Dynamical renormalization group analysis of $O(n)$ model in steady shear flow

By incorporating strong anisotropy into a dynamical renormalization group analysis, this study identifies a new stable Gaussian fixed point for the $O(n)$ model under steady shear flow, revealing that shear flow stabilizes long-range order in two dimensions and alters the upper critical dimensions for both conserved and non-conserved order parameters, thereby violating the equilibrium Hohenberg-Mermin-Wagner theorem.

The Gist

Imagine a crowded dance floor where everyone is trying to move in sync. In a calm room (equilibrium), if the music stops, the dancers might freeze in place, or if they try to form a pattern, they might get jostled apart by the sheer number of people bumping into each other. In physics, this is like a material trying to decide whether to be ordered (like a magnet) or disordered (like a gas).

Now, imagine someone starts pushing the entire dance floor sideways, creating a steady "shear flow." The dancers are no longer just bumping randomly; they are being swept along in a specific direction. This paper asks: How does this constant pushing change the way the dancers organize themselves?

The authors, Harukuni Ikeda and Hiroyoshi Nakano, used a sophisticated mathematical tool called the "Renormalization Group" (think of it as a microscope that zooms in and out to see how patterns change at different sizes) to study this. They looked at two types of dancers:

1. Model A: Dancers who can move freely and change their spots easily (non-conserved).
2. Model B: Dancers who are stuck in a grid and can only swap places with neighbors (conserved).

Here are the main discoveries, explained simply:

1. The "Magic" of the Push

In a normal, calm room, there are strict rules about how small a room can be before the dancers can't form a big, organized pattern.

• The Old Rule: In a 2D room (like a flat floor), if the dancers are trying to break symmetry (like choosing a specific direction to face), the Hohenberg-Mermin-Wagner theorem says it's impossible. The random jostling is too strong, and the pattern breaks. You need at least a 3D room for this to work.
• The New Discovery: The authors found that when you apply that steady "push" (shear flow), the rules change completely. The push actually stabilizes the pattern. Even in a flat 2D room, the dancers can now form a perfect, long-range order. The "push" suppresses the chaotic jostling that usually ruins the party.

2. The "New Normal" (The Fixed Point)

In physics, systems often settle into a "fixed point"—a state where the rules of the game stop changing no matter how much you zoom in or out.

• Without the push: The system tries to settle into a "Gaussian fixed point" (a standard, predictable state), but the push makes this state unstable. It's like trying to balance a pencil on its tip while someone is shaking the table.
• With the push: The authors found a new, stable fixed point. Because the push is so strong, the system finds a new way to balance. This new state is "Gaussian" (simple and predictable), but it behaves very differently from the calm state.

3. The Dimensions Shrink

The paper introduces two critical numbers:

• Upper Critical Dimension ($d_{up}$): The size of the room where the "simple" rules (mean-field theory) start working perfectly.

  • Before: You needed a 4D room for the simple rules to work.
  • After: With the push, the simple rules work even in a 2D room (for Model A) and even in a 0D room (for Model B, which implies they work everywhere).
  • Analogy: It's as if the push makes the dancers so coordinated that they act like they are in a much larger, simpler world, even when they are in a tiny, cramped space.

• Lower Critical Dimension ($d_{low}$): The smallest room size where order is possible.

  • Before: You needed a room larger than 2D to have order.
  • After: With the push, order is possible even if the room is smaller than 2D (the math says $d_{low} < 2$).
  • Analogy: The push is so effective at organizing the crowd that they can stay in line even in a hallway that is too narrow for them to stand normally.

4. The "Stretching" Effect

The most interesting visual change is how the dancers move.

• In a calm room: If you look at the distance between dancers, it's the same in all directions (isotropic).
• In the push: The dancers stretch out. Along the direction of the push, they become very long and thin; perpendicular to the push, they stay short.
• The Result: The "correlation" (how much one dancer's move predicts another's) changes. In the direction of the push, the connection becomes weaker and follows a strange, fractional power law (like $1/|q|^{2/3}$ instead of the usual $1/|q|^2$). It's like the dancers are holding hands in a long, stretched-out chain rather than a tight circle.

5. Why Previous Experiments Were Confused

The authors mention that computer simulations in the past gave confusing results. Some said the order parameter (how organized the group is) was 0.37, others said 0.48, and the "simple" theory predicted 0.5.

• The Explanation: The authors suggest that the "stretching" (anisotropy) is so extreme that standard computer simulations weren't big enough to see the true pattern.
• The Analogy: Imagine trying to photograph a very long, thin snake. If your camera frame is square, you might cut off the tail or the head, making it look like a short, stubby worm. To see the whole snake, you need a camera that is 100 times wider than it is tall. The authors argue that previous simulations used "square cameras" on a "snake-like" system, leading to wrong measurements.

Summary

This paper claims that steady shear flow acts like a powerful organizer. It breaks the old rules of physics that said "you can't have order in 2D." Instead, the flow creates a new, stable state where order is easier to achieve, the rules become simpler (mean-field), and the system stretches out dramatically along the flow direction. The authors believe this explains why some experiments see "mean-field" behavior and why others get confused—they just haven't accounted for this extreme stretching.

