Abstract
We first investigate the lower bound for higher eigenvalues \(\lambda _i\) of the Laplace operator on a bounded domain and obtain a sharp lower bound. Then, we extent this estimate of the eigenvalues to general cases. Finally, we study the eigenvalues \(\Gamma _i\) for the clamped plate problem and deliver a sharp bound for the clamped plate problem for arbitrary dimension.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
Let \(\Omega \) be a bounded domain with piecewise smooth boundary \(\partial \Omega \) in an n-dimensional Euclidean space \(\textbf{R}^n\). First of all, we focus on the following Dirichlet eigenvalue problem of Laplacian
It is well known that the spectrum of eigenvalue problem (1.1) is real and discrete (cf. [2, 6, 12, 15, 21])
where each \(\lambda _i\) has finite multiplicity which is counted by its multiplicity.
Let \(V(\Omega )\) be the volume of \(\Omega \), and \(\omega _n\) the volume of the unit ball in \({R}^n\). Then the following well-known Weyl’s asymptotic formula holds
which implies that
In 1961, Pólya [23] proved that, if \(n=2\) and \(\Omega \) is a tiling domain in \({R}^2\), then
Based on the result above, he proposed the famous conjecture:
Conjecture of Pólya. If\(\Omega \) is a bounded domain in \({R}^n\), then k-th eigenvalue \(\lambda _k\) of the eigenvalue problem (1.1) satisfies
During the past six decades, many mathematicians have focused on this problem and the related topics, there are a lot of important results on this aspect (cf. [4, 5, 7, 10, 11, 13, 14, 16, 18]) and we suggest that readers refer [25, 29] for more details. In 1983, Li and Yau [17] verified the famous Li-Yau inequality
It’s seen from the asymptotic formula (1.2), that Li-Yau’s inequality is the best possible in the sense of the average of eigenvalues. From (1.3), one can derive
which gives a partial solution to the Pólya conjecture with a factor \(\frac{n}{n+2}\). This conjecture is still open up to now.
In [20], Melas obtained the following beautiful estimate which improves (1.3) for \(n\ge 1\) and \(k\ge 1\)
where \(c_n\) is a positive constant depending only on n and
is called the moment of inertia of \(\Omega \). In fact \(c_n\le \frac{1}{24(n+2)}\). Obviously,
In the formula (2.27) of [20], Males requires \(c\le \min \{\frac{1}{6},\frac{(2\pi )^2}{{\omega _n}^{\frac{4}{n}}}\}\). According to \(\frac{\omega _n^{\frac{4}{n}}}{(2\pi )^2}\le \frac{1}{2}\), we get \(c\le \frac{1}{6}\). Putting \(c\le \frac{1}{6}\) into the formula (2.27) of [20], we get \(c_n\le \frac{1}{24(n+2)}\) in (1.4).
Afterwards, Kovařík, Vugalter and Weidl [13] improved this results when \(n=2\). They proved that
where \(C(a_0)\) is a positive constant depending on \(a_0\in [0,1]\) and the length of the smooth parts of \(\partial \Omega \), \(\varepsilon (k)=\frac{2}{\sqrt{\log _2(\frac{2\pi k}{c})}}\) and \(c=\sqrt{\frac{3\pi }{14}}10^{-11}\).
The first purpose of this paper is to improve Melas’s estimate (1.4) by giving a sharper polynomial inequality, see Corollary 2.4. For more general cases, where \(n\ge m\ge 2\) and \(k\ge 1\), we obtain a lower bound for eigenvalues in Sect. 3, and we should mention that our result gives a sharp lower bounds by comparing Lemma 2.2 with the polynomial inequality in [20]. As a consequence of our result, we prove the Theorem 3.1. An interesting problem is to investigate the similar problem in a Cartan-Hadamard manifold and we recommend readers to refer to [27, 28] for details.
