A generic strategy for dynamic load balancing of distributed memory parallel computational mechanics using unstructured meshes

A large class of computational problems are characterised by frequent synchronisation, and computational requirements which change as a function of time. When such a problem is solved on a message passing multiprocessor machine [5], the combination of these characteristics leads to system performance which deteriorate in time. As the communication performance of parallel hardware steadily improves so load balance becomes a dominant factor in obtaining high parallel efficiency. Performance can be improved with periodic redistribution of computational load; however, redistribution can sometimes be very costly. We study the issue of deciding when to invoke a global load re-balancing mechanism. Such a decision policy must actively weigh the costs of remapping against the performance benefits, and should be general enough to apply automatically to a wide range of computations. This paper discusses a generic strategy for Dynamic Load Balancing (DLB) in unstructured mesh computational mechanics applications. The strategy is intended to handle varying levels of load changes throughout the run. The major issues involved in a generic dynamic load balancing scheme will be investigated together with techniques to automate the implementation of a dynamic load balancing mechanism within the Computer Aided Parallelisation Tools (CAPTools) environment, which is a semi-automatic tool for parallelisation of mesh based FORTRAN codes.

Load-balancing for parallel adaptive unstructured grids

A parallel method for the dynamic partitioning of unstructured meshes is outlined. The method includes diffusive load-balancing techniques and an iterative optimisation technique known as relative gain optimisationwhich both balances theworkload and attempts to minimise the interprocessor communications overhead. It can also optionally include amultilevel strategy. Experiments on a series of adaptively refined meshes indicate that the algorithmprovides partitions of an equivalent or higher quality to static partitioners (which do not reuse the existing partition) and much more rapidly. Perhaps more importantly, the algorithm results in only a small fraction of the amount of data migration compared to the static partitioners.

Parallel dynamic load-balancing for adaptive unstructured meshes

This chapter describes a parallel optimization technique that incorporates a distributed load-balancing algorithm and provides an extremely fast solution to the problem of load-balancing adaptive unstructured meshes. Moreover, a parallel graph contraction technique can be employed to enhance the partition quality and the resulting strategy outperforms or matches results from existing state-of-the-art static mesh partitioning algorithms. The strategy can also be applied to static partitioning problems. Dynamic procedures have been found to be much faster than static techniques, to provide partitions of similar or higher quality and, in comparison, involve the migration of a fraction of the data. The method employs a new iterative optimization technique that balances the workload and attempts to minimize the interprocessor communications overhead. Experiments on a series of adaptively refined meshes indicate that the algorithm provides partitions of an equivalent or higher quality to static partitioners (which do not reuse the existing partition) and much more quickly. The dynamic evolution of load has three major influences on possible partitioning techniques; cost, reuse, and parallelism. The unstructured mesh may be modified every few time-steps and so the load-balancing must have a low cost relative to that of the solution algorithm in between remeshing.

Dynamic load-balancing for parallel adaptive unstructured meshes

A parallel method for dynamic partitioning of unstructured meshes is described. The method employs a new iterative optimisation technique which both balances the workload and attempts to minimise the interprocessor communications overhead. Experiments on a series of adaptively refined meshes indicate that the algorithm provides partitions of an equivalent or higher quality to static partitioners (which do not reuse the existing partition) and much more quickly. Perhaps more importantly, the algorithm results in only a small fraction of the amount of data migration compared to the static partitioners.

Load balancing for aspect ratio in adaptive finite element simulations

In parallel adaptive finite element simulations the work load on the individual processors may change frequently. To (re)distribute the load evenly over the processors a load balancing heuristic is needed. Common strategies try to minimise subdomain dependencies by optimising the cutsize of the partitioning. However for certain solvers cutsize only plays a minor role, and their convergence is highly dependent on the subdomain shapes. Degenerated subdomain shapes cause them to need significantly more iterations to converge. In this work a new parallel load balancing strategy is introduced which directly addresses the problem of generating and conserving reasonably good subdomain shapes in a dynamically changing Finite Element Simulation. Geometric data is used to formulate several cost functions to rate elements in terms of their suitability to be migrated. The well known diffusive method which calculates the necessary load flow is enhanced by weighting the subdomain edges with the help of these cost functions. The proposed methods have been tested and results are presented.

Load balancing for parallel adaptive FEM

We present a dynamic distributed load balancing algorithm for parallel, adaptive finite element simulations using preconditioned conjugate gradient solvers based on domain-decomposition. The load balancer is designed to maintain good partition aspect ratios. It calculates a balancing flow using different versions of diffusion and a variant of breadth first search. Elements to be migrated are chosen according to a cost function aiming at the optimization of subdomain shapes. We show how to use information from the second step to guide the first. Experimental results using Bramble's preconditioner and comparisons to existing state-of-the-art balancers show the benefits of the construction.

