# RCAIDE/Methods/Noise/Frequency_Domain_Buildup/Rotor/broadband_noise.py
#
#
# Created: Sep 2024, Niranjan Nanjappa
# ----------------------------------------------------------------------------------------------------------------------
# IMPORT
# ----------------------------------------------------------------------------------------------------------------------
# RCAIDE
from RCAIDE.Framework.Core import Units
from RCAIDE.Library.Methods.Noise.Frequency_Domain_Buildup.Rotor.BPM_boundary_layer_properties import BPM_boundary_layer_properties
from RCAIDE.Library.Methods.Noise.Frequency_Domain_Buildup.Rotor.LBL_VS_broadband_noise import LBL_VS_broadband_noise
from RCAIDE.Library.Methods.Noise.Frequency_Domain_Buildup.Rotor.TBL_TE_broadband_noise import TBL_TE_broadband_noise
from RCAIDE.Library.Methods.Noise.Frequency_Domain_Buildup.Rotor.TIP_broadband_noise import TIP_broadband_noise
from RCAIDE.Library.Methods.Noise.Frequency_Domain_Buildup.Rotor.noise_directivities import noise_directivities
# Python Package imports
import numpy as np
# ----------------------------------------------------------------------------------------------------------------------
# Compute Broadband Noise
# ----------------------------------------------------------------------------------------------------------------------
[docs]
def broadband_noise(conditions,coordinates,rotor,settings,Noise,cpt):
'''This computes the trailing edge noise compoment of broadband noise of a propeller or
lift-rotor in the frequency domain. Boundary layer properties are computed using RCAIDE's
panel method.
Assumptions:
Boundary layer thickness (delta) appear to be an order of magnitude off at the trailing edge and
correction factor of 0.1 is used. See lines 255 and 256
Source:
Li, Sicheng Kevin, and Seongkyu Lee. "Prediction of Urban Air Mobility Multirotor VTOL Broadband Noise
Using UCD-QuietFly." Journal of the American Helicopter Society (2021).
Inputs:
freestream - freestream data structure [m/s]
angle_of_attack - aircraft angle of attack [rad]
bspv - rotor blade section trailing position edge vectors [m]
velocity_vector - velocity vector of aircraft [m/s]
rotors - data structure of rotors [None]
aeroacoustic_data - data structure of acoustic data [None]
settings - accoustic settings [None]
res - results data structure [None]
Outputs
res. *acoustic data is stored and passed in data structures*
SPL_prop_broadband_spectrum - broadband noise in blade passing frequency spectrum [dB]
SPL_prop_broadband_spectrum_dBA - dBA-Weighted broadband noise in blade passing frequency spectrum [dbA]
SPL_prop_broadband_1_3_spectrum - broadband noise in 1/3 octave spectrum [dB]
SPL_prop_broadband_1_3_spectrum_dBA - dBA-Weighted broadband noise in 1/3 octave spectrum [dBA]
p_pref_azimuthal_broadband - azimuthal varying pressure ratio of broadband noise [Unitless]
p_pref_azimuthal_broadband_dBA - azimuthal varying pressure ratio of dBA-weighted broadband noise [Unitless]
SPL_prop_azimuthal_broadband_spectrum - azimuthal varying broadband noise in blade passing frequency spectrum [dB]
SPL_prop_azimuthal_broadband_spectrum_dBA - azimuthal varying dBA-Weighted broadband noise in blade passing frequency spectrum [dbA]
Properties Used:
N/A
'''
aeroacoustic_data = conditions.energy.converters[rotor.tag]
num_cpt = 1
num_mic = len(coordinates.X[0,:,0,0,0])
num_blades = len(coordinates.X[0,0,:,0,0])
num_sec = len(coordinates.X[0,0,0,:,0])
frequency = settings.center_frequencies
num_cf = len(frequency)
num_az = aeroacoustic_data.number_azimuthal_stations
# ----------------------------------------------------------------------------------
# Trailing Edge Noise
# ----------------------------------------------------------------------------------
freestream = conditions.freestream
speed_of_sound = freestream.speed_of_sound # speed of sound
density = freestream.density # air density
dyna_visc = freestream.dynamic_viscosity
alpha = aeroacoustic_data.disc_effective_angle_of_attack/Units.degrees
alpha_tip = aeroacoustic_data.disc_effective_angle_of_attack[:,-1]/Units.degrees
