Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “PERFORMANCE PREDICTION”

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 955 records · Page 53

Propulsion-induced interference effects on jet-lift VTOL aircraft

Recent investigations of propulsion-induced aerodynamics are reviewed. Experimental and analytical studies of the fundamental properties of lifting jet wakes and wind-tunnel investigations of VTOL configuration variables are discussed. The jet-wake flow studies will provide the data base for evaluating theoretical performance prediction methods; while the configuration studies will provide guidance for optimizing airframe/propulsion system integration. The influence of results to date on high-performance VTOL designs emphasizing minimum jet-induced losses and improved thrust vectoring in forward flight (VIFF) capability is illustrated.

Winston, M. M.↗

Adaptive proximate time-optimal servomechanisms - Continuous time case

A Proximate Time-Optimal Servo (PTOS) is developed, along with conditions for its stability. An algorithm is proposed for adapting the PTOS (APTOS) to improve performance in the face of uncertain plant parameters. Under ideal conditions APTOS is shown to be uniformly asymptotically stable. Simulation results demonstrate the predicted performance.

Workman, M. L.↗

Thermal Analysis of an In-Space Heat Shield Exposed to a Rocket Plume During Stage Separation

Thermal analyses were conducted to evaluate the thermal response of various individual materials and multi-layer configurations for an in-space heat shield exposed to a rocket plume during stage separation. Frequently used, readily available thermal protection materials consisting of cork, carbon cloth phenolic (CCP), and silica cloth phenolic (SCP) were selected for an initial screening analysis. SCP was chosen as the primary material for further evaluation using an updated thermal environment calculated using computational fluid dynamics (CFD) results. A multi-layer configuration consisting of SCP providing erosion resistance and a Nomex honeycomb providing thermal protection was then evaluated as a solution for more mass-efficient performance. Pyrolysis depth, bondline temperature, back-side temperature, and a thermal factor of safety were used to define the necessary material thickness profile. Adhesives were included in the analyses with considerations for maximum temperature and application method. One dimensional thermal analyses at several stations along the heat shield surface were performed for the initial screening and follow-on analyses. The methodology used standard processes and computer programs applicable to solid rocket motor internal insulation and nozzle thermal analysis such as Chemical Equilibrium and Applications (CEA), Aerotherm Chemical Equilibrium (ACE), Momentum/Energy Integral Technique (MEIT), and Insulation Thermal Response and Ablation Code (ITRAC). The baseline SCP/Nomex honeycomb multi-layer configuration thermal performance predictions satisfied the objectives with an acceptable mass estimate for the design maturity. The EA9673 film adhesive and SCP resin were the best performing bonding agents considered for the SCP/Nomex honeycomb and Nomex honeycomb/structural substrate interfaces. The baseline mass estimate was lowest when the resin was the bonding agent due to the higher maximum service temperature. These thermal analyses evaluated typical internal and external nozzle materials using a calculated thermal environment and heat loads that were lower than a typical rocket motor internal environment but greater than standard external aeroheating.

Andrew T Hiatt↗

Thermal-hydraulic performance of the TREAT Multi-SERTTA for reactivity initiated accident experiments

Experimental testing in the Multi-Static Environment Rodlet Transient Test Apparatus (SERTTA) will lead the rebirth of transient fuel testing in the United States as part of the Accident Tolerant Fuels (ATF) program. The Multi-SERTTA is comprised of 4 isolated pressurized environments capable of a wide variety of working fluids and thermal conditions. Ultimately, the TREAT reactor as well as the Multi-SERTTA test vehicle serve the purpose of providing desired thermal-hydraulic boundary conditions to the test specimen. The initial ATF testing in TREAT will focus on reactivity insertion accident (RIA) events using both gas and water environments including typical PWR operating pressures and temperatures. The heat transfer conditions during the high energy power pulses of RIA events remains a subject of large uncertainty and great importance for fuel performance predictions. To support transient experiments, the Multi-SERTTA vehicle has been modeled using RELAP5 with a baseline test specimen composed of fresh UO{sub 2} fuel in zircaloy cladding. The results show that for energy depositions less than 1380 J/g-UO{sub 2}, the vessel environment remains sub-supercritical. Two extremes of gap conductance show significant effect on event timing and peak cladding temperature but little effect on end state conditions. Inclusion of oxidation has increasing impact for increasing energy depositions. The effect of oxidation is non-negligible for planned TREAT experiments in the range of explored energy depositions. Pulse width has little impact on thermal hydraulic conditions under the assumptions used in this analysis. The model used in this work has been coupled to the fuel performance code, BISON, to accomplish best estimate predictive capability for Multi-SERTTA experiments. The results show the great flexibility and capacity of the TREAT Multi-SERTTA test vehicle to provide testing under a wide range of prototypic thermal-hydraulic conditions as never done before.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Scalable Prediction of Energy Consumption using Incremental Time Series Clustering

