Hello everyone,
I tried to record the magnitude of shear stress using linear sensor mask to see how longitudinal wave can generate shear wave in a layered medium. But we failed to record the shear stress in sensors. How can I record the magnitude of shear stress as well as the pressure of longitudinal wave?
Thank you
Lenny
This is my code
clear all
close all
% create the computational grid
Nx = 100; % number of grid points in the x (row) direction
Ny = 200; % number of grid points in the y (column) direction
dx = 0.1e-3; % grid point spacing in the x direction [m]
dy = 0.1e-3; % grid point spacing in the y direction [m]
kgrid = kWaveGrid(Nx, dx, Ny, dy);
% define the properties of the upper layer of the propagation medium
medium.sound_speed_compression = 1500 * ones(Nx, Ny); % [m/s]
medium.sound_speed_shear = zeros(Nx, Ny); % [m/s]
medium.density = 1000 * ones(Nx, Ny); % [kg/m^3]
% define the properties of the lower layer of the propagation medium
medium.sound_speed_compression(Nx/2:end, :) = 2000; % [m/s]
medium.sound_speed_shear(Nx/2:end, :) = 800; % [m/s]
medium.density(Nx/2:end, :) = 1200; % [kg/m^3]
% create the time array
cfl = 0.05; % Courant-Friedrichs-Lewy number
t_end = 5e-6; % [s]
kgrid.makeTime(max(medium.sound_speed_compression(:)), cfl, t_end);
% define a source
source.s_mask = makeLine(Nx, Ny, [Nx/2,1],[1,Ny/2]);
%source.u_mask = makeLine(Nx, Ny, [Nx/2,1],[1,Ny/2]);
%create initial pressure distribution using makeDisc
disc_magnitude = 400; % [Pa]
source.p0 = disc_magnitude*source.s_mask ;
% define a time varying source
%source_freq = 1e6; % [Hz]
%source_mag = 0.00002;
%source.ux = source_mag * sin(2 * pi * source_freq * kgrid.t_array);
%source.uy = source_mag * sin(2 * pi * source_freq * kgrid.t_array);
%display_mask=source.u_mask;
sensor.mask=makeLine(Nx,Ny,[40,40],[60,60]);
% define input arguments
input_args = {'PlotScale', [-100, 100, -70, 70], 'PlotPML', false,...
'DisplayMask', sensor.mask, 'DataCast', 'single'};
% run the simulation
sensor_data = pstdElastic2D(kgrid, medium, source, sensor, input_args{:});
imagesc(sensor_data);
colormap(getColorMap);
colorbar;
Hi Lenny,
You can define which parameters you would like to record using the sensor.record input (see the documentation for details). Unfortunately, it's not currently possible to record the components of the stress (only the particle velocity or pressure). If you're comfortable with MATLAB, it is pretty easy to hack into the code (add it after line 814 in V1.2). It's on the list for a future release.
Brad.
Hi Brad,
Could you give a little bit more advice on the above hack? I'm modeling the typical NDT case of a longitudinal wave launched in an acrylic wedge, which gets converted to shear wave in a steel part. I'd like to extract the pressure (or velocity) from the shear wave in the steel, but there is also a longitudinal wave there (a couple, actually). So the "u" or "u_max" data would map both waves together. The shear stress, if it could be extracted and saved, would solve that problem.
Thanks,
Andy
Hi Andy,
For k-Wave V1.2, to hack it into the pstdElastic2D in the simplest way, immediately before the time loop (line 617) add:
sensor_data.sxy = castZeros([num_sensor_points, num_recorded_time_points]);
Then after line 814, add the following:
sensor_data.sxy(:, file_index) = sxy_split_x(sensor_mask_index) + sxy_split_y(sensor_mask_index);
Similarly for 3D. Note, I haven't tried this, so let me know if it doesn't work for some reason.
Brad.
Hi,
I had the same problem.
Thank you Brad for explanations, they work, but additionally in pstdElastic2D.m it need to add after 1017 line someting like:
SS.p=sensor_data.p;
SS.sxy=sensor_data.sxy;
sensor_data = SS;%sensor_data.p;
Thanks,
Sergiy
Hello,
Has the ability to record the components of shear stress been implemented in the current release? If so, how can this be done?
Thanks,
PC
As a follow-up. I modified the 3D elastic code to record the normal components of shear stress. Am I correct in using sxx = sxx_split_x + sxx_split_y + sxx_split_z? (and similarly for syy and szz and of course with the correct indexing)
Thanks again,
PC
Hi Patrick,
Unfortunately, it is still not in the current release. Yes, that looks correct.
Brad.
You must log in to post.