Layer time control for large scale additive manufacturing using high performance computing
Explore the source record for details and available documents.
SEARCH · Engineering Papers
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.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Air-conditioning (cooling) for buildings is the single largest use of electricity in the United States (U.S.). This drives summer peak electric demand in much of the U.S. Improved air-conditioning technology thus has the greatest potential impact on the electric grid compared to other technologies that use electricity. Thermally-activated technologies (TAT), such as natural gas engine-driven heat pumps (GHP), can provide overall peak load reduction and electric grid relief for summer peak demand. GHP offers an attractive opportunity for commercial building owners to reduce electric demand charges and operating expenses. Engine-driven systems have several potential advantages over conventional single-speed or single-capacity electric motor-driven units. Among them are variable speed operation, high part load efficiency, high temperature waste heat recovery from the engine, and reduced annual operating costs (SCGC 1998). Although gas engine-driven systems have been in use since the 1960s, current research is resulting in better performance, lower maintenance requirements, and longer operating lifetimes. Gas engine-driven systems are typically more expensive to purchase than comparable electric motor-driven systems, but they typically cost less to operate, especially for commercial building applications. Operating cost savings for commercial applications are primarily driven by electric demand charges. GHP operating costs are dominated by fuel costs, but also include maintenance costs. The reliability of gas cooling equipment has improved in the last few years and maintenance requirements have decreased (SCGC 1998, Yahagi et al. 2006). Another advantage of the GHP over electric motor-driven is the ability to use the heat rejected from the engine during heating operation. The recovered heat can be used to supplement the vapor compression cycle during heating or to supply other process loads, such as water heating. The use of the engine waste heat results in greater operating efficiency compared to conventional electric motor-driven units (SCGC 1998). In Japan, many hundreds of thousands of natural gas-driven heat pumps have been sold (typically 40,000 systems annually) (Yahagi et al. 2006). The goal of this program is to develop dependable and energy efficient GHPs suitable for U.S. commercial rooftop applications (the single largest commercial product segment). This study describes the laboratory performance evaluation of an integrated 10-ton GHP rooftop unit (a 900cc Daihatsu-Aisin natural gas engine) which uses R410A as the refrigerant (GEDAC No.23). ORNL Thermally-Activated Heat Pump (TAHP) Environmental Chambers were used to evaluate this unit in a controlled laboratory environment.
In a multiyear research agreement with Tenix Investments Pty. Ltd., Sandia has been developing field deployable technologies for detection of biotoxins in water supply systems. The unattended water sensor or UWS employs microfluidic chip based gel electrophoresis for monitoring biological analytes in a small integrated sensor platform. This instrument collects, prepares, and analyzes water samples in an automated manner. Sample analysis is done using the {mu}ChemLab{trademark} analysis module. This report uses analysis results of two datasets collected using the UWS to estimate performance of the device. The first dataset is made up of samples containing ricin at varying concentrations and is used for assessing instrument response and detection probability. The second dataset is comprised of analyses of water samples collected at a water utility which are used to assess the false positive probability. The analyses of the two sets are used to estimate the Receiver Operating Characteristic or ROC curves for the device at one set of operational and detection algorithm parameters. For these parameters and based on a statistical estimate, the ricin probability of detection is about 0.9 at a concentration of 5 nM for a false positive probability of 1 x 10{sup -6}.
The modeling of solids is most naturally placed within a Lagrangian framework because it requires constitutive models which depend on knowledge of the original material orientations and subsequent deformations. Detailed kinematic information is needed to ensure material frame indifference which is captured through the deformation gradient F. Such information can be tracked easily in a Lagrangian code. Unfortunately, not all problems can be easily modeled using Lagrangian concepts due to severe distortions in the underlying motion. Either a Lagrangian/Eulerian or a pure Eulerian modeling framework must be introduced. We discuss and contrast several Lagrangian/Eulerian approaches for keeping track of the details of material kinematics.
