Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Operator approximation”

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 1,387 records · Page 77
All sourcesNASA NTRSDOE OSTI
NASA NTRS2011

Remote Sensing of Terrestrial Water Storage with GRACE and Future Gravimetry Missions

The Gravity Recovery and Climate Experiment (GRACE) has demonstrated that satellite gravimetry can be a valuable tool for regional to global water cycle observation. Studies of ice sheet and glacier mass losses, ocean bottom pressure and circulation, and variability of water stored on and in the land including groundwater all have benefited from GRACE observations, and the list of applications and discoveries continues to grow. As the mission approaches its tenth anniversary of launch on March 12,2012, it has nearly doubled its proposed lifetime but is showing some signs of age. In particular, degraded battery capacity limits the availability of power in certain orbital configurations, so that the accelerometers must be turned off for approximately one month out of six. The mission managers have decided to operate the spacecrafts in a manner that maximizes the remaining lifetime, so that the longest possible climate data record is available from GRACE. Nevertheless, it is not unlikely that there will be a data gap between GRACE and the GRACE Follow On mission, currently proposed for launch in 2016. In this presentation we will describe recent GRACE enabled science, GRACE mission health, and plans for GRACE Follow On and other future satellite gravimetry missions.

Rodell, Matt↗
NASA NTRS2016

Integrated Refrigeration and Storage for Advanced Liquid Hydrogen Operations

NASA has used liquefied hydrogen (LH2) on a large scale since the beginning of the space program as fuel for the Centaur and Apollo upper stages, and more recently to feed the three space shuttle main engines. The LH2 systems currently in place at the Kennedy Space Center (KSC) launch pads are aging and inefficient compared to the state-of-the-art. Therefore, the need exists to explore advanced technologies and operations that can drive commodity costs down, and provide increased capabilities. The Ground Operations Demonstration Unit for Liquid Hydrogen (GODU-LH2) was developed at KSC to pursue these goals by demonstrating active thermal control of the propellant state by direct removal of heat using a cryocooler. The project has multiple objectives including zero loss storage and transfer, liquefaction of gaseous hydrogen, and densification of liquid hydrogen. The key technology challenge was efficiently integrating the cryogenic refrigerator into the LH2 storage tank. A Linde LR1620 Brayton cycle refrigerator is used to produce up to 900W cooling at 20K, circulating approximately 22 g/s gaseous helium through the hydrogen via approximately 300 m of heat exchanger tubing. The GODU-LH2 system is fully operational, and is currently under test. This paper will discuss the design features of the refrigerator and storage system, as well as the current test results.

Liquefaction↗
DOE OSTI2022

Dynamic mode decomposition with core sketch

With the increase in collected data volumes, either from experimental measurements or high fidelity simulations, there is an ever-growing need to develop computationally efficient tools to process, analyze, and interpret these datasets. Modal analysis techniques have gained great interest due to their ability to identify patterns in the data and extract valuable information about the system being considered. Dynamic mode decomposition (DMD) relies on elements of the Koopman approximation theory to compute a set of modes, each associated with a fixed oscillation frequency and a decay/growth rate. Extracting these details from large datasets can be computationally expensive due to the need to implement singular value decomposition of the input data matrix. Sketching algorithms have become popular in numerical linear algebra where statistical theoretic approaches are utilized to reduce the cost of major operations. A sketch of a matrix is another matrix, which is significantly smaller, but still sufficiently approximates the original system. We put forth an efficient DMD framework, SketchyDMD, based on a core sketching algorithm that captures information about the range and corange (their mutual relationship) of input data. The proposed sketching-based framework can accelerate various portions of the DMD routines, compared to classical methods that operate directly on the raw input data. We conduct numerical experiments using the spherical shallow water equations as a prototypical model in the context of geophysical flows. In conclusion, we show that the proposed SketchyDMD is superior to existing randomized DMD methods that are based on capturing only the range of the input data.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
NASA NTRS2000

The fluid dynamic and shear environment in the NASA/JSC rotating-wall perfused-vessel bioreactor

