Abstract Meeting the rising global demand for liquefied hydrogen will require a scale-up of liquefaction infrastructure. Higher plant capacities increase the viability of novel cycles and components, which can achieve improved performance. It has been shown that switching the final hydrogen expansion from a Joule-Thomson valve to a radial turboexpander (sub-cooled liquid phase) in series with a Joule?Thomson valve (two-phase) increases both yield and efficiency. This paper describes the design of a prototype turboexpander from an aerodynamic, manufacturing and stress perspective. The aerodynamic design is performed using an extended version of the open-source turbomachinery design code turbigen. Using a radial turbine meanline code and geometry parameters, the annulus and blade geometry are sent to a RANS solver with real gas property tables from CoolProp. This integrated process enables rapid investigation of the design space. Despite the challenging working fluid conditions, this paper shows that a conventional design methodology (developed for ideal gas radial turbines) can still be used, achieving an isentropic efficiency in excess of 90% for the baseline case. The aerodynamic design is then assessed against mechanical and manufacturability constraints. The design is modified by increasing blade thickness, by aft-loading, by adding fillets and by finding the optimum blade number. Incorporating the final turbine performance into a liquefaction cycle model confirms increases in yield of 10.7% and exergetic efficiency of 3.4% compared to the same cycle with a single Joule-Thomson valve expansion.
Gomez et al. (Fri,) studied this question.