Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Edge computing”

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 1,315 records · Page 73

Numerical simulation of separated and vortical flows on bodies at large angles of attack

Navier-Stokes computations were carried out for high-angle-of-attack flow about bodies of revolution in supersonic and subsonic flow. Numerical and physical factors affecting flow predictions are discussed. Accurate prediction of the complex three-dimensional separated flow is found to depend on using grids sufficiently fine to resolve the details of both the viscous boundary layer and the off-surface separated flow structures, and, for turbulent flow cases, use of an eddy-viscosity turbulence model which accounts for the leeward vortical flow structures. An example of high-incidence flow computed about a more complex geometry, the F-18 fighter fuselage forebody and leading edge extension, is also presented.

Schiff, Lewis B.↗

The design and implementation of a parallel unstructured Euler solver using software primitives

This paper is concerned with the implementation of a three-dimensional unstructured grid Euler-solver on massively parallel distributed-memory computer architectures. The goal is to minimize solution time by achieving high computational rates with a numerically efficient algorithm. An unstructured multigrid algorithm with an edge-based data structure has been adopted, and a number of optimizations have been devised and implemented in order to accelerate the parallel communication rates. The implementation is carried out by creating a set of software tools, which provide an interface between the parallelization issues and the sequential code, while providing a basis for future automatic run-time compilation support. Large practical unstructured grid problems are solved on the Intel iPSC/860 hypercube and Intel Touchstone Delta machine. The quantitative effect of the various optimizations are demonstrated, and we show that the combined effect of these optimizations leads to roughly a factor of three performance improvement. The overall solution efficiency is compared with that obtained on the CRAY-YMP vector supercomputer.

Das, R.↗

Leading-edge effects on boundary-layer receptivity

Numerical calculations are presented for the incompressible flow over a parabolic cylinder. The computational domain extends from a region upstream of the body downstream to the region where the Blasius boundary-layer solution holds. A steady mean flow solution is computed and the results for the scaled surface vorticity, surface pressure and displacement thickness are compared to previous studies. The unsteady problem is then formulated as a perturbation solution starting with and evolving from the mean flow. The response to irrotational time harmonic pulsation of the free-stream is examined. Results for the initial development of the velocity profile and displacement thickness are presented. These calculations will be extended to later times to investigate the initiation of instability waves within the boundary-layer.

Gatski, Thomas B.↗

Interferometry and computational studies of an oscillating airfoil compressible dynamic stall flow field

A unique comparison has been made between real time interferograms and full Navier-Stokes computations of the density field over an oscillating airfoil undergoing dynamic stall for compressible flow conditions. Good agreement was found until a dynamic stall vortex formed in the flow. Subsequent evolution of the flow field was found to be very different in the computations. The reasons for this difference have been explained in terms of the leading edge region flow physics and the refined flow modeling that needs to be used for the post-stall flow field.

Chandrasekhara, M. S.↗

Computation and Validation of Fluid/Structure Twin Tail Buffet Response

The buffet response of the flexible twin-tail/delta wing configuration, a multidisciplinary problem, is solved using three sets of equations on a multi-block grid structure. The first set is the unsteady, compressible, full Navier-Stokes equations which are used for obtaining the flow-field vector and the aerodynamic loads on the twin tails. The second set is the coupled aeroelastic equations which are used for obtaining the bending and torsional deflections of the twin tails. The third set is the grid-displacement equations which are used for updating the grid coordinates due to the tail deflections. The computational model is similar to the one used by Washburn et. al. which consists of a delta wing of aspect ratio one and twin tails with taper ratios of 0.23. The vortex of the twin tails are located at the wing trailing edge. The configuration is pitched at 30 deg angle of attack, and the freestream Mach number and Reynolds number are 0.3 and 1.25 million, respectively. With the twin tails fixed as rigid surfaces, the problem is solved for the initial flow conditions. Next, the problem is solved for the twin tail response for uncoupled bending and torsional vibrations due to the unsteady loads produced by the vortex breakdown flow of the leading-edge vortex cores. The configuration is investigated for three spanwise positions of the twin tails; inboard, midspan and outboard locations. The computational results are validated and are in very good agreement with the experimental data of Washburn, et. al.

