As noted above, NPLs include lasers with active media that are excited directly using nuclear radiation, or with the use of intermediate nuclear-optical converters. There is also another method of pumping gas lasers using nuclear radiation, in which this radiation is used only for pre-ionization of the laser medium. In this case, the nuclear radiation performs an auxiliary function (creating the conditions for uniform volume pumping the gas medium), while direct pumping occurred through gas discharge. Such lasers (sometimes they were called electroionization lasers), which radiate at transitions of CO2, CO, XeF molecules and of Xe and Ar atoms, are not classified as NPLs, and are not considered in this book. Information about them may be found, for example, in the survey papers [54, 56, 61].

There are three basic sources of nuclear radiation which can be used to pump NPLs:

1. Nuclear explosive devices.

2. Radioactive isotopes.

3. Laboratory neutron sources.

At present, direct nuclear pumping is implemented for gas laser media in which the populating of lasing levels occurs as a result of processes occurring in a high — pressure low-temperature plasma formed by ionizing radiation. Sometimes this plasma is called nuclear-excited plasma.

A plasma formed during the stopping of high-energy charged particles in a gas at atmospheric pressure is a thermodynamically non-equilibrium plasma. A plasma of this type, in which the concentration of electrons substantially exceeds its equilib­rium value, is called a supercooled or recombination plasma [1, 2]. In a nuclear — excited plasma (in contrast to a gas-discharge plasma with an ionization type of non-equilibrium), the majority of electrons have energies close to thermal, and recombination processes play a key role in the kinetics of the charged particles.

While conducting investigations associated with development of NPLs and the study of their characteristics, it becomes necessary to determine the total energy absorbed in the gas mixture, and the spatial distribution of the energy deposition through the gas volume. Such data make it possible to assess and to optimize the physical characteristics, including the laser efficiency, the threshold pumping power, the optimal pressure for the given geometry and composition of the gas mixture, the space-time structure of the optical inhomogeneities, and so forth. Both the total energy absorbed in the gas mixture and the specific energy deposition in each point depend on the geometry of the laser channel, the composition and density of the mixture, and the composition and thickness of the active layer containing the fissionable material. Usually lasers excited by fission fragments have the shape of an elongated circular cylinder or rectangular channel (see Chap. 2). In the first case, the active fissionable material is applied in a thin layer (5U = 1-10 qm) to the internal surface of the cylindrical laser body; in the second, to two opposing parallel internal surfaces of the rectangular channel.

NPLs are characterized by transverse dimensions comparable to the range of the fragment in the gas laser mixture. Therefore, the profile of distribution of the specific energy deposition in the transverse cross section of the laser markedly depends on the transverse coordinates. The dimensions of the lasers in the direction of the optical axis of the system exceed many times over the range of the fission fragments in the utilized mixtures. This means that in the overwhelming portion of the volume of the laser channel, the longitudinal distribution of the specific energy deposition is determined only by the shape of the longitudinal distributions of the gas density and neutron flux irradiating the laser.

One of the chief deficiencies of early methods [8—14] of calculating the energy deposition of heavy ions, including fission fragments, was that the density of the excited gas was assumed to be uniform. In principle, the Monte Carlo method makes it possible to calculate the energy deposition with allowance for the non-uniformity of spatial distribution of the medium density.

Study [12] employs an alternate “two region” approach to address this problem. However, the spatial distribution of the energy deposition is given in histogram form [14] (calculation of the energy deposition is done for a finite selected volume in the neighborhood of the given point), which creates evident difficulties for restoration of a continuous dependence of the target value on coordinates. Breaking up each volume element into smaller ones, with the goal of clarifying the behavior of the spatial dependence, requires a significant increase in the calculating time and operational memory of the computer to retain the desired statistical accuracy of the calculation.

Matyev [15—17] developed an analytical method of calculating the energy deposition to media with an arbitrary distribution of density of the excited gas.

The space-time distribution of the specific power deposition to the gas F r. t is determined as in [15, 16]

, ч, 4 R P Г. t, 4

f(7’1) = . OR * p0 ‘(?4 (u)

Here q is the average fission power for the uranium-containing active layer per unit of volume of this layer; R0, R; are the mean range of a fission fragment in the gas

and in the material of the uranium-containing layer respectively; p ^ r, t^ is the density of the gas at the point in question at the time t; p0 is the average (non-perturbed) gas density; and Z^r, t^ is a dimensionless functional, defined by the distribution of gas density and the geometry of the cell. The correlations for r, t^ obtained in [15, 16] come to operations involving double integration.

Although they have a very cumbersome appearance, it then becomes possible with comparatively little effort to conduct a precise calculation of the energy deposition in non-uniform media with a finite transverse dimension.

As in other studies [1—3, 8—14] dedicated to calculations of the energy deposition when there is slowing down of heavy ions, including fission fragments, in the gas, the dependence of the energy E, of a decelerating particle of given mass on the range x in the model [15—17] is presented in the form

Ei = Е,- o(1 — x)n. (7.8)

The exponent n determines the “slowing down law”: for n = 1, one speaks of a “linear law” for slowing down; for n = 2, a “square law,” for n = 3/2, a “3/2 law.” The linear law is usually used to describe the deceleration of light ions—protons and a-particles [1]. It is rarely applied to fission fragments [10]. Usually n = 2 and n = 3/2 are used to describe the deceleration of fission fragments. Both values of the exponent n were discussed in study [18]. The square law of deceleration [18], which is also the one used in most studies, is considered the most adequate for fission fragments.

