Abstract
In this paper, we investigate the stability of the 2-dimensional (2D) Taylor–Couette (TC) flow for the incompressible Navier–Stokes equations. The explicit form of velocity for 2D TC flow is given by \(u=(Ar+\frac{B}{r})(-\sin \theta , \cos \theta )^T\) with \((r, \theta )\in [1, R]\times \mathbb {S}^1\) being an annulus and A, B being constants. Here, A, B encode the rotational effect and R is the ratio of the outer and inner radii of the annular region. Our focus is the long-term behavior of solutions around the steady 2D TC flow. While the laminar solution is known to be a global attractor for 2D channel flows and plane flows, it is unclear whether this is still true for rotating flows with curved geometries. In this article, we prove that the 2D Taylor–Couette flow is asymptotically stable, even at high Reynolds number (\(Re\sim \nu ^{-1}\)), with a sharp exponential decay rate of \(\exp (-\nu ^{\frac{1}{3}}|B|^{\frac{2}{3}}R^{-2}t)\) as long as the initial perturbation is less than or equal to \(\nu ^\frac{1}{2} |B|^{\frac{1}{2}}R^{-2}\) in Sobolev space. The powers of \(\nu \) and B in this decay estimate are optimal. It is derived using the method of resolvent estimates and is commonly recognized as the enhanced dissipative effect. Compared to the Couette flow, the enhanced dissipation of the rotating Taylor–Couette flow not only depends on the Reynolds number but also reflects the rotational aspect via the rotational coefficient B. The larger the |B|, the faster the long-time dissipation takes effect. We also conduct space-time estimates describing inviscid-damping mechanism in our proof. To obtain these inviscid-damping estimates, we find and construct a new set of explicit orthonormal basis of the weighted eigenfunctions for the Laplace operators corresponding to the circular flows. These provide new insights into the mathematical understanding of the 2D Taylor–Couette flows.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
Reynolds’s famous experiment [27] inspired the study of hydrodynamic stability at high Reynolds number. In this regime, the laminer flows could become unstable and transition to turbulence [10, 29, 32, 36].
With the low Reynolds number, Serrin [30] demonstrated that all equilibria of the forced Navier–Stokes equation on bounded domains are linearly stable. At high Reynolds number, even in the absence of boundaries, the viscosity can significantly complicate the linear problem.
In this paper, we study the 2D (two-dimensional) incompressible Navier–Stokes (NS) equations:
where \(v=(v_1,v_2)\in \mathbb {R}^2\) is the fluid velocity, \(x=(x_1,x_2)\in \Omega \) represents the space variables, \(\Omega \in \mathbb {R}^2\) is an annular region and \(t\ge 0\) represents the time variable. The unknowns in the equation are the velocity field \(v(t,x)=(v^1(t,x),v^2(t,x))\) and the pressure \(p(t,x)\in \mathbb {R}\). The constant \(\nu >0\) is known as the kinematic viscosity, which is very small in our paper. And the Reynolds number Re is proportional to the inverse of \(\nu \).
In the 2D case, the vorticity field
is a scalar. By taking the 2D curl of the Navier–Stokes equations, Eq. (1.2) can be transformed into its vorticity formulation:
In this paper, we employ (1.3) to study the asymptotics of solutions to (1.1) around a Taylor–Couette (TC) flow.
1.1 Taylor–Couette flow
TC flow describes a steady circular solution of the viscous fluid bounded between two rotating infinitely long coaxial cylinders. The understanding of solutions’ asymptotics around a TC flow has board applications, including desalination, magnetohydrodynamics and viscometric analysis. Despite being a simple type of rotating solutions, perturbation of the TC flow has proven to be a challenging subject and has been extensively studied experimentally, theoretically, and numerically for a long time. Many questions related to TC flow remain unanswered, making it active research field in fluid mechanics [8, 11, 18, 26]. However, rigorous mathematical proofs in this field are still insufficient, even for the 2D case.
In the following, we adopt the convention \(r = |x|\) and denote the radial vorticity \(\omega (x) = \omega (r)\) and stream function \(\phi (x) = \phi (r)\). In 2D, the vorticity and the velocity field then take the form
We first derive the explicit form of the steady TC flow. First noting that if \(\omega =const\), then \(\omega \) is a solution to the 2D NS equation (1.3). For this case, the stream function \(\phi (r)\) can be determined from (1.4), which yields
By employing the polar coordinates, one can relabel \(v(x_1,x_2)\) and \(\omega (r)\) as \(U(r,\theta )\) and \(\Omega (r)\), respectively. This allows us to solve Eq. (1.5) and obtain the expressions for U and \(\Omega \) as follows:
Here A, B are constants and spatial variables \((r, \theta )\) belong to a domain \(\mathcal {D}=[1,R]\times \mathbb {S}^1\). The function \(U(r,\theta )\) in (1.6) is hence a steady state for the 2D incompressible NS equation (1.1), which is commonly called the Taylor–Couette flow.
1.2 Hydrodynamic stability at high Reynolds number
Reynolds well-known experiment [27] revealed that small perturbations can cause significant change of the flow due to the nonlinear nature of the Navier–Stokes equations. In experiments [9, 31] for the TC flow, a small perturbation could lead to instability and transition to turbulence at high Reynolds numbers. This phenomenon is called the subcritical transition, which is a central topic in fluid mechanics.
Numerous efforts have been made to comprehend the subcritical transition mechanism [6]. Kelvin initially suggested that as the Reynolds number \(Re\rightarrow \infty \), the basin of attraction for the laminar flow diminishes, allowing the flow to become nonlinearly unstable for small but finite perturbations [20]. In [32], Trefethen, Trefethen, Reddy and Driscoll proposed a way to determine the threshold amplitude by considering a perturbation of size \(Re^{-\beta }\) with \(\beta \ge 0\) and \(Re\rightarrow +\infty \). In other words, they aimed to identify the lowest possible value of \(\beta \), which could result in a transition to turbulence with a perturbation of size \(O(Re^{-\beta })\). Bedrossian, Germain, and Masmoudi [2] presented a rigorous mathematical formulation of this approach using the fact that \(Re^{-1}\sim \nu \).
Given a norm \(\Vert \cdot \Vert _X\), we now hope to determine a nonnegative number \(\beta =\beta (X)\) so that
The exponent \(\beta \) is referred to as the transition threshold in the applied literature. Significant advancements and findings related to the subcritical transition regime for the 2D incompressible NS equations have emerged recently. These developments include researches on Couette flow without boundaries [2, 4, 5, 23, 24] and with boundaries [3, 7], on Lamb–Oseen vortices [12,13,14,15, 21] and on periodic Kolmogorov flow [34, 35]. Recently, Guo, Pausader and Widmayer [16] has made significant progress in the stability problem of the Euler flow with rotation.
1.3 Main theorems
The aim of this paper is to explore both linear and nonlinear stability mechanism of the 2D Taylor–Couette flow using rigorous analytical tools. Specifically, we prove nonlinear asymptotic stability of the TC flows for the 2D NS equations. Furthermore, we identify and trace a critical parameter reflecting the rotating properties of this system.
Our focus in this paper is on the stability of the TC flow, hence we study the below perturbation equation. Setting \(w = \omega - \Omega \), \(u = v - U\), with \(\varphi \) being the stream function satisfying \(\Delta \varphi =w\) and \(u=(-\partial _2\varphi ,\partial _1\varphi )\), we concert (1.3) to polar coordinates:
Remark 1.1
Note that constants A and B in (1.6), (1.7) and constant R in (1.7) will serve as parameters in our later arguments. In particular, B will correspond to the rotating effect.
Our main results are summarized below.
Theorem 1.1
The fully nonlinear two-dimensional incompressible Navier–Stokes equations exhibit asymptotic stability around the Taylor–Couette flow under perturbations of size \(\nu ^\frac{1}{2} |B|^{\frac{1}{2}}R^{-2}\) in the \(H^1\) space, with the vorticity controllable at any time by the initial data.
The more precise mathematical statements are given in the following two theorems. The first main result can be directly derived from Proposition 4.3 and Lemma 5.4, and describes the asymptotic stability of the linearized TC flow.
Theorem 1.2
There exist constants \(C,c>0\) being independent of \(\nu ,A,B,R\), such that the solution w for the linear system
exists globally in time and for any \(t\ge 0\). Furthermore, the following stability estimates hold
where \(\bar{w}(t)=\frac{1}{2\pi }\int _0^{2\pi }w(t,r,\theta )d\theta \).
The second main result describes the asymptotic stability of the nonlinear Navier–Stokes equations around the TC flow. Reynolds experiments tell us that, with small viscosity coefficients, even tiny initial perturbations could cause the flow to be chaotic. Therefore, in order to establish the nonlinear asymptotic stability of Navier–Stokes equations, the initial perturbations must be restricted to a certain range with respect to the viscosity coefficient \(\nu \). The specific formulation of the result is as follows.
Theorem 1.3
Assume that \(0<\log R\lesssim \nu ^{-\frac{1}{3}}|B|^{\frac{1}{3}}\). There exist constants \(\nu _0\) and \(c_0, C,c'>0\) independent of \(\nu ,A,B,R\) such that for any \(0<\nu \le \nu _0\), if the initial data satisfies
then the solution \(w(t,r,\theta )\) to the system (1.7) is global in time. Moreover, the following stability estimates hold
where \(\bar{w}(t)=\frac{1}{2\pi }\int _0^{2\pi }w(t,r,\theta )d\theta \).
By setting the radius R to a constant value, such as \(R=2\), we can immediately derive the following corollary from Theorem 1.3.
Theorem 1.4
There exist constants \(\nu _0\) and \(c_0, C,c'>0\) independent of \(\nu ,A,B\) so that for any \(0<\nu \le \nu _0\) and \(|B|\ge \nu _0\), if
then the system (1.7) admits a global-in-time solution \(w(t,r,\theta )\) satisfying the following stability estimates
where \(\bar{w}(t)=\frac{1}{2\pi }\int _0^{2\pi }w(t,r,\theta )d\theta \).
1.4 Difficulties, new ingredients and the sketch of the proof
1.4.1 Rotational effect and enhanced dissipation
Compared to Couette flow, the Taylor–Couette flow involves additional coefficients A and B that account for rotational effects. Both Theorem 1.2 and Theorem 1.3 demonstrate that for both linearized and nonlinear equations, only the rotation coefficient B affects the stability of the system, and the energy of vorticity is irrelevant to the coefficient A.
The impact of the rotation coefficient B on stability is mainly manifested in the so-called enhanced dissipation effect. The decay rate of the heat equation is \(e^{-\nu R^{-2} t}\), but in the presence of B, the dissipation rate becomes faster and can be described by \(e^{-c\nu ^{\frac{1}{3}}|B|^{\frac{2}{3}}R^{-2}t}\) (Notice that we are interested in the regime when \(0<\nu \ll 1\)). This implies that while the heat equation exhibits decay only after the time scale of \(\nu ^{-1}\), the antisymmetric part \(ik\frac{B}{r^2}\) in the linearized equation results in the energy decay after a shorter time scale of \(\nu ^{-\frac{1}{3}}|B|^{-\frac{2}{3}}R^{2}\). This phenomenon is known as the enhanced dissipation effect. A shear or a diffusion averaging mechanism to trigger it were investigated in [22, 28]. Here we prove this effect in a different setting with rotations.
Theorem 1.2 and Theorem 1.3 both indicate that the enhanced dissipation effect becomes stronger as |B| increases, which is quantified by a faster asymptotic decay rate \(e^{-c\nu ^{\frac{1}{3}}|B|^{\frac{2}{3}}R^{-2}t}\). This enhanced dissipation has a timescale of \(t\sim \nu ^{-\frac{1}{3}}|B|^{-\frac{2}{3}}R^{2}\) around the Taylor–Couette flow, with \(t\sim \nu ^{-\frac{1}{3}}\) being consistent with Couette flow [2,3,4,5, 7, 23, 24].
The enhanced dissipation decay rate \(e^{-c\nu ^{\frac{1}{3}}|B|^{\frac{2}{3}}R^{-2}t}\) for the TC flow in this paper is optimal. In Sect. 3, we derived the sharp resolvent estimates, which are also known as the optimal pseudospectral bound. It is important to note that although the enhanced dissipation decaying rates for the TC flow (\(\nu ^{\frac{1}{3}}|B|^{\frac{2}{3}}R^{-2}\)) and Couette flow (\(\nu ^{\frac{1}{3}}\)) have the same power of \(\nu \), it does not suggest that the corresponding enhanced dissipation can be directly derived through a scaling argument from results of Couette flow. This is because the (normalized) antisymmetric parts of linearized operators for these two flows are still different. Furthermore, with the aid of dimensional analysis or scaling transformations, it can be shown that the sum of the power of \(\nu \) and B is always 1. To see this, assume that the decay rate is in the form of \(\exp (-c\nu ^{\alpha }|B|^{\beta }R^{\gamma }t)\). Noting that the dimension of \(\nu , B\) and R are \(\textsf{T}^{-1}\textsf{L}^2, \textsf{T}^{-1}\textsf{L}^2\) and L, respectively (Here \(\textsf{T}\) represents the time and \(\textsf{L}\) denotes the length). Since the exponent \(\nu ^{\alpha }|B|^{\beta }R^{\gamma }t=\textsf{T}^{1-\alpha -\beta } \textsf{L}^{2\alpha +2\beta +\gamma }\) is dimensionless, it must hold \(\alpha +\beta =1\) and \(\gamma =-2\). Thus, the factor \(|B|^{\frac{2}{3}}\) is also optimal.
Additionally, inspired by the approach used to handle Oseen vortices [13, 15, 21], in [1] the authors employed self-similar variables and derived an enhanced dissipation decay rate that is independent of the outer radius R.
