Search NASA⌕ Search

SEARCH · Search NASA

Results for “LINEAR PROGRAMMING”

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 1,117 records · Page 62

SU-G-JeP2-04: Comparison Between Fricke-Type 3D Radiochromic Dosimeters for Real-Time Dose Distribution Measurements in MR-Guided Radiation Therapy

Purpose: To assess MR signal contrast for different ferrous ion compounds used in Fricke-type gel dosimeters for real-time dose measurements for MR-guided radiation therapy applications. Methods: Fricke-type gel dosimeters were prepared in 4% w/w gelatin prior to irradiation in an integrated 1.5 T MRI and 7 MV linear accelerator system (MR-Linac). 4 different ferrous ion (Fe2?) compounds (referred to as A, B, C, and D) were investigated for this study. Dosimeter D consisted of ferrous ammonium sulfate (FAS), which is conventionally used for Fricke dosimeters. Approximately half of each cylindrical dosimeter (45 mm diameter, 80 mm length) was irradiated to ∼17 Gy. MR imaging during irradiation was performed with the MR-Linac using a balanced-FFE sequence of TR/TE = 5/2.4 ms. An approximate uncertainty of 5% in our dose delivery was anticipated since the MR-Linac had not yet been fully commissioned. Results: The signal intensities (SI) increased between the un-irradiated and irradiated regions by approximately 8.6%, 4.4%, 3.2%, and 4.3% after delivery of ∼2.8 Gy for dosimeters A, B, C, and D, respectively. After delivery of ∼17 Gy, the SI had increased by 24.4%, 21.0%, 3.1%, and 22.2% compared to the un-irradiated regions. The increase in SI with respect to dose was linear for dosimeters A, B, and D with slopes of 0.0164, 0.0251, and 0.0236 Gy{sup −1} (R{sup 2} = 0.92, 0.97, and 0.96), respectively. Visually, dosimeter A had the greatest optical contrast from yellow to purple in the irradiated region. Conclusion: This study demonstrated the feasibility of using Fricke-type dosimeters for real-time dose measurements with the greatest optical and MR contrast for dosimeter A. We also demonstrated the need to investigate Fe{sup 2+} compounds beyond the conventionally utilized FAS compound in order to improve the MR signal contrast in 3D dosimeters used for MR-guided radiation therapy. This material is based upon work supported by the National Science Foundation Graduate Research Fellowship Program under Grant No. LH- 102SPS.

60 APPLIED LIFE SCIENCES↗

Monte Carlo Simulations of the Dissolution of Borosilicate and Aluminoborosilicate Glasses in Dilute Aqueous Solutions

The aim of this study was to provide atomic-level insights into the dissolution behavior of borosilicate and aluminoborosilicate glasses to complement and help interpret previous experimental work on the NeB glass series studied by Pierce et al. [Pierce E. M., Reed L. R., Shaw W. J., McGrail B. P., Icenhower J. P., Windisch C. F., Cordova E. A. and Broady J. (2010) Experimental determination of the effect of the ratio of B/Al on glass dissolution along the nepheline (NaAlSiO4) - Malinkoite (NaBSiO4) join. Geochim. Cosmochim. Acta 74, 2634-2654]. The composition of these glasses was 50 mol% SiO2 - 25 mol% Na2O - (25-x) mol% Al2O3 - x mol% B2O3, with x varying from 0 to 20 mol%. In the first part of this work, the different structural features of these glasses (e.g., presence of non-bridging oxygens, partition of boron between trigonal and tetrahedral bonding environments, and formation of boroxol rings), identified in the study of Pierce et al., were implemented in the Monte Carlo program. Their effects on the dissolution of borosilicate and aluminosilicate glasses were then evaluated individually and led to the following conclusions. (1) The dependence of the dissolution rate on the amount of non-bridging oxygens was found to be linear at all Si/B ratios and the accelerating effect of non-bridging oxygens was shown to increase with increasing Si/B ratio. (2) The formation of boroxol rings and of clusters of boroxol rings resulted in an increase of the dissolution rate at all Si/B ratios and, again, the extent of the rate increase was strongly dependent on the Si/B ratio. (3) For aluminosilicate glasses, the implementation of the aluminum avoidance rule was found to increase the rate of dissolution relative to that obtained for a random distribution. In the second part of this work, Monte Carlo simulations were performed to model the dissolution of the NeB glasses in dilute conditions. One of the conclusions that emerged from the study of Pierce et al. was that either the rupture of the Al-O bonds or that of the Si-O bonds was the rate-limiting step controlling the dissolution of the NeB glasses. The Monte Carlo simulations carried out in this work enabled us to refine this conclusion. Indeed, the simulations showed that, at low B/Al ratios, the rupture of both Al-O-Si and Si-O-Si linkages contributed to the dissolution rate whereas, at high B/Al ratios, the dissolution rate was independent of the rupture of Al-O-Si linkages and was controlled by S1 sites (silicon sites at the glass-water interface with one bond to nearest-neighbor sites) and dissolution via detachment of clusters.

