Edwin Van Der Weide - Academia.edu (original) (raw)
Papers by Edwin Van Der Weide
Numerical simulations of active flow control have been carried out for the flow around the NACA00... more Numerical simulations of active flow control have been carried out for the flow around the NACA0018 profile for Mach = 0.15, Re = 2× 106, a = 15o using the Unsteady Reynolds Averaged Navier-Stokes (URANS) equations. Two types of flow control, zero-net-mass jets (synthetic jets) and continuously blowing jets, have been considered to delay the onset of separation. The synthetic jets have been applied to the 2D situation, i.e. infinitely long slits in the spanwise, for which the angle between the jet and surface normal has been varied to study the effect on the separation. For the continuously blowing jets the effect of 3D mixing is taken into account and an optimization of several jet parameters has been carried out to obtain the best result possible.
Recent developments in the SBP-SAT method have made available high-order interpolation operators ... more Recent developments in the SBP-SAT method have made available high-order interpolation operators (Mattsson and Carpenter, SIAM J Sci Comput 32(4):2298–2320, 2010). Such operators allow the coupling of different SBP methods across nonconforming interfaces of multiblock grids while retaining the three fundamental properties of the SBP-SAT method: strict stability, accuracy, and conservation. As these interpolation operators allow a more flexible computational mesh, they are appealing for complex geometries. Moreover, they are well suited for problems involving sliding meshes, like rotor/stator interactions, wind turbines, helicopters, and turbomachinery simulations in general, since sliding interfaces are (almost) always nonconforming. With such applications in mind, this paper presents an accuracy analysis of these interpolation operators when applied to fluid dynamics problems on moving grids. The classical problem of an inviscid vortex transported by a uniform flow is analyzed: the...
AIAA Journal, 2021
A structured Reynolds-averaged Navier-Stokes solver is directly coupled to a linear stability the... more A structured Reynolds-averaged Navier-Stokes solver is directly coupled to a linear stability theory (LST) solver to include the effect of laminar-turbulent transition in the flow simulations. The flowfield variables of the flow solver are used to both find streamlines along which transition can be predicted and to provide the LST code with the required boundary-layer profiles. Instabilities included in the analysis are of the Tollmien-Schlichting and crossflow nature relevant to high-Reynolds-number flows in low turbulence environments. The coupling is fully automated and can therefore be used efficiently in the analysis and design of geometries with external flows. The Technical University of Braunschweig's sickle wing with spanwise-varying crossflow and the natural laminar flow version of the Common Research Model are simulated under various conditions. Applications to these relevant three-dimensional test cases showcase the capability of the method to model the real flow physics. Advantages and challenges of the approach with regard to future design endeavors are discussed.
34th AIAA Applied Aerodynamics Conference, 2016
ρ = p RT0 The flow is governed by the Navier-Stokes equations with a Prandtl number of 0.71, spec... more ρ = p RT0 The flow is governed by the Navier-Stokes equations with a Prandtl number of 0.71, specific heat ratio γ = 1.4 and the bulk viscosity is assumed to be zero. Furthermore, the Mach number V0/c0 = 0.1 and the Reynolds number Re = ρ0V0L μ = 1600. The initial temperature is uniform, T0 = p0 ρ0R . The solution is computed on the periodic domain Ω = {−πL ≤ x, y, z ≤ πL} which is discretized using four uniform structured grids containing 65, 129, 257 and 513 vertices respectively. For the 65, 129 and 257 grids it was possible to use our local Linux cluster, while the 513 grid was run on up to 512 processors of the LISA machine of SARA, the Dutch Supercomputer Center. With a convective time scale tc = L V0 , the final time in the simulation is tfinal = 20tc. The classical 4 th
AIAA Scitech 2019 Forum, 2019
Achieving highly accurate solutions is one of major motivations in simulating fluid flows. Among ... more Achieving highly accurate solutions is one of major motivations in simulating fluid flows. Among various methods in Computational Fluid Dynamics, the discontinuous Galerkin (DG) method has gained popularity due to the ability to obtain higher order solutions on arbitrary grids. However, the method is not widely adopted (yet) in industry due to several challenges in the implementation. One of those challenges is capturing a discontinuity. Many researchers developed strategies to capture this abrupt jump in a flow robustly in the DG method. Most of them are using information from neighboring cells to sense a discontinuity, which violates the local character of the DG algorithm and therefore increases the complexity. Furthermore, additional communication is inevitable which degrades the scaling performance, especially on massively parallel platforms. In this work, we propose a simple discontinuity sensor completely local to an element. By using self-information, one can minimize commun...