The rotating-wall perfused-vessel (RWPV) bioreactor, used for both microgravity and Earth-based cell science experiments, is characterized in terms of the fluid dynamic and fluid shear stress environment. A numerical model of the flow field is developed and verified with laser Doppler velocimeter measurements. The effects of changes in operating conditions, including rotation rates and fluid perfusion rates, are investigated with the numerical model. The operating conditions typically used for ground-based experiments (equal rotation of the inner and outer cylinders) leads to flow patterns with relatively poor mass distribution characteristics. Approximately 50% of the inlet-perfused fluid bypasses the bulk of the fluid volume and flows to the perfusion exit. For operating conditions typical in microgravity, small differential rotation rates between the inner and outer cylinders lead to greatly improved flow distribution patterns and very low fluid shear stress levels over a large percentage of the fluid volume. Differences in flow patterns for the different operating conditions are explored. Large differences in the hydrodynamic environments for operating conditions typical of true microgravity and ground-based "microgravity simulations" are demonstrated.

Bioreactors↗
NASA NTRS2002

Integrated Stirling Convertor and Hall Thruster Test Conducted

An important aspect of implementing Stirling Radioisotope Generators on future NASA missions is the integration of the generator and controller with potential spacecraft loads. Some recent studies have indicated that the combination of Stirling Radioisotope Generators and electric propulsion devices offer significant trip time and payload fraction benefits for deep space missions. A test was devised to begin to understand the interactions between Stirling generators and electric thrusters. An electrically heated RG- 350 (350-W output) Stirling convertor, designed and built by Stirling Technology Company of Kennewick, Washington, under a NASA Small Business Innovation Research agreement, was coupled to a 300-W SPT-50 Hall-effect thruster built for NASA by the Moscow Aviation Institute (RIAME). The RG-350 and the SPT-50 shown, were installed in adjacent vacuum chamber ports at NASA Glenn Research Center's Electric Propulsion Laboratory, Vacuum Facility 8. The Stirling electrical controller interfaced directly with the Hall thruster power-processing unit, both of which were located outside of the vacuum chamber. The power-processing unit accepted the 48 Vdc output from the Stirling controller and distributed the power to all the loads of the SPT-50, including the magnets, keeper, heater, and discharge. On February 28, 2001, the Glenn test team successfully operated the Hall-effect thruster with the Stirling convertor. This is the world's first known test of a dynamic power source with electric propulsion. The RG-350 successfully managed the transition from the purely resistive load bank within the Stirling controller to the highly capacitive power-processing unit load. At the time of the demonstration, the Stirling convertor was operating at a hot temperature of 530 C and a cold temperature of -6 C. The linear alternator was producing approximately 250 W at 109 Vac, while the power-processing unit was drawing 175 W at 48 Vdc. The majority of power was delivered to the Hall thruster discharge circuit operating at 115 Vdc and 0.9 A. Testing planned for late 2001 will examine the possibility of directly driving the Hall thruster discharge circuit using rectified and filtered output from the Stirling alternator.

Mason, Lee S.↗
DOE OSTI2023

Time-Dependent Boundary Modeling to Inform Design of SPARC Diagnostic and Actuators

The purpose of this work was to inform the design of diagnostic and actuator systems for a new experimental facility for Commonwealth Fusion Systems (CFS), called SPARC [Creely2020], using time-dependent plasma boundary simulations with the Scrape Off Layer Plasma Simulator (SOLPS) code [Wiesen2015]. The SPARC tokamak is a device that aims to demonstrate fusion energy production and high-field plasma operating scenarios. The developed technologies will be incorporated into a net-electricity pilot that will demonstrate approximately 200MW of electric power. A major difference from other private industry fusion companies is the use of a standard aspect ratio, but at a high magnetic field strength, which provides a well-established physics basis and a modest extrapolation to reactor operation. Time-dependent simulations were used to inform the design of main-chamber plasma facing components, which are protected by feedback control of the heat and particle loads during high power operation. The simulations provide the timescales and magnitudes of plasma and neutral particle response useful for determining the position and operation of diagnostics and actuators that will enable control. The plasma response and phase-space diagrams also provide input into advanced model-based control schemes, which are envisioned for later SPARC operation. Model-based control can reduce the risk of component failure, which would require expensive in-situ repairs and the associated delays, by predicting the system response to actuators over a short time-horizon.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

Decomposition of high-rank factorized unitary coupled-cluster operators using ancilla and multiqubit controlled low-rank counterparts

