Research Portfolio
Our research portfolio covers experimental and numerical investigations of thermofluids, aerodynamics and turbomachinery.
Novel endwall feature for secondary losses
Secondary flows are a major source of aerodynamic loss in turbomachinery. This research explores a novel endwall feature consisting of a channel embedded within the hub, extending from upstream of the blade passage to downstream of the trailing edge. 3D CFD simulations show that the concept substantially weakens secondary-flow structures, reducing core aerodynamic losses by approximately 30% relative to a smooth (baseline) endwall.
Mixed-convective flows during turbine shutdown
Increasing renewable penetration is driving conventional power-generation turbines toward more flexible operation with frequent start-ups and shutdowns. During shutdown, mixed-convection flows within turbine casing cavities can cause non-uniform cooling, thermal gradients, and differential radial growth. A novel experimental facility has been developed to investigate these transient flow and heat-transfer mechanisms under engine-representative conditions.
Thermal plume suppression inside turbine cavities
During turbine shutdown, mixed-convection flows within casing cavities can generate strong thermal plumes, causing non-uniform cooling, thermal stresses, and clearance variations. This research investigates passive flow-control strategies to mitigate these effects. Experiments, RANS, and wall-resolved LES demonstrate that an upstream blockage plate suppresses plume development and improves thermal uniformity by redistributing axial momentum while limiting the required ventilation flow.
MDO for aerodynamics, stability and performance of Missiles
Missile Aerodynamics Prediction and Parametric Optimization Tool is developed to enable designers to predefine aerodynamic, stability, and geometric feature variables as either targets or constraints. After optimization, a fully-defined STP file is ready to be exported.
Cast surface influence on heat transfer and aerodynamics
Realistic cast-surface roughness can significantly influence turbomachinery heat transfer but is difficult to represent in conventional CFD. Wall-resolved LES using profilometry-derived cast-iron topography shows increased wall shear, heat transfer, and near-wall turbulent transport compared with a smooth surface. Conventional RANS using an equivalent sand-grain roughness height underpredicts these effects, highlighting the limitations of simplified roughness modelling.
NNR model for turbulent heat flux and Reynolds stress treatment
RANS turbulence models remain the industry standard despite known limitations in complex mixed-convection flows. This research develops a physics-based nonlinear RANS framework for improved prediction of Reynolds stresses, turbulent heat fluxes, and heat transfer across varying buoyancy conditions. Developed using turbine shutdown flows, the model improves numerical robustness and predictive capability without introducing additional transport equations.