Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “multiple control objectives”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 307 records · Page 17

Robot-based Additive Manufacturing of Lego-type Modular Molds for Wind Blades

The objective of this project is to reduce the cost and lead time of horizontal wind turbine blade mold tooling and blade transportation, while maintaining the highest standards of blade quality. The solution involves a smart-design family of modular molds that are easily transportable to fabrication sites near the place of service. Key innovations include the use of additive manufacturing (AM) to integrate conformal thermal management channels, offering enhanced control over the thermal profiles tailored to specific blade materials. This approach enables in-situ quality assurance during mold fabrication, significantly improves mold life, and allows for reuse across multiple production cycles. Ultimately, the solution aims to optimize both tooling and transportation costs, contributing to the scalability of wind turbine blade production. A significant barrier to scaling up the production of large wind turbine blades lies in the high costs associated with tooling and the transportation of blades. Traditional molds are expensive, bulky, and difficult to transport, adding considerable lead time and cost to the overall manufacturing process. Additionally, transporting blades to distant locations for final assembly further exacerbates these challenges. The project aims to address these inefficiencies by demonstrating a modularized, additive-manufactured mold that meets all necessary blade specification requirements, specifically for blade lengths between 120m and 150m.

17 WIND ENERGY↗

Low-temperature vapour-liquid-solid (VLS) growth of vertically aligned silicon oxide nanowires using concurrent ion bombardment.

Vertically aligned silicon oxide nanowires can be synthesized over a large area by a low-temperature, ion-enhanced, reactive vapour-liquid-solid (VLS) method. Synthesis of these randomly ordered arrays begins with a thin indium film deposited on a Si or SiO{sub 2} surface. At the processing temperature of 190 C, the indium film becomes a self-organized seed layer of molten droplets, receiving atomic silicon from a DC magnetron sputtering source rather than from the gaseous precursors used in conventional VLS growth. Simultaneous vigorous ion bombardment aligns the objects vertically and expedites mixing of oxygen and silicon into the indium. Silicon oxide precipitates from each droplet in the form of multiple thin strands having diameters as small as 5 nm. These strands form a single loose bundle growing normal to the surface, eventually consolidating to form one nanowire. The vertical rate of growth can reach 300 nm min{sup -1} in an environment containing argon, hydrogen, and traces of water vapour. This paper discusses the physical and chemical factors leading to the formation of the nanostructures. It also demonstrates how the shape of the resulting nanostructures can be further controlled by sputtering, during both VLS growth and post-VLS processing. Key technological advantages of the developed process are nanowire growth at low substrate temperatures and the ability to form aligned nanostructure arrays, without the use of lithography or templates, on any substrate onto which a thin silicon film can be deposited.

08 HYDROGEN↗

‘Flux+Mutability’: a conditional generative approach to one-class classification and anomaly detection

Abstract Anomaly Detection is becoming increasingly popular within the experimental physics community. At experiments such as the Large Hadron Collider, anomaly detection is growing in interest for finding new physics beyond the Standard Model. This paper details the implementation of a novel Machine Learning architecture, called Flux+Mutability, which combines cutting-edge conditional generative models with clustering algorithms. In the ‘flux’ stage we learn the distribution of a reference class. The ‘mutability’ stage at inference addresses if data significantly deviates from the reference class. We demonstrate the validity of our approach and its connection to multiple problems spanning from one-class classification to anomaly detection. In particular, we apply our method to the isolation of neutral showers in an electromagnetic calorimeter and show its performance in detecting anomalous dijets events from standard QCD background. This approach limits assumptions on the reference sample and remains agnostic to the complementary class of objects of a given problem. We describe the possibility of dynamically generating a reference population and defining selection criteria via quantile cuts. Remarkably this flexible architecture can be deployed for a wide range of problems, and applications like multi-class classification or data quality control are left for further exploration.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Genome and Transcriptome of Clostridium phytofermentans, Catalyst for the Direct Conversion of Plant Feedstocks to Fuels

