Calibrated Low-Order Transient Thermal and Flow Models for Robust Test Facility Design

Abstract

This paper describes an upgrade to high temperature operation of the Engine Component AeroThermal (ECAT) facility, an established engine-parts facility at the University of Oxford. The facility is used for high-TRL research and development, new technology demonstration, and for component validation (typically large civil-engine HP NGVs). In current operation the facility allows Reynolds number, Mach number, and coolant-to-mainstream pressure ratio to be matched to engine conditions. Rich-burn or lean-burn temperature, swirl and turbulence profiles can also be simulated. The upgrade will increase the maximum inlet temperature to 600 K, allowing coolant-to-mainstream temperature ratio to be matched to engine conditions. This will allow direct validation of temperature ratio scaling methods in addition to providing a test bed in which all important non-dimensional parameters for aero-thermal behaviour are exactly matched. To accurately predict the operating conditions of the upgraded facility, a low order transient thermal model was developed in which the air delivery system and working section are modelled as a series of distributed thermal masses. Nusselt number correlations were used to calculate convective heat transfer to and from the fluid in the pipes and working section. The correlation was tuned and validated with experimental results taken from tests conducted in the existing facility. This modelling exercise informed a number of high-level facility design decisions, and provides an accurate estimate of the running conditions of the upgraded facility. We present detailed results from the low-order modelling, and discuss the key design decisions. We also present a discussion of challenges in the mechanical design of the working section, which is complicated by transient thermal stress induced in the working section components during facility start-up. The high-temperature core is unusually high-TRL for a research organisation, and we hope both the development and methodology will be of interest to engine designers and the research community.