Hi;
I am trying to simulate acoustic shear wave propagation using k-wave pstdElastic2D function.
In most soft tissue, the shear wave speed is much lower than compressional wave speed.
In other words, the shear wave speed is often below 10 m/s compared to compressional wave speed, which is around 1500 m/s.
As a result, my k-grid.t_array becomes really large.
Is there a way to optimize the time steps to view shear wave propagation at longer times?
Hi Sama,
If you want to run a simulation long enough to capture both the compressional and shear waves running from one side of the grid to the other, unfortunately there is no way around this (the same is true if you were conducting an experiment). The time step is defined to meet stability constraints based on the maximum speed.
One thought: if you are only interested in the shear wave and the medium is homogeneous, you could artificially set the compressional wave speed to be much lower, thereby increasing the required dt. In a homogeneous medium, there is no coupling between the two wave types.
Brad.
Hi Bradley;
Thank you for the hint.
It is really helpful.
How does this issue translate into real world experiment?
Regarding the simulation, if I am just interested in the shear wave in a heterogeneous media, it is not going to work without a long and fine time array?
Could I run the code for a fine temporal grid and a short time to capture the compressional wave and then use the sensor_data pressure signal at the last temporal point as a source for a coarser temporal grid over longer time? Does it make sense?
Hi Sama,
In the real world, if you wanted to record the propagation of both compressional and shear waves, you would need to record for long enough to capture the slower (in your case, much slower) propagating shear wave.
If you have a heterogeneous medium, then there will be conversion between the compressional and shear waves at the interfaces. In this case, you will need to run the simulation for the fine temporal grid and longer time. Recording and re-transmitting the compressional wave won't help, as it will still contain the same frequency content and propagate at the same speed.
Unfortunately, unless you can decouple the two wave types, it's not an easy problem to solve!
Brad.
You must log in to post.