The second purpose of this paper is to estimate eigenvalues of the following clamped plate problem. Let \(\Omega \) be a bounded domain in \({R}^n\). We consider the following clamped plate problem, which describes characteristic vibrations of a clamped plate:
where \(\Delta \) is the Laplacian operator and \(\nu \) denotes the outward unit normal to the boundary \(\partial \Omega \). As is known, this problem has a real and discrete spectrum (cf. [1])
where each \(\Gamma _i\) has finite multiplicity which is repeated according to its multiplicity.
For the eigenvalues of the clamped plate problem, Agmon [1] and Pleijel [22] gave the following asymptotic formula
This implies that
Furthermore, Levine and Protter [16] proved that the eigenvalues of the clamped plate problem satisfy
The formula (1.6) shows that the coefficient of \(k^{\frac{4}{n}}\) is the best possible in the sense of the average of eigenvalues. Later, Cheng and Wei [8] improved the above estimate as follows:
where \(n\ge 1\) and \(k\ge 1\).
Recently, by using a different method, Cheng and Wei [9] got better lower bounds for eigenvalues of the clamped plate problem and proved that
where \(n\ge 2\) and \(k\ge 1\).
Furthermore, they gave upper bounds for the sum of \(\Gamma _i\),
where \(k\ge V(\Omega )r_0^n\), and
In [30], Yildirim and Yolcu improved Cheng and Wei’s estimates by replacing the last term in the right hand of (1.7) by a positive term of \(k^{\frac{1}{n}}\). For any bounded open set \(\Omega \subseteq R^n\), where \(n\ge 2\) and \( k \ge 1\), Yildirim and Yolcu got the following inequality
where
In Sect. 4, we will improve Yildirim and Yolcu’s [30] estimate (1.8) by giving a shaper polynomial inequality when \(n\ge 3\), see Corollary 4.4.
2 Lower bounds for sums of Dirichlet eigenvalues
In this section we prove the following theorem.
Theorem 2.1
For any bounded domain \(\Omega \subseteq R^n\), \(n\ge 2\) we have
where
\(\alpha \), \(\rho \) are defined by (1.9) and a is defined by (2.16).
Firstly, we introduce some notations and definitions. For a bounded domain \(\Omega \), the moment of inertia of \(\Omega \) is defined by
By a translation of the origin and a suitable rotation of axes, we can assume that the center of mass is the origin and
We now fix a \(k \ge 1\) and let \(u_1,\ldots ,u_k\) denote an orthonormal set of eigenfunctions of (1.1) corresponding to the set of eigenvalues \(\lambda _1(\Omega ),\ldots ,\lambda _k(\Omega )\). We consider the Fourier transform of each eigenfunction
It seems from Plancherel’s Theorem that \(f_1,.\ldots , f_k\) is an orthonormal set in \(R^n\). Since these eigenfunctions \(u_1,\ldots ,u_k\) are also orthonormal in \(L_2(\Omega )\), Bessel’s inequality implies that for every \(\xi \in R^n\)
Since
we have
By the boundary condition, we get
for each \(1\le j\le k\). Set
From (2.1), we have
for each \(\xi \in R^n\). We also get
Assume (by approximating F) that the decreasing function \(\phi : [0,+\infty )\rightarrow [0,(2\pi )^{-n}V(\Omega )]\) is absolutely continuous. Let \(F^*(\xi ) = \phi (|\xi |)\) denote the decreasing radial rearrangement of F. Put \(\mu (t)=|\{F^*>t \}|=|\{F>t\}|\). It follows from the coarea formula that
Since \(F^*\) is radial, we have \(\mu (\phi (s))=|\{F^*>\phi (s) \}|=\omega _n s^n\). Differentiating both side of the above equality, we have \(n\omega _n s^{n-1}={\mu }'(\phi (s))\phi '(s)\) for almost all s. This together with (2.3), \(\rho =2(2\pi )^{-n}\sqrt{V(\Omega )I(\Omega )}\) and the isoperimetric inequality implies
For almost all s, we have
Since the map \(\xi \mapsto |\xi |^2\) is radial and increasing, applying (2.5), we get
and
The following lemma will be used in the proof of Theorem 2.1.
