Manycore Algorithms for Batch Scalar and Block Tridiagonal Solvers

A Scalable Infrastructure for Online Performance Analysis on CFD Application

This paper gives a discussion on the Computational Fluid Dynamics (CFD) applications in offshore engineering. Accurate hydrodynamic quantities are essential for engineering design. Offshore structures are generally subject to high Reynolds number flows. These high Reynolds number flow conditions (Re > 106 ) are difficult and expensive to achieve in an experimental setup. Therefore, it is attractive to use CFD to provide the essential hydrodynamic quantities for practical design. Verification and validation studies are important for determining the validity of the CFD prediction. A procedure of performing CFD simulation is shown. Different types of turbulence modelling are discussed. Three examples of high Reynolds number CFD simulations, covering flow around offshore structural components and waves past partially-submerged horizontal cylinders, are shown and discussed.

<div class="htmlview paragraph">Development of Caterpillar Fuel Systems' MEUI-B injector has involved application of Computational Fluid Dynamics (CFD) in order to improve performance of the high pressure spill valve. Initial performance bench testing with concept stage experimental injectors indicated that the chamber pressure was decaying at an unacceptably slow rate, and the valve demonstrated erratic behavior at some operating conditions. The slow pressure decay and inconsistent spill valve motion were believed to be caused by flow forces generated during the low lift portion of the spill valve opening event. This theory was pursued by utilizing CFD to design two valves for testing in the next phase of the injector development cycle: A baseline geometry, similar to the original concept injector valve, and a new design incorporating localized seat geometry changes for inducing flow force assisted valve opening. Bench test performance data clearly shows that the modified design significantly improves the spill valve opening event and subsequent chamber pressure decay rate. This paper documents the CFD analysis which led to the development of the new MEUI-B high pressure spill valve design.</div>

The application of computational fluid dynamics(CFD) in the food industry such as drying, thermal sterilization, mixing, refrigeration and humidification of cold storage was reviewed. The results from previous studies have shown that CFD was a powerful numerical tool that is applied to model fluid flow situations and aid in the optimal design of engineering equipment and food process. With the development of computer technology, it is conceivable that CFD will continue to provide more explanations for physical modeling of fluid flow and process system design for the food industry.

Cart3D is a computational fluid dynamics (CFD) application aimed at conceptual and preliminary design of aerospace vehicles with complex geometries. It is widely used by design engineers at NASA, Department of Defense and aerospace companies in the USA. We present detailed performance analysis of Cart3D using two tools SGI MPInside and op_scope that collects hardware counter data from Intel Performance Monitoring Unit (PMU) on supercomputers based on Nehalem micro-architecture. Using these tools, we have done dynamic profiling of Cart3D (compute time, communication time and I/O time), along with dynamic profiling of MPI functions (MPI_Sendrecv, MPI_Bcast, MPI_Isend, MPI_Irecv, MPI_Allreduce, MPI_Barrier, etc.) with respect to message size of each rank and time consumed by each function. MPI communication is further analyzed by studying the performance of MPI functions used in this application as a function of message size and number of cores. Using these tools we have also studied efficiency of the processor to measure its effective utilization, efficiency of the floating-point units, percentage of vectorization and percentage of data coming from L2 cache, L3 cache, and main memory. This study was performed on two computing sub-systems based on quad-core Nehalem-EP and hex-core West mere-EP processors that are part of Pleiades an SGI Altix ICE at NASA Ames Research Center.

Computational fluid dynamics (CFD) techniques have become an important tool for investigating and predicting flow behavior in many industrial and engineering processes. In the last two decades, CFD has been used for the study, design, and improvement of irrigation equipment. Numerical simulations can be used to predict fluid flow, heat transfer, and chemical reactions within complex systems. The objective of this review is to provide an overview of the uses of CFD in irrigation engineering applications. The paper is organized into two main sections: fundamentals of CFD and irrigation engineering applications. The first section presents the main methods used in numerical simulations, basic equations to predict fluid flow parameters, meshing concerns, and convergence criteria. In the second section, we present applications related to friction and local head losses in pipes, liquid and solid-liquid flow simulation in drippers, chemical scaling, filters, sprinklers, direct-acting pressure-regulating valves, and Venturi injectors. The briefly described applications indicated that CFD modeling can be an accurate, quick, and feasible method for the investigation of flow parameters in irrigation pipes, fittings, emitters, and accessories. The CFD simulations can be useful for designing new products as well as for improving and optimizing existing products. Computational fluid dynamics uses in irrigation engineering must be encouraged, particularly for innovation purposes resulting from the cooperation between academia and irrigation companies.

