{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Simple FEM-BEM coupling for the Helmholtz equation with FEniCSx"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Background"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"For this problem, you will need FEniCSx installed alongside Bempp. If FEniCSx is not available on your system you can use the Docker image from bempp.com\n",
"\n",
"In this tutorial, we will solve the problem of a wave travelling through a unit cube, $\\Omega = [0,1]^3$ with different material parameters inside and outside the domain. The incident wave is given by\n",
"\n",
"$$\n",
"u^\\text{inc}(\\mathbf{x})=\\mathrm{e}^{\\mathrm{i} k \\mathbf{x}\\cdot\\mathbf{d}},\n",
"$$\n",
"\n",
"where $\\mathbf{x}=(x,y,z)$ and $\\mathbf{d}$ is the direction of the incident wave. In the implementation we use, $\\mathbf{d} = \\frac{1}{\\sqrt{3}}(1,1,1)$.\n",
"\n",
"The PDE is\n",
"\n",
"$$\n",
"\\Delta u + n(\\mathbf{x})^2 k^2 u = 0, \\quad \\text{ in } \\Omega\\\\\n",
"\\Delta u + k^2 u = 0, \\quad \\text{ in } \\mathbb{R}^3 \\backslash \\Omega\n",
"$$\n",
"\n",
"In this example, we use \n",
"\n",
"$$\n",
"n(\\mathbf{x}) = 0.5\n",
"$$\n",
"Since the interior wavenumber is constant one could have also used a BEM/BEM coupling approach. However, here we demonstrate the use of FEM for the interior problem using the FEniCSx finite element package."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### FEM Part"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In $\\Omega$, the FEM part is formulated as\n",
"\n",
"$$\n",
"\\int_\\Omega \\nabla u\\cdot\\nabla v -k^2\\int_\\Omega n^2uv - \\int_{d\\Omega} v\\frac{\\partial u}{\\partial \\nu} = 0,\n",
"$$\n",
"\n",
"or\n",
"\n",
"$$\n",
"\\langle\\nabla u,\\nabla v\\rangle_\\Omega - k^2\\langle n^2u,v\\rangle_\\Omega - \\langle \\lambda,v\\rangle_\\Gamma=0,\n",
"$$\n",
"\n",
"where $\\lambda=\\frac{\\partial u}{\\partial \\nu}$.\n",
"\n",
"Later, we will write this as the following operator equation\n",
"\n",
"$$\n",
"\\mathsf{A}u-k^2 \\mathsf{M}u-\\mathsf{M}_\\Gamma \\lambda = 0\n",
"$$"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### BEM Part"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In $\\mathbb{R}^3 \\backslash \\Omega$, we let $u = u^\\text{inc}+u^\\text{s}$, where $u^\\text{inc}$ is the incident wave and $u^\\text{s}$ is the scattered wave. As given in Integral equation methods in scattering theory by Colton & Kress,\n",
"\n",
"$$\n",
"0 = \\mathcal{K}u^\\text{inc}-\\mathcal{V}\\frac{\\partial u^{inc}}{\\partial \\nu},\\\\[2mm]\n",
"u^\\text{s} = \\mathcal{K}u^\\text{s}-\\mathcal{V}\\frac{\\partial u^{s}}{\\partial \\nu},\n",
"$$\n",
"where $\\mathcal{K}$ and $\\mathcal{V}$ are the double single layer potential operators. Adding these, we get\n",
"\n",
"$$\n",
"u^\\text{s} = \\mathcal{K}u-\\mathcal{V}\\lambda.\n",
"$$\n",
"\n",
"This representation formula will be used to find $u^\\text{s}$ for plotting later.\n",
"\n",
"Taking the trace on the boundary gives\n",
"\n",
"$$\n",
"u-u^\\text{inc} = \\left(\\tfrac{1}{2}\\mathsf{Id}+\\mathsf{K}\\right)u -\\mathsf{V}\\lambda.\n",
"$$\n",
"\n",
"This rearranges to\n",
"\n",
"$$\n",
"u^\\text{inc} = \\left(\\tfrac{1}{2}\\mathsf{Id}-\\mathsf{K}\\right)u+\\mathsf{V}\\lambda.\n",
"$$"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Full Formulation"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The full blocked formulation is\n",
"\n",
"$$\n",
"\\begin{bmatrix}\n",
" \\mathsf{A}-k^2 \\mathsf{M} & -\\mathsf{M}_\\Gamma\\\\\n",
" \\tfrac{1}{2}\\mathsf{Id}-\\mathsf{K} & \\mathsf{V}\n",
"\\end{bmatrix}\n",
"\\begin{bmatrix}\n",
" u\\\\\n",
" \\lambda\n",
"\\end{bmatrix}=\\begin{bmatrix}\n",
" 0\\\\\n",
" u^\\text{inc}\n",
"\\end{bmatrix}.\n",
"$$\n",
"\n",
"This formulation is not stable for all frequencies due to the possibility of interior resonances. But it is sufficient for this example and serves as a blueprint for more complex formulations."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Implementation"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We begin by importing DOLFINx (the FEniCSx python library), UFL (FEniCS's unified form language), MPI, Bempp and NumPy."
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"import dolfinx\n",
"from dolfinx.fem import functionspace, function\n",
"from dolfinx.mesh import create_unit_cube\n",
"import dolfinx.geometry\n",
"import ufl\n",
"from mpi4py import MPI\n",
"import bempp.api\n",
"import numpy as np"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Next, we set the wavenumber ``k`` and the direction ``d`` of the incoming wave."
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"k = 6.\n",
"d = np.array([1., 1., 1])\n",
"d /= np.linalg.norm(d)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We create a DOLFINx mesh. Later, the boundary mesh will be extracted from this."
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"mesh = create_unit_cube(MPI.COMM_WORLD, 10, 10, 10)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Next, we make the DOLFINx and Bempp function spaces.\n",
"\n",
"The function ``fenics_to_bempp_trace_data`` will extract the trace space from the DOLFINx space and create the matrix ``trace_matrix``, which maps between the dofs (degrees of freedom) in DOLFINx and Bempp."
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"INFO:bempp:Created grid with id c8ce5f52-7d2c-4a7b-8127-ae6c4c774573. Elements: 1200. Edges: 1800. Vertices: 602\n",
"TIMING:bempp:Grid.__init__ : 3.452e+00s\n",
"TIMING:bempp:_compute_p1_dof_map : 1.514e+00s\n",
"TIMING:bempp:p1_continuous_function_space : 1.521e+00s\n",
"TIMING:bempp:p0_discontinuous_function_space : 8.383e-04s\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"FEM dofs: 1331\n",
"BEM dofs: 1200\n"
]
}
],
"source": [
"from bempp.api.external import fenicsx\n",
"\n",
"fenics_space = functionspace(mesh, (\"CG\", 1))\n",
"trace_space, trace_matrix = \\\n",
" fenicsx.fenics_to_bempp_trace_data(fenics_space)\n",
"bempp_space = bempp.api.function_space(trace_space.grid, \"DP\", 0)\n",
"\n",
"fem_size = fenics_space.dofmap.index_map.size_global\n",
"bem_size = bempp_space.global_dof_count\n",
"\n",
"print(\"FEM dofs: {0}\".format(fem_size))\n",
"print(\"BEM dofs: {0}\".format(bem_size))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We create the boundary operators that we need."
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
"id_op = bempp.api.operators.boundary.sparse.identity(\n",
" trace_space, bempp_space, bempp_space)\n",
"mass = bempp.api.operators.boundary.sparse.identity(\n",
" bempp_space, bempp_space, trace_space)\n",
"dlp = bempp.api.operators.boundary.helmholtz.double_layer(\n",
" trace_space, bempp_space, bempp_space, k)\n",
"slp = bempp.api.operators.boundary.helmholtz.single_layer(\n",
" bempp_space, bempp_space, bempp_space, k)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We create the DOLFINx function spaces and the function (or in this case constant) ``n``."
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
"u = ufl.TrialFunction(fenics_space)\n",
"v = ufl.TestFunction(fenics_space)\n",
"n = 0.5"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We make the vectors on the right hand side of the formulation."
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [],
"source": [
"@bempp.api.complex_callable\n",
"def u_inc(x, n, domain_index, result):\n",
" result[0] = np.exp(1j * k * np.dot(x, d))\n",
"u_inc = bempp.api.GridFunction(bempp_space, fun=u_inc)\n",
"\n",
"# The rhs from the FEM\n",
"rhs_fem = np.zeros(fem_size)\n",
"# The rhs from the BEM\n",
"rhs_bem = u_inc.projections(bempp_space)\n",
"# The combined rhs\n",
"rhs = np.concatenate([rhs_fem, rhs_bem])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We are now ready to create a ``BlockedLinearOperator`` containing all four parts of the discretisation of\n",
"$$\n",
"\\begin{bmatrix}\n",
" \\mathsf{A}-k^2 \\mathsf{M} & -\\mathsf{M}_\\Gamma\\\\\n",
" \\tfrac{1}{2}\\mathsf{Id}-\\mathsf{K} & \\mathsf{V}\n",
"\\end{bmatrix}.\n",
"$$"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"TIMING:bempp:Start operation: \n",
"TIMING:bempp:Finished Operation: : 3.2153611183166504s\n",
"TIMING:bempp:assemble_sparse : 3.217e+00s\n",
"TIMING:bempp:Start operation: \n",
"TIMING:bempp:Finished Operation: : 2.544266700744629s\n",
"TIMING:bempp:assemble_sparse : 2.545e+00s\n",
"TIMING:bempp:Start operation: Regular assembler:helmholtz_double_layer_boundary:opencl\n",
"INFO:bempp:OpenCL CPU Device set to: pthread-Intel(R) Core(TM) i9-9900K CPU @ 3.60GHz\n",
"TIMING:bempp:Finished Operation: Regular assembler:helmholtz_double_layer_boundary:opencl: 1.7337913513183594s\n",
"TIMING:bempp:Start operation: Singular assembler:helmholtz_double_layer_boundary:opencl\n",
"TIMING:bempp:Finished Operation: Singular assembler:helmholtz_double_layer_boundary:opencl: 0.5207102298736572s\n",
"TIMING:bempp:Start operation: Regular assembler:helmholtz_single_layer_boundary:opencl\n",
"TIMING:bempp:Finished Operation: Regular assembler:helmholtz_single_layer_boundary:opencl: 1.0467305183410645s\n",
"TIMING:bempp:Start operation: Singular assembler:helmholtz_single_layer_boundary:opencl\n",
"TIMING:bempp:Finished Operation: Singular assembler:helmholtz_single_layer_boundary:opencl: 0.49677538871765137s\n"
]
}
],
"source": [
"from bempp.api.assembly.blocked_operator import BlockedDiscreteOperator\n",
"from scipy.sparse.linalg.interface import LinearOperator\n",
"blocks = [[None,None],[None,None]]\n",
"\n",
"trace_op = LinearOperator(trace_matrix.shape, lambda x:trace_matrix @ x)\n",
"\n",
"A = fenicsx.FenicsOperator((ufl.inner(ufl.grad(u), ufl.grad(v)) - k**2 * n**2 * ufl.inner(u, v)) * ufl.dx)\n",
"\n",
"blocks[0][0] = A.weak_form()\n",
"blocks[0][1] = -trace_matrix.T * mass.weak_form().to_sparse()\n",
"blocks[1][0] = (.5 * id_op - dlp).weak_form() * trace_op\n",
"blocks[1][1] = slp.weak_form()\n",
"\n",
"blocked = BlockedDiscreteOperator(np.array(blocks))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Next, we solve the system, then split the solution into the parts assosiated with u and λ. For an efficient solve, preconditioning is required."