A comparison of results of calculations by the methods of [15—17] with the published results of calculations by the Monte Carlo method [14] shows their close identity. Figures 7.3 and 7.4 show the results of calculation of the efficiency for absorption of the kinetic energy of fission fragments in a gas medium. Results are for both cylindrical and flat uranium layers. The figures include the designations: e, the ratio of fission fragment kinetic energy absorbed in the gas medium to their total kinetic energy, released in the uranium layer; D0 = d/R0, the normalized transverse dimension of the gas volume (d is the diameter in the case of a cylindrical uranium layer or thickness of gas in the case of a flat layer); and D; = 8U/R1, the normalized thickness of the uranium-containing layer.

Є, %

50

40 30 20 10 0

The ranges of the uranium fission fragments in certain materials are provided in Table 7.1. To describe the mass spectrum of fission fragments, frequently an approximation is used, which assumes that when a 235U nucleus is split by thermal neutrons, two fragments are formed. These are a light and heavy one, which have average masses of 96 amu and 140 amu, and kinetic energies of 100 MeV and

68 MeV respectively. An empirical relationship [14] can be used to determine the mass ranges R (mg/cm2) of fission fragments in various media:

R = (0.0391 + 0.0202As/Z]/2) Е3/3, (7.9)

where As and Zs are the mass number and the charge of atoms of the medium; Ef, MeV is the initial kinetic energy of the fragment. Sometimes a simpler relationship is used, which makes it possible to determine the so-called mean mass range R without dividing the fragments into light and heavy [14]:

R = 0.755 + 0.388AS x Z;1/2. (7.10)

Prior to development of the computational methods [15—17] during the early computational investigations of the dynamics of development of optical inhomo­geneities in sealed laser cells, the influence of changes in gas density on the profile of distribution of the energy deposition into the gas was taken into account by tying the specific energy deposition sources to Lagrangian coordinates [19—21]. But the distribution itself of these sources was determined by the method described in [14]. In this case, the power deposition of the decelerating fragments in a gas portion of fixed mass, calculated relative to the intensity of uranium nuclear fissions, does not depend on the time, and remains equal to its initial value in a non-perturbed gas. That approach provides a precise description of the spatial distribution of the energy deposition only for an infinitely extended flat geometry [16, 22]. For other types of geometry of the cell, in particular for cylindrical or rectangular laser channels, this technique does not adequately allow for the influ­ence of redistribution of density on the energy deposition. Thus, in the case of cylindrical cells with an axially symmetrical distribution of medium density, in which the diameter is equal to the mean range of a fragment in a non-perturbed gas, or exceeds it, with an increase in the energy deposition, the gas particles are concentrated toward the axis of the channel, where when there are very large energy deposition (significantly exceeding those realistically attainable even in the most powerful pulsed reactors), a region inaccessible to the fragments may form [23]. In [23] this phenomenon is called “axial screening.” Nonetheless, special investigations showed that for cylindrical cells, with pumping durations of t1/2 < 3 ms, fixation of the energy deposition in Lagrangian coordinates makes it possible to obtain gas temperature and density deviations for any point of the volume which are no more than 10 % of the corresponding values obtained in calculations with precise allowance for the influence of redistribution of density on the specific energy deposition [16, 21].

10.1 Main Types of Stationary Reactor Lasers

Chapter 6 gives a brief overview of designs for nuclear laser setups with a pulse duration of ~10 ms (OKUYAN and LIRA setups). Their operation is based on the “initiating reactor-subcritical laser block” scheme. The literature has repeatedly suggested different design schemes for stationary nuclear laser facilities or reactor — laser (RL) based on gas laser media. The term RL was apparently first proposed in article [1], where the potential of removing energy in the form of laser radiation from the reactor core during the relaxation stage was noted in principle. The most interesting version of the RL is a device where, as a consequence of spatial overlapping of the nuclear fuel and laser medium, a significant portion of the nuclear energy is converted into laser radiation bypassing the intermediate thermal energy stage.

The idea of creating a stationary nuclear-laser facility based on using a gas medium was expressed in 1964 before the appearance of the first nuclear-pumped lasers (NPLs). In this study [2], a reactor core was examined that consisted of 1,500 aluminum tubes surrounded by moderators. These tubes had layers of 235U depos­ited on their inner surfaces and were filled with a He-Ne mixture at ~10 Torr of pressure. There are some schemes of stationary RLs that are differentiated by the nuclear fuel used for reactor core operation and pumping of the laser medium (Fig. 10.1): (a) a gas compound UF6, uranium vapor, or uranium aerosol particles; (b) thin layers of uranium or its compounds on metal substrates.

Interest arose in UF6, because it is practically the only gas-phase compound of uranium at low temperatures, and the energy deposition efficiency for fission fragments in the “UF6 + laser gas” mixture approaches 100 %. In addition, a series experiments [3, 4] were carried out that show the potential for creating gas-core nuclear reactors with comparatively low critical masses of UF6 (~6 kg).