Dynamic load balancing of distributed memory parallel computational mechanics using unstructured meshes for multi-physical modelling

As the complexity of parallel applications increase, the performance limitations resulting from computational load imbalance become dominant. Mapping the problem space to the processors in a parallel machine in a manner that balances the workload of each processors will typically reduce the run-time. In many cases the computation time required for a given calculation cannot be predetermined even at run-time and so static partition of the problem returns poor performance. For problems in which the computational load across the discretisation is dynamic and inhomogeneous, for example multi-physics problems involving fluid and solid mechanics with phase changes, the workload for a static subdomain will change over the course of a computation and cannot be estimated beforehand. For such applications the mapping of loads to process is required to change dynamically, at run-time in order to maintain reasonable efficiency. The issue of dynamic load balancing are examined in the context of PHYSICA, a three dimensional unstructured mesh multi-physics continuum mechanics computational modelling code.

Mesh partitioning and load balancing for distributed memory parallel systems

A method is outlined for optimising graph partitions which arise in mapping unstructured mesh calculations to parallel computers. The method employs a relative gain iterative technique to both evenly balance the workload and minimise the number and volume of interprocessor communications. A parallel graph reduction technique is also briefly described and can be used to give a global perspective to the optimisation. The algorithms work efficiently in parallel as well as sequentially and when combined with a fast direct partitioning technique (such as the Greedy algorithm) to give an initial partition, the resulting two-stage process proves itself to be both a powerful and flexible solution to the static graph-partitioning problem. Experiments indicate that the resulting parallel code can provide high quality partitions, independent of the initial partition, within a few seconds. The algorithms can also be used for dynamic load-balancing, reusing existing partitions and in this case the procedures are much faster than static techniques, provide partitions of similar or higher quality and, in comparison, involve the migration of a fraction of the data.

Automatic implementation of dynamic load balancing strategies for structured computational mechanics codes

This paper presents a new dynamic load balancing technique for structured mesh computational mechanics codes in which the processor partition range limits of just one of the partitioned dimensions uses non-coincidental limits, as opposed to using coincidental limits in all of the partitioned dimensions. The partition range limits are 'staggered', allowing greater flexibility in obtaining a balanced load distribution in comparison to when the limits are changed 'globally'. as the load increase/decrease on one processor no longer restricts the load decrease/increase on a neighbouring processor. The automatic implementation of this 'staggered' load balancing strategy within an existing parallel code is presented in this paper, along with some preliminary results.

Development of error correction techniques for nitrate-N load estimation models

Excess nutrient loads carried by streams and rivers are a great concern for environmental resource managers. In agricultural regions, excess loads are transported downstream to receiving water bodies, potentially causing algal blooms, which could lead to numerous ecological problems. To better understand nutrient load transport, and to develop appropriate water management plans, it is important to have accurate estimates of annual nutrient loads. This study used a Monte Carlo sub-sampling method and error-corrected statistical models to estimate annual nitrate-N loads from two watersheds in central Illinois. The performance of three load estimation methods (the seven-parameter log-linear model, the ratio estimator, and the flow-weighted averaging estimator) applied at one-, two-, four-, six-, and eight-week sampling frequencies were compared. Five error correction techniques; the existing composite method, and four new error correction techniques developed in this study; were applied to each combination of sampling frequency and load estimation method. On average, the most accurate error reduction technique, (proportional rectangular) resulted in 15% and 30% more accurate load estimates when compared to the most accurate uncorrected load estimation method (ratio estimator) for the two watersheds. Using error correction methods, it is possible to design more cost-effective monitoring plans by achieving the same load estimation accuracy with fewer observations. Finally, the optimum combinations of monitoring threshold and sampling frequency that minimizes the number of samples required to achieve specified levels of accuracy in load estimation were determined. For one- to three-weeks sampling frequencies, combined threshold/fixed-interval monitoring approaches produced the best outcomes, while fixed-interval-only approaches produced the most accurate results for four- to eight-weeks sampling frequencies.

Serviços dos ecossistemas prestados pelos estuários : especiação e capacidade de retenção do fósforo