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"TIMING:bempp:_Solver.__init__ : 2.485e-02s\n",
"TIMING:bempp:Start operation: \n",
"TIMING:bempp:Finished Operation: : 1.1819589138031006s\n",
"TIMING:bempp:assemble_sparse : 1.184e+00s\n",
"WARNING:py.warnings:/usr/lib/python3/dist-packages/scipy/sparse/linalg/dsolve/linsolve.py:296: SparseEfficiencyWarning: splu requires CSC matrix format\n",
" warn('splu requires CSC matrix format', SparseEfficiencyWarning)\n",
"\n",
"TIMING:bempp:_Solver.__init__ : 3.294e-03s\n",
"TIMING:bempp:_Solver.solve : 2.603e-03s\n",
"TIMING:bempp:_Solver.solve : 3.290e-05s\n",
"TIMING:bempp:_Solver.solve : 3.338e-05s\n",
"TIMING:bempp:_Solver.solve : 8.276e-04s\n",
"TIMING:bempp:_Solver.solve : 8.810e-04s\n",
"TIMING:bempp:_Solver.solve : 2.587e-04s\n",
"TIMING:bempp:_Solver.solve : 1.340e-04s\n",
"TIMING:bempp:_Solver.solve : 1.208e-03s\n",
"TIMING:bempp:_Solver.solve : 8.800e-04s\n",
"TIMING:bempp:_Solver.solve : 2.425e-04s\n",
"TIMING:bempp:_Solver.solve : 2.391e-04s\n",
"TIMING:bempp:_Solver.solve : 1.192e-03s\n",
"TIMING:bempp:_Solver.solve : 8.972e-04s\n",
"TIMING:bempp:_Solver.solve : 2.308e-04s\n",
"TIMING:bempp:_Solver.solve : 2.244e-04s\n",
"TIMING:bempp:_Solver.solve : 1.179e-03s\n",
"TIMING:bempp:_Solver.solve : 8.926e-04s\n",
"TIMING:bempp:_Solver.solve : 2.258e-04s\n",
"TIMING:bempp:_Solver.solve : 2.530e-04s\n",
"TIMING:bempp:_Solver.solve : 1.243e-03s\n",
"TIMING:bempp:_Solver.solve : 9.072e-04s\n",
"TIMING:bempp:_Solver.solve : 2.186e-04s\n",
"TIMING:bempp:_Solver.solve : 1.905e-04s\n",
"TIMING:bempp:_Solver.solve : 1.084e-03s\n",
"TIMING:bempp:_Solver.solve : 9.575e-04s\n",
"TIMING:bempp:_Solver.solve : 3.076e-05s\n",
"TIMING:bempp:_Solver.solve : 2.680e-04s\n",
"TIMING:bempp:_Solver.solve : 9.916e-04s\n",
"TIMING:bempp:_Solver.solve : 1.271e-03s\n",
"TIMING:bempp:_Solver.solve : 1.290e-04s\n",
"TIMING:bempp:_Solver.solve : 1.674e-04s\n",
"TIMING:bempp:_Solver.solve : 1.214e-03s\n",
"TIMING:bempp:_Solver.solve : 9.804e-04s\n",
"TIMING:bempp:_Solver.solve : 3.219e-05s\n",
"TIMING:bempp:_Solver.solve : 3.099e-05s\n",
"TIMING:bempp:_Solver.solve : 9.716e-04s\n",
"TIMING:bempp:_Solver.solve : 7.665e-04s\n",
"TIMING:bempp:_Solver.solve : 1.566e-04s\n",
"TIMING:bempp:_Solver.solve : 1.156e-04s\n",
"TIMING:bempp:_Solver.solve : 1.028e-03s\n",
"TIMING:bempp:_Solver.solve : 1.048e-03s\n",
"TIMING:bempp:_Solver.solve : 1.366e-04s\n",
"TIMING:bempp:_Solver.solve : 1.922e-04s\n",
"TIMING:bempp:_Solver.solve : 1.075e-03s\n",
"TIMING:bempp:_Solver.solve : 1.153e-03s\n",
"TIMING:bempp:_Solver.solve : 1.676e-04s\n",
"TIMING:bempp:_Solver.solve : 1.938e-04s\n",
"TIMING:bempp:_Solver.solve : 1.096e-03s\n",
"TIMING:bempp:_Solver.solve : 9.286e-04s\n",
"TIMING:bempp:_Solver.solve : 9.680e-05s\n",
"TIMING:bempp:_Solver.solve : 1.898e-04s\n",
"TIMING:bempp:_Solver.solve : 1.013e-03s\n",
"TIMING:bempp:_Solver.solve : 9.339e-04s\n",
"TIMING:bempp:_Solver.solve : 1.729e-04s\n",
"TIMING:bempp:_Solver.solve : 1.745e-04s\n",
"TIMING:bempp:_Solver.solve : 1.074e-03s\n",
"TIMING:bempp:_Solver.solve : 8.385e-04s\n",
"TIMING:bempp:_Solver.solve : 1.476e-04s\n",
"TIMING:bempp:_Solver.solve : 1.962e-04s\n",
"TIMING:bempp:_Solver.solve : 1.004e-03s\n",
"TIMING:bempp:_Solver.solve : 1.191e-03s\n",
"TIMING:bempp:_Solver.solve : 1.345e-04s\n",
"TIMING:bempp:_Solver.solve : 1.667e-04s\n",
"TIMING:bempp:_Solver.solve : 1.216e-03s\n",
"TIMING:bempp:_Solver.solve : 8.729e-04s\n",
"TIMING:bempp:_Solver.solve : 1.292e-04s\n",
"TIMING:bempp:_Solver.solve : 1.559e-04s\n",
"TIMING:bempp:_Solver.solve : 9.711e-04s\n",
"TIMING:bempp:_Solver.solve : 8.779e-04s\n",
"TIMING:bempp:_Solver.solve : 1.321e-04s\n",
"TIMING:bempp:_Solver.solve : 1.857e-04s\n",
"TIMING:bempp:_Solver.solve : 1.228e-03s\n",
"TIMING:bempp:_Solver.solve : 8.600e-04s\n",
"TIMING:bempp:_Solver.solve : 1.309e-04s\n",
"TIMING:bempp:_Solver.solve : 1.640e-04s\n",
"TIMING:bempp:_Solver.solve : 9.894e-04s\n",
"TIMING:bempp:_Solver.solve : 9.060e-04s\n",
"TIMING:bempp:_Solver.solve : 1.671e-04s\n",
"TIMING:bempp:_Solver.solve : 2.706e-04s\n",
"TIMING:bempp:_Solver.solve : 1.153e-03s\n",
"TIMING:bempp:_Solver.solve : 7.885e-04s\n",
"TIMING:bempp:_Solver.solve : 1.364e-04s\n",
"TIMING:bempp:_Solver.solve : 2.036e-04s\n",
"TIMING:bempp:_Solver.solve : 9.329e-04s\n",
"TIMING:bempp:_Solver.solve : 9.768e-04s\n",
"TIMING:bempp:_Solver.solve : 1.335e-04s\n",
"TIMING:bempp:_Solver.solve : 1.578e-04s\n",
"TIMING:bempp:_Solver.solve : 9.742e-04s\n",
"TIMING:bempp:_Solver.solve : 8.216e-04s\n",
"TIMING:bempp:_Solver.solve : 1.314e-04s\n",
"TIMING:bempp:_Solver.solve : 1.636e-04s\n",
"TIMING:bempp:_Solver.solve : 9.387e-04s\n",
"TIMING:bempp:_Solver.solve : 8.545e-04s\n",
"TIMING:bempp:_Solver.solve : 3.242e-05s\n",
"TIMING:bempp:_Solver.solve : 1.965e-04s\n",
"TIMING:bempp:_Solver.solve : 9.096e-04s\n",
"TIMING:bempp:_Solver.solve : 8.411e-04s\n",
"TIMING:bempp:_Solver.solve : 1.564e-04s\n",
"TIMING:bempp:_Solver.solve : 1.407e-04s\n",
"TIMING:bempp:_Solver.solve : 9.649e-04s\n",
"TIMING:bempp:_Solver.solve : 8.235e-04s\n",
"TIMING:bempp:_Solver.solve : 1.719e-04s\n",
"TIMING:bempp:_Solver.solve : 1.581e-04s\n",
"TIMING:bempp:_Solver.solve : 9.449e-04s\n",
"TIMING:bempp:_Solver.solve : 9.179e-04s\n",
"TIMING:bempp:_Solver.solve : 1.583e-04s\n",
"TIMING:bempp:_Solver.solve : 1.473e-04s\n",
"TIMING:bempp:_Solver.solve : 8.063e-04s\n",
"TIMING:bempp:_Solver.solve : 9.191e-04s\n",
"TIMING:bempp:_Solver.solve : 2.220e-04s\n",
"TIMING:bempp:_Solver.solve : 1.483e-04s\n",
"TIMING:bempp:_Solver.solve : 9.310e-04s\n",
"TIMING:bempp:_Solver.solve : 1.012e-03s\n",
"TIMING:bempp:_Solver.solve : 1.302e-04s\n",
"TIMING:bempp:_Solver.solve : 1.152e-04s\n",
"TIMING:bempp:_Solver.solve : 8.609e-04s\n",
"TIMING:bempp:_Solver.solve : 8.743e-04s\n",
"TIMING:bempp:_Solver.solve : 3.338e-05s\n",
"TIMING:bempp:_Solver.solve : 2.022e-04s\n",
"TIMING:bempp:_Solver.solve : 7.946e-04s\n",
"TIMING:bempp:_Solver.solve : 1.317e-03s\n",
"TIMING:bempp:_Solver.solve : 1.807e-04s\n",
"TIMING:bempp:_Solver.solve : 6.413e-05s\n",
"TIMING:bempp:_Solver.solve : 1.010e-03s\n",
"TIMING:bempp:_Solver.solve : 1.074e-03s\n",
"TIMING:bempp:_Solver.solve : 1.519e-04s\n",
"TIMING:bempp:_Solver.solve : 1.464e-04s\n",
"TIMING:bempp:_Solver.solve : 8.152e-04s\n",
"TIMING:bempp:_Solver.solve : 8.495e-04s\n",
"TIMING:bempp:_Solver.solve : 1.750e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 1.019e-03s\n",
"TIMING:bempp:_Solver.solve : 8.304e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 7.467e-04s\n",
"TIMING:bempp:_Solver.solve : 1.166e-03s\n",
"TIMING:bempp:_Solver.solve : 3.982e-05s\n",
"TIMING:bempp:_Solver.solve : 3.505e-05s\n",
"TIMING:bempp:_Solver.solve : 7.129e-04s\n",
"TIMING:bempp:_Solver.solve : 8.085e-04s\n",
"TIMING:bempp:_Solver.solve : 3.266e-05s\n",
"TIMING:bempp:_Solver.solve : 3.123e-05s\n",
"TIMING:bempp:_Solver.solve : 7.393e-04s\n",
"TIMING:bempp:_Solver.solve : 7.834e-04s\n",
"TIMING:bempp:_Solver.solve : 2.298e-04s\n",
"TIMING:bempp:_Solver.solve : 1.037e-04s\n",
"TIMING:bempp:_Solver.solve : 9.568e-04s\n",
"TIMING:bempp:_Solver.solve : 9.155e-04s\n",
"TIMING:bempp:_Solver.solve : 3.123e-05s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 6.881e-04s\n",
"TIMING:bempp:_Solver.solve : 8.810e-04s\n",
"TIMING:bempp:_Solver.solve : 3.028e-05s\n",
"TIMING:bempp:_Solver.solve : 2.861e-05s\n",
"TIMING:bempp:_Solver.solve : 6.635e-04s\n",
"TIMING:bempp:_Solver.solve : 9.358e-04s\n",
"TIMING:bempp:_Solver.solve : 3.910e-05s\n",
"TIMING:bempp:_Solver.solve : 3.099e-05s\n",
"TIMING:bempp:_Solver.solve : 7.758e-04s\n",
"TIMING:bempp:_Solver.solve : 8.059e-04s\n",
"TIMING:bempp:_Solver.solve : 1.156e-04s\n",
"TIMING:bempp:_Solver.solve : 1.330e-04s\n",
"TIMING:bempp:_Solver.solve : 7.875e-04s\n",
"TIMING:bempp:_Solver.solve : 7.737e-04s\n",
"TIMING:bempp:_Solver.solve : 3.052e-05s\n",
"TIMING:bempp:_Solver.solve : 2.646e-05s\n",
"TIMING:bempp:_Solver.solve : 6.828e-04s\n",
"TIMING:bempp:_Solver.solve : 1.284e-03s\n",
"TIMING:bempp:_Solver.solve : 3.982e-05s\n",
"TIMING:bempp:_Solver.solve : 3.886e-05s\n",
"TIMING:bempp:_Solver.solve : 8.783e-04s\n",
"TIMING:bempp:_Solver.solve : 1.065e-03s\n",
"TIMING:bempp:_Solver.solve : 3.099e-05s\n",
"TIMING:bempp:_Solver.solve : 2.813e-05s\n",
"TIMING:bempp:_Solver.solve : 7.150e-04s\n",
"TIMING:bempp:_Solver.solve : 9.940e-04s\n",
"TIMING:bempp:_Solver.solve : 3.147e-05s\n",
"TIMING:bempp:_Solver.solve : 2.766e-05s\n",