The factorized form of the unitary coupled-cluster (UCC) approximation is one of the most promising methodologies to prepare trial states for strongly correlated systems within the variational quantum eigensolver framework. The factorized form of the UCC Ansatz can be systematically applied to a reference state to generate the desired entanglement. The difficulty associated with such an approach is the requirement of simultaneously entangling a growing number of qubits, which quickly exceeds the hardware limitations of today's quantum machines. In particular, while circuits for singles and double excitations can be performed on current hardware, higher-rank excitations require too many gate operations. In this work, we propose a set of schemes that trade off using extra qubits for a reduced gate depth to decompose high-rank UCC excitation operators into significantly lower depth circuits. These results will remain useful even when fault-tolerant machines are available to reduce the overall state-preparation circuit depth.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2021

Inverse Design of Plasma Metamaterial Devices for Optical Computing

We apply inverse-design methods to produce two-dimensional plasma metamaterial (PMM) devices. Backpropagated finite-difference frequency-domain (FDFD) simulations are used to design waveguides and demultiplexers operating under both TE and TM modes. Demultiplexing and waveguiding are demonstrated for devices composed of plasma elements with reasonable plasma densities of approximately 7 GHz, allowing for future in situ training and experimental realization of these designs. We also explore the possible applicability of PMMs to nonlinear boolean operations for use in optical computing. Functionally complete logical connectives (or and and) are achieved in the TM mode.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
NASA NTRS1990

Navigational utility of high-precision radio interferometer for Galileo's approach to Jupiter

The effect of very long baseline interferometry (VLBI) measurements of 2-nanoradian (nrad) accuracy has been studied for use in Galileo's approach to Jupiter's moon Io. Of particular interest is reducing the error in the minimum altitude above Io's surface. The nominal tracking strategy includes Doppler, range, and onboard optical data, in addition to VLBI data with 25-nrad accuracy. For nominal data, the altitude error is approximately 250 km with a data cutoff of 19 days before closest approach to Io. A limited number (two to four) of 2-nrad VLBI measurements, simulating a demonstration of improved VLBI data, were found to reduce the altitude error by 10 to 40 percent. Improving the accuracy of the VLBI measurements of the nominal tracking strategy to 2 nrads, to simulate the results from an operational few-nrad VLBI capability, was found to reduce the altitude error by an approximate factor of four. This reduction in altitude error is attributed to the ability that VLBI data give to help determine the along-track component of Jupiter's ephemeris. This capability complements the ability of the onboard optical data to determine the radial and cross-track components of Jupiter's ephemeris.

Folkner, W. M.↗
DOE OSTI2025

Coefficient-to-Basis Network: a fine-tunable operator learning framework for inverse problems with adaptive discretizations and theoretical guarantees

We propose a Coefficient-to-Basis Network (C2BNet), a novel framework for solving inverse problems within the operator learning paradigm. C2BNet efficiently adapts to different discretizations through fine-tuning, using a pre-trained model to significantly reduce computational cost while maintaining high accuracy. Unlike traditional approaches that require retraining from scratch for new discretizations, our method enables seamless adaptation without sacrificing predictive performance. Furthermore, we establish theoretical approximation and generalization error bounds for C2BNet by exploiting low-dimensional structures in the underlying datasets. Our analysis demonstrates that C2BNet adapts to low-dimensional structures without relying on explicit encoding mechanisms, highlighting its robustness and efficiency. To validate our theoretical findings, we conducted extensive numerical experiments that showcase the superior performance of C2BNet on several inverse problems. The results confirm that C2BNet effectively balances computational efficiency and accuracy, making it a promising tool to solve inverse problems in scientific computing and engineering applications.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Allocation and Operation of A Hydropneumatic Energy Storage with Building Microgrid

Behind-meter application for a recently invented hydropneumatic Ground-Level Integrated Diverse Energy Storage (GLIDES) is focused in this research. Considering possible uncertainties from solar radiation and electricity load in a building microgrid, two-stage stochastic programming is adopted for 15-min operation and a mixed-integer nonlinear stochastic model is built with integrated energy systems of GLIDES, solar panel and power generating unit. In order to make the model computationally tractable, we use sample average approximate method to decrease associated complexity with sample size and improve solving efficiency. The experimental results in deterministic and stochastic operation indicate that best candidate solution from SAA are superior to the solution from expected value problem in terms of optimality gap and variance.