Lemma 2.2
Let \(n\ge 2\), \(\rho >0\), \(A>0\). If \(\psi : [0,+\infty )\rightarrow [0,+\infty )\) is a decreasing function (and absolutely continuous) satisfying
and
Then
where
Proof
We choose the function \(\alpha \psi (\beta t)\) for appropriate \(\alpha , \beta >0\), such that \(\rho = 1\) and \(\psi (0) = 1\). By [20] we can also assume that \(B=\int _0^\infty s^{n+1}\psi (s)ds <\infty \). If we let \(q(s)=-\psi ^{'}(s)\) for \(s\ge 0\), we have \(0\le q(s)\le 1\) and \(\int _0^\infty q(s)=\psi (0)=1.\) Moreover, integration by parts implies that
and
Next, let \(0\le a < +\infty \) satisfies that
By the same argument as in Lemma 1 of [17], such real number a exists. From [20], we have
To estimate the last integral we take \(\tau > 0\) to be chosen later. Applying (2.11) and integrating the both sides of the following inequality
we get
where
Putting, \(\tau = (nA)^{1/n}\) we get
This proves Lemma 2.2.
To prove (2.12), we need to show that for any \(\tau >0\) we have
Taking \(t=\frac{s}{\tau }\), we define f(t) (for \(t>0\)) by
Differentiating, f(t) we have
It follows from the above formula that if \(n\ge 2\), then \(t=1\) is the minimum point of f and \(f\ge \min \{f(1)=0,f(0)=0\}\). This implies
\(\square \)
Next we will give the proof of Theorem 2.1.
Proof of Theorem 2.1
Applying Lemma 2.2 to the function \(\phi \) with \(A=(n\omega _n)^{-1}k, \rho =2(2\pi )^{-n}\sqrt{V(\Omega )I(\Omega )}\) and submitting it to (2.8), we obtain
where \(0<c_1\le 1\) is a constant and a is defined by
We observe the following facts
-
(i)
\(0<\psi (0)\le (2\pi )^{-n}V(\Omega )\),
-
(ii)
if R is a positive constant such that \(\omega _n R^n=V(\Omega )\), then
It follows from the above properties
On the other hand, we consider the following function
for \(t\in (0, (2\pi )^{-n}V(\Omega )]\), where
and
Then we have
By a direct calculation, we see from \(\omega _n=\frac{2\pi ^{\frac{n}{2}}}{n\Gamma (\frac{n}{2})}\) that
where \(\Gamma (t)\) is the Gamma function.
Therefore, in view of (2.18), if
then \(g_2(t)\) is decreasing on \((0, (2\pi )^{-n}V(\Omega )]\). Now we consider another estimate. Setting
where
and
we have
Therefore, we conclude that if
then G(t) is decreasing on \((0, (2\pi )^{-n}V(\Omega )]\). Finally, we obtain
where \(\alpha \), \(\rho \) are defined in the (1.9) and
\(\square \)
Note that \(\lambda _1<\lambda _2\le \lambda _3\le \cdots \). This together with the above lemma implies the following estimate for higher eigenvalues.
Corollary 2.3
For any bounded domain \(\Omega \subseteq R^n\), \(n\ge 2\) and any \( k \ge 1\) we have
where
\(\alpha \), \(\rho \) are defined by (1.9).
In fact, if we choose special a in (2.13), we also have the following result.