Clostridium phytofermentans was isolated from forest soil and is distinguished by its capacity to directly ferment plant cell wall polysaccharides into ethanol as the primary product, suggesting that it possesses unusual catabolic pathways. The objective of our present study was to understand the molecular mechanisms of biomass conversion to ethanol in a single organism, Clostridium phytofermentans, by analyzing its complete genome and transcriptome during growth on plant carbohydrates. The saccharolytic versatility of C. phytofermentans is reflected in a diversity of genes encoding ATP-binding cassette sugar transporters and glycoside hydrolases, many of which may have been acquired through horizontal gene transfer. These genes are frequently organized as operons that may be controlled individually by the many transcriptional regulators identified in the genome. Preferential ethanol production may be due to high levels of expression of multiple ethanol dehydrogenases and additional pathways maximizing ethanol yield. The genome also encodes three different proteinaceous bacterial microcompartments with the capacity to compartmentalize pathways that divert fermentation intermediates to various products. Lastly, these characteristics make C. phytofermentans an attractive resource for improving the efficiency and speed of biomass conversion to biofuels.

09 BIOMASS FUELS↗

RELAP-7 Beta Release: Summary of Capabilities

RELAP-7 is a nuclear systems safety analysis code being developed at the Idaho National Laboratory (INL). Building upon the decades of software development at the INL, we began the development of RELAP-7 in 2011 to support the Risk Informed Safety Margins Characterization (RISMC) Pathway. As part of this development, the first lines of RELAP-7 code were committed to the software revision control repository on November 7th, 2011. The overall design goal of RELAP-7 is to take advantage of the previous thirty years of advancements in computer architecture, software design, numerical methods, and physical models in order to provide capabilities needed for the RISMC methodology and to support nuclear power safety analysis. RELAP-7 is built using the INL’s modern scientific software development framework, MOOSE (Multi-physics Object Oriented Simulation Environment). MOOSE provides improved numerical calculations (including higher-order integration in both space and time, yielding converged second-order accuracy). The RELAP-7 code structure is based on multiple physical component models such as pipes, junctions, pumps, etc. Each component can have options for different fluid models such as single- and two-phase flow. This component-based and physics-based software architecture allows RELAP-7 to adopt different physical models for different applications. A relatively new two-phase hydrodynamic model, termed the ''7-Equation model'' for two phasic pressures, velocities, energies, and volumetric fraction, is incorporated into RELAP-7 for liquid-gas (water-steam) flows. This new model allows second-order integration because it is well-posed, which will reduce the numerical error associated with traditional systems analysis codes. In this paper, we provide a RELAP-7 capability list describing analysis features, range of applicability, and reactor components that will be available for the December 15th, 2014 beta release of the software.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Robust Assessment of Real-World Adversarial Examples

We explore rigorous, systematic, and controlled experimental evaluation of adversarial examples in the real world and propose a testing regimen for evaluation of real world adversarial objects. We show that for small scene/ environmental perturbations, large adversarial performance differences exist. Current state of adversarial reporting exists largely as a frequency count over a dynamic collections of scenes. Our work underscores the need for either a more complete report or a score that incorporates scene changes and baseline performance for models and environments tested by adversarial developers. We put forth a score that attempts to address the above issues in a straight-forward exemplar application for multiple generated adversary examples. We contribute the following: 1. a testbed for adversarial assessment, 2. a score for adversarial examples, and 3. a collection of additional evaluations on testbed data.

