Radial void fraction measurement of a random multisized pebble stacking

This report describes the research regarding the radial void fraction of a pebble-bed reactor. The experimental model is a scaled version (1:10) of the active HTR-10 reactor in China. In a pebble-bed reactor the void fraction distribution is important in calculating the correct criticality. In the near wall region the radial void fraction shows an oscillatory behavior named flow channeling. This effect has consequences on the mechanics of heat transfer and on the flow and pressure drop of the coolant through the reactor.Earlier studies have been performed on this topic since their are various ways to determine the porosity profile. In this report the void fraction is acquired using a gamma-transmission technique. With this method the packing fraction can be obtained while keeping the pebble stacking intact. In the experiments the gamma radiation from the 60 [keV] peak of an Am-241 source is used to determine the decay of the beam through a pebble-bed. Two collimators with a width of 1.0 and 2.0 [mm] are used to create the correct resolution. The measurement time varies from a day to a week, depending on various experimental parameters like the amount of counts and the pebble stacking. Compensating for build up effect was not necessary according to calibration measurements. The radial void fraction is obtained for different pebble stackings. The profile for an unisized pebble-bed comply with other research. The void fraction distribution of the shaken bed, measured to determine the influence of an earthquake, showed a oscillatory behavior not seen before. The multisized pebble-bed with spheres of two sizes were difficult to produce. Three experiments with a different pebble multisized distribution showed a average void fraction significantly lower than the unisized packed bed. Mixing the smaller spheres in the bed slightly decreased the amplitude of the oscillations. Flow channeling will probably still be observed in a multisized bed. In the experiments uncertainties range from 1% to 10%, dependent on the used collimator and the void fraction. The average void fraction was lower than expected in the multisized experiments. This could be improved by measuring the background during the measurements of by improving the calibration method. Furthermore a method is developed to determine the walls boundary. It is defined as the point with an distance of 1 [mm] from the point of highest void fraction in the profile. This definition was proven useful according to the experiments.