it.uu.se
... 46 ENUMATH 2009 ENUMATH Stabilized Finite Element Method for Compressible-Incompressible Flow... more ... 46 ENUMATH 2009 ENUMATH Stabilized Finite Element Method for Compressible-Incompressible Flows Marie Billaud1, Gérard GALLICE1 and Boniface NKONGA2 1.CEA - CESTA, Le Barp, France 2.Université JA Dieudonné, Nice, France Marie Billaud marie.billaud@cea.fr ...
Energies, 2019
The discontinuous Galerkin (DG) method has become popular in Computational Fluid Dynamics mainly ... more The discontinuous Galerkin (DG) method has become popular in Computational Fluid Dynamics mainly due to its ability to achieve high-order solution accuracy on arbitrary grids, its high arithmetic intensity (measured as the ratio of the number of floating point operations to memory references), and the use of a local stencil that makes scalable parallel solutions possible. Despite its advantages, several difficulties hinder widespread use of the DG method, especially in industrial applications. One of the major challenges remaining is the capturing of discontinuities in a robust and accurate way. In our previous work, we have proposed a simple shock detector to identify discontinuities within a flow solution. The detector only utilizes local information to sense a shock/discontinuity ensuring that one of the key advantages of DG methods, their data locality, is not lost in transonic and supersonic flows. In this work, we reexamine the shock detector capabilities to distinguish betwee...
We show how a stable and accurate hybrid procedure for fluid flow can be constructed. Two separat... more We show how a stable and accurate hybrid procedure for fluid flow can be constructed. Two separate solvers, one using high order finite difference methods and another using the node-centered unstructured finite volume method are coupled in a truly stable way. The two flow solvers run independently and receive and send information from each other by using a third coupling code. Exact solutions to the Euler equations are used to verify the accuracy and stability of the new computational procedure. We also demonstrate the capability of the new procedure in a calculation of the flow in and around a model of a coral.
A hybrid method composed of finite difference-finite difference and finite differencefinite volum... more A hybrid method composed of finite difference-finite difference and finite differencefinite volume schemes for the time-dependent Navier-Stokes equations has been developed. A detailed analysis of the stability of the proposed algorithms, paying special attention to the stability of the interfaces between the subdomains is performed. We prove that the interface coupling is stable and conservative. This techniques makes it possible to combine the efficiency of the finite difference method and the flexibility of the finite volume schemes. We exemplify the procedure by using finite difference methods. The computational results corroborate the theoretical analysis.
Wind turbine aerodynamics can be broadly classified in the high Reynolds number and low Mach numb... more Wind turbine aerodynamics can be broadly classified in the high Reynolds number and low Mach number regime. Flows in this regime are generally incompressible and have large regions where they can be considered as inviscid. Thus, a great number of tools have been developed with incompressible and inviscid flow assumptions. However, as wind turbines designs become more complicated and more efficient, higher fidelity and more accurate ools like CFD are necessary. In this paper, a new open source pressure based incompressible RANS solver for wind turbine applications is introduced. The new solver is implemented within the open source multiphysics CFD suite SU2. A second order finite volume method is used for the space discretization and Euler implicit and explicit schemes for the time integration. Two turbulence models - the k-w mean shear stress model (SST) and the Spalart-Allamaras model, are available. A verification and validation study is carried out on the solver based on a number...
