Characterisation of the MIRI spectrometer, an instrument for the James Webb Space Telescope

The MIRI-MRS is a future space based Medium Resolution Spectrometer and one of four instruments to be integrated onto The James Webb Space Telescope. The Medium Resolution Spectrometer is designed to be diffraction limited across its entire passband of 5 - 28.3 microns. It achieves this through the spectral filtering of the passband into four channels with each one containing an integral field unit optimised for minimal diffraction losses. The integral field unit enables the simultaneous measurement of the spectral data across the entire field of view. The design of the Medium Resolution Spectrometer is outlined with particular reference to the choice of slice widths used for each channel to minimise the diffraction losses from the slicing mechanism. The slice widths are also used to derive the extent of the field of view and combined with the along slice plate scale at the detector the technique required for complete spatial sampling of the spectrometer is outlined. The operation of the Channel 1 image slicer component was tested cryogenically at 5 microns for diffraction losses due to the slicing of the point spread function. This was so that the actual diffraction losses could be measured and compared with the optical model. From the resulting analysis I concluded that the operation of the image slicers were well understood for diffraction losses. Performance tests were required on the instrument because of its novel design. This was the first implementation of an integral field unit operating between 5 - 28.3 microns and it was necessary to ensure that the operation of the image slicer did not induce unacceptable diffraction losses into the instrument. Tests were required on the assembled instrument to verify the optical design. A Verification Model of MIRI was built to enable test verification of the optical design. This testing was carried out in advance of the MIRI Flight Model assembly so that changes could be made to the Flight Model design if necessary. This testing phase was also designed to define the calibration process necessary to prepare the MIRI Flight Model for scientific operations. For the testing phase it was necessary to create an astronomical source simulator. This MIRI Telescope Simulator was constructed in Madrid where I spent two months ensuring the point source movement across the field of view would be sufficient to investigate the Medium Resolution Spectrometer. My contribution was to help assemble both the Verification and Flight Models. I also participated in the Verification Model testing phase from the test design phase to the test implementation and data analysis. My role in the analysis was to investigate the field of view of the Medium Resolution Spectrometer Verification Model and whether the field of view requirements for the spectrometer were met. During this analysis I also verified that the diffraction effects of the end-to-end instrument were well understood by the optical model. The Medium Resolution Spectrometer Verification Model field of view compromised the field of view requirement for the spectrometer. A similar analysis for the Flight Model showed that there would be a low probability that the field of view requirement would be met. As a result of the analysis I defined a new slit mask design that would align the field of view sampled by Channel 1 to increase the aligned field of view. As a result there is a high probability that the field of view requirement for the Flight Model will be exceeded. The test analysis discovered a magnification effect within the spectrometer which must be properly characterised to enable accurate field of view reconstruction. I designed a test necessary for the calibration phase of the Flight Model to enable full spatial alignment of the Medium Resolution Spectrometer. I also measured an excess flux level in the Channel 1 observations at the detector and there was a ghost detected in the Channel 1 images. Whilst the origin of either the excess flux or the ghost could not be completely determined I investigated the possibility that they will not be present in the Flight Model due to the slight design differences. If present however they will not increase the background level of an observation above the requirement outlined for Channel 1