Thus, there is currently no doubt of the potential for creating gas-core reactors based on UF6, although these devices are rather complicated and dangerous.

© Springer Science+Business Media New York 2015 373

S. P. Melnikov et al., Lasers with Nuclear Pumping,

DOI 10.1007/978-3-319-08882-2_10

Laser beams

Fig. 10.1 Schematics of the main variants for a stationary RL [6]: (a) a gas-core reactor whose core is a mixture of uranium fuel (UF6, uranium aerosol, or vapor) and laser gas; (b) an RL using thin-layer uranium fuel

However, when choosing laser gas media together with UF6, a number of difficul­ties arise, which are caused by the chemical aggressiveness of UF6 and its radiolysis products, high losses in the active medium due to absorption of the laser radiation by UF6 molecules, and the high quenching rate of the excited atoms and molecules in collisions with UF6 molecules.

As laser media for gas-core RL based on UF6, these mixtures have been suggested: UF6-CO2-N2-He [6], UF6-He, UF6-TlF-F2 [7], UF6-He-Xe [8], UF6- Ar-Xe [9], etc. However, there are currently no laser media using UF6, for which the feasibility of pumping by nuclear radiation would have been established experimentally. Experimental [10] and theoretical [11] studies have shown that, at the present time, the most effective laser media with nuclear pumping at the infrared transition of atoms M = Xe, Kr, and Ar may not operate at UF6 concen­trations higher than 1 % due to a reduction in the ionized atom M+ concentrations and the high rate of collisional quenching of the excited atoms M* by the UF6 molecules. One paper [9] concludes that designs of gas-core RLs using UF6 are apparently unrealistic because it is difficult to provide the conditions for laser operation with sufficiently high concentrations of UF6, which are necessary to attain reactor criticality. The authors of study [9] came to the same conclusion when analyzing the possibility for creating an RL using aerosol fuel with uranium particles several micrometers in diameter.

The design of an RL using uranium vapors was not examined in detail. In the authors’ opinion [12], the main difficulty for creating such facilities is the necessity to maintain the active medium at a high temperature (~4,000 K). This substantially complicates the construction of the device and severely limits the choice of laser medium.

At present, the most realistic designs are heterogeneous RLs using thin-film uranium fuel. The core of this RL is, in essence, a specific quantity of laser cells with uranium layers appropriately placed in a neutron moderator matrix. Calcula­tions show that, with the appropriate selection of components, this system has a sufficient neutron multiplication factor when only uranium layers inside laser cells are used, that is, the conditions for the operation of the RL are provided without utilizing additional fuel (uranium). The number of laser cells may vary from a hundred to several thousand. The total amount of uranium may vary from 5.7 to 40­70 kg, and characteristic linear dimensions are 2.5 m.

Figure 10.2 shows the two main alternatives for laser cells: (a) a cylindrical cell capable of longitudinal gas circulation relative to the axis of the cell; (b) a rectan­gular cell with transverse gas circulation relative to the axis of the cell (in the direction indicated by the arrows). Longitudinal gas circulation is possible with low power deposition and slow (<10 m/s) gas velocities. At high power, transverse circulation with the energy transfer to internal radiators is more effective (Fig. 10.2b). Some characteristics of laser cells with transverse gas flow are examined in Sect. 9.3 of Chap. 9. This laser cell design was proposed at the All-Union Scientific Research Institute for Experimental Physics [VNIIEF] [13, 14] and verified experimentally on the LM4/BIGR complex (see Chap. 6, Sect. 6.1).

Stationary RLs using thin-layer uranium fuel have been considered in Russia [15—18] and abroad [2, 6, 19]. The main characteristics of these systems and the issues that arise with their design are considered below in examples of the RLs that are discussed in studies [15—18].

During the earlier work on NPLs in the United States, a number of groups provided important insightful theoretical studies that clarified the physics of this new field. This work generally tackled one of the fundamental issues involved—the details of the source term created by the incoming radiation, the transfer of energy to the laser media, the reaction kinetics of the species in the lasing medium, and the optics associated with the laser, especially special effects such as thermal blooming encountered in such lasers. There is a complex matrix where these issues overlap with applications and a variety of possible laser media such as noble gases, excimers, molecular gases, liquids, etc. This matrix is far from full, so much work remains to be done to fully understand all of the possible routes that NPLs may take. The main goals are to obtain improved efficiencies, and ways to take advantage of the large energy but relatively low peak powers available with this form of pumping the laser.

Mark A. Kushner reported a number of important theoretical studies of nuclear — pumped lasers while at the University of Illinois. A summary of his early studies with NPLs is presented next in the section.

Nuclear-pumped lasers are characterized by high-threshold neutron fluxes and low gain. The closer the fissile gas can be coupled to the actual lasing process, the more efficient the laser will be. One approach is to seek laser systems such as chemical lasers that have low threshold power requirements. Kushner presented a model for a nuclear-pumped HF laser using a 235UF6-H2 gas mixture [30]. This model predicted a peak gain of about 50 %/m and a threshold neutron flux of < 1014/ cm2. He provided recommendations concerning the optimum use of this unique system. However, as discussed in other work by Wilson and DeYoung [31] and Maceda [43], this and other lasers requiring 235UF6 are difficult to achieve experi­mentally, so the practicality of such lasers remains to be determined.

