Al-driven automotive material selection and structural design for manufacturing

人工智能驱动的汽车制造材料选择和结构设计

• 批准号: 10083425
• 金额: $ 26.06万
• 依托单位: MULTI-X SOLUTIONS LIMITED
• 依托单位国家: 英国
• 项目类别: Collaborative R&D
• 财政年份: 2023
• 资助国家: 英国
• 起止时间: 2023 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=10083425
• 关键词: Al; driven; automotive; material; selection

The automotive industry and road transport contribute significantly to the UK's net greenhouse emissions, with one-third attributable to this sector. Lightweight vehicle design plays a crucial role in reducing CO2 emissions, necessitating the intelligent use of materials and advancements in vehicle component manufacturing technologies. Component stamping, representing approximately 11% of total vehicle cost, is a critical area for innovation.Accurate simulation and prediction of stamping-induced defects in the sheet metal stamping industry rely on material formability data. While simulations offer cost and time savings by utilising material formability data, their complexity and limited accessibility to designers create challenges. Integrating Artificial Intelligence (AI) can enhance accuracy by capturing material forming behaviours, but the scarcity of high-fidelity material testing data needed for training remains an obstacle.Multi-X has developed leading material formability testing solutions for reproducing real manufacturing conditions, such as hot stamping. These tests have enabled the curation of high-fidelity datasets of formability properties for structural materials used in various manufacturing processes and conditions. This project aims to leverage Multi-X's datasets to train and enhance the cutting-edge AI models developed by Dr Li and her Advanced Manufacturing Group at Imperial College London to extend their applicability to efficiently and accurately predicting vehicle component manufacturability under real conditions. By harnessing the high-fidelity data from Multi-X's testing, the AI models can ensure the optimal design of lightweight components within safe strain limits of metals thereby mitigating undesirable failures efficiently.This project will revolutionise vehicle manufacturing by introducing a game-changing, user-accessible design tool, offered through the Software-as-a-Service (SaaS) delivery mode. This tool will optimise complex-shaped component design, select lightweight materials, reduce development time, and lower costs. With accurate AI-based manufacturability evaluations replacing traditional simulations, the project will contribute to reducing CO2 emissions through the creation of efficient and lightweight component designs.

汽车工业和道路运输对英国的净温室气体排放量贡献很大，其中三分之一可归因于该部门。轻量化车辆设计在减少二氧化碳排放方面起着至关重要的作用，需要智能使用材料和改进车辆部件制造技术。零部件冲压约占汽车总成本的11%，是创新的关键领域。在钣金冲压行业中，精确模拟和预测冲压引起的缺陷依赖于材料的可成形性数据。虽然模拟通过利用材料可成形性数据节省了成本和时间，但其复杂性和对设计师的有限可访问性带来了挑战。集成人工智能（AI）可以通过捕捉材料成形行为来提高精度，但训练所需的高保真材料测试数据的稀缺仍然是一个障碍。Multi-X开发了领先的材料成形性测试解决方案，用于再现真实的制造条件，例如热冲压。这些测试使在各种制造工艺和条件下使用的结构材料的可成形性的高保真数据集的管理成为可能。该项目旨在利用Multi-X的数据集来训练和增强由李博士和她在伦敦帝国理工学院的先进制造小组开发的尖端人工智能模型，以扩展其适用性，从而在真实的条件下有效准确地预测车辆部件的可制造性。通过利用Multi-X测试的高保真数据，人工智能模型可以确保在金属的安全应变范围内优化轻量化部件的设计，从而有效地减少不良故障。该项目将通过引入改变游戏规则的用户可访问的设计工具，通过软件即服务（SaaS）交付模式提供，从而彻底改变汽车制造业。该工具将优化复杂形状的部件设计，选择轻质材料，减少开发时间并降低成本。通过准确的基于人工智能的可制造性评估取代传统的模拟，该项目将通过创建高效和轻量化的组件设计来减少二氧化碳排放。