Perforate-Wash-Cement (P/W/C) is a well plugging technique extensively used by ConocoPhillips in the Greater Ekofisk Area over the last ten years. In Jet-type P/W/C, Tubing Conveyed Perforating (TCP) guns perforate casing in the target interval before the zone is washed and a cross-sectional cement plug is installed using a specialized Bottom Hole Assembly (BHA). The goal is to maximize the cement plug quality through optimizing the BHA and operational process parameters within constraints imposed by the operating conditions. This paper describes application of Computational Fluid Dynamics (CFD) to achieve this objective. CFD is well suited for modeling wash and cement processes including the associated non-Newtonian fluids. A CFD model of these processes employs unsteady multiphase Reynolds Averaged Navier Stokes (RANS) based Volume of Fluid (VOF) approach with a Shear Stress Transport (SST) K-omega turbulence model. BHA translation and rotation are simulated using moving deforming-layering mesh with interface approach. Physical properties of non-Newtonian fluids selected are based on in-situ conditions and derived via lab tests. CFD-specific considerations such as domain size, turbulence model, mesh type and size, and computational timestep are discussed in this paper. The magnitude and duration of the jet-induced "pressure pulse" in the annular space between the formation wall and the casing are a key to efficient displacement during the wash and cementing processes. Displacement efficiency, in terms of the percentage of mud or cement in a control volume as a function of time, depends on perforation size and density, nozzle size/number, flow rate, fluid properties, as the ROP/BHA pulling speed and tool RPM. Key findings from a parametric study to optimize these parameters are presented. To validate CFD nozzle flow predictions and to rule out potential for false results due to cavitation, laboratory tests were performed under pressurized conditions. A comparison of results from CFD simulations using traditional versus optimized parameters are presented, demonstrating significant efficiency improvement achieved. The study initially focused on North Sea application. Encouraged by substantial improvement in process efficiency through parametric optimization, the technique was extended to UK operations and is planned for the Bayu-Undan field. Cementing simulation results for Bayu-Undan specific well configurations are presented, showing how CFD modeling helped reveal less than ideal cementing efficiency. Ultimately, this work provides significant time savings and quality improvement for P&A projects while maintaining a high safety standard.

Most of us are not familiar with the concept that an internal combustion (IC) engine is working on a four-stroke six-event principle. The six events are suction, compression, combustion, expansion, blow-down and exhaust. However, the expansion and blow-down happen in the power stroke and they can be clubbed together. Therefore, we can say that an engine, whether diesel or gasoline, is working on four-stroke five-event principle. The purpose of this chapter is to make the reader to familiarize with the complexities involved in the working of a four-stroke engine. The five events which are completed in four strokes are: suction, compression, combustion, expansion and exhaust. Application of computational fluid dynamics (CFD) principles for each process mentioned above is a challenging job. The difficulty in understanding the working of an IC engine is due to the fact that we cannot see what is happening inside the cylinder piston arrangement. All that is described in textbooks is based on our knowledge gained over a period of time by conducting experiments. There is no doubt in saying that “seeing is believing”. As it is next to impossible to have a complete experimental flow visualization, nowadays CFD comes in handy to have theoretical flow visualization. Well-developed software is available for the simulation in 3D geometries. In this chapter, we explain step-by-step the details required for the CFD simulation of various processes in an IC engine. Extensive results obtained over a period of 20 years of research by the application of CFD in analysing the flow in engines are comprehensively presented and discussed. CFD can be very well applied for analysing any particular process. It can also be used for the modification of the existing engine design or can also be employed for a new design of an engine. It is hoped that readers may be benefitted in understanding the application of CFD for fluid flow analysis and engine design by reading this chapter. Therefore, the main aim of this chapter is to make the reader appreciate how exactly CFD can be applied for design of an engine. As it is application oriented, we are not going deep into the equations, modelling, etc. A number of case studies are presented and discussed.