In an effort to implement non-reflective boundary conditions (NRBCs) in the context of the high-o... more In an effort to implement non-reflective boundary conditions (NRBCs) in the context of the high-order discontinuous Galerkin (dG) finite element method (FEM), the perfectly matched layer (PML) and the Navier-Stokes characteristic boundary condition (NSCBC) are considered for the compressible Navier-Stokes and Euler equations. A conservative-formulation Cartesian-based two-dimensional nodal dG solver and an entropy-formulation curvilinear-based three-dimensional modal dG solver are used. For the first, a low-storage fourth-order Runge-Kutta (LSRK4) is employed for time marching, while for the second, a strong-stability-preserving third-order Runge-Kutta (SSPRK3) is selected. Results include classical problems such as the isentropic vortex and the Kelvin-Helmholtz instability for the nodal solver, while a spherical pressure disturbance and a flow past a hump are considered for the modal solver. Both PML and NSCBC prove very promising in the context of the dG method. Future work will entail the development and testing of the PML in their viscous-term inclusion, as well as the compatibility conditions on edges and corners for the NSCBC on more rigorous test cases
This classic test case aims at characterizing the solver’s ability to preserve vorticity in an in... more This classic test case aims at characterizing the solver’s ability to preserve vorticity in an inviscid flow. The unsteady 2D Euler equations govern the simulation, which consists in a 2D vortex transported by a uniform flow across a rectangular computational domain of dimensions (x, y) = (0, Lx)× (0, Ly). The initial configuration of the vortex, centered in (xc, yc) and superimposed onto the uniform (infinity) flow, is given by the following equations:
We submit 5 sets of results: 1 for the inviscid subsonic case, 2 for the viscous case (with sharp... more We submit 5 sets of results: 1 for the inviscid subsonic case, 2 for the viscous case (with sharp and with rounded trailing edge), 2 for the transonic case (with and without shock capturing, see below for the detailed description of the shock capturing scheme). For this test case, we generated a fine O-grid of 577 × 513 vertices using the hyperbolic grid generation capabilities of the commercial software Pointwise [1]. The farfield is located at 1000 chords, as requested. The trailing edge is sharp, unless stated otherwise. For the subsonic configurations, the vertex distribution on the airfoil is the same on pressure and suction side, as shown in fig. 1. Instead, for the transonic configuration, vertices are clustered on the suction side, in particular close to the shock region, fig. 2. Initial guesses are obtained via grid sequencing, where appropriate. The coarser grids are obtained by deleting every other grid line from the finer grid (regular coarsening).
The decoupled solver was used to get an initial solution on the coarse grid and as main solver on... more The decoupled solver was used to get an initial solution on the coarse grid and as main solver on the medium and fine grids (see below for a description of the grid and the refinement strategy). The coupled solver was mainly used on the coarse grid. Since all development, tests, and runs have been performed in a relative short time, no tuning for perfomance has been carried out; hence, we expect the solvers not to perform at peak efficiency. For this test case, a fine O-grid of 513 × 169 vertices was generated with the commercial software Pointwise R © [2]. The farfield is located at 240 chords, and all results are obtained by applying a vortex correction [3]. This correction turned out to be crucial in order to limit the effect of the farfield position on lift and drag coefficients to less than 0.01 counts. For the given location, this is indeed the case. However, when the vortex correction is not applied, the farfield should be located several thousands of chord lengths away in or...
Out of a host of experimental aerodynamic data available for the DLR-F6 configuration a set of re... more Out of a host of experimental aerodynamic data available for the DLR-F6 configuration a set of reference test cases has been chosen for the second AIAA drag prediction workshop. Participants of the workshop were invited to use state-of-the-art tools for computational aerodynamics to compute these test cases and provide numerical data for integrated force coefficients and flow phenomena with the aim to assess the quality of today's flow solvers for transonic viscous flow about complex geometries. We present results for the block-structured flow solvers FLO107-MB and TFLO2000, which are based on numerical schemes by the third author. An incidence sweep and a grid refinement study at design conditions are shown, as well as an evaluation of the prediction of flow phenomena. Two different numerical schemes for convective fluxes and four different turbulence models have been used in this work. A comparison between the different numerical methods and turbulence models is particularly e...
