Graduation Semester and Year
Spring 2026
Language
English
Document Type
Thesis
Degree Name
Master of Science in Aerospace Engineering
Department
Mechanical and Aerospace Engineering
First Advisor
Dr. Vijay Gopal
Second Advisor
Dr. Kamesh Subbarao
Third Advisor
Dr. Luca Maddalena
Abstract
Operation of a blowdown supersonic wind tunnel to achieve desired testing conditions requires the development of an effective control system. For cold supersonic wind tunnels, regulation of total pressure within the plenum chamber is essential to achieve the target dynamic pressure. In such facilities without active heating, the total temperature decays throughout the experimental run. In this work, the primary objective is to develop a control system for a high unit Reynolds number supersonic wind tunnel at the University of Texas at Arlington, capable of operating in three modes: constant dynamic pressure, constant Reynolds number, and constant number density. This effort involved the development of a detailed gas-dynamic model of the wind tunnel, a dynamic model of the control valve, a hybrid control strategy, as well as data acquisition and signal generation using National Instruments hardware and the MATLAB platform. Two primary challenges were addressed during this process: runtime optimization under high mass flow rates with moderate storage tank capacity, and mitigation of the effects of dead time on system stability.
Achieving optimal performance was challenging, particularly due to the presence of control valve dead time. In the present facility, the dead time is comparable to the characteristic time of the wind tunnel, which inherently makes control of the system challenging. The control valve imposes a lower bound on the achievable characteristic time, as the system can only evolve as fast as permitted by the control valve dynamics. In facilities with longer runtimes, less aggressive gains can be used, reducing the impact of dead time. However, for the present system, conventional real-time PID control becomes difficult to tune and may lead to instability.
To address this issue, a hybrid control strategy has been developed. A detailed virtual model of the wind tunnel, incorporating control valve dynamics and non-ideal effects such as real gas effects, heat transfer, and pressure losses, is constructed in Simulink and used in conjunction with a custom optimization routine to generate controller gains without dead time effects in the virtual environment. The hybrid control system developed, implemented in LabVIEW, operates in two stages. The first stage is passive, using a precomputed control trajectory from the optimized virtual simulation to reduce the error between the process variable and the setpoint. Once the error is sufficiently small, a second, active PID stage is engaged to refine the response. This approach mitigates integral windup and improves controllability despite the presence of dead time. The hybrid strategy appears to be a novel approach in the context of wind tunnel control.
The control framework is further extended to support multimode operation, enabling the system to maintain constant plenum pressure, or constant test section Reynolds number, or constant test section number density. The presented experimental results demonstrate significant improvement in steady-state performance of the wind tunnel across all operating modes considered.
Keywords
gas dynamic, multimode control system, dead time, hybrid, PID, blowdown supersonic wind tunnel, stability, matlab, simulink, optimization
Disciplines
Aerodynamics and Fluid Mechanics | Aerospace Engineering | Engineering | Navigation, Guidance, Control and Dynamics
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Recommended Citation
Prasad, Harshwardhan, "Gas Dynamic Modeling of Blowdown Supersonic Wind Tunnel and Multimode Control System Development" (2026). Mechanical and Aerospace Engineering Theses. 3.
https://mavmatrix.uta.edu/mechaerospace_theses2/3
Included in
Aerodynamics and Fluid Mechanics Commons, Navigation, Guidance, Control and Dynamics Commons
Comments
The author would like to acknowledge the advisor, Dr. Vijay Gopal, and committee members, Dr. Kamesh Subbarao and Dr. Luca Maddalena for their support and guidance.