aluminosilicates↗

Fermilab proton accelerator complex status and improvement plans

Fermilab carries out an extensive program of accelerator-based high energy particle physics research at the Intensity Frontier that relies on the operation of 8 GeV and 120 GeV proton beamlines for a n umber of fixed target experiments. Routine operation with a world-record 700kW of average 120 GeV beam power on the neutrino target was achieved in 2017 as the result of the Proton Improvement Plan (PIP) upgrade. There are plans to further increase the power to 900 – 1000 kW. The next major upgrade of the FNAL accelerator complex, called PIP-II, is under development. It aims at 1.2MW beam power on target at the start of the LBNF/DUNE experiment in the middle of the next decade and assumes replacement of the existing 40-years old 400 MeV normal-conducting Linac with a modern 800 MeV superconducting RF linear accelerator. There are several concepts to further double the beam power to >2.4MW after replacement of the existing 8 GeV Booster synchrotron. In this article we discuss current performance of the Fermilab proton accelerator complex, the upgrade plans for the next two decades and the accelerator R&D program to address cost and performance risks for these upgrades.

43 PARTICLE ACCELERATORS↗

Data-driven gradient optimization for field emission management in a superconducting radio-frequency linac

Field emission can cause significant problems in superconducting radio-frequency linear accelerators (linacs). When cavity gradients are pushed higher, radiation levels within the linacs may rise exponentially, causing degradation of many nearby systems. This research aims to utilize machine learning with uncertainty quantification to predict radiation levels at multiple locations throughout the linacs and ultimately optimize cavity gradients to reduce field emission-induced radiation while maintaining the total linac energy gain necessary for the experimental physics program. The optimized solutions show over 40% reductions for both neutron and gamma radiation from the standard operational settings. Published by the American Physical Society 2025

43 PARTICLE ACCELERATORS↗

Optimal Sizing and Operation of a Hybrid Clean Energy Center

In this paper, we consider the deployment of battery energy storage systems (BESSs) paired with photovoltaics (PV) to form a hybrid clean energy center (HCEC), where the BESS is equipped for two main purposes: i) reducing the volatility of PV output and ii) shifting energy to meet an export limit. We develop a linear dynamical model to characterize the output of an HCEC and express the volatility and export power as functions of the battery's dispatch operations. We also propose an optimal sizing method via second-order cone programming (SOCP) to maximize the net benefits of the HCEC. Numerical studies are performed to illustrate and validate the proposed method.

Battery energy storage systems, photovoltaics, sec↗

Modelling feedbacks between human and natural processes in the land system

Abstract. The unprecedented use of Earth's resources by humans, in combination with increasing natural variability in natural processes over the past century, is affecting the evolution of the Earth system. To better understand natural processes and their potential future trajectories requires improved integration with and quantification of human processes. Similarly, to mitigate risk and facilitate socio-economic development requires a better understanding of how the natural system (e.g. climate variability and change, extreme weather events, and processes affecting soil fertility) affects human processes. Our understanding of these interactions and feedback between human and natural systems has been formalized through a variety of modelling approaches. However, a common conceptual framework or set of guidelines to model human–natural-system feedbacks is lacking. The presented research lays out a conceptual framework that includes representing model coupling configuration in combination with the frequency of interaction and coordination of communication between coupled models. Four different approaches used to couple representations of the human and natural system are presented in relation to this framework, which vary in the processes represented and in the scale of their application. From the development and experience associated with the four models of coupled human–natural systems, the following eight lessons were identified that if taken into account by future coupled human–natural-systems model developments may increase their success: (1) leverage the power of sensitivity analysis with models, (2) remember modelling is an iterative process, (3) create a common language, (4) make code open-access, (5) ensure consistency, (6) reconcile spatio-temporal mismatch, (7) construct homogeneous units, and (8) incorporating feedback increases non-linearity and variability. Following a discussion of feedbacks, a way forward to expedite model coupling and increase the longevity and interoperability of models is given, which suggests the use of a wrapper container software, a standardized applications programming interface (API), the incorporation of standard names, the mitigation of sunk costs by creating interfaces to multiple coupling frameworks, and the adoption of reproducible workflow environments to wire the pieces together.

