This thesis describes the thermal design and thermal testing of the development model radiative cooler for the Composite Infra-Red Spectrometer (CIRS) due for launch on the Cassini spacecraft in 1997. The radiative cooler is used to cool the instrument's Focal Plane Assembly (FPA) to approximately 80K. The FPA holds two arrays of HgCdTe detectors for the mid infra-red spectrometer of the instrument which covers the wavelength range 7μm to 17μm. The FPA is mounted from the optics on a titanium alloy tripod and is cooled conductively by the radiator via a flexible link and a cold finger. A range of thermal models of the system have been developed ranging from a simple, analytical model to a finite difference numerical model. A calorimeter was designed to perform heat leak measurements on samples of Multi- Layer Insulation (MLI) blankets to determine the number and type of shields required for the MLI blanket covering the back of the cooler radiator. A test facility incorporating a vacuum system, a space simulator target, and a simulator for the CIRS instrument was designed and constructed for testing the assembled cooler. Various configurations of the Development Model (DM) CIRS cooler were tested as components became available and the results obtained compared to the thermal model predictions. It was found that the cooler will attain a temperature of 80K in operation, but with less excess cooling power than predicted by the thermal models.

Barium peroxide was selected as oxidant in a fundamental physico-chemical study of binary pyrotechnic systems on account of its apparently simple decomposition stoichiometry. With this selection, the choice of fuel was governed by the requirements of a self-sustaining reaction at combustion temperatures below the melting point of the platinum/rhodium thermocouples (∼1760°C) used for recording temperature-time profiles during burning, and at burning rates not exceeding the response of the sensors used to monitor combustion. Both manganese and molybdenum metal powders satisfied the above requirements. Strontium peroxide was also available as an oxidant and so the combustion of binary metal/oxidant systems using both fuels and both BaO₂ and SrO₂ oxidants was investigated. The Mn/BaO₂, MoBaO₂ and Mn/SrO₂ systems burnt over a wide range of compositions, but the range of ignitable compositions for the Mo/SrO₂ system was very limited. The linear burning rates, for all these systems, ranged from 2 to 12 mm s⁻¹ and burning rates were increased by the use of smaller particle-sizes of fuel and greater loading pressures. Inert additives generally decreased the burning rate. Temperature-time profiles were recorded for all the compositions which sustained combustion. Kinetic parameters were estimated from the shapes of these profiles using procedures developed by Hill et al and Boddington and Laye. Activation energies derived from the profiles were low (3 to 40 kJ mol⁻¹) and support suggestions that reactions at high temperatures are controlled by diffusion processes. Thermal analysis was used to identify the processes occurring in the four systems. The main exothermic events were observed to correspond approximately with the onset of oxidant decomposition. A pre-ignition reaction was tentatively identified in the Mn/BaO₂ system. Oxidation of the metal fuels was generally incomplete, probably because of the formation of protective layers of product. Activation energies, derived from thermal analysis results, were in the range of from 70 to 720 kJ mol⁻¹. Ternary systems containing either mixed fuels or mixed oxidants were also examined. No interactions hetween the fuels or between the oxidants were observed. Other techniques used included bomb calorimetry, measurement of thermal conductivity, X-ray powder diffraction, infrared spectroscopy and scanning electron microscopy.