1. Asecond difficulty in obtaining efficient laser operation is that as a result of the constant pressure requirement the static gas temperature in the generator rises considerably as power is extracted. To quantify.this effect, the one-dimensional gas conservation equations were numerically solved for the present plasma and generator conditions. Two calculations were pe_rformed. In one case O'EFF was fixed at 3. 2 Mhos/m (corresponding to a2constant generator current density J y of 3 A/cm and a nearly constant Ne of 2. 5 x 1019m-3). In the other case, 2. The diode was a conventional laser tube 2 m. long and 2. 5 cm in diameter. The test gas, a mixture of He, Nz, and COz or He and COz,flowed over a cesium bath at a rate which produced a flow velocity of about 10 m/ sec. in the diode. The diode was cooled with oil at a temperature of 100° C to prevent the cesium from depositing on the walls. Somewhat unexpectedly it was observed that the liquid cesium reacts strongly with COz . At a static pressure of 10torr in a He: N2:co 2mixtureof proportion 1. 0: 1. 0: 8, 0, the total voltage was about 8 KV at a current of 50 m. a. and the laser output was 25 watts. The low output was a result of the use of brewster angle windows which protected the mirrors from cesium attack, The COz was injected downstream of the cesium bath. For a bath temperature of 190° C no cesium effects were discernable in the discharge. ]for bath temperatures of 225° to 245° C and a 10 torr. pressure of helium the discharey color was violet and the cesium doublet at 4550 A was very prominent in a small spectroscope. With the cesium, the total voltage was 800 V and the current 100 ma Without the cesium, the voltage was 1800 V at 100 ma. current. At the above bath temperatures, the cesium vapor pressure is a few tenths of a torr. However, it is likely that only a fraction of this amount was carried by the helium gas. The introduction of 0. 1 torr. of Nz quenched the cesium doublet radiation and changed the color of the discharge to orange, which is characteristic of the Nz bands. The voltage rose about 100o/o. If instead, 0. 1 torr. CO2 was injected into the He + Cs mixture, the voltage rose to a few K. V., the discharge color turned white and the cesium doublet lines disappeared. No laser radiation was obtained until the CO2 pressure reached 1 torr. at which time 5 watts output was measured at an identical voltage as in the absence of cesium. The same laser power output at the same voltage and current level was obtained in a 1:1:8 mixture of Nz, CO2, and He, with and without the cesium. A wide range of gas mixtures were tested with identical results. Raising the helium pressure to 50 torr •. resulted in a constricted unstable discharge. There are two possible explanations for these results. One, the COz attacks the cesium vapor, as it does the cesium liquid. This is not too probable since in the shock tube experiments, cesium and COz are mixed for periods of a few minutes, yet successful experiments have been performed. Nevertheless, a special experiment will be performed to determine the presence of a cesium - CO2 reaction in the ga.seous state. Two, the gas and electrical characteristics of the static diode laser in the presence of cesium are not in the proper range to produce a population inversion. One would, nevertheless, expect that cesium should p;roduce some improvement in the production of a population inversion since the electron energy in a cesium plasma is in the energy range where the CO21aser transition cross-sections at a maximum. This question can be resolved in part by theoretical calculations, Finally, it is to be noted that the laser diode experiment verifies the theoretical discussion in that it shows that molecular quenching of the cesium excitation is a dominant factor in this type of gas mixture.
3. The supersonic, shock tunnel - MHD facility has been described elsewhere(4,5). For the present study a differen MHD channel was used. Its geometry was the same as the previous channel. The nozzle throat area was 9. 8 x 15 cm 2. the supersonic nozzle length was 20. 3 cm, the generator length was 75 cm, and its entrance and exit areas were 9. 8 x 16. 5 cm 2 and 19. 1 x 25. 4 cm 2 respectively. The channel divergence was linear. The only difference was that in the present channel heated tungsten wire electrodes were used in the generator. The wires were suspended in the stream, outside the electrode wall boundary layer, in a direction parallel to the B field. There were 37 electrode pairs at an average axial spacing of 2 cm. The cathodes were heated to 1200° - 1800°K. However, they were not thermionically emitting. In cesium seeded, argon experiments electrode currents of 100 Amp, corresponding to an ave 2age electrode surface current density of 30 A/cm, were obtained. As already notec, the present eleclirieaJ.ly loaded M HD generator operated at Mach numbers of 1.5 to 2, which is considerably below 4. 5 value needed for the laser gas mixture. To simulate the high Mach number condition, the MHD expe·riments were performed at the downstream end of the generator where to open-circuit Mach number ranged from 2. 5 to 3.
4. The first experiment, Run ,ff 291, was performed in mixture A (with N 2) at zero magnetic field, An average current of 17Amperes per electrode was obtained during the 4 m. s, test time at all seven electrode pairs (see Figure 3), The currnt level was free of high frequency fluctuations, Figure 4 shows the measured equipotential distribution for this run. The measurements were obtained with probes located on the insulate,r wall, Their axial spacing was 12 cm. Thus, the spatia,l resolution for this series of experiments is not too good. Neverthless, for the zero B field run, the equipotentials in Ile inter electrode region are reasonably parallel. Thus, one could assume that in this case the discharge is not significantly blown downstream. This was verified from continuum radiation measurements which showed that the radiation in the interelectrode region was 100 times greater than the upstream radiation, and two times greater than the radiation 5 cm. downstream of electrode ,/134. One also notes in figure 4 that the electric field gradient increases near the electrode wall 1;1, indicating the existance of curJ"ent constriction. This constriction is to be expected since the current carrying electrodes were spa,ced about 4 cm. apart. From figure 4 one obtains for Run #291 an average electric field of 110 V/crp. From figure 3 one computes an average curreqt density, based on the electrode wall area, of 0, 24 amp/cmZ. From E and j one obtains a scalar conductivity of 2.1 x 10-3 Mhos/cm. Using the Maxwell averaged rose-sections (IS) one computes from a an Ne of about 2 x 1017m-3at Te of zooo°K to 3000°K. It should be noted that due to the large molecular concentration the electron density cannot be corpputed from the Saha equation at Te. Using the electron energy equation (Eq. 1) one obtains an experimental O of 2000, The computed6EFF for this gas miJ"r (see eq. 1) is about 20. A detailed analyses as described in section ·2. 2 was performed. The sum of the elastic energy loss.vibrational excitation loss of COz, vibrational excitation loss of Nz, Cs guenching loss to N2 and CO2 was compared
5. to j 2/cr.jZ/cr was equal to 26 watts/cm 3, The sum of the energy losses was 4. 4 W/cm 3 at Te= 30000K, 2 W /cm 3 at Te= 2S00OK, and 0, 8 W /cm 3 at Te = Z000OK. At Te = 2000°K, CO2 vibrational excitation is the dominant energy loss, accounting for 70% of the electron energy loss, At Te • 3000°K, quenching by CO2 and Nz of the cesium