Comparison of Simultaneous Shock Temperature Measurements from Three Different Pyrometry Systems

Abstract

Pyrometry is one of the most prevalent techniques for measuring temperature in shock physics experiments. However, the challenges of applying pyrometry in such highly dynamic environments produces multiple sources of uncertainty that require investigation. An outstanding question is the degree of agreement between different pyrometers and different experiments. Here we report a series of novel plate impact experiments with simultaneous thermal radiance measurements using three different multi-wavelength optical pyrometry systems, each with different spatial and temporal resolutions, on samples shocked to identical states. We compare the temperatures measured by each system and their associated uncertainties using a number of emissivity assumptions. The results shown that the measurements from all three systems agree within uncertainty. Some non-thermal light contamination was observed despite a number of prevention measures.

Appendix 1

During the analysis presented here, an uncertainty in reflectivity in percentage terms is used. It is useful to discuss how uncertainty in reflectivity is translated to emissivity uncertainty.

In fixed uncertainties, because reflectivity and emissivity are complementary, see Eq. 2, the uncertainties are equal but opposite,

$$\Delta \varepsilon = - \Delta r,$$

where Δ denotes an absolute uncertainty. However, if proportional errors are defined such that,

$$\delta \varepsilon = \frac{\Delta \varepsilon }{\varepsilon };\;\delta r = \frac{\Delta r}{r},$$

where δ denotes proportional uncertainty, the relation is different. Rearranging the above expressions,

$$\varepsilon \delta \varepsilon = - r\delta r.$$

This expression can be used to determine,

$$\delta \varepsilon = \frac{r\delta r}{1 - r}.$$

Thus a 10% error in reflectivity on a baseline reflectivity of 0.9 will give a proportional emissivity error of 90%.