Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Integrated Computational Materials Design for Alloy Additive Manufacturing: Introducing Data-Driven Approach to Physical Metallurgy

Wang, Xin (2023) Integrated Computational Materials Design for Alloy Additive Manufacturing: Introducing Data-Driven Approach to Physical Metallurgy. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

Download (5MB) | Preview


Additive manufacturing (AM) attracts broad interest due to its ability to produce complex geometries, fast prototyping, and in-situ repair. However, AM involves many parameters and uncertainties, which lead to products property variation. For instance, the influence of composition variation on AM component is an important issue that has not been thoroughly studied. Moreover, micro-segregation in AM caused by the high-cooling rate makes the as-built structure and properties vary locally within the prints. The alloy bulk properties may change with different AM processing parameters. Such variations must be studied while the experimental study is time- and cost-consuming.
This thesis introduced the data-driven approach, such as statistical analysis, machine learning, and Bayesian inference combined with integrated computational materials engineering (ICME), to address the AM property variation challenges. Firstly, the process-structure-property-performance (PSPP) relationships for AM high-strength low-alloy (HSLA) 115 steel with post-treatment were established to study feedstock composition impact on the print performance. High-throughput calculations of the ICME framework quantified uncertainties in critical properties, such as yield strength, printability, and low-temperature ductility, with the feedstock composition variation. Moreover, the machine learning approach was implemented to surrogate the ICME model framework for an accelerated simulation for a more comprehensive study and robust feedstock composition optimization. Finally, the printed optimized HSLA 115 steel showed excellent properties even though the printed composition differs from the designed composition, which proves the successfulness of the design with uncertainty in composition. This thesis studied the impact of AM 316L stainless steel segregation on the deformation mechanism and mechanical properties. A machine learning-based stacking fault energy (SFE) predictor, which surpassed the conventional thermodynamic and empirical models, was developed to predict the SFE change with segregation in AM. This data-driven model successfully explained the twinning behavior in as-built AM 316L. Finally, a Bayesian-based model calibration was applied to understand the considerable variation in the mechanical properties of CoCrFeMnNi HEAs with different AM techniques and processing parameters. It revealed the importance of dislocation density and grain size in strengthening the AM products, and correlation analysis was conducted to find the relationship between the strengthening mechanism and processing parameters.


Social Networking:
Share |


Item Type: University of Pittsburgh ETD
Status: Unpublished
CreatorsEmailPitt UsernameORCID
Wang, Xinxiw114@pitt.eduxiw1140000-0002-4287-3956
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairXiong,
Committee MemberGleeson,
Committee MemberTo,
Committee MemberWiezorek, Jö
Committee MemberBedewy,
Committee MemberChen,
Date: 14 September 2023
Date Type: Publication
Defense Date: 7 April 2023
Approval Date: 14 September 2023
Submission Date: 8 May 2023
Access Restriction: No restriction; Release the ETD for access worldwide immediately.
Number of Pages: 148
Institution: University of Pittsburgh
Schools and Programs: Swanson School of Engineering > Materials Science and Engineering
Degree: PhD - Doctor of Philosophy
Thesis Type: Doctoral Dissertation
Refereed: Yes
Uncontrolled Keywords: ICME, Data-Driven, Modeling, Additive Manufacturing
Date Deposited: 14 Sep 2023 13:35
Last Modified: 14 Sep 2023 13:35


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item