Corollary 2.4
For any bounded domain \(\Omega \subseteq R^n\), \(n\ge 2\) and any \( k \ge 1\) we have
Proof
Combining with the formula (2.25) in [20] and (57) in [30], we have
By using similar discussion in the proof of Theorem 2.1, we get
where \(0<C_1\le \frac{1}{6}\) and \(0<C_2\le \frac{1}{9}\) are two constants which will be determined later. We consider the following function
which would be decreasing on \((0, (2\pi )^{-n}V(\Omega )]\) if \(g'((2\pi )^{-n}V(\Omega ))\le 0\). In view of (2.18), the degression of g(t) is equal to the following inequality
Since \(k\ge 1\) and \(\frac{\omega _n^{\frac{4}{n}}}{(2\pi )^2}\le \frac{1}{2}\), we can choose \(C_1=\frac{1}{6}\). Therefore, \(C_2\) satisfies
where
Obviously, \( {\tilde{C}}_2\ge \frac{1}{9}\). Hence, we complete our proof. \(\square \)
3 Lower bounds for Dirichlet eigenvalues in higher dimensions
In this section we will give a universal lower bound on the sum of eigenvalues for \(n\ge m+1\), where \(m\ge 2\).
Theorem 3.1
For any bounded domain \(\Omega \subseteq R^n\), \(n\ge m+1\ge 3\) and \( k \ge 1\), we have
where
\(\alpha =\frac{V(\Omega )}{(2\pi )^n},\,\,\,\,\rho =2(2\pi )^{-n}\sqrt{V(\Omega )I(\Omega )}\) and a is defined by (2.16).
The following lemma will be used in the proof of Theorem 3.1.
Lemma 3.2
For an integer \(n\ge m+1\ge 0\) and positive real numbers s and \(\tau \) we have the following inequality:
Proof
Setting \(t=\frac{s}{\tau }\), and putting
for \(t\ge 0\), we get
It follows from the above formula that if \(n \ge m+1\), then \(t = 1\) is the minimum point of f(t) and \(f \ge \min \{f(1) = 0,f(0) = 0\}\). So, we get
\(\square \)
Next we will give the proof of Theorem 3.1.
Proof of Theorem 3.1
For \(l\ge 0\), \(\tau \ge \frac{1}{2}\) and \(a \ge 0\), we have
where
Therefore, we get
From
and
we obtain
Choosing \(\tau =(nA)^{\frac{1}{n}}\), we get
It follows from (3.3) that
From (2.8), we know
In view of \(A=\frac{k}{n\omega _n}\), we have
where \(0<c_2\le 1\) is a constant.
When \(m=1\), we complete the proof of Theorem 2.1 in Sect. 2. We assume that \(m\ge 2\). Putting
where
and
we have
When
we get that \(g_2(t)\) is decreasing on \((0, (2\pi )^{-n}V(\Omega )]\) by using the following formulas
Hence g(t) is also decreasing on \((0, (2\pi )^{-n}V(\Omega )]\). This implies
where
and
\(\square \)
From the above lemma, we have the following universal lower bounds for higher eigenvalues.
Corollary 3.3
For any bounded domain \(\Omega \subseteq R^n\), \(n\ge m+1\ge 3\) and \( k \ge 1\) we have
where
\(\alpha =\frac{V(\Omega )}{(2\pi )^n},\,\,\,\,\rho =2(2\pi )^{-n}\sqrt{V(\Omega )I(\Omega )}\) and a is defined by (2.16).
Due to the similar discussion to Corollary 2.4, we have
Corollary 3.4
For any bounded domain \(\Omega \subseteq R^n\), \(n\ge 3\) and any \( k \ge 1\) we have
Proof
According to Lemma 3.2, we have
By using similar discussion in the proof of Theorem 2.1, we get
where \(0<C_3\le \frac{3}{80}\) is a constant which will be chosen. By using the similar discussion in the proof of Corollary 4.4, one can choose \(C_3=\frac{3}{80}\). Hence, we complete our proof. \(\square \)
4 A universal lower bound on eigenvalues of the clamped plate problem
In this section, let \(\phi (z)\) be the decreasing radial rearrangement of h(z) where h(z) is defined as (4.9). Then, a is defined by
We will give a universal lower bounds on the sum of eigenvalues for \(n\ge m\), where \(m\ge 1\).