Kushner also sought other classes of lasers well suited to nuclear pumping. In study [32], he proposed study of a class of NPLs in which the laser levels are populated by negative ion-positive ion neutralization. He analyzed two particular systems, 3He/N2O/O2 and 3He/N2O/N2. Laser action was predicted in atomic oxygen and nitrogen as a result of O~ — O+ and O~ — N+ neutralizations. The source of negative ions is a dissociative attachment to N2O. A threshold neutron flux of «1015 cm-2 s-1, peak power of >500 W/cm2, and a fission power to laser power conversion efficiency of «0.4 % were predicted. The electron density in the attachment process was evaluated, and the highly selective nature of the ion-ion neutralization in producing product-excited states.

Kushner studied the excitation source term along with average energies going into excitation and ionization [W-values] for nuclear pumping with fission fragments. This is particularly important to nuclear pumping because of the unique highly non-Maxwellian characteristic of the electron energy distribution created during the slowing down process. In study [33] Kushner and Moratz described a method for calculating excitation and ionization rates in a plasma generated by the slowing of fission fragments in a gaseous medium such as neon. The energy distribution of the fission fragments as they slow down and the excitation and ionization of neon due to collisions with the fission fragments was calculated. Effective W values for ion­ization and excitation of neon directly by the fission fragments of uranium were found to be about 71.6 and 11 eV, respectively. To illustrate the non-equilibrium characteristic of this process, they compared the source function for electrons produced by an e-beam to the source function of electrons produced by heavy ions which are shown to have had a lower average energy.

In another study [34], Kushner and colleagues presented a model for the heavy — ion pumping of a XeF (B! X) laser by uranium fission fragments. The model is a self-consistent accounting of the generation and transport of the fission fragments through the fission foils, slowing of the fragments in the gas, evolution of the secondary-electron-source function and distribution, and the XeF laser plasma kinetics. By stimulating the same quantities for an e-beam-pumped plasma, direct comparisons can be made for laser performance. Kushner and colleagues found that the secondary-electron source generated by the e-beam was more energetic than that for direct ionization by fission fragments due to a more favorable mass ratio for momentum transfer collisions with orbital electrons. This difference in the electron — source functions significantly affects W values (energy/ion pair) and excitation fractions. The impact on laser performance, though, is not large due to the high efficiency of channeling deposited energy to the upper level in XeF lasers. For conditions typical of fission fragment excitation (power deposition 1-3 kW cm-3, pulse length «200 qs), this model predicted 10-15 % higher gain with e-beam excitation than heavy-ion excitation.

Later, Kushner et al. pointed out that excimer lasers excited by electron or ion beams having energy deposition of 100 s J/1-atm over many microseconds experi­ence a temperature rise of hundreds of degrees (K) [35]. The increase in gas temperature can significantly impact both the kinetics and spectroscopic para­meters. In this study, Kushner and colleagues considered the high-temperature (<900 K) plasma kinetics and absorption in He and Ne buffered gas mixtures for particle beam pumped XeF lasers. They found both gain and absorption depended differently on gas temperature in these mixtures (absorption decreased in He mixtures, but increased in Ne mixtures). The differences were attributed to a reduction in diatomic absorbing species with increasing temperature and differences in the temperature dependence of the optical absorption cross-sections for NeXe+ and Xe2+.

Later, Kushner and colleagues discussed the effects of He addition on the performance of fission fragment excited Ar/Xe lasers [36]. This was viewed as an important laser system because both Russian and Sandia researchers were studying such systems experimentally. The intrinsic power efficiency of the atomic xenon laser depends upon the electron density because of the mixing of the laser levels by electron collisions while the electron density in high-pressure particle-beam excited plasmas increases with increasing gas temperature. Therefore, in order to reduce the amount of electron collisional mixing when operating at high-energy loadings (>100 s J/1-atm), mixtures having a high-heat capacity are required. In particle — beam excited Ar/Xe mixtures, which typically yielded the highest intrinsic laser efficiencies, increasing the gas pressure to increase the heat capacity was not always practical due to the high-stopping power of the gas mixture. For this reason, Kushner and colleagues investigated adding He to Ar/Xe mixtures for a fission — fragment excited atomic xenon laser. Adding He increased the heat capacity without appreciably perturbing the favorable kinetics resulting in efficient opera­tion of the laser in Ar/Xe mixtures. They found that when adding He to Ar/Xe mixtures the dominant laser transition wavelength switched from 1.73 to 2.03 ^m without significantly decreasing the efficiency. The laser pulse length also increased, an effect attributed to a lowering of both the electron temperature and gas temperatures.

Subsequent studies of the infrared atomic xenon laser (5d! 6p) by Kushner et al. found it to be an attractive candidate for fission fragment excitation, providing low-power deposition (l—100 W cm-3), long pulse lengths (1-10 ms), and high — energy deposition (100’s J/l) [37]. The optical gain at 1.73 and 2.03 ^m was measured in a reactor-excited xenon laser and yielded values exceeding 0.03­0.05 cm-1 at power depositions of less than 10 s W/cm3. The gain was also found to rapidly terminate before the peak of the pump pulse for some experimental conditions. A computer model was developed to predict gain in fission-fragment- excited xenon lasers. It was found that the termination of gain was most likely attributable to gas heating which increased the electron density, leading to electron collision quenching. The specific dependence of gain on pump rate suggested that a reduced rate of recombination of molecular ions with increasing gas temperature was partly responsible for this behavior.