"TIMING:bempp:_Solver.solve : 7.133e-04s\n",
"TIMING:bempp:_Solver.solve : 8.228e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 6.821e-04s\n",
"TIMING:bempp:_Solver.solve : 7.975e-04s\n",
"TIMING:bempp:_Solver.solve : 1.431e-04s\n",
"TIMING:bempp:_Solver.solve : 1.299e-04s\n",
"TIMING:bempp:_Solver.solve : 7.699e-04s\n",
"TIMING:bempp:_Solver.solve : 8.125e-04s\n",
"TIMING:bempp:_Solver.solve : 1.333e-04s\n",
"TIMING:bempp:_Solver.solve : 1.321e-04s\n",
"TIMING:bempp:_Solver.solve : 7.551e-04s\n",
"TIMING:bempp:_Solver.solve : 9.153e-04s\n",
"TIMING:bempp:_Solver.solve : 3.123e-05s\n",
"TIMING:bempp:_Solver.solve : 3.052e-05s\n",
"TIMING:bempp:_Solver.solve : 7.625e-04s\n",
"TIMING:bempp:_Solver.solve : 7.913e-04s\n",
"TIMING:bempp:_Solver.solve : 3.099e-05s\n",
"TIMING:bempp:_Solver.solve : 3.076e-05s\n",
"TIMING:bempp:_Solver.solve : 7.832e-04s\n",
"TIMING:bempp:_Solver.solve : 9.110e-04s\n",
"TIMING:bempp:_Solver.solve : 3.028e-05s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 6.731e-04s\n",
"TIMING:bempp:_Solver.solve : 7.787e-04s\n",
"TIMING:bempp:_Solver.solve : 1.178e-04s\n",
"TIMING:bempp:_Solver.solve : 1.371e-04s\n",
"TIMING:bempp:_Solver.solve : 7.672e-04s\n",
"TIMING:bempp:_Solver.solve : 8.984e-04s\n",
"TIMING:bempp:_Solver.solve : 1.154e-04s\n",
"TIMING:bempp:_Solver.solve : 1.299e-04s\n",
"TIMING:bempp:_Solver.solve : 7.405e-04s\n",
"TIMING:bempp:_Solver.solve : 1.014e-03s\n",
"TIMING:bempp:_Solver.solve : 3.099e-05s\n",
"TIMING:bempp:_Solver.solve : 1.905e-04s\n",
"TIMING:bempp:_Solver.solve : 7.696e-04s\n",
"TIMING:bempp:_Solver.solve : 9.317e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.646e-05s\n",
"TIMING:bempp:_Solver.solve : 6.814e-04s\n",
"TIMING:bempp:_Solver.solve : 1.054e-03s\n",
"TIMING:bempp:_Solver.solve : 3.958e-05s\n",
"TIMING:bempp:_Solver.solve : 1.614e-04s\n",
"TIMING:bempp:_Solver.solve : 8.748e-04s\n",
"TIMING:bempp:_Solver.solve : 8.898e-04s\n",
"TIMING:bempp:_Solver.solve : 3.171e-05s\n",
"TIMING:bempp:_Solver.solve : 2.742e-05s\n",
"TIMING:bempp:_Solver.solve : 7.162e-04s\n",
"TIMING:bempp:_Solver.solve : 8.552e-04s\n",
"TIMING:bempp:_Solver.solve : 3.123e-05s\n",
"TIMING:bempp:_Solver.solve : 7.629e-05s\n",
"TIMING:bempp:_Solver.solve : 7.591e-04s\n",
"TIMING:bempp:_Solver.solve : 1.103e-03s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 6.757e-04s\n",
"TIMING:bempp:_Solver.solve : 9.403e-04s\n",
"TIMING:bempp:_Solver.solve : 3.386e-05s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 7.639e-04s\n",
"TIMING:bempp:_Solver.solve : 8.948e-04s\n",
"TIMING:bempp:_Solver.solve : 2.313e-04s\n",
"TIMING:bempp:_Solver.solve : 1.936e-04s\n",
"TIMING:bempp:_Solver.solve : 1.116e-03s\n",
"TIMING:bempp:_Solver.solve : 7.756e-04s\n",
"TIMING:bempp:_Solver.solve : 3.028e-05s\n",
"TIMING:bempp:_Solver.solve : 2.718e-05s\n",
"TIMING:bempp:_Solver.solve : 7.844e-04s\n",
"TIMING:bempp:_Solver.solve : 1.004e-03s\n",
"TIMING:bempp:_Solver.solve : 8.106e-05s\n",
"TIMING:bempp:_Solver.solve : 2.789e-05s\n",
"TIMING:bempp:_Solver.solve : 7.961e-04s\n",
"TIMING:bempp:_Solver.solve : 8.743e-04s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 1.931e-04s\n",
"TIMING:bempp:_Solver.solve : 8.113e-04s\n",
"TIMING:bempp:_Solver.solve : 9.673e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 5.364e-05s\n",
"TIMING:bempp:_Solver.solve : 7.625e-04s\n",
"TIMING:bempp:_Solver.solve : 7.482e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 2.694e-05s\n",
"TIMING:bempp:_Solver.solve : 7.658e-04s\n",
"TIMING:bempp:_Solver.solve : 7.985e-04s\n",
"TIMING:bempp:_Solver.solve : 3.743e-05s\n",
"TIMING:bempp:_Solver.solve : 2.766e-05s\n",
"TIMING:bempp:_Solver.solve : 7.753e-04s\n",
"TIMING:bempp:_Solver.solve : 1.381e-03s\n",
"TIMING:bempp:_Solver.solve : 3.314e-05s\n",
"TIMING:bempp:_Solver.solve : 3.552e-05s\n",
"TIMING:bempp:_Solver.solve : 7.865e-04s\n",
"TIMING:bempp:_Solver.solve : 1.002e-03s\n",
"TIMING:bempp:_Solver.solve : 3.338e-05s\n",
"TIMING:bempp:_Solver.solve : 3.219e-05s\n",
"TIMING:bempp:_Solver.solve : 8.779e-04s\n",
"TIMING:bempp:_Solver.solve : 1.071e-03s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.599e-05s\n",
"TIMING:bempp:_Solver.solve : 6.752e-04s\n",
"TIMING:bempp:_Solver.solve : 1.159e-03s\n",
"TIMING:bempp:_Solver.solve : 3.386e-05s\n",
"TIMING:bempp:_Solver.solve : 3.028e-05s\n",
"TIMING:bempp:_Solver.solve : 7.625e-04s\n",
"TIMING:bempp:_Solver.solve : 1.010e-03s\n",
"TIMING:bempp:_Solver.solve : 6.747e-05s\n",
"TIMING:bempp:_Solver.solve : 5.937e-05s\n",
"TIMING:bempp:_Solver.solve : 9.420e-04s\n",
"TIMING:bempp:_Solver.solve : 8.259e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.599e-05s\n",
"TIMING:bempp:_Solver.solve : 6.890e-04s\n",
"TIMING:bempp:_Solver.solve : 8.676e-04s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 7.906e-04s\n",
"TIMING:bempp:_Solver.solve : 9.704e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.694e-05s\n",
"TIMING:bempp:_Solver.solve : 6.683e-04s\n",
"TIMING:bempp:_Solver.solve : 7.670e-04s\n",
"TIMING:bempp:_Solver.solve : 2.813e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 6.487e-04s\n",
"TIMING:bempp:_Solver.solve : 1.050e-03s\n",
"TIMING:bempp:_Solver.solve : 3.219e-05s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 7.055e-04s\n",
"TIMING:bempp:_Solver.solve : 9.456e-04s\n",
"TIMING:bempp:_Solver.solve : 3.910e-05s\n",
"TIMING:bempp:_Solver.solve : 3.290e-05s\n",
"TIMING:bempp:_Solver.solve : 7.801e-04s\n",
"TIMING:bempp:_Solver.solve : 1.014e-03s\n",
"TIMING:bempp:_Solver.solve : 2.813e-05s\n",
"TIMING:bempp:_Solver.solve : 2.599e-05s\n",
"TIMING:bempp:_Solver.solve : 7.193e-04s\n",
"TIMING:bempp:_Solver.solve : 8.996e-04s\n",
"TIMING:bempp:_Solver.solve : 1.199e-04s\n",
"TIMING:bempp:_Solver.solve : 1.156e-04s\n",
"TIMING:bempp:_Solver.solve : 7.031e-04s\n",
"TIMING:bempp:_Solver.solve : 7.637e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.742e-05s\n",
"TIMING:bempp:_Solver.solve : 7.741e-04s\n",
"TIMING:bempp:_Solver.solve : 7.689e-04s\n",
"TIMING:bempp:_Solver.solve : 2.766e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 6.800e-04s\n",
"TIMING:bempp:_Solver.solve : 1.248e-03s\n",
"TIMING:bempp:_Solver.solve : 1.805e-04s\n",
"TIMING:bempp:_Solver.solve : 3.028e-05s\n",
"TIMING:bempp:_Solver.solve : 7.985e-04s\n",
"TIMING:bempp:_Solver.solve : 8.991e-04s\n",
"TIMING:bempp:_Solver.solve : 3.290e-05s\n",
"TIMING:bempp:_Solver.solve : 2.069e-04s\n",
"TIMING:bempp:_Solver.solve : 8.328e-04s\n",
"TIMING:bempp:_Solver.solve : 8.013e-04s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 2.599e-05s\n",
"TIMING:bempp:_Solver.solve : 7.248e-04s\n",
"TIMING:bempp:_Solver.solve : 9.654e-04s\n",
"TIMING:bempp:_Solver.solve : 3.052e-05s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 6.781e-04s\n",
"TIMING:bempp:_Solver.solve : 8.404e-04s\n",
"TIMING:bempp:_Solver.solve : 5.388e-05s\n",
"TIMING:bempp:_Solver.solve : 1.266e-04s\n",
"TIMING:bempp:_Solver.solve : 7.939e-04s\n",
"TIMING:bempp:_Solver.solve : 8.543e-04s\n",
"TIMING:bempp:_Solver.solve : 2.718e-05s\n",
"TIMING:bempp:_Solver.solve : 2.646e-05s\n",
"TIMING:bempp:_Solver.solve : 6.206e-04s\n",
"TIMING:bempp:_Solver.solve : 8.314e-04s\n",
"TIMING:bempp:_Solver.solve : 1.214e-04s\n",
"TIMING:bempp:_Solver.solve : 1.163e-04s\n",
"TIMING:bempp:_Solver.solve : 7.126e-04s\n",
"TIMING:bempp:_Solver.solve : 7.923e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 1.779e-04s\n",
"TIMING:bempp:_Solver.solve : 8.543e-04s\n",
"TIMING:bempp:_Solver.solve : 7.830e-04s\n",
"TIMING:bempp:_Solver.solve : 1.476e-04s\n",
"TIMING:bempp:_Solver.solve : 1.469e-04s\n",
"TIMING:bempp:_Solver.solve : 8.256e-04s\n",
"TIMING:bempp:_Solver.solve : 7.710e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 8.707e-04s\n",
"TIMING:bempp:_Solver.solve : 9.985e-04s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 2.694e-05s\n",
"TIMING:bempp:_Solver.solve : 6.664e-04s\n",
"TIMING:bempp:_Solver.solve : 9.685e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.718e-05s\n",
"TIMING:bempp:_Solver.solve : 7.331e-04s\n",
"TIMING:bempp:_Solver.solve : 9.124e-04s\n",
"TIMING:bempp:_Solver.solve : 3.076e-05s\n",
"TIMING:bempp:_Solver.solve : 2.646e-05s\n",
"TIMING:bempp:_Solver.solve : 6.707e-04s\n",
"TIMING:bempp:_Solver.solve : 8.008e-04s\n",
"TIMING:bempp:_Solver.solve : 3.147e-05s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 6.807e-04s\n",
"TIMING:bempp:_Solver.solve : 7.713e-04s\n",