deep learning, adversarial machine learning, machi↗

High Energy Storage Capacity Low Cost Iron Flow Battery

A new flow battery was proposed that utilizes low cost materials: iron as the only active element, cheap aqueous electrolytes, and inexpensive separators. During charging, ferrous iron (Fe 2+ ) is oxidized to ferric iron (Fe 3+ ) at the positive electrode while it (Fe 2+ ) is reduced to form iron metal (Fe 0 ) at the negative electrode. Because iron is plated at the negative electrode during charging, conventional electrode structures couple the energy storage capacity and the power rating of the battery. In order to decouple the energy and power ratings and regain the economic advantages of a flow battery, a slurry electrode design was proposed for use as the negative electrode. The slurry electrode is made by flowing electrically conductive particles in an electrolyte containing the dissolved iron species. On charging, iron is plated onto the particles. The particles can then carry the iron metal out of the cell to be stored in external reservoirs. This project looked to develop a slurry for the negative electrode of the all-iron flow battery that can operate at 200 mA/cm 2 with <100 mV overpotential, be pumped with parasitic energy costs < 10%, and promote plating on the particles. A scalable and stackable cell design of 1100 cm 2 was developed. The project was initially under contract in mid-January 2013. The project was converted from a one-year proof of concept seedling to an additional two-year program, then extended with an 18-month plus-up, and finally with an additional 6-month approved cost-overrun. During the seeding project, the concept of plating on carbon particles in a slurry was demonstrated. However, this was accomplished with high-cost multi-walled carbon nanotubes as the slurry material. The following two-year program discovered several low-cost carbon blacks that could also be used in this application. These findings were supported by an analysis that demonstrated the effectiveness of the carbon that takes into account not just the effective electronic conductivity of the slurry, but also the surface area per volume of the slurry. It was demonstrated that stable, flowable slurries could be created with these carbons, even when the carbons where fully loaded with iron during charging. During the two-year program, laboratory-scale 50 cm 2 single-cells demonstrated charge-discharge cycling at current densities of 150 mA/cm 2 and energy efficiencies of >50%. Meanwhile, cost modeling continued to show promise for this energy storage approach. Following the conversion period, the overall objective of the nominally 18-month plus up period was to close the technical gap of a 25X scale-up and reduce risk to enable commercial investment required to develop a slurry iron flow battery product. A specific objective was to demonstrate 150 cycles of performance exceeding 70% energy efficiency with a 10-cell stack of 1156 cm 2 cells operating at 100 mA/cm 2 . The 10-cell short stack was designed to have a power delivery rating of approximately 1 kW, which represents a size commensurate with the requirements of our commercialization partner at that time. A complete battery support system for the full-size cell was designed, fabricated, and tested. This includes pumps (multiple types were considered, and constant volume hose pumps were found to be acceptable for performance and cost), electrolyte rebalance reactor, and sensors and control systems. The support system proved to be well-behaved and worked well. A new low-cost membrane was developed for this system based on a microporous polymer support with a layer of hydrogel polymer. The membrane had low resistivity, and significantly reduced pressure driven fluid flow across the membrane (as compared to the un-coated microporous support). This membrane approach was scaled-up in conjunction with a commercial toll coating company, and a large amount of membrane was manufactured to support large cell and stack testing. A new concept for an electrolyte rebalance reactor to react hydrogen gas from the negative electrolyte (from plating inefficiency) with excess ferric ion in the positive electrolyte was developed. This design was very simple, low cost, and effective. The full scale 1156 cm 2 single-cell was designed, modeled, fabricated and tested. Modeling showed that uniform flow-distribution should be obtained with the designed flow distributor. The design took into consideration the need to stack the cells, and components were fabricated for a ten-cell stack. Testing of the cell for charge and discharge demonstrated current densities of 50 mA/cm 2 with efficiencies of 50%. However, optimization and long cycling could not be accomplished because the slurry electrode would plug within the cell and block the flow. This would shut down the cell after several charge-discharge cycles. Various approaches were pursued to address this problem with some progress. However, there was not adequate time to find solutions to the problem although several promising approaches were identified for future work. The final deliverable for this proposed 18-month effort was to be a 1 kW, 6 kWh slurry flow battery, with complete balance of plant, ready for testing at an ARPA-E ‘Charges’ site with an estimated capital cost less than $\$ $45/kWh for a six-hour system (excluding power electronics). The deliverable milestone was not achieved for reasons mentioned above related to the slurry plugging. At the completion of this project, it was determined that the cost targets are achievable. The cost estimate for a 1 MW system with a one-hour capacity was estimated at $\$ $140/kWh (including power electronics) and for a six-hour system was estimated at $\$ $40/kWh (including power electronics). These costs include the battery, the balance of plant, the power conditioning and a concrete pad. This program resulted in three patents including the slurry-based iron flow battery, the rebalance reactor, and the membrane separator. At the time of this report, slurry-based iron flow battery patents have been issued in the US, China, Japan, Korea, and Europe. The other patent applications were under various stages of examination. A licensee was identified and the company initiated a scale-up and development program to take this technology to the marketplace, and CWRU staff have transferred their know-how and knowledge base to the company. This technology will enhance the economic and energy security of the United States by enabling the use of intermittent renewable energy technology, such as wind and solar power, as dispatchable resources. This report summarizes the results of this research and development program.

