Figure 1.The heating chamber, showing the vapour trap and twin-wheel arrangement to prevent contamination of the measurement window.
Résumé Les mesures pyrométriques ne perturbant pas le milieu étudié sont délicates à réaliser lorsqu'il y a évaporation, comme c'est le cas dans les essais de fabrication de cristaux, en raison des erreurs que provoque la moindre contamination d'une fenêtre située sur le chemin optique. Pour se prémunir contre cette contamination, il a été mis au point trois techniques utilisables soit isolément soit en association les unes avec les autres. Elles permettent de mesurer pendant de longues durées la température de métaux fondus, chauffés sous vide poussé à 200 K au-dessus de leur point de fusion, avec une finesse de résolution de l'ordre de ±0.1 K.
Contractor:
Alenia Spazio (I)
KE, Forschungsinstitut für Kerntechnik und Energiewandlung
e.V., (D)
Methods such as ratio or multi-spectral pyrometry, which are some times proposed for in situ corrections of emissivity, or radiation absorption by windows cannot be used if the radiation properties are wavelength dependent. Because the wavelength dependency of a contamination layer varies with its chemical composition and thickness, the only way to get reliable radiation temperature measurements of evaporative objects is by protecting the window against contamination. Three techniques have been developed for this purpose and have been successfully tested.
Very simple protection may be realised with a twin wheel arrangement. Each wheel contains four holes; three holes have a window and one does not, a total of 6 protective windows. The transmittance of each window is measured in situ by briefly turning the wheel to the window-free position to compare the signals obtained with and without a window. This technique is only good for checking window transmittance under conditions of low contamination. It may also be used, in combination with the vapour trap technique described later, to self-check its efficiency and to measure the remaining contamination.
To obtain continuous protection, a special Laval nozzle was constructed which produced a supersonic transverse gas flow whose width was matched to the cross-section of the optical path of the pyrometer.(Figure 1). The evaporation chamber is thus divided into two sections, maintained at different pressures - pH in the heating section and Pso in the vapour-free section. The efficiency of the vapour trap depends on the pressure Po at the entrance to the Laval nozzle and thus the corresponding flow rate.
Figure 1.The heating chamber, showing the vapour trap and
twin-wheel arrangement to prevent contamination of the
measurement window.
With this arrangement, copper and silver melts were placed in the heating section held at high vacuum (10-7 bar), and heated to temperatures of 1650 K and 1400 K respectively, approximately 200 K above their melting points. Without protection, window contamination would have decreased the measured temperature by more than 250 K in one hour. Results taken with vapour trap protection are shown in Figure 2. At a flow rate of 0.2 l/h (red curve), the drift in measured temperature is 26 K per hour and at the optimum flow rate of 0.4 l/h errors due to contamination are practically eliminated. After a 5-hour experiment run a drift of approximately 2 K per hour was observed. These measurements were checked using the twin wheel arrangement, also shown in Figure 1.
Figure 2. Drift of the pyrometer signal due to window
contamination versus vapour trap protection for various argon
flow rates. The test object was a silver melt heated to 1350 K
in high vacuum 10-7 bar.
In principle, the protective glass windows can be exchanged for a strip of plastic foil with the advantage that a long strip can provide between 200 to 300 replaceable windows (Figure 3). Unfortunately, the transmittance of these foils is not homogeneous, and a random temperature measurement error in the range of 1 to 2 K may result as the uncalibrated foil is wound on. To overcome this problem, a new technique has been developed in which:
Figure 3. Foil drive system of the window protection unit, showing the components
for exact positioning of the foil.
Thorough reproducibility tests have demonstrated that the technique is accurate and the remaining uncertainty in the pyrometric temperature measurement, caused by the foil transmittance, is of the order of ±0.1 K when the specimen is at a temperature of 1500 K.
The switching frequency may be reduced by combining the foil technique with a slight counterflow of argon gas. Using this combination, the temperature of a silver melt at 1300 K was measured and the results are shown in Figure 4. The first foil section was held in the optical path for 20 minutes, and the contamination accumulated during that time produced a decrease in measured temperature of approximately 4 K. Without counterflow the temperature drift would have been approximately 20K under the same conditions. By switching from one section to the next every 10 minutes, the temperature reading of pyrometer was stable within ±0.5 K, which is in the same order of magnitude as the fluctuations in the thermocouple temperature. By reading only the values of starting temperature after each changing a section and interpolating between these values, the fluctuations in the measured temperature are held below ±0.2 K.
Figure 4. Contamination protection by the replaceable foil
technique combined with slight argon counterflow at 0.04
l/h during measurement of a silver melt in vacuum at
5x10-7 bar. The circles represent the pyrometer
signals immediately after switching to the next foil section,
before it has been contaminated
The existing experimental foil device is very large and heavy and, because it is mounted on a flange between heater and window, it is therefore difficult to align with the optical axis of the pyrometer. For routine applications, a new cassette type foil drive system has been proposed. In this concept, the drive system can be mounted directly on the inner side of the window flange with automatic centring along a window axis which is identical to the axis of the optical path. This gives a flexible system which can be easily adapted to any heating chamber for space or terrestrial applications.
Both vapour trap and foil technique can be used to protect a window of a vacuum chamber against contamination. However a method which combines the foil technique with a counterflow of an inert gas offers high flexibility, compactness and low energy consumption. Even at very high evaporation rates, a temperature resolution of the order of ±0.1 K can be realised, making possible measurement times up to 100 hours with a single foil coil, a small half litre bottle of argon gas and a laminar flow at 100 bar pressure.
Preparing for the Future Vol. 7 No. 1