Ilana Lu

Date of Award

Spring 2019

Document Type

Open Access Thesis


Mechanical Engineering

First Advisor

Anthony Reynolds


As a solid state process, friction stir welding (FSW) is an advantageous joining method for higher strength aluminum alloys such as 2050 Al-Cu-Li which are not readily fusion welded. This study explores the potential of friction stir lap welding (FSLW) as an additive manufacturing process for this aluminum-lithium alloy. To accomplish this, lap weld joints were performed so that each additional layer would penetrate the preceding weld pass which creates a large, stacked build. Excess material is machined away leaving only weld. Cast AA2050 is explored in this study as a more cost-efficient feedstock for this process as compared to wrought plate.

Three builds of 2050 aluminum alloy were studied for weld response variables, and properties. These variations included: (i) 25mm length pin and cast 2050 aluminum alloy in three layers including 2050-T3 substrate and top layers, (ii) 12.85mm length pin and cast 2050 in seven total layers with a 2050-T3 substrate/top, and (iii) 12.85mm length pin and 2050-T3 in all seven layers. A finite element analysis (FEA) was also performed to compare beams made from the AM process and from monolithic 2050 plate.

All builds exhibited inhomogeneous hardness distributions with minimums in the overlapping heat affected zones. In addition to the hardness results, transverse tensile tests of two of the builds showed that strain tended to concentrate in the soft, heat affected zones but overall the builds possess moderately good ductility and weld efficiency. Fracture tests of weld material demonstrated complex behavior—fracture resistance of the welds was higher than that of the parent material; however, the presence of inhomogeneous hardness and process-created distributions added a level of irregularity. Overall, it was found that there is not a disadvantage to using cast material in friction stir additive manufacturing (FSAM) however the practicality of this process is not ideal at the current technology level.


© 2019, Ilana Lu