Search NASA⌕ Search

SEARCH · Search NASA

Results for “Data Variables”

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 505 records · Page 28

Statistical optimisation techniques in fatigue signal editing problem

Success in fatigue signal editing is determined by the level of length reduction without compromising statistical constraints. A great reduction rate can be achieved by removing small amplitude cycles from the recorded signal. The long recorded signal sometimes renders the cycle-to-cycle editing process daunting. This has encouraged researchers to focus on the segment-based approach. This paper discusses joint application of the Running Damage Extraction (RDE) technique and single constrained Genetic Algorithm (GA) in fatigue signal editing optimisation.. In the first section, the RDE technique is used to restructure and summarise the fatigue strain. This technique combines the overlapping window and fatigue strain-life models. It is designed to identify and isolate the fatigue events that exist in the variable amplitude strain data into different segments whereby the retention of statistical parameters and the vibration energy are considered. In the second section, the fatigue data editing problem is formulated as a constrained single optimisation problem that can be solved using GA method. The GA produces the shortest edited fatigue signal by selecting appropriate segments from a pool of labelling segments. Challenges arise due to constraints on the segment selection by deviation level over three signal properties, namely cumulative fatigue damage, root mean square and kurtosis values. Experimental results over several case studies show that the idea of solving fatigue signal editing within a framework of optimisation is effective and automatic, and that the GA is robust for constrained segment selection.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Popnet : computer vision based deep learning model for forecasting gridded population

Here, this study introduces Popnet, a deep learning model for forecasting 1 km-gridded populations, integrating U-Net, ConvLSTM, a Spatial Autocorrelation module and deep ensemble methods. Using spatial variables and population data from 2000 to 2020, Popnet predicts South Korea’s population trends by age groups (under 14, 15-64 and over 65) up to 2040. In validation, it outperforms traditional machine learning and state-of-the-art computer vision models. The output of this model discovered significant polarisation: population growth in urban areas, especially the capital region, and severe depopulation in rural areas. Popnet is a robust tool for offering significant insights to policymakers and related stakeholders about the detailed future population, which allows them to establish detailed, localised planning and resource allocations.

computer vision↗

Laser interference structuring of Cu for adhesive joining of HVAC equipment

Due to potential energy and cost savings benefits, adhesive joining has been recently considered for heating, ventilation, air conditioning, and refrigeration (HVAC&R) systems. HVAC adhesive bonding requires cost-effective and adequate surface preparations of adherents. Here, this article investigates the use of abrasion, traditional single-beam laser, and a laser-interference technique as surface preparations of copper surfaces. Surface morphology is characterized using scanning electron microscopy (SEM) and atomic force microscopy (AFM). Effective submicrometer peak-to-valley structuring with a periodicity of ∼2.7μ⁢m was demonstrated for laser-interference processing. Single-lap shear tests were conducted for 89 joints made with bondline thicknesses of 0.15 and 0.3 mm. Data on process variables and measured variables included open-time, bond length, maximum load, displacement at failure, shear lap strength, and failure mode. A statistical analysis was conducted on each lot to determine the lower limit with 95% confidence intervals for displacements at failure and shear lap strengths. A comparison is presented between the properties of laser-structured joints with respect to those prepared by abrasion, which is considered the baseline surface preparation technique. Based on this comparison, one single-beam laser technique and two laser-interference techniques were shown to exhibit vastly superior performance over the joints made with abraded specimens.