The research documented herein was funded by a research contract between the Research Reactors Division (RRD) of Oak Ridge National Laboratory (ORNL) and the University of Tennessee, Knoxville (UTK) Mechanical, Aerospace and Biomedical Engineering Department (MABE). The research was governed by a statement of work (SOW) which clearly defines nine specific tasks. This report is outlined to follow and document the results of each of these nine specific tasks. The primary goal of this phase of the research is to demonstrate, through verification and validation methods, that COMSOL is a viable simulation tool for thermal-hydraulic modeling of the High Flux Isotope Reactor (HFIR) core. A secondary goal of this two-dimensional phase of the research is to establish methodology and data base libraries that are also needed in the full three-dimensional COMSOL simulation to follow. COMSOL version 3.5a was used for all of the models presented throughout this report.
Prior to the COVID-19 pandemic, US hospitals relied on static projections of future trends for long-term planning and were only beginning to consider forecasting methods for short-term planning of staffing and other resources. With the overwhelming burden imposed by COVID-19 on the health care system, an emergent need exists to accurately forecast hospitalization needs within an actionable timeframe. Our goal was to leverage an existing COVID-19 case and death forecasting tool to generate the expected number of concurrent hospitalizations, occupied intensive care unit (ICU) beds, and in-use ventilators 1 day to 4 weeks in the future for New Mexico and each of its five health regions. We developed a probabilistic model that took as input the number of new COVID-19 cases for New Mexico from Los Alamos National Laboratory’s COVID-19 Forecasts Using Fast Evaluations and Estimation tool, and we used the model to estimate the number of new daily hospital admissions 4 weeks into the future based on current statewide hospitalization rates. The model estimated the number of new admissions that would require an ICU bed or use of a ventilator and then projected the individual lengths of hospital stays based on the resource need. By tracking the lengths of stay through time, we captured the projected simultaneous need for inpatient beds, ICU beds, and ventilators. We used a postprocessing method to adjust the forecasts based on the differences between prior forecasts and the subsequent observed data. Thus, we ensured that our forecasts could reflect a dynamically changing situation on the ground. Forecasts made between September 1 and December 9, 2020, showed variable accuracy across time, health care resource needs, and forecast horizon. Forecasts made in October, when new COVID-19 cases were steadily increasing, had an average accuracy error of 20.0%, while the error in forecasts made in September, a month with low COVID-19 activity, was 39.7%. Across health care use categories, state-level forecasts were more accurate than those at the regional level. Although the accuracy declined as the forecast was projected further into the future, the stated uncertainty of the prediction improved. Forecasts were within 5% of their stated uncertainty at the 50% and 90% prediction intervals at the 3- to 4-week forecast horizon for state-level inpatient and ICU needs. However, uncertainty intervals were too narrow for forecasts of state-level ventilator need and all regional health care resource needs. Real-time forecasting of the burden imposed by a spreading infectious disease is a crucial component of decision support during a public health emergency. Our proposed methodology demonstrated utility in providing near-term forecasts, particularly at the state level. This tool can aid other stakeholders as they face COVID-19 population impacts now and in the future.
The MAJORANA DEMONSTRATOR is an experiment constructed to search for Neutrinoless double-beta decay in 76Ge and to demonstrate the feasibility to deploy a large-scale experiment in a phased and modular fashion. It consists of two modular arrays of natural and 76Ge-enriched germanium detectors totalling 44.1 kg, operating on the 4850’ level of the Sanford Underground Research Facility in Lead, South Dakota, USA. Data acquisition in the commissioning phase began in June 2015, followed by data acquisition with the full detector array in August 2016. The ultralow background and record energy resolution achieved by the MAJORANA DEMONSTRATOR enable a sensitive neutrinoless double-beta decay search, as well as additional searches for physics beyond the Standard Model. I will discuss the design elements that enable these searches, along with our latest results, focusing on the neutrinoless double-beta decay search. I will also discuss the current status and the future plans of the MAJORANA DEMONSTRATOR, as well as the plans for a future tonne-scale 76Ge experiment.
