Karen A. Miller’s research while affiliated with Los Alamos National Laboratory and other places

The blossoming of sensing solutions based on the use of carbon materials and the pervasive exploration of compressed sensing (CS) for developing structural health monitoring applications suggest the possibility of combining these two research areas in a novel family of smart structures. Specifically, the authors propose an architecture for security-related applications that leverages the tunable electrical properties of a graphite oxide (GO) paper-based tamper-evident seal with a compressed-sensing (CS) encryption/authentication protocol. The electrical properties of GO are sensitive to the traditional methods that are commonly used to remove and replace paper-based tamper-evident seals (mechanical lifting, solvents, heat/cold temperature changes, steam). The sensitivity of the electro-chemical properties of GO to such malicious insults is exploited in this architecture. This is accomplished by using GO paper to physically realize the measurement matrix required to implement a compressive sampling procedure. The proposed architecture allows the seal to characterize its integrity, while simultaneously providing an encrypted/authentication feature making the seal difficult to counterfeit, spoof, or remove/replace. Traditional digital encryption/authentication techniques are often bit sensitive making them difficult to implement as part of a measurement process. CS is not bit sensitive and can tolerate deviation caused by noise and allows the seal to be robust with respect to environmental changes that can affect the electrical properties of the GO paper during normal operation. Further, the reduced amount of samples that need to be stored and transmitted makes the proposed solution highly attractive for power constrained applications where the seal is interrogated by a remote reader.

Structural Health Monitoring (SHM) technology has great unexplored potential in security applications. Specifically breakthroughs in graphene-oxide (GO) damage-detecting skins coupled with nonlinear, sparse signal processing techniques being used for SHM lend themselves to addressing the need for low-power remotely-readable tamper-evident seals. Assessing the integrity of a tamper-evident seal is inherently an SHM problem. In this case damage is caused by a human adversary, not the environment. This paper presents a novel architecture that leverages the tunable electrical properties of a GO-paper-based seal with a compressed-sensing (CS) acquisition protocol. This architecture allows the seal to characterize its integrity, while simultaneously providing an encrypted authentication feature making the seal difficult to counterfeit and/or spoof. The electrical properties of GO are sensitive to the traditional methods used to attack paper-based seals (mechanical lifting, solvents, heat/cold, steam). This property of GO allows us to determine if a seal has been tampered with simply by measuring its electrical properties. Specific areas of focus addressed by this work include the quantitative analysis of the encryption/authentication capabilities provided by CS, and methods for enhancing the detection of cracks/cuts propagating through the sensitive GO paper.

The ability to detect the diversion of a significant quantity of nuclear material is the objective of international nuclear safeguards, and safeguarding uranium enrichment technology is especially important in preventing the spread of nuclear weapons. In gas centrifuge enrichment plants, the uranium is in the form of uranium hexafluoride (UF6), which regularly transitions between phases during processing and storage due to its low triple point. The variation of material distribution in Uf6 cylinders has a significant effect on enrichment measurement uncertainty, especially with increasing accuracy of passive neutron assay measurements, motivating interest in non-destructive methods of examining UF6 cylinder fill distributions. For this study, a novel steady-state scanning laser Doppler vibrometer (LDV) system is used to determine the material and phase distributions within small UF6 cylinders. This technology induces steady-state ultrasonic waves throughout an object and then uses a scanning laser to measure vibrational wavenumber at pixilated points on the object. Different materials and phases are excited by the ultrasonic waves to different extents, changing the local measured wavenumbers and allowing fill distributions of the storage cylinders to be mapped. Through non-destructive examination of the variation in typical fill profiles of UF6 cylinders, the uncertainty associated with neutron measurements of UF6 cylinders may be better characterized.

