Emissivity measurements of shocked tin using a multi-wavelength integrating sphere

Pyrometric measurements of radiance to determine temperature have been performed on shock physics experiments for decades. However, multi-wavelength pyrometry schemes sometimes fail to provide credible temperatures in experiments, which incur unknown changes in sample emissivity, because an emissivity change also affects the spectral radiance. Hence, for shock physics experiments using pyrometry to measure temperatures, it is essential to determine the dynamic sample emissivity. The most robust way to determine the normal spectral emissivity is to measure the spectral normal-hemispherical reflectance using an integrating sphere. In this paper, we describe a multi-wavelength (1.6-5.0 μ m) integrating sphere system that utilizes a "reversed" scheme, which we use for shock physics experiments. The sample to be shocked is illuminated uniformly by scattering broadband light from inside a sphere onto the sample. A portion of the light reflected from the sample is detected at a point 12 deg from normal to the sample surface. For this experiment, we used the system to measure emissivity of shocked tin at four wavelengths for shock-stress values between 17 and 33GPa. The results indicate a large increase in effective emissivity upon shock release from tin when the shock is above 24-25GPa, a shock stress for which the sample is partially melted when the shock releases. We also recorded an IR image of one of the shocked samples through the integrating sphere, and the emissivity inferred from the image agreed well with the integrating-sphere, pyrometer-detector data. Here, we discuss experimental data, uncertainties, and a data analysis process. We also describe unique emissivity-measurement problems arising from shock experiments and methods to overcome such problems.