Learning Forum

[Loic Feukam Takoundjou]

I want to calculate the losses of a curved waveguide.
After defining the source, I have defined the monitors to give the transmission coefficient at different sections of the curved structure.
The power I get from the monitor placed at the end of the bend is negative and is of the order of e-41 and also has an imaginary part (see screenshot).
Is this a programming error? 
How can I deduce the losses from the complex powers I get at the end?

Thank you in advance.

[Guilin Sun]

Somehow my reply was not posted yesterday.

There are a few things:

1: negative power means the power flow is toward the negative axis.

2: the data you viewed is the raw power data without normalization.

3: Power is usually normalized to sourcepower transmission - Script command

4:For such bend waveguide, it is better to use FDE in MODE to get the loss contant:

Solving bent waveguides in FDE and FEEM

Bent waveguide analysis

5: you can also use the mode source to get the bending loss: choose "user select" and calculate the mode

 

Choose PML boundary conditions:

set the bending radius:

specify the wavelength or frequency, then "calculate modes":

Please try.

For the half circule bending, we do not recommend to simulate it using FDTD. If you really want, please use mode expansion monitor, or better port to get the modal S paramter.

 

Please write a new post if you have other questions.

 

 

 

[Loic Feukam Takoundjou]

Thank you for the informations, I have followed your instructions for calculating losses

[Ahmad Murad]

I'm experiencing a similar issue, so I attempted to follow your instructions for calculating losses. However, I'm still encountering negative values for losses, specifically when analyzing different modes at varying radii. My design involves a bent waveguide, and I'm trying to create a graph that depicts the loss as a function of the bending radius. I would greatly appreciate it if you could explain in detail the steps to achieve this plot

[Guilin Sun]

Sall negative values are due to numerical error. usually neff can be accurate up to 1e-5. Higher accuracy needs careful settings and testing.

Please write a new post as this post will not be monitored from today.