We report recent progress on the on-going NuSTAR observation campaign of 8 TeV-detected middle-aged pulsar wind nebulae (PWNe). This campaign constitutes a major part of our NuSTAR study of some of the most energetic TeV sources in our Galaxy detected by VERITAS and HAWC. NuSTAR is the first focusing X-ray telescope operating above 10 keV in space with sub-arcminute angular resolution. Broad-band X-ray imaging and spectroscopy data, obtained by NuSTAR, allow us to probe sub-PeV electron populations through detecting synchrotron X-ray radiation. Our targets include PeVatron candidates detected by HAWC, the Boomerang nebula, PWNe crushed by supernova remnant shocks (or relic PWNe) and G0.9‚0.1 in the Galactic Center. Combined with Fermi-LAT data and available TeV data, we aim to provide a complete, multi-wavelength view of a diverse class of middle-aged (~10–100 kyrs old) PWNe. Our NuSTAR analysis detected hard X-ray emission from the Eel and Boomerang PWNe and characterized their broad-band X-ray spectra most accurately. We plan to apply both time-evolution and multi-zone PWNe models to multi-wavelength spectral energy distribution (SED) data over the radio, X-ray, GeV and TeV bands. In this proceeding, we will review our observation campaign and discuss the preliminary results for some PWNe.
Recent work shows that quantum signal processing (QSP) and its multi-qubit lifted version, quantum singular value transformation (QSVT), unify and improve the presentation of most quantum algorithms. QSP/QSVT characterize the ability, by alternating ansätze, to obliviously transform the singular values of subsystems of unitary matrices by polynomial functions; these algorithms are numerically stable and analytically well-understood. That said, QSP/QSVT require consistent access to a single oracle, saying nothing about computing joint properties of two or more oracles; these can be far cheaper to determine given an ability to pit oracles against one another coherently. This work introduces a corresponding theory of QSP over multiple variables: M-QSP. Surprisingly, despite the non-existence of the fundamental theorem of algebra for multivariable polynomials, there exist necessary and sufficient conditions under which a desired stable multivariable polynomial transformation is possible. Moreover, the classical subroutines used by QSP protocols survive in the multivariable setting for non-obvious reasons, and remain numerically stable and efficient. Up to a well-defined conjecture, we give proof that the family of achievable multivariable transforms is as loosely constrained as could be expected. The unique ability of M-QSP to obliviously approximate joint functions of multiple variables coherently leads to novel speedups incommensurate with those of other quantum algorithms, and provides a bridge from quantum algorithms to algebraic geometry.
Quantum systems of infinite dimension, such as bosonic oscillators, provide vast resources for quantum sensing. Yet, a general theory on how to manipulate such bosonic modes for sensing beyond parameter estimation is unknown. We present a general algorithmic framework, quantum signal processing interferometry (QSPI), for quantum sensing at the fundamental limits of quantum mechanics by generalizing Ramsey-type interferometry. Our QSPI sensing protocol relies on performing nonlinear polynomial transformations on the oscillator's quadrature operators by generalizing quantum signal processing (QSP) from qubits to hybrid qubit-oscillator systems. We use our QSPI sensing framework to make efficient binary decisions on a displacement channel in the single-shot limit. Theoretical analysis suggests the sensing accuracy, given a single-shot qubit measurement, scales inversely with the sensing time or circuit depth of the algorithm. We further concatenate a series of such binary decisions to perform parameter estimation in a bit-by-bit fashion. Numerical simulations are performed to support these statements. Our QSPI protocol offers a unified framework for quantum sensing using continuous-variable bosonic systems beyond parameter estimation and establishes a promising avenue toward efficient and scalable quantum control and quantum sensing schemes beyond the NISQ era.
We introduce a physics guided data-driven method for image-based multi-material decomposition for dual-energy computed tomography (CT) scans. The method is demonstrated for CT scans of virtual human phantoms containing more than two types of tissues. The method is a physics-driven supervised learning technique. We take advantage of the mass attenuation coefficient of dense materials compared to that of muscle tissues to perform a preliminary extraction of the dense material from the images using unsupervised methods. We then perform supervised deep learning on the images processed by the extracted dense material to obtain the final multi-material tissue map. The method is demonstrated on simulated breast models with calcifications as the dense material placed amongst the muscle tissues. The physics-guided machine learning method accurately decomposes the various tissues from input images, achieving a normalized root-mean-squared error of 2.75%.