Massey, Steven J.↗

MIT-NASA Workshop: Transformational Technologies

As a space faring nation, we are at a critical juncture in the evolution of space exploration. NASA has announced its Vision for Space Exploration, a vision of returning humans to the Moon, sending robots and eventually humans to Mars, and exploring the outer solar system via automated spacecraft. However, mission concepts have become increasingly complex, with the potential to yield a wealth of scientific knowledge. Meanwhile, there are significant resource challenges to be met. Launch costs remain a barrier to routine space flight; the ever-changing fiscal and political environments can wreak havoc on mission planning; and technologies are constantly improving, and systems that were state of the art when a program began can quickly become outmoded before a mission is even launched. This Conference Publication describes the workshop and featured presentations by world-class experts presenting leading-edge technologies and applications in the areas of power and propulsion; communications; automation, robotics, computing, and intelligent systems; and transformational techniques for space activities. Workshops such as this one provide an excellent medium for capturing the broadest possible array of insights and expertise, learning from researchers in universities, national laboratories, NASA field Centers, and industry to help better our future in space.

Mankins, J. C.↗

Numerical Comparison of Periodic MoM (Method of Moments) and BMIA (Banded Matrix Iteration Method)

The most popular numerical technique in rough surface scattering is the Method of Moments (MoM). Since the scattering patch size is finite, the edge current must be suppressed to obtain accurate scattering cross sections. Two standard ways to minimize the edge current are periodic boundary conditions and incident wave tapering. We compare the accuracy & computational requirements of these methods.

Method of Moments Banded Matrix Iteration Method M↗

Processing circuitry for single channel radiation detector

Processing circuitry is provided for a high voltage operated radiation detector. An event detector utilizes a comparator configured to produce an event signal based on a leading edge threshold value. A preferred event detector does not produce another event signal until a trailing edge threshold value is satisfied. The event signal can be utilized for counting the number of particle hits and also for controlling data collection operation for a peak detect circuit and timer. The leading edge threshold value is programmable such that it can be reprogrammed by a remote computer. A digital high voltage control is preferably operable to monitor and adjust high voltage for the detector.

Holland, Samuel D.↗

Receptivity and Transition of Supersonic Boundary Layers Over Swept Wings

The receptivity, stability, and transition of three-dimensional supersonic boundary layers over (1) a swept cylinder, (2) a swept wing with a sharp leading edge, and (3) a swept wing with a blunt leading edge are numerically investigated for a free-stream Mach number of 3. These computations are compared to an earlier experimental and computational study performed by Archambaud et al.1 The steady flow fields with and without roughness elements are obtained by solving the full Navier-Stokes equations. The N-factors computed in this study at the transition onset locations reported in Ref. 1 for flow over the swept cylinder are approximately 16.5 for traveling crossflow disturbances and 9 for stationary disturbances. The N-factors for the traveling crossflow are high based on our past experiences. However, they are comparatively smaller than those reported by Archambaud et al., who found N-factor values in the range of 20 to 25 for traveling disturbances and 13 to 20 for stationary disturbances. Similarly, the N-factors computed in this study for the traveling and stationary disturbances for the flow over the sharp wing are approximately 7 and 2.5, respectively, and for the flow over the blunt wing are 6.5 and 4.8, respectively. Using the envelope method, Archambaud et al. obtained values of approximately 8.0 and 4.0 for the sharp wing case and 16.0 and 12.0 for the blunt wing case.

Balakumar, Ponnampalam↗

Novel Micro Strip-to-Waveguide Feed Employing a Double-Y Junction

Previous micro strip -to -waveguide transitions either required a hermetically sealed waveguide configuration, or a balun that needed to be tuned according to the frequency band of interest. In this design, the balun is realized using a double -Y junction to transition from microstrip to coplanar strip feeding a quasi-Yagi dipole array (see figure). The length of the feed (Lf) extending into the waveguide is 15.54 mm. The length of the ground plane below the ULTRALAM substrate is 7.75 mm. The lengths L1 , L2, and L3 are 8.50 mm, 4.38 mm, and 2.14 mm, respectively. These lengths were computed via a preliminary optimization aimed at improving the return loss at the band edges. The waveguide feed was designed to excite the TE10 mode in a WR-90 waveguide, and to operate over the recommended frequencies of 8.2 to 12.4 GHz. The feed employs a Rogers 6010 substrate (dielectric constant Er approx. equals 10.2) bonded with a Rogers ULTRALAM substrate (Er approx. equals 2.5). The ULTRALAM substrate serves to provide mechanical strength for 6010 substrate, and to mitigate loses due to parasitic modes (the ground plane is etched on the bottom of this layer due to the topology of the double -Y balun). The double-Y balun transitioning from an unbalanced microstrip line to a balanced coplanar strip (CPS) line does not provide inherent impedance transformation; hence, Klopfenstein impedance tapers were synthesized to transition from 50 to 77 Ohms in the microstrip section and from 77 to 110 Ohms in the CPS section. At the balun junction, the CPS stub lengths were chosen such that the X/ 8 resonance is pushed outside the bandwidth of operation. Also, the smallest allowable conductor width and gap spacing were chosen to meet acceptable manufacturing tolerances.

