The FORTRAN version of the simulation code to model the current nd breakup apparatus uses as a foundation the code developed for the simulation and analysis of the previous nd breakup experiment which measured the neutron scattering length, ann, at 19 MeV at TUNL using the recoil-proton geometry [i]. Although this measurement focused on the scattering length, and the current measurement focuses on the cross sections, the apparatus and detector configurations are similar. The FORTRAN program has been modified to correctly simulate the current detector configuration, including neutron detectors that are out of the plane defined by the beam, and proton detectors.
The interaction probabilities, or cross sections, for nd scattering at a specific energy and angle are generated by the theoretical calculations and tabulated for use by the computer program. These theoretical cross section libraries at 19 MeV for this experiment using a CD-Bonn potential have been created. Several or more grid points were chosen for each angle acceptance to cover the entire detector region, and three beam energies (18.6, 19.0, 19.4 MeV) were chosen to characterize the beam energy spread. The spread in beam particle energy results from transmission through the deuterium gas cell where the beam neutrons are created, through the vacuum chamber windows and through the deuterium target. The library at 19 MeV was created for each combination of angles and beam energy and for each detector configuration (CS or SS). Similarly, the libraries at 16 MeV are being developed. Note that an NCCU student contributed to the effort of developing the 19 MeV cross section library before the start of the CREST program.
In addition to modifying the FORTRAN version, we started developing the GEANT4 based computer model of the experiment. We intend to take advantage of the well-developed GEANT4 geometry modeling capabilities to simulate this experiment using a more realistic geometry than is feasible in the FORTRAN version. The basic geometry model of the experiment has been developed and coded which includes the target, a CD2 foil with a small (1.7%) amount of CH2, two proton detectors, two “transmission foil detectors” for the proton arm, and four neutron detectors. Figure 1 shows the basic configuration of this experiment using the GEANT4 geometry simulation package.
Figure 10. Basic geometry of the nd breakup experiment and particle trajectories. Two larger purple (four smaller dark green) cylinders are proton (neutron) detectors. Green lines (blue lines) represent the trajectories of neutral-charge (positive-charge) particles. [Note: Green (blue) lines are shown in lighter (darker) lines in black and white printing.]
For convenience, we are currently using one of the built-in hadronic physics lists in GEANT4 called QGSP_BIC_HP which includes the high-precision neutron interaction data library at low neutron energy. The same cross sections used in the FORTRAN version will be implemented in a user defined cross section library.
The current version of the GEANT4 package code is able to generate nd breakup events. However, due to very small cross sections, or interaction probabilities, such an event rarely happens. Therefore, the current version of this simulation code is too inefficient to simulate this experiment. We are now working to solve this difficulty by changing the way events are generated in order to create a faster program. We will also construct a more precise geometry by adding the target chamber and enclosures for the proton and neutron detectors as well as the supporting structure and lead shielding wall.
We are also considering simulating this experiment using a more realistic neutron beam generated from the 2H(d,n)3He reaction in the deuteron gas cell. To do so, the gas cell and collimator will be added to the current geometry. With this detailed geometry, we will be able to study the energy loss of recoil protons in the target and in the transmission foil detectors.
Both the FORTRAN version and the GEANT4 version will be used to compare data with simulated results.