Sabau, Adrian S. [Oak Ridge National Laboratory (O↗

QUASI-BIENNIAL MODULATION OF GALACTIC COSMIC RAYS

The time variability of the cosmic-ray (CR) intensity at three different rigidities has been analyzed through the empirical mode decomposition technique for the period 1964-2004. Apart from the {approx}11 yr cycle, quasi-biennial oscillations (QBOs) have been detected as a prominent scale of variability in CR data, as well as in the heliomagnetic field magnitude at 1 AU and in the sunspot area. The superposition of the {approx}11 yr and QBO contributions reproduces the general features of the CR modulation, such as most of the step-like decreases and the Gnevyshev Gap phenomenon. A significant correlation has also been found between QBOs of the heliospheric magnetic field and the CR intensity during even solar activity cycles, suggesting that the former are responsible for step-like decreases in CR modulation, probably dominated by the particle diffusion/convection in such periods. In contrast, during odd-numbered cycles, no significant correlation is found. This could be explained with an enhanced drift effect also during the solar maximum or a greater influence of merged interaction regions at great heliocentric distances during odd cycles. Moreover, the QBOs of CR data are delayed with respect to sunspot activity, the lag being shorter for A > 0 periods of even cycles ({approx}1-4 months) than for A < 0 periods of odd cycles ({approx}7-9 months); we suggest that solar QBOs also affect the recovery of the CR intensity after the solar activity maximum.

79 ASTRONOMY AND ASTROPHYSICS↗

THE DENSEST GALAXY

We report the discovery of a remarkable ultra-compact dwarf galaxy around the massive Virgo elliptical galaxy NGC 4649 (M60), which we call M60-UCD1. With a dynamical mass of 2.0 × 10{sup 8} M {sub ☉} but a half-light radius of only ∼24 pc, M60-UCD1 is more massive than any ultra-compact dwarfs of comparable size, and is arguably the densest galaxy known in the local universe. It has a two-component structure well fit by a sum of Sérsic functions, with an elliptical, compact (r{sub h} = 14 pc; n ∼ 3.3) inner component and a round, exponential, extended (r{sub h} = 49 pc) outer component. Chandra data reveal a variable central X-ray source with L{sub X} ∼ 10{sup 38} erg s{sup –1} that could be an active galactic nucleus associated with a massive black hole or a low-mass X-ray binary. Analysis of optical spectroscopy shows the object to be old (∼> 10 Gyr) and of solar metallicity, with elevated [Mg/Fe] and strongly enhanced [N/Fe] that indicates light-element self-enrichment; such self-enrichment may be generically present in dense stellar systems. The velocity dispersion (σ ∼ 70 km s{sup –1}) and resulting dynamical mass-to-light ratio (M/L{sub V} = 4.9 ± 0.7) are consistent with—but slightly higher than—expectations for an old, metal-rich stellar population with a Kroupa initial mass function. The presence of a massive black hole or a mild increase in low-mass stars or stellar remnants is therefore also consistent with this M/L{sub V} . The stellar density of the galaxy is so high that no dynamical signature of dark matter is expected. However, the properties of M60-UCD1 suggest an origin in the tidal stripping of a nucleated galaxy with M{sub B} ∼ –18 to –19.

79 ASTRONOMY AND ASTROPHYSICS↗

IDENTIFICATION OF THE HIGH-ENERGY GAMMA-RAY SOURCE 3FGL J1544.6–1125 AS A TRANSITIONAL MILLISECOND PULSAR BINARY IN AN ACCRETING STATE

We present X-ray, ultraviolet, and optical observations of 1RXS J154439.4–112820, the most probable counterpart of the unassociated Fermi-LAT source 3FGL J1544.6–1125. The optical data reveal rapid variability, which is a feature of accreting systems. The X-rays exhibit large-amplitude variations in the form of fast switching (within ∼10 s) between two distinct flux levels that differ by a factor of ≈10. The detailed optical and X-ray behavior is virtually identical to that seen in the accretion-disk-dominated states of the transitional millisecond pulsar (MSP) binaries PSR J1023+0038 and XSS J12270–4859, which are also associated with γ-ray sources. Based on the available observational evidence, we conclude that 1RXS J154439.4–112820 and 3FGL J1544.6–1125 are the same object, with the X-rays arising from intermittent low-luminosity accretion onto an MSP and the γ-rays originating from an accretion-driven outflow. 1RXS J154439.4–112820 is only the fourth γ-ray-emitting low-mass X-ray binary system to be identified and is likely to sporadically undergo transformations to a non-accreting rotation-powered pulsar system.