58 GEOSCIENCES↗

Noisy-Intermediate-Scale Quantum Electromagnetic Transients Program

Quantum-empowered electromagnetic transients program (QEMTP) is a promising paradigm for tackling EMTP's computational burdens. Nevertheless, no existing studies truly achieve a practical and scalable QEMTP operable on today's noisy-intermediate-scale quantum (NISQ) computers. The strong reliance on noise-free and fault-tolerant quantum devices--which appears to be decades away--hinder practical applications of current QEMTP methods. Here, we devise a NISQ-QEMTP methodology which for the first time transitions the QEMTP operations from ideal, noise-free quantum simulators to real, noisy quantum computers. The main contributions lie in: (1) a shallow-depth QEMTP quantum circuit for mitigating noises on NISQ quantum devices; (2) practical QEMTP linear solvers incorporating executable quantum state preparation and measurements for nodal voltage computations; (3) a noise-resilient QEMTP algorithm leveraging quantum resources logarithmically scaled with power system dimension; (4) a quantum shifted frequency analysis (QSFA) for accelerating QEMTP by exploiting dynamic phasor simulations with larger time steps; (5) a systematical analysis on QEMTPs performance under various noisy quantum environments. Extensive experiments systematically verify the accuracy, efficacy, universality and noise-resilience of QEMTP on both noise-free simulators and IBM real quantum computers.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Monte Carlo modeling and simulations of the High Definition (HD120) micro MLC and validation against measurements for a 6 MV beam

Purpose: The most recent Varian micro multileaf collimator (MLC), the High Definition (HD120) MLC, was modeled using the BEAMNRC Monte Carlo code. This model was incorporated into a Varian medical linear accelerator, for a 6 MV beam, in static and dynamic mode. The model was validated by comparing simulated profiles with measurements. Methods: The Varian Trilogy (2300C/D) accelerator model was accurately implemented using the state-of-the-art Monte Carlo simulation program BEAMNRC and validated against off-axis and depth dose profiles measured using ionization chambers, by adjusting the energy and the full width at half maximum (FWHM) of the initial electron beam. The HD120 MLC was modeled by developing a new BEAMNRC component module (CM), designated HDMLC, adapting the available DYNVMLC CM and incorporating the specific characteristics of this new micro MLC. The leaf dimensions were provided by the manufacturer. The geometry was visualized by tracing particles through the CM and recording their position when a leaf boundary is crossed. The leaf material density and abutting air gap between leaves were adjusted in order to obtain a good agreement between the simulated leakage profiles and EBT2 film measurements performed in a solid water phantom. To validate the HDMLC implementation, additional MLC static patterns were also simulated and compared to additional measurements. Furthermore, the ability to simulate dynamic MLC fields was implemented in the HDMLC CM. The simulation results of these fields were compared with EBT2 film measurements performed in a solid water phantom. Results: Overall, the discrepancies, with and without MLC, between the opened field simulations and the measurements using ionization chambers in a water phantom, for the off-axis profiles are below 2% and in depth-dose profiles are below 2% after the maximum dose depth and below 4% in the build-up region. On the conditions of these simulations, this tungsten-based MLC has a density of 18.7 g cm{sup -3} and an overall leakage of about 1.1 {+-} 0.03%. The discrepancies between the film measured and simulated closed and blocked fields are below 2% and 8%, respectively. Other measurements were performed for alternated leaf patterns and the agreement is satisfactory (to within 4%). The dynamic mode for this MLC was implemented and the discrepancies between film measurements and simulations are within 4%. Conclusions: The Varian Trilogy (2300 C/D) linear accelerator including the HD120 MLC was successfully modeled and simulated using the Monte Carlo BEAMNRC code by developing an independent CM, the HDMLC CM, either in static and dynamic modes.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Adiabatic Quantum Programming: Minor Embedding With Hard Faults