In another study pertaining to this system, Kushner and colleagues predicted gain of the 1.73 ^m [5d(3/2)1-6p(5/2)2] and 2.03 ^m [5d(3/2)1-6p(3/2)1] atomic xenon transitions for gas temperatures between 290 and 590 K [38]. Fission — fragment excitation was used to generate the gain in Ar/Xe, He/Ar/Xe, and Ne/Ar/Xe gas mixtures at a pump power of 8 W/cm3. For constant gas density, the gain exhibited an approximate T~ngas dependence with n between 2 and 3. The primary reactions responsible for the temperature dependence of the gain were identified as dimer formation, dissociative recombination, and enhanced electron collisional mixing of the laser manifold due to an increased electron density.

In a following study [39], Kushner and Shon discussed excitation mechanisms and gain modeling of the high-pressure atomic Ar laser in He/Ar mixtures. The high-pressure (>0.5 atm) atomic Ar laser (3d! 4p) oscillated on four infrared transitions (1.27-2.4 qm). Quasicontinuous oscillation on the 1.79 qm transition was obtained using electron beam and fission fragment excitation over a wide range of power deposition and gas pressure. In this regard, a computer model was developed to investigate excitation mechanisms of the Ar laser. Results from the model suggest that the upper laser level of the 1.79 qm transition [Ar(3d[1/2]1)] is dominantly populated by dissociative recombination of HeAr+. In contrast, the dissociative recombination of Ar+2 was believed to predominantly produce Ar (4 s) states. Electroionization from Ar metastables at moderate to high pump rates was likely to be responsible for the high efficiency of the Ar laser. Gain and laser oscillation were discussed and compared to experiments for He/Ar gas mixtures using various Ar mole fractions and total pressures. These results showed that the optimum Ar mole fractions in He/Ar mixtures are ~0.1-5 % for quasicontinuous pumping.

In a study aimed at understanding the pumping source term for coated wall type lasers, Miley and colleagues studied the transport of heavy charged particles produced by the 10B (n, a) nuclear reaction using a mean-range straight-flight model [23]. The slowing down of these particles in a gas adjacent to the coating where they are born is described in terms of their flux energy spectrum, scalar flux, average energy, and energy-loss rate. These results were used in a plasma kinetics model which was then compared to measurements of metastable excited state densities in helium and neon plasmas created by the heavy charged particles. The resulting plasma kinetic model was shown to agree reasonably well with measure­ments involving 10B-coated gas containers inserted in the Illinois pulsed TRIGA reactor. Ne(1S3) and Ne(1S5) densities ranging from 2 x 108 to 5 x 109 atom/cm3 and 3 x 109 to 4 x 1010 atom/cm3, respectively, were obtained in neon at 16.5 Torr with fluxes from 1012 to 1015 n/(cm2 s). Maximum He(21S) densities of about 10~4 atom/(cm3 unit flux) were obtained in helium at pressures above 200 Torr. Such measurements, not previously available, were the first attempt to directly evaluate excitation processes in a plasma created in a nuclear reactor core.

In a related study, P. E. Theiss and Miley developed a method to evaluate the spatial dependence of the energy spectrum of charged particle currents entering a process fluid from a bordering solid fuel region [40]. This enabled calculation of spatially dependent ionization-excitation rates including the general case where the formation rate varies with particle energy. Non-dimensional plots permit appli­cation of the results to various fluid-fuel combinations. Also, with adjustment of two parameters, these results can also be used for various particles including fission fragments, betas, alphas, and protons. The example of excitation of helium by alpha irradiation is also presented in some detail in study [40].

Another concept that is important to future NPLs involves uranium-pumped nuclear plasmas. In a study focusing on this, Maceda and Miley developed a model for the atomic levels above ground state in neutral, U0, and singly ionized, U+, uranium based on identified atomic transitions [41]. They found 168 states in U0 and 95 in U+. A total of 1,581 atomic transitions were used to complete this process in this study. Atomic inverse lifetimes and line widths for the radiative transitions as well as the electron collisional cross sections were presented. They identified 285 transitions in U0 and 278 transitions in U+, and during the course of the research employed 1,581 atomic transitions. The final compilation of these data is given in study [41].

As one approach to the use of an uranium plasma for NPLs, Y. R. Shaban and Miley proposed a hybrid uranium-core-laser reactor (UCLR) for advanced space missions that could radiantly transfer energy to a propellant or alternately active laser action [42]. The propellant mode would be employed in the phases of the mission requiring a higher thrust. However, for the bulk of the travel, the propellant would be turned off and the ultrahigh specific impulse laser mode of operation would be employed. The concept was further developed, research and development issues were identified, and the different operational modes were discussed in study [42].

