k-Wave

1. huangchao
   

   Member
   Joined: Nov '10
   Posts: 57

   Hi Dr. Cox and Dr. Treeby,

   I did some PAT simulations with strong absorption media, and the results blew up. Here are my codes:

   Nx=256;
   dx=1e-3;
   kgrid=makeGrid(Nx,dx,Nx,dx);

   medium.sound_speed=1500;
   medium.density=1000;
   medium.alpha_coeff=1;
   medium.alpha_power=3;

   source.p0=makeDisc(Nx,Nx,Nx/2,Nx/2,10);

   sensor.mask=makeCartCircle(50e-3,100);

   input_args = {'CartInterp','linear', 'PMLInside', true, 'PMLSize', 20, 'DataCast', 'gsingle', 'PlotPML', false, 'PlotSim', false};

   p=kspaceFirstOrder2D(kgrid,medium,source,sensor,input_args{:});

   However, if I set medium.alpha_power=2 and medium.alpha_coeff=1, the results were reasonable. But if I increased the medium.alpha_coeff to 1000 (which is not impractical though), the results would blow up again.

   So I wonder if the parameters alpha_power and alpha_coeff are restricted to some ranges for stable simulations.

   Thanks,
   Chao

   Posted 12 years ago #

2. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,643

   Hi Chao,

   The value for y must be defined as 0 < y < 3 with y ~= 1. If you change y to 2.99 in your example, it should run without problems. We will update the error checking to give an error for the next release.

   In the case of extremely high absorption values, it's likely due to numerical rounding errors. There are no set guidelines I'm afraid, but it shouldn't be a problem for physically relevant absorption values. Incidentally, the implementation of the absorption term in kspaceSecondOrder is more accurate than in the first order codes, so this will be more stable for very high absorption values.

   Hope that helps,

   Brad.

   Posted 12 years ago #

RSS feed for this topic

Reply

You must log in to post.