Refractometry of air is a central problem for interferometer-based dimensional measurements. Refractometry at the 10−9 level is only valid if air temperature gradients are controlled at the millikelvin (mK) level. Very precise tests of secondgeneration National Institute of Standards and Technology (NIST) refractometers involve comparing two instruments (two optical cavities made from ultralow expansion glass) that are located in nominally the same environment; temperature gradients must be kept below a few millikelvin to achieve satisfactory precision of these tests. This paper describes a thermal stabilization scheme that maintains < 1 mK thermal gradients over 100 h in a 0.5 m x 0.15 m x 0.15 m volume. The approach uses passive (aluminum envelopes and foam insulation) and active (thermistors, foil heaters, and proportional-integral-derivative control) temperature stabilization. Thermal gradients are sensed with thermocouples and a nanovoltmeter and switch; the reference junctions of the thermocouples being in thermal contact with a thermistor temperature standard. Indirect evidence shows that performance is better than the < 1 mK uncertainty of measuring temperature gradients, the uncertainty of which is due to limitations of the nanovoltmeter and switch.
Citation: NCSL International Measure
Pub Type: Journals