Numerical simulations of active flow control have been carried out for the flow around the NACA00... more Numerical simulations of active flow control have been carried out for the flow around the NACA0018 profile for Mach = 0.15, Re = 2× 106, a = 15o using the Unsteady Reynolds Averaged Navier-Stokes (URANS) equations. Two types of flow control, zero-net-mass jets (synthetic jets) and continuously blowing jets, have been considered to delay the onset of separation. The synthetic jets have been applied to the 2D situation, i.e. infinitely long slits in the spanwise, for which the angle between the jet and surface normal has been varied to study the effect on the separation. For the continuously blowing jets the effect of 3D mixing is taken into account and an optimization of several jet parameters has been carried out to obtain the best result possible.
Recent developments in the SBP-SAT method have made available high-order interpolation operators ... more Recent developments in the SBP-SAT method have made available high-order interpolation operators (Mattsson and Carpenter, SIAM J Sci Comput 32(4):2298–2320, 2010). Such operators allow the coupling of different SBP methods across nonconforming interfaces of multiblock grids while retaining the three fundamental properties of the SBP-SAT method: strict stability, accuracy, and conservation. As these interpolation operators allow a more flexible computational mesh, they are appealing for complex geometries. Moreover, they are well suited for problems involving sliding meshes, like rotor/stator interactions, wind turbines, helicopters, and turbomachinery simulations in general, since sliding interfaces are (almost) always nonconforming. With such applications in mind, this paper presents an accuracy analysis of these interpolation operators when applied to fluid dynamics problems on moving grids. The classical problem of an inviscid vortex transported by a uniform flow is analyzed: the...
AIAA Journal, 2021
A structured Reynolds-averaged Navier-Stokes solver is directly coupled to a linear stability the... more A structured Reynolds-averaged Navier-Stokes solver is directly coupled to a linear stability theory (LST) solver to include the effect of laminar-turbulent transition in the flow simulations. The flowfield variables of the flow solver are used to both find streamlines along which transition can be predicted and to provide the LST code with the required boundary-layer profiles. Instabilities included in the analysis are of the Tollmien-Schlichting and crossflow nature relevant to high-Reynolds-number flows in low turbulence environments. The coupling is fully automated and can therefore be used efficiently in the analysis and design of geometries with external flows. The Technical University of Braunschweig's sickle wing with spanwise-varying crossflow and the natural laminar flow version of the Common Research Model are simulated under various conditions. Applications to these relevant three-dimensional test cases showcase the capability of the method to model the real flow physics. Advantages and challenges of the approach with regard to future design endeavors are discussed.
34th AIAA Applied Aerodynamics Conference, 2016
ρ = p RT0 The flow is governed by the Navier-Stokes equations with a Prandtl number of 0.71, spec... more ρ = p RT0 The flow is governed by the Navier-Stokes equations with a Prandtl number of 0.71, specific heat ratio γ = 1.4 and the bulk viscosity is assumed to be zero. Furthermore, the Mach number V0/c0 = 0.1 and the Reynolds number Re = ρ0V0L μ = 1600. The initial temperature is uniform, T0 = p0 ρ0R . The solution is computed on the periodic domain Ω = {−πL ≤ x, y, z ≤ πL} which is discretized using four uniform structured grids containing 65, 129, 257 and 513 vertices respectively. For the 65, 129 and 257 grids it was possible to use our local Linux cluster, while the 513 grid was run on up to 512 processors of the LISA machine of SARA, the Dutch Supercomputer Center. With a convective time scale tc = L V0 , the final time in the simulation is tfinal = 20tc. The classical 4 th
AIAA Scitech 2019 Forum, 2019
Achieving highly accurate solutions is one of major motivations in simulating fluid flows. Among ... more Achieving highly accurate solutions is one of major motivations in simulating fluid flows. Among various methods in Computational Fluid Dynamics, the discontinuous Galerkin (DG) method has gained popularity due to the ability to obtain higher order solutions on arbitrary grids. However, the method is not widely adopted (yet) in industry due to several challenges in the implementation. One of those challenges is capturing a discontinuity. Many researchers developed strategies to capture this abrupt jump in a flow robustly in the DG method. Most of them are using information from neighboring cells to sense a discontinuity, which violates the local character of the DG algorithm and therefore increases the complexity. Furthermore, additional communication is inevitable which degrades the scaling performance, especially on massively parallel platforms. In this work, we propose a simple discontinuity sensor completely local to an element. By using self-information, one can minimize commun...
