Publications
129 results found
Lamb AR, Gorman GJ, Elsworth D, 2013, A fracture mapping and extended finite element scheme for coupled deformation and fluid flow in fractured porous media, INTERNATIONAL JOURNAL FOR NUMERICAL AND ANALYTICAL METHODS IN GEOMECHANICS, Vol: 37, Pages: 2916-2936, ISSN: 0363-9061
- Author Web Link
- Cite
- Citations: 31
Lange M, Gorman G, Weiland M, et al., 2013, Benchmarking mixed-mode PETSc performance on high-performance architectures
The trend towards highly parallel multi-processing is ubiquitous in allmodern computer architectures, ranging from handheld devices to large-scale HPCsystems; yet many applications are struggling to fully utilise the multiplelevels of parallelism exposed in modern high-performance platforms. In order torealise the full potential of recent hardware advances, a mixed-mode betweenshared-memory programming techniques and inter-node message passing can beadopted which provides high-levels of parallelism with minimal overheads. Forscientific applications this entails that not only the simulation code itself,but the whole software stack needs to evolve. In this paper, we evaluate themixed-mode performance of PETSc, a widely used scientific library for thescalable solution of partial differential equations. We describe the additionof OpenMP threaded functionality to the library, focusing on sparsematrix-vector multiplication. We highlight key challenges in achieving goodparallel performance, such as explicit communication overlap using task-basedparallelism, and show how to further improve performance by explicitly loadbalancing threads within MPI processes. Using a set of matrices extracted fromFluidity, a CFD application code which uses the library as its linear solverengine, we then benchmark the parallel performance of mixed-mode PETSc acrossmultiple nodes on several modern HPC architectures. We evaluate the parallelscalability on Uniform Memory Access (UMA) systems, such as the FujitsuPRIMEHPC FX10 and IBM BlueGene/Q, as well as a Non-Uniform Memory Access (NUMA)Cray XE6 platform. A detailed comparison is performed which highlights thecharacteristics of each particular architecture, before demonstrating efficientstrong scalability of sparse matrix-vector multiplication with significantspeedups over the pure-MPI mode.
Buchan AG, Pain CC, Tollit TS, et al., 2013, Simulated spatially dependent transient kinetics analysis of the Oak Ridge Y12 plant criticality excursion, Progress in Nuclear Energy, Vol: 63, Pages: 12-21
In June 1958 an accidental nuclear excursion occurred in the C-1 Wing of building 9212 in a process facility designed to recover enriched Uranium U(93) from various solid wastes. The accident was caused by the inadvertent flow of enriched uranyl nitrate into a 55 gallon drum which established a prompt critical nuclear excursion. Following the initial fission spike the nuclear system oscillated in power. The reaction was eventually terminated by the additional water which was flowing into the drum. The criticality excursion was estimated to have lasted approximately 20 min based upon nearby radiation measurement equipment with an estimated total fission yield of 1.3 × 1018 fissions of which the first fission spike contributed 6 × 1016 fissions.The traces from the radiation measurement devices indicated that most of the fissions occurred in the first 2.8 min, in which case the average power required for the observed fission yield was approximately 220 kW. After the first 2.8 min the system was postulated to have boiled causing a sharp decrease in density and reactivity of the system. This boiling probably reduced the power output from the system to a low level for the final 18 min of the excursion. This paper will aim to investigate the subsequent evolution of the Y12 excursion using the fundamentally based spatially dependent neutron/multiphase CFD kinetics simulation tool - FETCH. The reconstruction of the Y12 excursion using FETCH will follow the evolution of the excursion up until the uranyl nitrate starts to boil. The results of the FETCH simulation are presented and compared against the known measurements of the excursion from the radiation detection instruments located near the drum.
