Gasket temperature: an alternate technique for estimating sample temperature in a multi-anvil apparatus

It is common for sample temperatures to be estimated by the power input to the furnace in multi-anvil experiments in which a thermocouple cannot be used or the thermocouple failed during heating. Uncertainties using this technique are often on the order of ±85°C or larger. This paper describes a new method for estimating sample temperature using a second thermocouple outside all pressure media. Temperatures recorded at this external (gasket) thermocouple trend linearly with the internal (sample) thermocouple temperature. Because of thermal lag, it is necessary to determine the first gasket temperature (T₀) corresponding to the desired sample temperature. Accurate prediction of T₀ for the desired sample temperature can come from relatively few (5–6) gasket-temperature measurements made during the initial temperature ramp over a small temperature range (500–600°C). Using this method and manually ramping to T₀ allows for uncertainties in sample temperature estimation as small as ±20°C.