1.4.2 Comparison with Couette flow from an operator perspective
Previous studies have investigated plane Couette flow with \((x,y)\in \mathbb {T}\times \mathbb {R}\) ( see [2, 4, 5, 23, 24]) and Couette flow in a finite channel with \((x,y)\in \mathbb {T}\times [-1,1]\) (see [3, 7]). In these works, the spatial variable x corresponding to the non-shear direction was defined on a torus. However, for the TC flow in this paper, the spatial domain is an annulus region \((r,\theta )\in [1,R]\times \mathbb {S}^1\).
As both x in Couette flow and \(\theta \) in TC flow are defined on a torus or \(\mathbb {S}^1\), it is natural to apply Fourier transform on the x or \(\theta \) direction. The corresponding linearized equations around the Couette flow and TC flow are given as below.
The linearized operators for the Couette flow and TC flow take the different forms:
These two operators exhibit different enhanced dissipation rates of \(\nu ^{\frac{1}{3}}\) and \(\nu ^{\frac{1}{3}}|B|^{\frac{2}{3}}R^{-2}\), respectively.
It can be observed that both the symmetric and antisymmetric parts of the linearized operator for the TC flow have different structures from those of the Couette flow. In the following discussion, we will demonstrate how these differences affect our results and the corresponding proofs for the TC flow.
In previous studies of the planar Couette flow [2, 4, 5, 23, 24], where the shear variable y is defined over the entire space \(\mathbb {R}\), mathematicians employed the Fourier transform in y to obtain ordinary differential equations with respect to time t. However, in the case of TC flow, the radial variable r is defined in the bounded domain [1, R] with boundaries, which makes it inconvenient to conduct the Fourier transform in r directly. Instead, we adopt the method of resolvent estimates to derive enhanced dissipation, with detailed results and proofs provided in Sect. 3. Note that this method has also been employed by Chen, Li, Wei and Zhang to study the Couette flow within a finite channel in [7].
1.4.3 Resolvent estimates
In Sect. 3, our key results are the resolvent estimates presented in Proposition 3.2. Define
and denote
Our initial goal is the following inequality:
which is a resolvent estimate from \(L^2\) space to \(L^2\) space, as given in Lemma 3.3. This estimate corresponds to the estimate obtained for the Couette flow in [7]:
However, due to the specific structure of the TC flow, the corresponding resolvent estimate is weighted in r. If we take \(c'\) as \(\frac{1}{2C}\), the inequality (1.9) directly yields the following estimate:
which takes the form of a resolvent estimate for \(rF-c'\nu ^{\frac{1}{3}} |kB|^{\frac{2}{3}}\frac{w}{r}\). Notice that \(rF-c'\nu ^{\frac{1}{3}} |kB|^{\frac{2}{3}}\frac{w}{r}\) can be written as
Here, \(rL_kr-c'\nu ^{\frac{1}{3}} |kB|^{\frac{2}{3}}\) represents the translation of \(rL_kr\) to the left by \(c'\nu ^{\frac{1}{3}} |kB|^{\frac{2}{3}}\), which leads to the enhanced dissipation rate for the following linear evolution equation
This allows us to obtain the dissipative factor \(e^{-c\nu ^{\frac{1}{3}} |kB|^{\frac{2}{3}}R^{-2}t}\) in space-time estimates for the fully nonlinear system, as shown in detail in Proposition 4.5.
Next, we derive the resolvent estimate for the stream function \(\varphi \):
The proof of this estimate relies on the resolvent estimate for \(\Vert w\Vert _{L^1}\), and its details are given in Lemma 3.4.
Finally, in Lemma 3.5 and Lemma 3.6 we establish the below estimates for the \(H^1\) norms of w and \(\varphi \), which are controlled by the \(H^{-1}\) norms of the resolvent equations, respectively:
The resolvent estimates presented above are crucial for our analysis, and are summarized in Proposition 3.2. Notably, all of the estimates in Proposition 3.2 have been carefully derived in preparation for the subsequent discussion in Sect. 4.
Based on the above conclusions, we summarize the following table which illustrates the relation between coefficients of the terms on the left of resolvent estimates and the norm of \(F-c'\nu ^{\frac{1}{3}}|kB|^{\frac{2}{3}}R^{-2}w\) on the right (Table 1).
The above powers of \(\nu \) in the coefficients are optimal and consistent with the 2D Couette flow in [7]. However, in this paper the coefficients are also dependent on B and R, and the powers of B are also sharp. As shown in the table, the coefficients on the left will be multiplied by \((\frac{\nu }{|kB|})^{\frac{1}{3}}\) if the regularity of \(F-c'\nu ^{\frac{1}{3}}|kB|^{\frac{2}{3}}R^{-2}w\) drops from \(L^2\) to \(H^{-1}_{r}\).
1.4.4 Pseudospectrum and enhanced dissipation
In the Sect. 4, the pseudospectral bound of L is defined as
Recall that in Proposition 3.2 we establish the following resolvent estimates
Since \(r\in [1, R]\), we have
This provides a lower bound for the pseudospectrum of \(L_k\):
Applying the Gearhart-Pr\(\ddot{u}\)ss type lemma established by Wei in [33] and by Helffer-Sjöstrand in [17], we can obtain the pointwise estimate of the semigroup bound in Proposition 4.3
where \(w_k^l(t)\) satisfies the homogeneous linear equation
It is worth noting that here, the exponential decay factor \(e^{-c(\nu k^2)^{\frac{1}{3}}|B|^{\frac{2}{3}}R^{-2}t}\) reflects the enhanced dissipation effect.
1.4.5 Integrated invisid damping
The second part of Sect. 4 presents estimations for the stream function \(\varphi ^l_k\) governed by the initial vorticity. Here \(\varphi _k^l\) satisfies the linear elliptic equation
The precise estimate is given in Lemma 4.6, which states that, for any \(k\in \mathbb {Z}\) and \(|k|\ge 1\), it holds
Noting that \(\nu \) is much smaller than 1, and B is a fixed constant, the leading terms on the right of above inequality are
Thus, both the coefficients in front of \(\varphi _k^l\) terms on the left and the main initial-vorticity terms on the right are independent of the viscosity coefficient \(\nu \). This phenomenon is commonly refer to as the effect of inviscid damping. Therefore, Lemma 4.6 presents a space-time version of linear inviscid damping around 2D TC flow.
To prove this result, in Lemma 4.8 we find and formulate a new set of explicit orthonormal basis of (weighted) eigenfunctions corresponding to the Laplace operator for circular flows, namely \(-\partial _r^2+(k^2-\frac{1}{4})/r^2\). This basis can be represented in the closed form
and satisfies the equation
Note that there is a weight \(1/r^2\) on the right and \(-(\frac{l\pi }{\log R} )^2-k^2\) is not the canonical eigenvalue. As far as we know, our explicit constructions here are new. With this basis, we move to prove a crucial estimate
where \(k\in \mathbb {Z}\backslash \{0 \}\), B and R are constants with \(R>1\), \(w_k(0) \in L^2([1, R])\), and \(\tilde{\varphi }^{l_0}_k\) is defined as the solution to the elliptic equations below:
Notice that the Laplace operator for the 2D Couette flow is given by \(-\partial _y^2+k^2\) within the domain \(y\in [-1,1]\). To find corresponding orthonormal eigenfunctions, it is natural to introduce the set \(\{\sin \left( \pi j(y+1)/2\right) \}_{j=1}^{\infty }\), as described in [7]. However, for the circular Laplace operator \(\partial _r^2-\frac{k^2-\frac{1}{4}}{r^2}\), we need to construct a set of more complicated eigenfunctions. The main issue is that if we directly consider the following eigenvalue problem:
it can be shown that there is no closed-form solution. In this paper, we instead solve the weighted eigenvalue problem:
with \(w(r)=\frac{1}{r^2}\). Utilizing change of variables, we obtain a family of closed-form solutions to this boundary value problem, which can be written in the following explicit expressions
These functions form an orthonormal basis of the weighted space \(L^2_w([1, R])\). Therefore, with the aid of \(\{ \psi _l \}_{l=1}^{\infty }\), we can evaluate
where \(\tilde{w}^{l_0}_k{:}{=}e^{-i kB\frac{t}{r^2}}w_k(0)\). Notice that
Thus, employing Fourier transform and Plancherel’s formula, the desired estimate (1.11) follows.
It is worth noting that Lemma 4.6 also plays an important role in improving the nonlinear transition threshold of Theorem 1.3.
1.4.6 Space-time estimates for the linearized Navier–Stokes equations
In Proposition 5.1 of Sect. 5, we establish the space-time estimates for the linearized 2D Navier–Stokes equation written in the vorticity formulation (2.2):
with the stream function \(\varphi _k\) satisfying
To achieve so, we employ the estimates in Sects. 3 and 4.
(1) The space-time estimate for the zero frequency part (i.e. \(k=0\)) can be directly obtained in view of the equation structure and integration by parts, as detailed in Lemma 5.4.
(2) For the nonzero frequency part with \(k\in \mathbb {Z}\backslash \{0 \}\), we decompose the solution \(w_k=w_k^l+w_k^{n}\), where \(w_k^l\) obeys the homogeneous linear equation
while \(w_k^{n}\) is the solution to the inhomogeneous linear equation with zero initial condition
Here we denote the nonlinear forms by \(f_1\) and \(f_2\), which are expressed as
As a result, we also decompose \(\varphi _k\) as \(\varphi _k=\varphi ^{l}_k+\varphi ^{n}_k\), where \(\varphi _k^l\) and \(\varphi _k^n\) fulfill
and
For our space-time estimate, the first step is to provide a control of \(w_k^l(t)\) based on the initial data \(w_k^l(0)\). With the aid of the pseudospectrum bounds and enhanced dissipation effect proved in Sect. 3, we establish the following inequality in Lemma 4.4, which holds for any \(k\in \mathbb {Z}\) and \(|k|\ge 1\):
To proceed, we derive an integrated linear invisid damping estimate, in which \(\varphi _k^l\) can be controlled with the initial data of the vorticity
as detailed in Lemma 4.6. Note that \(\nu \) is much smaller than 1, and B is a fixed constant, so the initial vorticity has an upper bound independent of the viscosity coefficient \(\nu \). By combining this observation with the estimates for \(w_k^l\), we obtain the following inequality
It is worth mentioning that the coefficients in front of the terms on both sides of the inequality are independent of viscosity. Therefore, similar estimates can also hold for the inviscid flow (i.e. \(\nu =0\)).
By applying the resolvent estimates from Proposition 3.2, we then arrive at the conclusion stated in Proposition 4.5 for the nonlinear part, which involves the inhomogeneous linear equation with zero initial data. This result allows us to bound \(w_k^n\) and \(\varphi ^{n}_k\) by the nonlinear terms \(f_1\) and \(f_2\) as follows
Finally, by combining the aforementioned three estimates, we derive the crucial proposition of Sect. 5, namely Proposition 5.1. This proposition states that the following inequality holds
We refer to it as the space-time estimates for the vorticity \(w_k\) and the stream function \(\varphi _k\).
1.4.7 Nonlinear stability
Finally, we close the energy estimate with a bootstrap argument. By utilizing the space-time estimates established in Proposition 5.1 from Sect. 5, we derive the result of Proposition 6.1:
Here, we define \(\mu _k{:}{=}\max \left\{ (\nu k^2)^{\frac{1}{3}} |B|^{\frac{2}{3}} R^{-2}, \nu k^2 R^{-2} \right\} \), as it is necessary to use distinct energy estimates for the high-frequency case of \(\nu k^2\ge |B|\) and the low-frequency case of \(\nu k^2\le |B|\) separately. In the case of 2D Couette flow [7], the discussions on different frequencies k correspond to \(\nu k^2\ge 1\) and \(\nu k^2\le 1\). It is important to mention that the frequency classification of k for the 2D TC flow also relies on the rotation parameter B.
We then define the energy functional \(E_k\) as follows:
According to Proposition 6.1, it can be infered that for \(k\in \mathbb {Z}\backslash \{0 \}\), it holds
Utilizing Lemma 6.3, we can control nonlinear forms \(f_1, f_2\) by
and
This allows us to obtain Lemma 6.4, which states that
Simultaneously, we can derive analogous estimates for \(E_0\) by introducing \(\mathcal {M}_0(0)=\Vert w_0(0)\Vert _{L^2}\). Incorporating all estimates above through a bootstrap argument, we finally arrive at the main theorem of this paper: Theorem 6.1.
1.4.8 Subcritical transition and transition threshold
Theorem 1.2 and Theorem 1.3 follow from Proposition 4.3 and Theorem 6.1, respectively. Here are some further remarks regarding these two main theorems.
On one hand, Theorem 1.2 shows that the linearized Navier–Stokes equations around the 2D Taylor–Couette flow is dynamically stable at any Reynolds number (including the inviscid case), and the vorticity at any time can be controlled by the given initial value in \(L^2\) norm. On the other hand, for the full Navier–Stokes equation, we only expect nonlinear asymptotic stability within the range of small perturbation of size \(\nu ^{\beta (s)}(\beta (s)>0)\) in \(H^s\) space. This proved in our Theorem 1.3, provided the initial perturbation does not exceed \(\nu ^\frac{1}{2} |B|^{\frac{1}{2}}R^{-2}\). The exponent \(\frac{1}{2}\) of \(\nu \) agrees with that of the 2D Couette flow [5, 7], and provides an upper bound \(\frac{1}{2}\) for the corresponding transition threshold \(\beta \). The dependency of the subcritical transition on the rotational speed B is captured by \(|B|^{\frac{1}{2}}\). This indicates that the system can achieve global stability under a wider range of initial data via taking account of the rotating effect.
2 Derivation of the Perturbative Equation
We take the Fourier transform in the \(\theta \) direction and denote the Fourier coefficients of w and \(\varphi \) by \(\hat{w}_k\) and \(\hat{\varphi }_k\), respectively. Using this notation, the Eq. (1.7) can be rewritten as follows:
with \(\hat{\varphi }_k\) satisfying \((\partial _r^2+\frac{1}{r}\partial _r-\frac{k^2}{r^2})\hat{\varphi }_k=\hat{w}_k\).
