Gaussian Beamlets Paper

[Jashcraf]

Issue tracking for "Hybrid Propagation Physics for The Design and Modeling of Astronomical Observatories Outfitted with Coronagraphs". Tentatively submitting to SPIE's JATIS

Figures

• [x] Illustration of Worku's GBD Algorithm
• [x] Gaussian Beamlets v.s. Fresnel, Zemax, CODE V, FRED

Sections

• [x] Write Conclusion Section
• [x] Write Worku's mathematics
• [x] Consider removing paraxial GBD?

Proof-Reading

• [x] Jaren - Introduction
• [x] Jaren - Methods
• [x] Jaren - Results
• [x] Jaren - Conclusion
• [x] Ewan
• [x] Daewook
• [x] A.J.
• [x] Trenton

[Jashcraf]

Try use the pickle package so that you can send raysets to the HPC and do GBD calculations on them.

[Jashcraf]

Tell the advising squad about this on Monday because I think it's a really good idea. We can compare PSF's through the MTF calculations.

The MTF is simple $MTF = |FFT[PSF]|$, but what it represents is pretty powerful. It's how well a given spatial frequency is transferred through an optical system, and it's on a linear scale. Plus, the solution for a circle is analytically known so the comparison should be really easy. Here's the difference of the MTF's for the gaussian beamlets

And here's the difference for Fresnel

The fractional difference is on the order of +/-1%. Here white is "well preserved" and blue/red is "not well preserved". The flower shape that shows up on the GBD one shows that the spatial frequencies along the distributed gaussian beamlets are "not well preserved", which are the axes of distribution of the gaussian beamlets where we expect the error to be!

[Jashcraf]

Also consider check in the Phase transfer function to make sure this is preserved as well

[Jashcraf]

Officially submitted to JATIS!