Electron accelerators with energies of 15-30 MeV are widely used as sources of not only fast electrons and bremsstrahlung but also neutrons. Fast neutrons with average energies -1.5 MeV are secondary particles produced in converter--targets via the reactions (3', n) and (3,, f). As a rule, heavy elements (Pb, U, and Pu) serve as the converter material [1]. The neutron yield for 30 MeV electrons equals several percent of the intensity of the incident electrons, and on account of neutron-induced fission it is approximately two times higher for fissioning elements than for lead. From the standpoint of radiation safety, a lead converter is preferable, since high residual activity, which is due to fission products, accumulates in the f'msioning elements. For 30 MeV electrons with an average current of 10/zA, the maximum thermal-neutron flux density in an water neutron moderator with a lead converter at the center equals 2-109 sec-l.cm -2. This is 10 thousand times lower than in a standard reactor, but the simplicity of a neutron source based on electron accelerators and the possibility of obtaining resonance and fast neutrons opens up wide possibilities for using such a neutron source to solve scientific and applied problems [2][3][4][5][6][7][8].For neutron activation of samples, cavities with a volume of several cubic centimeters are ordinarily used in the moderator. At the same time, if a large cavity is present in the moderator, the range of problems that can be solved can be greatly expanded and not only large solid samples but also liquids can be analyzed, which is difficult and sometimes impossible to do in reactors in the case of corrosive, hot, or volatile substances.The difficulty of producing a large cavity in the moderator lies in the fact that the presence of any channel unavoidably results in neutron leakage and a large distortion of the neutron field. So, a moderator always contains a channel for transporting electrons to the converter, which, as a rule, lies at the center of the moderator. According to the experiment of [9], this decreases the flux near the converter by approximately 30% compared with a homogeneous moderator and strongly distorts the thermal-neutron field. In addition, a photoneutron converter is not a quasipoint source of fast neutrons [9, 10]; this should also result in a distortion of the neutron field. In addition, in a hydrogen-containing moderator the neutron flux density depends strongly on the distance to the converter [9], and for this reason, one would think, the neutron distribution in the large cavity should also be just as nonuniform, which in this case completely negates it.However, these arguments are actually groundless. As far as the extended nature of the photoneutron sources concerned, the calculations of [6] showed that the nonuniformity of the neutron distribution over the volume of the converter affects very little the computational results for the flux density of slow neutrons in the moderator, since even in polyethylene the neutron moderation length -5 cm appro...
