David S. Marx

Hybrid Lyot coronagraph for WFIRST: high-contrast broadband testbed demonstration

Hybrid Lyot Coronagraph (HLC) is one of the two operating modes of the Wide-Field InfraRed Survey Telescope (WFIRST) coronagraph instrument. Since being selected by National Aeronautics and Space Administration (NASA) in December 2013, the coronagraph technology is being matured to Technology Readiness Level (TRL) 6 by 2018. To demonstrate starlight suppression in presence of expecting on-orbit input wavefront disturbances, we have built a dynamic testbed in Jet Propulsion Laboratory (JPL) in 2016. This testbed, named as Occulting Mask Coronagraph (OMC) testbed, is designed analogous to the WFIRST flight instrument architecture: It has both HLC and Shape Pupil Coronagraph (SPC) architectures, and also has the Low Order Wavefront Sensing and Control (LOWFS/C) subsystem to sense and correct the dynamic wavefront disturbances. We present upto-date progress of HLC mode demonstration in the OMC testbed. SPC results will be reported separately. We inject the flight-like Line of Sight (LoS) and Wavefront Error (WFE) perturbation to the OMC testbed and demonstrate wavefront control using two deformable mirrors while the LOWFS/C is correcting those perturbation in our vacuum testbed. As a result, we obtain repeatable convergence below 5 × 10−9 mean contrast with 10% broadband light centered at 550 nm in the 360 degrees dark hole with working angle between 3 λ/D and 9 λ/D. We present the key hardware and software used in the testbed, the performance results and their comparison to model expectations.

Dynamic testbed demonstration of WFIRST coronagraph low order wavefront sensing and control (LOWFS/C)

To maintain the required performance of WFIRST Coronagraph in a realistic space environment, a Low Order Wavefront Sensing and Control (LOWFS/C) subsystem is necessary. The LOWFS/C uses a Zernike wavefront sensor (ZWFS) with the phase shifting disk combined with the starlight rejecting occulting mask. For wavefront error corrections, WFIRST LOWFS/C uses a fast steering mirror (FSM) for line-of-sight (LoS) correction, a focusing mirror for focus drift correction, and one of the two deformable mirrors (DM) for other low order wavefront error (WFE) correction. As a part of technology development and demonstration for WFIRST Coronagraph, a dedicated Occulting Mask Coronagraph (OMC) testbed has been built and commissioned. With its configuration similar to the WFIRST flight coronagraph instrument the OMC testbed consists of two coronagraph modes, Shaped Pupil Coronagraph (SPC) and Hybrid Lyot Coronagraph (HLC), a low order wavefront sensor (LOWFS), and an optical telescope assembly (OTA) simulator which can generate realistic LoS drift and jitter as well as low order wavefront error that would be induced by the WFIRST telescope’s vibration and thermal changes. In this paper, we will introduce the concept of WFIRST LOWFS/C, describe the OMC testbed, and present the testbed results of LOWFS sensor performance. We will also present our recent results from the dynamic coronagraph tests in which we have demonstrated of using LOWFS/C to maintain the coronagraph contrast with the presence of WFIRST-like line-of-sight and low order wavefront disturbances.

Electric field conjugation in the presence of model uncertainty

The Wide-Field Infrared Survey Telescope (WFIRST) is a 2.4m diameter space telescope NASA program. The payload will include a coronagraph instrument (CGI). The CGI designs under development use deformable mirrors (DM) to create a point spread function (PSF) with a dark region around the obscured star object. Electric field conjugation (EFC) is an iterative nonlinear optimization procedure that uses measurements of the electric field at the image to determine the DM displacements that modify the PSF to create the region of high contrast in the image. EFC requires a numerical model of the coronagraph to calculate the Jacobian of the system, which is used, along with regularization, to solve for the DM displacements for each iteration of the nonlinear optimization. Ideally, the coronagraph is aligned and calibrated, and the calibration data are used in the numerical model for calculating the Jacobian. However, calibration and alignment measurements always contain uncertainty resulting in calibration error. Therefore, the Jacobian calculated from the numerical model is not an exact representation of the physical coronagraph, and the resulting DM solution for an EFC iteration does not have the exact impact on the electric field of the coronagraph as predicted by the EFC. The result is slow convergence, and, as will be shown, the necessity of more restrictive regularization. Using Monte Carlo trials, we investigate the effect of calibration error on EFC convergence and regularization. Comparison to results from the High Contrast Imaging Testbed Hybrid Lyot Coronagraph are also presented.

