Development and optimisation of a method for assessment of porosity in additively manufactured parts using x-ray tomography

Student thesis: Doctoral Thesis

Abstract

Additive manufacturing (AM) is a process whereby components are built layer-by-layer by means of powder or wire. The AM process provides clear benefits over conventional subtractive machining primarily in terms of part complexity. The technology continues to be the subject of rapid development and consequently, the geometrical repeatability and mechanical properties of AM parts are subject of research. AM technologies have evolved rapidly in the last decade enabling manufacturers to acquire additive manufacturing machines at costs comparable to multi-axis computer numerical control (CNC) machines. Additionally, the process offers the possibility of utilising high-performance engineering materials and super alloys, such as Titanium and Inconel; attracting special interest from aerospace and medical (orthopaedic implant) applications.

There are several drawbacks in the process preventing manufacturers from widely adopting this technology. A significant barrier is the lack of understanding and assessment technologies for the structural integrity of AM components. The evaluation of a component functional performance is the most important stage of any component development. Generally, the mechanical performance could be assessed by means of destructive and non-destructive testing methods. However, the destructive testing methods can be both expensive and time consuming.

In this thesis, the non- destructive testing method used is X-ray computed tomography (XCT), which is being utilised in characterising and measuring external and internal defects/features of metallic components made by AM technologies. XCT is capable of providing information about pore position, location and volume. While XCT is promising, there are several difficulties stopping industry from utilizing the technology. One of the main obstacles is the level of subjectivity within the process; this is problematic for assessment of internal process defect features such as process porosity.

This thesis reports on the process of developing the inspection techniques, which are capable of detecting and characterising internal defects; approaching the size of a single AM powder particle. In order to duplicate internal AM defects and features, a series of small porosity characterisation specific artefacts were developed. These artefacts contained micro defects, semi-fused powder and unfused powder. The artefacts were used to identify the challenges in the XCT process and determine the optimum XCT scan parameters for detecting and measuring internal defect/features. Furthermore, in this thesis, the limitations and challenges faced when attempting to identify such features in AM component are reported, and the XCT results repeatability and reproducibility was investigated. Additionally, SLM AM build chamber internal feature printing resolution was investigated, and defect analysis results were verified.

The results of studies/experiments carried out as part of this thesis had shown that XCT is an effective method for porosity detection, with a resolution approaching a single powder particle (15µm). Furthermore, based the different studies reported in this thesis; the recommended scan strategy was using a magnification as low as 38-μm voxel size to identify the location of the defects and then confirm the area where the defect is located using a high magnification scan.

It was also noted that surface determination is playing a vital rule in the results reliability. The results of several experiments reported that the ISO50% surface determination threshold is not the appropriate threshold for porosity analysis and custom surface determination protocol was developed. The results showed that selecting the appropriate scan parameters combined with the optimisation of surface determination could enhance the XCT defect characterisation process.

Based on the experience and knowledge gained from this research, a best practice method was developed, the best practice guide enhanced the efficacy of the method was demonstrated on and industrial NIKON XTH 225 machine.

The work carried out as part of this thesis were designed to be industry relevant and representative of actual internal defects, especially those found in AM components. This research was intended to be universal for all XCT machines and all powder-based AM processes. Although XCT process is capable of detecting internal defects within AM components the process requires high level of experience. At this stage, although highly informative, is not suitable in high volume manufacturing environment that requires 100% inspection, it is more suitable for low volume high value components found in medical and aerospace applications.
Date of Award28 Jan 2022
Original languageEnglish
SupervisorPaul Bills (Main Supervisor), Liam Blunt (Co-Supervisor) & Radu Racasan (Co-Supervisor)

Cite this

'