Active spectroscopy on HID lamps : exploration of various methods

Abstract

In order to improve our understanding of plasma processes in metal halide (MH) lamps, numerical models are under construction. Verification of these models requires quantitative data of the species density distributions inside these lamps. Since density profiles are dictated to local chemical processes and transport mechanisms, both strongly depending on the temperature, accurate temperature information is of great importance. Therefore the first goal of this thesis is to develop diagnostic techniques and to employ these techniques to determine two different temperatures, namely the gas temperature and the electron temperature. Based on the temperature data, the presence of LTE can be verified. This validation of the LTE-assumption will have a significant impact on modelling. To achieve the first goal, two main diagnostic tools have been designed and explored, namely X-ray absorption (XRA) and Thomson scattering (TS). XRA gives the gas temperature while TS provides data on the electron temperature and electron density. In a sidetrack we also did a feasibility study to determine the elemental density of sodium or iodine by means of -spectroscopy. However, by tracing radioactive isotope 24Na generated by the irradiation of the natural isotope 23Na, it was found that it is impossible to obtain the elemental density distribution unless the lamp could be placed in a nuclear reactor. Moreover, a simple calculation shows that filling the lamp with a salt containing 22Na is not an option to obtain the spatial distribution of sodium. One promising method would be the SPECT technique which has been successfully applied in nuclear medicine diagnostics to trace radioactive nuclei. By introducing 123I-labelled NaI into the lamp, it is in principle possible to obtain the spatial distribution of iodine with the SPECT system with a spatial resolution of 0.5 mm. It requires further investigation and an in-depth study in order to apply this method to metal halide lamps. Both XRA and TS measurements were performed on the same lamps, high pressure Hg lamps with 15 mg and 50 mg Hg fillings. The results show that the electron temperature is higher than the gas temperature. This means that the LTE-assumption in these lamps is not valid. Moreover, a deviation from Saha equilibrium was found by comparing the ground state Hg density using the XRA results and the ground state Saha density using the results from TS. It has been found that the deviation from Saha equilibrium is increasing with the distance to the center. TS has also been performed on an Ar model lamp in order to investigate the plasma properties in front of the electrodes. It has been found that the plasma is strongly ionizing in front of the electrodes and thus not in LTE. There are several challenges when TS is performed on the high pressure Hg lamps. Firstly, the TS signal has to compete with the substantial signal of stray light which is introduced by the plasma surroundings: the lamp envelope. Secondly, the TS signal has to compete with the strong plasma radiation in the Hg lamp. Thirdly, the TS signal is limited by the laser power since the laser should not break the quartz tube and not create a laser-induced plasma. If one can not meet these three requirements at the same time, then it is impossible to do TS measurements. Fortunately, by means of several adjustments on the existing TS-setup we have successfully obtained the clear TS signal on high pressure Hg lamps. The results obtained in this way on pure Hg plasmas may offer a firm platform from which TS on metal halide lamps can be conducted in the near future. However, it should be realized that the presence of metal halides will lead to an increase of the plasma background radiation so that the application of TS will be increasingly difficult. The execution of XRA measurements on HID lamps is found to be far from trivial, but possible. The experimental setup required covers a high dynamic range CCD camera, a high flux X-ray source and a near monochromatic source. After lamp-on and lamp-off images are taken, a simple subtraction of the logarithm of lamp-off and lamp-on image intensities will not bring us automatically the correct temperature profile. The final temperature profile depends on correct procedures for signal extraction, image process procedures and Abel inversion. To demonstrate the capabilities of this technique, XRA measurements were performed on various kinds of HID lamps. It was shown that this method is well suited for gas temperature measurements on HID lamps. It was also shown that the presence of the metal additives in the arc leads to constriction of the temperature profile. Moreover, the constriction of temperature profile indicates the presence of axial segregation effects. Even for the worst scenario, a PCA HID lamp with low Hg pressure and small diameter, it has been proven that this method can yield reliable results. To conclude, the XRA method is a reliable and sensitive diagnostic method for gas temperature measurement in HID lamps. TS is another important technique for the electron gas properties. By combining these two methods, we have obtained a better insight in the plasma conditions.