To eliminate the first derivative \(\frac{1}{r}\partial _r\), we introduce the weight \(r^{\frac{1}{2}}\) and define \(w_k\) and \(\varphi _k\) as \(w_k{:}{=}r^{\frac{1}{2}}e^{ikAt}\hat{w}_k\) and \(\varphi _k{:}{=}r^{\frac{1}{2}} e^{ikAt}\hat{\varphi }_k\). As a consequence, the Eq. (2.1) is transferred into
where \(\varphi _k\) satisfies
Note that the Navier boundary condition
implies that \(\partial _t w_k\) and the nonlinear terms in (2.2) vanish on the boundary of interval [1, R]. This forces the second derivative of \(w_k\) to be zero on the boundary, i.e.
Up to this point, we convert our problem to exploring the dynamics and long-time behaviors of (2.2).
Remark 2.1
In below, for two quantities A and B, we frequently use \(A\lesssim B\) in short to stand for the inequality \(A\le C B\) with some universal constant \(C>0\), that is independent of \(\nu , k, B, \lambda \) and R. Additionally, we also write \(A\approx B\) to indicate that both \(A\lesssim B\) and \(B\lesssim A\) are true.
3 Resolvent Estimates
To establish decays of the linearized equation of (2.2) as below
the key step is to derive the resolvent estimates that will be presented in Proposition 3.2. More precisely, we want to study the resolvent equation subject to the following (Navier) boundary condition for any \(\lambda \in \mathbb {R}\). After taking the Fourier transform with respect to time t, the resolvent equation becomes
The domain of the operator is defined as
Note that for any \(|k|\ge 1\), it holds
We also introduce
Here \(\langle \, \ \rangle \) represents the canonical inner product in \(L^2(\mathbb {R}_{+},dr)\).
In this section, we will present the resolvent estimates for \(w'\), \(\frac{w}{r}\), \(\varphi '\) and \(\frac{\varphi }{r}\) with respect to F in both \(L^2\) norm and \(H_r^{-1}\) norm.
3.1 Coercive estimates of the real part
We start with the coercive estimates for the real part of \( \langle F, w \rangle \), which will be used to obtain the desired resolvent estimates for w.
Lemma 3.1
For any \(|k|\ge 1\) and \(w\in D_k\), it holds
Proof
Via integration by parts, one can directly check
\(\square \)
3.2 Resolvent estimates
For Eq. (3.1), this subsection aims to establish upper bounds for w and \(\varphi \) in terms of F. We prove the following inequalities.
Proposition 3.2
For any \(|k|\ge 1\), \(\lambda \in \mathbb {R}\) and \(w\in D_k\), there exist constants \(C,c>0\) independent of \(\nu ,k,B,\lambda ,R\), such that for any \(0\le c'\le c\), we have
Moreover, it also holds
The proof of this proposition can be separated into four parts. We first derive the resolvent estimates for w from \(L^2\) to \(L^2\).
Lemma 3.3
For any \(|k|\ge 1\), \(\lambda \in \mathbb {R}\) and \(w\in D_k\), there exist constants \(C>0\) independent of \(\nu ,k,B,\lambda ,R\), such that the following estimate holds
Moreover, there exist constants \(C,c>0\) independent of \(\nu ,k,B,\lambda ,R\), such that for any \(0\le c'\le c\), it is also true that
Proof
We first prove (3.3). The second inequality readily follows if one selects \(c=\frac{1}{2C}\) since for all \(c'\in [0, c]\) it holds
By utilizing Lemma 3.1 and Cauchy–Schwarz inequality, we obtain
To proceed, we then first prove the statement
The mathematical discussion can be divided into three cases:
Here \(0<\delta \ll 1\) is a small constant, which will be determined later.
-
(1)
Case of \(\lambda \in (-\infty ,\frac{1-\delta }{R^2}]\cup [1+\delta ,\infty )\). One can check
$$\begin{aligned} |\Im \langle F,w\rangle |=|kB|\big |\langle (\frac{1}{r^2}-\lambda )w,w\rangle \big |=|kB|\big |\langle (1-\lambda r^2)\frac{w}{r},\frac{w}{r}\rangle \big |\ge |kB|\delta \Vert \frac{w}{r}\Vert _{L^2}^2. \end{aligned}$$Taking \(\delta =\frac{\nu ^{\frac{1}{3}}}{|kB|^{\frac{1}{3}}}\), it yields
$$\begin{aligned} \Vert rF\Vert _{L^2}\ge \nu ^{\frac{1}{3}}|kB|^{\frac{2}{3}}\Vert \frac{w}{r}\Vert _{L^2}. \end{aligned}$$(3.6) -
(2)
Case of \(\lambda \in [\frac{1-\delta }{R^2},\frac{1}{R^2}]\) or \(\lambda \in [1,1+\delta ]\). i)\(\lambda \in [\frac{1-\delta }{R^2},\frac{1}{R^2}]\). Noting that \(\sqrt{\frac{1-\delta }{\lambda }}\in [1,R]\), we first obtain
$$\begin{aligned} |\Im \langle F,w\rangle |&=|kB|\int _1^R(1-\lambda r^2)|\frac{w}{r}|^2dr\\&\ge |kB|\int _1^{\sqrt{\frac{1-\delta }{\lambda }}}(1-\lambda r^2)|\frac{w}{r}|^2dr\ge |kB|\delta \Vert \frac{w}{r}\Vert _{L^2(1,\sqrt{\frac{1-\delta }{\lambda }})}^2. \end{aligned}$$The rest part of \(\Vert \frac{w}{r}\Vert _{L^2}\) can be controlled as below
$$\begin{aligned}&\Vert \frac{w}{r}\Vert _{L^2(\sqrt{\frac{1-\delta }{\lambda }},R)}^2=\int _{\sqrt{\frac{1-\delta }{\lambda }}}^R|\frac{w}{r}|^2dr\le \int _{\sqrt{\frac{1-\delta }{\lambda }}}^R\frac{1}{r^2}dr\Vert w\Vert _{L^{\infty }}^2\\&\quad =(\sqrt{\frac{\lambda }{1-\delta }}-\frac{1}{R})\Vert w\Vert _{L^{\infty }}^2 \le (\sqrt{\frac{1}{1-\delta }}-1)\frac{\Vert w\Vert _{L^{\infty }}^2}{R}\lesssim \frac{\delta }{R}\Vert w\Vert _{L^{\infty }}^2\le \delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2. \end{aligned}$$Combining these two bounds, we deduce
$$\begin{aligned} \Vert \frac{w}{r}\Vert _{L^2}^2&=\Vert \frac{w}{r}\Vert _{L^2(1,\sqrt{\frac{1-\delta }{\lambda }})}^2+\Vert \frac{w}{r}\Vert _{L^2(\sqrt{\frac{1-\delta }{\lambda }},R)}^2\\&\lesssim |kB|^{-1}\delta ^{-1}|\Im \langle F,w\rangle |+\delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2. \end{aligned}$$It is then inferred from Lemma A.1 and Lemma 3.1 that
$$\begin{aligned} \Vert \frac{w}{r}\Vert _{L^2}^2&\lesssim |kB|^{-1}\delta ^{-1}|\Im \langle F,w\rangle |+\delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2\\&\lesssim |kB|^{-1}\delta ^{-1}\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\delta \Vert \frac{w}{r}\Vert _{L^2}\Vert r^{\frac{1}{2}}\partial _r(\frac{w}{r^{\frac{1}{2}}})\Vert _{L^2}\\&\lesssim |kB|^{-1}\delta ^{-1}\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\delta \Vert \frac{w}{r}\Vert _{L^2}(\Vert w'\Vert _{L^2}+\Vert \frac{w}{r}\Vert _{L^2})\\&\lesssim |kB|^{-1}\delta ^{-1}\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\nu ^{-\frac{1}{2}}\delta \Vert rF\Vert _{L^2}^{\frac{1}{2}}\Vert \frac{w}{r}\Vert _{L^2}^{\frac{3}{2}}. \end{aligned}$$This leads to
$$\begin{aligned} \Vert rF\Vert _{L^2}\gtrsim \min \{|kB|\delta ,\nu \delta ^{-2}\}\Vert \frac{w}{r}\Vert _{L^2}. \end{aligned}$$The desired estimate (3.5) follows by picking \(\delta =\frac{\nu ^{\frac{1}{3}}}{|kB|^{\frac{1}{3}}}\). ii)\(\lambda \in [1,1+\delta ]\). Observing \(\sqrt{\frac{1+\delta }{\lambda }}\in [1, R]\), we have
$$\begin{aligned} |\Im \langle F,w\rangle |&=|kB|\int _1^R(\lambda r^2-1)|\frac{w}{r}|^2dr\\&\ge |kB|\int _{\sqrt{\frac{1+\delta }{\lambda }}}^R(\lambda r^2-1)|\frac{w}{r}|^2dr\ge |kB|\delta \Vert \frac{w}{r}\Vert _{L^2(\sqrt{\frac{1+\delta }{\lambda }},R)}^2. \end{aligned}$$When \(r\in (1,\sqrt{\frac{1+\delta }{\lambda }})\), we bound \(\frac{w}{r}\) by the \(L^\infty \) norm of \(\frac{w}{r^{\frac{1}{2}}}\) and it holds
$$\begin{aligned}&\Vert \frac{w}{r}\Vert _{L^2(1,\sqrt{\frac{1+\delta }{\lambda }})}^2=\int _1^{\sqrt{\frac{1+\delta }{\lambda }}}|\frac{w}{r}|^2dr \le \int _1^{\sqrt{\frac{1+\delta }{\lambda }}}\frac{1}{r}dr\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2\\&\quad =(\sqrt{\frac{1+\delta }{\lambda }}-1)\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2 \le (\sqrt{1+\delta }-1)\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2\le \delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2. \end{aligned}$$Summing two inequalities above renders
$$\begin{aligned} \Vert \frac{w}{r}\Vert _{L^2}^2&=\Vert \frac{w}{r}\Vert _{L^2(1,\sqrt{\frac{1+\delta }{\lambda }})}^2+\Vert \frac{w}{r}\Vert _{L^2(\sqrt{\frac{1+\delta }{\lambda }},R)}^2\\&\lesssim \delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2+|kB|^{-1}\delta ^{-1}|\Im \langle F,w\rangle |. \end{aligned}$$In view of Lemma A.1 and Lemma 3.1, we further deduce
$$\begin{aligned} \Vert \frac{w}{r}\Vert _{L^2}^2&\lesssim |kB|^{-1}\delta ^{-1}|\Im \langle F,w\rangle |+\delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2\\&\lesssim |kB|^{-1}\delta ^{-1}\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\delta \Vert \frac{w}{r}\Vert _{L^2}\Vert r^{\frac{1}{2}}\partial _r(\frac{w}{r^{\frac{1}{2}}})\Vert _{L^2}\\&\lesssim |kB|^{-1}\delta ^{-1}\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\delta \Vert \frac{w}{r}\Vert _{L^2}(\Vert w'\Vert _{L^2}+\Vert \frac{w}{r}\Vert _{L^2})\\&\lesssim |kB|^{-1}\delta ^{-1}\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\nu ^{-\frac{1}{2}}\delta \Vert rF\Vert _{L^2}^{\frac{1}{2}}\Vert \frac{w}{r}\Vert _{L^2}^{\frac{3}{2}}. \end{aligned}$$This yields
$$\begin{aligned} \Vert rF\Vert _{L^2}\gtrsim \min \{|kB|\delta ,\nu \delta ^{-2}\}\Vert \frac{w}{r}\Vert _{L^2}. \end{aligned}$$By setting \(\delta =\frac{\nu ^{\frac{1}{3}}}{|kB|^{\frac{1}{3}}}\), we can optimize the inequality above and thus obtain
$$\begin{aligned} \Vert rF\Vert _{L^2}\gtrsim \nu ^{\frac{1}{3}}|kB|^{\frac{2}{3}}\Vert \frac{w}{r}\Vert _{L^2}. \end{aligned}$$(3.7) -
(3)
Case of \(\lambda \in [\frac{1}{R^2},1]\). Using the fact
$$\begin{aligned} |1-\lambda r^2|\le \delta \Rightarrow 1-\delta \le \lambda r^2\le 1+\delta \Rightarrow \sqrt{\frac{1-\delta }{\lambda }}\le r\le \sqrt{\frac{1+\delta }{\lambda }}, \end{aligned}$$for small \(\delta >0\), it can be seen that
