where c is the velocity of light. Here, tc = 5 x 10-8 s. The quantity T, the stimulated emission rate, is equal to n<jecAN where ae is the stimulated emission cross-section. Then where Xe is the emitted wavelength, A (~1) is the Einstein coefficient for spontaneous emission from the upper laser level, and AXe is the emission bandwidth. With AXe taken as the bandwidth for Doppler broadening at 300 °K, then ae = 1.6 x 10~17 cm2. Equations 2 through 9 were solved by computer, and are compared with the experimental results of Zapata in Figs. 2a and b. In the experiment, the light intensity from the xenon lamp was measured with a photodetector, and varied approximately as IV. COMPARISON WITH EXPERIMENT The idealized model differed from the experiment in a number of ways. The theory assumed absorption occurred one-dimensionally in a thickness d. Experimentally the pumping source was a xenon lamp placed parallel to the laser tube. Its light was focussed onto the laser tube axis by a reflecting cylinder of elliptical cross-section. The illumination increased towards the axis, but as the source was distributed and the reflector was not perfectly elliptical, the maximum value of the pumping photon flux density was uncertain. The pumping density was equated to C times the solar radiance. The values of the equivalent C in the experiments were probably between 2000 and 40,000 for the various runs. There were also differences in the spectral distribution of the xenon lamp and the solar spectrum assumed. Figure 2 compares the measured light output from a 1-m long IBr laser pumped by a xenon discharge lamp powered from a capacitor bank, with the computer solution of Eqs. 2 through 9. The IBr pressure was 3 torr and C was about 5000 in the experiment, which are the values assumed in the calculations. The overall shape of the experimental and calculated laser output are similar. Both start about the same time, and consist of a sharp spike followed by a tail. The ratio of spike to tail is higher in the calculated version, and the oscillations in the theoretical wave shape are not evident in the experiment. Estimates of the time constant of the InAs photodetector circuit showed it was high enough to reduce the high frequency components of the signal. The duration of the experimental pulse for this and all other runs was always shorter than the calculated values by a factor of about two. We believe this was due to the lasant being heated near the axis. Rough calculations of the IBr gas temperature, assuming C = 5000, indicated a temperature rise of several hundred °K in 50 /zs. The temperature rise would cause dissociation of IBr and deplete the lasant. This view is consistent with an experimental measurement of IBr density on the laser axis during a pulse, which showed a drop of about 50%, a value too high to have been
RkJQdWJsaXNoZXIy MTU5NjU0Mg==