RoseNNa: A performant, portable library for neural network inference with application to computational fluid dynamics

In the industrial application of Computational Fluid Dynamics (CFD), the quality of simulations and the level of reliability of the results have emerged as major issues of concern. This paper describes a recent initiative of ERCOFTAC (European Research Community on Flow Turbulence and Combustion), involving European code developers, code vendors, academics and industrial users, to set up a practical set of standard guidelines to improve the quality of CFD simulations and to help in the assessment of the trustworthiness of the CFD results. The initiative led to the publication of the first edition of the ERCOFTAC Best Practice Guidelines (BPG) in January 2000. The BPG provide generic advice on how to perform quality CFD calculations. Some examples of typical guidelines from the BPG are provided in this paper.

Preconditioned Krylov subspace methods for solving nonsymmetric matrices from CFD applications

Computational fluid dynamics (CFD), alone or coupled with the most advanced imaging tools, allows for the assessment of blood flow patterns in cardiovascular disease to both understand their pathophysiology and anticipate the results of their surgical or interventional repair. CFD is a mathematical technique that characterizes fluid flow using the laws of physics. Through the utilization of specific software and numerical procedures based on virtual simulation and/or patient data from computed tomography, resonance imaging, and 3D/4D ultrasound, models of circulation for most CHDs can be reconstructed. CFD can provide insight into the pathophysiology of coronary artery anomalies, interatrial shunts, coarctation of the aorta and bicuspid aortic valve, tetralogy of Fallot and univentricular heart, valvular heart disease, and aortic disease. In some cases, CFD may be able to simulate different types of surgical or interventional repairs, allowing for the tailoring of treatment accordingly.

This paper looks at the application of Computational Fluid Dynamics (CFD) and integral approaches to the study of effects of obstacles on pollutant dispersion from a point source placed within an idealised urban area (MUST). This study is part of a modelling exercise within the COST Action 732. Numerical results are compared with wind tunnel data. We use the CFD code FLUENT and the dispersion model ADMS-Urban. The CFD model predicts concentrations more accurately than the integral model. However, both models results satisfy accepted statistical criteria, showing that those criteria should not be the only way of evaluating a model.

Rotor broadband noise can be predicted using various numerical and analytical methods. This paper explores the application of computational fluid dynamics (CFD) combined with Lee's wall pressure spectrum and Amiet's trailing edge noise model to predict broadband noise, comparing the results to different experiments. A simple, ideally twisted rotor blade with the NACA 0012 airfoil in hover is employed to establish the appropriate methodology for integrating CFD with other models. Two approaches are utilized: the first is a direct input approach, where flow data from the CFD solution is directly input into the models, while the second approach employs CFD to inform XFOIL inputs, which subsequently provide inputs for the wall pressure spectrum and acoustic model. The CFD/XFOIL approach demonstrates better agreement with experimental measurements and is less sensitive to CFD pre- and post-processing setups. The Straight Up Imagining (SUI) Endurance rotor is also validated for forward flight, followed by the prediction of broadband noise generated by a propeller and wing configuration. Observed trends during the validations and predictions suggest that airfoil selection and blade geometry play a more significant role in the magnitude of trailing edge noise than thrust.

This paper presents an application of Computational Fluid Dynamics (CFD) technique for designing of protective fences for shelter from snow and winds commonly used on offshore rigs and other facilities in cold regions. The CFD simulation of real situations provided detailed information about the flow characteristics behind the fence and the effective shelter zone. Such flow details are usually very expensive, time consuming and difficult to obtain by physical experiments. The numerical simulations found that the shape of pores and surface shear stress has significant impact on the reattachment length and the size of effective fence zone. The effective shelter distance can be increased by avoiding sharp angular corners in the porous pores and reducing surface shear stress. The 3D modelling and CFD simulation provide detailed aerodynamics flow characteristics and stress distributions that allow evaluating a number of designs cost effectively. The technique can be integrated with small-scale intelligent manufacturing system (SIMS) for flexible design and manufacturing of fences.