Inverse elastic obstacle scattering problems by monotonicity method

Mengjiao Bai, Huaian Diao, Weisheng Zhou · 2025

Read Inverse elastic obstacle scattering problems by monotonicity method by Mengjiao Bai, Huaian Diao, Weisheng Zhou (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-04655.txt.

Introduction

Bai, Diao, and Zhou tackle an imaging problem in a mathematics paper published in 2025: finding hidden rigid obstacles by bouncing elastic waves off them. Only the far-off echo pattern can be measured. From that pattern they must recover where each obstacle sits and what shape it takes. Their method rests on monotonicity, a one-way mathematical relationship. Test regions are compared against the true scatterer through the wave operator, which they split into factors.

Special wave functions that concentrate in small spots make the test exact. The criterion decides, for any trial region, whether it lies inside or outside the obstacle. Repeating the test paints the boundary. No initial guess and no knowledge of the background material are required. Such methods serve nondestructive testing of structures and medical imaging alike. The paper proves the identification is unique.

Catalog record

CREATORMengjiao Bai, Huaian Diao, Weisheng Zhou
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE54 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-04655

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Inverse elastic obstacle scattering problems by monotonicity method Mengjiao Bai, Huaian Diao, Weisheng Zhou Abstract . We consider the elastic wave scattering problem involving rigid obstacles. This work addresses the inverse problem of reconstructing the position and shape of such obstacles using far-field measurements. A novel monotonicity-based approach is developed for this purpose. By factorizing the far-field operator and utilizing the existence of localized wave functions, we derive a shape characterization criterion for the obstacle boundary. The proposed method employs monotonicity tests to determine the geometric relationship between any given test domain and the actual scatterer. As a result, the shape and location of rigid elastic obstacles can be uniquely identified without requiring any initial guesses or prior knowledge of the physical parameters of the homogeneous background medium. Keywords: inverse elastic obstacle problem, factorization method, localized wave functions 1. Introduction 1.1. Mathematical setup Let D\subset\mathbb{R}^{2} be a bounded elastic obstacle such that its exterior \mathbb{R}^{2}\setminus\overline{D} is connected. The elastic medium surrounding the obstacle is assumed to be homogeneous, isotropic, and occupies the entire plane \mathbb{R}^{2} . The background medium is characterized by the Lamé constants \lambda and \mu , which satisfy the strong ellipticity conditions : \mu>0\quad\text{and}\quad\lambda+\mu>0. We assume that the medium has unit (normalized) density. The obstacle D is illuminated by an incident time-harmonic elastic plane wave {\mathbf{u}}^{i} , which is a superposition of a compressional (P-) wave and a shear (S-) wave. The incident wave is given by (1.1) \displaystyle{\mathbf{u}}^{i}(\mathbf{x})=a_{p}\mathbf{d}\,e^{\mathrm{i}k_{p}\mathbf{d}\cdot\mathbf{x}}+a_{s}\mathbf{d}^{\perp}\,e^{\mathrm{i}k_{s}\mathbf{d}\cdot\mathbf{x}}, where \mathbf{d},\mathbf{d}^{\perp}\in\mathbb{S}^{1} are unit vectors such that \mathbf{d} is the direction of wave propagation, and \mathbf{d}^{\perp} is orthogonal to \mathbf{d} . The coefficients a_{p},a_{s}\in\mathbb{C} are not simultaneously zero, i.e., (a_{p},a_{s})\neq(0,0) . The wavenumbers for the compressional and shear waves are defined by k_{p}=\frac{\omega}{\sqrt{\lambda+2\mu}},\quad k_{s}=\frac{\omega}{\sqrt{\mu}}, where \omega>0 is the angular