$$\begin{aligned} \sqrt{\frac{1+\delta }{\lambda }}-\sqrt{\frac{1-\delta }{\lambda }}=\frac{2\delta }{\sqrt{\lambda }(\sqrt{1+\delta }+\sqrt{1-\delta })}\lesssim \frac{\delta }{\sqrt{\lambda }}. \end{aligned}$$Now we choose \(r_{-}\in (\sqrt{\frac{1-\delta }{\lambda }}-\frac{\delta }{\sqrt{\lambda }},\sqrt{\frac{1-\delta }{\lambda }})\) and \(r_{+}\in (\sqrt{\frac{1+\delta }{\lambda }},\sqrt{\frac{1+\delta }{\lambda }}+\frac{\delta }{\sqrt{\lambda }})\) such that the following inequalities hold
$$\begin{aligned} |w'(r_{-})|^2\le \frac{\Vert w'\Vert _{L^2}^2}{\delta /\sqrt{\lambda }},\quad |w'(r_{+})|^2\le \frac{\Vert w'\Vert _{L^2}^2}{\delta /\sqrt{\lambda }}. \end{aligned}$$(3.8)In order to control \(\Vert \frac{w}{r}\Vert _{L^2((1, r_{-})\cup (r_{+},R))}\), we examine the inner product of F and \(w(\chi _{(1,r_{-})}-\chi _{(r_{+},R)})\). Via integration by parts, we write
$$\begin{aligned}&\langle F,w(\chi _{(1,r_{-})}-\chi _{(r_{+},R)})\rangle \\&\quad =-\nu \int _1^{r_{-}}w''\overline{w}dr+\nu \int _{r_{+}}^Rw''\overline{w}dr+\nu (k^2-\frac{1}{4})[\int _1^{r_{-}}\frac{|w|^2}{r^2}dr-\int _{r_{+}}^R\frac{|w|^2}{r^2}dr]\\&\qquad +ikB\big (\int _1^{r_{-}}(\frac{1}{r^2}-\lambda )|w|^2dr+\int _{r_{+}}^R(\lambda -\frac{1}{r^2})|w|^2dr\big )\\&\quad =\nu \int _1^{r_{-}}|w'|^2dr-\nu \int _{r_{+}}^R|w'|^2dr-\nu [w'(r_{-})\overline{w}(r_{-})+w'(r_{+})\overline{w}(r_{+})]\\&\qquad +\nu (k^2-\frac{1}{4})[\int _1^{r_{-}}\frac{|w|^2}{r^2}dr-\int _{r_{+}}^R\frac{|w|^2}{r^2}dr]\\&\qquad +ikB\big (\int _1^{r_{-}}(1-\lambda r^2)\frac{|w|^2}{r^2}dr+\int _{r_{+}}^R(\lambda r^2-1)\frac{|w|^2}{r^2}dr\big ). \end{aligned}$$Taking the imaginary part of above equality, it indicates that
$$\begin{aligned}&|kB|\Big |\int _1^{r_{-}}(1-\lambda r^2)\frac{|w|^2}{r^2}dr+\int _{r_{+}}^R(\lambda r^2-1)\frac{|w|^2}{r^2}dr\Big |\\&\quad \le \Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\nu \big (|w'(r_{-})\overline{w}(r_{-})|+|w'(r_{+})\overline{w}(r_{+})|\big ). \end{aligned}$$Thanks to the choice of \(r_-\) and \(r_+\), we can derive the estimate for \(\Vert \frac{w}{r}\Vert _{L^2\big ((1, r{-})\cup (r_{+},R)\big )}\) as below
$$\begin{aligned}&\Vert \frac{w}{r}\Vert _{L^2((1, r_{-})\cup (r_{+},R))}^2\nonumber \\&\quad \le |kB|^{-1}\delta ^{-1}\Big [\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\nu \big (|w'(r_{-})\overline{w}(r_{-})|+|w'(r_{+})\overline{w}(r_{+})|\big )\Big ]. \end{aligned}$$(3.9)According to (3.8), the second term on the right-hand side of (3.9) can be bounded by
$$\begin{aligned}&\nu \big (|w'(r_{-})\overline{w}(r_{-})|+|w'(r_{+})\overline{w}(r_{+})|\big )\le \frac{\nu \lambda ^{\frac{1}{4}}}{\delta ^{\frac{1}{2}}}\Vert w'\Vert _{L^2}\Big (|\overline{w}(r_{-})|+|\overline{w}(r_{+})|\Big )\\&\quad \le \frac{\nu \lambda ^{\frac{1}{4}}}{\delta ^{\frac{1}{2}}}(r_{-}^{\frac{1}{2}}+r_{+}^{\frac{1}{2}})\Vert w'\Vert _{L^2}\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}\lesssim \frac{\nu \lambda ^{\frac{1}{4}}}{\delta ^{\frac{1}{2}}}\lambda ^{-\frac{1}{4}}\Vert w'\Vert _{L^2}\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}\\&\quad =\frac{\nu }{\delta ^{\frac{1}{2}}}\Vert w'\Vert _{L^2}\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}. \end{aligned}$$Thus, we obtain the following estimate for \(\Vert \frac{w}{r}\Vert _{L^2}\):
$$\begin{aligned}&\Vert \frac{w}{r}\Vert _{L^2}^2=\Vert \frac{w}{r}\Vert _{L^2((1, r_{-})\cup (r_{+},R))}^2+\Vert \frac{w}{r}\Vert _{L^2(r_{-},r_{+})}^2\\&\quad \le |kB|^{-1}\delta ^{-1}\Big (\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\frac{\nu }{\delta ^{\frac{1}{2}}}\Vert w'\Vert _{L^2}\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}\Big )+\int _{r_{-}}^{r_{+}}\frac{1}{r}dr\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^\infty }^2 \\&\quad \lesssim |kB|^{-1}\delta ^{-1}\Big (\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\frac{\nu }{\delta ^{\frac{1}{2}}}\Vert w'\Vert _{L^2}\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}\Big )+\delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^\infty }^2, \end{aligned}$$where in the last line we employ Lemma A.6 from the Appendix. Utilizing Lemma 3.1 and Lemma A.1, it can be further deduced that
$$\begin{aligned} \Vert \frac{w}{r}\Vert _{L^2}^2&\lesssim |kB|^{-1}\delta ^{-1}\Big (\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\frac{\nu }{\delta ^{\frac{1}{2}}}\Vert w'\Vert _{L^2}\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}\Big )+\delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^\infty }^2\\&\lesssim |kB|^{-1}\delta ^{-1}\Big (\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\frac{\nu ^{\frac{1}{2}}}{\delta ^{\frac{1}{2}}}\Vert rF\Vert _{L^2}^{\frac{1}{2}}\Vert \frac{w}{r}\Vert _{L^2}^{\frac{1}{2}}\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}\Big )+\delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^\infty }^2\\&\lesssim |kB|^{-1}\delta ^{-1}\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+|kB|^{-2}\delta ^{-4}\nu \Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^\infty }^2\\&\lesssim |kB|^{-1}\delta ^{-1}\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+|kB|^{-2}\delta ^{-4}\nu \Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\delta \Vert \frac{w}{r}\Vert _{L^2}\Vert r^{\frac{1}{2}}\partial _r(\frac{w}{r^{\frac{1}{2}}})\Vert _{L^2}\\&\lesssim |kB|^{-1}\delta ^{-1}\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+|kB|^{-2}\delta ^{-4}\nu \Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}\\&\quad +\delta \Vert \frac{w}{r}\Vert _{L^2}(\Vert w'\Vert _{L^2}+\Vert \frac{w}{r}\Vert _{L^2})\\&\lesssim |kB|^{-1}\delta ^{-1}\Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+|kB|^{-2}\delta ^{-4}\nu \Vert rF\Vert _{L^2}\Vert \frac{w}{r}\Vert _{L^2}+\delta \nu ^{-\frac{1}{2}}\Vert rF\Vert _{L^2}^{\frac{1}{2}}\Vert \frac{w}{r}\Vert _{L^2}^{\frac{3}{2}}. \end{aligned}$$This leads to the conclusion
$$\begin{aligned} \Vert rF\Vert _{L^2}\gtrsim \min \{|kB|\delta ,|kB|^{2}\delta ^{4}\nu ^{-1},\delta ^{-2}\nu \}\Vert \frac{w}{r}\Vert _{L^2}. \end{aligned}$$One can obtain an optimized form of the above inequality if \(\delta \) is chosen as \(\delta =\frac{\nu ^{\frac{1}{3}}}{|kB|^{\frac{1}{3}}}\). It then follows that
$$\begin{aligned} \Vert rF\Vert _{L^2}\gtrsim \nu ^{\frac{1}{3}}|kB|^{\frac{2}{3}}\Vert \frac{w}{r}\Vert _{L^2}. \end{aligned}$$
Therefore, we arrive at
Notice that Lemma 3.1 also implies
Along with (3.10) this yields the desired estimate on \(w'\):
The remaining task is to control \(\Vert r(\frac{1}{r^2}-\lambda )w\Vert _{L^2}\), and we appeal to exploring the imaginary part of \(\langle F,r^2(\frac{1}{r^2}-\lambda )w\rangle \). Integration by parts gives
Then we take the imaginary part of both sides and rewrite \(-\lambda rw = r(\frac{1}{r^2}-\lambda )w-\frac{w}{r}\) to deduce
In view of (3.10) and (3.11), the \(L^2\) norm of \(r(\frac{1}{r^2}-\lambda )w\) obeys the following bounds:
provided that \(\frac{\nu }{|kB|} \ll 1\). This completes the proof. \(\square \)
Relying on the above \(L^2\) estimates for w and \(w'\), we can derive the below bounds for \(\varphi \) and \(\varphi '\).
Lemma 3.4
For any \(|k|\ge 1\), \(\lambda \in \mathbb {R}\) and \(w\in D_k\), there exist constants \(C,c>0\) independent of \(\nu ,k,B,\lambda ,R\), such that for any \(0\le c'\le c\), it holds
Proof
Denote \(\tilde{F}:=rF-c'\nu ^{\frac{1}{3}} |kB|^{\frac{2}{3}}\frac{w}{r}\). For \(\delta =(\frac{\nu }{|kB|})^{\frac{1}{3}}\ll 1\), we set
Employing Lemma 3.3, Lemma A.1 and Lemma A.8, we have
On the other hand, utilizing Lemma 3.3 again as well as Lemma A.7, it can be inferred that
Thus we deduce
The controls for \(\varphi '\) and \(\frac{\varphi }{r}\) also follow from Lemma A.3. And we have
This completes the proof. \(\quad \square \)
Now we turn to estimate the \(H^{-1}_r\) norm of w through the \(H^1_r\) norm of F.
Lemma 3.5
For any \(|k|\ge 1\), \(\lambda \in \mathbb {R}\) and \(w\in D_k\), there exist a constants \(C,c>0\) independent of \(\nu ,k,B,\lambda \), such that for \(0\le c'\le c\), the following estimate holds
Proof
By Lemma 3.1, one first obtains
which gives
Denote \(G{:}{=}F-c'\nu ^{\frac{1}{3}} |kB|^{\frac{2}{3}}R^{-2}w\). In below, we utilize a similar method as in the proof of Proposition 3.3 to demonstrate
To achieve this, we will examine the following three cases:
where \(0<\delta \ll 1\) is a small constant to be determined later.
-
(1)
Case of \(\lambda \in (-\infty ,\frac{1-\delta }{R^2}]\cup [1+\delta ,\infty )\). One can check
$$\begin{aligned}&|\Im \langle G,w\rangle |=|\Im \langle F,w\rangle |=|kB|\big |\langle (\frac{1}{r^2}-\lambda )w,w\rangle \big |=|kB|\big |\langle (1-\lambda r^2)\frac{w}{r},\frac{w}{r}\rangle \big |\ge |kB|\delta \Vert \frac{w}{r}\Vert _{L^2}^2. \end{aligned}$$Taking \(\delta =\frac{\nu ^{\frac{1}{3}}}{|kB|^{\frac{1}{3}}}\), together with (3.12), we obtain
$$\begin{aligned} C(\Vert G\Vert _{H^{-1}_r}+c'\nu ^{\frac{2}{3}} |kB|^{\frac{1}{3}}\Vert \frac{w}{r}\Vert _{L^2})\ge \nu ^{\frac{2}{3}} |kB|^{\frac{1}{3}}\Vert \frac{w}{r}\Vert _{L^2}. \end{aligned}$$Choosing \(Cc'\le \frac{1}{2}\), we then prove (3.13).