"TIMING:bempp:_Solver.solve : 2.861e-05s\n",
"TIMING:bempp:_Solver.solve : 2.766e-05s\n",
"TIMING:bempp:_Solver.solve : 8.893e-04s\n",
"TIMING:bempp:_Solver.solve : 8.109e-04s\n",
"TIMING:bempp:_Solver.solve : 5.388e-05s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 7.088e-04s\n",
"TIMING:bempp:_Solver.solve : 9.913e-04s\n",
"TIMING:bempp:_Solver.solve : 3.099e-05s\n",
"TIMING:bempp:_Solver.solve : 1.893e-04s\n",
"TIMING:bempp:_Solver.solve : 9.179e-04s\n",
"TIMING:bempp:_Solver.solve : 8.721e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 7.043e-04s\n",
"TIMING:bempp:_Solver.solve : 7.699e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 2.646e-05s\n",
"TIMING:bempp:_Solver.solve : 6.700e-04s\n",
"TIMING:bempp:_Solver.solve : 8.259e-04s\n",
"TIMING:bempp:_Solver.solve : 3.028e-05s\n",
"TIMING:bempp:_Solver.solve : 2.317e-04s\n",
"TIMING:bempp:_Solver.solve : 9.050e-04s\n",
"TIMING:bempp:_Solver.solve : 7.606e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 6.416e-04s\n",
"TIMING:bempp:_Solver.solve : 1.021e-03s\n",
"TIMING:bempp:_Solver.solve : 3.314e-05s\n",
"TIMING:bempp:_Solver.solve : 2.410e-04s\n",
"TIMING:bempp:_Solver.solve : 9.196e-04s\n",
"TIMING:bempp:_Solver.solve : 7.873e-04s\n",
"TIMING:bempp:_Solver.solve : 3.052e-05s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 7.713e-04s\n",
"TIMING:bempp:_Solver.solve : 8.078e-04s\n",
"TIMING:bempp:_Solver.solve : 1.566e-04s\n",
"TIMING:bempp:_Solver.solve : 1.392e-04s\n",
"TIMING:bempp:_Solver.solve : 8.557e-04s\n",
"TIMING:bempp:_Solver.solve : 7.634e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 6.127e-04s\n",
"TIMING:bempp:_Solver.solve : 9.358e-04s\n",
"TIMING:bempp:_Solver.solve : 3.219e-05s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 7.930e-04s\n",
"TIMING:bempp:_Solver.solve : 8.845e-04s\n",
"TIMING:bempp:_Solver.solve : 1.106e-04s\n",
"TIMING:bempp:_Solver.solve : 1.678e-04s\n",
"TIMING:bempp:_Solver.solve : 9.129e-04s\n",
"TIMING:bempp:_Solver.solve : 8.597e-04s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 1.089e-03s\n",
"TIMING:bempp:_Solver.solve : 8.030e-04s\n",
"TIMING:bempp:_Solver.solve : 3.147e-05s\n",
"TIMING:bempp:_Solver.solve : 2.260e-04s\n",
"TIMING:bempp:_Solver.solve : 9.875e-04s\n",
"TIMING:bempp:_Solver.solve : 8.254e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 1.886e-04s\n",
"TIMING:bempp:_Solver.solve : 7.875e-04s\n",
"TIMING:bempp:_Solver.solve : 8.574e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.789e-05s\n",
"TIMING:bempp:_Solver.solve : 7.846e-04s\n",
"TIMING:bempp:_Solver.solve : 8.905e-04s\n",
"TIMING:bempp:_Solver.solve : 5.770e-05s\n",
"TIMING:bempp:_Solver.solve : 7.892e-05s\n",
"TIMING:bempp:_Solver.solve : 7.753e-04s\n",
"TIMING:bempp:_Solver.solve : 8.838e-04s\n",
"TIMING:bempp:_Solver.solve : 1.109e-04s\n",
"TIMING:bempp:_Solver.solve : 1.447e-04s\n",
"TIMING:bempp:_Solver.solve : 7.291e-04s\n",
"TIMING:bempp:_Solver.solve : 9.143e-04s\n",
"TIMING:bempp:_Solver.solve : 1.495e-04s\n",
"TIMING:bempp:_Solver.solve : 1.504e-04s\n",
"TIMING:bempp:_Solver.solve : 8.612e-04s\n",
"TIMING:bempp:_Solver.solve : 7.725e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 1.674e-04s\n",
"TIMING:bempp:_Solver.solve : 7.346e-04s\n",
"TIMING:bempp:_Solver.solve : 8.109e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 2.766e-05s\n",
"TIMING:bempp:_Solver.solve : 7.334e-04s\n",
"TIMING:bempp:_Solver.solve : 8.266e-04s\n",
"TIMING:bempp:_Solver.solve : 3.052e-05s\n",
"TIMING:bempp:_Solver.solve : 3.242e-05s\n",
"TIMING:bempp:_Solver.solve : 7.696e-04s\n",
"TIMING:bempp:_Solver.solve : 8.099e-04s\n",
"TIMING:bempp:_Solver.solve : 1.256e-04s\n",
"TIMING:bempp:_Solver.solve : 1.121e-04s\n",
"TIMING:bempp:_Solver.solve : 7.157e-04s\n",
"TIMING:bempp:_Solver.solve : 7.975e-04s\n",
"TIMING:bempp:_Solver.solve : 1.366e-04s\n",
"TIMING:bempp:_Solver.solve : 1.233e-04s\n",
"TIMING:bempp:_Solver.solve : 7.782e-04s\n",
"TIMING:bempp:_Solver.solve : 7.451e-04s\n",
"TIMING:bempp:_Solver.solve : 2.742e-05s\n",
"TIMING:bempp:_Solver.solve : 2.789e-05s\n",
"TIMING:bempp:_Solver.solve : 7.544e-04s\n",
"TIMING:bempp:_Solver.solve : 9.091e-04s\n",
"TIMING:bempp:_Solver.solve : 3.409e-05s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 7.663e-04s\n",
"TIMING:bempp:_Solver.solve : 8.328e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.813e-05s\n",
"TIMING:bempp:_Solver.solve : 7.994e-04s\n",
"TIMING:bempp:_Solver.solve : 9.403e-04s\n",
"TIMING:bempp:_Solver.solve : 2.861e-05s\n",
"TIMING:bempp:_Solver.solve : 5.794e-05s\n",
"TIMING:bempp:_Solver.solve : 1.014e-03s\n",
"TIMING:bempp:_Solver.solve : 1.041e-03s\n",
"TIMING:bempp:_Solver.solve : 3.552e-05s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 6.859e-04s\n",
"TIMING:bempp:_Solver.solve : 7.877e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 6.056e-05s\n",
"TIMING:bempp:_Solver.solve : 7.474e-04s\n",
"TIMING:bempp:_Solver.solve : 8.681e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 2.503e-05s\n",
"TIMING:bempp:_Solver.solve : 7.019e-04s\n",
"TIMING:bempp:_Solver.solve : 8.023e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.544e-04s\n",
"TIMING:bempp:_Solver.solve : 8.557e-04s\n",
"TIMING:bempp:_Solver.solve : 8.774e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 6.733e-04s\n",
"TIMING:bempp:_Solver.solve : 8.707e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 2.599e-05s\n",
"TIMING:bempp:_Solver.solve : 7.441e-04s\n",
"TIMING:bempp:_Solver.solve : 7.961e-04s\n",
"TIMING:bempp:_Solver.solve : 1.431e-04s\n",
"TIMING:bempp:_Solver.solve : 1.075e-04s\n",
"TIMING:bempp:_Solver.solve : 8.585e-04s\n",
"TIMING:bempp:_Solver.solve : 8.388e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.077e-04s\n",
"TIMING:bempp:_Solver.solve : 7.350e-04s\n",
"TIMING:bempp:_Solver.solve : 9.706e-04s\n",
"TIMING:bempp:_Solver.solve : 3.028e-05s\n",
"TIMING:bempp:_Solver.solve : 2.072e-04s\n",
"TIMING:bempp:_Solver.solve : 8.371e-04s\n",
"TIMING:bempp:_Solver.solve : 8.144e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 6.566e-04s\n",
"TIMING:bempp:_Solver.solve : 8.376e-04s\n",
"TIMING:bempp:_Solver.solve : 3.767e-05s\n",
"TIMING:bempp:_Solver.solve : 7.629e-05s\n",
"TIMING:bempp:_Solver.solve : 8.314e-04s\n",
"TIMING:bempp:_Solver.solve : 7.715e-04s\n",
"TIMING:bempp:_Solver.solve : 1.252e-04s\n",
"TIMING:bempp:_Solver.solve : 1.321e-04s\n",
"TIMING:bempp:_Solver.solve : 7.391e-04s\n",
"TIMING:bempp:_Solver.solve : 1.102e-03s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 7.222e-04s\n",
"TIMING:bempp:_Solver.solve : 9.072e-04s\n",
"TIMING:bempp:_Solver.solve : 3.052e-05s\n",
"TIMING:bempp:_Solver.solve : 7.939e-05s\n",
"TIMING:bempp:_Solver.solve : 1.105e-03s\n",
"TIMING:bempp:_Solver.solve : 8.852e-04s\n",
"TIMING:bempp:_Solver.solve : 1.209e-04s\n",
"TIMING:bempp:_Solver.solve : 1.409e-04s\n",
"TIMING:bempp:_Solver.solve : 7.353e-04s\n",
"TIMING:bempp:_Solver.solve : 8.545e-04s\n",
"TIMING:bempp:_Solver.solve : 3.076e-05s\n",
"TIMING:bempp:_Solver.solve : 2.646e-05s\n",
"TIMING:bempp:_Solver.solve : 7.460e-04s\n",
"TIMING:bempp:_Solver.solve : 7.846e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 6.840e-04s\n",
"TIMING:bempp:_Solver.solve : 1.024e-03s\n",
"TIMING:bempp:_Solver.solve : 3.481e-05s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 1.011e-03s\n",
"TIMING:bempp:_Solver.solve : 8.836e-04s\n",
"TIMING:bempp:_Solver.solve : 3.409e-05s\n",
"TIMING:bempp:_Solver.solve : 3.099e-05s\n",
"TIMING:bempp:_Solver.solve : 7.524e-04s\n",
"TIMING:bempp:_Solver.solve : 8.304e-04s\n",
"TIMING:bempp:_Solver.solve : 1.159e-04s\n",
"TIMING:bempp:_Solver.solve : 1.442e-04s\n",
"TIMING:bempp:_Solver.solve : 7.365e-04s\n",
"TIMING:bempp:_Solver.solve : 8.807e-04s\n",
"TIMING:bempp:_Solver.solve : 1.700e-04s\n",
"TIMING:bempp:_Solver.solve : 2.742e-05s\n",
"TIMING:bempp:_Solver.solve : 8.399e-04s\n",
"TIMING:bempp:_Solver.solve : 8.776e-04s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 2.599e-05s\n",
"TIMING:bempp:_Solver.solve : 6.721e-04s\n",
"TIMING:bempp:_Solver.solve : 1.028e-03s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.646e-05s\n",
"TIMING:bempp:_Solver.solve : 7.219e-04s\n",
"TIMING:bempp:_Solver.solve : 8.123e-04s\n",
"TIMING:bempp:_Solver.solve : 1.709e-04s\n",
"TIMING:bempp:_Solver.solve : 1.123e-04s\n",
"TIMING:bempp:_Solver.solve : 8.078e-04s\n",
"TIMING:bempp:_Solver.solve : 8.316e-04s\n",
