PhD With Me 🔬

Keeping Organised (digitally)

PhD epoch 1 > Actual Organisation > CST2024 > September U-23-24

Contains simulations with a very specific naming convention

(Never delete or overwrite simulated data unless you know that it is required)

Starter_hail mary.cst was the template starting point.

The geometry matched that of Ming2017. But with the simplified difference of an 'off-centre' unit cell. Which shouldn't matter in the metasurface.

From that create: 1_VO2..... .cst
2_Al...... .cst
3_TiN..... .cst

and, inevitably,
22_Al....3.70.cst
222_Al...3.80.cst
fixing_22_Al....3.70.cst

From there I export the s11 data to a text file saved in:
PhD epoch 1 > Actual Project

That data (after remembering to delete the header) is now in the path of the .m file another_absorpance_converted.m

Because I always create duplicates of original files to tinker around with.

The naming convention for the text file to be used in the .m file is close to the cst name to not get confused (but it could be exact next time).
e.g. 2_Al_SiO2_375.txt

From that point I currently copy the generated image into a powerpoint slide show which is also embedded within that Fortnight's Dropbox paper document. (Also for OS agnostic viewing).

But the next step in the workflow I am planning to improve (and there is also a pre-workflow for parameter/materials selection that I've yet to discuss) is to numerically compare the exported data in various ways.

I'm interested in MSE values to compare how good the output design is when compared to the target one (in this case the input is a grey body spectrum from AlN-covered Silicon pv cell).
And I'm also interested in evaluating which design is better considering the sensitivities of lab or commercial fabrication.

In the coming days I'll make a cloud deployed object for uploading or pasting .txt or .csv data (appropriate folder management workflows also pending). \
And that will give me an analysis of the outputted absorpance spectrum: Q-values, Resonance peaks (maybe user stated), MSE, for now that would be fine.
But ultimately I would like to simulate the entire system: - heat source - absorption - TEG. And also think about ways that this heat flow might occur one way (porous aerogels?), or the alternative idea of reflecting back just a 980nm 'laser' pulse and to eliminate the need for the Thermo Electric Generator.