it.uu.se
... 46 ENUMATH 2009 ENUMATH Stabilized Finite Element Method for Compressible-Incompressible Flow... more ... 46 ENUMATH 2009 ENUMATH Stabilized Finite Element Method for Compressible-Incompressible Flows Marie Billaud1, Gérard GALLICE1 and Boniface NKONGA2 1.CEA - CESTA, Le Barp, France 2.Université JA Dieudonné, Nice, France Marie Billaud marie.billaud@cea.fr ...
Energies, 2019
The discontinuous Galerkin (DG) method has become popular in Computational Fluid Dynamics mainly ... more The discontinuous Galerkin (DG) method has become popular in Computational Fluid Dynamics mainly due to its ability to achieve high-order solution accuracy on arbitrary grids, its high arithmetic intensity (measured as the ratio of the number of floating point operations to memory references), and the use of a local stencil that makes scalable parallel solutions possible. Despite its advantages, several difficulties hinder widespread use of the DG method, especially in industrial applications. One of the major challenges remaining is the capturing of discontinuities in a robust and accurate way. In our previous work, we have proposed a simple shock detector to identify discontinuities within a flow solution. The detector only utilizes local information to sense a shock/discontinuity ensuring that one of the key advantages of DG methods, their data locality, is not lost in transonic and supersonic flows. In this work, we reexamine the shock detector capabilities to distinguish betwee...
We show how a stable and accurate hybrid procedure for fluid flow can be constructed. Two separat... more We show how a stable and accurate hybrid procedure for fluid flow can be constructed. Two separate solvers, one using high order finite difference methods and another using the node-centered unstructured finite volume method are coupled in a truly stable way. The two flow solvers run independently and receive and send information from each other by using a third coupling code. Exact solutions to the Euler equations are used to verify the accuracy and stability of the new computational procedure. We also demonstrate the capability of the new procedure in a calculation of the flow in and around a model of a coral.
A hybrid method composed of finite difference-finite difference and finite differencefinite volum... more A hybrid method composed of finite difference-finite difference and finite differencefinite volume schemes for the time-dependent Navier-Stokes equations has been developed. A detailed analysis of the stability of the proposed algorithms, paying special attention to the stability of the interfaces between the subdomains is performed. We prove that the interface coupling is stable and conservative. This techniques makes it possible to combine the efficiency of the finite difference method and the flexibility of the finite volume schemes. We exemplify the procedure by using finite difference methods. The computational results corroborate the theoretical analysis.
Wind turbine aerodynamics can be broadly classified in the high Reynolds number and low Mach numb... more Wind turbine aerodynamics can be broadly classified in the high Reynolds number and low Mach number regime. Flows in this regime are generally incompressible and have large regions where they can be considered as inviscid. Thus, a great number of tools have been developed with incompressible and inviscid flow assumptions. However, as wind turbines designs become more complicated and more efficient, higher fidelity and more accurate ools like CFD are necessary. In this paper, a new open source pressure based incompressible RANS solver for wind turbine applications is introduced. The new solver is implemented within the open source multiphysics CFD suite SU2. A second order finite volume method is used for the space discretization and Euler implicit and explicit schemes for the time integration. Two turbulence models - the k-w mean shear stress model (SST) and the Spalart-Allamaras model, are available. A verification and validation study is carried out on the solver based on a number...