-
(2)
Case of \(\lambda \in [\frac{1-\delta }{R^2},\frac{1}{R^2}]\) or \(\lambda \in [1,1+\delta ]\). i) \(\lambda \in [\frac{1-\delta }{R^2},\frac{1}{R^2}]\). Observing that
$$\begin{aligned} 1-\lambda r^2\ge \delta \Longleftrightarrow 1-\delta \ge \lambda r^2\Longleftrightarrow 1\le r\le \sqrt{\frac{1-\delta }{\lambda }}, \end{aligned}$$we then deduce
$$\begin{aligned} |\Im \langle G,w\rangle |&=|\Im \langle F,w\rangle |=|kB|\int _1^R(1-\lambda r^2)|\frac{w}{r}|^2dr\\&\ge |kB|\int _1^{\sqrt{\frac{1-\delta }{\lambda }}}(1-\lambda r^2)|\frac{w}{r}|^2dr\ge |kB|\delta \Vert \frac{w}{r}\Vert _{L^2(1,\sqrt{\frac{1-\delta }{\lambda }})}^2. \end{aligned}$$The remaining part of \(\Vert \frac{w}{r}\Vert _{L^2}\) can be bounded in terms of \(\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}\) as below
$$\begin{aligned}&\Vert \frac{w}{r}\Vert _{L^2(\sqrt{\frac{1-\delta }{\lambda }},R)}^2=\int _{\sqrt{\frac{1-\delta }{\lambda }}}^R|\frac{w}{r}|^2dr\le \int _{\sqrt{\frac{1-\delta }{\lambda }}}^R\frac{1}{r^2}dr\Vert w\Vert _{L^{\infty }}^2\\&\quad =(\sqrt{\frac{\lambda }{1-\delta }}-\frac{1}{R})\Vert w\Vert _{L^{\infty }}^2 \le (\sqrt{\frac{1}{1-\delta }}-1)\frac{\Vert w\Vert _{L^{\infty }}^2}{R}\lesssim \frac{\delta }{R}\Vert w\Vert _{L^{\infty }}^2\le \delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2. \end{aligned}$$Combining these two estimates, together with (3.12), we get
$$\begin{aligned} \Vert \frac{w}{r}\Vert _{L^2}^2&=\Vert \frac{w}{r}\Vert _{L^2(1,\sqrt{\frac{1-\delta }{\lambda }})}^2+\Vert \frac{w}{r}\Vert _{L^2(\sqrt{\frac{1-\delta }{\lambda }},R)}^2\\&\lesssim |kB|^{-1}\delta ^{-1}(\nu ^{-1}\Vert G\Vert _{H^{-1}_r}^2+c'\nu ^{\frac{1}{3}}|kB|^{\frac{2}{3}}\Vert \frac{w}{r}\Vert _{L^2}^2)+\delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2. \end{aligned}$$Applying Lemma A.1 and (3.12), we provide the estimate for \(\Vert \frac{w}{r}\Vert _{L^2}\) by
$$\begin{aligned} \Vert \frac{w}{r}\Vert _{L^2}^2&\lesssim |kB|^{-1}\delta ^{-1}(\nu ^{-1}\Vert G\Vert _{H^{-1}_r}^2+c'\nu ^{\frac{1}{3}}|kB|^{\frac{2}{3}}\Vert \frac{w}{r}\Vert _{L^2}^2)+\delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2\\&\lesssim |kB|^{-1}\delta ^{-1}(\nu ^{-1}\Vert G\Vert _{H^{-1}_r}^2+c'\nu ^{\frac{1}{3}}|kB|^{\frac{2}{3}}\Vert \frac{w}{r}\Vert _{L^2}^2)+\delta \Vert \frac{w}{r}\Vert _{L^2}\Vert r^{\frac{1}{2}}\partial _r(\frac{w}{r^{\frac{1}{2}}})\Vert _{L^2}\\&\lesssim |kB|^{-1}\delta ^{-1}(\nu ^{-1}\Vert G\Vert _{H^{-1}_r}^2+c'\nu ^{\frac{1}{3}}|kB|^{\frac{2}{3}}\Vert \frac{w}{r}\Vert _{L^2}^2)\\&\quad +\delta \Vert \frac{w}{r}\Vert _{L^2}(\Vert w'\Vert _{L^2}+\Vert \frac{w}{r}\Vert _{L^2})\\&\lesssim |kB|^{-1}\delta ^{-1}(\nu ^{-1}\Vert G\Vert _{H^{-1}_r}^2+c'\nu ^{\frac{1}{3}}|kB|^{\frac{2}{3}}\Vert \frac{w}{r}\Vert _{L^2}^2)\\&\quad +\delta \Vert \frac{w}{r}\Vert _{L^2}(\nu ^{-\frac{1}{2}}\Vert G\Vert _{H^{-1}_r}+\sqrt{c'}\nu ^{\frac{1}{6}}|kB|^{\frac{1}{3}}\Vert \frac{w}{r}\Vert _{L^2})+\delta \Vert \frac{w}{r}\Vert _{L^2}^2 \\&\lesssim |kB|^{-1}\delta ^{-1}(\nu ^{-1}\Vert G\Vert _{H^{-1}_r}^2+c'\nu ^{\frac{1}{3}}|kB|^{\frac{2}{3}}\Vert \frac{w}{r}\Vert _{L^2}^2)+|kB|\delta ^3\Vert \frac{w}{r}\Vert _{L^2}^2+\delta \Vert \frac{w}{r}\Vert _{L^2}^2. \end{aligned}$$Picking \(\delta =\frac{\nu ^{\frac{1}{3}}}{|kB|^{\frac{1}{3}}}\), we then obtain
$$\begin{aligned}&\Vert \frac{w}{r}\Vert _{L^2}^2 \le C\nu ^{-\frac{4}{3}}|kB|^{-\frac{2}{3}}\Vert G\Vert _{H^{-1}_r}^2+Cc'\Vert \frac{w}{r}\Vert _{L^2}^2+C\nu \Vert \frac{w}{r}\Vert _{L^2}^2+C\delta \Vert \frac{w}{r}\Vert _{L^2}^2. \end{aligned}$$The additional \(\Vert \frac{w}{r}\Vert _{L^2}\) terms on the right can be absorbed to the left by choosing a sufficiently small \(c'>0\), and noting that \(0<\nu , \delta \ll 1\). Hence, we also get the desired estimate (3.13). ii) \(\lambda \in [1,1+\delta ]\). Using that
$$\begin{aligned} \lambda r^2-1\ge \delta \Longleftrightarrow r^2\ge \frac{1+\delta }{\lambda } \Longleftrightarrow \sqrt{\frac{1+\delta }{\lambda }}\le r\le R, \end{aligned}$$we can bound the imaginary part of \(\langle G,w\rangle \) as below
$$\begin{aligned}&|\Im \langle G,w\rangle |=|\Im \langle F,w\rangle |=|kB|\int _1^R(\lambda r^2-1)|\frac{w}{r}|^2dr\\&\quad \ge |kB|\int _{\sqrt{\frac{1+\delta }{\lambda }}}^R(\lambda r^2-1)|\frac{w}{r}|^2dr\ge |kB|\delta \Vert \frac{w}{r}\Vert _{L^2(\sqrt{\frac{1+\delta }{\lambda }},R)}^2. \end{aligned}$$We further control \(\Vert \frac{w}{r}\Vert _{L^2(1,\sqrt{\frac{1+\delta }{\lambda }})}\) via
$$\begin{aligned} \Vert \frac{w}{r}\Vert _{L^2(1,\sqrt{\frac{1+\delta }{\lambda }})}^2=&\int _1^{\sqrt{\frac{1+\delta }{\lambda }}}|\frac{w}{r}|^2dr \le \int _1^{\sqrt{\frac{1+\delta }{\lambda }}}\frac{1}{r}dr\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2\\ =&(\sqrt{\frac{1+\delta }{\lambda }}-1)\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2 \le (\sqrt{1+\delta }-1)\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2\le \delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2. \end{aligned}$$Together with (3.12), this implies
$$\begin{aligned} \Vert \frac{w}{r}\Vert _{L^2}^2=&\Vert \frac{w}{r}\Vert _{L^2(1,\sqrt{\frac{1+\delta }{\lambda }})}^2+\Vert \frac{w}{r}\Vert _{L^2(\sqrt{\frac{1+\delta }{\lambda }},R)}^2\\ \lesssim&\delta \Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^{\infty }}^2+|kB|^{-1}\delta ^{-1}(\nu ^{-1}\Vert G\Vert _{H^{-1}_r}^2+c'\nu ^{\frac{1}{3}}|kB|^{\frac{2}{3}}\Vert \frac{w}{r}\Vert _{L^2}^2). \end{aligned}$$Thus, we obtain the same bound for \(\Vert \frac{w}{r}\Vert _{L^2}\) as in the scenario with \(\lambda \in [\frac{1-\delta }{R^2},\frac{1}{R^2}]\). Applying Lemma A.1, inequality (3.12) and choosing \(\delta =\frac{\nu ^{\frac{1}{3}}}{|kB|^{\frac{1}{3}}}\) and \(c'>0\) sufficiently small, we then derive the desired bound (3.13).
-
(3)
Case of \(\lambda \in [\frac{1}{R^2},1]\). We now use
$$\begin{aligned} |1-\lambda r^2|\le \delta \Rightarrow 1-\delta \le \lambda r^2\le 1+\delta \Rightarrow \sqrt{\frac{1-\delta }{\lambda }}\le r\le \sqrt{\frac{1+\delta }{\lambda }}. \end{aligned}$$Given \(0<\delta \ll 1\), it then holds
$$\begin{aligned} \sqrt{\frac{1+\delta }{\lambda }}-\sqrt{\frac{1-\delta }{\lambda }}=\frac{2\delta }{\sqrt{\lambda }(\sqrt{1+\delta }+\sqrt{1-\delta })}\lesssim \frac{\delta }{\sqrt{\lambda }}. \end{aligned}$$We further choose \(r_{-}\in (\sqrt{\frac{1-\delta }{\lambda }}-\frac{\delta }{\sqrt{\lambda }},\sqrt{\frac{1-\delta }{\lambda }})\) and \(r_{+}\in (\sqrt{\frac{1+\delta }{\lambda }},\sqrt{\frac{1+\delta }{\lambda }}+\frac{\delta }{\sqrt{\lambda }})\) satisfying the following inequalities
$$\begin{aligned} |w'(r_{-})|^2\le \frac{\Vert w'\Vert _{L^2}^2}{\delta /\sqrt{\lambda }},\quad |w'(r_{+})|^2\le \frac{\Vert w'\Vert _{L^2}^2}{\delta /\sqrt{\lambda }}. \end{aligned}$$(3.14)For the next step, we will construct an appropriate multiplier. To do so, we first define a piecewise \(C^1\) cutoff function \(\rho \) with domain (1, R) as follows
$$\begin{aligned} \rho (r)=\left\{ \begin{aligned}&1,\quad r\in (1,r_{-}-\frac{\delta }{\sqrt{\lambda }}),\\&\sin \big (\frac{\pi }{2}\frac{\sqrt{\lambda }}{\delta }(r_{-}-r)\big ),\quad r\in (r_{-}-\frac{\delta }{\sqrt{\lambda }},r_{-}),\\&0,\quad r\in (r_{-},r_{+}),\\&\sin \big (\frac{\pi }{2}\frac{\sqrt{\lambda }}{\delta }(r_{+}-r)\big ),\quad r\in (r_{+},r_{+}+\frac{\delta }{\sqrt{\lambda }}),\\&-1,\quad r\in (r_{+}+\frac{\delta }{\sqrt{\lambda }},R). \end{aligned} \right. \end{aligned}$$Via integration by parts, we obtain the following expression
$$\begin{aligned} -\Im \langle G,w\rho \rangle =\Im \langle F,w\rho \rangle&=\Im \langle -\nu \partial _r^2w+ikB(\frac{1}{r^2}-\lambda )w,w\rho \rangle \\&=\Im \langle ikB(\frac{1}{r^2}-\lambda )w,w\rho \rangle +\nu \Im \langle w',w\rho '\rangle . \end{aligned}$$Taking the imaginary part of above equality leads to
$$\begin{aligned}&|kB|\Big |\int _1^{r_{-}-\frac{\delta }{\sqrt{\lambda }}}(1-\lambda r^2)\frac{|w|^2}{r^2}dr+\int _{r_{+}+\frac{\delta }{\sqrt{\lambda }}}^R(\lambda r^2-1)\frac{|w|^2}{r^2}dr\Big |\\&\quad \le \Vert G\Vert _{H^{-1}_r}\Vert w\rho \Vert _{H^1_r}+\nu \Vert w'\Vert _{L^2}\Vert w\rho '\Vert _{L^2}. \end{aligned}$$Based on our choice of \(\rho \) and the definition of the \(H_r^1\) norm, it follows that
$$\begin{aligned} \Vert w\rho \Vert _{L^2}\lesssim \Vert (w\rho )'\Vert _{L^2}+\Vert \frac{w}{\rho }{r}\Vert _{L^2} \le&\Vert w'\Vert _{L^2}+\Vert w\rho '\Vert _{L^2}+\Vert \frac{w}{r}\Vert _{L^2}, \end{aligned}$$and
$$\begin{aligned} \Vert w\rho '\Vert _{L^2}\lesssim \frac{\sqrt{\lambda }}{\delta }\Vert w \Vert _{L^2\big ((r_{-}-\frac{\delta }{\sqrt{\lambda }},r_{-})\cup (r_{+},r_{+}+\frac{\delta }{\sqrt{\lambda }})\big )} \lesssim \delta ^{-1} \Vert \frac{w}{r}\Vert _{L^2\big ((r_--\frac{\delta }{\sqrt{\lambda }}, r_++\frac{\delta }{\sqrt{\lambda }})\big )}. \end{aligned}$$Hence, plugging in (3.12) and (3.14), we then derive the estimate
$$\begin{aligned}&\Vert \frac{w}{r}\Vert _{L^2\big ((1, r_{-}-\frac{\delta }{\sqrt{\lambda }})\cup (r_{+}+\frac{\delta }{\sqrt{\lambda }},R)\big )}^2\\&\quad \lesssim |kB|^{-1}\delta ^{-1}\Big [\Vert G\Vert _{H^{-1}_r}(\Vert w'\Vert _{L^2}+\delta ^{-1} \Vert \frac{w}{r}\Vert _{L^2\big ((r_--\frac{\delta }{\sqrt{\lambda }}, r_++\frac{\delta }{\sqrt{\lambda }})\big )})+\nu \Vert w'\Vert _{L^2}\cdot \delta ^{-1}\Vert \frac{w}{r}\Vert _{L^2}\Big ]\\&\quad \lesssim |kB|^{-1}\delta ^{-1}\Big [\Vert G\Vert _{H^{-1}_r}(\nu ^{-1}\Vert G\Vert _{H^{-1}_r}+\nu ^{-1}\sqrt{c'}\nu ^{\frac{2}{3}}|kB|^{\frac{1}{3}}\Vert \frac{w}{r}\Vert _{L^2}+\delta ^{-1} \Vert \frac{w}{r}\Vert _{L^2\big ((r_--\frac{\delta }{\sqrt{\lambda }}, r_++\frac{\delta }{\sqrt{\lambda }})\big )})\\&\qquad +(\Vert G\Vert _{H^{-1}_r}+\sqrt{c'}\nu ^{\frac{2}{3}}|kB|^{\frac{1}{3}}\Vert \frac{w}{r}\Vert _{L^2})\cdot \delta ^{-1} \Vert \frac{w}{r}\Vert _{L^2\big ((r_--\frac{\delta }{\sqrt{\lambda }}, r_++\frac{\delta }{\sqrt{\lambda }})\big )}\Big ]\\&\quad \lesssim |kB|^{-1}\delta ^{-1}\Big (\nu ^{-1}\Vert G\Vert _{H^{-1}_r}^2+c'\nu ^{\frac{1}{3}}|kB|^{\frac{2}{3}}\Vert \frac{w}{r}\Vert _{L^2}^2+\nu \delta ^{-2} \Vert \frac{w}{r}\Vert _{L^2\big ((r_--\frac{\delta }{\sqrt{\lambda }}, r_++\frac{\delta }{\sqrt{\lambda }})\big )}\Big ). \end{aligned}$$Meanwhile, it can be deduced from Lemma A.6, Lemma A.1 and (3.12) that
$$\begin{aligned}&\Vert \frac{w}{r}\Vert _{L^2\big ((r_--\frac{\delta }{\sqrt{\lambda }}, r_++\frac{\delta }{\sqrt{\lambda }})\big )}^2\\&\quad \le \int _{r_{-}-\frac{\delta }{\sqrt{\lambda }}}^{r_{+}+\frac{\delta }{\sqrt{\lambda }}}\frac{1}{r}dr\Vert \frac{w}{r^{\frac{1}{2}}}\Vert _{L^\infty }^2 \lesssim \delta \Vert \frac{w}{r}\Vert _{L^2}\Vert r^{\frac{1}{2}}\partial _r(\frac{w}{r^{\frac{1}{2}}})\Vert _{L^2} \\&\quad \lesssim \delta \Vert \frac{w}{r}\Vert _{L^2}(\Vert w'\Vert _{L^2}+\Vert \frac{w}{r}\Vert _{L^2})\\&\quad \lesssim \delta \Vert \frac{w}{r}\Vert _{L^2}(\nu ^{-1}\Vert G\Vert _{H^{-1}_r}+\nu ^{-1}\sqrt{c'}\nu ^{\frac{2}{3}}|kB|^{\frac{1}{3}}\Vert \frac{w}{r}\Vert _{L^2}+\Vert \frac{w}{r}\Vert _{L^2})\\&\quad \lesssim \frac{1}{\sqrt{c'}}\nu ^{-2}\delta ^{2}\Vert G\Vert _{H^{-1}_r}^2+\sqrt{c'}(\nu ^{-\frac{1}{3}}|kB|^{\frac{1}{3}}\delta +1)\Vert \frac{w}{r}\Vert _{L^2}^2+\delta \Vert \frac{w}{r}\Vert _{L^2}^2. \end{aligned}$$Taking \(\delta =\frac{\nu ^{\frac{1}{3}}}{|kB|^{\frac{1}{3}}}\), we can then bound the \(L^2\) norm of \(\frac{w}{r}\) as below
$$\begin{aligned} \Vert \frac{w}{r}\Vert _{L^2}^2&=\Vert \frac{w}{r}\Vert _{L^2\big ((1, r_{-}-\frac{\delta }{\sqrt{\lambda }})\cup (r_{+}+\frac{\delta }{\sqrt{\lambda }},R)\big )}^2+\Vert \frac{w}{r}\Vert _{L^2\big (( r_{-}-\frac{\delta }{\sqrt{\lambda }},r_{+}+\frac{\delta }{\sqrt{\lambda }})\big )}^2\\&\lesssim |kB|^{-1}\delta ^{-1}\Big (\nu ^{-1}\Vert G\Vert _{H^{-1}_r}^2+c'\nu ^{\frac{1}{3}}|kB|^{\frac{2}{3}}\Vert \frac{w}{r}\Vert _{L^2}^2\Big )\\&\quad + \Vert \frac{w}{r}\Vert _{L^2\big ((r_--\frac{\delta }{\sqrt{\lambda }}, r_++\frac{\delta }{\sqrt{\lambda }})\big )}^2\\&\lesssim (1+\frac{1}{\sqrt{c'}})\nu ^{-\frac{4}{3}}|kB|^{-\frac{2}{3}}\Vert G\Vert _{H^{-1}_r}^2+(c'+\sqrt{c'}+\delta )\Vert \frac{w}{r}\Vert _{L^2}^2. \end{aligned}$$Picking \(c'>0\) sufficiently small and noting that \(0<\delta \ll 1\), we can proceed to the conclusion
$$\begin{aligned} \Vert G\Vert _{H^{-1}_r}\gtrsim \nu ^{\frac{2}{3}}|kB|^{\frac{1}{3}}\Vert \frac{w}{r}\Vert _{L^2}. \end{aligned}$$(3.15)
We have therefore proved (3.13) in all three scenarios. This together with (3.12) yields the desired bounds for \(\Vert w\Vert _{H^{1}_r}\):
\(\square \)
For future use, we also establish the resolvent estimate for \(\varphi '\).
