EVALUATING THE ACCURACY OF A DIGITAL CORRELATION SYSTEM TO PREDICT THE FATIGUE FAILURE OF ADDITIVELY MANUFACTURED PARTS

• Shaun Botha North-West University https://orcid.org/0000-0002-4975-0721
• Jaundrie Fourie North-West University https://orcid.org/0000-0001-9796-4153
• Cornelius Petrus Kloppers North-West University https://orcid.org/0000-0001-5942-0251

DOI:

https://doi.org/10.7166/34-1-2736

Keywords:

Additive manufacturing, fatigue testing, fatigue life prediction

Abstract

Internal defects that occur during the additive manufacturing process of direct metal laser sintering can have a significant influence on the fatigue performance of Ti-6Al-4V additively manufactured parts compared with parts that are made using traditional manufacturing methods. For the adoption of additively manufactured parts in the medical, automotive, and aerospace industries, the expected fatigue life of parts must be predictable and repeatable for successful application. Micro-CT scanning is commonly used to provide insight into the expected fatigue life of parts by examining internal defects that have occurred during melting. However, using a micro-CT for every part in the supply chain is not always feasible, so alternatives must be considered. One approach would be to analyse parts by applying a short-run fatigue test coupled with a digital image correlation (DIC) system to detect areas on parts where excessive distortion is taking place. This study investigated the level of accuracy that a DIC system could obtain as an alternative non-destructive test to predict where and when a test specimen would fail. To determine whether this alternative was viable, tests were carried out until the specimens failed. The DIC images were analysed at 50% of the fatigue life and at the point just before failure to determine whether the DIC system accurately indicated the strain concentration at the same point where the specimens failed. Experimental data from the study shows that the DIC system could accurately predict the point of failure at the fatigue half-life in only 10% of the test specimens that were investigated. The DIC system was able accurately to predict the point of failure right before failure occurred in only 25% of the test specimens that were investigated.