Mercurial > repos > public > sbplib
view +scheme/Heat2dCurvilinear.m @ 1198:2924b3a9b921 feature/d2_compatible
Add OpSet for fully compatible D2Variable, created from regular D2Variable by replacing d1 by first row of D1. Formal reduction by one order of accuracy at the boundary point.
author | Martin Almquist <malmquist@stanford.edu> |
---|---|
date | Fri, 16 Aug 2019 14:30:28 -0700 |
parents | 8d73fcdb07a5 |
children |
line wrap: on
line source
classdef Heat2dCurvilinear < scheme.Scheme % Discretizes the Laplacian with variable coefficent, curvilinear, % in the Heat equation way (i.e., the discretization matrix is not necessarily % symmetric) % u_t = div * (kappa * grad u ) % opSet should be cell array of opSets, one per dimension. This % is useful if we have periodic BC in one direction. properties m % Number of points in each direction, possibly a vector h % Grid spacing grid dim order % Order of accuracy for the approximation % Diagonal matrix for variable coefficients KAPPA % Variable coefficient D % Total operator D1 % First derivatives % Second derivatives D2_kappa H, Hi % Inner products e_l, e_r d1_l, d1_r % Normal derivatives at the boundary alpha % Vector of borrowing constants % Boundary inner products H_boundary_l, H_boundary_r % Metric coefficients b % Cell matrix of size dim x dim J, Ji beta % Cell array of scale factors % Numerical boundary flux operators flux_l, flux_r end methods function obj = Heat2dCurvilinear(g ,order, kappa_fun, opSet) default_arg('opSet',{@sbp.D2Variable, @sbp.D2Variable}); default_arg('kappa_fun', @(x,y) 0*x+1); dim = 2; kappa = grid.evalOn(g, kappa_fun); m = g.size(); m_tot = g.N(); % 1D operators ops = cell(dim,1); for i = 1:dim ops{i} = opSet{i}(m(i), {0, 1}, order); end I = cell(dim,1); D1 = cell(dim,1); D2 = cell(dim,1); H = cell(dim,1); Hi = cell(dim,1); e_l = cell(dim,1); e_r = cell(dim,1); d1_l = cell(dim,1); d1_r = cell(dim,1); for i = 1:dim I{i} = speye(m(i)); D1{i} = ops{i}.D1; D2{i} = ops{i}.D2; H{i} = ops{i}.H; Hi{i} = ops{i}.HI; e_l{i} = ops{i}.e_l; e_r{i} = ops{i}.e_r; d1_l{i} = ops{i}.d1_l; d1_r{i} = ops{i}.d1_r; end %====== Assemble full operators ======== KAPPA = spdiag(kappa); obj.KAPPA = KAPPA; % Allocate obj.D1 = cell(dim,1); obj.D2_kappa = cell(dim,1); obj.e_l = cell(dim,1); obj.e_r = cell(dim,1); obj.d1_l = cell(dim,1); obj.d1_r = cell(dim,1); % D1 obj.D1{1} = kron(D1{1},I{2}); obj.D1{2} = kron(I{1},D1{2}); % -- Metric coefficients ---- coords = g.points(); x = coords(:,1); y = coords(:,2); % Use non-periodic difference operators for metric even if opSet is periodic. xmax = max(ops{1}.x); ymax = max(ops{2}.x); opSetMetric{1} = sbp.D2Variable(m(1), {0, xmax}, order); opSetMetric{2} = sbp.D2Variable(m(2), {0, ymax}, order); D1Metric{1} = kron(opSetMetric{1}.D1, I{2}); D1Metric{2} = kron(I{1}, opSetMetric{2}.D1); x_xi = D1Metric{1}*x; x_eta = D1Metric{2}*x; y_xi = D1Metric{1}*y; y_eta = D1Metric{2}*y; J = x_xi.*y_eta - x_eta.*y_xi; b = cell(dim,dim); b{1,1} = y_eta./J; b{1,2} = -x_eta./J; b{2,1} = -y_xi./J; b{2,2} = x_xi./J; % Scale factors for boundary integrals beta = cell(dim,1); beta{1} = sqrt(x_eta.^2 + y_eta.^2); beta{2} = sqrt(x_xi.