WFIRST coronagraph optical modeling

End-to-end numerical optical modeling of the WFIRST coronagraph incorporating wavefront sensing and control is used to determine the performance of the coronagraph with realistic errors, including pointing jitter and polarization. We present the performance estimates of the current flight designs as predicted by modeling. We also describe the release of a new version of the PROPER optical propagation library, our primary modeling tool, which is now available for Python and Matlab in addition to IDL.

Laser metrology in the micro-arcsecond metrology testbed

The Space Interferometer Mission (SIM), scheduled for launch in 2009, is a space-born visible light stellar interferometer capable of micro-arcsecond-level astrometry. The Micro-Arcsecond Metrology testbed (MAM) is the ground-based testbed that incorporates all the functionalities of SIM minus the telescope, for mission-enabling technology development and verification. MAM employs a laser heterodyne metrology system using the Sub-Aperture Vertex-to-Vertex (SAVV) concept. In this paper, we describe the development and modification of the SAVV metrology launchers and the metrology instrument electronics, precision alignments and pointing control, locating cyclic error sources in the MAM testbed and methods to mitigate the cyclic errors, as well as the performance under the MAM performance metrics.

WFIRST/AFTA coronagraph contrast performance sensitivity studies: simulation versus experiment

The WFIRST/AFTA 2.4 m space telescope currently under study includes a stellar coronagraph for the imaging and the spectral characterization of extrasolar planets. The coronagraph employs sequential deformable mirrors to compensate for phase and amplitude errors. Using the optical model of an Occulting Mask Coronagraph (OMC) testbed at the Jet Propulsion Laboratory, we have investigated through modeling and simulations the sensitivity of dark hole contrast in a Hybrid Lyot Coronagraph (HLC) for several error cases, including lateral and longitudinal translation errors of two deformable mirrors, DM1 and DM2, lateral and/or longitudinal translation errors of an occulting mask and a Lyot-Stop, clocking errors of DM1 and DM2, and the mismatch errors between the testbed and the model sensitivity matrices. We also investigated the effects of a control parameter, namely the actuator regularization factor, on the control efficiency and on the final contrast floor. We found several error cases which yield contrast results comparable to that observed on the HLC testbed. We present our findings in this paper.

WFIRST low order wavefront sensing and control dynamic testbed performance under the flight like photon flux

To maintain the required performance for the WFIRST Coronagraph Instrument (CGI) in a realistic space environment, a Low Order Wavefront Sensing and Control (LOWFS/C) subsystem is necessary. The WFIRST CGI LOWFS/C subsystem will use the Zernike wavefront sensor, which has a phase-shifting disk combined with the coronagraph’s focal plane mask, to sense the low-order wavefront drift and line-of-sight (LoS) error using the rejected starlight. The dynamic tests on JPL’s Occulting Mask Coronagraph (OMC) Testbed have demonstrated that LOWFS/C can maintain coronagraph contrast to better than 10-8 in presence of WFIRST-like line of sight and low order wavefront disturbances in both Shaped Pupil Coronagraph (SPC) and Hybrid Lyot Coronagraph (HLC) modes. However, the previous dynamic tests have been done using a bright source with photon flux equivalent to stellar magnitude of MV = -3.5. The LOWFS/C technology development on the OMC testbed has since then concentrated in evaluating and improving the LOWFS/C performance under the realistic photon flux that is equivalent to WFIRST Coronagraph target stars. Our recent testbed tests have demonstrated that the LOWFS/C can work cohesively with the stellar light suppression wavefront control, which brings broad band coronagraph contrast from ~1x10-6 to 6x10-9, while LOWF/C is simultaneously suppressing the WFIRST like LoS and low order wavefront drift disturbances on a source that photon flux is equivalent to a MV = 2 star. This lab demonstration mimics the CGI initial dark hole establish process on a bright reference star. We have also demonstrated on the testbed that LOWFS/C can maintain the coronagraph contrast by suppressing the WFIRST like line-of-sight disturbances on a fainter MV = 5 star. This mimics scenario of CGI science target observations. In this paper we will present the recent dynamic testbed performance results of LOWFS/C LoS loops and low order wavefront error correction loop on the flight like photon flux.