25 ENERGY STORAGE↗

Integrated Simulation of Weld Residual Stress Evolution and Crack Propagation Using XFEM

Nuclear power plant components operate in environments that promote multiple degradation mecha- nisms, several of which involve crack initiation and growth. An ongoing effort in the U.S. Department of Energy’s Nuclear Energy Advanced Modeling and Simulation (NEAMS) program is developing a general capability within the Multiphysics Object Oriented Simulation Environment (MOOSE) framework for simulating three-dimensional crack growth under a range of driving conditions, including fatigue, stress corrosion cracking (SCC), brittle fracture, and stress-relaxation cracking. This report demonstrates an end-to-end workflow that uses this capability to model weld-residual-stress-driven SCC in the J-groove weld of a pressurized-water reactor control rod drive mechanism penetration in the vessel head. The workflow consists of a thermomechanical welding simulation with temperature-dependent plasticity, followed by cooldown to ambient conditions, and a restart of the simulation using the MOOSE extended finite element method (XFEM) module to propagate a three-dimensional crack through the residual stress field. New welding capabilities were developed to properly initialize newly activated elements in the weld region, and robustness improvements were made to the mesh-based algorithm for defining cutting planes in the 3D XFEM algorithm, allowing it to handle complex crack fronts and stress fields. Together these advances allowed the simulated SCC crack to grow from an initial elliptical flaw in the weld, across the weld, through the tube wall, and almost to the triple point (where the weld, tube, and reactor pressure vessel head intersect) over roughly 36 years of simulated service. These results demonstrate a workflow that can be extended to fully three-dimensional welding simulations and more complex crack interaction problems.

42 - ENGINEERING↗

CO2 MONITORING FOR DEMAND CONTROLLED VENTILATION IN COMMERCIAL BUILDINGS

Carbon dioxide (CO{sub 2}) sensors are often deployed in commercial buildings to obtain CO{sub 2} data that are used, in a process called demand-controlled ventilation, to automatically modulate rates of outdoor air ventilation. The objective is to keep ventilation rates at or above design specifications and code requirements and also to save energy by avoiding excessive ventilation rates. Demand controlled ventilation is most often used in spaces with highly variable and sometime dense occupancy. Reasonably accurate CO{sub 2} measurements are needed for successful demand controlled ventilation; however, prior research has suggested substantial measurement errors. Accordingly, this study evaluated: (a) the accuracy of 208 CO{sub 2} single-location sensors located in 34 commercial buildings, (b) the accuracy of four multi-location CO{sub 2} measurement systems that utilize tubing, valves, and pumps to measure at multiple locations with single CO{sub 2} sensors, and (c) the spatial variability of CO{sub 2} concentrations within meeting rooms. The field studies of the accuracy of single-location CO{sub 2} sensors included multi-concentration calibration checks of 90 sensors in which sensor accuracy was checked at multiple CO{sub 2} concentrations using primary standard calibration gases. From these evaluations, average errors were small, -26 ppm and -9 ppm at 760 and 1010 ppm, respectively; however, the averages of the absolute values of error were 118 ppm (16%) and 138 ppm (14%), at concentrations of 760 and 1010 ppm, respectively. The calibration data are generally well fit by a straight line as indicated by high values of R{sup 2}. The Title 24 standard specifies that sensor error must be certified as no greater than 75 ppm for a period of five years after sensor installation. At 1010 ppm, 40% of sensors had errors greater than {+-}75 ppm and 31% of sensors has errors greater than {+-}100 ppm. At 760 ppm, 47% of sensors had errors greater than {+-}75 ppm and 37% of sensors had errors greater than {+-}100 ppm. A significant fraction of sensors had errors substantially larger than 100 ppm. For example, at 1010 ppm, 19% of sensors had an error greater than 200 ppm and 13% of sensors had errors greater than 300 ppm. The field studies also included single-concentration calibration checks of 118 sensors at the concentrations encountered in the buildings, which were normally less than 500 ppm during the testing. For analyses, these data were combined with data from the calibration challenges at 510 ppm obtained during the multi-concentration calibration checks. For the resulting data set, the average error was 60 ppm and the average of the absolute value of error was 154 ppm. Statistical analyses indicated that there were statistically significant differences between the average accuracies of sensors from different manufacturers. Sensors with a 'single lamp single wavelength' design tended to have a statistically significantly smaller average error than sensors with other designs except for 'single lamp dual wavelength' sensors, which did not have a statistically significantly lower accuracy. Sensor age was not consistently a statistically significant predictor of error.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Control oriented model order reduction for the hydraulic system of an autonomous wheel loader

