Derive the Impedances Numerically
Posted: Thu Feb 19, 2026 7:14 am
Dear STKO team,
I hope you are doing well.
I would like to kindly request your assistance regarding the following numerical example.
I am trying to learn how to derive impedance functions through a numerical model, so I started with a very simple case: a stiff, massless foundation bonded to a homogeneous half-space soil, where the solution is already available in the literature Gazetas.
I started with the vertical case. I modeled the soil with E = 2e+8, v = 0.25, and rho = 1800 kg/m³. I placed absorbing boundaries at the corners with G = 8e+8, v = 0.25, and rho = 1800 kg/m³. The foundation was modeled with E = 2e+19 and v = 0.4.
To excite a wide range of frequencies, I applied a Ricker wavelet. The procedure I followed to derive the impedance functions is as follows:
1. I extracted the displacement at the foundation node (basically the input) and performed a Fourier transform.
2. I summed all the reactions at the contact nodes between soil and foundation and then performed a Fourier transform.
3. I divided the FFT of the reaction forces by the FFT of the displacement.
4. I normalized the result with respect to (rho * vs**2 / b).
the results of the first and second points are like follow:
then deriving the two parts of the impedance function real and imaginary :
and compare it Gazetas formula to the proposed solution , the results are like follow:
I am not sure why, in my simulation, the static stiffness appears to be low ( Re(K) = 1.857862e+05 N/m) compared to the analytical results ( Re(K) = 1.4e+09 N/m).
Here is my model:
Looking forward to your answers.
I hope you are doing well.
I would like to kindly request your assistance regarding the following numerical example.
I am trying to learn how to derive impedance functions through a numerical model, so I started with a very simple case: a stiff, massless foundation bonded to a homogeneous half-space soil, where the solution is already available in the literature Gazetas.
I started with the vertical case. I modeled the soil with E = 2e+8, v = 0.25, and rho = 1800 kg/m³. I placed absorbing boundaries at the corners with G = 8e+8, v = 0.25, and rho = 1800 kg/m³. The foundation was modeled with E = 2e+19 and v = 0.4.
To excite a wide range of frequencies, I applied a Ricker wavelet. The procedure I followed to derive the impedance functions is as follows:
1. I extracted the displacement at the foundation node (basically the input) and performed a Fourier transform.
2. I summed all the reactions at the contact nodes between soil and foundation and then performed a Fourier transform.
3. I divided the FFT of the reaction forces by the FFT of the displacement.
4. I normalized the result with respect to (rho * vs**2 / b).
the results of the first and second points are like follow:
then deriving the two parts of the impedance function real and imaginary :
and compare it Gazetas formula to the proposed solution , the results are like follow:
I am not sure why, in my simulation, the static stiffness appears to be low ( Re(K) = 1.857862e+05 N/m) compared to the analytical results ( Re(K) = 1.4e+09 N/m).
Here is my model:
Looking forward to your answers.