"TIMING:bempp:_Solver.solve : 1.166e-04s\n",
"TIMING:bempp:_Solver.solve : 1.328e-04s\n",
"TIMING:bempp:_Solver.solve : 7.341e-04s\n",
"TIMING:bempp:_Solver.solve : 8.926e-04s\n",
"TIMING:bempp:_Solver.solve : 1.616e-04s\n",
"TIMING:bempp:_Solver.solve : 1.342e-04s\n",
"TIMING:bempp:_Solver.solve : 7.768e-04s\n",
"TIMING:bempp:_Solver.solve : 7.553e-04s\n",
"TIMING:bempp:_Solver.solve : 1.092e-04s\n",
"TIMING:bempp:_Solver.solve : 1.409e-04s\n",
"TIMING:bempp:_Solver.solve : 7.980e-04s\n",
"TIMING:bempp:_Solver.solve : 8.533e-04s\n",
"TIMING:bempp:_Solver.solve : 7.772e-05s\n",
"TIMING:bempp:_Solver.solve : 2.599e-05s\n",
"TIMING:bempp:_Solver.solve : 7.429e-04s\n",
"TIMING:bempp:_Solver.solve : 9.735e-04s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 2.034e-04s\n",
"TIMING:bempp:_Solver.solve : 8.402e-04s\n",
"TIMING:bempp:_Solver.solve : 8.748e-04s\n",
"TIMING:bempp:_Solver.solve : 1.204e-04s\n",
"TIMING:bempp:_Solver.solve : 1.400e-04s\n",
"TIMING:bempp:_Solver.solve : 7.346e-04s\n",
"TIMING:bempp:_Solver.solve : 8.864e-04s\n",
"TIMING:bempp:_Solver.solve : 3.290e-05s\n",
"TIMING:bempp:_Solver.solve : 2.067e-04s\n",
"TIMING:bempp:_Solver.solve : 8.714e-04s\n",
"TIMING:bempp:_Solver.solve : 7.923e-04s\n",
"TIMING:bempp:_Solver.solve : 1.223e-04s\n",
"TIMING:bempp:_Solver.solve : 1.254e-04s\n",
"TIMING:bempp:_Solver.solve : 7.582e-04s\n",
"TIMING:bempp:_Solver.solve : 8.876e-04s\n",
"TIMING:bempp:_Solver.solve : 2.813e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 8.633e-04s\n",
"TIMING:bempp:_Solver.solve : 8.719e-04s\n",
"TIMING:bempp:_Solver.solve : 1.130e-04s\n",
"TIMING:bempp:_Solver.solve : 1.366e-04s\n",
"TIMING:bempp:_Solver.solve : 7.782e-04s\n",
"TIMING:bempp:_Solver.solve : 8.402e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 6.719e-04s\n",
"TIMING:bempp:_Solver.solve : 8.302e-04s\n",
"TIMING:bempp:_Solver.solve : 3.076e-05s\n",
"TIMING:bempp:_Solver.solve : 2.718e-05s\n",
"TIMING:bempp:_Solver.solve : 7.412e-04s\n",
"TIMING:bempp:_Solver.solve : 7.615e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 1.683e-04s\n",
"TIMING:bempp:_Solver.solve : 7.601e-04s\n",
"TIMING:bempp:_Solver.solve : 1.001e-03s\n",
"TIMING:bempp:_Solver.solve : 3.266e-05s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 7.408e-04s\n",
"TIMING:bempp:_Solver.solve : 8.438e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 7.393e-04s\n",
"TIMING:bempp:_Solver.solve : 8.175e-04s\n",
"TIMING:bempp:_Solver.solve : 2.766e-05s\n",
"TIMING:bempp:_Solver.solve : 1.469e-04s\n",
"TIMING:bempp:_Solver.solve : 6.795e-04s\n",
"TIMING:bempp:_Solver.solve : 9.582e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 7.069e-04s\n",
"TIMING:bempp:_Solver.solve : 7.644e-04s\n",
"TIMING:bempp:_Solver.solve : 2.789e-05s\n",
"TIMING:bempp:_Solver.solve : 2.599e-05s\n",
"TIMING:bempp:_Solver.solve : 6.888e-04s\n",
"TIMING:bempp:_Solver.solve : 8.492e-04s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 2.646e-05s\n",
"TIMING:bempp:_Solver.solve : 7.894e-04s\n",
"TIMING:bempp:_Solver.solve : 8.767e-04s\n",
"TIMING:bempp:_Solver.solve : 2.789e-05s\n",
"TIMING:bempp:_Solver.solve : 2.327e-04s\n",
"TIMING:bempp:_Solver.solve : 9.317e-04s\n",
"TIMING:bempp:_Solver.solve : 7.668e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 6.886e-04s\n",
"TIMING:bempp:_Solver.solve : 8.273e-04s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 7.353e-04s\n",
"TIMING:bempp:_Solver.solve : 8.230e-04s\n",
"TIMING:bempp:_Solver.solve : 1.190e-04s\n",
"TIMING:bempp:_Solver.solve : 1.605e-04s\n",
"TIMING:bempp:_Solver.solve : 8.678e-04s\n",
"TIMING:bempp:_Solver.solve : 8.283e-04s\n",
"TIMING:bempp:_Solver.solve : 1.221e-04s\n",
"TIMING:bempp:_Solver.solve : 1.402e-04s\n",
"TIMING:bempp:_Solver.solve : 7.391e-04s\n",
"TIMING:bempp:_Solver.solve : 9.961e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 7.477e-04s\n",
"TIMING:bempp:_Solver.solve : 9.437e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.551e-05s\n",
"TIMING:bempp:_Solver.solve : 6.814e-04s\n",
"TIMING:bempp:_Solver.solve : 1.156e-03s\n",
"TIMING:bempp:_Solver.solve : 3.219e-05s\n",
"TIMING:bempp:_Solver.solve : 2.043e-04s\n",
"TIMING:bempp:_Solver.solve : 8.123e-04s\n",
"TIMING:bempp:_Solver.solve : 8.984e-04s\n",
"TIMING:bempp:_Solver.solve : 1.557e-04s\n",
"TIMING:bempp:_Solver.solve : 1.249e-04s\n",
"TIMING:bempp:_Solver.solve : 7.737e-04s\n",
"TIMING:bempp:_Solver.solve : 8.090e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 1.972e-04s\n",
"TIMING:bempp:_Solver.solve : 7.770e-04s\n",
"TIMING:bempp:_Solver.solve : 8.225e-04s\n",
"TIMING:bempp:_Solver.solve : 2.394e-04s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 1.014e-03s\n",
"TIMING:bempp:_Solver.solve : 9.124e-04s\n",
"TIMING:bempp:_Solver.solve : 1.113e-04s\n",
"TIMING:bempp:_Solver.solve : 1.342e-04s\n",
"TIMING:bempp:_Solver.solve : 7.157e-04s\n",
"TIMING:bempp:_Solver.solve : 8.974e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.148e-04s\n",
"TIMING:bempp:_Solver.solve : 8.397e-04s\n",
"TIMING:bempp:_Solver.solve : 7.880e-04s\n",
"TIMING:bempp:_Solver.solve : 2.813e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 6.855e-04s\n",
"TIMING:bempp:_Solver.solve : 9.236e-04s\n",
"TIMING:bempp:_Solver.solve : 3.576e-05s\n",
"TIMING:bempp:_Solver.solve : 3.147e-05s\n",
"TIMING:bempp:_Solver.solve : 7.629e-04s\n",
"TIMING:bempp:_Solver.solve : 8.502e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 2.129e-04s\n",
"TIMING:bempp:_Solver.solve : 8.337e-04s\n",
"TIMING:bempp:_Solver.solve : 9.201e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 7.441e-04s\n",
"TIMING:bempp:_Solver.solve : 8.171e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 6.809e-04s\n",
"TIMING:bempp:_Solver.solve : 7.994e-04s\n",
"TIMING:bempp:_Solver.solve : 9.131e-05s\n",
"TIMING:bempp:_Solver.solve : 1.366e-04s\n",
"TIMING:bempp:_Solver.solve : 8.008e-04s\n",
"TIMING:bempp:_Solver.solve : 8.862e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 3.145e-04s\n",
"TIMING:bempp:_Solver.solve : 1.088e-03s\n",
"TIMING:bempp:_Solver.solve : 7.896e-04s\n",
"TIMING:bempp:_Solver.solve : 3.195e-05s\n",
"TIMING:bempp:_Solver.solve : 1.004e-04s\n",
"TIMING:bempp:_Solver.solve : 1.154e-03s\n",
"TIMING:bempp:_Solver.solve : 8.876e-04s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 9.277e-04s\n",
"TIMING:bempp:_Solver.solve : 9.077e-04s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 2.861e-05s\n",
"TIMING:bempp:_Solver.solve : 7.095e-04s\n",
"TIMING:bempp:_Solver.solve : 9.389e-04s\n",
"TIMING:bempp:_Solver.solve : 3.290e-05s\n",
"TIMING:bempp:_Solver.solve : 2.127e-04s\n",
"TIMING:bempp:_Solver.solve : 8.843e-04s\n",
"TIMING:bempp:_Solver.solve : 9.282e-04s\n",
"TIMING:bempp:_Solver.solve : 2.718e-05s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 6.180e-04s\n",
"TIMING:bempp:_Solver.solve : 9.253e-04s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 2.742e-05s\n",
"TIMING:bempp:_Solver.solve : 7.198e-04s\n",
"TIMING:bempp:_Solver.solve : 8.132e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 6.266e-04s\n",
"TIMING:bempp:_Solver.solve : 9.935e-04s\n",
"TIMING:bempp:_Solver.solve : 3.505e-05s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 7.300e-04s\n",
"TIMING:bempp:_Solver.solve : 8.905e-04s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 7.117e-04s\n",
"TIMING:bempp:_Solver.solve : 9.747e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.742e-05s\n",
"TIMING:bempp:_Solver.solve : 6.325e-04s\n",
"TIMING:bempp:_Solver.solve : 1.001e-03s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 1.588e-04s\n",
"TIMING:bempp:_Solver.solve : 7.246e-04s\n",
"TIMING:bempp:_Solver.solve : 8.693e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 5.102e-05s\n",
"TIMING:bempp:_Solver.solve : 7.000e-04s\n",
"TIMING:bempp:_Solver.solve : 9.401e-04s\n",
"TIMING:bempp:_Solver.solve : 3.099e-05s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 7.117e-04s\n",
"TIMING:bempp:_Solver.solve : 7.513e-04s\n",
"TIMING:bempp:_Solver.solve : 3.648e-05s\n",
"TIMING:bempp:_Solver.solve : 2.789e-05s\n",
"TIMING:bempp:_Solver.solve : 7.908e-04s\n",
"TIMING:bempp:_Solver.solve : 7.832e-04s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 2.813e-05s\n",
"TIMING:bempp:_Solver.solve : 6.981e-04s\n",
"TIMING:bempp:_Solver.solve : 9.274e-04s\n",
"TIMING:bempp:_Solver.solve : 3.171e-05s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 7.205e-04s\n",