Theorem 4.1
For any bounded domain \(\Omega \subseteq R^n\), \(n\ge m\ge 1\) and \( k \ge 1\) we have
-
(1)
When \(n=1\) and
$$\begin{aligned} \frac{2\sqrt{2}S_3}{5} \le k, \end{aligned}$$we have
$$\begin{aligned} \begin{aligned} \sum _{i=1}^k\Gamma _i \ge&\omega _n^{-\frac{4}{n}} {\alpha ^{-\frac{4}{n}}k^{1+\frac{4}{n}}} -\omega _n^{\frac{m-4}{n}} \frac{4S_{m+2}}{(n+4)\rho ^{m}}\alpha ^{\frac{mn+m-4}{n}}k^{\frac{n-m+4}{n}}\\&+\omega _n^{\frac{m-3}{n}} \frac{4mS_{m+2}}{(n+4)(m+2)\rho ^{m+1}}\alpha ^{\frac{(m+1)n+m-3}{n}}k^{\frac{n-m+3}{n}}, \end{aligned} \end{aligned}$$(4.2)where \(\alpha \), \(\rho \) are defined by (1.9) and
$$\begin{aligned} S_l=(a+1)^l-a^l. \end{aligned}$$ -
(2)
When \(m\ge 2\), we have
$$\begin{aligned} \begin{aligned} \sum _{i=1}^k\Gamma _i \ge&\omega _n^{-\frac{4}{n}} {\alpha ^{-\frac{4}{n}}k^{1+\frac{4}{n}}} -\omega _n^{\frac{m-4}{n}} \frac{4S_{m+2}}{(n+4)\rho ^{m}}\alpha ^{\frac{mn+m-4}{n}}k^{\frac{n-m+4}{n}}\\&+c_3\omega _n^{\frac{m-3}{n}} \frac{4mS_{m+2}}{(n+4)(m+2)\rho ^{m+1}}\alpha ^{\frac{(m+1)n+m-3}{n}}k^{\frac{n-m+3}{n}}, \end{aligned} \end{aligned}$$where
$$\begin{aligned} c_3\le \min \left\{ 1, \frac{2^{\frac{m+1}{2}}(n+2)(m+2)}{S_{m+2}[(m+1)n+m-3]}k^{\frac{m+1}{n}} \right\} . \end{aligned}$$
Next, we recall the definition and serval properties of the symmetric decreasing rearrangements. Let \(\Omega \subset R^n \) be a bounded domain. Its symmetric rearrangement \(\Omega ^*\) is the open ball with the same volume as \(\Omega \),
By using a symmetric rearrangement of \(\Omega \), we have
Then we have
The following lemma is useful in the proof of Theorem 4.1.
Lemma 4.2
For integers \(n\ge m\ge 1\) and positive real numbers s and \(\tau \), we have the following inequality:
Proof
Taking \(t=\frac{s}{\tau }\), and putting f(t)
for \(t\ge 0\), we get
From the above formula, it is clear that when \(n \ge m\), we have \(t = 1\) is the minimum point of f(t) and then \(f \ge \min \{f(1) = 0,f(0) = 0\}\). We get
\(\square \)
Now, we will give the proof of Theorem 4.1.
