Exponentially Convergent Sparse Discretizations and Application to Near Surface Geophysics Murthy N. Guddati North Carolina State University November 9, 2017
Outline Part 1: Impedance Preserving Discretization Part 2: Absorbing Boundary Conditions (1-sided DtN map) Joint work with Tassoulas, Druksin, Lim, Zahid, Savadatti, Thirunavukkarasu Part 3: Complex-length FEM for finite domains (2-sided DtN map) Joint work with Druskin, Vaziri Astaneh Part 4: Inversion for Near-surface Geophysics Joint work with Vaziri Astaneh 2
Part1. Impedance Preserving Discretization
Model Problem 2 2 2 2 u u u 1 u 3D wave equation in free space 0 2 2 2 2 2 x y z c t ik y ik z i t Fourier transform in t, y, z, with u Ue y z 2 2 U 2 2 2 k U 0 k k k is complex valued 2 y z 2 x c Exact solution: ikx ikx U Ae Be , where is the horizontal wavenumber k . is a plane/evanescent wave i kx k y k z ( t ) ikx U e u e y z 4
Finite Element Solution on a Uniform Grid in x 2 U FE discretization of: 2 k U 0 2 x Element contribution matrix with uniform element size of h : 2 2 1 1 k h 1 1 A h 3 1 1 A B 1 3 6 2 k k h elem h 1 1 1 1 B A 2 2 1 k h B 1 6 3 h 6 Assembly results in the difference equation: BU 2 AU BU 0 j 1 j j 1 5
Changing Mesh Size: Reflections A simple analysis using two uniform meshes with different element sizes (h, H), but the same material What happens when a right propagating wave hits the interface? Exact solution – just passes through Finite element solution – reflections due to impedance mismatch Z : discrete impedance of left domain Z Z h H h R Z Z Z : discrete impedance of right domain H H h 6
Computing Discrete Impedance (Half-space Stiffness) Basic idea: discrete half-space + finite element = discrete half-space A B U Z U A Z B U 0 0 h 0 h 0 2 2 2 A Z B h B A Z U 0 B A Z U 0 h 1 h 1 2 2 1 1 k h A 2 kh h 3 2 2 Z A B ik 1 h 2 2 1 2 1 k h B 1 h 6 Error term . depends on element size, resulting impedance mismatch when Z h the element size changes, resulting in reflections 7
Optimal Integration for Minimizing Reflection Error Minimize the error in impedance by using generalized integration rules 2 1 1 2 2 A 1 k h 2 kh h 4 2 2 2 Z A B ik 1 h 4 2 1 1 2 2 B 1 k h Error term h 4 Minimize the error term by choosing 0 The error in impedance is completely eliminated! No more reflections Formally valid for more general 2 nd order equations (anisotropic, visco- elasticity etc., electromagnetics etc. – G, 2006, CMAME) Linear elements + midpoint integration = Impedance Preserving Discretization 8
Part 2. Absorbing Boundary Conditions Perfectly Matched Discrete Layers
Perfectly Matched Discrete Layers …Impedance Preserving Discretization of PML Perfectly Matched Layers (PML) (Berenger, 1994; Chew et.al. 1995) Step 1: Bend the domain into complex space Reduced reflection into the interior e 2 ikL R PML Interior (real x) PML Region (imaginary or complex x) Step 2: discretize PMDL domain (in complex space) Impedance is no longer preserved; perfect matching is destroyed Requires a large number of carefully chosen PML layers Impedance preserving discretization comes to the rescue! Impedance is preserved/matched, irrespective of element length, small, large, real, complex – Perfectly Matching Discrete Layers (PMDL) 2 Discretize with 3-5 complex-length linear finite elements j nlayer 1 ikL / 2 j R PMDL No discretization error, but truncation causes 1 ikL / 2 j 1 j reflections. The reflection coefficient is derived as 10
PMDL vs PML: Effectiveness of Midpoint Integration PML with 3 layers PMDL with 3 layers 11
PMDL: Some More Old Results Impedance preservation property is valid for any equation that is linear and second order in space (G, CMAME, 2006) Elastic and other complicated wave equations (G, Lim & Zahid, 2007) PMDL with 5 layers PML with 5 layers Evanescent waves can be treated effectively Padded PMDL – contains large real lengths with midpoint integration (Zahid & G, CMAME, 2006) 12
Salient Features of PMDL 2 j nlayer k k Exponential convergence j R k k j 1 j Near optimal discretization Optimal: need staggered grids (with Druskin et al., 2003) Links PML to rational ABCs Lindman, Engquist-Majda, Higdon and variants (e.g. CRBC) We started this from E-M/Higdon ABCs (G, Tassoulas, 2000) Extensions to corners is straightforward Additional advantage: Provides solutions to some difficult cases Backpropagating waves: anisotropy PML for discrete/periodic media 13
PMDL for Backpropagating Waves Opposing signs of phase and group velocities Backpropagating waves grow in the PML region PML cannot work! (Bécache, Fauqueux and Joly, 2003) Reduced Reflection into the interior PML result: radiation in anisotropic media PML Region Interior A counter-intuitive idea: make the reflections in PML region decay faster than the growth of the incident wave Works only with PMDL: needs impedance preserving discretization! Result from PMDL after the fix Savadatti & G (2012), J Comp. Phys. 14
Anisotropic elasticity – Tilted Elliptic Case Arbitrary parameters Ideal Slowness Stable parameters Savadatti & G (2012), J Comp. Phys. 15
Anisotropic elasticity – Non-elliptic Case Traditional mesh Two different coordinated materials Savadatti & G (2012), J Comp. Phys. 16
PMDL for Periodic Media (after discretization) Periodic media has internal reflections and transmissions Constructive interference leads to long-range propagation PML’s complex stretching spoils this balance and internal reflections and transmissions get mixed up! PML for Lattice Waves: 7% reflections w/20 PML Basic Ideas (Discrete/Periodic PMDL): layers Periodic media = Discrete vector wave equation (vector size = ndof in a cell) Discrete vector equation = impedance preserving discretization of more complicated wave equation Apply PMDL on the complicated wave equation results in impedance matching for periodic media Discrete PMDL: less than Open problem: stability for complex problems 1% error w/ 4 PMDL layers G & Thirunavukkarasu, JCP (2009), Waves 2011 17
Part 3. Two-Sided DtN Map Complex-length Finite Element Method
Facilitating the Approximation of 2-Sided DtN Map 2 u Consider the equation: 2 u 0, z (0, ) L 2 z A B Exact 2-sided DtN map: K = exact B A By definition, exact DtN Map is impedance preserving: 2 2 2 A B Z exact Consider impedance preserving discretization of the interval: A B 2 2 2 K = , A B Z exact exact B A Error in A and B would be similar since: 2 2 2 2 2 A B Z A B exact Approximating two-sided map reduces to approximating one-sided map Better derivation based on Crank-Nicolson discretization of the propagator 19
1D Helmholtz Equation 2 u 2 u 0 2 z z 0 exp(- i L ) exp(+ i L ) 1 st Order Form v u / z z L u 0 i i u i Eigenvalues z v i / i 0 v i / Downwardwaves: z 20
Recommend
More recommend