Prediction of the Unsteady Transition Behavior in Low Pressure Turbine Flows Using Time and Frequency Domain Methods

Abstract

The transient transition behavior of a two-stage low pressure turbine rig facility is investigated in terms of numerical studies. The surface of a second stage stator vane is equipped with a thin film sensor array along its suction side providing time-resolved measurement data of the underlying boundary layer. The measurement data indicate a laminar behavior over the accelerated region of the stator vane. At the decelerated region close to the vane’s trailing edge, alternating transition mechanisms of both bypass and separation induced transition combined with subsequent reattachment can be observed. The measurement data in combination with numerical results from a time-marching full-wheel simulation are used to assess the results from an unsteady flow simulation based on the harmonic balance approach in the frequency domain. For both time and frequency domain simulations, the turbulence behavior is considered by application of Wilcox’ k–ω two equation model in combination with Menter and Langtrys γ-Reθ transition model. The numerical results regarding transient distributions of intermittency and associated shape factor of the boundary layer are compared with proper measurement quantities in order to evaluate the capability of the applied harmonic balance solver to predict the unsteady transition behavior over the investigated vane’s suction side.

Prediction of Transient Pressure Fluctuations within a Low-Pressure Turbine Cascade Using a Lanczos-Filtered Harmonic Balance Method

“…This is achieved by comparing time-resolved measurement data conducted by surface thin film gauges within a 2-stage LPT test facility. The results presented in [10] support the harmonic balance general capability to reflect the unsteady transition behavior while taking advantage of the Lanczos-filter approach. However, in contrast to the pure correlation-based transition model applied in this effort, transition is modeled in [10] by relying on the solution of two additional transport equations.…”

“…The results presented in [10] support the harmonic balance general capability to reflect the unsteady transition behavior while taking advantage of the Lanczos-filter approach. However, in contrast to the pure correlation-based transition model applied in this effort, transition is modeled in [10] by relying on the solution of two additional transport equations. As usual in the context of Menter's γ − Re Θ t framework [30], one equation represents the intermittency γ and the second models the transition Reynolds number Re Θ t .…”

“…In recent studies of [10], the general capability to predict the unsteady transition if the Lanczos-filter method is applied, has been investigated. This is achieved by comparing time-resolved measurement data conducted by surface thin film gauges within a 2-stage LPT test facility.…”

“…Nevertheless, the question remains how taking advantage of the Lanczos-filter method affects the capability of the HB method to predict unsteady pressure fluctuations and transient turbulence mechanisms. The latter aspect is discussed in [10] where the prediction of unsteady transition mechanisms is validated by time-resolved hot film measurement data. Furthermore, the results from [10] support previous studies of [7,11] where the HB method's general capability to reproduce the transient transition behavior is assessed by data based on fast-response pressure transducers.…”

“…The latter aspect is discussed in [10] where the prediction of unsteady transition mechanisms is validated by time-resolved hot film measurement data. Furthermore, the results from [10] support previous studies of [7,11] where the HB method's general capability to reproduce the transient transition behavior is assessed by data based on fast-response pressure transducers.…”