Proof of Theorem 4.1
Let \(\{u_j\}_{j=1}^{\infty }\) be the eigenfunction corresponding to the eigenvalue \(\Gamma _j\), \(j=1,2.\ldots \) which satisfy
Thus, \(\{ u_j \}_{j=1}^{\infty }\) forms an orthonormal basis of \(L^2(\Omega )\). We define a function \(\varphi _j\) by
Denote by \({\widehat{\varphi }}_j(z)\) the Fourier transform of \(\varphi _j (x)\). For any \(z \in \textbf{R}^n\), we have
By the Plancherel formula, we have
for any i, j. Since \(\{ u_j \}_{j=1}^{\infty }\) is an orthonormal basis in \(L^2(\Omega )\), the Bessel inequality implies that
For each \(j = 1,\ldots ,k\), we deduce from the divergence theorem and \(u_j|_{\partial \Omega }=\frac{\partial u_j}{\partial \nu }|_{\partial \Omega }=0\) that
It follows from the Parseval’s identity that
Since
we obtain
Putting
one derives from (4.6) that \(0 \le h(z) \le (2\pi )^{-n}V(\Omega )\). It follows from (4.8) and the Cauchy-Schwarz inequality that
for every \(z\in \textbf{R }^n\). From the Parseval’s identity, we derive
Applying the symmetric decreasing rearrangement to h(z) and noting that \(\zeta =\sup |\nabla h|\le 2(2\pi )^{-n}\sqrt{V(\Omega )I(\Omega )}:=\eta ,\) we see from (2.6)
for almost every s. According to (4.4) and (4.7), we infer
In order to apply Lemma 4.2, from (4.4) and the definition of A, we take
From (4.3), we deduce that
On the other hand, \(0 < \phi (0) \le \sup h^*(z) = \sup h(z) \le (2\pi )^{-n} V(\Omega )\).
For any \(k\ge 1\) and \(a \ge 0\), we have
where
Let \(D'=\int ^{a+1}_a s^{n+4}ds\), from the above lemma, integrating the both sides of (4.5) over \([a,a+1]\), we get
From
and
we get
This implies that
Taking \(\tau =(nA)^{\frac{1}{n}}\), we get
Then, we get
According to (4.4), (4.7) and the above inequality, we conclude
For \(m=1\) and \(n=1\), we define f(t) as follows
on \((0,(2\pi )^{-n}V(\Omega )]\), where
and
for \(0<\xi \le 1\). Then
When
we prove that f(t) decreases on \((0,(2\pi )^{-n}V(\Omega )]\) by using
and
Therefore, if
we get
where
and
Noting that \(A=\frac{k}{n\omega _n}\), we obtain the following inequality
When \(m\ge 2\), F(t) is defined by
for \(t\in (0,(2\pi )^{-n}V(\Omega )]\), where
for \(0<c_3\le 1\) and
This implies
So, if
we obtain that F(t) decreases on \((0,(2\pi )^{-n}V(\Omega )]\), which yields that
where
\(\square \)
For higher eigenvalues, we have the following universal lower bounds
Corollary 4.3
For any bounded domain \(\Omega \subseteq R^n\), \(n\ge m\ge 1\) and any \( k \ge 1\) we have
-
(1)
When \(n=1\) and
$$\begin{aligned} \frac{2\sqrt{2}S_3}{5} \le k, \end{aligned}$$we have
$$\begin{aligned} \begin{aligned} \Gamma _k \ge&\omega _n^{-\frac{4}{n}} {\alpha ^{-\frac{4}{n}}k^{\frac{4}{n}}} -\omega _n^{\frac{m-4}{n}} \frac{4S_{m+2}}{(n+4)\rho ^{m}}\alpha ^{\frac{mn+m-4}{n}}k^{\frac{-m+4}{n}}\\&+\omega _n^{\frac{m-3}{n}} \frac{4mS_{m+2}}{(n+4)(m+2)\rho ^{m+1}}\alpha ^{\frac{(m+1)n+m-3}{n}}k^{\frac{-m+3}{n}}, \end{aligned} \end{aligned}$$(4.17)where \(\alpha \), \(\rho \) are defined by (1.9) and
$$\begin{aligned} S_l=(a+1)^l-a^l. \end{aligned}$$ -
(2)
When \(m\ge 2\), we have
$$\begin{aligned} \begin{aligned} \Gamma _k \ge&\omega _n^{-\frac{4}{n}} {\alpha ^{-\frac{4}{n}}k^{\frac{4}{n}}} -\omega _n^{\frac{m-4}{n}} \frac{4S_{m+2}}{(n+4)\rho ^{m}}\alpha ^{\frac{mn+m-4}{n}}k^{\frac{-m+4}{n}}\\&+c_3\omega _n^{\frac{m-3}{n}} \frac{4mS_{m+2}}{(n+4)(m+2)\rho ^{m+1}}\alpha ^{\frac{(m+1)n+m-3}{n}}k^{\frac{-m+3}{n}}, \end{aligned} \end{aligned}$$(4.18)where
$$\begin{aligned} c_3\le \min \left\{ 1, \frac{2^{\frac{m+1}{2}}(n+2)(m+2)}{S_{m+2}[(m+1)n+m-3]}k^{\frac{m+1}{n}} \right\} . \end{aligned}$$
According to Lemma 4.2 and the proof of Theorem 4.1, we also have the following result.
