Logo of the Physikalisch-Technische Bundesanstalt

Transforming the high-energy neutron beam at iThemba LABS into a neutron reference radiation field has begun – optimizing the shielding of the neutron production target


The upgrade of the high‑energy neutron beam at iThemba LABS is planned to take place within the scope of a cooperation project between the European metrology institutes NPL, IRSN and PTB and their South African partners iThemba LABS, the University of Cape Town and the South African metrology institute NMISA. For this purpose, the first beamtime was logged at the neutron beam of iThemba LABS in January 2019. During this beamtime, the distribution of the neutron background present for the shielding configuration in place at that time was measured for neutron energies of 66 MeV and 200 MeV. These measurements were performed by means of the HERMEIS Bonner sphere spectrometer of IRSN. Since then, the NPL and IRSN teams have evaluated those measurements. At the same time, a Monte Carlo model of the neutron beam facility was set up to verify whether the measurement results obtained complied with the neutron fluence rates computed with the MCNP6 particle tracking code. An improved shielding configuration was then designed based on the MCNP6 simulations. Figure 1 shows the neutron beam facility from above with a color‑coding scheme for the shielding materials used.

privious shielding of the target area

Figure 1:   Results of the investigation of the neutron production target’s previous shielding element against the experimental area of the neutron beam facility at iThemba LABS. The proton beam energy was 66 MeV and the thickness of the lithium target 8 mm. The red and black histograms show the spectral neutron fluence rates obtained from the data of IRSN’s Bonner sphere spectrometer with different prior data. The black histograms represent the result of the Monte Carlo simulations with MCNP6 for measurement positions #1 - #4.

The challenge in computing such a shielding configuration is the strong attenuation of the neutron fluence rate behind the shield. For a non‑elastic attenuation coefficient of approx. 0.08 cm-1, the iron shield with a thickness of 200 cm (light blue area in Figure 1), which is located between the target and the experimental area, reduces the primary fluence rate of high‑energy neutrons by a factor of 10-7. Due to inelastic scattering and (n,xn) reactions, a large share of the primary neutrons is, however, “converted” into low‑energy neutrons; thus, the total neutron fluence rate transmitted is much higher behind the shield.

Modeling the particle tracks behind such thick shields requires specific Monte Carlo methods for variance reduction as well as sufficient computing power. The results shown in Figure 1 for a 66 MeV proton beam and a lithium neutron production target with a thickness of 8 mm could only be achieved by using PTB’s high‑performance computing cluster in Berlin‑Charlottenburg. At the positions shown, the (C/E) ratios of the computed and measured total fluence rates of background neutrons vary between 0.6 and 1.4. In view of the complex geometry and of the solid shields, this represents a good agreement. An important reason for deviations might also be the neutron production in the aluminum block that acts as a beam dump for the 66 MeV proton beam. This neutron production is not sufficiently well known. To solve this problem, improved cross‑section data for neutron production would be very helpful. The C/E ratios obtained for the 200 MeV neutron beam were between 0.3 and 0.8. Similar to the 66 MeV beam, the largest deviation occurred at position #2, which is closest to the aluminum beam dump.

The local team in Cape Town has used a shutdown of the cyclotron to reinforce the shielding of the neutron production target at positions that were clearly critical. It was originally planned to investigate the impact of these modifications in the spring of 2020. Unfortunately, due to the Covid‑19 pandemic, these investigations had to be postponed for an indefinite period of time. Simulations of the improved shielding have shown that the fluence rate of the background neutrons should be reduced by a factor of 3. Figure 2 shows the improved shielding near the beam dump and near the access to the target area.

 improved target shielding

Figure 2:   Current setup with improved shielding of the target area.


Opens local program for sending emailR. Nolte, Department 6.4, Working Group 6.42