Time series datasets are a canonical form of high velocity Big Data, and often generated by pervasive sensors, such as found in smart infrastructure. Performing predictive analytics on time series data can be computationally complex, and requires approximation techniques. In this paper, we motivate this problem using a real application from the smart grid domain. We propose an incremental clustering technique, along with a novel affinity score for determining cluster similarity, which help reduce the prediction error for cumulative time series within a cluster. We evaluate this technique, along with optimizations, using real datasets from smart meters, totaling ~700,000 data points, and show the efficacy of our techniques in improving the prediction error of time series data within polynomial time.

Simmhan, Yogesh↗

Thrust Augmentation with Mixer/Ejector Systems

Older commercial aircraft often exceed FAA (Federal Aviation Administration) sideline noise regulations. The major problem is the jet noise associated with the high exhaust velocities of the low bypass ratio engines on such aircraft. Mixer/ejector exhaust systems can provide a simple means of reducing the jet noise on these aircraft by mixing cool ambient air with the high velocity engine gases before they are exhausted to ambient. This paper presents new information on thrust performance predictions, and thrust augmentation capabilities of mixer/ejectors. Results are presented from the recent development program of the patented Alternating Lobe Mixer Ejector Concept (ALMEC) suppressor system for the Gulfstream GII, GIIB and GIII aircraft. Mixer/ejector performance procedures are presented which include classical control volume analyses, compound compressible flow theory, lobed nozzle loss correlations and state of the art computational fluid dynamic predictions. The mixer/ejector thrust predictions are compared to subscale wind tunnel test model data and actual aircraft flight test measurements. The results demonstrate that a properly designed mixer/ejector noise suppressor can increase effective engine bypass ratio and generate large thrust gains at takeoff conditions with little or no thrust loss at cruise conditions. The cruise performance obtained for such noise suppressor systems is shown to be a strong function of installation effects on the aircraft.

Presz, Walter M., Jr.↗

Prediction of DIII-D Pedestal Structure From Externally Controllable Parameters

The sharp increase of pressure at the edge of a high confinement mode (H-mode) plasma, the pedestal, strongly impacts overall plasma performance. Predicting the pedestal is a necessity to control and optimize tokamak operations. Here, an experimental data-driven machine learning (ML) approach is presented that predicts the pedestal heights and widths of electron density (n e ) and electron temperature (T e ) profiles as well as the separatrix ne from externally controllable parameters such as the plasma shape, heating method and power, and gas puff rate and integrated gas puff. The OMFIT framework was used with DIII-D data to efficiently, robustly, and automatically build a database of pedestal parameters to train machine learning models. Database creation was enabled by the search engine tool for DIII-D data, TokSearch, which parallelizes data fetching, enabling fast searches through basic signals of thousands of DIII-D shots and selection of relevant time intervals. Principal Component Analysis (PCA) separated the database into three clusters that represent classes of plasma shapes that are regularly used in DIII-D. The most important parameters for setting the pedestal structure were plasma current (I p ), toroidal magnetic field (B Φ ), neutral beam heating power (P NBI ) and shaping quantities. The Deep Jointly Informed Neural Networks (DJINN) algorithm was applied to identify suitable neural network (NN) architectures that appropriately capture the features of the pedestal database. Separate NNs were implemented for each pedestal parameter, and ensembling methods were used to improve the prediction accuracy and allowed estimation of the prediction uncertainty. The pedestal predictions of the test dataset lie within the measurement uncertainties of the pedestal parameters. The NN outperformed simple Linear Regression (LR) analysis, indicating non-linear dependencies in the pedestal structure. The presented achievements illustrate a promising path for future research, using feature extraction to infer experimental trends and thereby improve pedestal models as well as deploying NN for a fast pedestal prediction in DIII-D scenario development.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Prediction of DIII-D Pedestal Structure from Externally Controllable Parameters