Corollary 4.4
For any bounded domain \(\Omega \subseteq R^n\), \(n\ge 3\) and any \( k \ge 1\) we have
Proof
By using Lemma 4.2, we get
In view of (4.15) and (57) in [30], integrating the both sides of the above inequality over \([a,a+1]\), we have
By using similar discussion in the proof of Theorem 4.1 and taking \(\tau =(nA)^{\frac{1}{n}}\), we get
Hence,we arrive at
where \(0<d_1\le 1\) is a constant to be determined. Let \(t\in (0,(2\pi )^{-n}V(\Omega )]\), we define
which would be decreasing on \((0, (2\pi )^{-n}V(\Omega )]\) if \(Q'((2\pi )^{-n}V(\Omega ))\le 0\). Obviously, \(Q'((2\pi )^{-n}V(\Omega ))\le 0\) is equal to
Due to (2.18) and \(\frac{\omega _n^{\frac{4}{n}}}{(2\pi )^2}\le \frac{1}{2}\), if
we have \(Q'((2\pi )^{-n}V(\Omega ))\le 0\), where
By direct computation, one has \(d_0>1\). Therefore, we obtain the following eigenvalue inequality
\(\square \)
References
Agmon, S.: On kernels, eigenvalues and eigenfunctions of operators related to elliptic problems. Commun. Pure Appl. Math. 18, 627–663 (1965)
Ashbaugh, M.S., Benguria, R.D.: Universal bounds for the low eigenvalues of Neumann Laplacian in \(n\) dimensions. SIAM J. Math. Anal. 24, 557–570 (1993)
Bandle, C.: Isoperimetric inequalities and applications. Pitman Monographs and Studiesin Mathematics, vol. 7. Pitman, Boston (1980)
Berezin, F.A.: Covariant and contravariant symbols of operators. Izv. Akad. Nauk SSSR Ser. Mat. 36, 1134–1167 (1972)
Bergweiler, W., Eremenko, A.: Proof of a conjecture of Pólya on the zeros of successive derivatives of real entire functions. Acta Math. 197, 145–166 (2006)
Chavel, I.: Eigenvalues in Riemannian Geometry. Academic Press, New York (1984)
Cheng, Q-M., Qi, X.R.: Lower bound etimates for eigenvalues of the Laplacian. arXiv:1104.5298 (2012)
Cheng, Q.-M., Wei, G.X.: A lower bound for eigenvalues of a clamped plate problem. Calc. Var. Part. Diff. Equ. 42, 579–590 (2011)
Cheng, Q.-M., Wei, G.X.: Upper and lower boundes for eigenvalues of the clamped plate problem. J. Differ. Equ. 255, 220–233 (2013)
Frank, L., Loss, M., Weidl, T.: Pólya’s conjecture in the presence of a constant magnetic field. J. Eur. Math. Soc. 2, 1365–1383 (2009)
Ilyin, A., Laptev, A.: Berezin-Li-Yau inequalities on domains on the sphere. J. Math. Anal. Appl. 6, 1253–1269 (2019)
Ji, Z.C.: New Bounds on Eigenvualues of Laplacian. Acta Math. Sci. Ser. B (Engl. Ed.) 2, 545–550 (2019)
Kovařík, H., Vugalter, S., Weidl, T.: Two-dimensional Berezin-Li-Yau inequalities with a correction term. Comm. Math. Phys. 3, 959–981 (2009)
Kwaśnicki, M., Laugesen, S., Siudeja, A.: Pólya’s conjecture fails for the fractional Laplacian. J. Spectr. Theory 1, 127–135 (2019)
Laptev, A.: Dirichlet and Neumann eigenvalue problems on domains in Euclidean spaces. J. Funct. Anal. 151, 531–545 (1997)