Control design for an autonomous wheel loader requires a high-fidelity and low-order dynamics model. The objective of order reduction is to decrease the number of states in the model while maintaining performance comparable to the original model. Of all the power components in a wheel loader, the model order reduction for steering and working hydraulic systems is most demanding due to its nonlinear and stiff dynamical nature. This paper describes a physics-inspired model order reduction method that can reduce the model order by close to thirty percent. It is achieved by discarding orders with fast dynamics and consolidating multiple states into less new states. Furthermore, this method can also be extended to other types of off-road vehicles such as excavators, bulldozers, etc. Simulation results demonstrate that the reduced-order model exhibits very similar performance to the full-order model under typical operating conditions of wheel loaders, with an output error of less than six percent.

42 ENGINEERING↗

A systems biology perspective on Nrf2-mediated antioxidant response

Cells in vivo are constantly exposed to reactive oxygen species (ROS) generated endogenously and exogenously. To defend against the deleterious consequences of ROS, cells contain multiple antioxidant enzymes expressed in various cellular compartments to scavenge these toxic species. Under oxidative stresses, these antioxidant enzymes are upregulated to restore redox homeostasis. Such an adaptive response results from the activation of a redox-sensitive gene regulatory network mediated by nuclear factor E2-related factor 2. To more completely understand how the redox control system is designed by nature to meet homeostatic goals, we have examined the network from a systems perspective using engineering approaches. As with man-made control devices, the redox control system can be decomposed into distinct functional modules, including transducer, controller, actuator, and plant. Cells achieve specific performance objectives by utilizing nested feedback loops, feedforward control, and ultrasensitive signaling motifs, etc. Given that endogenously generated ROS are also used as signaling molecules, our analysis suggests a novel mode of action to explain oxidative stress-induced pathological conditions and diseases. Specifically, by adaptively upregulating antioxidant enzymes, oxidative stress may inadvertently attenuate ROS signals that mediate physiological processes, resulting in aberrations of cellular functions and adverse consequences. Lastly, by simultaneously considering the two competing cellular tasks-adaptive antioxidant defense and ROS signaling-we re-examine the premise that dietary antioxidant supplements is generally beneficial to human health. Our analysis highlights some possible adverse effects of these widely consumed antioxidants.