We introduce an efficient distributed sequence parallel approach for training transformer-based deep learning image segmentation models. The neural network models are comprised of a combination of a Vision Transformer encoder with a convolutional decoder to provide image segmentation mappings. The utility of the distributed sequence parallel approach is especially useful in cases where the tokenized embedding representation of image data are too large to fit into standard computing hardware memory. To demonstrate the performance and characteristics of our models trained in sequence parallel fashion compared to standard models, we evaluate our approach using a 3D MRI brain tumor segmentation dataset. We show that training with a sequence parallel approach can match standard sequential model training in terms of convergence. Furthermore, we show that our sequence parallel approach has the capability to support training of models that would not be possible on standard computing resources.
Recent advancements in AI models, like ChatGPT, are structured to retain user interactions, which could inadvertently include sensitive healthcare data. In the healthcare field, particularly when radiologists use AI-driven diagnostic tools hosted on online platforms, there is a risk that medical imaging data may be repurposed for future AI training without explicit consent, spotlighting critical privacy and intellectual property concerns around healthcare data usage. Addressing these privacy challenges, a novel approach known as Unlearnable Examples (UEs) has been introduced, aiming to make data unlearnable to deep learning models. A prominent method within this area, called Unlearnable Clustering (UC), has shown improved UE performance with larger batch sizes but was previously limited by computational resources (e.g., a single workstation). To push the boundaries of UE performance with theoretically unlimited resources, we scaled up UC learning across various datasets using Distributed Data Parallel (DDP) training on the Summit supercomputer. Our goal was to examine UE efficacy at high-performance computing (HPC) levels to prevent unauthorized learning and enhance data security, particularly exploring the impact of batch size on UE’s unlearnability. Utilizing the robust computational capabilities of the Summit, extensive experiments were conducted on diverse datasets such as Pets, MedMNist, Flowers, and Flowers102. Our findings reveal that both overly large and overly small batch sizes can lead to performance instability and affect accuracy. However, the relationship between batch size and unlearnability varied across datasets, highlighting the necessity for tailored batch size strategies to achieve optimal data protection. The use of Summit’s high-performance GPUs, along with the efficiency of the DDP framework, facilitated rapid updates of model parameters and consistent training across nodes. Our results underscore the critical role of selecting appropriate batch sizes based on the specific characteristics of each dataset to prevent learning and ensure data security in deep learning applications. The source code is publicly available at https: // github. com/ hrlblab/ UE_ HPC .
Large language models (LLMs) have demonstrated remarkable success as foundational models, benefiting various downstream applications through fine-tuning. Loss scaling studies have demonstrated the superior performance of larger LLMs compared to their smaller counterparts. Nevertheless, training LLMs with billions of parameters poses significant challenges and requires considerable computational resources. For example, training a one trillion parameter GPT-style model on 20 trillion tokens requires a staggering 120 million exaflops. This research explores efficient distributed training strategies to extract this computation from Frontier, the world's first exascale supercomputer. We enable and investigate various model and data parallel training techniques, such as tensor parallelism, pipeline parallelism, and sharded data parallelism, to facilitate training a trillion-parameter model on Frontier. We empirically assess these techniques and their associated parameters to determine their impact on memory footprint, communication latency, and GPU's computational efficiency. We analyze the complex interplay among these techniques and find a strategy to combine them to achieve high throughput through hyperparameter tuning. We have identified efficient strategies for training large LLMs of varying sizes through empirical analysis and hyperparameter tuning. For 22 Billion, 175 Billion, and 1 Trillion parameters, we achieved GPU throughputs of 38.38%, 36.14%, and 31.96%, respectively. For the training of the 175 Billion parameter model and the 1 Trillion parameter model, we achieved 100% weak scaling efficiency on 1024 and 3072 Mi250X GPUs, respectively. We also achieved strong scaling efficiencies of 89% and 87% for these two models. We trained these models only tens of iterations instead of training till completion.
Explore the source record for details and available documents.
Explore the source record for details and available documents.