Levine, H.A., Protter, M.H.: Unrestricted lower bounds for eigenvalues for classes of elliptic equations and systems of equations with applications to problems in elasticity. Math. Methods Appl. Sci. 7, 210–222 (1985)
Li, P., Yau, S.T.: On the Schrödinger equations and the eigenvalue problem. Comm. Math. Phys. 88, 309–318 (1983)
Lieb, E.: The number of bound states of one-body Schrödinger operators and the Weyl problem. Proc. Symp. Pure Math. 36, 241–252 (1980)
Lin, F.H.: Extremum problems of Laplacian eigenvalues and generalized Pólya conjecture. Chin. Ann. Math. Ser. B 2, 497–512 (2017)
Melas, A.D.: A lower bound for sums of eigenvalues of the Laplacian. Proc. Am. Math. Soc. 131, 631–636 (2003)
Payne, L.E., Pólya, G., Weinberger, H.F.: On the ratio of consecutive eigenvalues. J. Math. Phys. 35, 289–298 (1956)
Pleijel, A.: On the eigenvalues and eigenfunctions of elastic plates. Commun. Pure Appl. Math. 3, 1–10 (1950)
Pólya, G.: On the eigenvalues of vibrating membranes. Proc. Lond. Math. Soc. 11, 419–433 (1961)
Pólya, G., Szegö, G.: Isoperimetric inequalities in mathematical physics. Annals of Mathematics Studies Number 27. Princeton university press, Princeton (1951)
Schoen, R., Yau, S.T.: Lectures on Differential Geometry. Int Press, Boston (1994)
Wang, Q.L., Xia, C.Y.: Universal bounds for eigenvalues of the biharmonic operator on Riemannian manifolds. J. Funct. Anal. 245, 334–352 (2007)
Xu, Z.Y., Xu, H.W.: A generalization of Pólya conjecture and Li-Yau inequalities for higher eigenvalues. Calc. Var. Part. Diff. Equ. 5, 19 (2020)
Xu, H. W.: Estimates of Higher Eigenvalues for Minimal Submanifolds. Differential Geometry, pp. 288-300. World Scientific, Singapore (1993)
Yau, S. T.: Problem section. Seminar on Differential Geometry, pp. 669-706, Ann. of Math. Stud., 102, Princeton Univ. Press, Princeton, N.J., (1982)
Yolcu, S.Y., Yolcu, T.: Estimates on the eigenvalues of the clamped plate problem on domains in Euclidean spaces. J. Math. Phys. 54, 1–13 (2013)
Acknowledgements
This work was supported by the National Natural Science Foundation of China, Grant No. 12071424, 11531012. The first author was also supported by the Postdoctoral Fund of Zhejiang Province, China, Grant No. ZJ2022004. Ji would like to thank Professor Kefeng Liu for his continued support, advice and encouragement.
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by Juergen Jost.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Ji, Z., Xu, H. Lower bounds for eigenvalues of Laplacian operator and the clamped plate problem. Calc. Var. 62, 175 (2023). https://doi.org/10.1007/s00526-023-02506-6
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s00526-023-02506-6