Lemma 3.6
For any \(|k|\ge 1\), \(\lambda \in \mathbb {R}\) and \(w\in D_k\), there exists a constant \(C>0\) independent of \(\nu ,k,B,\lambda , R\), such that there holds
Proof
If \(\nu |k|^3> |B|\), by Lemma 3.5 and Lemma A.3 we obtain
Now we work under the case \(\nu |k|^3\le |B|\). Let
with \(\delta =(\frac{\nu }{|kB|})^{\frac{1}{3}}\ll 1\). Denote \(a= \Vert \varphi '\Vert _{L^2}+|k|\Vert \frac{\varphi }{r}\Vert _{L^2}\). Utilizing Lemma A.3 from the Appendix, we immediately have
Employing Lemma 3.5, Lemma A.1 and Lemma A.8, the first term can be controlled through
To bound the second term on the right of (3.16), we use a piecewise \(C^1\) cut-off function \(\chi : (0, \infty ) \rightarrow \mathbb {R}\) from Lemma A.8, which is defined as
with \(r_{\pm }=\sqrt{\frac{1\pm \delta }{\lambda }}\). Conducting integration by parts allows us to write
Due to the fact that \(\chi (\frac{1}{r^2}-\lambda )=1\) on \(E^c\), it can be inferred that
Applying Lemma A.1 and Lemma A.8, we further have
and
Plugging the above inequalities into (3.17), together with Lemma 3.5, we then deduce
Therefore, the term \(a= \Vert \varphi '\Vert _{L^2}+|k|\Vert \frac{\varphi }{r}\Vert _{L^2}\) now obeys
Substituting \(\delta =(\frac{\nu }{|kB|})^{\frac{1}{3}}\) and noticing \(\nu |k|^3\le |B|\), we thus arrive at
This completes the proof of this lemma. \(\quad \square \)
3.3 Sharpness of the resolvent bound
In the last subsection, we prove that the resolvent bound \(\nu ^{\frac{1}{3}} |kB|^{\frac{2}{3}}\) in Lemma 3.3 is sharp. We have
Lemma 3.7
With \(F=-\nu (\partial _r^2-\frac{k^2-\frac{1}{4}}{r^2})w+ikB(\frac{1}{r^2}-\lambda )w\), there exist \(\lambda \in \mathbb {R}, R>1\) and a non-zero function \(w\in H_0^2([1,R])\) such that for \(|k|=1\)
Proof
Choose \(\lambda =\frac{1}{r_0^2}\) and \(\frac{|B|}{\nu }=r_0^3\) for some \(r_0\ge 1\). Setting \(R=r_0+\frac{1}{r_0}\), we construct
One can verify for \(k=1\)
In addition, we have
Thus we deduce
Together with \(\frac{|B|}{\nu }=r_0^3\), this implies
as desired. \(\quad \square \)
4 Enhanced Dissipation and Invisid Damping
4.1 Pseudospectral bound
The direct implication of the resolvent estimates is to provide controls for the semigroup by using pseudospectral bounds.
As in [25], we call an operator L in a Hilbert space H is accretive if
The operator L is said to be m-accretive if, in addition, all \(\Re \lambda <0\) belong to the resolvent set of L (see [19] for more details). The pseudospectral bound of L is defined as
Consider the operator \(L_k\)
in the domain
Note that \(-\partial _r^2\) is an operator with the compact resolvent. Since \(L_k\) is a relatively compact perturbation of \(-\nu \partial _r^2\) in the domain \(D_k\), it is hence clear that the operators \(L_k\) also has the compact resolvent, this indicates that the operator \(L_k\) has only point spectrum.
In Lemma 3.1 we have obtained
The above inequality indicates \(L_k\) being accretive, and furthermore m-accretive. Recall that in Proposition 3.2 we establish the following resolvent estimates
In view of the fact \(r\in [1, R]\), this provides a lower bound for the pseudospectrum of \(L_k\):
We summarize it into
Lemma 4.1
Let \(\Psi \) be defined as in (4.1). There exists some \(C>0\) independent of \(\nu ,k, B, R\) such that
4.2 Semigroup bound and enhanced dissipation
It is convenient to use the below space-time norm
To obtain decaying semigroup bounds from pseudospectral bounds, we appeal to the following Gearhart–Pr\(\ddot{u}\)ss type lemma established by Wei in [33]. (See also [17] by Helffer || Sjöstrand for relevant reference.)
Lemma 4.2
[33]. Let L be a m-accretive operator in a Hilbert space X. Then it holds
We proceed to studying the homogeneous linear equation
Utilizing semigroup theory, we can express \(w_k^l\) as
We start to derive the space-time estimate for \(w_k^l\).
Proposition 4.3
Let \(w_k^l\) be the solution to (4.2) with \(w_k(0)\in L^2\). Then for any \(k\in \mathbb {Z}\) and \(|k|\ge 1\), there exist constants \(C,c>0\) being independent of \(\nu ,k,B,R\), such that the following inequality holds
Moreover, for any \(c'\in (0,c)\), we have
Proof
The semigroup bounds (4.3) readily follows from Lemma 4.1 and Lemma 4.2. Hence for any \(c'\in (0,c)\), multiplying \(e^{c'(\nu k^2)^{\frac{1}{3}}|B|^{\frac{2}{3}}R^{-2}t}\) on both sides of (4.3) yields
Then we integrate above inequality with respect to t to deduce
This implies the estimate in (4.4). \(\square \)
4.3 Homogeneous linear equation with nonzero initial data
Based on the results of Proposition 4.3 and the structure of the linear equation itself, we can obtain the following \(L^{\infty }L^2\) energy estimate for the homogeneous linear equation with respect to \(w_k^l\).
Lemma 4.4
Let \(w_k^l\) be the solution to (4.2) with initial data \(w_k(0)\in L^2\). Considering c to be the same as in Proposition 4.3 and \(c'\in (0,c)\). For any \(k\in \mathbb {Z}\) and \(|k|\ge 1\), it holds
Proof
We first conduct the integration by parts and get
By multiplying \(e^{2c'(\nu k^2)^{\frac{1}{3}}|B|^{\frac{2}{3}}R^{-2}t}\) on both sides, we deduce
This implies a space-time estimate for \(w_k^l\):
Combining with Proposition 4.3, we arrive at
This completes the proof of Lemma 4.4. \(\quad \square \)
4.4 Inhomogeneous linear equation with zero initial data
We then derive the space-time estimates for inhomogeneous equations with zero initial conditions, which will be frequently used in later sections. The derivation of this lemma heavily relies on established resolvent estimates in Proposition 3.2 from Sect. 3.
Proposition 4.5
Given functions \(h_1(t, r), h_2(t, r)\) and g(r), assume w solves
and let the stream function \(\varphi \) satisfy
With c being the same constant as in Proposition 4.3, then for any \(c'\in [0, c]\), the following inequalities hold for w and \(\varphi \):
and
Proof
We introduce the weighted quantities
and
Via a direct check, we can see that
We then take the Fourier transform in t and defineFootnote 1
The inhomogeneous equation (4.7) can thus be transferred into the form
We further decompose \(\hat{w}\) as \(\hat{w}=\hat{w}^{1}+\hat{w}^{2}\), where \(\hat{w}^{1}\) and \(\hat{w}^{2}\) solve
and
We also define the corresponding stream functions \(\hat{\varphi }^1\) and \(\hat{\varphi }^2\) via the below linear elliptic equations
This enables us to write \(\hat{\varphi }=\hat{\varphi }^1+\hat{\varphi }^2\). In view of Proposition 3.2, we then obtain the estimate for \(w^1\) and \(\varphi ^1\):
as well as the control for \(w^2\) and \(\varphi ^2\):
Here we utilize the definition of \(H^{-1}_r\) norm to bound
Combining the above inequalities, we arrive at
and
According to the Plancherel’s theorem, we have the following equivalence relations based on \(L^2\) norms
Plugging in all estimates above, we thus deduce
and
Applying the integration by parts, we also get
By employing the Cauchy–Schwarz inequality, we further obtain
Together with (4.8), the above inequality yields
This finishes the proof of Proposition 4.5. \(\quad \square \)
4.5 Integrated invisid damping
This subsection is devoted to providing estimates for the stream function \(\varphi _k^l\) in terms of initial vorticity \(w_k(0)\), with \(\varphi _k^l\) admitting the linear elliptic equation
For inviscid fluids governed by the Euler equation, there is the concept of so-called inviscid damping. In fact, the effect of inviscid damping extends to viscous fluids as well. Lemma 4.6 presents a space-time version of linear inviscid damping around 2D TC flow.
Lemma 4.6
Let \(w_k^l\) be the solution to (4.2) with initial data \(w_k(0)\in L^2\) and \(\varphi _k^l\) solves (4.9). Considering c to be the same as in Proposition 4.3 and \(c'\in (0,c)\). For any \(k\in \mathbb {Z}\) and \(|k|\ge 1\), the following inequality holds
Remark 4.7
Note that \(\nu \) is much smaller than 1, and B is a fixed constant, so (4.10) can also be written in the following expression
We can see that both sides of the above inequality are independent of viscosity coefficient \(\nu \), which indicates that the result is also valid for the Euler equation. This type of estimate is called inviscid damping, and here we establish its integrated form. Thus we refer to it as the integrated invisid damping estimates.
In order to prove (4.10), we perform a suitable decomposition of \(\varphi ^l_k\). Define
and denote \(\tilde{w}^{l_2}_k\) to be the solution to the inhomogeneous linear equation equation with zero initial conditions as below:
As a result, the corresponding stream functions can be defined by
Then it is easily seen that \(w^l_k=\tilde{w}^{l_1}_k+\tilde{w}^{l_2}_k\) and \(\varphi ^l_k=\tilde{\varphi }^{l_1}_k+\tilde{\varphi }^{l_2}_k\).
We start with the estimates for \(\tilde{\varphi }^{l_1}_k\).