Adiabatic quantum programming defines the time-dependent mapping of a quantum algorithm into the hardware or logical fabric. An essential programming step is the embedding of problem-specific information into the logical fabric to define the quantum computational transformation. We present algorithms for embedding arbitrary instances of the adiabatic quantum optimization algorithm into a square lattice of specialized unit cells. Our methods are shown to be extensible in fabric growth, linear in time, and quadratic in logical footprint. In addition, we provide methods for accommodating hard faults in the logical fabric without invoking approximations to the original problem. These hard fault-tolerant embedding algorithms are expected to prove useful for benchmarking the adiabatic quantum optimization algorithm on existing quantum logical hardware. We illustrate this versatility through numerical studies of embeddabilty versus hard fault rates in square lattices of complete bipartite unit cells.

graph theory↗

Energy Recovery Linacs for Light Source Applications

Energy Recovery Linacs are being considered for applications in present and future light sources. ERLs take advantage of the continuous operation of superconducting rf cavities to accelerate high average current beams with low losses. The electrons can be directed through bends, undulators, and wigglers for high brightness x ray production. They are then decelerated to low energy, recovering power so as to minimize the required rf drive and electrical draw. When this approach is coupled with advanced continuous wave injectors, very high power, ultra-short electron pulse trains of very high brightness can be achieved. This paper will review the status of worldwide programs and discuss the technology challenges to provide such beams for photon production.

43 PARTICLE ACCELERATORS↗

ATF2 Summary and Status

The Accelerator Test Facility (ATF) [1] at KEK was originally constructed as a Damping Ring (DR) demonstration facility for the next generation of electron linear colliders. It consists of a 1.3 GeV s-band linac and a 136m circumference damping ring which can routinely generate 1.6 nC electron bunches with very small vertical emittances (<10pm.rad). Completed in December of 2008, a new extraction line and final focus optics was built, this is referred to as ATF2 [2]. The location of ATF in-relation to the KEK site can be seen in Figure 1. ATF is a general-purpose facility for accelerator physics related R and D activities, with approximately 50% of the available beam time currently devoted to the ATF2 program. The ATF2 extraction line (EXT) and final focus system (FFS) extends about 90 meters from the DR extraction point, a sketch of this system can be seen in figure 2, along with the location of key instrumentation. The primary purpose of the ATF2 beam line is to demonstrate a new style of magnetic focusing optics intended for use at future linear colliders, such as ILC or CLIC based on a local chromaticity correction scheme. The primary goal of the collaboration is to demonstrate the ability of this optics to focus an electron beam of similar emittance to the ILC using a FFS of the same design chromaticity. A comparison of some of the ATF2 design parameters with those of the ILC and CLIC are shown in table 1. With the 1.3 GeV beam energy of ATF2, the target spot size to prove the operability of this optics for ILC is 37nm. This is referred to as 'Goal 1'. In addition, Goal 1 requires the demonstration of stable operations with such a beam size at the Interaction Point (IP). 'Goal 2' is to demonstrate nm-level stability with respect to the IP beam monitors, such as is required for stable beam collisions at ILC or CLIC. A further goal is to attempt to increase the chromaticity of the final focus optics, reducing the IP beta functions to a level where the system more closely resembles the CLIC FFS, and also for some of the newly proposed ILC parameters. It is anticipated that the difficulty of successfully tuning the FFS to deliver design IP beam spot sizes scales with the chromaticity of the system. The design of the achromatic focusing system involves fine cancelation of first, second and third order aberrations at the IP. One needs more accurate cancelation the tighter one attempts to focus the beam, also including more higher order terms. By gradually pushing the system to smaller IP beta functions and higher chromaticity we hope to learn what the limits of such a design are when applied practically. This will give confidence in the design philosophy behind future high-luminosity colliders and in the simulations used to predict their behaviour.

43 PARTICLE ACCELERATORS↗

Evaluation of a semiautomated lung mass calculation technique for internal dosimetry applications