In an effort to implement non-reflective boundary conditions (NRBCs) in the context of the high-o... more In an effort to implement non-reflective boundary conditions (NRBCs) in the context of the high-order discontinuous Galerkin (dG) finite element method (FEM), the perfectly matched layer (PML) and the Navier-Stokes characteristic boundary condition (NSCBC) are considered for the compressible Navier-Stokes and Euler equations. A conservative-formulation Cartesian-based two-dimensional nodal dG solver and an entropy-formulation curvilinear-based three-dimensional modal dG solver are used. For the first, a low-storage fourth-order Runge-Kutta (LSRK4) is employed for time marching, while for the second, a strong-stability-preserving third-order Runge-Kutta (SSPRK3) is selected. Results include classical problems such as the isentropic vortex and the Kelvin-Helmholtz instability for the nodal solver, while a spherical pressure disturbance and a flow past a hump are considered for the modal solver. Both PML and NSCBC prove very promising in the context of the dG method. Future work will entail the development and testing of the PML in their viscous-term inclusion, as well as the compatibility conditions on edges and corners for the NSCBC on more rigorous test cases
This classic test case aims at characterizing the solver’s ability to preserve vorticity in an in... more This classic test case aims at characterizing the solver’s ability to preserve vorticity in an inviscid flow. The unsteady 2D Euler equations govern the simulation, which consists in a 2D vortex transported by a uniform flow across a rectangular computational domain of dimensions (x, y) = (0, Lx)× (0, Ly). The initial configuration of the vortex, centered in (xc, yc) and superimposed onto the uniform (infinity) flow, is given by the following equations:
We submit 5 sets of results: 1 for the inviscid subsonic case, 2 for the viscous case (with sharp... more We submit 5 sets of results: 1 for the inviscid subsonic case, 2 for the viscous case (with sharp and with rounded trailing edge), 2 for the transonic case (with and without shock capturing, see below for the detailed description of the shock capturing scheme). For this test case, we generated a fine O-grid of 577 × 513 vertices using the hyperbolic grid generation capabilities of the commercial software Pointwise [1]. The farfield is located at 1000 chords, as requested. The trailing edge is sharp, unless stated otherwise. For the subsonic configurations, the vertex distribution on the airfoil is the same on pressure and suction side, as shown in fig. 1. Instead, for the transonic configuration, vertices are clustered on the suction side, in particular close to the shock region, fig. 2. Initial guesses are obtained via grid sequencing, where appropriate. The coarser grids are obtained by deleting every other grid line from the finer grid (regular coarsening).
The decoupled solver was used to get an initial solution on the coarse grid and as main solver on... more The decoupled solver was used to get an initial solution on the coarse grid and as main solver on the medium and fine grids (see below for a description of the grid and the refinement strategy). The coupled solver was mainly used on the coarse grid. Since all development, tests, and runs have been performed in a relative short time, no tuning for perfomance has been carried out; hence, we expect the solvers not to perform at peak efficiency. For this test case, a fine O-grid of 513 × 169 vertices was generated with the commercial software Pointwise R © [2]. The farfield is located at 240 chords, and all results are obtained by applying a vortex correction [3]. This correction turned out to be crucial in order to limit the effect of the farfield position on lift and drag coefficients to less than 0.01 counts. For the given location, this is indeed the case. However, when the vortex correction is not applied, the farfield should be located several thousands of chord lengths away in or...
Out of a host of experimental aerodynamic data available for the DLR-F6 configuration a set of re... more Out of a host of experimental aerodynamic data available for the DLR-F6 configuration a set of reference test cases has been chosen for the second AIAA drag prediction workshop. Participants of the workshop were invited to use state-of-the-art tools for computational aerodynamics to compute these test cases and provide numerical data for integrated force coefficients and flow phenomena with the aim to assess the quality of today's flow solvers for transonic viscous flow about complex geometries. We present results for the block-structured flow solvers FLO107-MB and TFLO2000, which are based on numerical schemes by the third author. An incidence sweep and a grid refinement study at design conditions are shown, as well as an evaluation of the prediction of flow phenomena. Two different numerical schemes for convective fluxes and four different turbulence models have been used in this work. A comparison between the different numerical methods and turbulence models is particularly e...