The Ferrofluid Deformable Mirror Concept

Abstract

The detection of exoplanets, planets in other solar systems, through direct imaging is a particularly intriguing field of research that holds many implications for the future. By performing spectroscopy, possibly habitable Earth-like planets can be located and characterized. However, there are many challenges associated with imaging exoplanets, one being image distortions due to wavefront errors. To remedy this problem mirrors with controllable surface shapes, known as deformable mirrors, are used to real-time correct wavefront errors. Current deformable mirrors (DMs) use miniature mechanical actuators to generate the specific surface shapes needed for correction. The fundamental drawbacks with traditional DMs include cost, fragility, and the need for continuous power. In addition, images from traditional DMs have an undesirable quilt pattern due to the discrete actuators beneath the mirror surface. As an alternative to traditional DMs, Dr. Tyler Groff of the High Contrast Imaging Laboratory (HCIL) at Princeton University proposes a DM concept where a solid reflective surface is continuously supported by magnetic ferrofluid. Permanent mag- nets and electromagnetic inductors are then used to actuate the DM. The ferrofluid deformable mirror (FDM) concept solves the quilting problem and aims to be more cost-effective, robust, and energy efficient than traditional DMs. We begin this thesis by introducing the problem at hand and motivating the FDM concept. We also discuss past DM concepts and how they are different from the FDM. Next, we explore the design process for both the FDM and the interferometer setup used to measure the surface shape of the mirror. We include key engineering decisions, finite element analyses, CAD models, and part drawings. In the third part of this thesis, we characterize the response behavior of the FDM to actuation. This includes a demonstration of repeatability, a demonstration of the superposition of multiple actuators, and a process for calculating system gain. Finally, we conclude this thesis with a summary of what has been accomplished up to this point and suggest design improvements and future testing.