Search

We have evaluated the performance of an illicit substance detection system that performs image reconstruction using the maximum likelihood algebraic reconstruction algorithm, a few number of projections, and relatively coarse projection and pixel resolution. This evaluation was done using receiver operator curves and simulated data from the fast-neutron transmission spectroscopy system operated in a mode to detect explosives in luggage. The results show that increasing the number of projection angles is more important than increasing the projection resolution, the reconstructed pixel resolution, or the number of iterations in the maximum likelihood algorithm. A 100% detection efficiency with essentially no false positives is possible for a square block of RDX explosive, a projection resolution of 2 cm, a reconstructed pixel size of 2/spl times/2 cm, and five projection angles. For rectangular shaped explosives more angles are required to obtain the same system performance.

A model of a plasma ion source has been developed that approximates the system of Poisson and Boltzman-Vlasov equations normally used to describe ion sources by an external electric field, a collective electric field due to the charge column, and the starting boundary conditions. The equations of this model can be used directly in the Lorentz force equation to calculate trajectories without iteration. >

A high-energy neutron detector has been developed using a semiconductor diode fabricated from bulk gallium arsenide wafers with a polyethylene neutron converter layer. Typical thickness of the diode layer is 250 to 300 μm with bias voltages of 30 to 150 volts. Converter thicknesses up to 2030 μm have been tested. GaAs neutron detectors offer many advantages over existing detectors including positional information, directional dependence, gamma discrimination, radiation hardness, and spectral tailoring. Polyethylene-coated detectors have been shown to detect 14 MeV neutrons directly from a D-T neutron generator without interference from gamma rays or scattered neutrons. An array of small diode detectors can be assembled to perform fast neutron radiography with direct digital readout and real-time display of the image produced. In addition, because the detectors are insensitive to gamma rays and low energy neutrons, highly radioactive samples (such as spent nuclear fuel or transuranic waste drums) could be radiographed.

A simple model of the optics of an ion source has been developed and compared with the exact Poisson solution calculated by the computer code SNOW. The model replaces the exact Poisson potential by a Laplacian potential that is calculated by substituting a planar plate for the plasma and plasma sheath. The Laplacian potential is then used with and without a linear space‐charge correction to calculate particle trajectories. In this model the only free parameter is the distance between the planar plate and the plasma electrode. In general, there is good agreement between emittance curves calculated by the model and the SNOW results even when no space‐charge correction is used. For the geometry we studied, this implies that the effects of the plasma sheath and non‐linear space‐charge forces are small compared to aberrations introduced by the plasma aperture.

As part of an ongoing program to improve the fuel-motion capability of the fast-neutron hodoscope, a response model has been developed for the Hornyakbutton detector. The model provides a description of the efficiency and gamma sensitivity of a Hornyak button in terms of detector parameters such as scintillator concentration, grain size, length, and transparency. It is possible using this model to quantitatively choose detector parameters for the Hornyak button.

The use of computer-related devices has resulted in significant improvements in word processing, permitting better accuracy and more efficient turnover of paperwork in an office. A selective combination of these components comprises a computer-assisted writing and editing system for professionals at Argonne National Laboratory, used primarily for technical reports. Text is entered and corrected at a terminal keyboard, viewed on a CRT, and stored on a magnetic disc; camera-ready copy is obtained on an electrostatic printer. Engineers and scientists enter original text and do creative editing; typewriter-qualified personnel quickly learn to handle other editing and corrections; and computer-support personnel provide periodic file-management services.

Joint LASL/ANL studies of the fuel motion performance capability of a fast-neutron hodoscope have been performed on a 91-pin fuel bundle. Optimization of the neutron detectors used in the hodoscope was also investigated. The neutron detectors tested were a Hornyak button and a stilbene detector with pulse shape discrimination. These studies show that a hodoscope can provide useful fuel motion detection capability for large bundles and that the amount of information provided is greatly increased by the use of neutron detectors with high neutron thresholds and high detection efficiency. These measurements indicate that the TREAT hodoscope can be upgraded to meet most of the fuel motion diagnostic requirements of new safety facilities.

The fuel-motion measurement capability of the fast-neutron hodoscope has been upgraded by the addition of a 360-detector proton-recoil proportional-counter array, which detects high-energy fission neutrons. The current sensitive amplifier/discriminator module for each detector fits into a 12.7- by 12.7- by 102-mm package and costs less than $100 per module. It has a 50-ns rise time, a noise level of 100 nA, and a deadtime per event of 200 ns. Provision has been provided for the independent adjustment of the input current versus discriminator voltage for each module. The new proportional-counters cost approximately $400 each. Each detector has been tested to have the same gain versus voltage response. A space-charge model relating count-rate changes to space-charge effects has also been developed. The new detector array has been operational for approximately two years and has become the main detector system in fuel-motion analysis. It has significantly improved the linearity, stability, count-rate capability, and setup ease of the hodoscope.

