The Aeroswift machine is a novel high-speed powder bed fusion machine developed through a collaborative effort between the CSIR, Aerosud and the DSI. Its novelty lies in the substantial increase in build rate achieved through the implementation of a 5kW IPG laser and faster laser scanning speeds employed during processing. It is capable of producing Ti6Al4V low volume, high value and high integrity components required by the aerospace industry. Commercial selective laser melting (SLM) systems are a good benchmark for the type of quality needed in the integrity of aerospace components although they don't always meet them. The biggest difference between commercial systems and the Aeroswift machine is the amount of heat input used to make components based on the laser powers. Heat input is the ratio of the laser power to the scanning speed and it plays a role in the thermal history of a built part, its thermal gradients and therefore its residual stress. Heat input also has a big influence on the microstructure produced which determines the resultant mechanical properties. The focus of this project was to investigate the effect of increased heat input on residual stress, the development of microstructure and mechanical properties of Ti6Al4V specimens produced by the Aeroswift high (400 J/m) heat input system and commercial SLM Solution M280 low (150 J/m) heat input machine. This was to be accomplished by comparing the tested results of Aeroswift built specimens (High Heat Input) to those built by a commercial SLM machine (Low Heat Input). The effect of preheating on these properties was also studied. The low heat input specimens had two sets of test specimens, where one set was built without preheating and the other built at a preheating temperature of 200°C. This was the maximum preheating temperature for the commercial system used in this study. Firstly, the cantilever specimen were used to measure the amount of distortions that processing caused for both systems. The measured spread of the cantilever gave an indication of the amount of distortion caused by each processing condition. Distortion was found to be similar between the high heat specimen and the low heat specimen. Preheating at 200°C also did not give an appreciable difference in the amount of distortion. X-ray Diffraction was used to measure very near surface residual stresses up to a penetration depth of 5 microns. Blocks of 20X20X22 mm3 for each processing condition were used with measurements taken at the top surface center of the blocks. The very near surface stresses were higher with an increase in heat input, where high heat input specimens had average tensile residual stress in excess of 650 MPa while the low heat input specimens had average tensile residual stresses below 400 MPa. The Incremental hole drilling technique was utilised to measure the stresses in the blocks up to a depth of 1 mm from the top surface. Holes were drilled at the top surface center of each block. The stress distribution for both the high heat input specimens and the low heat specimens increased from 0.2 mm to a similar range of 500-600 MPa between 0.3 mm to 0.8 mm depth. Preheating at 200 °C yielded the same amount of stress. The microstructural analysis involved imaging from Optical Microscopy, Scanning Electron Microscopy and Electron Backscattered Diffraction. This combination of techniques confirmed a martensitic microstructure morphology of α' laths within prior β grains for all the specimens. The α' laths were arranged in the form of basket-weaves as well as colonies. The high heat input specimen prior β grains were columnar having grown across several layers in the build direction. For the low heat input specimens both with no preheating and with 200°C preheating, the prior β grains were atypically discontinuous. A hexagonal-titanium phase was identified in all the specimens as the dominant phase, with essentially no presence of the cubic phase. Dog-bone tensile specimens built in the z-direction (build direction) were used to test for static mechanical properties. The Yield Strength and the Ultimate Tensile Strength were above 1000 MPa and 1200 MPa respectively for all specimens. The average elongation of 11.2% in the low heat input specimen with no preheating was significantly higher than the 4.3% achieved by the high heat input specimen. The effect of the observed micro porosity under the microscope is thought to have contributed to this behaviour. Compact tension specimens for fracture toughness and fatigue crack growth rate testing were built in the ZX direction as per ASTM E399-17 labelling. The high heat input specimens had an average fracture toughness of 43 MPa√m compared to the less than 38 MPa√m achieved by the low heat input specimens. The high heat input specimens also had a better crack growth resistance than the low heat input specimens. The low heat input specimens without preheating had better crack initiation resistance. The results show that an increase in heat input does not have a substantial effect on the integrity and quality of parts. In fact, it produces comparable results to commercial SLM processing deployed in this study with respect to the properties studied, with the exception of a lower ductility. This brings about even more confidence on the advantage of high-speed processing. Future work should include testing at other orientations as well as testing higher preheating temperatures.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:uct/oai:localhost:11427/35920 |
| Date | 04 March 2022 |
| Creators | Motibane, Londiwe Portia |
| Contributors | Knutsen, Robert |
| Publisher | Faculty of Engineering and the Built Environment, Department of Mechanical Engineering |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Master Thesis, Masters, MSc |
| Format | application/pdf |
Page generated in 0.0026 seconds