Purpose: The authors sought to evaluate a simple, semiautomated lung mass estimation method using computed tomography (CT) scans obtained using a variety of acquisition techniques and reconstruction parameters for mass correction of medical internal radiation dose-based internal radionuclide radiation absorbed dose estimates.Methods: CT scans of 27 patients with lung cancer undergoing stereotactic body radiation therapy treatment planning with PET/CT were analyzed retrospectively. For each patient, free-breathing (FB) and respiratory-gated 4DCT scans were acquired. The 4DCT scans were sorted into ten respiratory phases, representing one complete respiratory cycle. An average CT reconstruction was derived from the ten-phase reconstructions. Mid expiration breath-hold CT scans were acquired in the same session for many patients. Deep inspiration breath-hold diagnostic CT scans of many of the patients were obtained from different scanning sessions at similar time points to evaluate the effect of contrast administration and maximum inspiration breath-hold. Lung mass estimates were obtained using all CT scan types, and intercomparisons made to assess lung mass variation according to scan type. Lung mass estimates using the FB CT scans from PET/CT examinations of another group of ten male and ten female patients who were 21–30 years old and did not have lung disease were calculated and compared with reference lung mass values. To evaluate the effect of varying CT acquisition and reconstruction parameters on lung mass estimation, an anthropomorphic chest phantom was scanned and reconstructed with different CT parameters. CT images of the lungs were segmented using the OsiriX MD software program with a seed point of about −850 HU and an interval of 1000. Lung volume, and mean lung, tissue, and air HUs were recorded for each scan. Lung mass was calculated by assuming each voxel was a linear combination of only air and tissue. The specific gravity of lung volume was calculated using the formula (lung HU − air HU)/(tissue HU − air HU), and mass = specific gravity × total volume × 1.04 g/cm{sup 3}.Results: The range of calculated lung masses was 0.51–1.29 kg. The average male and female lung masses during FB CT were 0.80 and 0.71 kg, respectively. The calculated lung mass varied across the respiratory cycle but changed to a lesser degree than did lung volume measurements (7.3% versus 15.4%). Lung masses calculated using deep inspiration breath-hold and average CT were significantly larger (p < 0.05) than were some masses calculated using respiratory-phase and FB CT. Increased voxel size and smooth reconstruction kernels led to high lung mass estimates owing to partial volume effects.Conclusions: Organ mass correction is an important component of patient-specific internal radionuclide dosimetry. Lung mass calculation necessitates scan-based density correction to account for volume changes owing to respiration. The range of lung masses in the authors’ patient population represents lung doses for the same absorbed energy differing from 25% below to 64% above the dose found using reference phantom organ masses. With proper management of acquisition parameters and selection of FB or midexpiration breath hold scans, lung mass estimates with about 10% population precision may be achieved.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

INL Senior Project

What did my team set out to accomplish:? Can I put a custom Machine Learning Model on FPGA?? Can I analyze network traffic in real time?? Can a QSFP port be used with an FPGA?? Does a visual representation of the latent space enhance our understanding of network traffic?? What is QSFP QSFP (Quad Small Form-Factor Pluggable)? QSFP supports transfer speeds generally up to 100Gb/s? Runs 4 parallel lines running up to 28 Gb/s? Why the latent space is important to our project ?Latent space is the compressed mapping of data points in a non-linear fashion? Create an understanding of the relationship of data collected? Can represent that relationship of a single network packet in 3 points (X, Y, Z) Project Outline FPGA? Custom Xilinx Petalinux Image for the operating system? Python program to collect packets and run them through the DPU (Data Processing Unit)? The program then sends the information over a socket to a computer? Display Program? Python Program that collects the information sent from the FPGA and display it in a graph

99 GENERAL AND MISCELLANEOUS↗

Differentiable programming for online training of a neural artificial viscosity function within a staggered grid Lagrangian hydrodynamics scheme

Lagrangian methods to solve the inviscid Euler equations produce numerical oscillations near shock waves. A common approach to reducing these oscillations is to add artificial viscosity (AV) to the discrete equations. The AV term acts as a dissipative mechanism that attenuates oscillations by smearing the shock across a finite number of computational cells. However, AV introduces several control parameters that are not determined by the underlying physical model, and hence, in practice are tuned to the characteristics of a given problem. We seek to improve the standard quadratic-linear AV form by replacing it with a learned neural function that reduces oscillations relative to exact solutions of the Euler equations, resulting in a hybrid numerical-neural hydrodynamic solver. Because AV is an artificial construct that exists solely to improve the numerical properties of a hydrodynamic code, there is no offline ‘viscosity data’ against which a neural network can be trained before inserting into a numerical simulation, thus requiring online training. We achieve this via differentiable programming, i.e. end-to-end backpropagation or adjoint solution through both the neural and differential equation code, using automatic differentiation of the hybrid code in the Julia programming language to calculate the necessary loss function gradients. A novel offline pre-training step accelerates training by initializing the neural network to the default numerical AV scheme, which can be learned rapidly by space-filling sampling over the AV input space. We find that online training over early time steps of simulation is sufficient to learn a neural AV function that reduces numerical oscillations in long-term hydrodynamic shock simulations. These results offer an early proof-of-principle that online differentiable training of hybrid numerical schemes with novel neural network components can improve certain performance aspects existing in purely numerical schemes.