^2 + y_xi.^2); J = spdiag(J); Ji = inv(J); for i = 1:dim beta{i} = spdiag(beta{i}); for j = 1:dim b{i,j} = spdiag(b{i,j}); end end obj.J = J; obj.Ji = Ji; obj.b = b; obj.beta = beta; %---------------------------- % Boundary operators obj.e_l{1} = kron(e_l{1},I{2}); obj.e_l{2} = kron(I{1},e_l{2}); obj.e_r{1} = kron(e_r{1},I{2}); obj.e_r{2} = kron(I{1},e_r{2}); obj.d1_l{1} = kron(d1_l{1},I{2}); obj.d1_l{2} = kron(I{1},d1_l{2}); obj.d1_r{1} = kron(d1_r{1},I{2}); obj.d1_r{2} = kron(I{1},d1_r{2}); % D2 coefficients kappa_coeff = cell(dim,dim); for j = 1:dim obj.D2_kappa{j} = sparse(m_tot,m_tot); kappa_coeff{j} = sparse(m_tot,1); for i = 1:dim kappa_coeff{j} = kappa_coeff{j} + b{i,j}*J*b{i,j}*kappa; end end ind = grid.funcToMatrix(g, 1:m_tot); % x-dir j = 1; for col = 1:m(2) D_kappa = D2{1}(kappa_coeff{j}(ind(:,col))); p = ind(:,col); obj.D2_kappa{j}(p,p) = D_kappa; end % y-dir j = 2; for row = 1:m(1) D_kappa = D2{2}(kappa_coeff{j}(ind(row,:))); p = ind(row,:); obj.D2_kappa{j}(p,p) = D_kappa; end % Quadratures obj.H = kron(H{1},H{2}); obj.Hi = inv(obj.H); obj.H_boundary_l = cell(dim,1); obj.H_boundary_l{1} = obj.e_l{1}'*beta{1}*obj.e_l{1}*H{2}; obj.H_boundary_l{2} = obj.e_l{2}'*beta{2}*obj.e_l{2}*H{1}; obj.H_boundary_r = cell(dim,1); obj.H_boundary_r{1} = obj.e_r{1}'*beta{1}*obj.e_r{1}*H{2}; obj.H_boundary_r{2} = obj.e_r{2}'*beta{2}*obj.e_r{2}*H{1}; %=== Differentiation matrix D (without SAT) === D2_kappa = obj.D2_kappa; D1 = obj.D1; D = sparse(m_tot,m_tot); d = @kroneckerDelta; % Kronecker delta db = @(i,j) 1-d(i,j); % Logical not of Kronecker delta % 2nd derivatives for j = 1:dim D = D + Ji*D2_kappa{j}; end % Mixed terms for i = 1:dim for j = 1:dim for k = 1:dim D = D + db(i,j)*Ji*D1{j}*b{i,j}*J*KAPPA*b{i,k}*D1{k}; end end end obj.D = D; %=========================================% % Normal flux operators for BC. flux_l = cell(dim,1); flux_r = cell(dim,1); d1_l = obj.d1_l; d1_r = obj.d1_r; e_l = obj.e_l; e_r = obj.e_r; % Loop over boundaries for j = 1:dim flux_l{j} = sparse(m_tot,m_tot); flux_r{j} = sparse(m_tot,m_tot); % Loop over dummy index for i = 1:dim % Loop over dummy index for k = 1:dim flux_l{j} = flux_l{j} ... - beta{j}\b{i,j}*J*KAPPA*b{i,k}*( d(j,k)*e_l{k}*d1_l{k}' + db(j,k)*D1{k} ); flux_r{j} = flux_r{j} ... + beta{j}\b{i,j}*J*KAPPA*b{i,k}*( d(j,k)*e_r{k}*d1_r{k}' + db(j,k)*D1{k} ); end end end obj.flux_l = flux_l; obj.flux_r = flux_r; % Misc. obj.m = m; obj.h = g.scaling(); obj.order = order; obj.grid = g; obj.dim = dim; obj.alpha = [ops{1}.borrowing.M.d1, ops{2}.borrowing.M.d1]; end % Closure functions return the operators applied to the own domain to close the boundary % Penalty functions return the operators to force the solution. In the case of an interface it returns the operator applied to the other doamin. % boundary is a string specifying the boundary e.g. 