Technical Summary

Problem Statement
The critical behavior of the $O(n)$ model under steady shear flow remains an open problem in non-equilibrium statistical mechanics. While equilibrium critical phenomena are well-understood via renormalization group (RG) methods, the introduction of shear flow breaks rotational symmetry and drives the system out of equilibrium. Previous theoretical work, such as De Gennes' mean-field analysis and Onuki and Kawasaki's RG study of Model H, suggested that shear flow suppresses critical fluctuations along the flow direction and potentially alters critical dimensions. However, numerical simulations of sheared Ising and $O(n)$ models have yielded conflicting results regarding critical exponents (specifically $\beta$) and the existence of long-range order in two dimensions ($d=2$). A key theoretical challenge has been the lack of a stable fixed point in RG analyses that correctly incorporates the strong anisotropy induced by shear flow, leading to discrepancies between theoretical predictions and numerical observations.

Methodology
The authors employ a dynamical renormalization group (RG) analysis tailored for anisotropic systems. They apply this framework to the $O(n)$ model in steady shear flow for two distinct cases:

1. Model A: Non-conserved order parameter dynamics.
2. Model B: Conserved order parameter dynamics.

The core methodological innovation lies in the adoption of an anisotropic scaling ansatz. Unlike earlier analyses that assumed isotropic scaling or treated anisotropy only through renormalized transport coefficients, this work explicitly assumes different scaling exponents for coordinates parallel ($x_1$) and perpendicular ($x_\perp$) to the flow direction. The scaling transformations are defined as:
$$x_1 = l^\zeta x'_1, \quad x_\perp = l x'_\perp, \quad t = l^z t', \quad \phi_a = l^\chi \phi'_a$$
where $\zeta$, $z$, and $\chi$ are independent critical exponents. The authors derive RG flow equations for the shear rate ($\dot{\gamma}$), transport coefficients, noise strength, and interaction terms by requiring the scale invariance of specific physical quantities (e.g., shear rate and noise strength) to identify stable Gaussian fixed points.

Key Contributions and Results
The primary contribution of the paper is the identification of a new stable Gaussian fixed point that exists specifically under the condition of steady shear flow. This fixed point is stable against the shear rate, unlike the equilibrium Gaussian fixed point which is destabilized by shear in any dimension.

The analysis yields the following specific results:

• Critical Dimensions:

  • Upper Critical Dimension ($d_{up}$): The paper finds that $d_{up} = 2$ for Model A and $d_{up} = 0$ for Model B. This implies that mean-field critical exponents are observed in physical dimensions $d=2$ and $d=3$ for both models.
  • Lower Critical Dimension ($d_{low}$): The analysis reveals $d_{low} = 0$ for Model A and $d_{low} = -2$ for Model B. Crucially, this indicates that continuous symmetry breaking (and thus long-range order) can occur even in $d=2$, a regime where the Hohenberg-Mermin-Wagner theorem prohibits such order in equilibrium systems.

• Critical Exponents:

  • For Model A, the exponents are $\zeta=3$, $z=2$, and $\chi = -d/2$. The order parameter scaling exponent is negative for all $d \geq 2$, confirming the stabilization of long-range order.
  • For Model B, the exponents are $\zeta=5$, $z=4$, and $\chi = -(d+2)/2$.
  • In both cases, the critical exponent for the order parameter is $\beta = 1/2$, consistent with mean-field theory.

• Correlation Functions:

  • The correlation functions exhibit distinct anisotropic scaling. In Fourier space, the correlation function along the flow direction ($q_1$) scales as $C(q_1) \sim |q_1|^{-2/3}$ for Model A and $C(q_1) \sim |q_1|^{-2/5}$ for Model B.
  • These fractional powers (smaller than the equilibrium value of 2) indicate a stronger suppression of critical fluctuations along the flow direction compared to equilibrium. The suppression is more pronounced in Model B than in Model A.

Significance and Claims
The paper claims that the resolution of the "controversy" surrounding mean-field characters in sheared models stems from the correct treatment of anisotropy in the scaling ansatz. By identifying a stable Gaussian fixed point with these specific anisotropic exponents, the authors provide a theoretical basis for:

1. The observation of mean-field exponents ($\beta=1/2$) in $d=2$ and $d=3$.
2. The existence of long-range order in $d=2$ for continuous symmetry breaking, which is impossible in equilibrium.
3. The explanation of why previous numerical simulations might have struggled to converge to these values: the large anisotropy exponent (particularly $\zeta=5$ for Model B) implies that simulations require extreme aspect ratios ($L_\parallel \sim L_\perp^5$) to capture the correct scaling, a condition often unmet in standard finite-size scaling analyses.

The authors position this work as a necessary step to reconcile theoretical RG predictions with numerical results, suggesting that the "missing" stable fixed point in earlier analyses was due to the neglect of strong anisotropy in the scaling hypothesis. They explicitly note that their results for Model A and B differ from the infinite shear-rate limit of Model H derived by Onuki and Kawasaki, attributing this to the different theoretical formalisms (explicit anisotropic scaling vs. isotropic scaling with anisotropic coefficients) and suggesting future work to bridge these frameworks.