interpreted as a conventional short-range 1D Ising chain perturbed weakly by finite-size corrections; rather, the mean-field term is structurally dominant in establishing long-range order [2511.07207].

## 3. Fluctuations, large deviations, and finite-size behavior

The Nagle–Kardar model admits an explicit large-deviation description of magnetization fluctuations. The equilibrium magnetization distribution satisfies
\[
P_N(M)\asymp e^{-N\psi(m)}, \tag{6}
\]
with \(\psi(m)\) a quasi-potential whose minimum determines the equilibrium magnetization [2511.07207]. Near \(m=0\), the form of \(\psi\) changes across the phase diagram. Away from criticality, \(\psi(m)\sim m^2\); on the critical line, \(\psi(m)\sim m^4\); and at the tricritical point, \(\psi(m)\sim m^6\) [2511.07207].

These forms directly determine finite-size scaling of equilibrium fluctuations:
\[
|m^*|\sim N^{-1/2}
\]
away from criticality,
\[
|m^*|\sim N^{-1/4} \tag{7}
\]
on the critical line, and
\[
|m^*|\sim N^{-1/6}. \tag{8}
\]
at the tricritical point [2511.07207]. Equivalently, with \(\langle |m|\rangle\sim N^{-\Delta_m}\),
\[
\Delta_m=\frac{\beta}{\nu d_u}, \tag{9}
\]
so that \(\Delta_m=1/4\) on the critical line and \(\Delta_m=1/6\) at tricriticality [2511.07207].

The joint distribution of magnetization and defect density also satisfies a large-deviation principle,
\[
P_N(M,2S)\asymp e^{-N U(m,s)}, \tag{10}
\]
with explicit quasi-potential
\[
\begin{aligned}
U(m,s) &= s\log s -\frac{1+m}{2}\log(1+m) -\frac{1-m}{2}\log(1-m) \\
&\quad +\frac{1+m-s}{2}\log(1+m-s) +\frac{1-m-s}{2}\log(1-m-s) \\
&\quad - \frac{1}{T}\Big(\frac{K}{2} m^2 + h m + J(1-2s) + f\Big).
\end{aligned} \tag{11}
\]
Minima of \(U(m,s)\) define locally stable macrostates, while saddle points define activation barriers between them [2511.07207, 2606.25919].

This two-variable structure is one of the model’s distinctive features. Magnetization \(m\) is the thermodynamic order parameter, but defect density \(s\) carries the domain-wall information required to track local rearrangements and activated transitions [2606.25919].

## 4. Glauber dynamics, Fokker–Planck reduction, and Model A criticality

The canonical dynamics most commonly studied for the Nagle–Kardar model is single-spin-flip Glauber dynamics with non-conserved magnetization [2511.07207]. At each update, one site is selected, a spin flip \(s_i\to -s_i\) is proposed, the associated energy change \(\Delta E\) is computed from the Hamiltonian, and the move is accepted with probability
\[
W(s_i\to -s_i)=\frac{1}{1+e^{\Delta E/T}}. \tag{12}
\]
The microscopic dynamics obeys detailed balance with respect to the canonical Gibbs distribution [2511.07207].

Because the Hamiltonian depends only on magnetization \(M\) and defects \(S\), the full Markov process can be coarse-grained to a master equation for the macrostate probability \(P_N(M,2S;t)\). In the large-\(N\) limit, one introduces
\[
m=\frac{M}{N},\qquad s=\frac{2S}{N},\qquad \epsilon=\frac{1}{N},
\]
and obtains a two-dimensional Fokker–Planck equation,
\[
\frac{\partial p(m,s;t)}{\partial t}
= \frac{\partial}{\partial m}\big[D_1(m,s)\,p\big]
+ \frac{\partial}{\partial s}\big[D_2(m,s)\,p\big]
+ \epsilon\,\frac{\partial^2}{\partial m^2}\big[D_{mm}(m,s)\,p\big]
+ \epsilon\,\frac{\partial^2}{\partial s^2}\big[D_{ss}(m,s)\,p\big]
+ 2\epsilon\,\frac{\partial^2}{\partial m\,\partial s}\big[D_{ms}(m,s)\,p\big]. \tag{13}
\]
The drift terms \(D_1,D_2\) govern deterministic relaxation of magnetization and defect density, while \(D_{mm},D_{ss},D_{ms}\) describe finite-size fluctuations [2511.07207, 2606.25919].

