Employing eye-tracking, screen capture and artifact analysis methods to characterize re-design for Additive Manufacturing behaviors
Open Access
- Author:
- Mehta, Priyesh
- Graduate Program:
- Additive Manufacturing and Design
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- October 24, 2019
- Committee Members:
- Catherine G P Berdanier, Thesis Advisor/Co-Advisor
Guhaprasanna Manogharan, Committee Member
Timothy W. Simpson, Program Head/Chair - Keywords:
- Additive Manufacturing
Powder bed fusion
metal Additive
Eye tracking
Qualitative analysis
Design for Additive Manufacturing
manufacturability matrix
design ideation
3D printing
cognitive design
engineering education
graduate engineering education
workforce development
design engineering
novice v/s expert designers - Abstract:
- Additive Manufacturing (AM) has attracted significant interest from industry and academia practitioners all around the globe. The design freedom offered by this technology along with a considerable reduction in lead times are key factors catalyzing its proliferation. Engineering designers are optimistically counting on AM to help them reinvent the product development life-cycle. A key influence to accelerate the adoption of AM is developing a multi-disciplinary workforce to meet the ever-increasing demand-supply gap. Designers of the future will require inter-disciplinary skill sets to adopt, evaluate and progress AM technologies. To help engineers optimally reap the benefits of the design freedom offered, a new design thinking approach coined as “Design for Additive Manufacturing” is recognized by industry and academia practitioners. Engineering designers are often tasked with re-designing a component or assembly for AM, which is traditionally intended for conventional near-net manufacturing processes. Existing research directives in the Design for AM arena focus on design optimization frameworks, worksheets and guidelines which provide a check-list for designers during the execution of the DfAM methodology. However, few research initiatives are invested towards understanding or characterizing the behaviors of designers performing a re-design activity. Such a synthesis would allow researchers to qualitatively understand design behaviors which can be then related to design success from the AM perspective. This thesis provides a systematic literature review of the state of the research in AM engineering education, followed by the first known qualitative characterization of the re-design for AM process employed by engineering designers during a design challenge. The design challenge essentially consists of a single objective optimization problem of re-designing an airplane bearing bracket for AM, particularly for the Laser Powder Bed Fusion (L-PBF) process. The complete re-design process is recorded using eye-tracking and screen capture methods using visual gaze pattern data. Design behaviors exhibited by engineers are characterized qualitatively using constant comparative methods derived from the traditional cognitive and human subject research literature. The designs generated by participants are analyzed using a novel manufacturability matrix, developed by the authors specifically for the L-PBF technology. The designs generated are then compared using a normalization approach, where a manufacturability index is derived to highlight participant performance. Aggregate behaviors from designers are compared and contrasted using content analysis methods to link designer behaviors with success in generating a design to match or exceed the manufacturability requirements for AM. The redesign for AM process is primarily driven by intuition, logical judgments, and application of engineering first principles. Results from this research highlight the differences in behaviors exhibited by novice and expert engineering designers challenged with the same optimization problem. It is observed that participants spend a significant portion of their total activity time on stress analysis and sketching related activities. A major portion of the total time spent by designers is invested in 2D Sketching related activities, which highlights the need for non-parametric software to suit the re-design for AM process. With data acquired from the participant approaches, a re-design for AM workflow is presented to encourage behaviors to correspond with success in manufacturability for the L-PBF process. Implications from this study will serve engineering designers to develop a comprehensive understanding of the approach and methods used during the (Re) Design for AM process. The designer-centric workflow presented in this study can be used as an aid by the engineering education and research community to help educate students, appreciate the re-design for AM workflow.