Guo X, Gorman G, Lange M, et al., 2013, Exploring the Thread-level Parallelisms for the Next Generation Geophysical Fluid Modelling Framework Fluidity-ICOM, Procedia Engineering, Vol: 61, Pages: 251 - 257-251 - 257, ISSN: 1877-7058
Lange M, Gorman G, Weiland M, et al., 2013, Acieving efficient strong scaling with PETSc using hybrid MPI/OpenMP optimisations, Publisher: Springer Berlin Heidelberg, Pages: 97-108
Milthaler FFM, Gorman GJ, Piggott MD, 2012, Reducing spurious drag forces when using mesh adaptivity in CFD, ECCOMAS 2012 - European Congress on Computational Methods in Applied Sciences and Engineering, e-Book Full Papers, Pages: 5621-5640
This work explores the impact of mesh adaptivity methods in combination with fixed as well as adaptive timestepping when modelling fluid dynamical systems that are sensitive to minor changes in the fluid's pressure and velocity. Here the diagnostic of interest for fluidsolid interaction modelling is the drag force. Depending on the solid's properties, even minor unphysical abrupt changes, so-called peaks, in the drag force - due to mesh adaptivity - could lead to a major disturbance in the model. For such systems the need naturally arises to reduce these peaks to a certain degree, until the sudden changes are small enough to be neglected. Hence, in this paper a variety of approaches are described and compared against one another, that aim to reduce these peaks. Moreover, further studies show the relation between the peaks in the drag force to the timestep, and pressure. The 3D-CFD software Fluidity, which uses an arbitrarily unstructured mesh, and a 3D mesh optimization algorithm was used for this case-study.
Buchan A, Eaton MD, Goddard AJH, et al., 2012, Simulated transient dynamics and heat transfer characteristics of the water boiler nuclear reactor SUPO with cooling coil heat extraction, Annals of nuclear energy, Vol: 48, Pages: 68-83
The term “water boiler” reactor refers to a type of aqueous homogeneous reactor (AHR) that was designed, built and operated by Los Alamos in the 1940s. This was the first type of liquid fuelled reactor and the first to be fuelled with enriched Uranium. For security reasons the term “water boiler” was adopted and three versions were built: LOPO (for low power), HYPO (for high power) and SUPO (for super power) which were spherical shaped reactor vessels. The name was appropriate as the reactors appeared to boil although this was actually due to the release of radiolytic gas bubbles; although SUPO was operated during some studies close to the boiling point of uranyl nitrate. The final water boiler “SUPO” was operated almost daily as a neutron source from 1951 until its deactivation in 1974-23 years of safe, reliable operation. Many of the key neutron measurements needed in the design of the early atomic weapons were made using LOPO, HYPO and SUPO. More recently SUPO has been considered as a benchmark for quasi-steady-state operation of AHRs with internal cooling structures.This paper presents modelling and analysis of the coupled neutronic and fluid time dependent characteristics of the SUPO reactor. In particular the quasi-steady-state dynamics of SUPO have been investigated together with its heat transfer characteristics. In the simulations presented the SUPO reactor is modelled using the spatially dependent neutron/multiphase CFD simulation tool, FETCH, at a quasi-steady-state power of 25 kW. SUPO also possessed a cooling coil system that fed cooling water through the reactor for the extraction of the fission and decay heat. This cooling system, and the heat extraction, is modelled in the simulations using a new sub-modelling approach that is detailed here. The results from this simulation, such as gas fraction, gas generation rate, coolant rate and average temperature, are compared against the available experimental information.
Nygaard ET, Pain CC, Eaton MD, et al., 2012, Steps Towards Verification and Validation of the FETCH Code for Level 2 Analysis, Design and Optimization of Aqueous Homogeneous Reactors, PHYSOR
Gorman GJ, Southern J, Farrell PE, et al., 2012, Hybrid OpenMP/MPI anisotropic mesh smoothing, International Conference on Computational Science (ICCS), Publisher: ELSEVIER SCIENCE BV, Pages: 1513-1522, ISSN: 1877-0509
- Author Web Link
- Cite
- Citations: 13
Weiland M, Mitchell L, Gorman G, et al., 2012, Mixed-mode implementation of PETSc for scalable linear algebra on multi-core processors
Gorman GJ, Southern J, Farrell PE, et al., 2012, Hybrid OpenMP/MPI Anisotropic Mesh Smoothing, Procedia Computer Science, Vol: 9, Pages: 1513-1522
Mesh smoothing is an important algorithm for the improvement of element quality in unstructured mesh finite element methods. A new optimisation based mesh smoothing algorithm is presented for anisotropic mesh adaptivity. It is shown that this smoothing kernel is very effective at raising the minimum local quality of the mesh. A number of strategies are employed to reduce the algorithm's cost while maintaining its effectiveness in improving overall mesh quality. The method is parallelised using hybrid OpenMP/MPI programming methods, and graph colouring to identify independent sets. Different approaches are explored to achieve good scaling performance within a shared memory compute node.