As new uranium enrichment plants are proposed and come online worldwide, interest in using neutron methods for uranium hexafluoride (UF6) cylinder assay has been growing; however, large discrepancies exist in published F(α,n) yields from uranium isotopes. Uncertainties in these data are propagated through the analysis of every UF6 measurement and have implications for safeguards conclusions drawn from them. In this paper, a value for the specific F(α,n) yield in UF6 from ²³⁴U is calculated from measurements of 30B cylinders containing bulk UF6 at the Rokkasho Enrichment Plant in Japan. The measurements were taken using the Uranium Cylinder Assay System. The yield was derived by combining the cylinder measurements with detailed Monte Carlo modeling, known isotopic composition, and inversion analysis. We calculated the ²³⁴U neutron emission rate in UF6 to be (474 ± 21) n/s·g⁻¹ with a 68% confidence level. The results obtained in this study will help enable an important class of nondestructive assay instruments to be applied with greater confidence and accuracy.

Uranium cylinder assay plays an important role in the nuclear material accounting at gas centrifuge enrichment plants. The Passive Neutron Enrichment Meter (PNEM) was designed to determine uranium mass and enrichment in 30B and 48Y cylinders using total neutron and coincidence counting in the passive mode. 30B and 48Y cylinders are used to hold bulk UF6 feed, product, and tails at enrichment plants. In this paper, we report the results of a Monte-Carlo-based feasibility study for an active uranium cylinder assay system based on the PNEM design. There are many advantages of the active technique such as a shortened count time and a more direct measure of 235U content. The active system is based on a modified PNEM design and uses a 252Cf source as the correlated, active interrogation source. We show through comparison with a random AmLi source of equal strength how the use of a correlated driver significantly boosts the active signal and reduces the statistical uncertainty. We also discuss ways in which an active uranium cylinder assay system can be optimized to minimize background from 238U fast-neutron induced fission and direct counts from the interrogation source.

Fluorine has a relatively large (α,n) production cross-section in the MeV range, the energy range of interest for special nuclear materials. In the uranium fuel cycle enriched UF6 in particular is a reasonably prolific source of (α,n) neutrons because along with 235U, 234U becomes enriched and it has a relatively short half-life. This enables the mass content of storage cylinders containing UF6 to be verified by neutron counting methods.In association with such measurements high resolution gamma-ray spectrometry (HRGS) measurements using a high-purity Ge detector are often undertaken to determine the 235U enrichment based off the intensity of the direct 186 keV line. The specific (α,n) neutron production, neutrons per second per gram of U, is sensitive to the relative isotopic composition, particularly the 234U concentration, and the traditional gross neutron counting approach is needed to quantitatively interpret the data.In addition to F(α,n) neutrons, α-induced reaction γ-rays are generated, notably at 110, 197, 582, 891, 1236 and 1275 keV. If one could observe 19F(α,xγ) gamma-lines in the HRGS spectra the thought was that perhaps the α-activity could be estimated directly, and in turn the 234U abundance obtained. For example, by utilizing the ratio of the detected 197–186 keV full energy peaks. However, until now there has been no readily available estimate of the expected strength of the reaction gamma-rays nor any serious consideration as to whether they might be diagnostic or not.In this work we compute the thick target yields of the chief reaction gamma-rays in UF6 using published thin target data. Comparisons are made to the neutron production rates to obtain γ/n estimates, and also to the 235U decay line at 186 keV which we take as a fiducial line. It is shown that the reaction gamma-rays are produced but are far too weak for practical safeguards purposes.Now that the underlying numerical data is readily available however, it can be used to support neutron and gamma production calculations in other fluorine compounds, for example impure plutonium reference materials where fluorine may be present only at the parts per million by weight level yet still present a serious nuisance addition to the neutron production rate.

