Learning Forum

[Odebowale]

I am working on NFTPV system energy conversion. I would like to ask if radiative heat transfer at nanoscale can be modelled in FDTD or DGTD or any of the lumerical simulation environment. I have searched online but can't find any example. 

[Guilin Sun]

FDTD and DGTD are optical solvers which need material data. You will need to simulate the device in HEAT, get the temperature distribution, and then import the data as an attribute, and give the material thermal coefficient. Then the optical solvers will consider the material data at each applied voltage or thermal power (which induce the refractive chage). Examples can be found online, such as
https://optics.ansys.com/hc/en-us/articles/360041686234-Photothermal-heating-in-plasmonic-nanostructures

https://optics.ansys.com/hc/en-us/articles/360034901773-Temperature-dependent-refractive-index-models

You also mentioned photocurrent. This will need to first simulate in optics to get the absorbed power, and then simulate in CHARGE to get current. Examples can be found onlie, such as https://optics.ansys.com/hc/en-us/articles/360042828913-Germanium-on-Silicon-lateral-photodetector

If both thermal and electric physics needs to be simulated at the same time, you can do a co-simulation.

The exact simulation workflow will depend on the original driving force and final results you desire. Please first try those examples. If you have questions for co-simulation please write a new post.

 

 

[Odebowale]

Thank you for your response. Allow me to provide a comprehensive explanation of my current work. 

I have been working on a project involving two key components: an emitter structure (metasurface) and a TPV (Thermophotovoltaic) cell structure. These structures were initially designed using the Finite-Difference Time-Domain (FDTD) method. Subsequently, I extracted the S-parameters for each of these structures using a script that I integrated into the near-field spectral heat flux calculation, as defined by equation 1. This allowed me to calculate the total heat flux by performing an integration. 

To further analyze the behavior of these structures, especially in the near-field regime (with a separation distance, d, of 10nm), I utilized the extracted S-parameters to calculate the photonic transmission coefficient, as defined by equation 2. This calculation ultimately led to the determination of the spectral flux, denoted as Q(w).

My next step in this project is to assess the impact of this spectral flux on the TPV cell. To better understand this process, I refer to an example provided here: (https://optics.ansys.com/hc/en-us/articles/360042165534). 

However, I have a specific question regarding the electrical calculations, particularly concerning the charge solver. In the provided example script, they used Psolar=100mW/m^2 as the input power and incorporated solar irradiance data in the optical solver to calculate Pabs (absorbed power) and the generation rate.

My inquiry is as follows: In my case, should I substitute the spectral flux and total flux obtained from equations 1 and 2 instead of using solar irradiance and Psolar? If this approach is appropriate, how can I incorporate the spectral flux into the solar generation script within the analysis group?

I hope this clarifies my query, and I appreciate your assistance with this matter.

Thanks 

[Guilin Sun]

This is a more theoretical question other than simulation question, isn't it?

The actual solar power depends on the location and the time of the day as well as other factors. The power spectra is from the "AM 1.5 Standard Spectrum". as for how much the actual power is, I believe  Psolar=100mW/m^2 is reasonable but just for illustration purpose. Please find some references to get the exact power for your design.

[Odebowale]

I know the exact power for my structure which I obtained from eq1 (both spectral power density and the total power density). In the example I provided, solar irradiation is used in the script in the analysis group to obtain absorbed power and generation rate which are imported into the charge solver for further analysis.
For my case, is there away I can import spectral flux I obtained into the analysis group instead of using solar irradiation?

Thanks

[Guilin Sun]

If the light source is not solar, sure you can use your own source. In addition, you can renormalize the source power from the result. You can also chage the scale in CHARGE to change the actual power used for CHARGE simulation after importing the absorption. If you use solar source, this is easier.