"TIMING:bempp:_Solver.solve : 7.789e-04s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 6.397e-04s\n",
"TIMING:bempp:_Solver.solve : 8.433e-04s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 2.742e-05s\n",
"TIMING:bempp:_Solver.solve : 6.897e-04s\n",
"TIMING:bempp:_Solver.solve : 7.684e-04s\n",
"TIMING:bempp:_Solver.solve : 3.076e-05s\n",
"TIMING:bempp:_Solver.solve : 1.335e-04s\n",
"TIMING:bempp:_Solver.solve : 7.365e-04s\n",
"TIMING:bempp:_Solver.solve : 9.315e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 2.742e-05s\n",
"TIMING:bempp:_Solver.solve : 6.349e-04s\n",
"TIMING:bempp:_Solver.solve : 9.620e-04s\n",
"TIMING:bempp:_Solver.solve : 3.076e-05s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 6.733e-04s\n",
"TIMING:bempp:_Solver.solve : 7.966e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.718e-05s\n",
"TIMING:bempp:_Solver.solve : 6.642e-04s\n",
"TIMING:bempp:_Solver.solve : 7.730e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 2.029e-04s\n",
"TIMING:bempp:_Solver.solve : 1.004e-03s\n",
"TIMING:bempp:_Solver.solve : 8.471e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 4.292e-05s\n",
"TIMING:bempp:_Solver.solve : 1.043e-03s\n",
"TIMING:bempp:_Solver.solve : 8.552e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 2.694e-05s\n",
"TIMING:bempp:_Solver.solve : 6.399e-04s\n",
"TIMING:bempp:_Solver.solve : 9.432e-04s\n",
"TIMING:bempp:_Solver.solve : 3.171e-05s\n",
"TIMING:bempp:_Solver.solve : 2.861e-05s\n",
"TIMING:bempp:_Solver.solve : 6.921e-04s\n",
"TIMING:bempp:_Solver.solve : 9.501e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 1.996e-04s\n",
"TIMING:bempp:_Solver.solve : 8.960e-04s\n",
"TIMING:bempp:_Solver.solve : 9.127e-04s\n",
"TIMING:bempp:_Solver.solve : 3.195e-05s\n",
"TIMING:bempp:_Solver.solve : 2.742e-05s\n",
"TIMING:bempp:_Solver.solve : 6.196e-04s\n",
"TIMING:bempp:_Solver.solve : 9.675e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 3.099e-05s\n",
"TIMING:bempp:_Solver.solve : 7.041e-04s\n",
"TIMING:bempp:_Solver.solve : 8.624e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.694e-05s\n",
"TIMING:bempp:_Solver.solve : 6.201e-04s\n",
"TIMING:bempp:_Solver.solve : 9.489e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 1.636e-04s\n",
"TIMING:bempp:_Solver.solve : 7.329e-04s\n",
"TIMING:bempp:_Solver.solve : 8.171e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.501e-04s\n",
"TIMING:bempp:_Solver.solve : 9.875e-04s\n",
"TIMING:bempp:_Solver.solve : 7.746e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 9.131e-05s\n",
"TIMING:bempp:_Solver.solve : 8.180e-04s\n",
"TIMING:bempp:_Solver.solve : 9.394e-04s\n",
"TIMING:bempp:_Solver.solve : 1.562e-04s\n",
"TIMING:bempp:_Solver.solve : 1.459e-04s\n",
"TIMING:bempp:_Solver.solve : 8.969e-04s\n",
"TIMING:bempp:_Solver.solve : 8.383e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.742e-05s\n",
"TIMING:bempp:_Solver.solve : 6.540e-04s\n",
"TIMING:bempp:_Solver.solve : 8.056e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 7.679e-04s\n",
"TIMING:bempp:_Solver.solve : 7.539e-04s\n",
"TIMING:bempp:_Solver.solve : 3.052e-05s\n",
"TIMING:bempp:_Solver.solve : 2.813e-05s\n",
"TIMING:bempp:_Solver.solve : 7.370e-04s\n",
"TIMING:bempp:_Solver.solve : 7.710e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 1.636e-04s\n",
"TIMING:bempp:_Solver.solve : 7.601e-04s\n",
"TIMING:bempp:_Solver.solve : 9.363e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.861e-05s\n",
"TIMING:bempp:_Solver.solve : 7.267e-04s\n",
"TIMING:bempp:_Solver.solve : 8.826e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 2.718e-05s\n",
"TIMING:bempp:_Solver.solve : 6.533e-04s\n",
"TIMING:bempp:_Solver.solve : 1.088e-03s\n",
"TIMING:bempp:_Solver.solve : 3.171e-05s\n",
"TIMING:bempp:_Solver.solve : 1.659e-04s\n",
"TIMING:bempp:_Solver.solve : 7.603e-04s\n",
"TIMING:bempp:_Solver.solve : 8.311e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.766e-05s\n",
"TIMING:bempp:_Solver.solve : 6.685e-04s\n",
"TIMING:bempp:_Solver.solve : 8.574e-04s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 3.028e-05s\n",
"TIMING:bempp:_Solver.solve : 8.593e-04s\n",
"TIMING:bempp:_Solver.solve : 9.232e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 2.646e-05s\n",
"TIMING:bempp:_Solver.solve : 6.733e-04s\n",
"TIMING:bempp:_Solver.solve : 8.557e-04s\n",
"TIMING:bempp:_Solver.solve : 1.099e-04s\n",
"TIMING:bempp:_Solver.solve : 1.299e-04s\n",
"TIMING:bempp:_Solver.solve : 7.119e-04s\n",
"TIMING:bempp:_Solver.solve : 9.627e-04s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 2.069e-04s\n",
"TIMING:bempp:_Solver.solve : 8.409e-04s\n",
"TIMING:bempp:_Solver.solve : 8.721e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 2.766e-05s\n",
"TIMING:bempp:_Solver.solve : 8.075e-04s\n",
"TIMING:bempp:_Solver.solve : 8.121e-04s\n",
"TIMING:bempp:_Solver.solve : 3.076e-05s\n",
"TIMING:bempp:_Solver.solve : 2.694e-05s\n",
"TIMING:bempp:_Solver.solve : 6.752e-04s\n",
"TIMING:bempp:_Solver.solve : 9.019e-04s\n",
"TIMING:bempp:_Solver.solve : 3.052e-05s\n",
"TIMING:bempp:_Solver.solve : 2.789e-05s\n",
"TIMING:bempp:_Solver.solve : 7.117e-04s\n",
"TIMING:bempp:_Solver.solve : 9.451e-04s\n",
"TIMING:bempp:_Solver.solve : 1.514e-04s\n",
"TIMING:bempp:_Solver.solve : 1.497e-04s\n",
"TIMING:bempp:_Solver.solve : 9.933e-04s\n",
"TIMING:bempp:_Solver.solve : 9.003e-04s\n",
"TIMING:bempp:_Solver.solve : 1.552e-04s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 9.794e-04s\n",
"TIMING:bempp:_Solver.solve : 8.979e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.599e-05s\n",
"TIMING:bempp:_Solver.solve : 7.212e-04s\n",
"TIMING:bempp:_Solver.solve : 8.316e-04s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 2.646e-05s\n",
"TIMING:bempp:_Solver.solve : 6.638e-04s\n",
"TIMING:bempp:_Solver.solve : 9.694e-04s\n",
"TIMING:bempp:_Solver.solve : 1.874e-04s\n",
"TIMING:bempp:_Solver.solve : 1.414e-04s\n",
"TIMING:bempp:_Solver.solve : 8.824e-04s\n",
"TIMING:bempp:_Solver.solve : 8.874e-04s\n",
"TIMING:bempp:_Solver.solve : 3.171e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 6.621e-04s\n",
"TIMING:bempp:_Solver.solve : 9.360e-04s\n",
"TIMING:bempp:_Solver.solve : 2.861e-05s\n",
"TIMING:bempp:_Solver.solve : 2.742e-05s\n",
"TIMING:bempp:_Solver.solve : 6.599e-04s\n",
"TIMING:bempp:_Solver.solve : 9.828e-04s\n",
"TIMING:bempp:_Solver.solve : 3.529e-05s\n",
"TIMING:bempp:_Solver.solve : 2.718e-05s\n",
"TIMING:bempp:_Solver.solve : 6.821e-04s\n",
"TIMING:bempp:_Solver.solve : 7.772e-04s\n",
"TIMING:bempp:_Solver.solve : 1.502e-04s\n",
"TIMING:bempp:_Solver.solve : 1.342e-04s\n",
"TIMING:bempp:_Solver.solve : 7.856e-04s\n",
"TIMING:bempp:_Solver.solve : 9.468e-04s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 2.646e-05s\n",
"TIMING:bempp:_Solver.solve : 7.041e-04s\n",
"TIMING:bempp:_Solver.solve : 7.648e-04s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 2.551e-05s\n",
"TIMING:bempp:_Solver.solve : 6.568e-04s\n",
"TIMING:bempp:_Solver.solve : 8.907e-04s\n",
"TIMING:bempp:_Solver.solve : 3.147e-05s\n",
"TIMING:bempp:_Solver.solve : 1.426e-04s\n",
"TIMING:bempp:_Solver.solve : 8.092e-04s\n",
"TIMING:bempp:_Solver.solve : 9.015e-04s\n",
"TIMING:bempp:_Solver.solve : 3.433e-05s\n",
"TIMING:bempp:_Solver.solve : 2.861e-05s\n",
"TIMING:bempp:_Solver.solve : 7.112e-04s\n",
"TIMING:bempp:_Solver.solve : 7.842e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.789e-05s\n",
"TIMING:bempp:_Solver.solve : 6.449e-04s\n",
"TIMING:bempp:_Solver.solve : 8.721e-04s\n",
"TIMING:bempp:_Solver.solve : 3.028e-05s\n",
"TIMING:bempp:_Solver.solve : 2.718e-05s\n",
"TIMING:bempp:_Solver.solve : 6.337e-04s\n",
"TIMING:bempp:_Solver.solve : 7.844e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.694e-05s\n",
"TIMING:bempp:_Solver.solve : 6.328e-04s\n",
"TIMING:bempp:_Solver.solve : 8.619e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.789e-05s\n",
"TIMING:bempp:_Solver.solve : 7.644e-04s\n",
"TIMING:bempp:_Solver.solve : 8.283e-04s\n",
"TIMING:bempp:_Solver.solve : 1.605e-04s\n",
"TIMING:bempp:_Solver.solve : 1.631e-04s\n",
"TIMING:bempp:_Solver.solve : 1.110e-03s\n",
"TIMING:bempp:_Solver.solve : 7.739e-04s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 2.694e-05s\n",
"TIMING:bempp:_Solver.solve : 6.695e-04s\n",
"TIMING:bempp:_Solver.solve : 7.908e-04s\n",
"TIMING:bempp:_Solver.solve : 3.147e-05s\n",
"TIMING:bempp:_Solver.solve : 2.766e-05s\n",
"TIMING:bempp:_Solver.solve : 7.350e-04s\n",
