Shah, Paras (2019) Optimisation and application of non-destructive metrology techniques to improve the quality, accuracy, and traceability of additively manufactured components. Doctoral thesis, University of Huddersfield.

The absence of repeatability and inadequate methods of precision verification of Additively Manufactured (AM) parts both externally and internally is a critical technical hurdle that discourages manufacturers from implementing such technologies. High-performance aerospace, automotive and medical industries have stringent requirements where failure from non-conformance and metallurgical defects could be catastrophic. Furthermore, post-processing and traditional Coordinate Measurement Systems (CMS) methods cause bottlenecks in the supply chain. Bespoke AM designs have the flexibility, allowing for Geometric Tolerancing and Dimensioning (GT&T) features to sophisticated lattice features used for weight reduction and strain relief, to match the speed of AM; this necessitates for adaptive metrology solutions which are essential for acceptable verification of the additive process. X-ray computed tomography (XCT) has come a long way since its first use in medical imaging, where physical density differences of less than 1% can be distinguished. Industrial XCT offers, non-destructive testing (NDT), defect detection, dimensional metrology, and scan times far superior to gold standard Coordinate Measurement Machines (CMM). Both AM and XCT are akin, regardless of complexity, the speed in manufacturing and inspection stays the same, respectively. Unlike their counterparts, Traditional Manufacturing (TM) and CMM where numbers of features drive speed and cost; but of course, these technologies are ‘tried and tested’ iteratively refined through centuries. AM’s ability to create complex internal geometries and XCT’s ability to inspect internal features are sometimes the only go-to option, but these methods are not standardised. The most critical factors for XCT are, XCT metrology is traceable, repeatable, and accurate, as well as, correct characterisation of AM features and structures; this brings the need for non-contact metrology but also brings forward a multitude of hurdles, lack of standards, measurement, and manufacturing uncertainties to name a few. The review of current state literature combined with recommendations and challenges from standards bodies, academia and industry has driven this research. A set of experiments, starting from understanding the fundamentals of XCT and its capability were performed to identify and tackle numerous uncertainties traditional CMM, AM and XCT is challenged with; this led to the development of a measurement artefact that can be manufactured using several AM methods and measured using CMM and XCT in a single setup. The Huddersfield Cylindrical artefact has been the centre of many capability studies, with design and methodology request from Lawrence Livermore National Laboratory (LLNL), University of California, Berkeley, National Physics Laboratory (NPL), Teddington, UK, National Metrology Institute of Japan (NMIJ), Japan, Department of Mechanical and Aerospace Engineering, University of California, San Diego, Renishaw, Gloucestershire, UK and Hexagon Manufacturing Intelligence, Telford, UK. The Huddersfield Cylindrical artefact has undergone, Measurement System Analysis (MSA) of CMM measurement strategy, dimensional measurement analysis of XCT and hardware/software parameter studies and, interlaboratory comparisons. Also, serval Statistical Process Control (SPC) and Six Sigma tools were applied, such as Taguchi, Design of Experiments (DoE), Individuals - Moving Range Charts (I-MR) and Monte Carlo methods. As industry moves closer to industry 4.0’s ethos, smart quality, and CAD the master, the work presented here, serves as a part in the foundation for traceable dimensional XCT for AM, increasing the bandwidth, and the throughput that AM offers while improving XCT dimensional measurement methodologies to detect part defects quickly, closing design, manufacturing and inspection loops.

FINAL THESIS - SHAH.pdf - Accepted Version
Restricted to Repository staff only until 29 November 2029.
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (11MB)
Add to AnyAdd to TwitterAdd to FacebookAdd to LinkedinAdd to PinterestAdd to Email