UK Consortium on Mesoscale Engineering Sciences (UKCOMES)
Lead Research Organisation:
University College London
Department Name: Mechanical Engineering
Abstract
The consortium aims to advance and lead in mesoscale science and engineering that are crucial to solving emerging societal challenges such as the net zero energy system, high-end manufacturing, healthcare and digital economy. This will be achieved by developing and exploiting cutting-edge mesoscopic modelling and simulation techniques with the aid of HEC (ARCHER2) and tier-2 GPU (Bede) systems. The UKCOMES community of academics, researchers, collaborators and end-users, already the largest and best in the world, will be consolidated and expanded to benefit the wider community and generate greater impact. Community codes as well as in-house codes will be further developed, disseminated and applied, using the best practices. The consortium, in working with CoSeC, will provide a stimulating, collaborative and interdisciplinary environment to train people in optimised use of current HEC and in preparation for the forthcoming exascale platforms, in order to conduct world-leading research and application.
Mesoscales refer to those in between atomistic and macroscales. Such scales exist in almost all physical, chemical, biological, biomedical, material, pharmaceutical and engineering phenomena and processes. Mesoscales bridge atomistic and macroscales, and thus span many orders of magnitude. To resolve mesoscales is a great computational challenge, which requires ever more powerful HEC platforms. Unsurprisingly, mesoscale modelling and simulation has grown in capability and popularity in tandem with the development of HEC. In particular, the lattice Boltzmann method (LBM) has had a phenomenal growth in recent decades. Other methods under study that share the philosophy and aim of mesoscale modelling and simulation include dissipative particle dynamics (DPD), smoothed particle hydrodynamics (SPH), discrete velocity method (DVM), direct simulation Monte Carlo (DSMC), kinetic Monte Carlo (kMC), and coarse-grained MD (CGMD). Due to the pervasiveness and complexity of mesoscopic problems, the research and end-user communities working in the field are diverse and multidisciplinary. The consortium not only acts as a focal point for the diverse communities but also allows efficient utilisation of HEC resources through coordination and training.
The remit of UKCOMES covers both simulation-methodology-orientated developments and application-driven research using HEC and tier-2 platforms. The work of the consortium will be pursued in the following work packages (WPs): (1) Community Codes Development, Optimisation & Dissemination; (2) Simulation & Optimisation of Net Zero Energy Systems; (3) Mesoscale Simulation & Design in Advanced Manufacturing; (4) Simulation & Application of Multiphase & Interfacial Flows; (5) Hemodynamics Simulation & Application in Healthcare; (6) VVUQ, Machine Learning & Data Analytics; (7) Engagement, Outreach, Dissemination and Impact Delivery.
The Management Committee (MC) of UKCOMES consists of the PI and WP leaders. Strategic inputs are provided by the Scientific Advisory Board (SAB) and Industrial Advisory Board (IAB) as well as CoSeC and EPCC representatives. The MC sets the scientific agenda, reviews progresses, allocates computing resources, advises on software development and releases, organises training of people and plans for impact delivery. A set of transparent criteria will be applied to allocation of computing resources on ARCHER2 and Bede. Best practices will be formulated and disseminated within the consortium regarding how to port, benchmark, optimise and run codes. Throughout the project, the membership of UKCOMES will stay open to anyone who has interest in or can benefit from mesoscale modelling and simulation. The consortium will work with UK HEC consortia, tier-2, CCPs and ExCALIBUR communities to strengthen the UK base in CSE and build software as an important UK infrastructure. The consortium is also committed to international and industrial engagement.