Venkatesan, Jaikrishina↗

Adapting the SpaceCube v2.0 Data Processing System for Mission-Unique Application Requirements

The SpaceCube (sup TM) v2.0 system is a superior high performance, reconfigurable, hybrid data processing system that can be used in a multitude of applications including those that require a radiation hardened and reliable solution. This paper provides an overview of the design architecture, flexibility, and the advantages of the modular SpaceCube v2.0 high performance data processing system for space applications. The current state of the proven SpaceCube technology is based on nine years of engineering and operations. Five systems have been successfully operated in space starting in 2008 with four more to be delivered for launch vehicle integration in 2015. The SpaceCube v2.0 system is also baselined as the avionics solution for five additional flight projects and is always a top consideration as the core avionics for new instruments or spacecraft control. This paper will highlight how this multipurpose system is currently being used to solve design challenges of three independent applications. The SpaceCube hardware adapts to new system requirements by allowing for application-unique interface cards that are utilized by reconfiguring the underlying programmable elements on the core processor card. We will show how this system is being used to improve on a heritage NASA GPS technology, enable a cutting-edge LiDAR instrument, and serve as a typical command and data handling (C&DH) computer for a space robotics technology demonstration.

SpaceCube↗

Utilizing Direct Numerical Simulations of Transition and Turbulence in Design Optimization

Design optimization methods that use the Reynolds-averaged Navier-Stokes equations with the associated turbulence and transition models, or other model-based forms of the governing equations, may result in aerodynamic designs with actual performance levels that are noticeably different from the expected values because of the complexity of modeling turbulence/transition accurately in certain flows. Flow phenomena such as wake-blade interaction and trailing edge vortex shedding in turbines and compressors (examples of such flows) may require a computational approach that is free of transition/turbulence models, such as direct numerical simulations (DNS), for the underlying physics to be computed accurately. Here we explore the possibility of utilizing DNS data in designing a turbine blade section. The ultimate objective is to substantially reduce differences between predicted performance metrics and those obtained in reality. The redesign of a typical low-pressure turbine blade section with the goal of reducing total pressure loss in the row is provided as an example. The basic ideas presented here are of course just as applicable elsewhere in aerodynamic shape optimization as long as the computational costs are not excessive.

Direct numerical simulations of transition and tur↗

Semi-Supervised Learning of Lift Optimization of Multi-Element Three-Segment Variable Camber Airfoil

This chapter describes a new intelligent platform for learning optimal designs of morphing wings based on Variable Camber Continuous Trailing Edge Flaps (VCCTEF) in conjunction with a leading edge flap called the Variable Camber Krueger (VCK). The new platform consists of a Computational Fluid Dynamics (CFD) methodology coupled with a semi-supervised learning methodology. The CFD component of the intelligent platform comprises of a full Navier-Stokes solution capability (NASA OVERFLOW solver with Spalart-Allmaras turbulence model) that computes flow over a tri-element inboard NASA Generic Transport Model (GTM) wing section. Various VCCTEF/VCK settings and configurations were considered to explore optimal design for high-lift flight during take-off and landing. To determine globally optimal design of such a system, an extremely large set of CFD simulations is needed. This is not feasible to achieve in practice. To alleviate this problem, a recourse was taken to a semi-supervised learning (SSL) methodology, which is based on manifold regularization techniques. A reasonable space of CFD solutions was populated and then the SSL methodology was used to fit this manifold in its entirety, including the gaps in the manifold where there were no CFD solutions available. The SSL methodology in conjunction with an elastodynamic solver (FiDDLE) was demonstrated in an earlier study involving structural health monitoring. These CFD-SSL methodologies define the new intelligent platform that forms the basis for our search for optimal design of wings. Although the present platform can be used in various other design and operational problems in engineering, this chapter focuses on the high-lift study of the VCK-VCCTEF system. Top few candidate design configurations were identified by solving the CFD problem in a small subset of the design space. The SSL component was trained on the design space, and was then used in a predictive mode to populate a selected set of test points outside of the given design space. The new design test space thus populated was evaluated by using the CFD component by determining the error between the SSL predictions and the true (CFD) solutions, which was found to be small. This demonstrates the proposed CFD-SSL methodologies for isolating the best design of the VCK-VCCTEF system, and it holds promise for quantitatively identifying best designs of flight systems, in general.