frequency. It can be directly verified that {\mathbf{u}}^{i} satisfies the two-dimensional Navier equation in \mathbb{R}^{2} : (1.2) \mu\Delta{\mathbf{u}}^{i}+(\lambda+\mu)\nabla\nabla\cdot{\mathbf{u}}^{i}+\omega^{2}{\mathbf{u}}^{i}={\mathbf{0}}. The interaction of the incident wave {\mathbf{u}}^{i} with the obstacle D gives rise to a scattered field {\mathbf{u}} , which also satisfies the Navier equation in the exterior domain: (1.3) \mu\Delta{\mathbf{u}}+(\lambda+\mu)\nabla\nabla\cdot{\mathbf{u}}+\omega^{2}{\mathbf{u}}={\mathbf{0}},\quad\text{in }\mathbb{R}^{2}\setminus\overline{D}. The total field is defined as the sum of the incident and scattered fields: {\mathbf{u}}^{\mathrm{tot}}:={\mathbf{u}}^{i}+{\mathbf{u}}. In this paper, we consider the case where the obstacle D is impenetrable and rigid . This implies that the total displacement field vanishes on the boundary of the obstacle: (1.4) {\mathbf{u}}^{\mathrm{tot}}={\mathbf{0}},\quad\text{on }\partial D. Additionally, the scattered field \mathbf{u} is required to satisfy the Kupradze radiation condition (1.5) \lim_{r\to\infty}r^{\frac{1}{2}}\left(\frac{\partial{\mathbf{u}}_{p}}{\partial r}-{\rm i}k_{p}{\mathbf{u}}_{p}\right)=0,\quad\lim_{r\to\infty}r^{\frac{1}{2}}\left(\frac{\partial{\mathbf{u}}_{s}}{\partial r}-{\rm i}k_{s}{\mathbf{u}}_{s}\right)=0,\ r:=|{\mathbf{x}}|, where {\mathbf{u}}_{p}=-\frac{1}{k_{p}^{2}}\nabla\nabla\cdot\mathbf{u},\quad{\mathbf{u}}_{s}=-\frac{1}{k_{s}^{2}}\mathbf{curl}curl\mathbf{u}, are known as the compressional and shear wave components of \mathbf{u} respectively. Given a vector function \mathbf{u}:=\left(u_{1},u_{2}\right)^{\top} and a scalar function u , the two-dimensional \mathbf{curl} and curl operators is defined as follows (1.6) {\rm curl}~\bm{u}=\partial_{1}u_{2}-\partial_{2}u_{1},\quad{\bf curl}~{u}=(\partial_{1}u,-\partial_{2}u)^{\top}. Then, we define the Helmholtz decomposition of the scattered wave {\mathbf{u}} in ( 1.4 ) as follows {\mathbf{u}}={\mathbf{u}}_{p}+{\mathbf{u}}_{s}. It is well-known that the scattered field satisfying the Kupradze radiation condition has the following asymptotic behavior (1.7) {\mathbf{u}}({\mathbf{x}})=\frac{{e}^{\mathrm{i}k_{p}|\mathbf{x}|}}{\sqrt{|\mathbf{x}|}}\mathbf{u}_{p}^{\infty}(\hat{{\mathbf{x}}})+\frac{{e}^{\mathrm{i}k_{s}|\mathbf{x}|}}{\sqrt{|\mathbf{x}|}}\mathbf{u}_{s}^{\infty}(\hat{{\mathbf{x}}})+O\left(\frac{1}{|\mathbf{x}|^{3/2}}\right),\quad|\mathbf{x}|\to\infty, where \mathbf{u}_{p}^{\infty} and \mathbf{u}_{s}^{\infty} defined on the unit sphere \mathbb{S}:=\left\{\hat{\mathbf{x}}\in\mathbb{R}^{2}\big||\hat{\mathbf{x}}|=1\right\} are vector-valued analytic functions and referred to the far-field pattern of \mathbf{u}_{p} and \mathbf{u}_{s} respectively. We typically consider ( \mathbf{u}_{p}^{\infty},\mathbf{u}_{s}^{\infty} ) as an element in \left[L^{2}(\mathbb{S})\right]^{2} space, and denote by \mathbf{u}^{\infty}\left(\cdot,\mathbf{d},a_{p},a_{s}\right) the far-field pattern associated with the scattered field produced by incident waves of the form ( 1.1 ) . In this work, our goal is to reconstruct the shape of the rigid scatterer from the far-field pattern \mathbf{u}^{\infty}\left(\cdot,\mathbf{d},a_{p},a_{s}\right) . 1.2. Connections to previous results and main findings Inverse scattering problems have been widely applied in various fields such as oil exploration and medical imaging. Over the past three decades, numerical methods for solving inverse scattering problems have undergone rapid development. Addressing inverse scattering problems typically requires overcoming their inherent nonlinearity and ill-posedness. To this end, iterative methods were initially proposed [ 17 ] . However, these methods typically require prior knowledge of certain a priori information and necessitate solving a forward