The main objective of this paper is development of a simple real-time transient performance model for jet engine control. A jet engine arrives to its most dangerous condition during transient operation that may be triggered by fast changes of the input fuel command signal. Thus, the control system specifications are formulated to specify the maximal variance of the fuel flow command (from idle to maximum power level) during transient maneuver. Linear and piecewise-linear techniques are not always convenient and appropriate for turbine engine controller design. An alternative quasilinear simple/fast engine model is discussed in this paper. This model has maximum accuracy for maximal variance of the fuel flow input command in accordance to the jet engine control system specifications. The fast model is obtained using the Novel Generalized Describing Function, proposed for investigation of nonlinear control systems. The paper presents the Novel Generalized Describing Function definition and then discusses the application of this technique for the development a fast turbine engine simulation suitable for control and real-time applications. Simulation results are compared between the conventional and fast models and found to provide good agreement.
This paper discusses an alternative prediction program. The basic DYNJET MATLAB prediction program is suitable for steady-state and transient performance of single-spool jet engine calculations with convergent nozzles without afterburners.The MATLAB program, available upon request, is flexible and may be applied to different research problems. The program is capable of calculating engine performance with a one-dimensional tabular or graphic presentation of a compressor map. The program uses Polar Coordinates for the compressor map description to increase accuracy of calculations. DYNJET includes two independent programs: DYNJET-1, a Matlab program for calculating engine steady-state performance, and DYNJET-2, a program for calculating transient performance. DYNJET-1 calculates steady-state operating lines and engine variables. The engine input is in terms of fuel pump voltage. DYNJET-2 calculates engine transient response to timedependent arbitrary fuel-pump voltage commands. The program output contains a compressor map with a transient line and time dependent variable graphs and tables, containing all output data. The dynamic engine model consists of nonlinear algebraic and differential equations including one for the fuel pump. Initial conditions (t = 0) are automatically calculated in the steady-state model for given values of Η and Μ and for initial input fuel pump voltage. Comparisons of simulation results with experimental data for jet engines ("Microjet", NPT-171 and SR-30) prove the model's validity.DYNJET programs are written in Matlab 5.31 and can be used without modifications in a PC or a UNIX network. They are modular and relatively easy to use, hence can serve as a bench-mark for engine control development as well as for educational purposes.
This paper discusses low cost and fast evaluation of effective performance maps for engine components such as nozzles, turbines and compressors using data acquired during transient engine operations. The objective of the present study is to re-evaluate effective engine performance maps when not all the engine component maps are known. The work is equally relevant to account for manufacturing tolerances of different engines resulting in small differences in their performance. Disassembly and rebuilding of an engine as well as degradation in performance due to extensive usage can cause small variations in their performance. If the engine maps do not represent particular engine operations with precision, its performance predictions will differ from the actual characteristics. The engine map evaluation is carried out using data acquired during aircraft transient operations at different altitudes, Mach numbers, angles of attack, side slip angles and during post-stall. The method is based on a conventional dynamic engine model. The unknown engine map characteristics are excluded from the model while the corresponding numbers of some measured values are added to the model input. The evaluated map values are calculated using an inverse solution of the equations of the “shortened” mathematical model. Some jet engine maps may be evaluated simultaneously. Two examples for fragments of compressor and compressor/nozzle map evaluations are presented in this paper.
Engine thrust vectoring (TV) is an emerging new technology for future military and civil aircraft in which the Technion has made significant contributions. Rapidly deflecting engine jets to maneuver the aircraft with or without conventional aerodynamic flight control (CAFC) significantly enhances the flight safety, agility, and combat kill-ratio capabilities of fighter aircraft in the near term and enhances the safety of civil transport jets in the long term. There are yet no realistic predictions of engine dynamic responses to yaw-pitch-roll TV commands in the public domain. Hence, the primary aim of this work is to provide such a first. The results obtained comprise, therefore, a required fundamental step for advanced aircraft/TV implementation. The selection of this work focuses on the Lockheed-Martin TV F-16/F-100 research study conducted at the Technion. A unique TV-engine computer algorithm has been developed that expands the conventional steady-state modeling capabilities of on-and off-design as well as the conventional transients (via throttle changes) to create realistic dynamic TV-engine simulations at various altitudes and Mach numbers. This work has been expanded to include predictions for TV in civil aircraft (via a fixed geometry nozzle) under the same conditions. It is concluded that the military TV configuration, as expected, produces no variations in engine performance while providing TV flight control benefits. It is also demonstrated that under the same dynamic conditions, the civil configuration provides an increase in thrust, enhancing the benefits available from TV in the civil domain.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.