Learning Forum

[marius.crouzier]

Hello,

I am doing FDTD simulation to study the phase shift of a plane wave on an resonator and see how it can change according the temperature of my material.

I have a plane wave source above my resonator. I put a monitor above this source (to collect the reflected light), I extract Ex from it and use unwrap(angle(..)) to get the phase.

I have experimental data of my material for different temperature that I change during a sweep.

I start by studying the phase shift according to different temperature as you can see on the next graph showing the phase according the wavelength for different material's temperature. 

When I got this graph, I thought that if I was doing a simulation with the same design but with a source with only one wavelength at 770 nm and changing the material's temperature, I should have a large phase shift of several radians. But, as you can see on the next graph showing the phase calculate at different temperature at 770 nm (Fig2), it didn’t happen. I also used unwrap(angle(…)).

So, my questions are:

How is the phase calculated ?

Why is there this discrepancy for the phase calculation between the two simulations ?

Thanks,

Marius

[Guilin Sun]

"How is the phase calculated ?" I guess you know how the phase is calcualted right? it uses "angle" script, similar to atan(Imag/real).

"Why is there this discrepancy for the phase calculation between the two simulations ?" THere are two main reasons:

1: Phase is a relative quantity. so its beginning reference point is important. When you got your first plot, it uses the phase at the lowest frequency as reference point. In the 2nd case, it uses the zero T as reference.

This is a frequenctly asked question. We have to understand the unique property of the phase: relative. anything larger than abs(pi) will be discarded.

2: accuracy change. when you change the wavelength from broadband to single, the mesh and the material will change. So you should use the same mesh and the material fitting for the single wavelength case, by ticking this

 

 

You may try to extract the the phase at the same exact wavelength as the single wavelength from the broadband simulation, and compare the case when both the mesh and the material fitting use the same settings from the broadband for the single wavelength simulation.