O fator nutricional presente nos sapais não é só importante para os organismos que nele habitam como também apresenta uma importância ecológica a uma maior escala. Exemplifica-se este fato com a capacidade de retenção do Fósforo (P) por parte das plantas podendo estas competir como produto químicos para sua imobilização através da incorporação da sua fração biodisponível para cumprimento dos mecanismos básicos de sobrevivência. O rizosedimento dessas plantas é uma importante fonte de nutrientes necessários para garantia básica da perpetuação de todos os serviços (regulação, provisão, cultural) gerados pelo ambiente. A perda de áreas de sapal foi o motivo principal para a elaboração desta tese visto a preocupação em relação aos impactes que a perda de serviços associados à área de estudo escolhida, o sistema lagunar da Ria de Aveiro, poderá vir sofrer caso aconteçam alterações hídricas (naturais ou antrópicas) mais acentuadas no futuro. Essa tese assume como compromisso chamar a atenção para essa questão como também investigar os mecanismos do ciclo do P este importante nutriente considerado essencial para a manutenção da vida em nosso planeta. Para este estudo foram escolhidas 3 espécies de plantas halófitas representantes do sapal amostral, são elas: Bolbochenous maritimus, Spartina maritima e de Juncus maritimus. As questões levantadas buscam compreender o papel do P nas transformações que ocorrem no rizosedimento das halófitas em 3 diferentes contextos: (i) quando comparados os valores da biomassa de P associado às halófitas estudadas (estudo realizado nos sapais povoados por Spartina maritima e Juncus maritimus) em toda extensão da Ria de Aveiro (ii) quando da disponibilidade de P perante a presença de poluentes inorgânicos no Largo do Laranjo, que corresponde a uma zona historicamente contaminada por metais (estudo realizado no rizosedimento das espécies Bolbochenous maritimus e Juncus maritimus) e (iii) quanto à intrusão superficial da água salgada da laguna (estudo focado na avaliação da perda do P próximo às raízes das halófitas que habitam a zona externa do dique do Baixo Vouga Lagunar (BVL) que corresponde à zona de confluência do Rio Vouga com a Ria de Aveiro). Por meio da análise do rizosedimento e da biomassa das diferentes halófitas predominantes do sapal foi possível verificar que, em relação à carga nutritiva fosfática, de Norte (Canal de São Jacinto/Ovar) a Sul (Canal de Mira) da Ria de Aveiro, a cota do sapal é mais importante do que a natureza espacial do sistema. Esta informação é primordial para a assertividade de futuras medidas de criação e reabilitação das áreas de sapal na Ria de Aveiro. Os estudos realizados nas áreas impactadas (Largo do Laranjo e BVL) incidiram na análise do perfil vertical rizosedimentar das halófitas pertencentes ao sapal médio-alto. Estes estudos revelaram que o rizosedimento dessas halófitas possui características peculiares que definem a dinâmica do ciclo do P de forma bastante característica. O rizosedimento das halófitas presentes no Largo do Laranjo apresentaram teores biodisponíveis de P equivalentes àqueles presentes no rizosedimento de área mais afastada da fonte pontual de contaminação para as mesmas espécies. Numa visão otimista, os resultados revelam que ambos sapais, Largo de Laranjo e Cais do Bico, apresentam condições equiparável para perpetuação dos serviços por eles gerados, que nesse caso se tornam ainda mais valiosos devido a garantia da fitoestabilização dos contaminantes (prevenção da entrada de contaminantes na coluna d’água e na cadeia alimentar). No Baixo Vouga Lagunar a análise espacial do sapal na zona exterior ao dique revelou a perda da diversidade das espécies bem como o aumento da área de vasa e consequentemente a perda dos teores de P associados ao rizosedimento das halófitas mais expostas aos efeitos da compressão costeira intensificados pelas ações naturais e antrópicas a que a Ria está sujeita. Em geral, a degradação dessa área de sapal devido aos fenômenos de assoreamento ou de erosão não pode ser separada dos processos sedimentares na área envolvente, principalmente no caso da Ria de Aveiro que tem sofrido constantes mudanças para atender às necessidades das atividades humanas. Essa tendência provavelmente não mudará no futuro próximo. Consequentemente, a evolução da laguna será principalmente dependente do resultado direto das ações humanas que deverão sempre encontrar formas de compensar os danos causados no âmbito de intervenções de reabilitação do sistema hídrico. A recriação de áreas de sapal pode ter lugar como medida mitigadora no âmbito de intervenções de reabilitação de áreas degradadas e ainda poderá servir como forma de captura de P para programas de utilização de fertilizantes naturais (componente desejável para a agricultura moderna o que é particularmente importante para as regiões altamente dependentes do mercado de importação). Deste modo, recomenda-se que a recuperação das áreas de sapal seja enquadrada no âmbito de intervenções mais amplas de valorização dos sistemas estuarinos e lagunares.

Influence of Cobalt on the Properties of Load-Sensitive Magnesium Alloys

In this study, magnesium is alloyed with varying amounts of the ferromagnetic alloying element cobalt in order to obtain lightweight load-sensitive materials with sensory properties which allow an online-monitoring of mechanical forces applied to components made from Mg-Co alloys. An optimized casting process with the use of extruded Mg-Co powder rods is utilized which enables the production of magnetic magnesium alloys with a reproducible Co concentration. The efficiency of the casting process is confirmed by SEM analyses. Microstructures and Co-rich precipitations of various Mg-Co alloys are investigated by means of EDS and XRD analyses. The Mg-Co alloys' mechanical strengths are determined by tensile tests. Magnetic properties of the Mg-Co sensor alloys depending on the cobalt content and the acting mechanical load are measured utilizing the harmonic analysis of eddy-current signals. Within the scope of this work, the influence of the element cobalt on magnesium is investigated in detail and an optimal cobalt concentration is defined based on the performed examinations.