79 ASTRONOMY AND ASTROPHYSICS↗

Using Weather Data and Climate Model Output in Economic Analyses of Climate Change

Economists are increasingly using weather data and climate model output in analyses of the economic impacts of climate change. This article introduces a set of weather data sets and climate models that are frequently used, discusses the most common mistakes economists make in using these products, and identifies ways to avoid these pitfalls. We first provide an introduction to weather data, including a summary of the types of datasets available, and then discuss five common pitfalls that empirical researchers should be aware of when using historical weather data as explanatory variables in econometric applications. We then provide a brief overview of climate models and discuss two common and significant errors often made by economists when climate model output is used to simulate the future impacts of climate change on an economic outcome of interest.

54 ENVIRONMENTAL SCIENCES↗

Investigation of magnetic nanoparticles for the rapid extraction and assay of alpha-emitting radionuclides from urine: Demonstration of a novel radiobioassay method

In the event of an accidental or intentional release of radionuclides into a populated area, three things must occur in a timely manner: food and drinking water supplies must be determined to be safe to eat / drink, civilians and/or military personnel must be surveyed to ensure that they do not have external contamination, and they must be screened to ensure that significant ingestion or inhalation of radionuclides has not occurred (this paper is concerned with the latter). In the event of such a disaster, the volume of radiobioassays to be performed would be tremendous. If the event released significant levels of β- or α-emitting radionuclides, in vivo assays would be ineffective. Therefore, highly efficient and rapid analytical methods for radionuclide detection from submitted spot urine samples (≤ 50 mL) would be required. At present, the quantitative determination of α-emitting radionuclides from urine samples is highly labor intensive, and requires significant sample preparation and analysis time. Sorbent materials that provide effective collection and enable rapid assay could significantly streamline the radioanalytical process. We have demonstrated the use of paramagnetic nanoparticles as a novel class of extracting media for four α-emitting radionuclides of concern (Po, Ra, Am, and U) from chemically unmodified and pH 2 human urine. Herein the initial experimental sorption results are presented along with a novel method that utilizes paramagnetic nanoparticles for the extraction of radionuclides from unmodified human urine followed by the magnetic field-induced collection of the particles for subsequent α-counting-source preparation. Additionally, we construct a versatile human dose model that determines the detector count times required to estimate internal human dose at specific protective action thresholds. The model provides a means to assess a method’s detection capabilities and use fundamental health physics parameters and actual experimental data as core variables. The modeling shows that with effective sorbent materials, rapid screening for internalized α-emitters is possible from a 50 mL spot urine sample volume collected within one week of exposure/intake.

59 BASIC BIOLOGICAL SCIENCES↗

Scaling properties of fractional momentum loss of high- p T hadrons in nucleus-nucleus collisions at s N N from 62.4 GeV to 2.76 TeV

We present measurements of the fractional momentum loss (S loss = delta pT / pT ) of high-transverse-momentum-identified hadrons in heavy-ion collisions. Using pi 0 in Au + Au and Cu + Cu collisions at √s NN = 62.4 and 200 GeV measured by the PHENIX experiment at the Relativistic Heavy Ion Collider and and charged hadrons in Pb + Pb collisions measured by the ALICE experiment at the Large Hadron Collider, we studied the scaling properties of S loss as a function of a number of variables: the number of participants, N part , the number of quark participants, N qp , the charged-particle density, dN ch /d η , and the Bjorken energy density times the equilibration time, epsilon Bj τ 0 . We also find that the p T , where S loss has its maximum, varies both with centrality and collision energy. Above the maximum, S loss tends to follow a power-law function with all four scaling variables. Finally, the data at √s NN = 200 GeV and 2.76 TeV, for sufficiently high particle densities, have a common scaling of S loss with dN ch /d η and ε Bj τ 0 , lending insight into the physics of parton energy loss.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Methodology for Estimating Thermal and Neutron Embrittlement of Austenitic Stainless Steel Welds During Service in Light Water Reactors

