https://www.mdu.se/

Compressor performance tests are mainly focused on the typical range of operation, resulting in limited knowledge of compressor behavior in the low-speed region. The main target of this work comprises the generation of compressor characteristics at low part-load by giving particular insight into the physical aspect of this operating condition. It is necessary for running transient and part-load performance simulation and can be considered as the first crucial step toward an optimal engine starting schedule. Modelling the low part-load operating regime requires accurate component performance maps extended to the low-speed area, where engine starting and altitude relight occur. In this work, a robust methodology for generating compressor maps in the low part-load operating regime is developed. Compressor geometry and typical operation range compressor map are required as inputs. Two different modelling processes are incorporated within this methodology. Extrapolation based on the principle of similarity laws with modified law exponents constitutes the first modelling process, which seems inaccurate when predicting compressor performance at fixed-rotor conditions. Interpolation based on the fixed-rotor characteristic constitutes the second modelling process, which can be either linear or adaptive. The adaptive interpolation scheme was developed by the authors and generates low-speed characteristics using the same allocation trend as the one obtained from given performance data. It is observed that performance data points of each β-line follow an exponential trend in mass flow differences while increasing rotational speed, with a calculated average relativized Root Mean Square (RMS) error of less than 5%. Adapting the same trend in mass flow to the low-speed region, a compressor performance map with continuous exponential trend in all characteristics (for part- and full-load conditions) can be achieved. Implementing the developed methodology on the High Pressure Compressor (HPC) of the Energy Efficient Engine (E3) project is also presented, showcasing its applicability and the merit of it being incorporated into any conventional performance prediction tool. Furthermore, a sensitivity analysis for input variables, namely compressor exit effective area and pressure loss model coefficients is carried out, demonstrating the significant impact of the former on the shape of the low part-load characteristics. Generation of compressor characteristics at low-speeds with this methodology can be viewed as an enabler for running credible transient starting simulation and transient diagnostics, thereby defining an optimal starting schedule, applicable to both power generation and aerospace industry.