"TIMING:bempp:_Solver.solve : 8.740e-04s\n",
"TIMING:bempp:_Solver.solve : 2.813e-05s\n",
"TIMING:bempp:_Solver.solve : 2.599e-05s\n",
"TIMING:bempp:_Solver.solve : 6.585e-04s\n",
"TIMING:bempp:_Solver.solve : 9.217e-04s\n",
"TIMING:bempp:_Solver.solve : 2.646e-05s\n",
"TIMING:bempp:_Solver.solve : 2.599e-05s\n",
"TIMING:bempp:_Solver.solve : 6.382e-04s\n",
"TIMING:bempp:_Solver.solve : 9.277e-04s\n",
"TIMING:bempp:_Solver.solve : 3.123e-05s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 7.150e-04s\n",
"TIMING:bempp:_Solver.solve : 9.112e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.718e-05s\n",
"TIMING:bempp:_Solver.solve : 6.676e-04s\n",
"TIMING:bempp:_Solver.solve : 9.582e-04s\n",
"TIMING:bempp:_Solver.solve : 3.076e-05s\n",
"TIMING:bempp:_Solver.solve : 2.766e-05s\n",
"TIMING:bempp:_Solver.solve : 6.936e-04s\n",
"TIMING:bempp:_Solver.solve : 8.311e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 6.919e-04s\n",
"TIMING:bempp:_Solver.solve : 8.330e-04s\n",
"TIMING:bempp:_Solver.solve : 1.481e-04s\n",
"TIMING:bempp:_Solver.solve : 1.342e-04s\n",
"TIMING:bempp:_Solver.solve : 8.111e-04s\n",
"TIMING:bempp:_Solver.solve : 8.171e-04s\n",
"TIMING:bempp:_Solver.solve : 1.671e-04s\n",
"TIMING:bempp:_Solver.solve : 1.531e-04s\n",
"TIMING:bempp:_Solver.solve : 8.595e-04s\n",
"TIMING:bempp:_Solver.solve : 7.720e-04s\n",
"TIMING:bempp:_Solver.solve : 1.142e-04s\n",
"TIMING:bempp:_Solver.solve : 1.256e-04s\n",
"TIMING:bempp:_Solver.solve : 7.246e-04s\n",
"TIMING:bempp:_Solver.solve : 8.538e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 2.000e-04s\n",
"TIMING:bempp:_Solver.solve : 8.044e-04s\n",
"TIMING:bempp:_Solver.solve : 9.112e-04s\n",
"TIMING:bempp:_Solver.solve : 2.861e-05s\n",
"TIMING:bempp:_Solver.solve : 2.599e-05s\n",
"TIMING:bempp:_Solver.solve : 6.788e-04s\n",
"TIMING:bempp:_Solver.solve : 9.756e-04s\n",
"TIMING:bempp:_Solver.solve : 3.242e-05s\n",
"TIMING:bempp:_Solver.solve : 3.099e-05s\n",
"TIMING:bempp:_Solver.solve : 9.980e-04s\n",
"TIMING:bempp:_Solver.solve : 8.419e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.005e-04s\n",
"TIMING:bempp:_Solver.solve : 8.228e-04s\n",
"TIMING:bempp:_Solver.solve : 7.663e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 3.242e-05s\n",
"TIMING:bempp:_Solver.solve : 6.719e-04s\n",
"TIMING:bempp:_Solver.solve : 9.720e-04s\n",
"TIMING:bempp:_Solver.solve : 3.123e-05s\n",
"TIMING:bempp:_Solver.solve : 1.769e-04s\n",
"TIMING:bempp:_Solver.solve : 7.579e-04s\n",
"TIMING:bempp:_Solver.solve : 8.426e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 1.822e-04s\n",
"TIMING:bempp:_Solver.solve : 7.722e-04s\n",
"TIMING:bempp:_Solver.solve : 9.813e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.742e-05s\n",
"TIMING:bempp:_Solver.solve : 6.967e-04s\n",
"TIMING:bempp:_Solver.solve : 8.464e-04s\n",
"TIMING:bempp:_Solver.solve : 1.926e-04s\n",
"TIMING:bempp:_Solver.solve : 2.694e-05s\n",
"TIMING:bempp:_Solver.solve : 8.197e-04s\n",
"TIMING:bempp:_Solver.solve : 8.035e-04s\n",
"TIMING:bempp:_Solver.solve : 3.219e-05s\n",
"TIMING:bempp:_Solver.solve : 2.813e-05s\n",
"TIMING:bempp:_Solver.solve : 7.024e-04s\n",
"TIMING:bempp:_Solver.solve : 8.850e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.694e-05s\n",
"TIMING:bempp:_Solver.solve : 8.922e-04s\n",
"TIMING:bempp:_Solver.solve : 8.266e-04s\n",
"TIMING:bempp:_Solver.solve : 4.697e-05s\n",
"TIMING:bempp:_Solver.solve : 2.718e-05s\n",
"TIMING:bempp:_Solver.solve : 7.942e-04s\n",
"TIMING:bempp:_Solver.solve : 8.368e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 7.305e-04s\n",
"TIMING:bempp:_Solver.solve : 8.373e-04s\n",
"TIMING:bempp:_Solver.solve : 1.242e-04s\n",
"TIMING:bempp:_Solver.solve : 1.469e-04s\n",
"TIMING:bempp:_Solver.solve : 7.470e-04s\n",
"TIMING:bempp:_Solver.solve : 8.626e-04s\n",
"TIMING:bempp:_Solver.solve : 1.595e-04s\n",
"TIMING:bempp:_Solver.solve : 1.640e-04s\n",
"TIMING:bempp:_Solver.solve : 9.151e-04s\n",
"TIMING:bempp:_Solver.solve : 1.066e-03s\n",
"TIMING:bempp:_Solver.solve : 1.893e-04s\n",
"TIMING:bempp:_Solver.solve : 1.204e-04s\n",
"TIMING:bempp:_Solver.solve : 8.221e-04s\n",
"TIMING:bempp:_Solver.solve : 9.294e-04s\n",
"TIMING:bempp:_Solver.solve : 1.266e-04s\n",
"TIMING:bempp:_Solver.solve : 1.159e-04s\n",
"TIMING:bempp:_Solver.solve : 7.150e-04s\n",
"TIMING:bempp:_Solver.solve : 8.457e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 7.491e-04s\n",
"TIMING:bempp:_Solver.solve : 8.650e-04s\n",
"TIMING:bempp:_Solver.solve : 8.726e-05s\n",
"TIMING:bempp:_Solver.solve : 1.953e-04s\n",
"TIMING:bempp:_Solver.solve : 8.335e-04s\n",
"TIMING:bempp:_Solver.solve : 7.718e-04s\n",
"TIMING:bempp:_Solver.solve : 1.278e-04s\n",
"TIMING:bempp:_Solver.solve : 1.190e-04s\n",
"TIMING:bempp:_Solver.solve : 7.644e-04s\n",
"TIMING:bempp:_Solver.solve : 8.132e-04s\n",
"TIMING:bempp:_Solver.solve : 1.228e-04s\n",
"TIMING:bempp:_Solver.solve : 1.106e-04s\n",
"TIMING:bempp:_Solver.solve : 7.172e-04s\n",
"TIMING:bempp:_Solver.solve : 8.841e-04s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 6.576e-04s\n",
"TIMING:bempp:_Solver.solve : 9.015e-04s\n",
"TIMING:bempp:_Solver.solve : 3.099e-05s\n",
"TIMING:bempp:_Solver.solve : 2.813e-05s\n",
"TIMING:bempp:_Solver.solve : 6.807e-04s\n",
"TIMING:bempp:_Solver.solve : 8.509e-04s\n",
"TIMING:bempp:_Solver.solve : 2.861e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 6.573e-04s\n",
"TIMING:bempp:_Solver.solve : 7.787e-04s\n",
"TIMING:bempp:_Solver.solve : 1.211e-04s\n",
"TIMING:bempp:_Solver.solve : 1.113e-04s\n",
"TIMING:bempp:_Solver.solve : 7.091e-04s\n",
"TIMING:bempp:_Solver.solve : 1.092e-03s\n",
"TIMING:bempp:_Solver.solve : 3.123e-05s\n",
"TIMING:bempp:_Solver.solve : 2.599e-05s\n",
"TIMING:bempp:_Solver.solve : 6.456e-04s\n",
"TIMING:bempp:_Solver.solve : 7.997e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.503e-05s\n",
"TIMING:bempp:_Solver.solve : 6.387e-04s\n",
"TIMING:bempp:_Solver.solve : 9.711e-04s\n",
"TIMING:bempp:_Solver.solve : 2.980e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 7.052e-04s\n",
"TIMING:bempp:_Solver.solve : 7.899e-04s\n",
"TIMING:bempp:_Solver.solve : 2.742e-05s\n",
"TIMING:bempp:_Solver.solve : 2.151e-04s\n",
"TIMING:bempp:_Solver.solve : 7.806e-04s\n",
"TIMING:bempp:_Solver.solve : 7.870e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 6.444e-04s\n",
"TIMING:bempp:_Solver.solve : 9.627e-04s\n",
"TIMING:bempp:_Solver.solve : 3.099e-05s\n",
"TIMING:bempp:_Solver.solve : 4.959e-05s\n",
"TIMING:bempp:_Solver.solve : 9.801e-04s\n",
"TIMING:bempp:_Solver.solve : 8.984e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.742e-05s\n",
"TIMING:bempp:_Solver.solve : 7.408e-04s\n",
"TIMING:bempp:_Solver.solve : 8.373e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 6.433e-04s\n",
"TIMING:bempp:_Solver.solve : 7.827e-04s\n",
"TIMING:bempp:_Solver.solve : 2.813e-05s\n",
"TIMING:bempp:_Solver.solve : 2.551e-05s\n",
"TIMING:bempp:_Solver.solve : 6.697e-04s\n",
"TIMING:bempp:_Solver.solve : 7.706e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 2.437e-04s\n",
"TIMING:bempp:_Solver.solve : 1.010e-03s\n",
"TIMING:bempp:_Solver.solve : 8.979e-04s\n",
"TIMING:bempp:_Solver.solve : 3.457e-05s\n",
"TIMING:bempp:_Solver.solve : 2.766e-05s\n",
"TIMING:bempp:_Solver.solve : 7.634e-04s\n",
"TIMING:bempp:_Solver.solve : 8.512e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.646e-05s\n",
"TIMING:bempp:_Solver.solve : 6.588e-04s\n",
"TIMING:bempp:_Solver.solve : 8.893e-04s\n",
"TIMING:bempp:_Solver.solve : 3.147e-05s\n",
"TIMING:bempp:_Solver.solve : 1.979e-04s\n",
"TIMING:bempp:_Solver.solve : 8.154e-04s\n",
"TIMING:bempp:_Solver.solve : 8.168e-04s\n",
"TIMING:bempp:_Solver.solve : 1.175e-04s\n",
"TIMING:bempp:_Solver.solve : 1.011e-04s\n",
"TIMING:bempp:_Solver.solve : 7.098e-04s\n",
"TIMING:bempp:_Solver.solve : 7.584e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 6.442e-04s\n",
"TIMING:bempp:_Solver.solve : 7.563e-04s\n",
"TIMING:bempp:_Solver.solve : 2.861e-05s\n",
"TIMING:bempp:_Solver.solve : 2.220e-04s\n",
"TIMING:bempp:_Solver.solve : 8.397e-04s\n",
"TIMING:bempp:_Solver.solve : 8.721e-04s\n",
"TIMING:bempp:_Solver.solve : 2.956e-05s\n",
"TIMING:bempp:_Solver.solve : 2.663e-04s\n",
"TIMING:bempp:_Solver.solve : 9.387e-04s\n",
"TIMING:bempp:_Solver.solve : 9.401e-04s\n",
"TIMING:bempp:_Solver.solve : 1.290e-04s\n",
"TIMING:bempp:_Solver.solve : 1.152e-04s\n",
"TIMING:bempp:_Solver.solve : 7.501e-04s\n",
"TIMING:bempp:_Solver.solve : 8.378e-04s\n",