The associated Langevin equation is
\[
d\mathbf{X}_t = \mathbf{D}(\mathbf{X}_t)\,dt + \sqrt{2\epsilon}\,\mathbf{B}(\mathbf{X}_t)\,d\mathbf{W}_t, \tag{14}
\]
with \(\mathbf{X}_t=(m_t,s_t)^{\mathrm T}\) and \(\mathbf{B}\mathbf{B}^{\mathrm T}\) equal to the diffusion matrix derived from the Fokker–Planck coefficients [2511.07207]. This mesoscopic description is sufficient to recover critical slowing down, metastable switching, and first-passage asymptotics.

Near criticality, center manifold theory reduces the deterministic dynamics to an effective one-dimensional evolution for the slow magnetization mode \(m(t)\). Along the second-order line,
\[
\frac{dm}{dt} = \frac{K^2 (K^2 - 3)}{3(1+K^2)}\, m^3 + O(m^5), \tag{15}
\]
which implies
\[
m(t)\sim t^{-1/2} \qquad (t\to\infty), \tag{16}
\]
whereas at the tricritical point the cubic coefficient vanishes and the leading term is quintic,
\[
\frac{dm}{dt}\sim \tilde c\,m^5, \tag{17}
\]
implying
\[
m(t)\sim t^{-1/4}. \tag{18}
\]
These decay laws yield the dynamic decay exponents
\[
\lambda_m=\frac12
\]
on the critical line and
\[
\lambda_m=\frac14
\]
at the tricritical point [2511.07207].

Using the dynamic scaling relation
\[
\lambda_m=\frac{\beta}{\nu z}, \tag{19}
\]
together with the static exponents, one finds in both cases
\[
z=2.
\]
This establishes that the critical dynamics of the Nagle–Kardar model belongs to the **Model A** universality class in the Hohenberg–Halperin classification: purely relaxational, Markovian dynamics with a non-conserved scalar order parameter [2511.07207, 2606.25919].

A common misconception is that the mean-field interaction should force mean-field dynamic behavior beyond Model A. The dynamical analyses instead show that, although the static critical structure is modified by the long-range term and tricriticality appears, the dynamic critical exponent remains \(z=2\) and the macroscopic relaxation mechanism is still that of non-conserved order-parameter dynamics [2511.07207].

## 5. Metastability, first-passage times, and Arrhenius behavior

Because the quasi-potential \(U(m,s)\) may have multiple minima, the Nagle–Kardar model exhibits bistability and metastability in the ordered region [2511.07207]. For \(h=0\), there are typically two symmetric minima at \(\pm m^*\); for \(h\neq 0\), symmetry is broken and one minimum becomes metastable [2511.07207, 2606.25919].

The average first-passage time \(\langle \tau \rangle\) for transitions between locally stable macrostates obeys an Arrhenius-type large-deviation law,
\[
\langle \tau \rangle \propto \exp\big[N\,\Delta U\big], \tag{20}
\]
where \(\Delta U\) is the barrier height between the initial minimum and the relevant saddle point in the quasi-potential landscape [2511.07207]. This is the natural finite-system analogue of activated barrier crossing, but with system size \(N\) multiplying the barrier because \(U\) is a large-deviation rate function.

Microscopic Glauber simulations and mesoscopic Langevin simulations both reproduce this scaling [2606.25919]. For symmetric bistable cases, left-to-right and right-to-left passage times coincide; with an external field they differ because the relevant barriers differ [2511.07207]. This direct correspondence between microscopic dynamics, macroscopic quasi-potential, and activated switching is one of the reasons the model is considered a useful testbed for nonequilibrium coarse-graining [2606.25919].

A plausible implication is that the Nagle–Kardar model occupies an intermediate position between exactly solvable equilibrium mixed-range spin systems and stochastic metastability models: it is simple enough to allow explicit quasi-potentials and coarse-grained drifts, but rich enough to display tricritical scaling and activated phase switching [2606.25919].

