Design of a Test Rig for Organic Vapours

A blow-down wind tunnel for real gas applications has been designed to characterize an organic vapour stream by independent measurements of pressure, temperature and velocity. Experiments are meant to investigate flow fields representative of expansions taking place in Organic Rankine Cycles (ORC) turbines. Strong real gas effects, high Mach numbers and approximations affecting the calculated properties of novel compounds, make the knowledge of ORC turbine blade passage flow field still rather limited. A significant enhancement of turbines efficiency is expected from detailed investigations on expansion streams. Despite Organic Cycles have attracted large research efforts in recent years, present days design tools still suffer from a lack of relevant experimental data. So far, ideal gas test cases and equilibrium measurements have supported separately CFD and thermodynamic model validations. These considerations prove the relevance of such a test rig. This paper discusses the design and the final layout of the facility, whose construction is currently in progress. A straight axis supersonic nozzle has been chosen as test section for early tests; investigations on blade cascades are foreseen in the future. Due to high stream densities and temperatures, a throat size compatible with probes intrusion made a continuous cycle plant unaffordable, requiring an input thermal power of around 2.5 MW. A reduction to 30 kW has been achieved by adopting a blow-down tunnel: the fluid, slowly vaporized in a high pressure vessel, feeds the nozzle at a lower pressure. The vapour is then collected in a low pressure tank and condensed. The loop is closed by liquid compression through a pump. Such a batch operating system also offers the option to select test/condensation pressures and temperatures, allowing experimentation of a wide variety of working fluids, even though new ORC compounds (e.g. Siloxanes, Fluorocarbons) remain of major interest. Maximum temperature and pressure are 400 °C and 50 bar. Despite the unsteady operational mode, the inlet nozzle pressure can be kept constant by a control valve. Depending on the fluid and test pressure, experiments may last from 20 seconds to several minutes, while their set-up requires a few hours. Fast response pressure transducers, pressure probes and thermocouples have been selected for thermodynamic measurements; Laser Doppler Velocimetry (LDV) and Schlieren techniques allow direct measurements of velocity and flow visualization. The design has been carried out with a lumped parameter approach, using Siloxane MDM and Hydrofluorocarbon R245fa as reference compounds and FluidProp® for properties calculation.