In the first part of this paper the equations and results for the transient models developed for the SAS components in isolation have been thoroughly explained together with the assumptions made and the limitations that arose subsequently. This second part explains the work carried out to couple the individual components into a single network with the aim of assembling a dynamic model for the whole engine air system. To the authors’ knowledge the models published hitherto are only valid for steady or quasi steady state. It is then the case that the differential equations that govern the fluid movement are not time discretised and thus can be solved in a relatively straightforward fashion. Unlike during transients, the flow is not supposed to reach sonic conditions anywhere within the network and most important, flow reversal cannot be accounted for. This study deals with the mathematical apparatus utilised and the difficulties found to integrate the single components into a network to predict the transient operation of the air system. The flow regime — subsonic or supersonic — and its direction have deemed the choice of the appropriate numerical and physical boundary conditions at the components’ interface for each time step particularly important. The integration is successfully validated against a known numerical benchmark — the De Haller test. A parametric analysis is then carried out to assess the effect of the length of the pipes that connect the system cavities on the pressure evolution in a downstream reservoir. Transient flow through connecting pipes is dependent on the fluid inertia and so it takes a certain time for the information to be transported from one end of the duct to the other. As it would be expected, the system with a longer pipe is found to have a longer settling time. Finally, the work concludes with the analysis of the flow evolution in the secondary air system during a shaft failure event. This work is intended to continue to address the limitations imposed by some of the assumptions made for an extended and more accurate applicability of the tool.

The amount of air drawn by the secondary air system (SAS) from the main gas path, although necessary, impairs the engine performance because it does not contribute to engine thrust. In order to quantify and minimize this pernicious effect, the usual practice is to model the air system with one-dimensional network solvers where the net nodes represent the various components of the system. For usual engine transients, it is sufficient to analyse the system performance with a quasi-steady approach because the time constant of the air system is two orders of magnitude smaller than the turbomachinery characteristic time. Nonetheless, the rapid changes that occur during certain transient or failure scenarios – particularly shaft failure events – call for a different approach to calculate the air system performance and the fluctuations of the turbomachinery endloads. However, there is no such approach available in the open literature to predict the transient response of the system. For steady-state conditions, the differential equations that govern the fluid evolution are not time discretized and thus can be solved in a relatively straightforward fashion. Moreover, unlike during transients, the flow is not supposed to reach sonic conditions anywhere within the network and, more importantly, flow reversal is not expected to occur. The aim of this study is to develop a model for gas turbine SAS dynamics capable of tackling the sudden changes in the flow properties that occur within the system in the rapid transient scenarios. The whole system is initially broken down into a series of chambers of a finite volume connected by pipes that are initially modelled in isolation and then interconnected. The single-component models and their assembly are successfully validated against numerical and experimental data. The resultant modular tool constitutes a baseline into which further improvements and modifications can be integrated in subsequent works. The usefulness of the computational tool is demonstrated through the analysis of the flow evolution within the SAS and the subsequent turbine endload during a shaft failure event.

The Secondary Air System (SAS) of a jet engine is an open system: there is a bleed off take from the compressor — at the lowest possible pressure compatible with the sink where the flow is to be discharged —, the air then travels through the internal cavities of the engine cooling down the compressor and turbine discs and sealing and cooling bearing chambers. Eventually, the air is discharged at the turbine rims preventing the air of the main gas path from entering the internal turbomachinery cavities that would damage the turbine assembly. Ultimately the system is also primarily responsible for determining the endloads exerted on the turbine discs. The amount of air bled from the main gas path, although necessary, impairs the engine performance because it is purged from the main engine cycle. In order to quantify and minimise its pernicious effect, the usual practice is to model the engine SAS in steady state conditions with 1D network solvers where the net nodes represent the various components of the system. For usual engine transients it is sufficient to analyse the system performance with a quasi steady approach because the time constant of the air system is insignificant compared with the turbomachinery characteristic time. Nonetheless, the rapid changes that occur during slam accelerations or failure scenarios — particularly shaft failure events — call for a different approach to calculate the endloads fluctuations. However, to the author’s knowledge, there is not such an approach to predict the transient response of the system available in the literature hitherto. The aim of the present research is to develop a dynamic model for gas turbine secondary air systems capable of tackling the sudden changes in the flow properties that occur within the system in the aforementioned cases. The whole system is initially broken down into a series of chambers of a finite volume connected by pipes that are initially modelled in isolation and then interconnected. The resultant tool constitutes a baseline onto which further improvements and modifications will be implemented in subsequent works. This first part of the paper explains the mathematical apparatus behind the model of the two main components of the SAS — chambers and pipes — isolated. Then the assumptions made and the limitations that arise as a result are described thoroughly. Finally, the computational results obtained are successfully compared against experimental data available in the public literature for validation.