I need to verify the results of the classical PEC or İmpedance half plane results obtained by the theory of Physical Optics by using HFSS or CST.
I attached the figures of the geometry and the total wave result for the incident wave angle of 60°, the total wave (=incident wave+geometrical optical wave+diffracted wave) and
diffracted wave in MATLAB.
How can I verify it in HFSS or CST or any suggestion?
This doesn't look like the theory of Physical Optics but the theory of Geometrical Optics and the Geometrical Theory of Diffraction. I don't think Physical Optics uses rays.
Is it GO and GTD that you are trying to check?
It is total field that includes all fields Incident+GO+diffracted.. PO directly gives total field with adding the incident field.. I have Total field as in figure, I am worried about how it is verified in HFSS or CST.
I presume you are plotting the modulus of the field amplitude, not the intensity, and it is in the near-field at a fixed distance from the edge.
In this case CST should be able to do this for you, and you will not need to model a semi-infinite plane, just a large one, or one with a curved edge at the other end. I expect HFSS will also do it.
In the figure, the field's absolute intensity is plotted with the observation point of rho=4*lambda distance. I made several studies with both HFSS and CST, the results only occur like diffracted field only. I added some figures related to HFSS and CST trials for lambda=0.1m and the observation point of the radiation box radius is 4*lambda=400mm.
What field source did you use? Was it plane wave illumination?
About intensity and amplitude.
None of the plots you have shown so far are intensity. Intensity is proportional to power. The three plots above are proportional to electric field (amplitude), which is proportional to the square root of power or intensity. The first plot you gave has the GTD field equal to half the step in the GO field. This is the relationship for amplitude or electric field (or magnetic field) not for power or intensity.
You can't see the step on the GO field because you only plot the total, which is the result of adding the GTD field to the GO field on the shadow side of the boundary, and subtracting it on the illuminated side of the boundary. For these to give a smooth field shape the GTD field amplitude has to be half the GO step, not half the GO power.
The three plots are not the GTD diffraction. The GTD diffraction has 180 degree phase change at the shadow and reflection boundary (120 deg and 240 deg), and also has a sign change in the slope at those places. It always comes to a point at the boundary. The CST time domain or frequency domain solvers will never give the GTD result. Similarly for HFSS.
Which polarization did you use? Was it parallel to the edge, or perpendicular to the edge?
In HFSS and CST, I tried many polarizations, but in the final trials, I always used the polarization is parallel to the surface or edge. I added the geometry figures from diffrent side views.
Electric field parallel to the edge will give the strongest diffracted field (the largest difference from the PO solution).
Did you use a plane wave?
How many wavelengths long and wide is your plate?
Yes, it is PW, I also tried other wave types such as cylindrical, dipole, Gaussian beam, even current source etc. However, results generally were similar. I get 50*lambda wide and long the surface. For radiation box, I set a sphere to the center of coordinate system with the radius distance of 4*lambda to measure radiation in every angle. I also tried them with Cube Radiation Box, results didn't change much. I also tried with PML, even the results didn't change much again.
Are you plotting the fields at a particular distance from the edge, or are these normalised far-field results? They look to me like the far-field scattering from the 50 wavelength square plate. I assume the plate is not 50 wavelengths square at 1, 2, and 3 GHz, but only at one of them, perhaps 3 GHz.
It is not possible to plot the far-field of a plane wave - it is a delta-function, so what you have is the polar diagram of the reflected wave, centred about 120 degrees, the polar diagram of the shadow the plate makes in the plane wave (see Babinet's principle), centred about 240 degrees, and low-level diffraction between these angles, and outside them too, but not so much.
I think your initial GO and GTD result (not PO) that you are trying to verify is in the near-field. The near-field for an infinitely long edge on a semi-infinite half-plane extends out to infinity.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Waves