## 6. Finite-size fluctuation-induced forces and ensemble dependence

Finite systems of the Nagle–Kardar model display strong ensemble dependence because the long-range interaction makes the model non-additive [2508.15430]. In the grand-canonical ensemble with periodic boundary conditions, the finite-size excess free energy defines a **Casimir force**,
\[
\beta F_{\rm Cas}^{(\zeta)}(L,T,h) = -\frac{\partial}{\partial L} f_{\rm ex}^{(\zeta)}(L,T,h), \tag{21}
\]
while in the canonical ensemble with fixed magnetization, the corresponding excess Helmholtz free energy defines a **Helmholtz force**,
\[
\beta F_{\rm H}^{(\zeta)}(L,T,M) = -\frac{\partial}{\partial L} f_{\rm ex}^{(\zeta)}(L,T,M). \tag{22}
\]
Here \(L=N\) for the one-dimensional ring [2508.15430].

Exact asymptotics show that on the critical line the critical Casimir amplitude is
\[
\Delta_{\rm Cas}^{\rm (cr)}=\frac14,
\]
while at the tricritical point it is
\[
\Delta_{\rm Cas}^{\rm (tr)}=\frac13.
\]
In both cases the corresponding critical Casimir force is **repulsive** near the singular regime [2404.18324, 2508.15430]. This is unusual because periodic boundary conditions are normally associated with attractive critical Casimir forces. The Nagle–Kardar model therefore violates the commonly cited “boundary condition rule” according to which equivalent boundary conditions imply attraction and conflicting ones imply repulsion [2404.18324, 2508.15430].

The Helmholtz force behaves differently. In the canonical ensemble with fixed magnetization, the contribution of the long-range term cancels exactly from the excess free energy, so the Helmholtz force is independent of \(K_l\) and coincides with that of the standard one-dimensional Ising model at fixed magnetization [2508.15430]. It nevertheless changes sign as a function of temperature and magnetization [2508.15430].

This ensemble dependence is not a minor finite-size correction. It directly reflects the non-additivity of the long-range interaction and the inequivalence of thermodynamic potentials in finite systems [2508.15430]. The Nagle–Kardar model has therefore become a controlled setting for comparing fluctuation-induced forces across conjugate ensembles.

## 7. Quantum extension and related nomenclature

A quantum extension of the Nagle–Kardar model adds a transverse field and is defined by
\[
H = -\frac{J}{2N} \sum_{i,j} S_i^z S_j^z
      - \frac{K}{2} \sum_i S_i^z S_{i+1}^z
      - h \sum_i S_i^x, \tag{23}
\]
with \(S_i^{x,z}\) Pauli matrices [2604.25837]. This quantum Nagle–Kardar model retains the competition between nearest-neighbor and mean-field interactions while introducing quantum fluctuations. The thermodynamic analysis shows that ensemble inequivalence present in the classical model is removed above a threshold transverse field \(h_c\): for \(h>h_c\), only second-order transition lines remain and ensemble equivalence is restored [2604.25837]. This suggests that quantum fluctuations can smooth out the first-order structure responsible for inequivalence in the classical model.

A separate terminological issue concerns the phrase “Nagle–Kardar” in the context of anisotropic stochastic growth and self-organized criticality. Some papers describe the continuous Hwa–Kardar model as “Hwa–Kardar (Nagle–Kardar)” [1508.00236]. That model is a stochastic height equation with anisotropic diffusion, nonlinear transport along a preferred direction, and white noise, used for self-organized critical systems such as driven sandpiles [1508.00236]. It is not the same object as the Ising-like mixed-range spin model governed by Hamiltonian (1). The overlap in naming is therefore historical or contextual rather than structural.

This distinction matters because the two model families belong to different research programs. The Ising-like Nagle–Kardar model concerns competing short- and long-range interactions, tricriticality, metastability, ensemble inequivalence, and Glauber dynamics [2511.07207, 2606.25919, 2508.15430]. The Hwa–Kardar model concerns anisotropic stochastic growth, self-organized criticality, and renormalization-group analysis under turbulent advection [1508.00236]. Conflating the two obscures both.

The Nagle–Kardar model, in its classical and quantum forms, is therefore best understood as an exactly tractable mixed-range Ising system whose combination of local domain-wall structure and global ferromagnetic coupling produces tricritical equilibrium behavior, Model A critical dynamics, activated metastability, and pronounced ensemble dependence [2511.07207, 2606.25919, 2508.15430, 2604.25837].

Source: https://www.emergentmind.com/topics/nagle-kardar-model