60 APPLIED LIFE SCIENCES↗

High-Silicon Steel Strip by Single-Step Shear Deformation Processing

It is well known that Fe-Si alloys with Si content higher than in conventional electrical sheet steels (>3.2% Si) can make a significant impact in improving the efficiency of electrical motors if they are available in sheet/foil (strip) forms at suitable cost. While the magnetic and electrical attributes (e.g., resistivity, core loss) of these high-Si Fe alloys, of relevance to electrical motor core laminations, are known to be exceptional, the alloys have limited workability, making them difficult to produce consistently in sheet/foil (strip) forms. Current processing techniques such as rolling, while adequate for producing conventional electrical steel sheet, do have important disadvantages - large energy consumption and emissions, limitations in processing of low-workability alloys (e.g., high-Si content steels), large-scale plant infrastructure, and less than adequate capability to engineer sheet metals with specific microstructures (e.g., fine-grained) and crystallographic textures (e.g., shear textures). It is therefore of interest to have an alternative commercial process that can produce sheet/foil (strip) from high-Si Fe alloys and which can also overcome some of the deficiencies of current multistage strip processes. The goal of the present project was design and demonstration of a new energy-efficient pilot process for producing high-Si electrical steel strip of commercial widths and thickness, and with superior electrical and magnetic properties than current electrical steels (Fe-3.2% Si as benchmark). The applications domain for these steels is electrical motor core laminations. We have addressed this goal by accomplishment of the following specific objectives and tasks: a) Development of an Fe-4Si-4Cr alloy with electrical resistivity >80 μΩ-cm, induction flux density >1.48 T at 5000 A/m and core loss 35% lower than the benchmark 3.2% Si alloy. The alloy which meets DOE target specifications for motor core attributes was designed with the Si content controlled for the electrical properties and the Cr content tailored to meet material/process workability requirements. b) A unique machining-based deformation processing system was designed and scaled-up to produce strip of commercial width (25 mm to 150 mm) and thickness (up to 0.5 mm) from the Fe-4Si-4Cr alloy and other alloys of varied workability including copper, Al6061-T6 and naval brass. The key attributes of the machining-based strip production are deformation processing by concentrated simple-shear; single-step production of strip from ingot using compact machine infrastructure; strip surface finish of Ra 0.35 to 1 micrometer that is comparable/superior to that of rolled strip; discrete production of strip that can potentially be done at point of use; and controllability of strip mechanical/formability properties by deformation control. c) The electrical, magnetic, surface quality, mechanical, formability, and metallurgical properties/attributes of the machining-based strip were established by direct ASTM standard or equivalent measurement techniques. d) Punching characteristics of the strip in terms of load, edge quality and macro defects were similar to those of conventional 3.2% Si electrical steels. These punching characteristics are critical from a manufacturability perspective for motor/transformer core applications. e) A modeling framework for energy analysis of multistage rolling and the machining-based deformation processing has been established. Application of this modeling to the two strip-processes showed that the machining-based process requires significantly lower specific energy for processing, ~ 25% of that for rolling. The modeling framework can be adapted for a range of sheet-metal forming, bulk metal forming, and machining processes. It can be used to identify key parameters controlling process specific energy. f) A comparative analysis of advantages and disadvantages of machining-based processing against rolling for strip production. The single stage machining-based processing, with compact infrastructure, represents a new manufacturing paradigm for sheet and foil manufacturing that can potentially also be applied to advanced titanium, aluminum, copper and magnesium alloys. The goals and objectives were accomplished by a cross-disciplinary project team comprising of personnel from Purdue University; M4 Sciences LLC, a small-business focused on advanced manufacturing technology development; the Pacific Northwest National Labs; and tool manufacturers. The team is currently in advanced discussions with multiple entities for future process development for commercialization.

36 MATERIALS SCIENCE↗

A Pareto analysis approach to assess relevant marginal CO{sub 2} footprint for petroleum products

Recently, linear programing (LP) models have been extended to track the marginal CO{sub 2} intensity of automotive fuels at the refinery gate. The obtained CO{sub 2} data are recommended for policy making because they capture the economic and environmental tensions as well as the processing effects related to oil products. However, they are proven to be extremely sensitive to small perturbations and therefore useless in practice. In this paper, we first investigate the theoretical reasons of this drawback. Then, we develop a multiple objective LP framework to assess relevant marginal CO{sub 2} footprints that preserve both defensibility and stability at a satisfactory level of acceptance. A case study illustrates this new methodology. - Highlights: • Refining LP models have limitations to provide useful marginal CO{sub 2} footprints. • A multi objective optimization framework is developed to assess relevant CO{sub 2} data. • Within a European Refinig industry, diesel is more CO{sub 2} intensive than gasoline.