Intelligent Systems↗

Magnet optical and beam matching issues in SNS MEBT

A Medium-Energy Beam Transport (MEBT) line is employed in the SNS linac. The MEBT lattice consists of fourteen electromagnetic quadrupoles and other devices. The quads have very small aspect ratios, and they are densely packed. Significant fringe fields and magnetic interference cause difficulties in beam matching. We have performed 3D simulations of the magnets, computed their optical properties, and compared their performance with what predicted by simple hard edge models. This paper reports our findings and a general solution to the problem.

43 PARTICLE ACCELERATORS↗

Roofline Analysis in the Intel® Advisor to Deliver Optimized Performance for applications on Intel® Xeon Phi™ Processor

In this session we show, in two case studies, how the roofline feature of Intel Advisor has been utilized to optimize the performance of kernels of the XGC1 and PICSAR codes in preparation for Intel Knights Landing architecture. The impact of the implemented optimizations and the benefits of using the automatic roofline feature of Intel Advisor to study performance of large applications will be presented. This demonstrates an effective optimization strategy that has enabled these science applications to achieve up to 4.6 times speed-up and prepare for future exascale architectures. # Goal/Relevance of Session The roofline model [1,2] is a powerful tool for analyzing the performance of applications with respect to the theoretical peak achievable on a given computer architecture. It allows one to graphically represent the performance of an application in terms of operational intensity, i.e. the ratio of flops performed and bytes moved from memory in order to guide optimization efforts. Given the scale and complexity of modern science applications, it can often be a tedious task for the user to perform the analysis on the level of functions or loops to identify where performance gains can be made. With new Intel tools, it is now possible to automate this task, as well as base the estimates of peak performance on measurements rather than vendor specifications. The goal of this session is to demonstrate how the roofline feature of Intel Advisor can be used to balance memory vs. computation related optimization efforts and effectively identify performance bottlenecks. A series of typical optimization techniques: cache blocking, structure refactoring, data alignment, and vectorization illustrated by the kernel cases will be addressed. # Description of the codes ## XGC1 The XGC1 code [3] is a magnetic fusion Particle-In-Cell code that uses an unstructured mesh for its Poisson solver that allows it to accurately resolve the edge plasma of a magnetic fusion device. After recent optimizations to its collision kernel [4], most of the computing time is spent in the electron push (pushe) kernel, where these optimization efforts have been focused. The kernel code scaled well with MPI+OpenMP but had almost no automatic compiler vectorization, in part due to indirect memory addresses and in part due to low trip counts of low-level loops that would be candidates for vectorization. Particle blocking and sorting have been implemented to increase trip counts of low-level loops and improve memory locality, and OpenMP directives have been added to vectorize compute-intensive loops that were identified by Advisor. The optimizations have improved the performance of the pushe kernel 2x on Haswell processors and 1.7x on KNL. The KNL node-for-node performance has been brought to within 30% of a NERSC Cori phase I Haswell node and we expect to bridge this gap by reducing the memory footprint of compute intensive routines to improve cache reuse. ## PICSAR is a Fortran/Python high-performance Particle-In-Cell library targeting at MIC architectures first designed to be coupled with the PIC code WARP for the simulation of laser-matter interaction and particle accelerators. PICSAR also contains a FORTRAN stand-alone kernel for performance studies and benchmarks. A MPI domain decomposition is used between NUMA domains and a tile decomposition (cache-blocking) handled by OpenMP has been added for shared-memory parallelism and better cache management. The so-called current deposition and field gathering steps that compose the PIC time loop constitute major hotspots that have been rewritten to enable more efficient vectorization. Particle communications between tiles and MPI domain has been merged and parallelized. All considered, these improvements provide speedups of 3.1 for order 1 and 4.6 for order 3 interpolation shape factors on KNL configured in SNC4 quadrant flat mode. Performance is similar between a node of cori phase 1 and KNL at order 1 and better on KNL by a factor 1.6 at order 3 with the considered test case (homogeneous thermal plasma).