disc_Re = aeroacoustic_data.disc_reynolds_number
disc_speed = aeroacoustic_data.disc_velocity
disc_Ma = aeroacoustic_data.disc_Mach_number
Vt = aeroacoustic_data.disc_tangential_velocity
Va = aeroacoustic_data.disc_axial_velocity
blade_chords = rotor.chord_distribution # blade chord
r = rotor.radius_distribution # radial location
Omega = aeroacoustic_data.omega # angular velocity
L = np.zeros_like(r)
del_r = r[1:] - r[:-1]
L[0] = del_r[0]
L[-1] = del_r[-1]
L[1:-1] = (del_r[:-1]+ del_r[1:])/2
if np.all(Omega == 0):
Noise.p_pref_broadband = np.zeros((num_cpt,num_mic,num_cf))
Noise.SPL_prop_broadband_spectrum = np.zeros_like(Noise.p_pref_broadband)
Noise.SPL_prop_broadband_spectrum_dBA = np.zeros_like(Noise.p_pref_broadband)
Noise.SPL_prop_broadband_1_3_spectrum = np.zeros((num_cpt,num_mic,num_cf))
Noise.SPL_prop_broadband_1_3_spectrum_dBA = np.zeros((num_cpt,num_mic,num_cf))
Noise.p_pref_azimuthal_broadband = np.zeros((num_cpt,num_mic,num_cf))
Noise.p_pref_azimuthal_broadband_dBA = np.zeros_like(Noise.p_pref_azimuthal_broadband)
Noise.SPL_prop_azimuthal_broadband_spectrum = np.zeros_like(Noise.p_pref_azimuthal_broadband)
Noise.SPL_prop_azimuthal_broadband_spectrum_dBA = np.zeros_like(Noise.p_pref_azimuthal_broadband)
else:
# dimension of matrices [control pt, microphone, # blades, # blade sections, # center frequencies, # azimuthal stations]
c = np.tile(blade_chords[None,None,None,:,None,None],(num_cpt,num_mic,num_blades,1,num_cf,num_az))
L = np.tile(L[None,None,None,:,None,None],(num_cpt,num_mic,num_blades,1,num_cf,num_az))
f = np.tile(frequency[None,None,None,None,:,None],(num_cpt,num_mic,num_blades,num_sec,1,num_az))
alpha_disk = np.tile(alpha[cpt,None,None,:,None,:],(1,num_mic,num_blades,1,num_cf,1))
V = np.zeros((num_cpt,num_mic,num_blades,num_sec,num_cf,num_az,3))
V[:,:,:,:,:,:,0] = -np.tile(Vt[cpt,None,None,:,None,:],(1,num_mic,num_blades,1,num_cf,1))
V[:,:,:,:,:,:,2] = np.tile(Va[cpt,None,None,:,None,:],(1,num_mic,num_blades,1,num_cf,1))
V_tot = np.linalg.norm(V, axis=6)
alpha_tip = np.tile(alpha_tip[cpt,None,None,None,None,:],(1,num_mic,num_blades,num_sec,num_cf,1))
c_0 = np.tile(speed_of_sound[cpt,:,None,None,None,None],(1,num_mic,num_blades,num_sec,num_cf,num_az))
rho = np.tile(density[cpt,:,None,None,None,None],(1,num_mic,num_blades,num_sec,num_cf,num_az))
mu = np.tile(dyna_visc[cpt,:,None,None,None,None],(1,num_mic,num_blades,num_sec,num_cf,num_az))
R_c = np.tile(disc_Re[cpt,None,None,:,None,:],(1,num_mic,num_blades,1,num_cf,1))
U = np.tile(disc_speed[cpt,None,None,:,None,:],(1,num_mic,num_blades,1,num_cf,1))
M = np.tile(disc_Ma[cpt,None,None,:,None,:],(1,num_mic,num_blades,1,num_cf,1)) # U/c_0
M_tot = V_tot/c_0
X_prime_r = np.tile(coordinates.X_prime_r[cpt,:,:,:,None,None,:],(1,1,1,1,num_cf,num_az,1))
cos_zeta_r = np.sum(X_prime_r*V, axis = 6)/(np.linalg.norm(X_prime_r, axis = 6)*V_tot)
r_er = np.tile(np.linalg.norm(coordinates.X_e_r[cpt], axis = 3)[None,:,:,:,None,None],(1,1,1,1,num_cf,num_az))
Phi_er = np.tile(coordinates.phi_e_r[cpt,:,:,:,None,None],(1,1,1,1,num_cf,num_az))
Theta_er = np.tile(coordinates.theta_e_r[cpt,:,:,:,None,None],(1,1,1,1,num_cf,num_az))
# flatten matrices
R_c = flatten_matrix(R_c,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
c = flatten_matrix(c,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
alpha_star = flatten_matrix(alpha_disk,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
alpha_tip = flatten_matrix(alpha_tip,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
U = flatten_matrix(U,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
cos_zeta_r = flatten_matrix(cos_zeta_r,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
M_tot = flatten_matrix(M_tot,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
f = flatten_matrix(f,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
c_0 = flatten_matrix(c_0,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
rho = flatten_matrix(rho,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
r_er = flatten_matrix(r_er,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