Lemma 4.8
Let \(w_k^l\) be the solution to (4.2) with initial data \(w_k(0)\in L^2\). Considering c to be the same as in Proposition 4.3, \(c'\in (0,c)\) and \(\tilde{\varphi }^{l_1}_k\) defined in (4.13). For any \(k\in \mathbb {Z}\) and \(|k|\ge 1\), the following inequality holds
Proof
Noting that \(\tilde{\varphi }^{l_1}_k=e^{-c(\nu k^2)^{\frac{1}{3}}|B|^{\frac{2}{3}}R^{-2} t}\tilde{\varphi }^{l_0}_k\) and \(c'\in (0,c)\), to obtain the above conclusion it suffices to prove
Here we develop some new ideas. With \(w=1/r^2\), we introduce the weighted Hilbert space \(L^2_w([1, R])\) with inner product
We find a set of explicit orthonormal basis for \(L^2_w([1, R])\) as below:
Here \(\psi _l\) satisfies the following property
As far as we know, our explicit constructions above are new. For the sake of convenience, we denote \(\alpha {:}{=}(\frac{2}{\log R})^{\frac{1}{2}}\), \(\beta {:}{=}\frac{\pi }{\log R}\) and \(\lambda _{k,l}{:}{=}(\beta l)^2+k^2\). Since
we can write
In view of the definition \(\tilde{w}^{l_0}_k=e^{-i kB\frac{t}{r^2}}w_k(0)\), we have the expression
where \(\mathcal {F}\) represents the canonical Fourier transform over \(\mathbb {R}\), and \(W^1_{k, l}(s)=\frac{1}{2s\sqrt{s}}w_k(0, \frac{1}{\sqrt{s}}) \psi _l (\frac{1}{\sqrt{s}})\).Footnote 2
Note that the Plancherels formula implies
This together with (4.17) yields
On the other hand, observing
we infer that
This completes the proof of this lemma. \(\quad \square \)
To control the \(L^2L^2\) norm of \(\tilde{\varphi }^{l_2}_k\), we utilize resolvent estimates in Sect. 3 together with Lemma 4.4.
Lemma 4.9
Let \(w_k^l\) be the solution to (4.2) with initial data \(w_k(0)\in L^2\). Considering c to be the same as in Proposition 4.3 and \(c'\in (0,c)\). For any \(k\in \mathbb {Z}\) and \(|k|\ge 1\), there holds
Proof
First recall that \(\tilde{w}^{l_2}_k\) satisfies the inhomogeneous linear equation
with the stream function \(\tilde{\varphi }^{l_2}_k\) solving the boundary value problem
Employing Proposition 4.5 with \(h_1=-\nu (k^2-\frac{1}{4})\frac{\tilde{w}^{l_1}_k}{r^2}+c(\nu k^2)^{\frac{1}{3}}|B|^{\frac{2}{3}}R^{-2} \tilde{w}^{l_1}_k\), \(h_2=v\partial _r \tilde{w}^{l_1}_k\) and \(g(r)=1\) yields
Notice that
and
Thus we arrive at
\(\square \)
Therefore, the desired bound (4.10) can be readily deduced from Lemma 4.8 and Lemma 4.9. This completes the proof of Lemma 4.6.
5 Space-Time Estimates for the Linearized Navier–Stokes Equations
In this section, we establish the space-time estimates for the linearized 2D Navier–Stokes equation in the vorticity formulation (1.7):
Recall in Sect. 2 we convert this equation to the following system
via introducing
Here \(L_k=-\nu (\partial _r^2-\frac{k^2-\frac{1}{4}}{r^2})+i\frac{kB}{r^2}\) and \(f_1, f_2\) stand for nonlinear forms given by
where \(\varphi _k\) satisfies
5.1 Space-time estimates for non-zero frequency
We decompose the solution \(w_k\) to (5.2) into two parts. Let \(w_k=w_k^l+w_k^{n}\), with \(w_k^l\) fulfilling the homogeneous linear equation
and \(w_k^{n}\) solving the inhomogeneous linear equation with zero initial data
Correspondingly, we also decompose \(\varphi _k\) as \(\varphi _k=\varphi ^{l}_k+\varphi ^{n}_k\), where \(\varphi _k^l\) and \(\varphi _k^n\) satisfy
and
Now we state the main conclusion of this section, which is also named as the space-time estimate for Eq. (5.2).
Proposition 5.1
Assume \(w_k\) is a solution to (5.2) with \(w_k(0)\in L^2\), then there exists a constant \(c'>0\) independent of \(\nu ,B,k,R\) so that it holds
Proof
Employing Lemma 4.4 and Lemma 4.6, we can bound the linear part \(w_k^l\) and \(\varphi _k^l\):
To obtain the desired bounds for the remainder terms \(w_k^n\) and \(\varphi _k^n\), we utilize Proposition 4.5 with the choices of \(h_1 = -\frac{1}{r}ikf_1\), \(h_2 = -r^{\frac{1}{2}}f_2\), and \(g(r) = r^{-\frac{1}{2}}\) to deduce
Combining these two estimates, we obtain the desired inequality. \(\quad \square \)
When \(\log R\lesssim \nu ^{-\frac{1}{3}}|B|^{\frac{1}{3}}\), the \((\frac{\nu }{|kB|})^{\frac{1}{3}}\log R\) term on the right for the above estimate can be eliminated, and thus it infers that
Proposition 5.2
Under the same conditions as in Proposition 5.1. If \(\log R\lesssim \nu ^{-\frac{1}{3}}|B|^{\frac{1}{3}}\), then there holds
Notice that when \(\nu k^2 R^{-2} \ge (\nu k^2)^{\frac{1}{3}}|B|^{\frac{2}{3}}R^{-2}\), i.e., \(\nu k^2\ge |B|\), the heat dissipation effect becomes more significant compared with the enhanced dissipation. Therefore, in the regime where \(\nu k^2\ge |B|\), the following space-time estimates offer a more precise control of \(w_k\).
Proposition 5.3
Let \(w_k\) be the solution to (5.2) with \(w_k(0)\in L^2\). Then there exists a constant \(c'>0\) independent of \(\nu ,B,k,R\), such that
Proof
Conducting the integration by parts, we obtain
It infers
By applying Cauchy–Schwarz inequality, we then deduce
Noticing that \(\Vert \frac{w_k}{r}\Vert _{L^2}\ge R^{-1}\Vert w_k\Vert _{L^2}\), it follows
Therefore, we can multiply \(e^{2c'\nu k^2R^{-2}t}\) on both sides of above inequality. With \(c'\) being a small constant independent of \(\nu ,B,k,R\), we obtain
This further implies
In view of Lemma A.3, it also holds
Combining these two estimates above yields
which completes the proof. \(\quad \square \)
5.2 Space-time estimates for zero frequency
To establish the space-time estimate for zero mode of solutions, we directly utilize the heat dissipative structure of equation and then perform integration by parts.
Recall that the nonlinear perturbation equation reads
where \(w|_{r=1,R}=0\) and \((\partial _r^2+\frac{1}{r}\partial _r+\frac{1}{r^2}\partial _{\theta }^2)\varphi =w\). Denote the zero mode of function \(f(r, \theta )\) to be
and set \(f_{\ne }{:}{=}f-f_{=}\). We then have that the zero frequency part of Eq. (5.6) takes the form
Subsequently, we derive the space-time estimate for \(w_{=}\).
Lemma 5.4
Let \(w_{=}\) be the solution to (5.7) and assume \(r^{\frac{1}{2}}w_{=}(0)\in L^2\). Denoting \(w_0{:}{=}r^{\frac{1}{2}} w_{=}\), then the following inequality holds
Here \(\varphi _l\) and \(w_{l}\) are defined in (5.3).
Proof
Employing integration by parts for Eq. (5.7), we deduce
Observe that
Conducting integration by parts again, we then arrive at
This leads to the following inequality
which renders
We proceed to integrate the above inequality in t variable and get
In view of the definition for \(w_k, \varphi _k\) from (5.3):
we can further express
This completes the proof of (5.8). \(\quad \square \)
6 Nonlinear Stability
This section is devoted to the proof of our main Theorem 1.3 using a bootstrap argument. Thanks to the transformation conducted in Sect. 2, we can focus on analyzing the system (5.2), which is presented in the following form
Here \(L_k=-\nu (\partial _r^2-\frac{k^2-\frac{1}{4}}{r^2})+i\frac{kB}{r^2}\) and \(f_1, f_2\) have the nonlinear structure of
Note that \(\varphi _k\) satisfies
For notational simplicity, we set \(\mu _k{:}{=}\max \left\{ (\nu k^2)^{\frac{1}{3}} |B|^{\frac{2}{3}} R^{-2}, \nu k^2 R^{-2} \right\} \). We then define the initial energy \(\mathcal {M}_0(0)=\Vert w_0(0)\Vert _{L^2}\) and for \(k\in \mathbb {Z}\backslash \{ 0 \}\) we let
As a result of Proposition 5.2 and Proposition 5.3, we readily deduce
Proposition 6.1
For \(k\in \mathbb {Z}\backslash \{0 \}\), let \(w_k\) be the solution to (5.2) with \(w_k(0)\in L^2\). Given \(\log R\lesssim \nu ^{-\frac{1}{3}}|B|^{\frac{1}{3}}\), then there exists a constant \(c'>0\) independent of \(\nu ,B,k,R\), such that it holds
We proceed to construct and control below the energy functionals, with
The derivation of a priori estimates for \(E_k\) crucially relies on the space-time estimates obtained in Sect. 5. We start from presenting the bound for the energy at zero frequency:
Lemma 6.2
Assume that \(\log R\lesssim \nu ^{-\frac{1}{3}}|B|^{\frac{1}{3}}\). We have the following energy inequality
Proof
Directly applying Lemma 5.4, we obtain
which proves Lemma 6.2. \(\quad \square \)
Next, we turn to control \(E_k\) for \(k\ne 0\). We first show that the nonlinear terms \(f_1\) and \(f_2\) obey the below \(L^2\) estimates.
Lemma 6.3
For \(k\in \mathbb {Z}\backslash \{0 \}\), we have
Proof
With the expressions of nonlinear terms
we get
where in the third line we employ Lemma A.3 and Lemma A.5.
And the \(L^2\) norm of \(f_2\) can be directly controlled by
\(\square \)
The above proposition leads to a control of \(E_k\) with \(k\in \mathbb {Z}\backslash \{0 \}\).
Lemma 6.4
Assume that \(\log R\lesssim \nu ^{-\frac{1}{3}}|B|^{\frac{1}{3}}\). For \(k\in \mathbb {Z}\backslash \{0 \}\), we have
Proof
Observe the following basic inequality
Thanks to the definition
this yields
by considering the scenarios \(\nu k^2<|B|\) and \(\nu k^2\ge |B|\) separately. In view of Proposition 6.1, Lemma A.1 and Lemma A.3, we now have
Utilizing Lemma 6.3, it holds
where in the last line we use the fact
Note that
Hence we obtain
To estimate the term involving \(f_2\) on the right of (6.2), we apply Lemma 6.3 again to deduce
Together with (6.2), the above two bounds give the desired control for \(E_k\) when \(|k|\ge 1\).
\(\square \)
Gathering all above estimates, we now conclude our main theorem:
Theorem 6.1
Assume that \(\log R\lesssim \nu ^{-\frac{1}{3}}|B|^{\frac{1}{3}}\). There exist constants \(\nu _0\) and \(c_0, C,c'>0\) independent of \(\nu ,B,R\), such that if
then the solution w to the system (2.2) exists globally in time and satisfies the decaying and stability estimates
Proof
Employing Lemma 6.2 and Lemma 6.4, we have
We perform a bootstrap argument to prove our main theorem. Assume that for a small constant \(c_0>0\) there holds
we then obtain
Choosing \(c_0\le \frac{1}{3} C^{-1}\), we arrive at
which improves the bootstrap assumption (6.5). By a standard continuity argument, we therefore prove that the system (2.2) admits a global-in-time solution, and its k-th mode \(w_k\) obeys the desired bounds
This completes the proof of our main theorem. \(\quad \square \)
Data availability
This article has no associated data.
Notes
Note that \(\varphi |_{r=1, R}=0\) in (4.6) guarantees the Fourier transforms here being well-defined for \(\lambda \in (0,\infty )\).
According to the boundary condition of \(w^l_k\), the Fourier transform of \(W^1_{k, l}(r)\) is well-defined.