97 MATHEMATICS AND COMPUTING↗

Scale–Resolving Simulations of a Fundamental Trailing–Edge Cooling Slot Using a Discontinuous–Galerkin Spectral–Element Method

The accurate prediction of turbulent mixing in high-pressure turbines that incorporate various airfoil surface-cooling strategies is becoming increasing critical to the design of modern gas turbine engines where the quest for improved efficiency is driving compressor overall pressure ratios and turbine inlet temperatures to much higher levels than ever before. In the present paper, a recently developed computational capability for accurate and efficient scale–resolving simulations of turbomachinery is extended to study the turbulent mixing mechanism of a simplified abstraction of an airfoil trailing-edge cooling slot - a plane wall jet with finite lip thickness discharging into an ambient flow. The computational capability is based on an entropy– stable, discontinuous–Galerkin approach that extends to arbitrarily high orders of spatial and temporal accuracy. The numerical results show that the present simulations capture the trends observed in the experiments. Discrepancies between the simulations and experiments are believed to be due to differences in the inflow profiles and tunnel side–wall effects. The thick lip configuration leads to a thicker wake and higher unsteadiness in the wall jet compared to the thin lip. A detailed comparison of the turbulent flowfields is presented to highlight differences arising due to lip thickness variations.

Scale Resolving↗

Flow fields within Jupiter's Great Red Spot and White Oval BC

Voyager 1 high-resolution images of Jupiter's Great Red Spot (GRS) and White Oval BC are used to map flow fields within these two areas. The relative vorticity is computed as a function of semi-major axis length and position angle in a coordinate system consisting of concentric ellipses of equal eccentricity. Wind speeds of 110-120 m/s are observed near the outer edge of both features, and along their minor axes relative vorticity profiles reach a maximum of 0.00006/s. Maximum Rossby numbers of 0.36 are computed for flows within both features, and are found to be low, indicating geostrophic constraints on the flow. The difference in streamline curvature within the GRS and the Oval BC is found to compensate for the difference in planetary vorticity at the respective latitudes of the features. Finally, motions within the central region of the GRS are slower and more random than around the spot's outer portion.

Mitchell, J. L.↗

Anomaly Detection in Dynamic Networks

Anomaly detection in dynamic communication networks has many important security applications. These networks can be extremely large and so detecting any changes in their structure can be computationally challenging; hence, computationally fast, parallelisable methods for monitoring the network are paramount. For this reason the methods presented here use independent node and edge based models to detect locally anomalous substructures within communication networks. As a first stage, the aim is to detect changes in the data streams arising from node or edge communications. Throughout the thesis simple, conjugate Bayesian models for counting processes are used to model these data streams. A second stage of analysis can then be performed on a much reduced subset of the network comprising nodes and edges which have been identified as potentially anomalous in the first stage. The first method assumes communications in a network arise from an inhomogeneous Poisson process with piecewise constant intensity. Anomaly detection is then treated as a changepoint problem on the intensities. The changepoint model is extended to incorporate seasonal behavior inherent in communication networks. This seasonal behavior is also viewed as a changepoint problem acting on a piecewise constant Poisson process. In a static time frame, inference is made on this extended model via a Gibbs sampling strategy. In a sequential time frame, where the data arrive as a stream, a novel, fast Sequential Monte Carlo (SMC) algorithm is introduced to sample from the sequence of posterior distributions of the change points over time. A second method is considered for monitoring communications in a large scale computer network. The usage patterns in these types of networks are very bursty in nature and don’t fit a Poisson process model. For tractable inference, discrete time models are considered, where the data are aggregated into discrete time periods and probability models are fitted to the communication counts. In a sequential analysis, anomalous behavior is then identified from outlying behavior with respect to the fitted predictive probability models. Seasonality is again incorporated into the model and is treated as a changepoint model on the transition probabilities of a discrete time Markov process. Second stage analytics are then developed which combine anomalous edges to identify anomalous substructures in the network.

97 MATHEMATICS AND COMPUTING↗