The effect of thermal aging on the degradation of fracture toughness and Charpy-impact properties of austenitic stainless steel (SS) welds has been characterized at reactor temperatures. The solidification behavior and the distribution and morphology of the ferrite phase in SS welds are described. Thermal aging of the welds results in moderate decreases in Charpy-impact strength and fracture toughness. The upper-shelf Charpy-impact energy of aged welds decreases by 50–80 J/cm2. The decrease in fracture toughness J-R curve, or JIc is relatively small. Thermal aging has minimal effect on the tensile strength. The fracture properties of SS welds are insensitive to filler metal; the welding process has a significant effect. The large variability in the data makes it difficult to establish the effect of the welding process on fracture properties of SS welds. Consequently, the approach used for evaluating thermal and neutron embrittlement of austenitic SS welds relies on establishing a lower-bound fracture toughness J-R curve for unaged and aged, and non-irradiated and irradiated, SS welds. The existing fracture toughness J-R curve data for SS welds have been reviewed and evaluated to define lower-bound J-R curve for submerged arc (SA)/shielded metal arc (SMA)/manual metal arc (MMA) welds and gas tungsten arc (GTA)/tungsten inert gas (TIG) welds in the unaged and aged conditions. At reactor temperatures, the fracture toughness of GTA/TIG welds is a factor of about 2.3 higher than that of SA/SMA/MMA welds. Thermal aging decreases the fracture toughness by about 20%. The potential combined effects of thermal and neutron embrittlement of austenitic SS welds are also described. Lower-bound curves are presented that define the change in coefficient C and exponent n of the power-law J-R curve and the JIc value for SS welds as a function of neutron dose. The potential effects of reactor coolant environment on the fracture toughness of austenitic SS welds are also discussed.

Chopra, O. K.↗

Energy spectrum control for modulated proton beams

In proton therapy delivered with range modulated beams, the energy spectrum of protons entering the delivery nozzle can affect the dose uniformity within the target region and the dose gradient around its periphery. For a cyclotron with a fixed extraction energy, a rangeshifter is used to change the energy but this produces increasing energy spreads for decreasing energies. This study investigated the magnitude of the effects of different energy spreads on dose uniformity and distal edge dose gradient and determined the limits for controlling the incident spectrum. A multilayer Faraday cup (MLFC) was calibrated against depth dose curves measured in water for nonmodulated beams with various incident spectra. Depth dose curves were measured in a water phantom and in a multilayer ionization chamber detector for modulated beams using different incident energy spreads. Some nozzle entrance energy spectra can produce unacceptable dose nonuniformities of up to {+-}21% over the modulated region. For modulated beams and small beam ranges, the width of the distal penumbra can vary by a factor of 2.5. When the energy spread was controlled within the defined limits, the dose nonuniformity was less than {+-}3%. To facilitate understanding of the results, the data were compared to the measured and Monte Carlo calculated data from a variable extraction energy synchrotron which has a narrow spectrum for all energies. Dose uniformity is only maintained within prescription limits when the energy spread is controlled. At low energies, a large spread can be beneficial for extending the energy range at which a single range modulator device can be used. An MLFC can be used as part of a feedback to provide specified energy spreads for different energies.

