Table of Contents
Fetching ...

Separation induced transition in a low pressure turbine under varying compressibility

Priya Pal, Abhijeet Guha, Aditi Sengupta

Abstract

The present study investigates influence of compressibility on separation induced transition in a low pressure turbine cascade using high fidelity direct numerical simulations of the T106A blade. Simulations are performed for inlet Mach numbers, Ms ranging from 0.15 to 0.35 at a fixed Reynolds number and high incidence, representative of off design LPT operation. A dispersion relation preserving numerical framework is employed to accurately capture instability waves, separation bubbles, and separation induced transition to turbulence. A comprehensive analysis is carried out using surface pressure and skin friction distributions, boundary layer integral parameters, spectral analyses, and budgets of compressible enstrophy. Increasing Ms systematically reduces streamwise extent of both leading edge and trailing edge separation bubbles and promotes earlier transition and reattachment, consistent with trends observed under increased free stream disturbances. Despite shorter separation regions, suction side momentum thickness at trailing edge increases from Ms = 0.15 to 0.35, indicating higher profile losses at elevated Ms. Spectral analyses demonstrate a redistribution of turbulent spatial and temporal scales, with energy injection occurring at progressively larger scales as Ms increases. Flow field visualizations reveal a transition pathway that shifts from two dimensional spanwise rolls and intermittent turbulent spots at low Ms to streak dominated, bypass like transition at higher Ms.

Separation induced transition in a low pressure turbine under varying compressibility

Abstract

The present study investigates influence of compressibility on separation induced transition in a low pressure turbine cascade using high fidelity direct numerical simulations of the T106A blade. Simulations are performed for inlet Mach numbers, Ms ranging from 0.15 to 0.35 at a fixed Reynolds number and high incidence, representative of off design LPT operation. A dispersion relation preserving numerical framework is employed to accurately capture instability waves, separation bubbles, and separation induced transition to turbulence. A comprehensive analysis is carried out using surface pressure and skin friction distributions, boundary layer integral parameters, spectral analyses, and budgets of compressible enstrophy. Increasing Ms systematically reduces streamwise extent of both leading edge and trailing edge separation bubbles and promotes earlier transition and reattachment, consistent with trends observed under increased free stream disturbances. Despite shorter separation regions, suction side momentum thickness at trailing edge increases from Ms = 0.15 to 0.35, indicating higher profile losses at elevated Ms. Spectral analyses demonstrate a redistribution of turbulent spatial and temporal scales, with energy injection occurring at progressively larger scales as Ms increases. Flow field visualizations reveal a transition pathway that shifts from two dimensional spanwise rolls and intermittent turbulent spots at low Ms to streak dominated, bypass like transition at higher Ms.
Paper Structure (9 sections, 13 equations, 14 figures, 3 tables)

This paper contains 9 sections, 13 equations, 14 figures, 3 tables.

Figures (14)

  • Figure 1: Schematic of flow inside a T106A low pressure turbine blade passage coloured with magnitude of vorticity contours.
  • Figure 2: Comparison of computed streamwise variation of time-averaged $C_p$ with the DNS results of Wissink wissink2003dns for $M_s = 0.15$. The top line represents the pressure distribution on the pressure surface, while the bottom line is the suction surface's $C_p$.
  • Figure 3: Iso-surfaces of Q-criterion, $Q = 50$ colored with the streamwise velocity, $u$ for test cases with $M_s = 0.15$, 0.20, 0.25 and 0.30, showing the pertinent coherent structures.
  • Figure 4: Iso-surfaces of spanwise vorticity, $\omega_z = -50$ colored with wall-normal vorticity, $\omega_y$ for test cases with $M_s = 0.15$, 0.20, 0.25 and 0.30, showing the relevant vortical structures.
  • Figure 5: Space-time plot of $\omega_y$ with indicated contour levels for test cases with (a) $M_s = 0.15$, (b) $M_s = 0.20$, (c) $M_s = 0.25$, (d) $M_s = 0.30$ and (e) $M_s = 0.35$ .
  • ...and 9 more figures