An open-source gaussian beamlet decomposition tool for modeling astronomical telescopes

Abstract

In the pursuit of directly imaging exoplanets, the high-contrast imaging community has developed a multitude of tools to simulate the performance of coronagraphs on segmented-aperture telescopes. As the scale of the telescope increases and science cases move toward shorter wavelengths, the required physical optics propagation to optimize high-contrast imaging instruments becomes computationally prohibitive. Gaussian Beamlet Decomposition (GBD) is an alternative method of physical optics propagation that decomposes an arbitrary wavefront into paraxial rays. These rays can be propagated expeditiously using ABCD matrices, and converted into their corresponding Gaussian beamlets to accurately model physical optics phenomena without the need of diffraction integrals. The GBD technique has seen recent development and implementation in commercial software (e.g. FRED, CODE V, ASAP) 1-3 but appears to lack an open-source platform. We present a new GBD tool developed in Python to model physical optics phenomena, with the goal of alleviating the computational burden for modeling complex apertures, many-element systems, and introducing the capacity to model misalignment errors. This study demonstrates the synergy of the geometrical and physical regimes of optics utilized by the GBD technique, and is motivated by the need for advancing open-source physical optics propagators for segmentedaperture telescope coronagraph design and analysis. This work illustrates GBD with Poisson's spot calculations and show significant runtime advantage of GBD over Fresnel propagators for many-element systems.

Hybrid propagation physics for the design and modeling of astronomical observatories: a coronagraphic example

“…To account for system misalignments, Q(z)−1 is a 2×2 matrix Q(z)−1 that encodes the complex curvature in two orthogonal directions and how they couple into each other, allowing for the beamlet to be generally astigmatic: 53 , 58 Q(z)−1=(Qfalse(zfalse)xx−1Qfalse(zfalse)xy−1Qfalse(zfalse)yx−1Qfalse(zfalse)yy−1).…”

“…To account for system misalignments, QðzÞ −1 is a 2 × 2 matrix QðzÞ −1 that encodes the complex curvature in two orthogonal directions and how they couple into each other, allowing for the beamlet to be generally astigmatic: 53,58 E Q -T A R G E T ; t e m p : i n t r a l i n k -; e 0 0 6 ; 1 1 7 ; 1 6 8…”

“…The ray transfer matrices for simple and paraxial optical elements (e.g., thin lens) are well known, 59 and were used in concert with GBD in a prior investigation with paraxial systems. 53 However, our aim is to use GBD to model non-paraxial optical system diffraction. Therefore, we need a method of computing the ray transfer matrix for an arbitrary skew ray path.…”

“…The decomposition of the field into Gaussian beams does not have an analytical solution. Therefore, undesirable artifacts can be introduced into the field if the decomposition is not well-understood and the sampling is insufficient 53 . To better understand the impact of a GBD PSF on high-contrast imaging simulations, we develop a hybrid propagation model to compare GBD to an equivalent Fraunhofer diffraction model.…”

Characterizing the instrumental polarization of SCExAO VAMPIRES

“…In order to determine the theoretical Mueller matrix of the system (with best guesses on the materials of some unknown coatings), polarization aberrations were simulated using the Poke polarization ray tracing package which has support for modeling the complex amplitude coefficients of multi-layer thin film stacks. 21 From the telescope mirrors to the end of the VAMPIRES optical train, the coatings assumed for all elements are as follows: aluminum for M1, silver for M2 and M3, protected silver for the first deformable mirror (DM), and protected silver on all OAPs and mirrors. Using 64 x 64 points across the pupil and λ = 750nm, the variation of each Mueller matrix model component is shown in Fig.…”