http://www.k-wave.org/forum/topic/why-the-simulation-with-different-z-step-in-3d-model-is-different Support for the k-Wave MATLAB toolbox en-US Wed, 13 May 2026 00:20:54 +0000 http://bbpress.org/?v=1.0.2 q http://www.k-wave.org/forum/search.php http://www.k-wave.org/forum/topic/why-the-simulation-with-different-z-step-in-3d-model-is-different#post-8570 Wed, 13 Jul 2022 17:05:00 +0000 bencox 8570@http://www.k-wave.org/forum/ <p>Hi Kevin, </p> <p>It's probably best to keep the grid spacings the same in all directions, ie. <code>dx = dy = dz</code>. If they are different then k-Wave will support a different range of frequencies in different directions, which might matter sometimes, eg. if the nonlinearity in your simulation generates harmonics, different directions may support different numbers of harmonics.</p> <p>Best wishes<br /> Ben </p> http://www.k-wave.org/forum/topic/why-the-simulation-with-different-z-step-in-3d-model-is-different#post-8548 Sun, 12 Jun 2022 14:08:09 +0000 kevin durant 8548@http://www.k-wave.org/forum/ <p>Dear author:</p> <p>I use the 'kspaceFirstOrder3D' function for 3D simulation, but the results are different when the different simulation step size in z-axis is set. I wonder if some parameters were wrong in the code. </p> <p>Here is the code. P0 is the input computed through Rayleigh-sommerfeld Integral of focused transducer at 20mm before the focus. The result is different with the <code>dz = 0.1e-3; Nz = 311</code>.</p> <pre><code>clear; clc; close all; % ========================================================================= % DEFINE THE K-WAVE GRID % ========================================================================= % initial condition [FileName, FileDir] = uigetfile(&#39;.mat&#39;, &#39;Initial Plane File&#39;); P0 = load(fullfile(FileDir, FileName)); P0 = cell2mat(struct2cell(P0)); % set total number of grid points not including the PML [Nx, Ny] = size(P0); % [grid points] Nz = 161; ; % [grid points] % calculate the spacing between the grid points dx = 0.2e-3; % [m] dy = dx; % [m] dz = 0.2e-3; % [m] % set desired grid size in the x-direction not including the PML x = Nx * dx; % [m] y = Ny * dy; z = Nz * dz; % [m] % create the k-space grid kgrid = kWaveGrid(Nx, dx, Ny, dy, Nz, dz); % ========================================================================= % DEFINE THE MEDIUM PARAMETERS % ========================================================================= sound_speed_Water = 1482; medium.sound_speed = sound_speed_Water * ones(Nx, Ny, Nz); % [m/s] medium.density = 994 * ones(Nx, Ny, Nz); % [kg/m^3] medium.alpha_coeff = 0 * ones(Nx, Ny, Nz); medium.alpha_power = 1; medium.alpha_mode = &#39;no_dispersion&#39;; medium.BonA = 5.2 * ones(Nx, Ny, Nz); pml_size = 10; % PML size pml_alpha = 2; % PML absorption coefficient [Np/grid point] % ========================================================================= % DEFINE THE INPUT SIGNAL % ========================================================================= % create the time array source_freq = 1.2e6; % [Hz] kgrid.makeTime(medium.sound_speed); dT = 1/source_freq; NumT = ceil(dT/kgrid.dt/10) * 10; T = kgrid.dt * kgrid.Nt; kgrid.dt = dT/NumT; Nt = ceil(T/kgrid.dt); kgrid.t_array = [0:1:Nt] * kgrid.dt; % define properties of the input signal source.p_mask = zeros(Nx, Ny, Nz); source.p_mask(:, :, 11) = 1; for i = 1:numel(P0) source.p(i,:) = abs(P0(i)) * sin(2*pi*source_freq*kgrid.t_array + angle(P0(i))); end % define a series of Cartesian points to collect the data sensor.mask = zeros(Nx, Ny, Nz); sensor.mask(ceil(Nx/2), :, :) = 1; sensor.record_start_index = ceil(numel(kgrid.t_array)/2); % define the field parameters to record sensor.record = {&#39;p&#39;, &#39;p_max_all&#39;,&#39;p_final&#39;}; input_args = {&#39;DisplayMask&#39;, sensor.mask, &#39;DataCast&#39;, &#39;single&#39;, ... &#39;CartInterp&#39;, &#39;nearest&#39;, &#39;PMLSize&#39;, [10 10 10], &#39;PMLAlpha&#39;, pml_alpha, &#39;PMLInside&#39;, false}; % run the simulation sensor_data = kspaceFirstOrder3D(kgrid, medium, source, sensor, input_args{:});</code></pre> <p>Best wishes<br /> kevin </p>