24 POWER TRANSMISSION AND DISTRIBUTION↗
NASA NTRS1996

CFD Study of Turbo-Ramjet Interactions in Hypersonic Airbreathing Propulsion System

Advanced airbreathing propulsion systems used in Mach 4-6 mission scenarios, usually involve turbo-ramjet configurations. As the engines transition from turbojet to ramjet, there is an operational envelope where both engines operate simultaneously. In the first phase of our study, an over/under nozzle configuration was analyzed. The two plumes from the turbojet and ramjet interact at the end of a common 2-D cowl, where they both reach an approximate Mach 3.0 condition and then jointly expand to Mach 3.6 at the common nozzle exit plane. For the problem analyzed, the turbojet engine operates at a higher nozzle pressure ratio than the ramjet, causes the turbojet plume overpowers the ramjet plume, deflecting it approximately 12 degrees downward and in turn the turbojet plume is deflected 6 degrees upward. In the process, shocks were formed at the deflections and a shear layer formed at the confluence of the two jets. This particular case was experimentally tested and the data were used to compare with a computational fluid dynamics (CFD) study using the PARC2D code. The CFD results were in good agreement with both static pressure distributions on the cowl separator and on nozzle walls. The thrust coefficients were also in reasonable agreement. In addition, inviscid relationships were developed around the confluence point, where the two exhaust jets meet, and these results compared favorably with the CFD results. In the second phase of our study, a 3-D CFD solution was generated to compare with the 2-D solution. The major difference between the 2-D and 3-D solutions was the interaction of the shock waves, generated by the plume interactions, on the sidewall. When a shock wave interacts with a sidewall and sidewall boundary layer, it is called a glancing shock sidewall interaction. These interactions entrain boundary layer flow down the shockline into a vortical flow pattern. The 3-D plots show the streamlines being entrained down the shockline. The pressure of the flow also decreases slightly as the sidewall is approached. Other difference between the 2-D and 3-D solutions were a lowering of the nozzle thrust coefficient value from 0.9850 (2-D) to 0.9807 (3-D), where the experimental value was 0.9790. In the third phase of our study, a different turbo-ramjet configuration was analyzed. The confluence of a supersonic turbojet and a subsonic ramjet in the turbine based combined-cycle (TBCC) propulsion system was studied by a 2-D CFD code. In the analysis, Mach 1.4 primary turbojet was mixed with the subsonic ramjet secondary flow in an ejector mode operation. Reasonable agreements were obtained with the supplied I-D TBCC solutions. For low downstream backpressure, the Fabri choke condition (Break-Point condition) was observed in the secondary flow within mixing zone. For sufficient high downstream backpressure, the Fabri choke no longer exist, the ramjet flow was reduced and the ejector flow became backpressure dependent. Highly non-uniform flow at ejector exit were observed, indicated that for smooth downstream combustion, the mixing of the two streams probably required some physical devices.

Chang, Ing↗
NASA NTRS1993

Application of fuzzy logic-neural network based reinforcement learning to proximity and docking operations: Special approach/docking testcase results

As part of the RICIS project, the reinforcement learning techniques developed at Ames Research Center are being applied to proximity and docking operations using the Shuttle and Solar Maximum Mission (SMM) satellite simulation. In utilizing these fuzzy learning techniques, we use the Approximate Reasoning based Intelligent Control (ARIC) architecture, and so we use these two terms interchangeably to imply the same. This activity is carried out in the Software Technology Laboratory utilizing the Orbital Operations Simulator (OOS) and programming/testing support from other contractor personnel. This report is the final deliverable D4 in our milestones and project activity. It provides the test results for the special testcase of approach/docking scenario for the shuttle and SMM satellite. Based on our experience and analysis with the attitude and translational controllers, we have modified the basic configuration of the reinforcement learning algorithm in ARIC. The shuttle translational controller and its implementation in ARIC is described in our deliverable D3. In order to simulate the final approach and docking operations, we have set-up this special testcase as described in section 2. The ARIC performance results for these operations are discussed in section 3 and conclusions are provided in section 4 along with the summary for the project.

