Development of a design methodology for transonic rotor blades with ORC application

Abstract

Organic Rankine Cycles (ORCs) are one of the technologies that can play an important role in the reduction of green house emissions. By converting low temperature energy sources in electricity, they are suitable for the exploitation of renewable sources (as solar and geothermal) and industrial waste heat. One of the most critical components in ORCs is the gas turbine, which usually has a radial inflow and one single expansion stage. The difficulties of the turbine design are related to real gas effects of the working fluids (organic compounds)
and high expansion ratios, which lead to a supersonic flow at the turbine exit.
The objective of this work is to develop a blade design methodology for a transonic turbine rotor. This is done by setting the focus on the blade passage and shaping it as a rotating nozzle. First, theory of rotating nozzles is developed, assuming the flow to be one dimensional and isentropic. Relations with respect to chocked conditions and the analytic solution to the flowfield are derived. Afterwards, the blade designmethodology is developed based on the rotating nozzle theory. Inputs of the methodology are total conditions, mass flowand static pressure at the rotor inlet, static pressure at the rotor outlet. Both the theory developed and the blade design methodology are validated. The relations derived for a one dimensional flow through an isentropic nozzle are valid for an ideal gas, while validation for real gas is not carried out. The location of the physical throat and its cross sectional area are determined and the analytic solution method of the flow field is proved to be precise. The blade design methodology is based on a one dimensional approximation and represents a first step towards a more precise blade design: the flow conditions along the nozzle mid line follow the expected trend. However, the boundary conditions are not respected and the flow varies considerably far from the mid line. Additional levels of complexity (as 2D approximation) have to be implemented in future work.