In another study concerning uranium plasma pumping, Maceda and Miley [43] computed the opacity and radiative-energy current due to line radiation was calcu­lated for a U235 plasma with a temperature range 5,000-8,000 K. Also, a variation in the neutron flux of 2 x 1012 to 2 x 1016 neutrons/(cm2 s) was considered. The plasma formed a cylinder with a diameter and height of 1 m. Because the electron states in uranium lie below 5 eV, recombination is the principal excitation mecha­nism. Inversions were found above 6,000 K and, at all temperatures, the line radiation at line center is greater than the corresponding blackbody radiation. An example of this is the 28,763-5,762-cm~1 transition in neutral uranium, where the Planck function at 5,000 K is 6.49 x 10-6 ergs/cm2, and the calculated radiative — energy current is 1.492 x 10~4 ergs/cm2. Negligible changes in the radiative-energy current were observed for changes in the neutron flux at a given temperature. As a benchmark, it is noted that the opacity at 5,000 K agrees well with earlier calcu­lations done by Parks and colleagues [44], while recombinational excitation explains the variation in the opacity with temperature.

Beam profile control is an extremely important problem for NPLs due to the lensing effect caused by heating during pumping, especially for a wall coating type system. Thus several theoretical studies of this effect were performed at Illinois [45]. NPLs potentially offer an attractive method for high power laser applications such as a space power beaming. However, thermal gradients created by the pumping of the ion medium produce a dynamic “thermal blooming” (lensing effect due to temperature profile effects) which must be understood and controlled for accurate beam focusing. In paper [45], Petra and Miley performed basic experi­mental studies of thermal blooming under a variety of conditions presented. It was found that the laser pulse length could be severely restricted by thermal blooming, but some kinetic control is possible by changing the driving pulse shape. One promising approach found for reducing the effect was to employ a combined volume-wall pumping technique. Petra provided calculation of optimal coating and 3He concentrations to achieve minimum thermal blooming.

In study [46], Petra and Miley considered the issue of control of the extracted beam quality of a NPL. One possible control method involved changing the spatial power deposition profile, which is a function of the gas pressure and the pumping

3 10 235

mechanism. Thus, we proposed a combined He and wall coating (e. g., B or U ) technique. Then, the pump power deposition profile could be adjusted by varying the 3He partial pressure. To examine this technique, measurements were performed with an in-core cell by imaging the radiation-induced fluorescence onto a linear diode array located some distance from the cell. Results from these measurements are presented in reference [46], and other related issues such as gas dynamics, index of refraction effects, and efficient beam extraction from a large-volume cavity are discussed.

In an effort to simplify calculational methods, R. H. Lo and Miley developed an integral balance technique for calculation of the electron energy distribution in a radiation-induced plasma [47]. Results predicted W-values reasonably well and compared favorably with more complicated Monte-Carlo calculations. As expected, results showed that the electron energy distribution was non-equilibrium with a high energy tail but lower average energy. This distribution differs from one using a normal electrical discharge and favors excitation over ionization. This then provides insights into how to seek NPLs which use a laser medium with energy levels receptive to such a distribution, illustrated in Fig. 13.4.

In addition to the source electron energy distribution, other reaction kinetics such as recombination play an important role in laser performance, as illustrated by a Monte Carlo simulation model for radiation-induced plasmas with nonlinear properties due to recombination [24]. This Monte Carlo model employed a piecewise-linearized predict-correct technique. Several variance reduction tech­niques were used, including antithetic variants. The resulting code was applied to the determination of the electron energy distribution for a noble-gas plasma created by alpha-particle irradiation. Results are presented in reference [21] for helium with an electron source rate from 1014 to 1018 electrons/(cm3 s), initial energies from 70 to 1,500 eV, pressures from 10 to 760 Torr, and electric-field-to-pressure ratios from 0 to 10 V/(cm Torr). The low-energy portion of the distribution function approached a Maxwellian for zero electric field and Druyvesteyn’s distribution with an applied electric field. However, above the ionization potential and extending to the source energy, a parabolic-shaped distribution (tail) occurred. This distribution is not too different from the one found in [47], although the effect of recombination on it is clearly discernable.

Due to pump source restrictions, even with a fast burst reactor NPLs typically exhibit relatively long (micro — to millisecond) pulse lengths with modest peak powers but with very high total energy [48]. These pump power restraints seriously limit the choice of laser media, favoring energy storage type lasers. One way to avoid this constraint is to employ a Nuclear-Driven Flashlamp (NDF) for the primary pumped element in the system. The fluorescence from this NDF can then be used for pumping a laser or for other high intensity light applications. The first experimental example of this approach was a 3He-XeBr2 NDF employed by Williams and Miley [49] to pump a small iodine laser. Study [50] discussed issues

ENERGY (»V)

involved in scaling such an NDF up to high power levels. Possible optimum configurations included use of microsphere or fiber pump elements dispersed in the NPF media. Analysis of such possibilities is presented along with consideration of special reflecting surface designs in references [49—51].