'l','r' or 'e','w','n','s'. % type is a string specifying the type of boundary condition. % data is a function returning the data that should be applied at the boundary. % neighbour_scheme is an instance of Scheme that should be interfaced to. % neighbour_boundary is a string specifying which boundary to interface to. function [closure, penalty] = boundary_condition(obj, boundary, type, symmetric, tuning) default_arg('type','Neumann'); default_arg('symmetric', false); default_arg('tuning',1.2); % nj: outward unit normal component. % nj = -1 for west, south, bottom boundaries % nj = 1 for east, north, top boundaries nj = obj.getBoundarySign(boundary); Hi = obj.Hi; [e, flux] = obj.getBoundaryOperator({'e', 'flux'}, boundary); H_gamma = obj.getBoundaryQuadrature(boundary); alpha = obj.getBoundaryBorrowing(boundary); Hi = obj.Hi; Ji = obj.Ji; KAPPA = obj.KAPPA; kappa_gamma = e'*KAPPA*e; switch type % Dirichlet boundary condition case {'D','d','dirichlet','Dirichlet'} if ~symmetric closure = -Ji*Hi*flux'*e*H_gamma*(e' ); penalty = Ji*Hi*flux'*e*H_gamma; else closure = Ji*Hi*flux'*e*H_gamma*(e' )... -tuning*2/alpha*Ji*Hi*e*kappa_gamma*H_gamma*(e' ) ; penalty = -Ji*Hi*flux'*e*H_gamma ... +tuning*2/alpha*Ji*Hi*e*kappa_gamma*H_gamma; end % Normal flux boundary condition case {'N','n','neumann','Neumann'} closure = -Ji*Hi*e*H_gamma*(e'*flux ); penalty = Ji*Hi*e*H_gamma; % Unknown boundary condition otherwise error('No such boundary condition: type = %s',type); end end function [closure, penalty] = interface(obj,boundary,neighbour_scheme,neighbour_boundary) % u denotes the solution in the own domain % v denotes the solution in the neighbour domain error('Interface not implemented'); end % Returns the boundary operator op for the boundary specified by the string boundary. % op -- string or a cell array of strings % boundary -- string function varargout = getBoundaryOperator(obj, op, boundary) assertIsMember(boundary, {'w', 'e', 's', 'n'}) if ~iscell(op) op = {op}; end for i = 1:numel(op) switch op{i} case 'e' switch boundary case 'w' e = obj.e_l{1}; case 'e' e = obj.e_r{1}; case 's' e = obj.e_l{2}; case 'n' e = obj.e_r{2}; end varargout{i} = e; case 'd' switch boundary case 'w' d = obj.d1_l{1}; case 'e' d = obj.d1_r{1}; case 's' d = obj.d1_l{2}; case 'n' d = obj.d1_r{2}; end varargout{i} = d; case 'flux' switch boundary case 'w' flux = obj.flux_l{1}; case 'e' flux = obj.flux_r{1}; case 's' flux = obj.flux_l{2}; case 'n' flux = obj.flux_r{2}; end varargout{i} = flux; end end end % Returns square boundary quadrature matrix, of dimension % corresponding to the number of boundary points % % boundary -- string function H_b = getBoundaryQuadrature(obj, boundary) assertIsMember(boundary, {'w', 'e', 's', 'n'}) switch boundary case 'w' H_b = obj.H_boundary_l{1}; case 'e' H_b = obj.H_boundary_r{1}; case 's' H_b = obj.H_boundary_l{2}; case 'n' H_b = obj.H_boundary_r{2}; end end % Returns the boundary sign. The right boundary is considered the positive boundary % boundary -- string function s = getBoundarySign(obj, boundary) assertIsMember(boundary, {'w', 'e', 's', 'n'}) switch boundary case {'e','n'} s = 1; case {'w','s'} s = -1; end end % Returns borrowing constant gamma*h % boundary -- string function gamm = getBoundaryBorrowing(obj, boundary) assertIsMember(boundary, {'w', 'e', 's', 'n'}) switch boundary case {'w','e'} gamm = obj.h(1)*obj.alpha(1); case {'s','n'} gamm = obj.h(2)*obj.alpha(2); end end function N = size(obj) N = prod(obj.m); end end end