97 MATHEMATICS AND COMPUTING↗

Multilane Automated Driving With Optimal Control and Mixed-Integer Programming

Road vehicle lane changes often initiate traffic disturbances and can therefore impact road networks’ energy and time efficiency. Furthermore, unexpected changes in traffic conditions may also render lane changes counterproductive for the lane-changing vehicle. Vehicle-to-vehicle connectivity combined with anticipative control could address these challenges via improved lane change decisions by automated vehicles. In a move toward this objective, receding horizon control cast as a mixed-integer quadratic program is used to plan lane changing and acceleration in a coupled optimization. A long-term pacing module, based on Pontryagin’s minimum principle from optimal control theory, sets terminal and input references for receding horizon control to target a user’s expected travel time. To remove nonlinear vehicle dynamics from the receding horizon controller, lane change commands are passed to a pure pursuit steering module whose response is approximated by a second-order linear model. Here, comparison against a rule-based reactive algorithm in arterial and highway scenarios shows an 8.9%–13.7% reduction in energy consumption and a 5.2%–10.3% reduction in the travel time, along with navigational improvements.

33 ADVANCED PROPULSION SYSTEMS↗

CSPlib: A performance portable parallel software toolkit for analyzing complex kinetic mechanisms

Computational singular perturbation (CSP) is a method to analyze dynamical systems. It targets the decoupling of fast and slow dynamics using an alternate linear expansion of the right-hand side of the governing equations based on eigenanalysis of the associated Jacobian matrix. This representation facilitates diagnostic analysis, detection and control of stiffness, and the development of simplified models. For this work, we have implemented CSP in a C++ open-source library CSPlib using the Kokkos parallel programming model to address portability across diverse heterogeneous computing platforms, i.e., multi/many-core CPUs and GPUs. We describe the CSPlib implementation and present its computational performance across different computing platforms using several test problems. Specifically, we test the CSPlib performance for a constant pressure ignition reactor model on different architectures, including IBM Power 9, Intel Xeon Skylake, and NVIDIA V100 GPU. The size of the chemical kinetic mechanism is varied in these tests. As expected, the Jacobian matrix evaluation, the eigensolution of the Jacobian matrix, and matrix inversion are the most expensive computational tasks. When considering the higher throughput characteristic of GPUs, GPUs performs better for small matrices with higher occupancy rate. CPUs gain more advantages from the higher performance of well-tuned and optimized linear algebra libraries such as OpenBLAS.

97 MATHEMATICS AND COMPUTING↗

High Performance Molecular Dynamic Simulation on Single and Multi-GPU Systems

The programming techniques supported and employed on these GPUs and Multi-GPUs systems are not sufficient to address problems exhibiting irregular, and unbalanced workload such as Molecular Dynamic (MD) simulations of systems with non-uniform densities. In this paper, we propose a task-based dynamic load-balancing solution to employ on MD simulations for single- and multi-GPU systems. The solution allows load balancing at a finer granularity than what is supported in existing APIs such as NVIDIA’s CUDA. Experimental results with a single-GPU configuration show that our fine-grained task solution can utilize the hardware more efficiently than the CUDA scheduler. On multi-GPU systems, our solution achieves near-linear speedup, load balance, and significant performance improvement over techniques based on standard CUDA APIs.

97 MATHEMATICS AND COMPUTING↗

A family of second-order fully explicit time integration schemes

A new family of fully explicit time integration methods is proposed, which has second-order accuracy for structures with and without damping. Using a diagonal mass matrix, the suggested scheme remains fully explicit not only for structures with a non-diagonal damping matrix, but also when the internal force vector is a nonlinear function of velocity. The present algorithm has an acceptable domain of stability, and it is self-starting. This technique introduces effectively numerical dissipation to suppress the high-frequency spurious modes, while at the same time the lower modes are not affected too much. In addition, numerical dispersion error of the scheme is considerably smaller than that of the central difference method. Solving several linear and nonlinear problems highlights the superior performance of the authors’ approach. Findings demonstrate that solution time for the suggested scheme is much less than that of the central difference technique. The related algorithm can be easily implemented into programs, which already contain the central difference method.

97 MATHEMATICS AND COMPUTING↗