Southern J, Gorman GJ, Piggott MD, et al., 2011, Parallel anisotropic mesh adaptivity with dynamic load balancing for cardiac electrophysiology, Journal of Computational Science, Vol: 3, Pages: 8-16
Rokos G, Gorman G, Kelly PHJ, 2011, Accelerating Anisotropic Mesh Adaptivity on nVIDIA's CUDA Using Texture Interpolation, 17th International Euro-Par Conference on Parallel Processing, Publisher: SPRINGER-VERLAG BERLIN, Pages: 387-398, ISSN: 0302-9743
- Author Web Link
- Cite
- Citations: 1
Farrell PE, Piggott MD, Gorman GJ, et al., 2011, Automated continuous verification for numerical simulation, Geoscientific Model Development, Vol: 4, Pages: 435-449
Fang F, Pain CC, Navon IM, et al., 2011, The independent set perturbation adjoint method: A new method of differentiating mesh-based fluids models, International Journal for Numerical Methods in Fluids, Vol: In review
Mitchell AJ, Allison PA, Gorman GJ, et al., 2011, Tidal circulation in an ancient epicontinental sea: The Early Jurassic Laurasian Seaway, Geology, Vol: 39, Pages: 207-210, ISSN: 0091-7613
Pavlidis D, Aristodemou E, Gomes JLMA, et al., 2010, Numerical Simulation of Air Flows in Street Canyons Using Mesh-Adaptive LES, Direct and large-eddy simulation VII, Editors: Armenio, Geurts, Fröhlich, Publisher: Springer Verlag, ISBN: 9789048136513
This book is of interest to scientists and engineers both at an early level in their career as well as at more senior levels.
Lamb A, Gorman G, Gosselin O, et al., 2010, Coupled Deformation and Fluid Flow in Fractured Porous Media using Dual Permeability and Explicitly Defined Fracture Geometry, 72nd EAGE Conference & Exhibition
We present a numerical scheme which combines the dual permeability method (DPM) and the extended finite element method (XFEM) to simulate coupled deformation and fluid flow in fractured porous media. The scheme incorporates spatial variability in fracture properties without requiring the fracture to be discretized or aligned with the computational mesh. DPM is used to describe the fluid flow interaction between the porous matrix and fractures, whilst XFEM is used to address the discontinuous displacement field within elements which intersect fractures. The method is strongly coupled and solves the stress and flow equations simultaneously for each time increment.
Southern J, Gorman GJ, Piggott MD, et al., 2010, Simulating cardiac electrophysiology using anisotropic mesh adaptivity, Journal of Computational Science, Vol: 1, Pages: 82-88
The simulation of cardiac electrophysiology requires small time steps and a fine mesh in order to resolve very sharp, but highly localized, wavefronts. The use of very high resolution meshes containing large numbers of nodes results in a high computational cost, both in terms of CPU hours and memory footprint. In this paper an anisotropic mesh adaptivity technique is implemented in the Chaste physiological simulation library in order to reduce the mesh resolution away from the depolarization front. Adapting the mesh results in a reduction in the number of degrees of freedom of the system to be solved by an order of magnitude during propagation and 2–3 orders of magnitude in the subsequent plateau phase. As a result, a computational speedup by a factor of between 5 and 12 has been obtained with no loss of accuracy, both in a slab-like geometry and for a realistic heart mesh with a spatial resolution of 0.125 mm.
