Experimental and numerical investigation of outlet guide vane and endwall heat transfer with various inlet flow angles
Artikel i vetenskaplig tidskrift, 2016
This paper investigates the heat transfer on the outlet guide vane (OGV) surface and its endwall region. The Reynolds number is fixed at 300,000 and the flow is subsonic. The inlet flow angle is varied from +25 degrees (on-design), to +40 degrees and -25 degrees (off-design). Experiments were conducted in a linear cascade test facility using thermochromic liquid crystal technique. Numerical simulations using RANS were carried out with three turbulence models, i.e., standard k-omega model (k-omega), baseline k-omega model (BSL), and shear stress transport k-omega model (SST). Both the experimental and numerical results are provided and compared. On the OGV surface, boundary layer transition and separation affect the heat transfer significantly and they vary with the inlet flow angle. The abilities of the three models to predict these heat transfer behaviors are revealed. For the on-design case, both BSL and SST models capture the main feature of the heat transfer variations due to transition, but the k-omega model fails. For off-design cases where separation occurs, there are discrepancies found between the calculations and experimental data. On the endwall region, the effects of a horseshoe vortex (HV) on the heat transfer are clearly noticed at the leading edge (LE). The three models perform well to simulate the pitchwise averaged Nusselt number while they always over-predict the strength and size of the HV, which leads to higher heat transfer there compared to the measurements. For off-design conditions, the HV becomes more energetic than that of the on design condition and the pressure side leg departs from the OGV at the inlet flow angle alpha = -25 degrees
Engineering
passage
Heat transfer measurements
free-stream turbulence
Outlet guide vane
Mechanics
Thermodynamics
cross-flow
linear turbine cascade
Numerical simulations
aerodynamics
Endwall
model
blade
transfer predictions