Recoil of Fission Products. II. In Heterogeneous Carbon Structures

Abstract

A previously reported investigation of the recoil behavior of ``light'' and ``heavy'' fission fragments in a uniform pyrocarbon structure has been extended to include five additional structures with porosities ranging from 5 to 23 vol%. As before, fragments were induced to penetrate target structures as nearly monoenergetic beams with random r‐space directions. Data were collected and arranged to produce plots of fractional activities remaining at various grinding levels along the z coordinate normal to thin (neutron‐activated) ²³⁵U source planes. Such plots are integrated forms of the z‐space probability‐density functions (called distribution functions); these, in turn, are specified by associated r‐space functions. An important facet of these studies was the development of integrated z‐space equations that would exhibit good agreement with data for porous structures. Selection of realistic r‐space functions was therefore mandatory. Application of the proper integrated equations permitted evaluation of several sets of range values that were nearly identical and independent of the structures involved. A compact expression for all results in revised form is $${\log }_{10}{\mathrm{(R\prime )}}_i\text{=0.478}{\log }_{10}{\mathrm{(E}}_0{)}_i\text{−0.465,}$$ where R′ (mg/cm²) is the range of any ith fragment with an initial fission energy E₀ (MeV). Although the range values for homogeneous and heterogeneous materials are the same, applicable distribution functions varied in form because of variations in pore‐size characteristics from one structure to the next. In some cases, the pore sizes were equal to or greater than the range. This resulted in large straggling effects and asymmetric r‐space distributions; furthermore, the R′ in the above expression corresponds to the most probable distance of r‐space travel. Range results for only two of the six structures investigated are correlated on the basis of a normal (or Gaussian) r‐space distribution, where the average and most probable range are synonymous. Results for all other structures are correlated by using general forms of the Maxwell distribution. In these cases, the average distance traveled is always greater than R′ (to an extent depending on the skewness of the distribution function selected for data correlation).