Case Study No: FTLCFD03
NUMERICAL INVESTIGATION OF FLOW AROUND DrivAer CAR NOTCHBACK MODEL
Introduction
The continuous development of automobile aerodynamics, including studies on the DrivAer car, aims to enhance performance while addressing environmental and energy concerns in transportation. Wind tunnel testing and CFD analysis are widely used to evaluate vehicle aerodynamics [1]. While wind tunnel tests provide realistic flow behavior, they are often expensive, time-consuming, and offer limited data. In contrast, CFD simulations can generate detailed insights into flow variables across the domain more efficiently. To ensure accuracy, CFD results are validated against experimental data, often using simplified models like the Ahmed body and SAE model. These models help in understanding fundamental flow structures by isolating specific aerodynamic phenomena. However, they cannot fully capture complex flow interactions, such as those occurring at the underbody and around the wheels, due to their oversimplified geometry [2].
Investigating these effects on production vehicles is often impractical due to limited access and short development cycles. To address this, the Institute of Aerodynamics and Fluid Mechanics at the Technische Universität München (TUM), in collaboration with Audi AG and BMW Group, introduced the DrivAer model [3]. This realistic, modular vehicle model features interchangeable components, including different rear designs, wheels, under-body details, and mirrors, allowing for diverse aerodynamic studies [1]. In this study, the notchback version of the DrivAer model is used to validate CFD results against wind tunnel data.
DrivAer Car Geometry
To validate CFD data for the notchback model with wind tunnel data, aerodynamic parameters like the coefficient of drag (Cd) and lift (Cl) are compared. Also, the value for the pressure coefficient (Cp) for the car surface is calculated.
Figure 1: DrivAer Car Notchback Model
Computational Domain
The CFD analysis is performed for the 1:4 scaled model, and the computational domain around the car is modelled as per the previous studies [1]. For the present study, a symmetric model was considered for analysis purposes. The domain dimensions are defined with respect to car length (L), width (W), and height (H).
Figure 2: Computational domain and nomenclature for the surfaces
Mesh
The ANSYS Fluent is used for the meshing of the geometry. A poly hex core mesh is selected for the meshing purpose, and a few bodies of influence (BOI) are created around the car to capture the wake behind the car.
Figure 3: Mesh for the car body
Figure 4: Sectional view of the mesh around the car body
| Minimum element size | Total number of elements |
|---|---|
| 3 mm | 44 Million |
Table 1: Mesh statistics
Boundary Conditions
Air is used as the working fluid, and the table below represents the boundary conditions used for the different surfaces.
| Surface | Boundary Condition |
|---|---|
| Inlet | Velocity Inlet: 30 [m/s] |
| Outlet | Gauge pressure outlet: 0 [Pa] |
| DrivAer Car | No Slip Wall |
| Ground Wall | No Slip Wall |
| Side Wall | Symmetry |
| Top Wall | Symmetry |
Table 2: Boundary conditions details
Results
Steady-state results
The CFD data for the steady-state simulation is compared with the wind tunnel data. The parameters used for comparison are calculated as follows,
![\[{C_d} = \frac{F_d}{0.5 \times \rho \times U^2 \times A} \]](data:image/svg+xml;base64,PHN2ZyB2aWV3Qm94PSIwIDAgMjExIDQ5IiB3aWR0aD0iMjExIiBoZWlnaHQ9IjQ5IiB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciPjwvc3ZnPg== "Rendered by QuickLaTeX.com")
![\[{C_l} = \frac{F_l}{0.5 \times \rho \times U^2 \times A} \]](data:image/svg+xml;base64,PHN2ZyB2aWV3Qm94PSIwIDAgMjA3IDQ5IiB3aWR0aD0iMjA3IiBoZWlnaHQ9IjQ5IiB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciPjwvc3ZnPg== "Rendered by QuickLaTeX.com")
![\[{C_p} = \frac{P - P0}{0.5 \times \rho \times U^2} \]](data:image/svg+xml;base64,PHN2ZyB2aWV3Qm94PSIwIDAgMTcwIDQ5IiB3aWR0aD0iMTcwIiBoZWlnaHQ9IjQ5IiB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciPjwvc3ZnPg== "Rendered by QuickLaTeX.com")
Where,
C
= Drag coefficient
C
= Lift coefficient
C
= Pressure coefficient
F = Total force (pressure and viscous force) in x-direction
F = Total force (pressure and viscous force) in z-direction
ρ = Free stream density [kg/m3]
U = Free stream velocity [m/s2]
A = Reference area [m]
P = Static pressure at given location [Pa]
P0 = Static pressure in freestream [Pa]
| Parameter | Experimental | CFD | Error [%] |
|---|---|---|---|
| Cd | 0.26 | 0.27 | 3.8 |
| Cl | 0.12 | 0.11 | 8.3 |
Table 3: CFD result comparison with wind tunnel
Figure 5: Velocity contour at the symmetry plane across the car
Figure 6: Streamline over the car body
Figure 7: Pressure coefficient along centerline (top and bottom surfaces)
Transient results
Transient simulation data provides more insights into the time-dependent effect on the flow variables. The image of the q-criterion with vorticity plot represents the vortex shedding behind the car.
Figure 8: Q-criterion with vorticity contour around the car body
Transient velocity contours at the central plane around DrivAer Car
References
- Qin, Peng, Alessio Ricci, and Bert Blocken. “CFD simulation of aerodynamic forces on the DrivAer car model: Impact of computational parameters.” Journal of Wind Engineering and Industrial Aerodynamics 248 (2024): 105711.
- Heft, Angelina I., Thomas Indinger, and Nikolaus A. Adams. “Experimental and numerical investigation of the DrivAer model.” Fluids Engineering Division Summer Meeting. Vol. 44755. American Society of Mechanical Engineers, 2012.
- Heft, Angelina I., Thomas Indinger, and Nikolaus A. Adams. Introduction of a new realistic generic car model for aerodynamic investigations. No. 2012-01-0168. SAE Technical Paper, 2012.
Explore our Case Studies and Publications
Gain insights from real-world CFD case studies demonstrating our technical expertise, problem-solving approach, and industry-driven solutions. Explore the full portfolio now.
Write to Us
If you are interested in consultancy services, please write to us with your requirements. We will get back to you and arrange for an online Teams/Zoom meeting to discuss further. Either fill out the form below or email us at consult@flowthermolab.com.