54 ENVIRONMENTAL SCIENCES↗

PJM Controller Testing with Prototypic PJM Nozzle Configuration

The U.S. Department of Energy (DOE) Office of River Protection’s Waste Treatment Plant (WTP) is being designed and built to pre-treat and then vitrify a large portion of the wastes in Hanford’s 177 underground waste storage tanks. The WTP consists of three primary facilities—pretreatment, low-activity waste (LAW) vitrification, and high-level waste (HLW) vitrification. The pretreatment facility will receive waste piped from the Hanford tank farms and separate it into a high-volume, low-activity liquid stream stripped of most solids and radionuclides and a much smaller volume of HLW slurry containing most of the solids and most of the radioactivity. Many of the vessels in the pretreatment facility will contain pulse jet mixers (PJM) that will provide some or all of the mixing in the vessels. Pulse jet mixer technology was selected for use in black cell regions of the WTP, where maintenance cannot be performed once hot testing and operations commence. The PJMs have no moving mechanical parts that require maintenance. The vessels with the most concentrated slurries will also be mixed with air spargers and/or steady jets in addition to the mixing provided by the PJMs. Pulse jet mixers are susceptible to overblows that can generate large hydrodynamic forces, forces that can damage mixing vessels or their internal parts. The probability of an overblow increases if a PJM does not fill completely. The purpose of the testing performed for this report was to determine how reliable and repeatable the primary and safety (or backup) PJM control systems are at detecting drive overblows (DOB) and charge vessel full (CVF) conditions. Testing was performed on the ABB 800xA and Triconex control systems. The controllers operated an array of four PJMs installed in an approximately 13 ft diameter × 15 ft tall tank located in the high bay of the Pacific Northwest National Laboratory (PNNL) 336 Building test facility. The PJMs were fitted with 4 inch diameter discharge nozzles representative of the nozzles to be used in the WTP. This work supplemented earlier controller tests done on PJMs with 2 inch nozzles (Bontha et al. 2007). Those earlier tests enabled the selection of appropriate pressure transmitters with associated piping and resulted in an alternate overblow detection algorithm that uses data from pressure transmitters mounted in a water flush line on the PJM airlines. Much of that earlier work was only qualitative, however, due to a data logger equipment failure that occurred during the 2007 testing. The objectives of the current work focused on providing quantitative determinations of the ability of the BNI controllers to detect DOB and CVF conditions. On both control systems, a DOB or CVF is indicated when the values of particular internal functions, called confidence values, cross predetermined thresholds. There are two types of confidence values; one based on a transformation of jet pump pair (JPP) drive and suction pressures, the other based on the pressure in the flush line. In the present testing, we collected confidence levels output from the ABB and Triconex controllers. These data were analyzed in terms of the true and noise confidence peaks generated during multiple cycles of DOB and CVF events. The distributions of peak and noise amplitudes were compared to see if thresholds could be set that would enable the detection of DOB and CVF events at high probabilities, while keeping false detections to low probabilities. Supporting data were also collected on PJM operation, including data on PJM pressures and levels, to provide direct experimental evidence of when PJMs were filling, full, driving, or overblowing.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

The relationship between urinary selenium levels and risk of gestational diabetes mellitus: A nested case–control study

Background Selenium (Se) is an essential trace element for the human body. Serum Se and urinary Se are also biomarkers to assess Se exposure status. However, studies focusing on the association between urinary Se and the risk of gestational diabetes mellitus (GDM) are rare. Objective To investigate the association between urinary Se and the risk of GDM. Methods A nested case–control study based on a prospective birth cohort in Wuhan, China, which focuses on the effects of prenatal environmental factors exposure on pregnant women and children’s health was conducted. Two hundred and twenty-six cases and 452 controls were included. Maternal urine samples were collected before GDM diagnosis, and the urinary Se levels were determined. We assessed the association of urinary Se with GDM by conditional logistic regression with maternal urinary Se level as a categorical variable, and estimated the association between Se and glucose levels by multiple linear regression. The potential modifier roles of maternal age and fetal sex have also been assessed. Results Lower urinary level of Se was significantly associated with a higher risk of GDM (OR = 2.35 for the tertile 1, 95% CI:1.36-4.06; adjusted OR = 1.79 for the tertile 2, 95%CI:1.09-2.95; p for trend = 0.01). Fetal sex had an interaction with Se in the association with GDM. The association was more pronounced among pregnant women with female fetuses than with male fetuses. Discussion Our study suggested a significant negative association between urinary Se and the risk of GDM, and this association may vary depending on the fetal sex.

Liu, Yuanxia↗

Secure Video Surveillance System (SVSS) for unannounced safeguards inspections.

The Secure Video Surveillance System (SVSS) is a collaborative effort between the U.S. Department of Energy (DOE), Sandia National Laboratories (SNL), and the Brazilian-Argentine Agency for Accounting and Control of Nuclear Materials (ABACC). The joint project addresses specific requirements of redundant surveillance systems installed in two South American nuclear facilities as a tool to support unannounced inspections conducted by ABACC and the International Atomic Energy Agency (IAEA). The surveillance covers the critical time (as much as a few hours) between the notification of an inspection and the access of inspectors to the location in facility where surveillance equipment is installed. ABACC and the IAEA currently use the EURATOM Multiple Optical Surveillance System (EMOSS). This outdated system is no longer available or supported by the manufacturer. The current EMOSS system has met the project objective; however, the lack of available replacement parts and system support has made this system unsustainable and has increased the risk of an inoperable system. A new system that utilizes current technology and is maintainable is required to replace the aging EMOSS system. ABACC intends to replace one of the existing ABACC EMOSS systems by the Secure Video Surveillance System. SVSS utilizes commercial off-the shelf (COTS) technologies for all individual components. Sandia National Laboratories supported the system design for SVSS to meet Safeguards requirements, i.e. tamper indication, data authentication, etc. The SVSS consists of two video surveillance cameras linked securely to a data collection unit. The collection unit is capable of retaining historical surveillance data for at least three hours with picture intervals as short as 1sec. Images in .jpg format are available to inspectors using various software review tools. SNL has delivered two SVSS systems for test and evaluation at the ABACC Safeguards Laboratory. An additional 'proto-type' system remains at SNL for software and hardware testing. This paper will describe the capabilities of the new surveillance system, application and requirements, and the design approach.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