mu = flatten_matrix(mu,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
L = flatten_matrix(L,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
M = flatten_matrix(M,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
Phi_er = flatten_matrix(Phi_er,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
Theta_er = flatten_matrix(Theta_er,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
# calculation of boundary layer properties, eqns 2 - 16
# boundary layer properies of tripped and untripped at 0 angle of attack
boundary_layer_data = BPM_boundary_layer_properties(R_c,c,alpha_star)
# define simulation variables/constants
Re_delta_star_p_untripped = boundary_layer_data.delta_star_p_untripped*U*rho/mu
Re_delta_star_p_tripped = boundary_layer_data.delta_star_p_tripped*U*rho/mu
# calculation of directivitiy terms , eqns 24 - 50
Dbar_h, Dbar_l = noise_directivities(Theta_er,Phi_er,cos_zeta_r,M_tot)
# calculation of turbulent boundary layer - trailing edge noise, eqns 24 - 50
SPL_TBL_TE_tripped = TBL_TE_broadband_noise(f,r_er,L,U,M,R_c,Dbar_h,Dbar_l,Re_delta_star_p_tripped,
boundary_layer_data.delta_star_p_tripped,
boundary_layer_data.delta_star_s_tripped,
alpha_star)
SPL_TBL_TE_untripped = TBL_TE_broadband_noise(f,r_er,L,U,M,R_c,Dbar_h,Dbar_l,Re_delta_star_p_untripped,
boundary_layer_data.delta_star_p_untripped,
boundary_layer_data.delta_star_s_untripped,
alpha_star)
# calculation of laminar boundary layer - vortex shedding, eqns 53 - 60
SPL_LBL_VS = LBL_VS_broadband_noise(R_c,alpha_star,boundary_layer_data.delta_star_p_untripped,r_er,L,M,Dbar_h,f,U)
# calculation of tip vortex noise, eqns 61 - 67
alpha_TIP = abs(alpha_tip)
SPL_TIP = TIP_broadband_noise(alpha_TIP,M,c,c_0,f,Dbar_h,r_er)
# TO DO : Compute BWI
# TO DO : Compute BVI - isnt this actually tonal ?
# Unflatten Matices
SPL_TBL_TE_tripped = unflatten_matrix(SPL_TBL_TE_tripped,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
SPL_TBL_TE_untripped = unflatten_matrix(SPL_TBL_TE_untripped,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
SPL_LBL_VS = unflatten_matrix(SPL_LBL_VS,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
SPL_TIP = unflatten_matrix(SPL_TIP,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
# Pressure from each Broadband source
P_BWI = np.zeros((num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)) # this will be replaced soon
P_TBL_TE_tripped = 10**(SPL_TBL_TE_tripped/10)
P_TBL_TE_untripped = 10**(SPL_TBL_TE_untripped/10)
P_LBL_VS = 10**(SPL_LBL_VS/10)
P_TIP = 10**(SPL_TIP/10)
P_TIP[:,:,:,:-1,:,:] = 0
# Sum broadband compoments along blade sections and blades to get self noise per rotor
P_b_7 = np.zeros((4,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az))
P_b_7[0,:,:,:,:,:,:] = P_BWI
P_b_7[1,:,:,:,:,:,:] = P_TBL_TE_tripped
P_b_7[2,:,:,:,:,:,:] = P_LBL_VS
P_b_7[3,:,:,:,:,:,:] = P_TIP
# Sum all components of broadband noise pressures
P_b_6 = np.sum(P_b_7, axis=0)
# Take product of total broadband noise with Doppler shift factor and weighting factor
M_tot = unflatten_matrix(M_tot,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
cos_zeta_r = unflatten_matrix(cos_zeta_r,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az)
Doppler_shift = 1/(1-M_tot*cos_zeta_r)
P_b_6_shifted = (1/num_az)*(1/Doppler_shift)*P_b_6
# Sum broadband noise along blade sections
P_b_5 = np.sum(P_b_6_shifted, axis=3)
# Sum broadband noise for all blades
P_b_4 = np.sum(P_b_5, axis=2)
# Sum broadband noise along all azimuthal stations
P_b_3 = np.sum(P_b_4, axis=3)
SPL_broadband = 10*np.log10(P_b_3)
# store results
Noise.SPL_prop_broadband_spectrum = SPL_broadband
Noise.SPL_prop_broadband_1_3_spectrum = Noise.SPL_prop_broadband_spectrum # already in 1/3 octave spectrum
return
[docs]
def flatten_matrix(x,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az):
return np.reshape(x,(num_cpt*num_mic*num_blades*num_sec*num_cf*num_az))
[docs]
def unflatten_matrix(x,num_cpt,num_mic,num_blades,num_sec,num_cf,num_az):
return np.reshape(x,(num_cpt,num_mic,num_blades,num_sec,num_cf,num_az))