61 RADIATION PROTECTION AND DOSIMETRY↗

A pencil beam dose calculation model for CyberKnife system

Purpose: CyberKnife system is initially equipped with fixed circular cones for stereotactic radiosurgery. Two dose calculation algorithms, Ray-Tracing and Monte Carlo, are available in the supplied treatment planning system. A multileaf collimator system was recently introduced in the latest generation of system, capable of arbitrarily shaped treatment field. The purpose of this study is to develop a model based dose calculation algorithm to better handle the lateral scatter in an irregularly shaped small field for the CyberKnife system. Methods: A pencil beam dose calculation algorithm widely used in linac based treatment planning system was modified. The kernel parameters and intensity profile were systematically determined by fitting to the commissioning data. The model was tuned using only a subset of measured data (4 out of 12 cones) and applied to all fixed circular cones for evaluation. The root mean square (RMS) of the difference between the measured and calculated tissue-phantom-ratios (TPRs) and off-center-ratio (OCR) was compared. Three cone size correction techniques were developed to better fit the OCRs at the penumbra region, which are further evaluated by the output factors (OFs). The pencil beam model was further validated against measurement data on the variable dodecagon-shaped Iris collimators and a half-beam blocked field. Comparison with Ray-Tracing and Monte Carlo methods was also performed on a lung SBRT case. Results: The RMS between the measured and calculated TPRs is 0.7% averaged for all cones, with the descending region at 0.5%. The RMSs of OCR at infield and outfield regions are both at 0.5%. The distance to agreement (DTA) at the OCR penumbra region is 0.2 mm. All three cone size correction models achieve the same improvement in OCR agreement, with the effective source shift model (SSM) preferred, due to their ability to predict more accurately the OF variations with the source to axis distance (SAD). In noncircular field validation, the pencil beam calculated results agreed well with the film measurement of both Iris collimators and the half-beam blocked field, fared much better than the Ray-Tracing calculation. Conclusions: The authors have developed a pencil beam dose calculation model for the CyberKnife system. The dose calculation accuracy is better than the standard linac based system because the model parameters were specifically tuned to the CyberKnife system and geometry correction factors. The model handles better the lateral scatter and has the potential to be used for the irregularly shaped fields. Comprehensive validations on MLC equipped system are necessary for its clinical implementation. It is reasonably fast enough to be used during plan optimization.

60 APPLIED LIFE SCIENCES↗

Learning to Predict Physical Properties Using Sums of Separable Functions

We present an algorithm for learning the function that maps a material structure to its value on some property, given the value of this function on several structures. We pose this problem as one of learning (regressing) a function of many variables from scattered data. Each structure is first converted to a weighted set of points by a process that removes irrelevant translations and rotations but otherwise retains full information about the structure. Then, incorporating a weighted average for each structure, we construct the multivariate regression function as a sum of separable functions, following the paradigm of separated representations. The algorithm can treat all finite and periodic structures within a common framework, and in particular does not require all structures to lie on a common lattice. We show how the algorithm simplifies when the structures do lie on a common lattice, and we present numerical results for that case.

36 MATERIALS SCIENCE↗

FLIT: A Generic Fortran Library based on Interfaces and Templates

This Fortran code consists of multiple modules with a focus on simplifying array operations, image processing, and numerical computation especially for computational geophysics applications. We intend to use this code to demonstrate the application and usefulness of Fortran interface and templates for generic programming, especially for computational geophysics and seismology applications. The code has several notable features. Firstly, it is based on a modularized structure, where each module contains multiple functions but with a focus of functionality. Secondly, it heavily uses interfaces and templates for improving the genericness and convenience of the resulting code, where a same function interface can enclose a group of functions that perform the same functionality but with inputs/output variables of different data types. Thirdly, it includes a variety of generic functions with an emphasis on array operations, such as rotation, flipping, cropping, padding, fast Fourier transform, Gaussian blurring, interpolation, and so on. We name this package FLIP – a generic Fortran Library based on Interfaces and Templates.

Gao, Kai↗

Improved Mobility Control for Carbon Dioxide (CO{sub 2}) Enhanced Oil Recovery Using Silica-Polymer-Initiator (SPI) Gels

