University of Michigan
I proudly graduated with a Masters in Mechanical Engineering with a focus on Design, in the class of 2008. It was one of the best experiences of my life! Michigan has given me more than I expected, and has made me a stronger better faster individual I am today.
I worked with Prof Kota, who heads the CSDL.Some of the work I did…
- Developed and modeled non-linear compact torsional springs using compliant members
- An amazing sector symmetric spring that could be made of stamped steel or polymers which displayed highly non-linear force-displacement behavior, based on geometry and changing contact regions changing the stiffness of the spring.
- Developed algorithms to optimize spring and material parameters
- I using ANSYS Classic, Optimus and MATLAB to obtain any desired non-linear force-deflection behavior. The algorithm used optimization routines to converge on any L-d curve by modifying either design or material design variables. This work was an amalgamation of my FEA class, and optimization class, and tons of perseverance to get the 3 softwares to talk to each other.
- Generated a library of behavioral curves thru optimization & DOE for further research and development
- Using the tools I developed, I explored the design space and identified key aspect ratios that result in interesting force responses of the spring. This is where I graduated and my work with Prof Kota ended.
I would have loved to continue working, perform experimental work and close the loop on the project.In retrospect, the work I did at the CSDL was the one I most enjoyed. When I joined the group, I hadn’t worked on any of those software packages. Within a years time, I was able to get non-linear large deflection contact based analysis running in Ansys running in batch mode, using APDL written by Matlab that ran optimization routines in parallel with DOE results from Optimus. The amount of satisfaction I got from the final results, and working codes; I cannot put into words.
I’ve studied the following courses at Michigan (in order of how much I’ve enjoyed them):
Finite Element Methods, Solid Continuum Mechanics, Intermediate Dynamics, Continuous Design Optimization, Automotive Body Structures, Smart Materials, Electromechanical System Design, Linear Algebra, Statistical Quality Control
Some of these courses were purely theoretical, while some of them had lab work and project work. Some of the best professors on the subjects taught these courses at the university… what a fantastic experience. To this day, I continue to refer to notes and revise the material to better my understanding of the material I learnt in school.
Some projects I’ve worked on…
Continuous Design Optimization:
Optimization of power consumption of Caterpillar Backhoe vehicle:
- Generated mathematical model of kinematic behavior of digging mechanism & hydraulic actuator system
- Performed multi-variable design optimization of power consumption by actuator sizing & selection of actuator pin locations resulting in lower power consumption
- Performed global optimization with a team working on hydraulic power train & controls of backhoe
Electromechanical System Design:
Mechatronic Gauge for Piano Regulation (Patent in progress):
Client: Piano Technology Department, U of M School of Music
- Designed & implemented a feedback control algorithm for constant velocity arm deflection and reaction force calculation in C++, LabVIEW using optical encoders & OOPIC microcontrollers
- Supervised undergraduate mechanical engineering team in design, analysis and fabrication of frictionless compliant linear actuator
- Surpassed expectations of client – 80% reduction in measurement time, 0.01 gm least count achieved
Engineering Research Center
Over the summer, I interned in the Reconfigurable Manufacturing Lab
at the ERC. I worked with Prof Moyne
and his group. This was some very challenging work for me: it involved learning JAVA and network protocols from scratch, designing the architecture for the work and implementing it within the 3 month timeframe.
As manufacturing operations become more precise from a timing perspective, there is an increasing need to utilize time sync across the factory. Precise time sync enables nodes to coordinate their actions & allows controllers to use time as an input for more accurate error correction. Accomplishing precision time sync over networks has been addressed by protocols like the IEEE 1588 std.I worked on the following at the lab:
- Developed a semiconductor factory network simulator to investigate IEEE 1588 time synchronization techniques in distributed systems in industrial Ethernet
- Implemented modules handling data requests & reports in XML, JAXB complying with SEMI standards
- Worked with a team of graduate students; achieved weekly goals in conjunction with a team at the National Institute of Standards and Technology, Maryland
We (our team) got two paper publications from the work done:
‘Precise Time Synchronization in Semiconductor Manufacturing’, IEEE International Symposium on Precision Clock Synchronization for Measurement, Control and Communication, 2007. ISPCS 2007
‘Time synchronization for diagnostics and control in Ethernet-based applications’, Submitted to Proceedings of the American Controls Conference, June 2008