WIP: Supporting Student Understanding of Finite Element Analysis and Computational Science: Classroom Scaffolding and ChatGPT

This work-in-progress research paper presents the preliminary results of a study exploring the effectiveness of using computational notebooks to enhance student learning in a Finite Element Analysis (FEA) course for undergraduate Mechanical Engineering students. Our previous work has shown that students often face difficulties in grasping abstract concepts from mechanics of materials while simultaneously learning computational modeling. However, students recognized several advantages of using MATLAB for FEA compared to manual calculations, including significant time savings, increased efficiency, and reduced errors. Nevertheless, they also faced challenges, including a steep learning curve for MATLAB and concerns about how this limitation hinders their conceptual understanding. Despite these drawbacks, they recognized the importance and value of developing computational skills for their future careers. In this study, we extended the scaffolds and changed the sequence of activities to address the challenges the students faced in the previous iteration of our work. Specifically, we provided worked examples that students needed to use, self-explain, and modify before they engaged in programming from scratch. Also, after developing a basic understanding of how to implement FEA in MATLAB, the students used ChatGPT to generate a code that would do the same task. This activity required them to evaluate and refine automatically generated MATLAB code, as ChatGPT may provide alternative solutions that might not always work correctly. We explore three main topics to understand student experiences with and perceptions of this approach: (1) the value of using computational methods compared to manual completion for FEA; (2) the challenges and support of using MATLAB for FEA; and (3) the effectiveness of simulation tools to learn FEA. The goal of this project is two-fold: (1) supporting student learning of intricate phenomena explored in mechanics of materials, like distribution of stress, and stiffness, and (2) fostering essential computational thinking skills through practical disciplinary coding experience. By implementing these elements, the study anticipates a substantial improvement in students' understanding of FEA principles and their ability to translate them into solutions for real-world engineering challenges.