A Multiscale Simulation Approach to Tackle Fuel Spray Atomisation and Combustion

作为产生全球80%电力的技术，化石燃料的燃烧将在未来几十年继续在能源生产中发挥关键作用，并对碳排放做出重大贡献。在用于公路、空中和水上运输的移动式内燃机以及用于发电站发电的固定式燃气轮机中，化石燃料的燃烧广泛通过液体燃料喷射来实现，这决定了燃料效率和排放。它涉及一系列复杂的多尺度、多物理场、多相现象，已被确定为这些燃烧装置的基础研究需求。随着越来越多的替代生物燃料进入燃料市场，这一需求变得越来越迫切，这使得燃料的特性变得更加复杂，液体燃料喷射的控制也变得更加困难。因此，未来的智能发动机需要精确控制各种燃料的喷射，这比目前所能实现的要微妙得多。目前对燃油喷射过程的数值研究可分为两大类：(1)用欧拉方法模拟浓喷状态下的初次破裂；(2)用拉格朗日方法模拟稀喷状态下的湍流-燃烧-液滴相互作用。然而，系统地研究燃油喷雾、雾化和燃烧是两种方法都无法实现的。由于小液滴和细韧带的分辨率要求，使用欧拉方法跟踪大量雾化液滴是困难的。对于拉格朗日方法，广泛使用的计算配置是均匀各向同性/剪切湍流，或时间/空间发展的混合层或满载点源液滴的射流。由于缺少由密集喷雾区初次破碎决定的雾化喷雾的重要初始条件，喷雾燃烧的研究目前还处于理论初级阶段，离指导和优化燃油喷射设计的预期目标还很遥远。本项目提出了一种想法，通过保持两种方法的优点并互补弥补彼此的缺点，弥合两种方法之间的差距来模拟燃油喷雾雾化和燃烧。集成的、多尺度的、欧拉-拉格朗日混合模拟方法可用于在当前和未来的超级计算机上执行燃料喷雾雾化和燃烧现象的高保真模拟，并研究雾化液体-燃料喷射、雾化蒸发液滴、燃烧和湍流之间复杂的多边相互作用。开发这样一个预测数值工具是实现设计和优化燃油效率和清洁发动机的完整、预测模拟能力目标的重要的第一步。它可以广泛影响包括非公路发动机在内的运输发动机的设计，并有助于加速燃料市场中使用的各种生物燃料，为英国生物能源的关键新兴产业做出贡献。液体喷雾工艺也广泛应用于其他研究学科，如医疗技术和先进制造。预测数值工具有助于提高这些优先研究领域中喷雾过程的科学理解、设计和控制。

As the technology to generate the world's 80% power, combustion of fossil fuels will continue to play a key role in energy production over the next several decades and contributes heavily to carbon emission. In mobile internal combustion engines for road, air and water transportation and stationary gas turbines for electricity generation in power stations, the burning of fossil fuels is widely achieved by liquid fuel injection, which dictates fuel efficiency and emissions. It involves a cascade of complex multiscale, multiphysics, multiphase phenomena, and has been identified as a basic research need for these combustion devices. The need is becoming more urgent as more alternative biofuels are introduced in the fuel market, making the fuel properties more complex and the control of liquid fuel injection more difficult. Therefore, future smart engines require precise control of the injection of a broad variety of fuels that is far more subtle than what can be achieved to date.Currently numerical research on the fuel spray process can be divided in two principal categories: (1) using an Eulerian approach to simulate primary breakup in the dense spray regime and (2) using a Lagrangian approach to simulate turbulence-combustion-droplets interaction in the dilute spray regime. However, to systematically study fuel spray atomisation and combustion cannot be achieved by either approach. To track a large amount of atomised droplets using an Eulerian approach is difficult due to the resolution requirement for small droplets and thin ligaments. For Lagrangian approaches, the widely used computational configurations are homogeneous isotropic/shear turbulence, or temporally/spatially developing mixing layers or jets laden with point-source droplets. Missing important initial conditions of the atomising spray which are determined by primary breakup in the dense spray zone, the research on spray combustion is currently in an early theoretical stage and far from the expected goal of guiding and optimising the design of fuel injection.This project proposes an idea to bridge the gap between the two approaches to simulate fuel spray atomisation and combustion, by keeping the advantages of the two approaches and complementarily remedying their disadvantages with each other. The integrated, multiscale, hybrid Eulerian-Lagrangian simulation approach can be used to perform high-fidelity simulation of the fuel spray atomisation and combustion phenomena and investigate complex multilateral interactions among an atomising liquid-fuel jet, atomised evaporating droplets, combustion, and turbulence on current and future supercomputers.Developing such a predictive numerical tool is an essential first step toward the goal of a complete, predictive simulation capability for the design and optimisation of fuel-efficient and clean engines. It can impact broadly the design of transportation engines including off-highway engines and help the acceleration of diverse biofuels being used in the fuel market, contributing to key emerging industries of bioenergy in the UK. A liquid spray process is also widely used in other research disciplines such as Healthcare Technologies and Advanced Manufacturing. A predictive numerical tool can contribute to improving the scientific understanding, design and control of the spray processes in these prioritised research areas.