SPI gels are multi-component silicate based gels for improving (areal and vertical) conformance in oilfield enhanced recovery operations, including water-floods and carbon dioxide (CO{sub 2}) floods, as well as other applications. SPI mixtures are like-water when pumped, but form light up to very thick, paste-like gels in contact with CO{sub 2}. When formed they are 3 to 10 times stronger than any gelled polyacrylamide gel now available, however, they are not as strong as cement or epoxy, allowing them to be washed / jetted out of the wellbore without drilling. This DOE funded project allowed 8 SPI field treatments to be performed in 6 wells (5 injection wells and 1 production well) in 2 different fields with different operators, in 2 different basins (Gulf Coast and Permian) and in 2 different rock types (sandstone and dolomite). Field A was in a central Mississippi sandstone that injected CO{sub 2} as an immiscible process. Field B was in the west Texas San Andres dolomite formation with a mature water-alternating-gas miscible CO{sub 2} flood. Field A treatments are now over 1 year old while Field B treatments have only 4 months data available under variable WAG conditions. Both fields had other operational events and well work occurring before/ during / after the treatments making definitive evaluation difficult. Laboratory static beaker and dynamic sand pack tests were performed with Ottawa sand and both fields’ core material, brines and crude oils to improve SPI chemistry, optimize SPI formulations, ensure SPI mix compatibility with field rocks and fluids, optimize SPI treatment field treatment volumes and methods, and ensure that strong gels set in the reservoir. Field quality control procedures were designed and utilized. Pre-treatment well (surface) injectivities ranged from 0.39 to 7.9 MMCF/psi. The SPI treatment volumes ranged from 20.7 cubic meters (m{sup 3}, 5460 gallons/ 130 bbls) to 691 m{sup 3} (182,658 gallons/ 4349 bbls). Various size and types of chemical/ water buffers before and after the SPI mix ensured that pre-gelled SPI mix got out into the formation before setting into a gel. SPI gels were found to be 3 to 10 times stronger than any commercially available cross-linked polyacrylamide gels based on Penetrometer and Bulk Gel Shear Testing. Because of SPI’s unique chemistry with CO{sub 2}, both laboratory and later field tests demonstrated that multiple, smaller volume SPI treatments maybe more effective than one single large SPI treatment. CO{sub 2} injectivities in injection well in both fields were reduced by 33 to 70% indicating that injected CO{sub 2} is now going into new zones. This reduction has lasted 1+ year in Field A. Oil production increased and CO{sub 2} production decreased in 5 Field A production wells, offsets to Well #1 injector, for a total of about 2,250 m{sup 3} (600,000 gallons/ 14,250 bbls) of incremental oil production- a $140 / SPI bbl return. Treated marginal production well, Field A Well #2, immediately began showing increased oil production totaling 238 m{sup 3} (63,000 gallons/ 1500 BBLs) over 1 year and an immediate 81% reduced gas-oil ratio.

01 COAL, LIGNITE, AND PEAT↗

Low-E Retrofit Demonstration and Educational Program

The objective of this project was to demonstrate the capability of low-emissivity (low-E) storm windows / panels and low-E retrofit glazing systems to significantly and cost effectively improve the energy efficiency of both existing residential and commercial buildings. The key outcomes are listed below: RESIDENTIAL CASE STUDIES: (a) A residential case study in two large multifamily apartment buildings in Philadelphia showed a substantial 18-22% reduction in heating energy use and a 9% reduction in cooling energy use by replacing old clear glass storm windows with modern low-E storm windows. Furthermore, the new low-E storm windows reduced the overall apartment air leakage by an average of 10%. (b) Air leakage testing on interior low-E panels installed in a New York City multifamily building over windows with and without AC units showed that the effective leakage area of the windows was reduced by 77-95%. (c) To study the use of low-E storm windows in a warmer mixed climate with a balance of both heating and cooling, 10 older homes near Atlanta with single pane windows were tested with three types of exterior storm windows: clear glass, low-E glass with high solar heat gain, and low-E glass with lower solar heat gain. The storm windows significantly reduced the overall home air leakage by an average of 17%, or 3.7 ACH50. Considerably high variability in the data made it difficult to draw strong conclusions about the overall energy usage, but for heating periods, the low-E storm windows showed approximately 15% heating energy savings, whereas clear storm windows were neutral in performance. For cooling periods, the low-E storm windows showed a wide range of performance from 2% to over 30% cooling energy savings. Overall, the study showed the potential for significantly more energy savings from using low-E glass versus no storm window or clear glass storm windows in warmer mixed climates, but it is difficult to conclusively say whether one type of low-E performed better than the other. COMMERCIAL CASE STUDIES: (a) A 12-story office building in Philadelphia was retrofitted by adding a double-pane low-E insulating glass unit to the existing single pane windows, to create a triple glazed low-E system. A detailed side-by-side comparison in two pairs of perimeter offices facing north and east showed a 39-60% reduction in heating energy use, a 9-36% reduction in cooling energy use, and a 10% reduction in peak electrical cooling demand. An analysis of utility bills estimated the whole building heating and cooling energy use was reduced by over 25%. Additionally, the retrofit window temperatures were commonly 20 degrees warmer on winter days, and 10-20 degrees cooler on summer days, leading to increased occupant comfort. (b) Two large 4-story office buildings in New Jersey were retrofitted with a similar system, but using two low-E coatings in the retrofit system. The energy savings are being monitored by a separate GPIC project; this work quantified the changes in glass surface temperatures, thermal comfort, and potential glass thermal stress. The low-E retrofit panels greatly reduced daily variations in the interior window surface temperatures, lowering the maximum temperature and raising the minimum temperature by over 20F compared to the original single pane windows with window film. The number of hours of potential thermal discomfort, as measured by deviation between mean radiant temperature and ambient air temperature by more than 3F, were reduced by 93 percent on the south orientation and over two-thirds on the west orientation. Overall, the low-E retrofit led to substantially improved occupant comfort with less periods of both overheating and feeling cold. (c) No significant thermal stress was observed in the New Jersey office building test window when using the low-E retrofit system over a variety of weather conditions. The surface temperature difference only exceeded 10F (500 psi thermal stress) for less than 1.5% of the monitored time, and in all cases, the maximum surface temperature difference never exceeded 35F (1,750 psi thermal stress). LOW-E STORM WINDOW OUTREACH AND EDUCATION PROGRAM: (a) The project team assisted the State of Pennsylvania in adding low-E storm windows as a cost effective weatherization measure on its priority list for the state weatherization assistance program. (b) No technical barriers that could hinder widespread application were identified in the case studies. However, educational barriers have been identified, in that weatherization personnel commonly misunderstand how the application of low-E storm windows is very different than much more expensive full window replacement. (c) A package of educational materials was developed to help communicate the benefits of low-E storm windows and retrofits as a cost effective tool for weatherization personnel. (d) Using detailed thermal simulations, more accurate U-factor and solar heat gain coefficient (SHGC) values were determined for low-E storm windows installed over different primary windows. IN SUMMARY, this work confirmed the potential for low-E storm windows, panels, and retrofit systems to provide significant energy savings, reductions in air leakage, and improvements in thermal comfort in both residential and commercial existing buildings.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