Another very important potential NPL laser involves the pumping of the iodine line. Miley reevaluated the existing data in study [52]. At the time, the chemically pumped O2-I2 transfer laser was gaining interest as a high-power laser for a variety of applications. Thus the nuclear-pumped version would be very significant and as already discussed seems to offer an electron energy distribution favorable to this system. Indeed, Garschadin had previously shown that an electrically pumped version faced obstacles due to competition between ionization and unwanted

ionization and excitation. (After years of work, additives to combat this problem have finally enabled an electrically driven version, termed the “e-COIL” [53].) An interest in alternate pumping techniques for some situations remained, however. In study [54] Miley reviewed earlier work on the potential use of nuclear pumping, either for direct pumping of O2 or for pumping of an excimer flashlamp in a photolytic iodine laser system. Such lasers are attractive for space power beaming systems where the compact, high-energy system achievable with a pulsed reactor is desired. For a lower pressure transfer laser system, direct nuclear pumping of O2 to create O2(1A) is the simplest, yet efficient approach. For higher-pressure lasers, e. g., ^ 1 atm, the use of a nuclear-pumped excimer appears to be attractive. In that case, a transfer laser can be developed through photolysis of O3 to generate O2(1 Д), or the flashlamp can be used directly in a photolytic iodine laser system. The choice among these options largely depends on the specific application as this translates into system requirements.

Studies of nuclear-pumped lasers at VNIIEF were continued after the 1975 devel­opment of the laser setup LUNA-2 M, developed to perform experiments in the lower hall of the VIR-2 reactor (see Fig. 2.5), which provides convenient conditions to carry out laser experiments: a sufficient spatial volume to irradiate large-scale

(100-200 cm) laser setups and a circular rail track with remote-controlled cars that deliver prepared laser setups to the irradiation zone. The presence of a hemispher­ical cavity makes it possible to irradiate laser setups in the lower hall with a spatially localized beam of intense neutron radiation; at the same time, the equip­ment outside of the beam is in a significantly smaller flux of the reactor radiation.

The LUNA-2 M setup (Fig. 2.11a) is one of the three main laser setups that make up the experimental complex near the VIR-2 M reactor [26, 33]. When this setup was used, VNIIEF performed the bulk of the experiments to study the characteris­tics of NPL active media in the visible and IR range of the spectrum [27, 28], as well as the problem of extracting the radiation from laser channels [51-53].

The experimental complex also includes the LUNA-2P (Fig. 2.11b) and LYaN — 2T setups [26, 28]. The basic characteristics of these setups are given in Table 2.5. Each setup consists of two laser cells (channels) that are identical in design and in irradiating conditions, making it possible to accelerate the rate of the investigations. The layout of the laser setup under the reactor core is shown in Fig. 2.12. The cars can be moved on the track and clamped under the reactor core remotely (from the control panel) or manually (directly in the lower reactor hall).

The LUNA-2P unit (Fig. 2.13) was developed in 1987 and was used to study laser characteristics and the optical quality of the gas medium in the gas flowing mode. The gas was circulated in the transverse direction relative to the optical axis at a speed of 5-10 m/s. Heat exchangers (radiators) were used to cool the gas— packets 100 cm long made of aluminum or beryllium plates 0.5-mm thick, with 0.5­mm gaps between them.

The LYaN-2T setup, developed in 1991 to study lasers operating on metal vapors and to study the influence of temperature on laser parameters, used quartz laser cells 3.2 cm in diameter with electrical heaters filled with a mixture of the isotope 3Не and laser gas. In this case, the active media were excited by the nuclear

8

12

10

Fig. 2.12 Diagram of laser setup in lower hall of reactor VIR-2 M [28]: (1) reactor core,

(2) neutron moderator, (3) plates with uranium layers, (4) cavity mirrors, (5) beam splitters, (6) focusing lens, (7) laser energy meter, (8) adjustment laser, (9) photoreceiver, (10) light filters, (11) cell gas mixture exhaust and filling system, (12) car, (13) protective shutter, (14 and 15) spectral devices

reaction products 3He(n, p)3H. A detailed description of the LYaN-2T setup is considered in study [54].

The energy parameters of the laser radiation were determined using the laser pulse energy meters IEK-1 and IMO-2 N, various photocells (F-28, FEK-11, FEK-14, and SDF-7) and photoresistors (PbS, PbSe). For measuring the spectra of laser radiation in UV, visible, and near-IR ranges of the spectrum, the spectro­graphs DFS-452 and MS257 were used, and in the mid-IR range, the monochro­mator MDR-2 with a PbSe photoresistor array. The absorption (reflection) spectra of optical materials and cavity mirrors were determined using various spectropho­tometers (IKS-14A, SF-9, etc.).

In the first stages of the investigations, electrical signals were registered using various multibeam oscillographs. At present an automated system is used for recording and processing measurement results using computer technologies.

From 1975 to 2010, around 3,000 pulses were produced on the VIR-2 and VIR-2 M reactors to study the characteristics of gas NPLs. As a result of these investigations, lasing was achieved and studied in the visible and IR ranges of the spectrum at roughly 40 transitions of the atoms Xe, Kr, Ar, Ne, C, O, N, and Cl and the ions Cd+ and Zn+ with excitation of gas mixtures at atmospheric pressure with uranium fission fragments and nuclear reaction products 3He(n, p)3H [1, 7, 27, 28]. In addition, a number of important tasks related to the study of optical and gas-dynamic characteristics of gaseous media were completed, and the information needed to develop CW multichannel NPLs was obtained [15, 16, 37].

