Monte Carlo simulations

PAOS is designed to easily accommodate customized simulations such as Monte Carlo runs to test the performance of an optical system with varying parameters. This is particularly useful as it overcomes one major drawback from using commercial propagation software such as Zemax OpticStudio, which requires preliminary knowledge of Zemax Programming Language (ZPL). Instead, PAOS can be used out of the box as a standard Python library, interleaved with user-written code to suit a specific simulation. Moreover, PAOS’s routines can be easily run in parallel by leveraging standard Python libraries such as joblib and tqdm, for computational efficiency.

Multi-wavelength simulations

PAOS uses the method parse_config to parse the .ini configuration file and return a list of optical chains, where each list entry is a dictionary of the optical surfaces in the file, estimated at the given wavelength.

This output can be readily used to run POP simulations at each different wavelength, to test that the system properties and optical performance are always compliant to specification. For instance, wavelength-dependent total throughput for systems with optical diaphragms and variations in effective focal ratio for systems with diffractive elements.

Wavefront error simulations

PAOS can be used to evaluate the performance of an optical system for a given number of wavefront error realizations, to test the compatibility of the aberrated PSFs with some performance requirement. For instance, PAOS provides an ensemble of wavefront error realizations (see Fig. 26) that are compatible or nearly compatible with the encircled energy (EE) requirement at the Ariel telescope exit pupil.

../../_images/wfe_realizations.png — Fig. 26 Wfe realizations table

The recommended way to access this dataset is using the astropy method ascii as in the following code example.

import os
from astropy.io import ascii

wfe_file = os.path.join('path/to/wfe_file.csv')
wfe = ascii.read(wfe_file)

The whole set provides an effective way to test subsystems optical performances ahead of a measurement of the surface deviation of the Ariel telescope assembly (TA).

For example, it has been used to derive the rEE (radius of encircled energy) requirement for the Ariel Optical Ground Support Equipment (OGSE), whose primary goal is to provide end-to-end testing of the integrated Ariel telescope, optical bench and spectrometers. To account for gravity effect (potato chip), vertical astigmatism was fixed to 3 \(\mu m\) root mean square (r.m.s.) as a rough estimate that will be replaced in the future with an input from Structural, Thermal and Optical Performance (STOP) analysis.

PAOS was used to simulate the wavefront propagation through the OGSE module at \(500 nm\), where diffraction effects are smallest. To simulate the OGSE beam, the Ariel primary mirror M1 was illuminated with a perfect beam with footprint \(1/4\) the M1 diameter and the OGSE beam expander was modeled as a lens doublet giving an expansion of \(4\).

Below, we report the histogram of aperture sizes that give an EE \(\sim 90 \%\) at the OGSE exit pupil. The difference between these aperture sizes and the TA rEE requirement informs on how aberrated the OGSE beam can be.

../../_images/ogse90.png — Fig. 27 Histogram of aperture sizes for OGSE