Mesoscales refer to those in between atomistic and macroscales. Such scales exist in almost all physical, chemical, biological, biomedical, material, pharmaceutical and engineering phenomena and processes. Mesoscales bridge atomistic and macroscales, and thus span many orders of magnitude. To resolve mesoscales is a great computational challenge, which requires ever more powerful HEC platforms. Unsurprisingly, mesoscale modelling and simulation has grown in capability and popularity in tandem with the development of HEC. In particular, the lattice Boltzmann method (LBM) has had a phenomenal growth in recent decades. Other methods under study that share the philosophy and aim of mesoscale modelling and simulation include dissipative particle dynamics (DPD), smoothed particle hydrodynamics (SPH), discrete velocity method (DVM), direct simulation Monte Carlo (DSMC), kinetic Monte Carlo (kMC), and coarse-grained MD (CGMD). Due to the pervasiveness and complexity of mesoscopic problems, the research and end-user communities working in the field are diverse and multidisciplinary. The consortium not only acts as a focal point for the diverse communities but also allows efficient utilisation of HEC resources through coordination and training.
The remit of UKCOMES covers both simulation-methodology-orientated developments and application-driven research using HEC and tier-2 platforms. The work of the consortium will be pursued in the following work packages (WPs): (1) Community Codes Development, Optimisation & Dissemination; (2) Simulation & Optimisation of Net Zero Energy Systems; (3) Mesoscale Simulation & Design in Advanced Manufacturing; (4) Simulation & Application of Multiphase & Interfacial Flows; (5) Hemodynamics Simulation & Application in Healthcare; (6) VVUQ, Machine Learning & Data Analytics; (7) Engagement, Outreach, Dissemination and Impact Delivery.
The Management Committee (MC) of UKCOMES consists of the PI and WP leaders. Strategic inputs are provided by the Scientific Advisory Board (SAB) and Industrial Advisory Board (IAB) as well as CoSeC and EPCC representatives. The MC sets the scientific agenda, reviews progresses, allocates computing resources, advises on software development and releases, organises training of people and plans for impact delivery. A set of transparent criteria will be applied to allocation of computing resources on ARCHER2 and Bede. Best practices will be formulated and disseminated within the consortium regarding how to port, benchmark, optimise and run codes. Throughout the project, the membership of UKCOMES will stay open to anyone who has interest in or can benefit from mesoscale modelling and simulation. The consortium will work with UK HEC consortia, tier-2, CCPs and ExCALIBUR communities to strengthen the UK base in CSE and build software as an important UK infrastructure. The consortium is also committed to international and industrial engagement.
Organisations
Publications
Bai Z
(2023)
Theoretical exploration on the performance of single and dual-atom Cu catalysts on the CO2 electroreduction process: a DFT study.
in Physical chemistry chemical physics : PCCP
Bai Z
(2023)
Understanding mechanisms of pyridine oxidation with ozone addition via reactive force field molecular dynamics simulations
in Chemical Engineering Science
Bai Z
(2023)
Impact of oxygen and nitrogen-containing species on performance of NO removal by coal pyrolysis gas
in Process Safety and Environmental Protection
Bai Z
(2023)
Effects of nitrogen-free species on NO removal performance by coal pyrolysis gas via reactive molecular dynamics simulations
in Journal of the Energy Institute
Bhagurkar A
(2023)
The relationship between electric processing condition and microstructure in the solidification of multicomponent oxides
in Ceramics International
Bhati AP
(2023)
Long Time Scale Ensemble Methods in Molecular Dynamics: Ligand-Protein Interactions and Allostery in SARS-CoV-2 Targets.
