Rim seals are critical in terms of limiting the temperature of highly-stressed engine components but function with a penalty to the power output and contribute to entropy gain stemming from mixing losses in the turbine. Ingress through rim seals is influenced by the presence of rotor blades and stator vanes, and the mainstream flow coefficient in the annulus that determines the corresponding swirl. This paper presents an experimental study of ingress upstream and downstream of the rotor disc in a 1.5-stage rig with double radial clearance rim seals. Two rotor discs were used, one with blades and one without, and two platforms were used downstream of the rotor, one with vanes and one without. Tests were conducted at two rotational speeds and a range of flow conditions was achieved by varying the annulus and sealing mass flow rates. Concentration effectiveness, swirl and steady pressure measurements separated, for the first time, the influence of the blades and vanes on ingress over a wide range of flow conditions. Measurements on the downstream stator platform provide added insight into the complex interaction between the egress and the mainstream. Measurements of unsteady pressure revealed the presence of large-scale structures, even in the absence of blades. The number and speed of the structures was shown to depend on the flow coefficient and the purge flow rate.

Rim seals are critical in terms of limiting the temperature of highly-stressed engine components but function with a penalty to the power output and contribute to entropy gain stemming from mixing losses in the turbine. Ingress through rim seals is influenced by the presence of rotor blades and stator vanes, and the mainstream flow coefficient in the annulus that determines the corresponding swirl. This paper presents an experimental study of ingress upstream and downstream of the rotor disc in a 1.5-stage rig with double radial clearance rim seals. Two rotor discs were used, one with blades and one without, and two platforms were used downstream of the rotor, one with vanes and one without. Tests were conducted at two rotational speeds and a range of flow conditions was achieved by varying the annulus and sealing mass flow rates. Concentration effectiveness, swirl and steady pressure measurements separated, for the first time, the influence of the blades and vanes on ingress over a wide range of flow conditions. Measurements on the downstream stator platform provide added insight into the complex interaction between the egress and the mainstream. Measurements of unsteady pressure revealed the presence of large-scale structures, even in the absence of blades. The number and speed of the structures was shown to depend on the flow coefficient and the purge flow rate.

Rim seals are fitted at the periphery of the stator and rotor discs to reduce the adverse effects of hot gas ingress on highly stressed turbine components limited by temperature. Ingress is induced by rotational effects such as disc pumping, as well as by asymmetric pressure-driven unsteady phenomena. These influences superpose to form a complex flow-physics problem that is a challenge for computational fluid dynamics. Engine designers typically use practical low-order models that require empirical validation and correlating parameters. This paper identifies the swirl ratio in the mainstream annulus as a dominant characterising parameter to predict ingress. This is a new interpretation that is supported by extending a low-order model based on turbulent transport using an effective eddy mixing length based on the difference in swirl between the annulus and seal clearance. Experimental measurements were made using a 1.5-stage turbine rig at low Reynolds number. The influence of annulus swirl ratio was investigated over a range of flow conditions and two rim-seal geometries, with the ingress quantified using CO2 tracer concentration in the sealing flow. The concentration data were complemented by measurements in the annulus using a five-hole aerodynamic probe.

Rim seals are fitted at the periphery of the stator and rotor discs to reduce the adverse effects of hot gas ingress on highly stressed turbine components limited by temperature. Ingress is induced by rotational effects such as disc pumping, as well as by asymmetric pressure-driven unsteady phenomena. These influences superpose to form a complex flow-physics problem that is a challenge for computational fluid dynamics. Engine designers typically use practical low-order models that require empirical validation and correlating parameters. This paper identifies the swirl ratio in the mainstream annulus as a dominant characterising parameter to predict ingress. This is a new interpretation that is supported by extending a low-order model based on turbulent transport using an effective eddy mixing length based on the difference in swirl between the annulus and seal clearance. Experimental measurements were made using a 1.5-stage turbine rig at low Reynolds number. The influence of annulus swirl ratio was investigated over a range of flow conditions and two rim-seal geometries, with the ingress quantified using CO2 tracer concentration in the sealing flow. The concentration data were complemented by measurements in the annulus using a five-hole aerodynamic probe.