Jani, Yashvant↗
NASA NTRS1948

Preliminary Investigation of a Ceramic Lining for a Combustion Chamber for Gas-turbine Use

Combustion chamber liners for gas turbines and experimental set-ups were tested for failure. A ceramic-lined test chamber was operated at fuel/air ratios up to 0.050. Thermal-shock evaluation indicated that a ceramic lining, which expands after firing, would crack but would not fall apart during operation. Refractoriness of the lining and the resistance to mechanical shock were adequate. In general, shell temperature reductions of approximately 400 deg F were effected by the use of this lining at fuel/air ratios of 0.016 and 0.050. The mechanism of failure of the ceramic lining was induced by sudden heating and cooling during operation.

Woodward, William H↗
DOE OSTI2020

The HRA/Solarium Project: Feedback Based on 20 Years of Experience in Treatment of Medium Level Waste - 20185

Since the 1980's the management of the historical site of Mol has been transferred to the National Agency for Radioactive Waste and Enriched Fissile Materials (ONDRAF/NIRAS). Belgoprocess is entrusted by ONDRAF/NIRAS with the operational waste management and site remediation. One of the major challenges has been the characterization, treatment and conditioning of approximately 200 m{sup 3} of medium level waste. These waste packages were stored in poor conditions in storage vaults ('HRA') or concrete containers ('Solarium'). They have been produced in various research programs and reactor operations at the Belgian nuclear energy research centre SCK.CEN, isotope production, decontamination and dismantling operations from the 1960's up to the 1980's. Despite the limited volume, this historical waste consists of a great variation of waste characteristics and waste configurations. To tackle these liabilities, a new processing facility was built in the 1990's to allow safe transfers, handling, characterization and treatment of these packages. Also some auxiliary facilities have been built to deal with the by-products, like emptied concrete containers. The engineering of the installation, safety procedures and (characterization) methodology that has been developed generically have been proved to be successful. Nevertheless, the specific nature of some items made it necessary to organize a step-by-step treatment in distinct campaigns in which some relevant extra (safety) measures had to be taken into account. This dynamic approach made it possible to process a great variety of waste types in order to to ensure that this legacy is no longer left for subsequent generations. After about 20 years of operations, there has been a great build-up of experience and feed-back concerning the waste management of these specific waste streams. Some cases will be highlighted to specify the approach followed for treatment of e.g.: - Na/NaK containing equipment; - spent radioactive sources; - fuel element residues; - medium level waste originating from activation experiments; - radium bearing medium level waste originating from Ac-227 production research. (authors)

07 ISOTOPE AND RADIATION SOURCES↗
DOE OSTI2024

High-volume tunable resonator for axion searches above 7 GHz

In this research, we present results from an experimental demonstration of a tunable thin-shell axion haloscope whose geometry decouples its overall volume from its resonant frequency, thereby evading the steep sensitivity degradation at high frequencies. An aluminum 2.6-l (41⁢λ 3 ) prototype, which tunes from 7.1 to 8.0 GHz, was fabricated and characterized at room temperature. An axion-sensitive, straightforwardly tunable TM 010 mode is clearly identified with a room-temperature quality factor, Q, of approximately 5000. The on-resonance E-field distribution is mapped and found to agree with numerical calculations. Anticipating future cryogenic operation, we develop an alignment protocol relying only on rf measurements of the cavity, maintaining a form factor of 0.57 across the full tuning range. These measurements demonstrate the feasibility of cavity-based haloscopes with operating volume V $\gg$ λ 3 . We discuss plans for future development and the parameters required for a thin-shell haloscope exploring the postinflationary axion parameter space (approximately 4 to 30 GHz) at Dine-Fischler-Srednicki-Zhitnitsky sensitivity.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
NASA NTRS1998