in Journal of chemical theory and computation
Chen J
(2023)
Pore-scale study of miscible density instability with viscosity contrast in porous media
in Physics of Fluids
Chen S
(2023)
Natural convection of large Prandtl number fluids: A controversy answered by a new thermal lattice Boltzmann model
in Case Studies in Thermal Engineering
Fei L
(2023)
Coupled lattice Boltzmann method-discrete element method model for gas-liquid-solid interaction problems
in Journal of Fluid Mechanics
| Description | Influence on National Engineering Policies |
| Geographic Reach | National |
| Policy Influence Type | Contribution to a national consultation/review |
| Title | CCDC 2255439: Experimental Crystal Structure Determination |
| Description | Related Article: Yuchong Yang, Tanya K. Ronson, Dingyu Hou, Jieyu Zheng, Ilma Jahovic´, Kai Hong Luo, Jonathan R. Nitschke|2023|J.Am.Chem.Soc.|145|19164|doi:10.1021/jacs.3c04228 |
| Type Of Material | Database/Collection of data |
| Year Produced | 2023 |
| Provided To Others? | Yes |
| URL | http://www.ccdc.cam.ac.uk/services/structure_request?id=doi:10.5517/ccdc.csd.cc2fpz37&sid=DataCite |
| Title | Development and performance of a HemeLB GPU code for human-scale blood flow simulation |
| Description | In recent years, it has become increasingly common for high performance computers (HPC) to possess some level of heterogeneous architecture - typically in the form of GPU accelerators. In some machines these are isolated within a dedicated partition, whilst in others they are integral to all compute nodes - often with multiple GPUs per node - and provide the majority of a machine's compute performance. In light of this trend, it is becoming essential that codes deployed on HPC are updated to execute on accelerator hardware. In this paper we introduce a GPU implementation of the 3D blood flow simulation code HemeLB that has been developed using CUDA C++. We demonstrate how taking advantage of NVIDIA GPU hardware can achieve significant performance improvements compared to the equivalent CPU only code on which it has been built whilst retaining the excellent strong scaling characteristics that have been repeatedly demonstrated by the CPU version. With HPC positioned on the brink of the exascale era, we use HemeLB as a motivation to provide a discussion on some of the challenges that many users will face when deploying their own applications on upcoming exascale machines. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2022 |
| Provided To Others? | Yes |
| Impact | An updated version is now published as: DOI: 10.1016/j.cpc.2022.108548 |
| URL | https://data.mendeley.com/datasets/jhdv4drxbx |
| Title | DL_MESO |
| Description | DL_MESO is UKCOMES's flagship open-source software for worldwide distribution: a general purpose, highly scalable mesoscale simulation package for both Lattice Boltzmann Method (LBM) and Dissipative Particle Dynamics (DPD) methods. It is written in C++ for LBM and Fortran90 for DPD. It is supplied with its own Java-based Graphical User Interface (GUI) and is capable of both serial and parallel execution using MPI and/or OpenMP. |
| Type Of Technology | Software |
| Year Produced | 2013 |
| Open Source License? | Yes |
| Impact | DL_MESO has been upgraded significantly through UKCOMES activities: (1) incorporation of the latest physical models; (2) assessment and validation of physical models and numerical methods in DL_MESO; (3) adaptation to new software and hardware environments. Briefly, work by UKCOMES is to ensure the accuracy, robustness and reliability of DL_MESO so that it stays at the cutting edge of mesoscale simulation. Since 2013, around 600 new licences of DL_MESO have been taken worldwide: about 31% are based in the UK, 16% mainland Europe, 18% China, 8% North America and 27% in the rest of the world. |
| URL | http://www.ukcomes.org/codes |
| Title | Processing Tools for HemeLB Simulations |
| Description | This repository contains the pre-processing and post-processing tools used in the simulations described in the referenced paper. Here FiveExit and ProfundaFemoris2 refer to the five-outlets model and the profunda femoris model described in the paper, respectively. The simulation cases studied are listed in the file experiments.csv in the corresponding sub-folders. For example, FiveExit_coarse refers to the simulations using the coarse grid and the five-outlets model. The simulations described in the paper were performed using HemePure, a HemeLB version with improved memory, compilation and scaling. The input files for these simulations were generated by the Python script writeInput.py. This script imports MyModules/InputOutput.py and reads experiments.csv and input_VfWKf.xml in the corresponding sub-folders, where the constant parameters for each simulation domain are provided in const_param.csv. To use this script, execute the command "python writeInput.py" with Python of version 3.6 or above. The outputs of HemeLB were first processed using paraviewPreprocess.sh to obtain human-readable outputs. These outputs were analysed using verification.py. The figures in the paper and its supplementary figures were generated using analysis.py. |
| Type Of Technology | Software |
| Year Produced | 2022 |
| Impact | The GPU version of HemeLB has been published in 2023, as DOI: 10.1016/j.cpc.2022.108548 |
| URL | https://rdr.ucl.ac.uk/articles/software/Processing_Tools_for_HemeLB_Simulations/21229733 |