Session: 19-03: Micro Gas Turbines.

Submission Number: 174244

Calibration of Microjet Gas Turbine Thermofluid Network Model Using Steady-State Measurements 

The accurate prediction of microjet gas turbine (MGT) performance requires models that can capture component-level interactions and operating conditions with high fidelity. However, discrepancies often arise between analytical cycle models and real engine behavior due to the accumulation of uncertainties in flow and heat transfer effects, and parasitic losses. This paper presents the calibration of a physics-based (white-box) steady-state thermofluid model of a MGT cycle based on mass, energy and momentum balances together with ideal gas properties. The calibration is performed using steady-state experimental datasets to enhance the model's predictive capabilities under varying load conditions (thrust levels). A 250N MGT operating on kerosene instrumented to measure shaft speed, thrust, fuel and intake air flow, exhaust gas temperature (EGT), and gas composition. The uncalibrated thermofluid Flownex SE model, based on component characteristic relationships taken from literature and standard thermodynamic assumptions, exhibited notable deviations in performance prediction compared to measurements. Calibration is performed through parameter tuning and correction factor adjustments using optimization algorithms. The methodology focuses on refining combustion efficiency using measured exhaust gas composition together with inlet air and fuel flow rates, and on adjusting turbomachinery performance maps based on additional measured performance parameters. Results demonstrate that the calibrated model has an overall improved predictive accuracy compared to the baseline. The study provides a modelling and calibration framework for enhancing MGT performance assessment methods and developing high-fidelity cycle models.

Presenting Author: Zargo Afonso Stellenbosch University

Presenting Author Biography: Zargo Afonso is a current Master’s student enrolled at the Department of Mechanical and Mechatronic Engineering at Stellenbosch University in South Africa. His research focus is on the development, validation, and application of a thermofluid network-based model of combustion and aerodynamics in a microjet gas turbine cycle. His work aims to improve the efficiency and reliability of steady-state and transient performance prediction modelling of small-scale turbine systems.

Authors:

Zargo Afonso Stellenbosch University
Pieter Rousseau Stellenbosch University
Ryno Laubscher Stellenbosch University