"TIMING:bempp:_Solver.solve : 2.861e-05s\n",
"TIMING:bempp:_Solver.solve : 2.933e-05s\n",
"TIMING:bempp:_Solver.solve : 8.886e-04s\n",
"TIMING:bempp:_Solver.solve : 8.478e-04s\n",
"TIMING:bempp:_Solver.solve : 2.766e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 8.099e-04s\n",
"TIMING:bempp:_Solver.solve : 1.044e-03s\n",
"TIMING:bempp:_Solver.solve : 5.722e-05s\n",
"TIMING:bempp:_Solver.solve : 1.850e-04s\n",
"TIMING:bempp:_Solver.solve : 9.687e-04s\n",
"TIMING:bempp:_Solver.solve : 8.442e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.575e-05s\n",
"TIMING:bempp:_Solver.solve : 7.226e-04s\n",
"TIMING:bempp:_Solver.solve : 1.020e-03s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 2.718e-05s\n",
"TIMING:bempp:_Solver.solve : 1.040e-03s\n",
"TIMING:bempp:_Solver.solve : 9.012e-04s\n",
"TIMING:bempp:_Solver.solve : 1.404e-04s\n",
"TIMING:bempp:_Solver.solve : 2.613e-04s\n",
"TIMING:bempp:_Solver.solve : 1.100e-03s\n",
"TIMING:bempp:_Solver.solve : 7.753e-04s\n",
"TIMING:bempp:_Solver.solve : 1.228e-04s\n",
"TIMING:bempp:_Solver.solve : 1.481e-04s\n",
"TIMING:bempp:_Solver.solve : 8.848e-04s\n",
"TIMING:bempp:_Solver.solve : 8.173e-04s\n",
"TIMING:bempp:_Solver.solve : 2.837e-05s\n",
"TIMING:bempp:_Solver.solve : 2.670e-05s\n",
"TIMING:bempp:_Solver.solve : 7.241e-04s\n",
"TIMING:bempp:_Solver.solve : 8.817e-04s\n",
"TIMING:bempp:_Solver.solve : 1.194e-04s\n",
"TIMING:bempp:_Solver.solve : 1.543e-04s\n",
"TIMING:bempp:_Solver.solve : 8.168e-04s\n",
"TIMING:bempp:_Solver.solve : 8.566e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 1.998e-04s\n",
"TIMING:bempp:_Solver.solve : 7.658e-04s\n",
"TIMING:bempp:_Solver.solve : 8.221e-04s\n",
"TIMING:bempp:_Solver.solve : 1.285e-04s\n",
"TIMING:bempp:_Solver.solve : 1.445e-04s\n",
"TIMING:bempp:_Solver.solve : 8.807e-04s\n",
"TIMING:bempp:_Solver.solve : 7.823e-04s\n",
"TIMING:bempp:_Solver.solve : 3.004e-05s\n",
"TIMING:bempp:_Solver.solve : 1.431e-04s\n",
"TIMING:bempp:_Solver.solve : 8.276e-04s\n",
"TIMING:bempp:_Solver.solve : 9.208e-04s\n",
"TIMING:bempp:_Solver.solve : 5.603e-05s\n",
"TIMING:bempp:_Solver.solve : 2.344e-04s\n",
"TIMING:bempp:_Solver.solve : 1.054e-03s\n",
"TIMING:bempp:_Solver.solve : 9.265e-04s\n",
"TIMING:bempp:_Solver.solve : 2.885e-05s\n",
"TIMING:bempp:_Solver.solve : 2.694e-05s\n",
"TIMING:bempp:_Solver.solve : 7.012e-04s\n",
"TIMING:bempp:_Solver.solve : 9.289e-04s\n",
"TIMING:bempp:_Solver.solve : 3.195e-05s\n",
"TIMING:bempp:_Solver.solve : 2.027e-04s\n",
"TIMING:bempp:_Solver.solve : 8.304e-04s\n",
"TIMING:bempp:_Solver.solve : 1.013e-03s\n",
"TIMING:bempp:_Solver.solve : 3.052e-05s\n",
"TIMING:bempp:_Solver.solve : 2.131e-04s\n",
"TIMING:bempp:_Solver.solve : 8.588e-04s\n",
"TIMING:bempp:_Solver.solve : 7.641e-04s\n",
"TIMING:bempp:_Solver.solve : 2.789e-05s\n",
"TIMING:bempp:_Solver.solve : 2.527e-05s\n",
"TIMING:bempp:_Solver.solve : 7.172e-04s\n",
"TIMING:bempp:_Solver.solve : 8.199e-04s\n",
"TIMING:bempp:_Solver.solve : 1.307e-04s\n",
"TIMING:bempp:_Solver.solve : 1.152e-04s\n",
"TIMING:bempp:_Solver.solve : 7.522e-04s\n",
"TIMING:bempp:_Solver.solve : 8.307e-04s\n",
"TIMING:bempp:_Solver.solve : 2.909e-05s\n",
"TIMING:bempp:_Solver.solve : 2.623e-05s\n",
"TIMING:bempp:_Solver.solve : 6.917e-04s\n",
"TIMING:bempp:_Solver.solve : 9.966e-04s\n",
"TIMING:bempp:_Solver.solve : 1.533e-04s\n",
"TIMING:bempp:_Solver.solve : 1.421e-04s\n",
"TIMING:bempp:_Solver.solve : 8.097e-04s\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Number of iterations: 312\n"
]
}
],
"source": [
"from bempp.api.assembly.discrete_boundary_operator import InverseSparseDiscreteBoundaryOperator\n",
"from scipy.sparse.linalg import LinearOperator\n",
"\n",
"# Compute the sparse inverse of the Helmholtz operator\n",
"# Although it is not a boundary operator we can use\n",
"# the SparseInverseDiscreteBoundaryOperator function from\n",
"# BEM++ to turn its LU decomposition into a linear operator.\n",
"P1 = InverseSparseDiscreteBoundaryOperator(\n",
" blocked[0,0].to_sparse().tocsc())\n",
"\n",
"# For the Laplace slp we use a simple mass matrix preconditioner. \n",
"# This is sufficient for smaller low-frequency problems.\n",
"P2 = InverseSparseDiscreteBoundaryOperator(\n",
" bempp.api.operators.boundary.sparse.identity(\n",
" bempp_space, bempp_space, bempp_space).weak_form())\n",
"\n",
"# Create a block diagonal preconditioner object using the Scipy LinearOperator class\n",
"def apply_prec(x):\n",
" \"\"\"Apply the block diagonal preconditioner\"\"\"\n",
" m1 = P1.shape[0]\n",
" m2 = P2.shape[0]\n",
" n1 = P1.shape[1]\n",
" n2 = P2.shape[1]\n",
" \n",
" res1 = P1.dot(x[:n1])\n",
" res2 = P2.dot(x[n1:])\n",
" return np.concatenate([res1, res2])\n",
"\n",
"p_shape = (P1.shape[0] + P2.shape[0], P1.shape[1] + P2.shape[1])\n",
"P = LinearOperator(p_shape, apply_prec, dtype=np.dtype('complex128'))\n",
"\n",
"# Create a callback function to count the number of iterations\n",
"it_count = 0\n",
"def count_iterations(x):\n",
" global it_count\n",
" it_count += 1\n",
"\n",
"from scipy.sparse.linalg import gmres\n",
"soln, info = gmres(blocked, rhs, M=P, callback=count_iterations)\n",
"\n",
"soln_fem = soln[:fem_size]\n",
"soln_bem = soln[fem_size:]\n",
"\n",
"print(\"Number of iterations: {0}\".format(it_count))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Next, we make DOLFINx and Bempp functions from the solution."
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [],
"source": [
"# Store the real part of the FEM solution\n",
"u = function(fenics_space)\n",
"u.vector[:] = np.ascontiguousarray(np.real(soln_fem))\n",
"\n",
"# Solution function with dirichlet data on the boundary\n",
"dirichlet_data = trace_matrix * soln_fem\n",
"dirichlet_fun = bempp.api.GridFunction(trace_space, coefficients=dirichlet_data)\n",
"\n",
"# Solution function with Neumann data on the boundary\n",
"neumann_fun = bempp.api.GridFunction(bempp_space, coefficients=soln_bem)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We now evaluate the solution on the slice $z=0.5$ and plot it. For the exterior domain, we use the respresentation formula\n",
"\n",
"$$\n",
"u^\\text{s} = \\mathcal{K}u-\\mathcal{V}\\frac{\\partial u}{\\partial \\nu}\n",
"$$\n",
"\n",
"to evaluate the solution."
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"# The next command ensures that plots are shown within the IPython notebook\n",
"%matplotlib inline\n",
"\n",
"Nx=200\n",
"Ny=200\n",
"xmin, xmax, ymin, ymax=[-1,3,-1,3]\n",
"plot_grid = np.mgrid[xmin:xmax:Nx*1j,ymin:ymax:Ny*1j]\n",
"points = np.vstack((plot_grid[0].ravel(),\n",
" plot_grid[1].ravel(),\n",
" np.array([0.5]*plot_grid[0].size)))\n",
"plot_me = np.zeros(points.shape[1], dtype=np.complex128)\n",
"\n",
"x,y,z = points\n",
"bem_x = np.logical_not((x>0) * (x<1) * (y>0) * (y<1) * (z>0) * (z<1))\n",
"\n",
"slp_pot= bempp.api.operators.potential.helmholtz.single_layer(\n",
" bempp_space, points[:, bem_x], k)\n",
"dlp_pot= bempp.api.operators.potential.helmholtz.double_layer(\n",
" trace_space, points[:, bem_x], k)\n",
"\n",
"plot_me[bem_x] += np.exp(1j * k * (points[0, bem_x] * d[0] \\\n",
" + points[1, bem_x] * d[1] \\\n",
" + points[2, bem_x] * d[2]))\n",
"plot_me[bem_x] += dlp_pot.evaluate(dirichlet_fun).flat\n",
"plot_me[bem_x] -= slp_pot.evaluate(neumann_fun).flat\n",
"\n",
"fem_points = points[:, np.logical_not(bem_x)].transpose()\n",
"try:\n",
" # Compatibility with older versions of DOLFINx\n",
" tree = dolfinx.geometry.BoundingBoxTree(mesh, 3)\n",
"except TypeError:\n",
" tree = dolfinx.geometry.bb_tree(mesh, 3)\n",
"midpoint_tree = dolfinx.geometry.create_midpoint_tree(\n",
" mesh, 3, list(range(mesh.topology.connectivity(3, 0).num_nodes))\n",
")\n",
"entities = []\n",
"for point in fem_points:\n",
" entities.append(dolfinx.geometry.compute_closest_entity(tree, midpoint_tree, mesh, point)[0])\n",
"fem_val = u.eval(fem_points, entities)\n",
"\n",
"plot_me[np.logical_not(bem_x)] += fem_val.T[0]\n",
"\n",
"plot_me = plot_me.reshape((Nx, Ny))\n",
"\n",
"plot_me = plot_me.transpose()[::-1]\n",
"\n",
"# Plot the image\n",
"from matplotlib import pyplot as plt\n",
"fig=plt.figure(figsize=(10, 8))\n",
"plt.imshow(np.real(plot_me), extent=[xmin, xmax, ymin, ymax])\n",
"plt.xlabel('x')\n",
"plt.ylabel('y')\n",
"plt.colorbar()\n",
"plt.title(\"FEM-BEM Coupling for Helmholtz\")\n",
"plt.show()"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.8.10"
}
},
"nbformat": 4,
"nbformat_minor": 4
}