Document Type
Thesis
Date of Award
Summer 2019
Degree Name
Master of Science in Mechanical Engineering - (M.S.)
Department
Mechanical and Industrial Engineering
First Advisor
Samuel Lieber
Second Advisor
Zhiming Ji
Third Advisor
Lu Lu
Fourth Advisor
Bob Tarantino
Abstract
The Skills Gap in Mechanical Engineering (ME) Design has been widening with the increasing number of baby boomers retiring (Silver Tsunami) and the lack of a new generation to acquire, practice and perfect their knowledge base. This growing problem has been addressed with several initiatives focused on attracting and retaining young talent; however, these types of initiatives may not be timely for this new group to be trained by an established Subject Matter Expert (SME) group. Automated Engineering Design provides a potential pathway to address not only the Skills Gap but also the transfer of information from SMEs to a new generation of engineers. Automation has been at the heart of the Advanced Manufacturing Industry, and has been successful at accomplishing repetitive tasks with processes, software and equipment. The next stage in Advanced Manufacturing is further integrating Machine Learning techniques (Artificial Intelligence (AI)) in order to mimic human decision making. These initiatives are clear for the type of mechanized systems and repetitive processes present in the manufacturing world, but the question remains if they can be effectively applied to the decision heavy area of ME Design. A collaboration with an industry partner New Jersey Precision Technologies (NJPT) was established in order to address this question. This thesis presents an ME Design Automation process involving a multi-stage approach: Design Definition, Task Differentiation, Workflow Generation and Expert System Development. This process was executed on plastic extrusion tooling design. A Computer Aided Design (CAD) based Expert System was developed for the Automation of design, and the generation of a database towards future Machine Learning work. This system was run on 6 extrusion product examples previously designed by NJPT through traditional methods. The time needed to generate the design was reduced by 95-98%. This thesis demonstrates the capability of automating ME design, the potential impact in industry and next steps towards the application of AI.
Recommended Citation
Prasad Varghese, Allen, "Mechanical design automation: a case study on plastic extrusion die tooling" (2019). Theses. 1686.
https://digitalcommons.njit.edu/theses/1686