SU-F-J-143: Initial Assessment of Image Quality of An Integrated MR-Linac System with ACR Phantom

Purpose/Objective(s): To assess the image quality of an integrated MR-Linac system and compare with other MRI systems that are primarily used for diagnostic purposes. Materials/Methods: An ACR MRI quality control (QC) phantom was used to evaluate the image quality of a fully integrated 1.5T MRI-Linac system recently installed at our institution. This system has a new split magnet design which gives the magnetic field strength of 1.5T. All images were acquired with a set of phased-array surface coils which are designed to have minimal attention of radiation beam. The anterior coil rests on a coil holder which keeps the anterior coil’s position consistent for QA purposes. The posterior coil is imbedded in the patient couch. Multiple sets of T1, T2/PD images were acquired using the protocols as prescribed by the ACR on three different dates, ranging 3 months apart. Results: The geometric distortion are within 0.5 mm in the axial scans and within 1mm in the saggital (z-direction) scans. Slice thickness accuracy, image uniformity, ghosting ratio, high contrast detectability are comparable to other 1.5T diagnostic MRI scanners. The low-contrast object detectability are lower comparatively, which is a result of using the body array coil. Additionally, the beam’s-eye-view images (oblique coronal and saggital images) have minimal geometric distortion at all linac gantry angles tested. No observable changes or drift in image quality is found from images acquired 3 month apart. Conclusion: Despite the use of a body array surface coil, the image quality is comparable to that of an 1.5T MRI scanner and is of sufficient quality to pass the ACR MRI accreditation program. The geometric distortion of the MRI system of the integrated MR-Linac is within 1mm for an object size similar to the ACR phantom, sufficient for radiation therapy treatment purpose. The authors received corporate sponsored research grants from Elekta which is the vendor for the MR-Linac evaluated in this study.

60 APPLIED LIFE SCIENCES↗

How Low Can You Go? Using Synthetic 3D Imagery to Drastically Reduce Real-World Training Data for Object Detection

Deep convolutional neural networks (DCNNs) currently provide state-of-the-art performance on image classification and object detection tasks, and there are many global security mission areas where such models could be extremely useful. Crucially, the success of these models is driven in large part by the widespread availability of high-quality open source data sets such as Image Net, Common Objects in Context (COCO), and KITTI, which contain millions of images with thousands of unique labels. However, global security relevant objects-of-interest can be difficult to obtain: relevant events are low frequency and high consequence; the content of relevant images is sensitive; and adversaries and proliferators seek to obscure their activities. For these cases where exemplar data is hard to come-by, even fine-tuning an existing model with available data can be effectively impossible. Recent work demonstrated that models can be trained using a combination of real-world and synthetic images generated from 3D representations; that such models can exceed the performance of models trained using real-world data alone; and that the generated images need not be perfectly realistic (Tremblay, et al., 2018). However, this approach still required hundreds to thousands of real-world images for training and fine tuning, which for sparse, global security-relevant datasets can be an unrealistic hurdle. In this research, we validate the performance and behavior of DCNN models as we drive the number of real-world images used for training object detection tasks down to a minimal set. We perform multiple experiments to identify the best approach to train DCNNs from an extremely small set of real-world images. In doing so, we: Develop state-of-the-art, parameterized 3D models based on real-world images and sample from their parameters to increase the variance in synthetic image training data; Use machine learning explainability techniques to highlight and correct through targeted training the biases that result from training using completely synthetic images; and Validate our results by comparing the performance of the models trained on synthetic data to one another, and to a control model created by fine-tuning an existing ImageNet-trained model with a limited number (hundreds) of real-world images.

97 MATHEMATICS AND COMPUTING↗