The sharp increase of pressure at the edge of a high confinement mode (H-mode) plasma, the pedestal, strongly impacts overall plasma performance. Predicting the pedestal is a necessity to control and optimize tokamak operations. An experimental data-driven machine learning (ML) approach is presented that predicts the pedestal heights and widths of electron density (ne) and electron temperature (Te) profiles as well as the separatrix ne from externally controllable parameters such as the plasma shape, heating method and power, and gas puff rate and integrated gas puff. The OMFIT framework was used with DIII-D data to efficiently, robustly, and automatically build a database of pedestal parameters to train machine learning models. Database creation was enabled by the search engine tool for DIII-D data, TokSearch, which parallelizes data fetching, enabling fast searches through basic signals of thousands of DIII-D shots and selection of relevant time intervals. Principal Component Analysis (PCA) separated the database into three clusters that represent classes of plasma shapes that are regularly used in DIII-D. The most important parameters for setting the pedestal structure were plasma current (Ip), toroidal magnetic field (Bφ), neutral beam heating power (PNBI) and shaping quantities. The Deep Jointly Informed Neural Networks (DJINN) algorithm was applied to identify suitable neural network (NN) architectures that appropriately capture the features of the pedestal database. Separate NNs were implemented for each pedestal parameter, and ensembling methods were used to improve the prediction accuracy and allowed estimation of the prediction uncertainty. The pedestal predictions of the test dataset lie within the measurement uncertainties of the pedestal parameters. The NN outperformed simple Linear Regression (LR) analysis, indicating non-linear dependencies in the pedestal structure. The presented achievements illustrate a promising path for future research, using feature extraction to infer experimental trends and thereby improve pedestal models as well as deploying NN for a fast pedestal prediction in DIII-D scenario development.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Physics vs structure: A systematic benchmark of learning strategies for multi-zone building thermal dynamics

Recent advances in physics-informed and data-driven machine learning promise improved thermal models for advanced building control, yet there is limited quantitative evidence on when added physics structure and architectural complexity are beneficial. Here, this work presents a systematic benchmark of five representative system identification methods for modeling multi-zone building thermal dynamics: linear state-space models, multi-layer perceptrons, neural state-space models, neural ordinary differential equations, and physically-consistent neural networks. The methods are evaluated across multiple data regimes and zone coupling strategies. Using a high-fidelity multi-zone commercial building emulator, we examine short-term and long-term prediction accuracy, computational efficiency, and ease of development. Our results reveal critical trade-offs between prediction performance, model complexity, and physical consistency. We demonstrate that decoupled, nonlinear black-box models consistently outperform coupled physics-constrained architectures in both predictive accuracy and out-of-distribution robustness in majority of the test cases for the building type considered in the study. Our findings quantify the cost of complexity in building thermal modeling and provide concrete, actionable, scenario-based guidelines for selecting model classes for control-oriented applications.

Building thermal modeling↗

A scalable algorithm for the optimization of neural network architectures

In this work, we propose a new scalable method to optimize the architecture of an artificial neural network. The proposed algorithm, called Greedy Search for Neural Network Architecture, aims to determine a neural network with minimal number of layers that is at least as performant as neural networks of the same structure identified by other hyperparameter search algorithms in terms of accuracy and computational cost. Numerical results performed on benchmark datasets show that, for these datasets, our method outperforms state-of-the-art hyperparameter optimization algorithms in terms of attainable predictive performance by the selected neural network architecture, and time-to-solution for the hyperparameter optimization to complete.

97 MATHEMATICS AND COMPUTING↗

Closed Crawl Space Performance: Proof of Concept in the Production Builder Marketplace

