from typing import Optional, Union
import numpy as np
from osaft.core.backgroundfields import WaveType
from osaft.core.frequency import Frequency
from osaft.core.functions import BesselFunctions as Bf
from osaft.core.geometries import Sphere
from osaft.core.variable import ActiveListVariable, ActiveVariable
from osaft.solutions.doinikov1994compressible.base import (
BaseDoinikov1994Compressible,
)
[docs]class CoefficientMatrix(BaseDoinikov1994Compressible):
"""Coefficient Matrix Doinikov (viscous fluid-viscous sphere; 1994)
:param f: Frequency [Hz]
:param R_0: Radius of the sphere [m]
:param rho_s: Density of the sphere [kg/m^3]
:param c_s: Speed of sound of in the sphere [m/s]
:param eta_s: shear viscosity of in the sphere [Pa s]
:param zeta_s: bulk viscosity of in the sphere [Pa s]
:param rho_f: Density of the fluid [kg/m^3]
:param c_f: Speed of sound of the fluid [m/s]
:param eta_f: shear viscosity [Pa s]
:param zeta_f: bulk viscosity [Pa s]
:param p_0: Pressure amplitude of the field [Pa]
:param position: Position within the standing wave field [m]
:param wave_type: Type of wave, traveling or standing
"""
def __init__(
self, f: Union[Frequency, float, int],
R_0: Union[Sphere, float, int],
rho_s: float, c_s: float, eta_s: float, zeta_s: float,
rho_f: float, c_f: float, eta_f: float, zeta_f: float,
p_0: float, wave_type: WaveType, position: float,
) -> None:
# init of parent method
super().__init__(
f=f, R_0=R_0,
rho_s=rho_s, c_s=c_s, eta_s=eta_s, zeta_s=zeta_s,
rho_f=rho_f, c_f=c_f, eta_f=eta_f, zeta_f=zeta_f,
p_0=p_0, wave_type=wave_type, position=position,
)
# Dependent Variables
self._x_hat = ActiveVariable(self._compute_x_hat, 'x_hat')
self._x_hat_v = ActiveVariable(self._compute_x_hat_v, 'x_v hat')
self._list_matrix_M_n = ActiveListVariable(
self._compute_matrix_M_n,
'Matrices M(n)',
)
self._list_vector_N_n = ActiveListVariable(
self._compute_vector_N_n,
'Vectors N(n)',
)
self._x_hat.is_computed_by(
self.scatterer._k_f, self.sphere._R_0,
)
self._x_hat_v.is_computed_by(
self.scatterer._k_v, self.sphere._R_0,
)
self._list_matrix_M_n.is_computed_by(
self._x, self._x_v,
self._x_hat, self._x_hat_v,
self.frequency._omega,
self.fluid._c_f, self.scatterer._c_f,
self.fluid._rho_f, self.scatterer._rho_f,
self.fluid._eta_f, self.scatterer._eta_f,
self.fluid._zeta_f, self.scatterer._zeta_f,
)
self._list_vector_N_n.is_computed_by(
self._x,
self.frequency._omega,
self.fluid._c_f, self.fluid._rho_f,
self.fluid._eta_f, self.fluid._zeta_f,
)
[docs] def det_M_n(self, n: int, column: Optional[int] = None) -> complex:
"""Determinant of the matrix `M` for the mode `n`
:param n: mode
:param column: the `l`th coefficient is replaced with the vector `N`
"""
matrix = self.M(n).copy()
vector = self.N(n)
if column is not None:
matrix[:, column] = vector
return np.linalg.det(matrix)
[docs] def M(self, n: int) -> np.ndarray:
"""Matrix M of order `n`
:param n: order
"""
return self._list_matrix_M_n.item(n)
[docs] def N(self, n: int) -> np.ndarray:
"""Vector `N`"""
return self._list_vector_N_n.item(n)
# -------------------------------------------------------------------------
# Properties
# -------------------------------------------------------------------------
@property
def x_hat(self) -> complex:
"""Product of :attr:`~.k_s` and :attr:`~.R_0`
:math:`\\hat{x}=k_s R_0`
"""
return self._x_hat.value
def _compute_x_hat(self) -> complex:
return self.k_s * self.R_0
@property
def x_hat_v(self) -> complex:
"""Product of :attr:`~.k_vs` and :attr:`~.R_0`
:math:`\\hat{x}_v=\\hat{k}_s R_0`
"""
return self._x_hat_v.value
def _compute_x_hat_v(self) -> complex:
return self.k_vs * self.R_0
# -------------------------------------------------------------------------
# Vector N(n)
# -------------------------------------------------------------------------
def _compute_vector_N_n(self, n: int) -> np.ndarray:
vector = np.zeros(4, dtype=complex)
vector[0] = -self.x * Bf.d1_besselj(n, self.x)
vector[1] = -Bf.besselj(n, self.x)
vector[2] = 1j * self.rho_f * self.c_f**2 / self.omega
vector[2] += self.zeta_f
vector[2] -= 2 / 3 * self.eta_f
vector[2] *= Bf.besselj(n, self.x)
vector[2] -= 2 * self.eta_f * Bf.d2_besselj(n, self.x)
vector[2] *= self.x**2
vector[3] = Bf.besselj(n, self.x)
vector[3] -= self.x * Bf.d1_besselj(n, self.x)
vector[3] *= 2 * self.eta_f
return vector
# -----------------------------------------------------------------------
# Matrix M(n)
# -----------------------------------------------------------------------
def _compute_matrix_M_n(self, n: int) -> np.ndarray:
# Initialize matrix
matrix = np.zeros((4, 4), dtype=complex)
# First row
matrix[0, 0] = self.x * Bf.d1_hankelh1(n, self.x)
matrix[0, 1] = -n * (n + 1) * Bf.hankelh1(n, self.x_v)
matrix[0, 2] = - self.x_hat * Bf.d1_besselj(n, self.x_hat)
matrix[0, 3] = n * (n + 1) * Bf.besselj(n, self.x_hat_v)
# Second Row
matrix[1, 0] = Bf.hankelh1(n, self.x)
matrix[1, 1] = -Bf.hankelh1(n, self.x_v)
matrix[1, 1] -= self.x_v * Bf.d1_hankelh1(n, self.x_v)
matrix[1, 2] = -Bf.besselj(n, self.x_hat)
matrix[1, 3] = Bf.besselj(n, self.x_hat_v)
matrix[1, 3] += self.x_hat_v * Bf.d1_besselj(n, self.x_hat_v)
# Third row
matrix[2, 0] = 1j * self.rho_f * self.c_f**2 / self.omega
matrix[2, 0] += self.zeta_f
matrix[2, 0] -= 2 / 3 * self.eta_f
matrix[2, 0] *= -Bf.hankelh1(n, self.x)
matrix[2, 0] += 2 * self.eta_f * Bf.d2_hankelh1(n, self.x)
matrix[2, 0] *= self.x**2
matrix[2, 1] = Bf.hankelh1(n, self.x_v)
matrix[2, 1] -= self.x_v * Bf.d1_hankelh1(n, self.x_v)
matrix[2, 1] *= 2 * n * (n + 1) * self.eta_f
matrix[2, 2] = 1j * self.rho_s * self.c_s**2 / self.omega
matrix[2, 2] += self.zeta_s
matrix[2, 2] -= 2 / 3 * self.eta_s
matrix[2, 2] *= Bf.besselj(n, self.x_hat)
matrix[2, 2] -= 2 * self.eta_s * Bf.d2_besselj(n, self.x_hat)
matrix[2, 2] *= self.x_hat**2
matrix[2, 3] = self.x_hat_v * Bf.d1_besselj(n, self.x_hat_v)
matrix[2, 3] -= Bf.besselj(n, self.x_hat_v)
matrix[2, 3] *= 2 * n * (n + 1) * self.eta_s
# Fourth row
matrix[3, 0] = self.x * Bf.d1_hankelh1(n, self.x)
matrix[3, 0] -= Bf.hankelh1(n, self.x)
matrix[3, 0] *= 2 * self.eta_f
matrix[3, 1] = self.x_v ** 2 * Bf.d2_hankelh1(n, self.x_v)
matrix[3, 1] += (n**2 + n - 2) * Bf.hankelh1(n, self.x_v)
matrix[3, 1] *= -self.eta_f
matrix[3, 2] = Bf.besselj(n, self.x_hat)
matrix[3, 2] -= self.x_hat * Bf.d1_besselj(n, self.x_hat)
matrix[3, 2] *= 2 * self.eta_s
matrix[3, 3] = self.x_hat_v ** 2 * Bf.d2_besselj(n, self.x_hat_v)
matrix[3, 3] += (n**2 + n - 2) * Bf.besselj(n, self.x_hat_v)
matrix[3, 3] *= self.eta_s
return matrix
if __name__ == '__main__':
pass