Development of a Single-Disk Neutron Chopper for the Time-Of-Flight Spectroscopy
| Participants: | Prof. Jack S. Brenizer |
| Prof. Kenan Ünlü | |
| J. Niderhaus, Graduate Student | |
| Services Provided: | Neutron Beam Laboratory |
| Sponsors: | U.S. Department of Energy-Innovations in Nuclear Infrastructure and Education (INIE) |
Introduction
A single-disk, “slow” chopper system 1,2 has been developed at the Penn State Radiation Science and Engineering Center (RSEC) for the purpose of energy spectrum measurements on thermal neutron beams. This was achieved by gating the beam with a single rotating narrowly-slotted disk of neutron-absorbing material (a “chopper”) and measuring the resultant time-of-flight (TOF) distribution of transmitted neutrons to an adjacent detector, located at a known separation distance. The TOF distribution was transformed to distributions in neutron speed and energy, which closely matched models based on the Maxwell-Boltzmann distribution. The advantage of this technique is that no measurement of the neutron or secondary particle energy deposition is necessary.
Experimental Setup and data Acquisition
Instead of measuring energy deposition, the TOF system recorded neutron counts and arrival times at the detector, relative to a known starting time. This starting time was established at the instant when the 1-mm-wide radial slit in the chopper disk reached full illumination as it passed through the beam. This occurred once per rotation, emitting a 384-ms-long pulse of neutrons in a thermal spread, which then traversed a known distance to the detector. The chopper disk rotated with a period of 115 ms, and was covered on both faces with 1-mm thick Cd sheet overlaid with Gd paint.
Neutrons transmitted by the chopper slit traversed a 2.18-m collimated flight path and interacted in a shielded LiI(Eu) scintillator by ( n , a) reactions. A photomultiplier tube/multi-channel scaler (MCS) system counted these events in 1024 20-ms-wide data channels, sweeping over all of the channels sequentially once per chopper rotation. The start of each sweep was synchronized to the opening of the chopper slit using a photodiode triggering device on the chopper apparatus
The chopper slit was centered in the RSEC primary neutron beam, at the exit of a 19.1-cm-diameter, 3.4-m-long beam tube (Figure 1-5). Neutrons were generated in the RSEC TRIGA Mark III reactor core, and moderated by a large D 2 O-filled tank coupled directly to the core. The beam tube used for this study had a direct view of the D 2 O tank. The beam diameter was reduced to 3.8 cm, and then to 1.9 cm, by two 36-cm-long cylindrical Pb/concrete collimator plugs, placed at the downstream end of the beam tube.
With the reactor operating at a power of 1 MWt, the neutron signature from the chopper was nearly undetectable due to the gamma field, unless appropriate measures of collimation were used. Two sets of 5-inch deep Pb bricks, closely spaced to match the width of the beam transmitted by the chopper, were inserted in a straight tube between the chopper and detector for this purpose. The background due to gammas still present in the beam was further reduced by pulse-height discrimination.
Figure 1. Schematic layout of Breazeale reactor core, beam port #4 and neutron chopper time-of flight system.
Figure 2. Schematic of chopper, collimator tube, and detector
Figure 3. Schematic of single-disk neutron chopper
Figure 4. A picture of single disk neutron chopper
Figure 5. A picture of chopper and collimator tube at beam port #4.
Data Transformation
The experimentally-measured data consisted of 1024 channels, each containing a number of accumulated counts, recorded over 150,000 MCS sweeps. A series of mathematical operations, performed on a channel-by-channel basis, was used to transform this dataset to distributions of differential flux in the time (TOF), velocity, and energy domains. The domain variables assigned to each channel were calculated as
where t i is the flight time, i is the channel index, dt is the MCS dwell time per channel (20 ms), D is the flight distance, and m is the neutron mass.
After background subtraction, the recording time per channel, dwell time, and detector efficiency were used to convert the recorded counts in each channel to the differential flux per unit time-of-flight ( df / dt ) i . Two chain rule factors were then applied to convert this value to differential flux per unit speed ( df / dv ) i , and differential flux per unit energy ( df / dE ) i . In this way, the total recorded flux was conserved.
Modeling and Results
The neutron beam is fully thermalized in the D 2 O moderator material. Hence, the differential number density of neutrons in the moderator with velocity v in dv can be described by the Maxwell-Boltzmann formula. Since flux is related to number density by 
, the flux of neutrons with velocity v in dv is described by the Maxwell-Boltzmann formula, modified by a factor of v and appropriately renormalized:
,
where f is the total flux, k is Boltzmann's constant, and T is the moderator temperature (20°C). Similarly, the Maxwell-Boltzmann distribution in energy is modified by a factor of E 1/2 to describe the flux distribution:
.
These two distributions constitute a Maxwell-Boltzmann-based model to which experimental results from the chopper can be compared.
Figs. 6 and 7 show the velocity and energy spectra derived from experimental data, along with the corresponding Maxwell-Boltzmann-based models. The “measured” data points are ( df / dv ) i and ( df / dE ) i , corrected for background, detection system dead time, and losses in air along the flight path. One data point per MCS channel is plotted, located at the velocity ( v i ) and energy ( E i ) corresponding to the midpoint of the time channel. The “model” data points are the Maxwell-Boltzmann-based formulae evaluated at v i and E i , after correction for losses in air in the beam tube.
Figure 6. Thermal neutron velocity spectrum.
Conclusions
The spectra generated by the chopper are in good agreement with physical models. These spectra fit the models to within one standard deviation. Since a slow-chopper spectrometer of this type is relatively simple and inexpensive to construct, it is hoped that future work will further develop this instrument for use in precise spectroscopic measurements on thermal neutron beams.
References
• T. Emoto, Ph.D. thesis, Cornell University , Ithaca , NY (1990).
• S. Spern, Ph.D. thesis, Cornell University , Ithaca , NY (1998).
Theses
John Henry J. Niederhaus, “A single-disk-chopper time-of flight spectrometer for thermalneutron beams.” A Master of Science Thesis in Nuclear Engineering, The Pennsylvania State University , August 2003.