This overview is intended to be a very concise, limited summary of the key project activities discussed in the detailed report that follows. Due to the large scope of this project, the detailed report is broken into three individually titled sections. Each section repeats key background information, with the goal that the sections will eventually stand alone as complete reports on the major activities of the project. The information presented herein comes from ongoing research, so please note that all observations, findings and recommendations presented are preliminary and subject to change in the future. We invite and welcome your comments and suggestions for improving the project. Advanced Energy completed its first jointly-funded crawl space research project with the Department of Energy in 2005. That project, funded under award number DE-FC26-00NT40995 and titled 'A Field Study Comparison of the Energy and Moisture Performance Characteristics of Ventilated Versus Sealed Crawl Spaces in the South' demonstrated the substantial energy efficiency and moisture management benefits that result from using properly closed crawl space foundations for residential construction instead of traditional wall vented crawl space foundations. Two activities of this first project included (1) an assessment of ten existing homes to document commonly observed energy and moisture failures associated with wall-vented crawl space foundations and (2) a detailed literature review that documented both the history of closed crawl space research and the historical lack of scientific justification for building code requirements for crawl space ventilation. The most valuable activity of the 2005 project proved to be the field demonstration of various closed crawl space techniques, which were implemented in a set of twelve small (1040 square feet), simply designed homes in eastern North Carolina. These homes had matched envelope, mechanical and architectural designs, and comparable performance characteristics with regard to infiltration and duct leakage. Researchers settled on two closed crawl space designs, one with insulation located in the framed floor structure above the crawl space and one with insulation on the crawl space perimeter wall, as the designs with the most widespread potential for application. Researchers based this assessment not only on the field performance, but also on input from residential builders, pest control professionals, code officials, installers, and building scientists active in the region. The key findings from the field demonstration were that (1) closed crawl spaces stay substantially drier than traditional wall-vented crawl spaces during humid climate conditions, and (2) the houses built on the closed crawl space foundations saved, on average, 15% or more on annual energy used for space heating and cooling. A comparison of the actual energy performance of the homes versus the performance predicted by a popular HERS software application showed that the software was unable to predict the demonstrated savings, in some cases predicting an energy penalty. Findings from the 2005 project were summarized in a publication titled Closed Crawl Spaces: An Introduction to Design, Construction and Performance. Since its release, the publication has received widespread use by builders, homeowners, installers and code officials concerned about crawl space construction. The findings were also used to create major revisions to the NC Residential Code, which were adopted in 2004 and immediately began to reduce the regulatory barriers to widespread commercialization of the technology in NC, particularly in new residential construction. Full project details are located at www.crawlspaces.org.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Performance evaluation of Space Shuttle SRB parachutes from air drop and scaled model wind tunnel tests

The aerodynamic performance characteristics have been determined for the Space Shuttle Solid Rocket Booster drogue, main, and pilot parachutes. The performance evaluation on the 20-degree conical ribbon parachutes is based primarily on air drop tests of full scale prototype parachutes. In addition, parametric wind tunnel tests were performed and used in parachute configuration development and preliminary performance assessments. The wind tunnel test data are compared to the drop test results and both sets of data are used to determine the predicted performance of the Solid Rocket Booster flight parachutes. Data from other drop tests of large ribbon parachutes are also compared with the Solid Rocket Booster parachute performance characteristics. Parameters assessed include full open terminal drag coefficients, reefed drag area, opening characteristics, clustering effects, and forebody interference.

Moog, R. D.↗

Analysis of turbulent boundary layer on cascade and rotor blades of turbomachinery

The momentum integral technique for predicting the boundary layer growth in three-dimensional flow has been extended to include the entrainment equation as the closure model. The numerical solution is compared with the cascade, inducer, compressor, and fan rotor blade data from various sources. The agreement is found to be excellent in all cases, with the exception of the separated flow. Both the momentum thickness and the limiting streamline angle predicted from this analysis compare well with the measured data for a rotor blade. The technique is extremely useful in engineering design, analysis, and performance prediction.

Lakshminarayana, B.↗

System structural test results - Six PODS III supports

Passive orbital disconnect struts can potentially reduce the support conductance a factor of 10 over state-of-the-art tension band nondisconnect supports and cut helium dewar weights in half for the same lifetime. A series of thermal and structural tests were performed to verify that this performance improvement is real. Structural tests on a PODS-III support system (consisting of six struts) is reported here. The results show the predicted performance improvements can be achieved and the PODS-III supports are ready for flight applications. For large tankage systems requiring higher side load capability, a PODS-IV version is currently being developed.

Parmley, Richard T.↗

A Recursive Multi-step Machine Learning Approach for Airport Configuration Prediction

Airport configuration selection is a complex decision-making process that involves several operational and human factors. In this paper we propose a novel recursive multi-step machine learning (ML) approach to predict airport configuration. The multi-step approach guarantees stability of the predicted configuration by taking as input the configuration predicted at the previous time step. The features of the proposed model include weather data, future arrival and departure counts and current configuration. Due to the importance of arrival and departure counts in predicting the airport configuration, arrival counts are calculated using landing time predictions selected from physics-based landing time predictions available in FAA System Wide Information Management data feeds for each flight. The selection rules were developed and refined to select the most accurate time for different phases of flight. The proposed model predicts the airport configurations up to 6 hours ahead. In this paper we show the predictive performance of the proposed model for six major US airports, including Charlotte Douglas International Airport (CLT), Dallas/Fort Worth International Airport (DFW), John F. Kennedy International Airport (JFK), Newark Liberty International Airport (EWR), LaGuardia Airport (LGA) and Dallas Love Field Airport (DAL). We trained and evaluated models on 2019 and 2020 data in order to study the effect of the pandemic and how changes in traffic patterns affected the performance of the proposed model. Results are compared with a baseline assuming no airport configuration changes. In our results for DFW, we obtained a prediction accuracy of 89.3% for 3 hours ahead prediction, and 82.8% for 6 hours ahead when applied on 2019 data.