Mixtures of rare gases with metal vapors are considered to be promising media for NPLs (see Chap. 3, Sect. 3.1), thus the studies that investigated their spectral — luminescence characteristics are of great interest. Unfortunately, the number of such studies is not great (see Table 4.14), and the information obtained refers to active media operating at comparatively low temperatures (up to ~800 K).

Calculations Based on the Approximation of Infinitely Extended Uranium Layers

If a cell’s working volume has the shape of a parallelepiped bounded by planar walls, and uranium layers are deposited to the inner surfaces of opposing walls, the transverse dimension of which noticeably exceeds the distance between them, and the cell itself has no buffer volumes (a cell of this type was used, for example, in studies [35, 36,41—43]), the transverse density distributions of the gas averaged along the length of the cell can then be calculated in an approximation of infinitely extended uranium layers. Using this approach, all the physical parameters are dependent only upon a single transverse coordinate. The equation system presented in the first section of this chapter for N = 1 describes the gas behavior in this approximation. Here, the specific source distribution, as indicated above, can be fixed in Lagrangian coordi­nates (see study [19]), which considerably simplifies the calculation procedure. The density inhomogeneity formation processes in cells of this type do not qualitatively differ from the processes occurring in cylindrical cells.

NPLs with direct pumping absorb the kinetic energy of nuclear particles (neutrons, Y-quanta, fission fragments) and convert it into laser radiation in the laser medium itself. Especially interesting are laser media with fissile elements uniformly distrib­uted in material (for example, 235U), because, in this case, higher specific energy depositions are possible. The main obstacle on the road to creating condensed — media NPLs is their radiation damage: the formation of displaced atoms and color centers in solid-state lasers (see, for example, [7]); radiolysis [8] and the formation of gas-vapor bubbles [9] on the tracks of nuclear particles in liquid lasers.

Solid-State Laser Media

To our knowledge, there have been no experiments on the direct pumping of solid — state laser media with nuclear radiation of pulsed reactors, although one study [2], for example, was devoted to investigation of the luminescent characteristics of some laser-active materials (ruby, glass with neodymium, CaWO4:Nd3+, Y3Al5O12: Nd3+, etc.) and the radiation resistance of ruby and neodymium glass lasers under the effect of the n, y-ray radiation of pulsed reactors. Some materials had rather high energy conversion efficiencies па, where nA is the ratio of the specific output of luminescent radiation to the specific energy deposition. Thus, with high-energy electron beam excitation the following values of ща were obtained: ща k 10 % (A k 611 nm) for Y2O3 doped with europium [10], and nA = 4 % (A = 900­1,100 nm) for the crystal Y3Al5O12:Nd3+ [11]. The value of nA depends on the type of charged particle. When excited by heavy charged particles (a-particles and fission fragments), the crystal Y3Al5O12:Nd3+ has lower values of ца: nA = 2 % in the range 900-1,100 nm, and nA = 0.28 % for the more intense laser line 1,063 nm [12]. We note that in another study [11], when the crystal Y3Al5O12:Nd3+ was excited by an electron beam, laser action or, at least, superluminescence was obtained.

Yet another study [13] examined the possibility of lasing when neodymium- doped, uranium-containing media is irradiated by the pulsed neutron flux of the BARS-6 reactor [14]. The Y3Al5O12 crystal and GLS2 silicate and phosphate (GLS22, GLS27, and LGS40) glasses were selected as the dielectric matrices. With U concentrations in these media at 5 x 10 cm, the specific energy deposition was 31 J/cm3 (the neutron pulse duration was 140 ps). We note that the possibility of introducing uranium into these media and its influence on laser characteristics have not yet been verified in practice. In the authors’ opinion [13], under these conditions and using direct pumping with fission fragments, a laser effect is possible for active media based on Y3Al5O12 and GLS27 and LGS40 glasses. Of the effects arising during a pumping pulse and resulting to additional losses, the calculations only took into account light radiation scattering on the fission fragment tracks. In fact, a substantially larger contribution to the total detrimental loss factor may be caused by radiation-induced (neutrons, y-ray emis­sion, and fission fragments) light absorption due to formation of color centers (see Chap. 2, Sect. 1). Some media (for example, Y3Al5O12 without 235U impurities) completely lose their transparency on the 1.06 pm line at the maximum of the reactor pulse when the dose rate is about 2 x 105 Gr/s. Therefore, conclusions of study [13] concerning the direct pumping of solid-state neodymium media with nuclear radiation in experiments on pulsed reactors are excessively optimistic.

The use of mixtures of rare gases and metal vapors as active media for NPLs is of interest first of all in connection with the possibility of obtaining lasing in the visible and UV ranges of the spectrum. We note that lasing at a large number of transitions of metal ions and atoms was obtained in recombination mode with the excitation of active media using an electron beam and in a gas discharge afterglow (for example, see [18,109], and the literature cited there). This information helps in the search for NPL active media.

Virtually all of the experimental and theoretical research known to us concerning metal vapor NPLs was carried out in Russia. At VNIIEF, the possibility of using mixtures of rare gases with vapors of metals (Li, Na, Mg, Hg, Cd, etc.) as the NPL active media was considered from 1975 to 1976 [3].