Date on Master's Thesis/Doctoral Dissertation

8-2020

Document Type

Doctoral Dissertation

Degree Name

Ph. D.

Department

Mechanical Engineering

Degree Program

Mechanical Engineering, PhD

Committee Chair

Chou, Kevin

Committee Member

Starr, Thomas

Committee Member

Berfield, Thomas

Committee Member

Wang, Hui

Author's Keywords

Additive manufacturing; laser powder bed fusion; melt pool; numerical modeling; porosity

Abstract

Powder bed additive manufacturing (PB-AM) process utilizes an electron beam or a laser as a heat source to melt the metallic powder particles. These processes have the capability of freeform fabrication, however certain defects such as porosity, high surface roughness, etc. would hinder its application. It is important to understand the effect of the process parameters and the underlying physical phenomena, which lead to the formation of such defects. In this regard, a three-dimensional (3D) thermo-fluid model is developed to study the effect of beam speed on the surface morphology during powder bed electron beam additive fabrication (PB-EBAF). Besides, the surfaces of PB-EBAF fabricated Ti-6Al-4V parts are analyzed using a white-light interferometer. The results show that in general, the build surface roughness along the beam moving direction slightly increases with the scanning speed. On the other hand, the hatch spacing noticeably affects the surface roughness in the transverse direction. In addition, the numerical model was modified to incorporate powder particles and study the effect of powder distribution towards the single-track formation during the laser powder bed fusion (LPBF) process. The numerical results show that the single-track morphology and density depend on the process parameters: scanning speed and laser power. Besides, micro-computed tomography (micro-CT) is utilized to characterize the pores formed during the LPBF process. Single tracks were fabricated with linear energy density (LED) ranging from 0.1 J/mm to 0.98 J/mm, and the samples were then scanned using micro-CT to measure keyhole porosity. The results show that the severity of the keyhole porosity increases with the increase of the LED. By keeping the LED constant in another single-track scanning experiment, different combinations of the power and the speed were tested to investigate the individual effect. The results show that for the same LED, the pore number and volume increased with increasing the power to a certain critical level, beyond which, the further increase and power resulted in fewer pore number and lower pore volume. The experimental results suggested that the dynamic phenomenon of a melt pool during the LPBF process is complex and sensitive to process parameters. Hence, a discrete element method (DEM) is utilized to obtain a powder distribution, which is then used to perform a thermo-fluid simulation using FLOW-3D software. The numerical results indicated that for a constant LED, the keyhole size increases with the increase in the laser power. The keyhole becomes stable at a higher power, which may reduce the occurrence of pores during laser scanning. In addition, a back and forth raster scanning is performed to form three tracks to investigate the effect of scan length on the melt pool size at different locations along the laser travel direction. Three scan speeds (375 mm/s, 750 mm/s, and 1500 mm/s) are used with laser power of 195 W, and three tracks are fabricated with 0.5 mm, 1 mm, and 1.5 mm scan lengths. Besides, hatch spacings of 80 µm and 120 µm are used. The fabricated samples are analyzed using white light interferometer and metallography. Moreover, a powder scale numerical model is developed to understand the residual heat effect on the melt pool. The results show that the region where the laser changes the direction is the most affected zone where a significant increase in the melt pool size is observed. The depth of the melt pool increased with increasing track number. The delay in the successive laser scan needed to minimize the residual heat effect is calculated for different process parameters.

Included in

Manufacturing Commons

Share

COinS