Abstract
The overall heat transfer coefficient of three types of thermopile solarimeters were determined from the rate of change of emf after an instantaneous change in the level of incident radiation. Conductive and radiative transfer coefficients were calculated from instrument dimensions, but experiments with Moll-Gorczynski solarimeters filled with different gases provided indirect proof of their accuracy. The empirical relation between Nusselt and Rayleigh numbers agreed reasonably with previous estimates for comparable systems. Increasing the overall heat transfer coefficient for a given type of thermopile element, by altering the gaseous content or the casing dimensions, will decrease sensitivity and linearity of response, increase the temperature coefficient and decrease the response time. Decreasing the thermal capacity of the element will decrease response time and increase sensitivity. A unique determination of sensitivity, linearity, temperature coefficient and response time of any instrument cannot be given, since all of these characteristics are functions of the incident radiation flux density.