Shaped pupil coronagraph: disk science mask experimental verification and testing

The Shaped Pupil Coronagraph (SPC) is one of the two operating modes of the baseline coronagraph instrument for the proposed WFIRST mission. While in SPC mode, multiple sets of shaped pupil masks and focal plane masks would be available for various imaging tasks. The disk science mask set (SPC-DSM) is designed for exozodiacal disk science. With a 360 degree high contrast field of view, extending up to 20 λ/D, the SPC-DSM provides a powerful tool to study exozodiacal dust clouds associated with stellar debris disks to gain insight of the exoplanet formation and stellar disk dynamics. We will describe the performance verification and demonstration of the SPC-DSM coronagraph as tested in the high contrast imaging testbed (HCIT) at JPL. The goal of the testbed demonstration is an average contrast of 5e-9 over a 10% bandwidth centered at 565nm, in a field of view extending from 6.5 λ/D to 20 λ/D. We will discuss electric field conjugation, performance metrics, and model agreement as applied to the SPC-DSM.

Hybrid lyot coronagraph for WFIRST: high contrast testbed demonstration in flight-like low flux environment

In order to validate required operation of the proposed Wide-Field InfraRed Survey Telescope (WFIRST) coronagraph instrument, we have built a testbed in Jet Propulsion Laboratory (JPL), which is analogous to the baseline WFIRST coronagraph instrument architecture. Since its birth in 2016, this testbed, named as Occulting Mask Coronagraph (OMC) testbed, has demonstrated several crucial technological milestones: Broadband high contrast demonstration in both Hybrid Lyot Coronagraph (HLC) and Shape Pupil Coronagraph (SPC) modes while the Low Order Wavefront Sensing and Control (LOWFS/C) subsystem senses and corrects the dynamic flight-like wavefront disturbances. In this paper, we present up-to-date progress of HLC mode demonstration in the OMC testbed. While injecting the flight-like low photon flux starlight with expected Line of Sight (LoS) and Wavefront Error (WFE) perturbation to the OMC testbed, we demonstrate generating high contrast dark hole images. We first study the expected photon flux in actual flight environment, and estimate detection noise and estimation accuracy of the complex electric field if the wavefront sensing algorithm is used based on the pair-wise difference imaging. Then, we introduce our improved scheme to mitigate this photon-starved flight-like low flux environment. As a result, we generate a dark hole that meets the WFIRST raw contrast requirements using the 2nd magnitude star light. We establish the key ideas, describe test setups, and demonstrate test results with data analysis.

Fast linearized coronagraph optimizer (FALCO) III: optimization of key coronagraph design parameters

Deformable mirrors (DMs) are increasingly becoming part of nominal coronagraph designs, such as the hybrid Lyot coronagraph, in addition to their role counteracting optical aberrations. Previous studies have investigated the effects of the inter-DM Fresnel number on achievable contrast, throughput, and tip/tilt sensitivity for apodized coronagraphs augmented with DMs to suppress diffraction from struts and segment gaps. In this paper, we build upon that previous work by directly suppressing tip/tilt sensitivity with the controller, both for coronagraphs with and without apodizers. We also explore the effects of other important design parameters such as actuator density and tip/tilt controller weighting on performance. These comprehensive coronagraph design studies are enabled by the Fast Linearized Coronagraph Optimizer (FALCO) software toolbox, which provides rapid re-calculation of the DM response matrix for a variety of coronagraphs.