http://www.k-wave.org/forum/topic/steering-linear-array Support for the k-Wave MATLAB toolbox en-US Tue, 12 May 2026 23:45:48 +0000 http://bbpress.org/?v=1.0.2 q http://www.k-wave.org/forum/search.php http://www.k-wave.org/forum/topic/steering-linear-array#post-7082 Fri, 18 Oct 2019 19:06:02 +0000 Bradley Treeby 7082@http://www.k-wave.org/forum/ <p>Hi Farnaz,</p> <p>They look the same, except for a constant offset. So they're both correct, but one will be slightly delayed compared to the other. Which one you want depends on which element you define the offsets relative to.</p> <p>Brad. </p> http://www.k-wave.org/forum/topic/steering-linear-array#post-7027 Wed, 28 Aug 2019 15:48:20 +0000 Farnaz 7027@http://www.k-wave.org/forum/ <p>Hi, </p> <p>I was wondering which of the following approaches for computing the delay manually for a linear array in 2D is correct: </p> <p>Assume I am modeling a transducer in which each element is consist of 4 grid points and the kerf is also 4 grid points.</p> <p>In the Steering A Linear Array Example the offset for toneBurst is computed with this:</p> <p>element_spacing = dx; % [m]<br /> element_index = -(num_elements - 1)/2:(num_elements - 1)/2;<br /> tone_burst_offset = element_spacing * element_index * ...<br /> sin(steering_angle * pi/180) / (medium.sound_speed * kgrid.dt);</p> <p>There is also an equation in manual p.41, eq. 3.1 that can be used to compute delays. </p> <p>In the example there is no pitch and each element is only one grid. If I want to adopt this to my problem I'd do this: </p> <p>element_spacing = 4*dx; % [m]<br /> tone_burst_freq = 5e6; % [Hz]<br /> element_index = 1:2:active_elements*2 ;<br /> tone_burst_offset = element_spacing * element_index * ...<br /> sin(steering_angle * pi/180) / (medium.sound_speed * kgrid.dt);<br /> for i = 1:active_elements<br /> source.ux(count,:) = signal(i,:) ;<br /> source.ux(count+1,:) = signal(i,:) ;<br /> source.ux(count+2,:) = signal(i,:) ;<br /> source.ux(count+3,:) = signal(i,:) ;<br /> count = count+4 ;<br /> end</p> <p>element spacing 4*dx because of of the number of grid points per element and element_index = 1:2:active_elements*2 because there is a kerf that should be skipped. </p> <p>When I apply this, simulation looks fine (as far as I see at least). </p> <p>Now With the equation: </p> <p>function delay = delay_pw_me(index_num,pitch,angle,c0,dt)<br /> delay = ((index_num*pitch*sin(pi*angle/180))/(c0*dt)) ;<br /> end</p> <p>pitch = 4*dx + 4*dx ;<br /> for i = 1 : active_elements<br /> tone_burst_offset2(1,i) = delay_pw_me(i,pitch,steering_angle,c0,kgrid.dt) ;<br /> end</p> <p>for the whole element a delay is computed and then I assign the delay same way (each computed delay to 4 sources). The simulation looks fine with these delays as well. </p> <p>But there is a shift between these two sets of computed delays, the question is which method is correct? </p> <p>Thanks in advance,<br /> Farnaz </p>