Wells MR, Allison PA, Piggott MD, et al., 2010, Tidal modeling of an ancient tide-dominated seaway, part 2: the Aptian Lower Greensand seaway of northwest Europe, Journal of Sedimentary Research, Vol: 80, Pages: 411-439, ISSN: 1527-1404
Wells MR, Allison PA, Piggott MD, et al., 2010, Tidal modeling of an ancient tide-dominated seaway, part 1: model validation and application to global early Cretaceous (Aptian) tides, Journal of Sedimentary Research, Vol: 80, Pages: 393-410, ISSN: 1527-1404
Mitchell AJ, Allison PA, Piggott MD, et al., 2010, Numerical modelling of tsunami propagation with implications for sedimentation in ancient epicontinental seas: the Lower Jurassic Laurasian Seaway, Journal of Sedimentary Geology, Vol: 228, Pages: 81-97, ISSN: 0037-0738
Garcia X, Pavlidis D, Gorman GJ, et al., 2010, A two-phase adaptive finite element method for solid–fluidcoupling in complex geometries, International Journal for Numerical Methods in Fluids
In this paper we present a method to solve the Navier–Stokes equations in complex geometries, suchas porous sands, using a finite-element solver but without the complexity of meshing the porous space.The method is based on treating the solid boundaries as a second fluid and solving a set of equationssimilar to those used for multi-fluid flow. When combined with anisotropic mesh adaptivity, it is possibleto resolve complex geometries starting with an arbitrary coarse mesh. The approach is validated bycomparing simulation results with available data in three test cases. In the first we simulate the flow pasta cylinder. The second test case compares the pressure drop in flow through random packs of sphereswith the Ergun equation. In the last case simulation results are compared with experimental data on theflow past a simplified vehicle model (Ahmed body) at high Reynolds number using large-eddy simulation(LES). Results are in good agreement with all three reference models.
Mitchell AJ, Ulicny D, Hampson GJ, et al., 2010, Modelling tidal current-induced bed shear stress and palaeocirculation in an epicontinental seaway: the Bohemian Cretaceous Basin, Central Europe, Sedimentology, Vol: 57, Pages: 359-388, ISSN: 0037-0746
Guo X, Gorman G, Ashworth M, et al., 2010, High Performance Computing Driven Software Development for Next-Generation Modelling of the Worlds Oceans, Cray Users Group Conference
The Imperial College Ocean Model (ICOM) is an open-source next generation ocean model build upon finite element methods and anisotropic unstructured adaptive meshing. Since 2009, a project has been funded by EPSRC to optimise the ICOM for the UK national high-end computing resource, HECToR (Cray XT4). Extensive use of profiling tools such as CrayPAT and Vampir, has been made in order to understand performance issues of the code on the Cray XT4. Of particular interest is the scalability of the sparse linear solvers and the algebraic multigrid preconditioners required to solve the system of equations.
Pavlidis D, Gorman GJ, Gomes JLMA, et al., 2010, Synthetic-Eddy Method for Urban Atmospheric Flow Modelling, Boundary-Layer Meteorology, Vol: 136, Pages: 285-299
Pavlidis D, Gomes JLMA, Gorman GJ, et al., 2010, Novel Mesh Adaptive LES Simulations for Multi-Scale Atmospheric Flows: Application to the Urban Environment, 30th NATO/SPS International Technical Meeting on Air Pollution Modeling and Its Applications, Publisher: SPRINGER, Pages: 63-+, ISSN: 1871-465X
- Author Web Link
- Cite
- Citations: 1
Farrell PE, Piggott MD, Gorman GJ, et al., 2010, Automated continuous verification and validation for numerical simulation, Geoscientific Model Development Discussions, Vol: 3, Pages: 1587-1623
Southern J, Gorman GJ, Piggott MD, et al., 2010, Anisotropic mesh adaptivity for cardiac electrophysiology, Pages: 929-938-929-938
Lamb AR, Gorman GJ, Gosselin OR, et al., 2010, Finite element coupled deformation and fluid flow in fractured porous media, 72nd European Association of Geoscientists and Engineers Conference and Exhibition 2010: A New Spring for Geoscience. Incorporating SPE EUROPEC 2010, Vol: 2, Pages: 1146-1161
We present a finite element scheme which combines the dual permeability method (DPM) and the extended finite element method (XFEM) to simulate coupled deformation and fluid flow in fractured porous media. The scheme incorporates spatial variability in fracture properties without requiring the fracture to be discretized or aligned with the computational mesh. DPM is used to describe the fluid flow interaction between the porous matrix and fractures, whilst XFEM is used to address the discontinuous displacement field within elements which intersect fractures. The method is strongly coupled and solves the stress and flow equations simultaneously for each time increment. DPM-XFEM uses the level set method (LSM) to define existing fractures and eliminates the need for their explicit discretization during simulation. The method performs well on coarse structured grids and does not require complex, difficult to generate meshes to conduct simulations. Comparison between the proposed method and the discrete fracture method (DFM) shows its ability to adequately determine the displacement and fluid pore pressure distribution within a fractured domain. © 2010, European Association of Geoscientists and Engineers.
This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.