The Passive Neutron Enrichment Meter (PNEM) is a nondestructive assay (NDA) system being developed at Los Alamos National Laboratory (LANL). It was designed to determine ²³⁵U mass and enrichment of uranium hexafluoride (UFâ) in product, feed, and tails cylinders (i.e., 30B and 48Y cylinders). These cylinders are found in the nuclear fuel cycle at uranium conversion, enrichment, and fuel fabrication facilities. The PNEM is a ³He-based neutron detection system that consists of two briefcase-sized detector pods. A photograph of the system during characterization at LANL is shown in Fig. 1. Several signatures are currently being studied to determine the most effective measurement and data reduction technique for unfolding ²³⁵U mass and enrichment. The system collects total neutron and coincidence data for both bare and cadmium-covered detector pods. The measurement concept grew out of the success of the Uranium Cylinder Assay System (UCAS), which is an operator system at Rokkasho Enrichment Plant (REP) that uses total neutron counting to determine ²³⁵U mass in UFâ cylinders. The PNEM system was designed with higher efficiency than the UCAS in order to add coincidence counting functionality for the enrichment determination. A photograph of the UCAS with a 48Y cylinder at REP is shown in Fig. 2, and the calibration measurement data for 30B product and 48Y feed and tails cylinders is shown in Fig. 3. The data was collected in a low-background environment, meaning there is very little scatter in the data. The PNEM measurement concept was first presented at the 2010 Institute of Nuclear Materials Management (INMM) Annual Meeting. The physics design and uncertainty analysis were presented at the 2010 International Atomic Energy Agency (IAEA) Safeguards Symposium, and the mechanical and electrical designs and characterization measurements were published in the ESARDA Bulletin in 2011.

Safeguarding sensitive fuel cycle technology such as uranium enrichment is a critical component in preventing the spread of nuclear weapons. A useful tool for the nuclear materials accountancy of such a plant would be an instrument that measured the uranium content of UF cylinders. The Uranium Cylinder Assay System (UCAS) was designed for Japan Nuclear Fuel Limited (JNFL) for use in the Rokkasho Enrichment Plant in Japan for this purpose. It uses total neutron counting to determine uranium mass in UF cylinders given a known enrichment. This paper describes the design of UCAS, which includes features to allow for unattended operation. It can be used on 30B and 48Y cylinders to measure depleted, natural, and enriched uranium. It can also be used to assess the amount of uranium in decommissioned equipment and waste containers. Experimental measurements have been carried out in the laboratory and these are in good agreement with the Monte Carlo modeling results.

Current safeguards approaches used by the International Atomic Energy Agency (IAEA) at gas centrifuge enrichment plants (GCEPs) need enhancement in order to verify declared low-enriched uranium (LEU) production, detect undeclared LEU production and detect highly enriched uranium (HEU) production with adequate detection probability using non destructive assay (NDA) techniques. At present inspectors use attended systems, systems needing the presence of an inspector for operation, during inspections to verify the mass and 235U enrichment of declared UF6 containers used in the process of enrichment at GCEPs. This paper contains an analysis of possible improvements in unattended and attended NDA systems including process monitoring and possible on-site destructive assay (DA) of samples that could reduce the uncertainty of the inspector's measurements. These improvements could reduce the difference between the operator's and inspector's measurements providing more effective and efficient IAEA GCEPs safeguards. We also explore how a few advanced safeguards systems could be assembled for unattended operation. The analysis will focus on how unannounced inspections (UIs), and the concept of information-driven inspections (IDS) can affect probability of detection of the diversion of nuclear materials when coupled to new GCEPs safeguards regimes augmented with unattended systems.

Current safeguards approaches used by the International Atomic Energy Agency (IAEA) at gas centrifuge enrichment plants (GCEPs) need enhancement in order to verify declared low-enriched (LEU) production, detect undeclared LEU production and detect highly enriched uranium (HEU) production with adequate detection probability using nondestructive assay (NDA) techniques. At present inspectors use attended systems, systems needing the presence of an inspector for operation, during inspections to verify the mass and ²³U enrichment of declared UF containers used in the process of enrichment at GCEPs. In verifying declared LEU production, the inspectors also take samples for off-site destructive assay (DA) which provide accurate data, with 0.1% to 0.5% measurement uncertainty, on the enrichment of the UF feed, tails, and product. However, taking samples of UF for off-site analysis is a much more labor and resource intensive exercise for the operator and inspector. Furthermore, the operator must ship the samples off-site to the IAEA laboratory which delays the timeliness of results and interruptions to the continuity of knowledge (CofK) of the samples during their storage and transit. This paper contains an analysis of possible improvements in unattended and attended NDA systems such as process monitoring and possible on-site analysis of DA samples that could reduce the uncertainty of the inspector's measurements and provide more effective and efficient IAEA GCEPs safeguards. We also introduce examples advanced safeguards systems that could be assembled for unattended operation.