1. therein) for 300 K< T(K) < 2000 A' and has tested them with a computer program designed for the determination of laser gain and maximum available power of a CO-2 -N?H^O GDL. He treats the modes u\ and 1/2 at equilibrium, but he recalls that there are elements that make this assumption controversial. He also considers that the levels N2(v - 1) and CO2(i/3) are at equilibrium. He uses the data given by Taylor and Bitterman [6]and calculations made by Sharma et al. [44]. A large scatter exists in most of the data. As a consequence, the calculations of gasdynamic laser gain and maximum available energy are subject to at least 25 percent inaccuracy. Four processes need more accurate rates, in particular process (5) with M - N2and H2O and process (2)with M = H2O. Volkov et al. Volkov et al. [51] use the compilation of Anderson with some changes for calculations of the laser gain. The law for process (5) with M = N2is given by Biryukov et al. [54]. CO2(vz) and N2(v - 1) are not considered at equilibrium. Taking into account the uncertainties in the relaxation constants, these authors determine the ranges for the optimal workingmixture composition and the ranges for the optimal stagnation temperatures and pressures. Taylor. Taylor [52] considers the vibrational relaxation of species CO2,N2, O2,CO and H2O. Hetakesintoaccount numerous experimental data obtained in various temperature ranges, in particular the one from Taylor and Bitterman [6]. Haixing. Haixing [53] uses a lot of experimental data, in particular the survey from Taylor and Bitterman [6] but mentions that it lacks data for many important processes and that some important problems in the relaxation process of the system are not thoroughly adressed. Hefinds largeerrors in the data deduced from experiment (one or two orders of magnitude). He uses a weighted least squared method to fit the data and give a greater weight to the most precise measures. For process (5), he proposes two laws, one according to the Taylor and Bitterman [6] assumption that CO2(v2) is at equilibrium with translational temperature, and the other without this assumption. The following process:
2. 2 Comparison of the models The temperature dependence laws given by the different authors are compared for the processes (1), (2), (3) and (5) with M = AV These reactions are of interest in this study where the CO2-N2mixture is considered. The results are presented in the figures from 2 to 5. For process (4) (Fermi resonance), there is no figure because no temperature dependence is available. We also have added in the figures some recent or significative experimental and theoretical data. Vibrational relaxation of CO2(v2) by JV2(V-T process (2)). This reaction is of substantial significance for the CO2-N2laser. However, up to 1969, it was subjected to considerable inaccuracies. Taylor and Bitterman [6]use the rate constant &^<7O2 ^orProcess(2) with M - CO-i for which a lot of data are available and propose to take fc^jv2=a-^?O2w^tna- 0.2. On the basis of new data, Anderson [50] and Volkov et al. [51] choose to take a = 0.5. Approximately at the same time, Nickerson [5] proposes a = 0.7. The curves corresponding to the different models are shown in figure 2. The reaction rate ^2/N2seemsto be relatively well known. However, process (2) is the rate determining step for the relaxation of CO-i because it is the lower vibrationalenergy level. A moreprecise determination of this reaction rate should be necessary. Vibrational relaxation of CO2(vz) by N2(V-V process (5)). For this process, all the following reactions are to be considered a priori:
3. (V -1). The quasi-resonant process (1) has been studied both experimentally and theoretically (cf. figure 4). Taylor and Bitterman [22]find from shock waves experiments that there is a minimum of the reaction rate at 1000 K. However, the least square determination by Nickerson [5] using several experimental data [7, 22, 24] shows a minimum at 600 K with lower reaction rates. The theoretical determination by Sharma and Brau [43]is ingood agreement with the experimental data of Moore et al. [7]and Rosser et al. [24] until 1000 K. For higher temperatures, this theoretical lawis not adequate. Taylor [52], Volkov et al. [51] and Haixing [53] use more recent data than Nickerson [5] which are in good agreement with Rosser [24]. The results provided by Vargin et al. [16] (spectrophonic method) present a minimumat 400 K and differ from the results of Rosser et al. [24]. The CARS measurement by Wang et al. [32] in 1992 leads to a lower reaction rate at 300 K, compared to the other results. However, Wang et al. [32] cite even lower determinations of this reaction rate by other authors. Here more precision is clearly needed. V-V process (4), Fermi resonance. Only the reaction rates at 300 K are available. The values given below (Table 1) are to be compared with the gas kinetic constant for COi in collision with another COi molecule, which isequal to 2.61010cm3/(molecule.s). The collision partners considered here are even Ny and Oi or CO- There are important disagreements in the data which