problem at each iteration step, which significantly increases computational costs. To enhance computational efficiency, non-iterative methods have been introduced, among which sampling-type methods have garnered considerable attention due to their computational efficiency and ease of implementation. Examples of such methods include the direct sampling method [ 3 ] , linear sampling method [ 5 ] , factorization method [ 13 ] and monotonicity method [ 18 , 11 ] . The factorization method, first introduced by Arens [ 3 ] for the theoretical analysis of two-dimensional rigid obstacle scattering. Additionally, the factorization method has been extended to penetrable scatterers [ 4 ] and crack detection [ 10 ] . Monotonicity methods, as a distinct category of non-iterative techniques, was originally introduced for solving electrical impedance tomography (EIT) problems [ 18 ] , where the authors developed a non-iterative inversion technique based on the monotonicity of the resistance matrix. Later, Harrach and Ullrich proposed a shape reconstruction method for EIT by leveraging the monotonicity of the Neumann-to-Dirichlet (NtD) operator with respect to conductivity [ 11 ] . In recent years, researchers further applied the monotonicity method to the reconstruction of elastic inclusions, proposing normalized and linearized monotonicity methods to recover the inhomogeneities in Lamé parameters and density relative to the background medium using the monotonicity of the NtD operator [ 6 , 7 , 8 ] . It is noteworthy that in these monotonicity-based methods for elastic inclusion reconstruction, authors typically rely on the monotonicity of the NtD operator to reconstruct inhomogeneities within the inclusions. Meanwhile, we note that references [ 1 , 2 ] proposed an innovative scheme based on the monotonicity method, achieving position reconstruction of acoustic and electromagnetic wave scatterers using far-field data. The implementation of this monotonicity method relies on both the advancement of the factorization method and the existence of localized wave functions—specifically, the existence of a sequence of functions that can attain arbitrarily large norms in certain designated regions while exhibiting arbitrarily small norms in other designated regions. In this study, we aim to reconstruct the shape of a rigid obstacle from far-field measurements using a monotonicity-based approach. By leveraging the factorization of the far-field operator and the existence of localized wave functions, we establish a shape characterization theorem for elastic obstacles. Specifically, the reconstruction is achieved by computing the number of positive (negative) eigenvalues of a linear combination of the far-field operator and a probe operator. This allows us to determine whether a test domain lies inside the obstacle, ultimately leading to a complete shape reconstruction. Finally, the remainder of the paper is structured as follows. In Section 2 , we present a factorization result for the far-field operator associated with rigid elastic scatterers and define key operators and notations used in subsequent proofs. Section 3 is devoted to proving the existence of localized wave functions, i.e., constructing a function sequence that diverges within a specified region while vanishing elsewhere. In Section 4 , we establish a shape characterization theorem for the reconstruction of rigid elastic scatterers, demonstrating that the shape of an impenetrable obstacle can be approximately reconstructed by computing the numbers of positive and negative eigenvalues of a specific operator combination. 2. Decomposition of The Far-field Operator This section presents the key properties of the single-layer potential operator, the factorization theorem for the far-field

Read the complete text