Measurement of spatial density distribution of primary-neutron generation in a pulse booster

Abstract

The spatial density distribution of the generation of primary neutrons by the (e, 3') and (y, n) reactions in an electron injector target and the rods surrounding it, placed in the core of a pulse booster, must be known in order to optimize the shape and location of the target so as to ensure a maximum neutron yield. A method of measuring such a distribution was proposed in 1989 on the basis that after a (y, n) reaction the lszw present in natural tungsten can transform into radioactive 181W, which does not occur in natural tungsten before irradiation; this isotope has a y-ray spectrum, unlike those of other radioactive nuclei, with E. r = 57.7, 66, 136.3, and 152.5 keV and a half-life of 121 and 271 days suitable for prolonged storage and for obtaining information about the distribution of primary-neutron generation [1]. A particularly useful feature is that the range of typical 181W photons in tungsten is less than 1-1.5 mm, less than 1 mm for the most intense lines with E v = 57.7 and 66 keV, which eliminates averaging of the results of measurement over the volume. The long half-life means that the undesirable gamma photons from the short-lived excited states of nuclei of other tungsten isotopes and impurities can be de-excited during weeks and even months after irradiation of the model without any appreciable loss of useful information. Essenee and Preliminary Experimental Verification of the Method. Irradiation of the target with 1017 35-MeV electrons or with 1016 150-MeV particles at the injector exit is sufficient for a subsequent reliable registration of the spatial distribution of lSlW formed in elements of the multiplying tungsten model of the booster. Both the target and all the rods of the model should consist of natural tungsten, each element of the model should consist of two parts, butt-jointed along the meridian symmetry plane, containing the axis of the electron-optical channel of the injector. The size and relative position of the target and rods should be the same as in the buffer itself. The model is set up at exactly the same place in the injector channel where the booster core and target are located and is irradiated for the time necessary for 1017-1016 electrons to impinge on the target (an exposure for 10 rain is sufficient in an LUte-40 linear electron accelerator). The model is irradiated for 1 week and then half of each element is extracted from it, and the y-ray density distribution in the plane of its cross-sectional surface and in the cylindrical (or other shape) lateral surface is measured from the lines E. r = 57.7, 66, 136.3, and 152.5 keV. It is sufficient to multiply the measured y-ray density by the known ratio of the primary-neutron yield per electron in the fissile material in the core and in the material of the booster target to the primaryneutron yield in natural tungsten (e.g., for the IBR-30 + LUll-40 booster with plutonium rods in the core the first multiplier is 3 and the multiplier for the tungsten target is 1) for the density distribution to be proportional to the real distribution in the buffer. We experimentally verified the efficiency of the method by using a rod of natural tungsten of length i40 mm and diameter 10 ram. This rod in a model of a tungsten target and several solid tungsten and lead imitators of rods of the core of the IBR-30 + LUI~-40 booster, with the same dimensions as the original, had been previously exposed (in 1988) to the LUt~40 beam to study the heat-release distribution and the current of primary electrons and electrons scattered by the elements of the booster. The density distribution of the primary-electron generation at various points on the surface of the solid tungsten rod (in the model it is put along with the target opposite its flat cut at an angle of 15 ° to the axis of the target; see Fig. 1) was