EVALUATION OF THE INTERRELATION BETWEEN KEY PROCESS PARAMETERS AND PRODUCT/MATERIAL PROPERTIES IN HYBRID ADDITIVE-SUBTRACTIVE PROCESS CHAINS

EPSRC产品组合领域：增材制造、后处理、数据驱动建模、表面完整性、机械性能增材制造（AM）为制造商提供了令人兴奋的新机会，使以前传统制造工艺无法实现的设计得以实现。包括航空航天、运输和生物医学工程在内的许多行业都对增材制造产生了兴趣，在这些行业中，增材制造具有提高系统效率、促进减轻重量、提高生物相容性和定制化水平的潜力。尽管对这一领域的兴趣日益浓厚，但目前在预测使用AM生产的物体的特性方面存在知识差距。虽然通过传统工艺制造的零件的性能可以准确预测，但关键增材制造工艺参数与最终产品/材料性能之间的相互关系尚未完全实现，导致目前大量的实验测试要求。本项目旨在通过实验和计算，建立增材制造参数和后处理条件与在混合增减法工艺链中生产的成品零件的表面、机械和材料性能之间的相关性，从而提高这种理解。该项目将分三个主要阶段进行：第一阶段将使用实验测试制度评估关键AM（基于激光的粉末床熔化）参数对感兴趣的输出数量（qoi）的影响，包括表面粗糙度/质量、残余应力、拉伸性能、疲劳、蠕变、腐蚀、显微硬度和成品零件的微观结构。随后将根据增材制造零件的功能要求进行后处理（激光纹理/抛光、机械微加工、热处理等），以确定这些工艺如何进一步影响组件的性能。该项目的最后阶段将涉及使用随机代理建模技术对增材制造过程进行数据驱动的分层建模，并对测量的qoi进行后处理参数。这将呈现在高维参数空间中创建qoi响应面，从而实现增材制造和后处理条件的鲁棒优化逆设计，以获得设计者指定的qoi值/分布。贝叶斯推理框架将被用来执行稳健的逆设计。这项研究将导致引入具有集成功能的组件，用于广泛的工业应用，如航空航天，汽车，铁路，电子和生物医学领域。

EPSRC Portfolio Areas: Additive manufacturing, post-processing, data driven modelling, surface integrity, mechanical propertiesAdditive manufacturing (AM) presents exciting new opportunities for manufacturers, enabling the realisation of designs which have previously been impossible to achieve with traditional manufacturing processes. Interest has developed in numerous industries, including aerospace, transportation, and biomedical engineering, where AM has the potential to improve system efficiencies, foster weight reduction and improve the biocompatibility and levels of customisation available.Despite the growing interest in this field, there is currently a knowledge gap in predicting the properties of objects produced using AM. Whereas the properties of a part manufactured through traditional processes can be accurately predicted, the interrelation between the key AM process parameters and the final product/material properties is yet to be fully realised, leading to extensive experimental testing requirements at present.This project aims to improve this understanding by establishing the correlation between AM manufacturing parameters, and post-processing conditions, with the surface, mechanical and material properties of finished parts produced in a hybrid additive-subtractive process chain, both experimentally and computationally.The project will be conducted in three major phases:The first phase will deal with evaluating the effects of key AM (laser-based powder bed fusion) parameters on the output quantities of interest (qoi) which include surface roughness/quality, residual stress, tensile properties, fatigue, creep, corrosion, microhardness and microstructure of the as-built parts, using an experimental test regime.This will be followed by post-processing (laser texturing/polishing, mechanical micromachining, heat treatment etc.) of the AM parts depending on their functional requirements to determine how such processes further affect the properties of the component.The final phase of the project will involve a data-driven hierarchical modelling of the AM process and post-processing parameters to the measured qoi using a stochastic surrogate modelling techniques. This will render the creation of a qoi response surface in a high dimensional parameter space, thus enabling the robust optimal inverse design of the AM and post-processing conditions to obtain designer-specified values/distributions of the qoi. A Bayesian inference framework would be utilised to perform the robust inverse design.The research will lead to the introduction of components with integrated functionalities for a broad range of industrial applications, such as aerospace, automotive, railway, electronic and biomedical sectors.