The study of the sedimentation mechanism in channels under the impact of tidal currents

The purpose of this research is to study sedimentation mechanism by mathematical modeling in access channels which are affected by tidal currents. The most important factor for recognizing sedimentation process in every water environment is the flow pattern of that environment. It is noteworthy that the flow pattern is affected by the geometry and the shape of the environment as well as the type of existing affects in area. The area under the study in this thesis is located in Bushehr Gulf and the access channels (inner and outer). The study utilizes the hydrodynamic modeling with unstructured triangular and non-overlapping grids, using the finite volume, From method analysis in two scale sizes: large scale (200 m to 7.5km) and small scale (50m to 7.5km) in two different time durations of 15 days and 3.5 days to obtain the flow patterns. The 2D governing equations used in the model are the Depth-Averaged Shallow Water Equations. Turbulence Modeling is required to calculate the Eddy Viscosity Coefficient using the Smagorinsky Model with coefficient of 0.3. In addition to the flow modeling in two different scales and the use of the data of 3.5 day tidal current modeling have been considered to study the effects of the sediments equilibrium in the area and the channels. This model is capable of covering the area which is being settled and eroded and to identify the effects of tidal current of these processes. The required data of the above mentioned models such as current and sediments data have been obtained by the measurements in Bushehr Gulf and the access channels which was one of the PSO's (Port and Shipping Organization) project-titled, "The Sedimentation Modeling in Bushehr Port" in 1379. Hydrographic data have been obtained from Admiralty maps (2003) and Cartography Organization (1378, 1379). The results of the modeling includes: cross shore currents in northern and north western coasts of Bushehr Gulf during the neap tide and also the same current in northern and north eastern coasts of the Gulf during the spring tide. These currents wash and carry fine particles (silt, clay, and mud) from the coastal bed of which are generally made of mud and clay with some silts. In this regard, the role of sediments in the islands of this area and the islands made of depot of dredged sediments should not be ignored. The result of using 3.5 day modeling is that the cross channels currents leads to settlement places in inner and outer channels in tidal period. In neap tide the current enters the channel from upside bend of the two channels and outer channel. Then it crosses the channel oblique in some places of the outer channel. Also the oblique currents or even almost perpendicular current from up slope of inner channel between No. 15 and No. 18 buoys interact between the parallel currents in the channel and made secondary oblique currents which exit as a down-slope current in the channel and causes deposit of sediments as well as settling the suspended sediments carried by these currents. In addition in outer channel the speed of parallel currents in the bend of the channel which is naturally deeper increases. Therefore, it leads to erosion and suspension of sediments in this area. The speed of suspended sediments carried by this current which is parallel to the channel axis decreases when they pass through the shallower part of the channel where it is in the buoys No.7 and 8 to 5 and 6 are located. Therefore, the suspended sediment settles and because of this process these places will be even shallower. Furthermore, the passing of oblique upstream leads to settlement of the sediments in the up-slope and has an additional effect on the process of decreasing the depth of these locations. On the contrary, in the down-slope channel, as the results of sediments and current modeling indicates the speed of current increases and the currents make the particles of down-slope channel suspended and be carried away. Thus, in a vast area of downstream of both channels, the sediments have settled. At the end of the neap tide, the process along with circulations in this area produces eddies which causes sedimentation in the area. During spring some parts of this active location for sedimentation will enter both channels in a reverse process. The above mentioned processes and the places of sedimentation and erosion in inner and outer channels are validated by the sediments equilibrium modeling. This model will be able to estimate the suspended, bed load and the boundary layer thickness in each point of both channels and in the modeled area.

THERMAL VACUUM TEST CORRELATION OF A ZERO PROPELLANT LOAD CASE THERMAL CAPACITANCE PROPELLANT GAUGING ANALYTICAL MODEL

This thesis describes the development and correlation of a thermal model that forms the foundation of a thermal capacitance spacecraft propellant load estimator. Specific details of creating the thermal model for the diaphragm propellant tank used on NASA’s Magnetospheric Multiscale spacecraft using ANSYS and the correlation process implemented are presented. The thermal model was correlated to within +/- 3 Celsius of the thermal vacuum test data, and was determined sufficient to make future propellant predictions on MMS. The model was also found to be relatively sensitive to uncertainties in applied heat flux and mass knowledge of the tank. More work is needed to improve temperature predictions in the upper hemisphere of the propellant tank where predictions were found to be 2-2.5 Celsius lower than the test data. A road map for applying the model to predict propellant loads on the actual MMS spacecraft in 2017-2018 is also presented.