Stability and Phase Noise Tests of Two Cryo-Cooled Sapphire Oscillators

A cryocooled Compensated Sapphire Oscillator (CSO), developed for the Cassini Ka-band Radio Science experiment, and operating in the 8K - 10K temperature range was previously demonstrated to show ultra-high stability of sigma(sub y) = 2.5 x 10 (exp -15) for measuring times 200 seconds less than or equal to tau less than or equal to 600 seconds using a hydrogen maser as reference. We present here test results for a second unit which allows CSO short-term stability and phase noise to be measured for the first time. Also included are design details of a new RF receiver and an intercomparison with the first CSO unit. Cryogenic oscillators operating below about 10K offer the highest possible short term stability of any frequency sources. However, their use has so far been restricted to research environments due to the limited operating periods associated with liquid helium consumption. The cryocooled CSO is being built in support of the Cassini Ka-band Radio Science experiment and is designed to operate continuously for periods of a year or more. Performance targets are a stability of 3-4 x 10 (exp -15) (1 second less than or equal to tau less than or equal to 100 seconds) and phase noise of -73dB/Hz @ 1Hz measured at 34 GHz. Installation in 5 stations of NASA's deep space network (DSN) is planned in the years 2000 - 2002. In the previous tests, actual stability of the CSO for measuring times tau less than or equal to 200 seconds could not be directly measured, being masked by short-term fluctuations of the H-maser reference. Excellent short-term performance, however, could be inferred by the success of an application of the CSO as local oscillator (L.O.) to the JPL LITS passive atomic standard, where medium-term stability showed no degradation due to L.O. instabilities at a level of (sigma)y = 3 x 10 (exp -14)/square root of tau. A second CSO has now been constructed, and all cryogenic aspects have been verified, including a resonator turn-over temperature of 7.907 K, and Q of 7.4 x 10 (exp 8). These values compare to a turn-over of 8.821 K and Q of 1.0 x 10 (exp 9) for the first resonator. Operation of this second unit provides a capability to directly verify for the first time the short-term (1 second less than or equal to tau less than or equal to 200 seconds) stability and the phase noise of the CSO units. The RF receiver used in earlier tests was sufficient to meet Cassini requirements for tau greater than or equal to 10 seconds but had short-term stability limited to 2-4 x 10 (exp -14) at tau = 1 second, a value 10 times too high to meet our requirements. A new low-noise receiver has been designed to provide approximately equal to 10-15 performance at 1 second, and one receiver is now operational, demonstrating again short-term CSO performance with H maser-limited stability. Short-term performance was degraded in the old receiver due to insufficient tuning bandwidth in a 100MHZ quartz VCO that was frequency-locked to the cryogenic sapphire resonator. The new receivers are designed for sufficient bandwidth, loop gain and low noise to achieve the required performance.

Dick, G. John↗
NASA NTRS2000

Performance of an Optimized Eta Model Code on the Cray T3E and a Network of PCs

In the year 2001, NASA will launch the satellite TRIANA that will be the first Earth observing mission to provide a continuous, full disk view of the sunlit Earth. As a part of the HPCC Program at NASA GSFC, we have started a project whose objectives are to develop and implement a 3D cloud data assimilation system, by combining TRIANA measurements with model simulation, and to produce accurate statistics of global cloud coverage as an important element of the Earth's climate. For simulation of the atmosphere within this project we are using the NCEP/NOAA operational Eta model. In order to compare TRIANA and the Eta model data on approximately the same grid without significant downscaling, the Eta model will be integrated at a resolution of about 15 km. The integration domain (from -70 to +70 deg in latitude and 150 deg in longitude) will cover most of the sunlit Earth disc and will continuously rotate around the globe following TRIANA. The cloud data assimilation is supposed to run and produce 3D clouds on a near real-time basis. Such a numerical setup and integration design is very ambitious and computationally demanding. Thus, though the Eta model code has been very carefully developed and its computational efficiency has been systematically polished during the years of operational implementation at NCEP, the current MPI version may still have problems with memory and efficiency for the TRIANA simulations. Within this work, we optimize a parallel version of the Eta model code on a Cray T3E and a network of PCs (theHIVE) in order to improve its overall efficiency. Our optimization procedure consists of introducing dynamically allocated arrays to reduce the size of static memory, and optimizing on a single processor by splitting loops to limit the number of streams. All the presented results are derived using an integration domain centered at the equator, with a size of 60 x 60 deg, and with horizontal resolutions of 1/2 and 1/3 deg, respectively. In accompanying charts we report the elapsed time, the speedup and the Mflops as a function of the number of processors for the non-optimized version of the code on the T3E and theHIVE. The large amount of communication required for model integration explains its poor performance on theHIVE. Our initial implementation of the dynamic memory allocation has contributed to about 12% reduction of memory but has introduced a 3% overhead in computing time. This overhead was removed by performing loop splitting in some of the high demanding subroutines. When the Eta code is fully optimized in order to meet the memory requirement for TRIANA simulations, a non-negligeable overhead may appear that may seriously affect the efficiency of the code. To alleviate this problem, we are considering implementation of a new algorithm for the horizontal advection that is computationally less expensive, and also a new approach for marching in time.

Kouatchou, Jules↗