machine learning↗

A Recursive Multi-step Machine Learning Approach for Airport Configuration Prediction

Airport configuration selection is a complex decision-making process that involves several operational and human factors. In this paper we propose a novel recursive multi-step machine learning (ML) approach to predict airport configuration. The multi-step approach guarantees stability of the predicted configuration by taking as input the configuration predicted at the previous time step. The features of the proposed model include weather data, future arrival and departure counts and current configuration. Due to the importance of arrival and departure counts in predicting the airport configuration, arrival counts are calculated using landing time predictions selected from physics-based landing time predictions available in FAA System Wide Information Management data feeds for each flight. The selection rules were developed and refined to select the most accurate time for different phases of flight. The proposed model predicts the airport configurations up to 6 hours ahead. In this paper we show the predictive performance of the proposed model for six major US airports, including Charlotte Douglas International Airport (CLT), Dallas/Fort Worth International Airport (DFW), John F. Kennedy International Airport (JFK), Newark Liberty International Airport (EWR), LaGuardia Airport (LGA) and Dallas Love Field Airport (DAL). We trained and evaluated models on 2019 and 2020 data in order to study the effect of the pandemic and how changes in traffic patterns affected the performance of the proposed model. Results are compared with a baseline assuming no airport configuration changes. In our results for DFW, we obtained a prediction accuracy of 89.3% for 3 hours ahead prediction, and 82.8% for 6 hours ahead when applied on 2019 data.

machine learning↗

Design Optimization of A-Type Heat Exchangers based on High Performance, Non-Round Tubes

Air-to-refrigerant heat exchangers (HXs) have been the topic of considerable research as they are the fundamental heat transfer components of HVAC&R systems. For residential and commercial applications, indoor heat exchangers are often in A-type configurations to reduce air-conditioner system footprint while still delivering the required cooling or heating. Classical HX design practices typically assume a uniform frontal air velocity profile to facilitate the use of conventional tube-fin airside heat transfer and pressure drop correlations for performance prediction. However, it is well-reported throughout the literature that A-type HXs experience significant airflow maldistribution which can severely degrade HX performance. Additionally, recent advancements in simulation software such as Computational Fluid Dynamics (CFD), Finite Element Analysis (FEA) and optimization algorithms have led researchers to consider primary heat transfer surface optimization to achieve highly compact HXs which do not require fins. In this paper, an A-type HX which utilizes shape-optimized non-round tubes is optimized for minimum duct size and airside pressure drop. This is achieved through coupling of an experimentally validated finite volume HX model with automated 2D CFD simulations of airflow through the HX for airside thermal-hydraulic performance evaluation. Preliminary results show that the optimal designs deliver similar capacity while achieving 20% reductions in airside pressure drop and duct cross-section area. Additionally, the optimal design apex angles in this study were similar to those from previous A-Type HX optimization studies for conventional tube-fin HX geometries, suggesting that the apex angle for optimal A-Type configuration performance may be independent of tube geometry.

42 ENGINEERING↗

Individual Data Sparsity in Smart Thermostat Big Data: Impacts on Modeling Thermostat Use Behavior Dynamics

This study explores the impacts of the sparsity of individual thermostat interaction data on modeling thermostat use behavior dynamics using a dataset of over 100,000 smart thermostats. In developing a data-driven model of Thermal Frustration Theory (TFT), we investigate the challenges and trade-offs in clustering occupant data to enhance predictive accuracy. Our findings reveal that a single, aggregated model fails to capture the diversity of occupant behaviors, resulting in extremely poor prediction performance. Conversely, excessive clustering exacerbates data sparsity, undermining model reliability. By identifying an optimal clustering strategy, we achieve a balance that significantly improves the prediction of manual setpoint changes during demand response (DR) events, enhancing energy management and occupant comfort

Fannon, David↗