Shock wave–boundary layer interaction
Hypersonic flows
Shock wave–boundary layer interactions on wedges, cones and wakes, from continuum to GPU solvers.
Research
HAVA Laboratory · Research
Research
We model flows across a wide spectrum: rarefied and kinetic, turbulent, magnetohydrodynamic, multiphase and hypersonic, with in-house GPU solvers and open-source codes, validated against experiments.
6research areas
–results
–verification & validation
In a nutshell
Our research on one page
The lab's research poster gives a one-page summary of our methods, solvers and results.
Open the poster (PDF)
Real-gas & nonequilibrium chemistry
Reacting flows
How hypersonic heating dissociates air and drives chemistry, and what it means for heat transfer and materials.
Free & impinging jets
Supersonic jets
Free and impinging underexpanded jets: shocks, expansion waves and shear-layer vortices.
Magnetohydrodynamics
MHD control
Using magnetic fields to reduce drag and heating on hypersonic vehicles during re-entry.
With CeFPaC at RPI
Flow control
Numerical flow control, validated with experiments at the Center for Flow Physics and Control.
Kinetic & continuum modelling
Rarefied flows
DSMC solutions of the Boltzmann equation and their comparison with Navier–Stokes turbulence models.
Results
Turbulent flow in a diffuser
Large-eddy simulation of subsonic turbulent diffuser flow, studied with the experiments of Prof. Amitay's group.
LESdiffuserCeFPaC
Turbulent flow in a diffuser
Subsonic turbulent flow in a diffuser computed with large-eddy simulation (LES), part of our work with the Center for Flow Physics and Control (CeFPaC) to validate numerical tools against the in-house experiments of Prof. Miki Amitay's group.
Shock wave–boundary layer interaction on a 5/30° double wedge
Three-dimensional Mach 7 flow over a double wedge at 781 Pa, showing spanwise velocity, numerical schlieren and surface heating.
Mach 7SWBLI3-D
Shock wave–boundary layer interaction on a 5/30° double wedge
Spatial distribution of the spanwise velocity W (m/s) on three x-constant planes, (a) x = 0.04 m, (b) x = 0.064 m and (c) x = 0.08 m, together with (d) the numerical schlieren in grayscale and the surface heat flux (W/m²) along the wedge surface, for a unit Reynolds number of 4.14 × 105 1/m (P∞ = 781 Pa) over the 5/30° double-wedge configuration.
Double cone in 2-D and 3-D
Translational temperature and surface heating from continuum SU2-NEMO 2-D and 3-D solutions at the lowest pressure.
double coneSU2-NEMOsurface heating
Double cone in 2-D and 3-D
Comparison of the surface parameters for a unit Reynolds number of 9.35 × 104 1/m: (a) spatial distribution of the translational temperature and (b) surface heating (kW/m²) from the continuum SU2-NEMO 2-D and 3-D solutions for the lowest-pressure case.
Three-dimensional hypersonic flow
Time-accurate three-dimensional hypersonic simulation.
3-Dunsteady
Three-dimensional hypersonic flow
Time-accurate three-dimensional hypersonic flow simulation.
Double-cone flow structure
Unsteady shock structures and their interaction with the flow around a double cone.
double coneshock interaction
Double-cone flow structure
Unsteady shock structures and their interaction with the surrounding flow around a double cone.
GPU wake simulation with HAVA-HYPERFDX1
A hypersonic cylinder wake on 0.3 billion grid points, run on a $3K desktop in under 15 hours.
HAVA-HYPERFDX1GPU0.3B grid points
GPU wake simulation with HAVA-HYPERFDX1
HAVA-HYPERFDX1 is the laboratory's in-house, high-fidelity GPU solver for hypersonic flows with strong shocks, separated regions and high-temperature gas effects. This wake simulation used about 0.3 billion grid points on a $3K desktop and finished in less than 15 hours, not on a large HPC cluster with thousands of nodes. More results are on the HAVA-HYPERFDX1 page.
2 figures
Magnetic drag reduction during re-entry
Placement of magnetic fields chosen by a design-driven simulation campaign to reduce viscous drag.
MHDdrag reductionre-entry
Magnetic drag reduction during re-entry
Through an extensive campaign of demanding simulations, ideal locations for the magnetic fields have been found with a design-driven approach. The drag reduction works by modifying viscous action at the surface, the main source of aerodynamic force on streamlined bodies at small angles of attack.
Rarefied cylinder wake, DSMC against Navier–Stokes
SPARTA DSMC and SU2-NEMO Navier–Stokes solutions for argon at Mach 7 and 100 Pa agree closely, also in the recirculation region.
SPARTASU2-NEMOturbulence models
Rarefied cylinder wake, DSMC against Navier–Stokes
Comparison of DSMC (SPARTA) and Navier–Stokes solutions for argon at a free-stream pressure of 100 Pa and Mach 7: (a) temperature field (K), (b) axial velocity (m/s), and (c) axial velocity along the stagnation line with several Navier–Stokes turbulence models in SU2. See the paper for details.
Real-gas chemistry in hypersonic flow
High-fidelity chemistry data built into our in-house and open-source solvers to benchmark real-gas effects.
dissociationreal gaschemistry
Real-gas chemistry in hypersonic flow
At hypersonic speeds, compression and friction heat the air strongly enough to dissociate molecules and start a range of chemical reactions. Our team integrates high-fidelity chemistry data into in-house and open-source solvers to benchmark these real-gas effects accurately.
2 figures
Underexpanded impinging jet
Shocks, expansion waves and shear-layer vortices of a supersonic jet impinging on a wall.
impinging jetNPRturbulence
Underexpanded impinging jet
Spatial variation of (a) the density gradient (kg/m⁴), (b) temperature (K), (c) Mach number and (d) turbulence kinetic energy (m²/s²) for the impinging-jet case.
Free expansion jet against Henderson (2005)
Velocity field and centreline profile of a free jet at NPR 4.03 compared with the experiment of Henderson (2005).
free jetNPR 4.03experiment
Free expansion jet against Henderson (2005)
Comparison of the velocity profiles for a free expansion jet with a nozzle pressure ratio of 4.03: (a) velocity magnitude (m/s) from the present work, (b) the experiment of Henderson (2005), and (c) the streamwise velocity along the centreline.
No matching items
Nothing matches these filters.