References
An, X., He, T., Li, T.: Enhanced dissipation and nonlinear asymptotic stability of the Taylor–Couette flow for the 2D Navier–Stokes equations. arXiv:2112.15357v1
Bedrossian, J., Germain, P., Masmoudi, N.: Stability of the Couette flow at high Reynolds number in 2D and 3D. Bull. Am. Math. Soc. (N.S.) 56, 373–414 (2019)
Bedrossian, J., He, S.: Inviscid damping and enhanced dissipation of the boundary layer for 2D Navier–Stokes linearized around Couette flow in a channel. Commun. Math. Phys. 379(1), 177–226 (2020)
Bedrossian, J., Masmoudi, N., Vicol, V.: Enhanced dissipation and inviscid damping in the inviscid limit of the Navier–Stokes equations near the two dimensional Couette flow. Arch. Ration. Mech. Anal. 219, 1087–1159 (2016)
Bedrossian, J., Wang, F., Vicol, V.: The Sobolev stability threshold for 2D shear flows near Couette. J. Nonlinear Sci. 28, 2051–2075 (2018)
Chapman, S.J.: Subcritical transition in channel flows. J. Fluid Mech. 451, 35–97 (2002)
Chen, Q., Li, T., Wei, D., Zhang, Z.: Transition threshold for the 2-D Couette flow in a finite channel. Arch. Ration. Mech. Anal. 238, 125–183 (2020)
Chossat, P., Iooss, G.: The Couette–Taylor Problem. Applied Mathematical Sciences, vol. 102. Springer-Verlag, New York (1994)
Daviaud, F., Hagseth, J., Bergé, P.: Subcritical transition to turbulence in plane Couette flow. Phys. Rev. Lett. 69, 2511–2514 (1992)
Drazin, P., Reid, W.: Hydrodynamic Stability. Cambridge Monographs on Applied and Computational Mathematics, Cambridge University Press, New York (1981)
Falkovich, G.: Fluid Mechanics. Cambridge University Press, Cambridge (2018)
Gallay, T.: Stability and interaction of vortices in two-dimensional viscous flows. Discr. Cont. Dyn. Syst. Ser. S 5, 1091–1131 (2012)
Gallay, T.: Enhanced dissipation and axisymmetrization of two-dimensional viscous vortices. Arch. Ration. Mech. Anal. 230, 939–975 (2018)
Gallay, T., Šverák, V.: Arnold’s variational principle and its application to the stability of planar vortices. arXiv:2110.13739
Gallay, T., Wayne, C.E.: Global stability of vortex solutions of the two-dimensional Navier–Stokes equation. Commun. Math. Phys. 255, 97–129 (2005)
Guo, Y., Pausader, B., Widmayer, K.: Global axisymmetric Euler flows with rotation. Invent. Math. 231, 169–262 (2023)
Helffer, B., Sjöstrand, J.: Improving semigroup bounds with resolvent estimates. Integral Equ. Oper. Theory 93(3), 36–41 (2021)
Kraichnan, R.H.: Inertial ranges in two-dimensional turbulence. Phys. Fluids 10, 1417–1423 (1967)
Kato, T.: Perturbation Theory for Linear Operators. Grundlehren der Mathematischen Wissenschaften, vol. 132. Springer, Berlin (1966)
Kelvin, L.: Stability of fluid motion-rectilinear motion of viscous fluid between two parallel plates. Phil. Mag. 24, 188–196 (1887)
Li, T., Wei, D., Zhang, Z.: Pseudospectral and spectral bounds for the Oseen vortices operator. Ann. Sci. Ec. Norm. Supér. 53, 993–1035 (2020)
Lundgren, T.: Strained spiral vortex model for turbulent fine structure. Phys. Fluids 25, 2193–2203 (1982)
Masmoudi, N., Zhao, W.: Stability threshold of two-dimensional Couette flow in Sobolev spaces. Ann. Inst. H. Poincaré C Anal. Non Linéaire 39(2), 245–325 (2022)
Masmoudi, N., Zhao, W.: Enhanced dissipation for the 2D Couette flow in critical space. Commun. Partial Differ. Equ. 45(12), 1682–1701 (2020)
Pazy, A.: Semigroups of Linear Operators and Applications to Partial Differential Equations. Applied Mathematical Sciences, vol. 44. Springer-Verlag, New York (1983)
Pope, B.: Turbulent Flows. Cambridge University Press, Cambridge (2000)
Reynolds, O.: An experimental investigation of the circumstances which determine whether the motion of water shall be direct or sinuous, and of the law of resistance in parallel channels. Proc. R. Soc. Lond. 35, 84 (1883)
Rhines, P.B., Young, W.R.: How rapidly is a passive scalar mixed within closed streamlines? J. Fluid Mech. 133, 133–145 (1983)
Schmid, P., Henningson, D.: Stability and Transition in Shear Flows. Applied Mathematical Sciences, vol. 142. Springer-Verlag, New York (2001)
Serrin, J.: On the stability of viscous fluid motions. Arch. Ration. Mech. Anal. 3, 1–13 (1959)
Tillmark, N., Alfredsson, P.H.: Experiments on transition in plane Couette flow. J. Fluid Mech. 235, 89–102 (1992)
Trefethen, L., Trefethen, A., Reddy, S., Driscoll, T.: Hydrodynamic stability without eigenvalues. Science 261, 578–584 (1993)
Wei, D.: Diffusion and mixing in fluid flow via the resolvent estimate. Sci. China Math. 64, 507–518 (2021)
Wei, D., Zhang, Z.: Enhanced dissipation for the Kolmogorov flow via the hypocoercivity method. Sci. China Math. 62(6), 1219–1232 (2019)
Wei, D., Zhang, Z., Zhao, W.: Linear inviscid damping and enhanced dissipation for the Kolmogorov flow. Adv. Math. 362, 106963 (2020)
Yaglom, A.: Hydrodynamic Instability and Transition to Turbulence, vol. 100. Springer-Verlag, New York (2012)
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Funding
XA is supported by MOE Tier 1 Grants A-0004287-00-00, A-0008492-00-00 and MOE Tier 2 Grant A-8000977-00-00. TL is supported by MOE Tier 1 Grant A-0004287-00-00.
Conflict of interest
The authors declare that there are no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix A. Basic Estimates
Appendix A. Basic Estimates
In the appendix, we provide the proof of several elementary but fundamental calculus lemmas which are continually utilized throughout this paper. The following two Sobolev-type inequalities are crucial in the derivation of resolvent estimates in Sect. 3.
Lemma A.1
For any \(w\in L^2(1,R)\) with \(R>1\), it holds
Moreover, if \(w|_{r=1,R}=0\), then
Proof
Choose \(r_0\in [1, R]\) such that
It then follows
And we have
which gives
When \(w|_{r=1,R}=0\), it can be inferred that
\(\square \)
For the weighted quantity \(r^{\frac{1}{2}} w\), we have
Lemma A.2
Let \(w\in L^2(1,R), w|_{r=1,R}=0\), where \(R>1\). Then the following inequality holds
Proof
Due to the vanishment of w on the boundary, it can be deduced
Meanwhile, through integration by part we can write
It then follows from (A.1) that
\(\square \)
To control the stream function \(\varphi _k\) with the vorticity \(w_k\), we develop a series of (weighted) elliptic estimates as below.
Lemma A.3
Suppose that \(|k|\ge 1\). Let \(w=(\partial _r^2-\frac{k^2-\frac{1}{4}}{r^2} )\varphi \) with \(\varphi |_{r=1, R}=0\). Then the following inequality holds
Proof
Noting that
we immediately obtain the first inequality. It then follows from Cauchy–Schwarz inequality that
which implies
and
On the other hand, in view of Cauchy–Schwarz inequality again and using Lemma A.1, it can be estimated
which yields the second inequality.
For the last inequality, we observe that
By applying Lemma A.1 and inequalities above, we then prove
and
\(\square \)
To establish the elliptic estimate for the zero mode of stream function, we require the below Poincaré-type inequalities.
Lemma A.4
For any \(w\in L^2(1,R), w|_{r=1,R}=0\) with \(R>1\), the following Poincaré-type inequalities hold
Proof
Via integration by parts, we immediately obtain
and
\(\square \)
Consequently, we obtain the following elliptic estimate which involves the outer radius R.
Lemma A.5
Let \(w=(\partial _r^2+\frac{1}{r}\partial _r)\varphi \) with \(\varphi |_{r=1, R}=w|_{r=1, R}=0\). The following inequality holds
Proof
Notice that
In view of Lemma A.4, it can be inferred
which yields
and
Applying Lemma A.1, we derive the \(L^\infty \) bound for \(\varphi '\) as below
This completes the proof of this lemma. \(\quad \square \)
Following is a collection of calculus inequalities which provide quantitative information of the interval \([1, R]\cap \{r> 0: \ |1-\lambda r^2|\le \delta \}\) where \(\delta >0\) is sufficiently small. These estimates play a significant role in the proof of resolvent estimates in Sect. 3.
Lemma A.6
For any \(\lambda \in [\frac{1}{R^2},1]\) and \(0<\delta \ll 1\) being a small constant, it holds
where \(r_{-}\in (\sqrt{\frac{1-\delta }{\lambda }}-\frac{\delta }{\sqrt{\lambda }},\sqrt{\frac{1-\delta }{\lambda }})\) and \(r_{+}\in (\sqrt{\frac{1+\delta }{\lambda }},\sqrt{\frac{1+\delta }{\lambda }}+\frac{\delta }{\sqrt{\lambda }})\).
Proof
A direct computation yields
and
\(\square \)
The \(L^2\) norm of \(r^{-\frac{1}{2}}(1-\lambda r^2)^{-1}\) in the region \(\{r\in [1, R]: \ |1-\lambda r^2|> \delta \}\) can be bounded in terms of R and \(\delta \).
Lemma A.7
For any \(\delta>0, R>1 \) and \(\lambda \in \mathbb {R}\). Let \(E=\{r> 0: \ |1-\lambda r^2|\le \delta \}\) and \(E^c=(0, \infty )\backslash E\). Then it holds
Proof
If \(\lambda \le 0\), then one has
Now we assume \(\lambda >0\). Denote \(r_{\pm }=\sqrt{\frac{1\pm \delta }{\lambda }}\). We consider the following five scenarios:
-
(1)
If \(1\le r_-<r_+\le R\), then a direct calculation implies
$$\begin{aligned} \begin{aligned}&\int _1^{r_-} \frac{dr}{r(1-\lambda r^2)^2}=\frac{1}{2} \int _{\delta }^{1-\lambda } \frac{ds}{(1-s)s^2} =\frac{1}{2}\left( \log |\frac{s}{s-1}|-\frac{1}{s} \right) \Big |^{1-\lambda }_{\delta }\\&\quad \lesssim \log (\lambda ^{-1}-1)+\delta ^{-1} \lesssim \log R+ \delta ^{-1}. \end{aligned} \end{aligned}$$(A.4)Here in the first line we use the change of variables \(s=1-\lambda r^2\) and the last line follows from the condition \(r_{-}\in [1, R]\). Similarly, taking into account the condition \(r_+\in [1, R]\), we deduce
$$\begin{aligned} \int _{r_+}^{R} \frac{dr}{r(1-\lambda r^2)^2} \lesssim \log R+ \delta ^{-1}. \end{aligned}$$(A.5)Combining these two inequalities above, we obtain the desired result.
-
(2)
If \(1\le r_-\le R<r_+\), then (A.3) directly follows from (A.4) and the fact that \([1,R]\cap E^c=[1, r_-]\).
-
(3)
If \(r_-<1\le r_+\le R\), then (A.3) can be readily obtained thanks to (A.5) and the fact that \([1,R]\cap E^c=[r_+, R]\).
-
(4)
If \(r_+<1\) or \(r_->R\), then \([1, R]\cap E^c=[1, R]\). Assume first \(r_->R\). It is hence clear that \(1-\lambda R^2\ge \delta \). A suitable change of variables implies
$$\begin{aligned} \begin{aligned}&\int _1^{R} \frac{dr}{r(1-\lambda r^2)^2}=\frac{1}{2} \int _{1-\lambda R^2}^{1-\lambda } \frac{ds}{(1-s)s^2} =\frac{1}{2}\left( \log (\frac{s}{s-1})-\frac{1}{s} \right) \Big |^{1-\lambda }_{1-\lambda R^2}\\&\quad \lesssim \log (R^2\frac{1-\lambda }{1-\lambda R^2})+\frac{1}{1-\lambda R^2} \lesssim \log R+ \delta ^{-1}. \end{aligned} \end{aligned}$$The case when \(r_+<1\) can be treated in a similar manner.
-
(5)
If \(r_-<1<R<r_+\), then immediately we see (A.3) holds since \([1, R]\cap E^c=\varnothing \).
\(\square \)
Finally, in order to prove the resolvent estimates for \(\varphi '\) in Lemma 3.6, it is necessary to construct a piecewise \(C^1\) cut-off function satisfying the following properties.
Lemma A.8
For any \(0<\delta \ll 1\) and \(\lambda \in \mathbb {R}\). Let \(E=\{r> 0: \ |1-\lambda r^2|\le \delta \}\) and \(E^c=(0, \infty )\backslash E\). Then there exists a piecewise \(C^1\) function \(\chi : (0, \infty ) \rightarrow \mathbb {R}\) satisfying the following properties:
-
(1)
\(\chi (r)=\frac{1}{\frac{1}{r^2}-\lambda }\) on \(E^c\).
-
(2)
there exists a constant \(C>0\) independent of \(\lambda \) and \(\delta \), such that it holds
$$\begin{aligned}\Vert \frac{\chi }{r^2}\Vert _{L^\infty (0, \infty )}+\delta ^{\frac{1}{2}}\Vert r^{\frac{1}{2}}(\frac{\chi }{r^2})'\Vert _{L^2(0, \infty )}\le C\delta ^{-1}.\end{aligned}$$ -
(3)
\(\Vert \frac{1}{r}\Vert _{L^1(E)}\lesssim \delta , \ \Vert (1-\lambda r^2)\frac{\chi }{r^{\frac{5}{2}}}\Vert _{L^2(E)}\lesssim \delta ^{\frac{1}{2}}\).
Proof
If \(\lambda \le 0\), then we have \(E=\varnothing \), since \(|1-\lambda r^2|\ge 1\). Setting \(\chi (r)=\frac{1}{\frac{1}{r^2}-\lambda }\), one can directly check that (1) and (2) are satisfied. Now we assume \(\lambda >0\). Denote \(r_{\pm }=\sqrt{\frac{1\pm \delta }{\lambda }}\). We can see that \(E=(r_-, r_+)\) and \(m(E)\approx \delta \lambda ^{-\frac{1}{2}}\).
Choose the piecewise \(C^1\) function \(\chi : (0, \infty ) \rightarrow \mathbb {R}\) as below:
Observing that \(\Vert \chi \Vert _{L^\infty (E)}\lesssim \delta ^{-1}\lambda ^{-1}\), \(\Vert \chi '\Vert _{L^\infty (E)}\lesssim \delta ^{-2}\lambda ^{-\frac{1}{2}}\), and \(r\approx \lambda ^{-\frac{1}{2}}\) on E, we then verify
and
It remains to establish
The first term on the left can be easily estimated due to the fact that \(\frac{\chi }{r^2}=\frac{1}{1-\lambda r^2}\) on \(E^c=\{r> 0: \ |1-\lambda r^2|> \delta \}\).
For the second term, a direct computation yields
which gives the desired inequality. \(\quad \square \)
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
An, X., He, T. & Li, T. Nonlinear Asymptotic Stability and Transition Threshold for 2D Taylor–Couette Flows in Sobolev Spaces. Commun. Math. Phys. 405, 146 (2024). https://doi.org/10.1007/s00220-024-05022-6
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s00220-024-05022-6