Current Research Opportunities in the TFCL

Current opportunities for conducting research in the Thermo-Fluids Control Lab will be posted here.  We are always looking for students whose interests match our vision, and whose abilities will be an asset to the group.  Unfortunately, natural limitations on my time mean that I can't advise every qualified graduate or undergraduate student.  If you are interested in a project listed below, or wish to discuss conducting your own idea for a research project, please send an email to Dr. Bryan Rasmussen (brasmussen@tamu.edu).

Undergraduate Research Projects

The TFCL is a strong advocate for undergraduate research experiences.  There are several undergraduate student worker positions, and virtually unlimited openings for undergraduate work-study students (MEEN 485).   Undergraduate students may choose to assist with one of our current research projects, or any project listed below.  Interested students should contact Dr. Bryan Rasmussen (brasmussen@tamu.edu).

Graduate Research Assistantships

There are currently no openings for additional graduate research assistants or graduate student workers.

Independent Graduate Research Projects

There are currently several opportunities for independent graduate research projects.  These projects are not currently funded.  Students wishing to pursue research in one of these areas will have access to the experimental facilities in the TFCL, and will be advised by Dr. Rasmussen.

Dynamic Modeling and Control of CO2 Refrigeration System

This project consists of the construction of an experimental CO2 refrigeration system, the development and validation of dynamic models, and the design and evaluation of control strategies. Much of the needed equipment is currently available, and the development of dynamic models is virtually complete. The immediate focus is on the development of the experimental system, with modeling and control research to follow.

Evaluation of Novel Rankine Cycle for Low Temperature Waste Heat Recovery

There is considerable interest in developing means of electricity generation from low temperature waste heat. This project will evaluate the potential advantages and potential control difficulties associated with a novel Rankine cycle. The immediate focus will be on computer simulation and evaluation. Should the analysis prove favorable, an experimental system will be constructed. Contact Dr. Rasmussen for more information.

Online Grey Box Identification of Vapor Compression System Dynamics

Many of the control-oriented models developed for vapor compression systems depend nonlinearly on several parameters that can not be obtained from steady state data. Experience has shown that the dominant dynamic response is highly sensitive to a subset of these parameters. This project will explore effective methods for online identification of these parameters.

Dynamic Fault Detection Using Real-Time Model Validation

This research seeks to develop on-line dynamic fault detection algorithm based on off-line model validation methods for robust control. Validation of dynamic models is generally approached from a statistical point of view. The statistical properties of the model residuals are calculated and evaluated leading to statistical confidence levels in the model accuracy. Alternatively, the model validation problem can be approached from a robust control viewpoint, using various computational tools to determine whether the data can be accounted for by an assumed level of measurement noise and model uncertainty. This approach has obvious benefits in the subsequent steps of robust control synthesis, where the aforementioned level of model uncertainty is used in the construction of the controller. Both static and dynamic fault detection algorithms have similarly used a statistical approach to determining online model deviations. However, a model uncertainty based approach to fault detection would be a much more powerful tool for assessing fault tolerant control strategies. The clear obstacle to this type of approach is the computational burden that is involved in converting these algorithms to operate in real-time. For more details regarding potential solutions to this challenge, please contact Dr. Rasmussen.