The shapes of the current pulses derived from fully depleted surface barrier detectors are observed and compared with theory. Separate measurements are made with low-energy alpha particles entering the front and back surfaces, so the pulses are due to the motion of electrons only, and holes only, respectively. These pulses should be of the same shape, but reversed in time and scaled in time in proportion to the mobilities. This is observed when there is a large over-voltage, but for lower voltages, straggling distorts the shape of the pulses from hole collection. The shapes of the electric fields in the detectors are surmised from the pulse shapes. At high voltage, the fields become much closer to uniform than is implied by the usual constant gradient assumption, and in two of the four detectors they become quite uniform.

($p, \mathrm{pn}$) reactions induced by 17-MeV protons on 12 different nuclei are studied by coincidence detection and energy measurements of the emitted neutron and proton. Probabilities of exciting various final states and of ($p, {p}^{\ensuremath{'}}\ensuremath{\gamma}$) reactions (i.e., no neutron emission) are measured as a function of the energy of the emitted particles, and attempts are made to explain the results with a model assuming that these are ($p, {p}^{\ensuremath{'}}$) reactions followed by neutron boil-off. There are many disagreements with the predictions of this model, and many inconsistencies between results from the different nuclei under this interpretation. It is shown that including contributions from ($p, n$) reactions followed by proton boil-off does little to improve the situation. The results might be explained if both particles are emitted in a direct reaction, although there is difficulty with the simplest such reaction, a single collision knockout, in that there is poor correlation between states excited in these reactions and in pickup reactions such as ($p, d$).

Coincidence experiments indicate that ($p,np$) reactions on ${\mathrm{Zr}}^{90}$ and ${\mathrm{Sn}}^{112}$ leave the residual nuclei in their ground states, whence neutron spectra can be deduced from measurements of proton spectra. Results are very different for the two nuclei. Analysis indicates that the variation of ${\ensuremath{\sigma}}_{i}$-the cross section for the inverse process-is larger than expected and in opposite directions for the two nuclei.

A method is described and demonstrated for a nondestructive determination of the hydrogen content inside materials in sheet form by measurement of the energy‐loss distribution from proton‐proton scattering events. The method is absolutely specific for hydrogen, and has a sensitivity of 1 ppm (atomic), or better. Materials analyzed (with ppm of hydrogen in parentheses) were: Ti (4200), Zr (5000), Al (20), Fe (∼ 10), C (13000), and Teflon (220). In the titanium and zirconium samples, the hydrogen content drops off near the surfaces. The method also detects oil or moisture on surfaces for thicknesses down to about 1 A; all materials studied had such surface films several hundred A thick.

A clean separation is made between the compound-nucleus and direct-reaction contributions to the low-energy continuum region of the energy spectra of protons from inelastic scattering. Bombarding energies are 12 and 17 MeV, and target nuclei are the isotopes of Ni, Zr, Cd, Mo, and Sn. The compound-nucleus contribution is studied by determining level-density parameters and the proton-to-neutron-emission ratios. These are in good agreement with predictions from statistical theory.

A method is suggested for determining the distribution of E l transition strengths from the ground state to excited states as a function of excitation energy and l ; it uses direct reaction inelastic scattering and a simplifying assumption which is tested and found to be reasonably accurate. The method depends on the reliability of DWBA calculations, and experimental tests of this reliability are devised and carried out; most of these tests are failed, so results on E l strength functions must await improvements in DWBA calculations.

The Weisskopf-Ewing assumption-that an ($x,yn$) reaction occurs with nearly unit probability if the first emitted particle, $y$, comes off with sufficiently low energy to make neutron emission energetically possible-is tested by coincidence studies of ($p,{p}^{\ensuremath{'}}n$) and ($p,{p}^{\ensuremath{'}}\ensuremath{\gamma}$) reactions on ${\mathrm{Sn}}^{119}$ and ${\mathrm{Ni}}^{61}$ and found to be grossly in error. In particular, ($p,{p}^{\ensuremath{'}}\ensuremath{\gamma}$) competes favorably with ($p,{p}^{\ensuremath{'}}n$) even when neutron emission is energetically possible by \ensuremath{\sim}2 MeV. In ${\mathrm{Sn}}^{119}$, the detailed shape of the energy spectrum of neutrons implies that spectra are drastically non-Maxwellian at low energies.