An Asynchronous Many-Task Implementation of In-Situ Statistical Analysis using Legion.

In this report, we propose a framework for the design and implementation of in-situ analy- ses using an asynchronous many-task (AMT) model, using the Legion programming model together with the MiniAero mini-application as a surrogate for full-scale parallel scientific computing applications. The bulk of this work consists of converting the Learn/Derive/Assess model which we had initially developed for parallel statistical analysis using MPI [PTBM11], from a SPMD to an AMT model. In this goal, we propose an original use of the concept of Legion logical regions as a replacement for the parallel communication schemes used for the only operation of the statistics engines that require explicit communication. We then evaluate this proposed scheme in a shared memory environment, using the Legion port of MiniAero as a proxy for a full-scale scientific application, as a means to provide input data sets of variable size for the in-situ statistical analyses in an AMT context. We demonstrate in particular that the approach has merit, and warrants further investigation, in collaboration with ongoing efforts to improve the overall parallel performance of the Legion system.

Pebay, Philippe Pierre↗

Reproducibility and Variability of I/O Performance on BG/Q: Lessons Learned from a Data Aggregation Algorithm

Reading and writing data efficiently from different tiers of storage is necessary for most scientific simulations to achieve good performance at scale. Many software solutions have been developed to decrease the I/O bottleneck. One wellknown strategy, in the context of collective I/O operations, is the two-phase I/O scheme. This strategy consists of selecting a subset of processes to aggregate contiguous pieces of data before performing reads/writes. In our previous work, we implemented the two-phase I/O scheme with a MPI-based topology-aware algorithm. Our algorithm showed very good performance at scale compared to the standard I/O libraries such as POSIX I/O and MPI I/O. However, the algorithm had several limitations hindering a satisfying reproducibility of our experiments. In this paper, we extend our work by 1) identifying the obstacles we face to reproduce our experiments and 2) discovering solutions that reduce the unpredictability of our results.

97 MATHEMATICS AND COMPUTING↗