Christina Royce: Analyzing Re-entry Systems by Melis A. '08

To manufacture the silicon chips that run your computer, silicon wafers must undergo repeated cycles of deposition, patterning, implantation of impurities, and heat treatments (1). Thus, the same piece of silicon must be processed using the same set of machinery several times. This type of manufacturing system is called a re-entry system, where a silicon wafer re-enters the system with a higher priority each time it is processed.

Christina Royce, a senior double majoring in Mechanical Engineering, with a concentration in Manufacturing through the 2A program, and Economics, is doing her senior thesis project on re-entry systems through the Lab for Manufacturing and Productivity (LMP). Her thesis advisor is Dr. Stanley Gershwin, who literally wrote the book on manufacturing systems engineering and is the associate director of the LMP.

To analyze these systems, Christina is developing computer simulations which will eventually be compared to mathematical approximations — which were developed by applying observations to fundamental theories. Her simulations receive input regarding process flow, equipment availabilities, and process times, and use it to calculate the probability that a certain machine will be working on a given part and the final output volume. These calculations are not a trivial matter considering the huge number of machines that are used to simultaneously process millions of parts in a given factory.

Christina’s research has several important real-world applications. For example, semiconductor manufactures are always looking to improve their process technology to create the next generation of smaller, and therefore more efficient and economical, chips. Intel just came out with the first 45 nm static random access memory (SRAM) chip, which will allow them to “fit twice as many transistors in a given area” (2). It would be economically infeasible for Intel to scrap all of their existing machinery, which is designed to make 65 nm chips, in favor of manufacturing the newer SRAM chip. Hence, it’s important to have flexible manufacturing systems that have the ability to apply different processes to a wide variety of parts, so companies can save a significant amount of time and resources. Also, by using simulations like the ones that Christina is developing, Intel can identify bottlenecks and increase their production.

This is Christina’s first faculty-supervised UROP. As a freshman, she participated in Terrascope, where students apply core science knowledge to solve complex, real-world problems that relate to Earth systems. As part of the program, she traveled to the Amazon and became interested in issues related to refugee camps. After returning from the trip, she and a group of three other students decided to continue the research that they began in Terrascope. So, in the fall of their sophomore year, they began the project by finding an advisor, monetary support, and a group of 20 students. They worked together to create a 60-page booklet with information about the current state of refugee camps, along with areas for improvement and future research. She describes it as a very challenging but rewarding experience.

Next year, Christina will be working at General Electric! They have a manufacturing division, but her first 6-month rotation will be in energy-related research. We wish her the best of luck =)

1. http://www.bell-labs.com/org/physicalsciences/projects/silicon/silicon.html
2. http://intel.com/technology/silicon/new_45nm_silicon.htm

Picture from http://www.absolutestockphoto.com/albums/userpics/10012/normal_Absolute_12_5416.jpg