Energy balance calculations have been made of a cylindrical high-pressure LTE discharge which contains small quantities of metal halides with a large excess of mercury. It is assumed that the partial pressures of emitting atoms remain at a constant value throughout the arc, and that the emission coefficient for atoms and ions can be calculated using the Corliss and Bozman (1962) transition probabilities. The model has been used to compare the effects of adding different elements (scandium and iron) and mixtures of elements on the temperature profile, and other properties of metal halide lamps. The self-absorbed radiation of the mercury atoms must be included in order to obtain realistic temperature profiles. It is assumed that for all lines emitted by the added elements, the arc is optically thin.

The net emission coefficient of a self-absorbed spectral line can be described by a semi-empirical formula which contains terms representing the generation and absorption of radiation. Four empirical constants are determined by applying least-squares fitting to the results of an exact radiation flux density calculation in a cylindrical arc tube. Calculations on the D-lines under conditions which are typical of high-pressure sodium lamps are described. In this case, the empirical formula predicts radiation flux densities which agree to better than 10% with the exact calculation even when the arc tube bore and arc temperature are changed by 15% from their original values. This method allows economical and accurate calculation of energy balances in arcs containing several self-absorbed spectral lines.

Operating wall temperature and electrode separation are important factors in the design of efficient high-pressure sodium lamps. The authors have measured and calculated how the arc efficacy varies with wall temperature and input power per unit length. In their experiments they controlled the wall temperature by passing gases with different thermal conductivities over the arc tube. Their experiments showed that the arc efficacy increases at rates of 3.75+or-0.15% per 100K rise in wall temperature and 3.5+or-0.3% per 1 Wmm-1 increase in electrical power input. The corresponding calculated average rates are 2.1% per 100K and 1.7% per 1 Wmm-1. The satisfactory agreement of their experiments with their calculations resolves a discrepancy between their computer model predictions and those of Waymouth and Wyner (1981). Their two-zone model predicted that the efficacy decreases as the electrical power input increases. Although they have found that the proportion of the total radiation emitted in the D-lines decreases as the power input increases, this is more than compensated by an increase in the total radiated power at the expense of thermal conduction.

In order to model high pressure discharge lamps which contain metal halides, the chemical composition in the plasma region is required. We have developed a procedure for calculating the composition of a plasma in local thermodynamic equilibrium, between 1500 and 6000 K, including the effect of radial diffusion. Mixtures of tin halides and sodium halides, when added to high pressure mercury discharges, produce highly efficient white light sources. We have determined the saturated vapor phase composition of some tin halide/sodium halide mixtures using the results from Knudsen effusion/ mass spectrometric, vapor pressure, and transpiration studies. We report herein a detailed study of the SnC12 + NaC1 (50:50) system and a preliminary investigation of the SnC12 + NaI (50:50) system together with the plasma compositions for the corresponding lamps. These results show that radial diffusion has a highly significant effect on the plasma composition.Our understanding of high pressure sodium lamps has been deepened through the development of a computer model (1). The model is based on solving the energy balance equation to give the radial temperature distribution of an arc. We assume that the electrical energy dissipated in an elemental volume equals the sum of the net heat and radiation losses. The variation of the chemical composition of the plasma with temperature stands at the heart of the model, and is used in the calculations of thermal and electrical conductivities and of radiation losses.A knowledge of the temperature profile enables us to predict the electrical and spectral characteristics of the arc. We have used our model to investigate the variation of efficacy for a high pressure sodium lamp with arc tube bore and current (2), and with arc tube temperature and electrical power dissipation (3). The model has thus contributed directly to the development and improvement of high pressure sodium lamps.Metal halide lamps are able to combine good color appearance and good color rendition with compact size, giving rise to highly efficient white light sources. A large variety of metal halide lamps are available on the market, these are generally confined to powers above 150W. Such lamps are employed in many applications: industrial and commercial interior lighting; exterior floodlighting; theater and stage lighting; set lighting; image projection systems; photoprinting; water sterilization; plant growth; street lighting. The present trend is towards lower power lamps, say below 150W, for use in display lighting. The development of successful low wattage metal halide lamps has proved to be a considerable challenge to the lamp manufacturing industry. To date, only a few low power metal halide lamps are available.The addition of tin halide/sodium halide mixtures to high pressure mercury discharges has been shown to produce lamps which combine high efficacy with good color rendition (4). The inclusion of metal halide systems in high pressure lamps gives rise to both atomic and molecular radiation, the latter being broad-ba...