• Cole, Ivan
• Bab-Hadiashar, Alireza
• Jun Toh, Rou
• Patel, Milan
• Asadi, Ehsan
• Awan, Sana
• Lomo, Felix N.
• Ye, Jiayu
• Gautam, Subash
• Hassan, Ul
• Alam, Nazmul
• Jose, J.

Abstract

Additive manufacturing has some significant advantages (low raw material waste, low net energy use, high design flexibility) that make it highly advantageous for specialised small-run production (i.e. ‘high-value, low-volume manufacturing), such as for medical and aerospace applications. However, three key factors limit its application in these and other manufacturing sectors: the low production output, the limited size of integral components that can be built, and uncertainties (e.g. incidence and spatial distribution of defects) in the builds. In this context, the term ‘defects’ is used broadly and could include loss of shape control, undesirable microstructures, voids and cracks, or residual stresses. The use of robotically controlled high-powered lasers in out-of-chamber directed energy deposition (DED) can allow for increased production rate and is also not limited by chamber size. However, such methods may also increase defects relative to smaller-scale production methods. Thus, in both established additive manufacturing methods, it is necessary to effectively monitor, control, and report any defects in a build, which may be achieved by several approaches:1) A battery of controlled experiments to understand the interactions between materials, and operational parameters, including tool paths and defects2) In-situ monitoring of operational parameters, shape, and melt-zone and thermal effects to predict the initiation of defects (with and without the use of machine learning) to develop and anchor new feedback control loops targeted at minimising defect growth3) Development of ‘virtual objects’ that represent both sensor data acquired during a build and its process parameter values, both to be used as an assurance of quality and for the automatic identification of manufacturing-related defects4) Characterisation of the forms of defects that occur in additive manufacturing and the consequence of different forms for a build’s performance, and finally, linking such performance prediction data into the aforementioned virtual object5) Iterative or evolutionary design based on computational materials and thermal and heat transfer modelling to digitally design additively manufactured forms with optimum performanceThis paper will summarise key research projects in the authors’ research teams addressing these five issues for laser-based deposition processes and cold spray additive manufacturing. In doing so, it will outline new approaches that can link digital models of additive form structure and manufacturing with the properties and performance of built objects. This will not only enhance both traditional and additive manufacturing processes but will also assist the use of newer and faster out-of-chamber methods.

Topics

• Deposition
• impedance spectroscopy
• microstructure
• experiment
• melt
• crack
• void
• directed energy deposition
• machine learning