Simulation of flow over a Mercedes-Benz Actros Truck (Calculate drag and lift coefficients)

To begin with the simulation set-up, we need to import the CAD and then close the geometry by filling up the holes at the windshield and two windows. This is done using the 'Fill Holes' option in the Surface Repair window.

Closing the geometry

The next step is to check for any surface errors. The surface repair was run to check for any surface defects.

Surface Errors

637 Pierced Faces and 159 Manifold Edges were found meaning the use of wrapping the geometry was necessary.

Before wrapping the geometry, the Wind Tunnel (WT) block was created around the truck.

Wind Tunnel around the truck

While making the WT, the compression of the tyres due to self-weight was also considered and was assumed to be 5mm.

Accounting for tyre compression

The surface was wrapped with a Base Size of 0.03m. The geometry obtained after the initial wrap was clean as the base size was small and hence, other controls such as contact prevention were not needed.

Wrapped Surface

The efficacy of the wrapper was checked by running the surface repair for the wrapped geometry. The geometry was found to be free of the initially found pierced faces and manifold edges errors. The face proximity and face quality errors also reduced in number and will be taken care during the remeshing.

Cleaned Surface

After cleaning the geometry surface, we need to work on setting up the Boundary Conditions (BCs) and creating a Mesh.

To start with, assign the geometry and the block to regions. Make sure to create 1 Region for both the Parts and 'Create Boundary for each Part surface'.

Expand the Regions section to Boundaries and give the 'Velocity Inlet' and 'Pressure Outlet' BCs to the WT inlet and outlet respectively.

The following meshing models were selected -

Surface Remesher

Automatic Surface Repair

Trimmed Cell Mesher

Prism Layer Mesher

Due to the large size of the geometry and WT and limited computing power, a larger base size was selected for the outer WT domain. To capture the wake region correctly, a smaller block was created to capture the wake and flow around the truck properly. Custom Volumetric Controls were used for the same. Base size of 1.8m was used whereas for the volumetric controls, the custom size of 5% of the base size was used (0.09m). Other parameters related to Prism Layers were also altered a little and the maximum cell size was limited to 100% of the base size.

[Note: The geometry size could've been scaled down to a small size and create a smaller domain to improve simulation speed. However, since this was my first attempt at simulating a full-size geometry, I wanted to keep the size as it is. I also got a good idea about balancing the meshing and simulation physics and the computing power.]

To view the mesh, two separate planes were created to account for the turning of the truck. One was created through and parallel to the tractor and the other was created through and parallel to the trailer.

Volume Mesh - Parallel through the tractor

Volume Mesh - Parallel through the trailer

The next step is to set up the physics models under the Continua tab. Following physics models were selected -

The inlet velocity value was set to 33.33 m/s which was given in the problem statement.

Stopping Criteria of 1000 iterations was set.

Before starting the simulation, reports to monitor drag and lift coefficients were set up. Frontal areas were calculated using frontal area reports and then provided as inputs to the force coefficient reports.

Two separate scalar scenes were also set up to capture the flow variables along the two planes created.

The simulation was then run till the stopping criteria was reached.

Results:

1. Residuals -

The residuals had reached a value between 1e-2 and 5e-4 and stabilised indicating that the steady-state was obtained.

2. Scalar Scenes -

Velocity and Pressure Contours along the plane parallel to the tractor

Velocity and Pressure Contours along the plane parallel to the trailer

Pressure Distribution along the truck body

The scalar plots help us to visualise the distribution of flow variables around the truck with regions of high and low pressures, recirculation zones.etc which are vital for analysis of exisiting models and designing improved versions.

3. Drag Coefficient -

The drag coefficient is a dimensionless number that is a ratio of the force acting on the body to the dynamic pressure. It helps us to quantify the drag or resistance on an object in a fluid environment. Higher the drag coefficient, higher is the drag/resistence and vice-versa. Drag coefficient plays a vital role in designing of wings or automobiles, as the drag coefficient for different surface areas and shapes of the object can be tested by keeping other flow variables constant. Cd = 0.99446

4. Lift Coefficient -

Similar to the drag coefficient, the lift coefficient quantifies the amount of lift generated. For aircrafts, a higher lift coefficient is desirable whereas for automobiles, especially sports cars, a lower lift coefficient is desirable. Cl = 0.25862

5. Pressure Coefficient -

Pressure Coefficient is a dimensionless number which describes the relative pressures throughout the fluid domain. In many situations in aerodynamics and hydrodynamics, the pressure coefficient at a point near a body is independent of body size. Consequently, an engineering model can be tested in a wind tunnel, pressure coefficients can be determined at critical locations around the model, and these pressure coefficients can be used with confidence to predict the fluid pressure at those critical locations around a full-size object.

where,

P = Static Pressure at the point at which pressure is calculated

P (infinity) = Static Pressure in freestream

P0 = Stagnation Pressure in freestream

A separate field function was created along with two probes, one in the freestream and the other in the stagnation zone. The above mentioned formula is fed in the Definition section of the newly created field function.

Location of the two probes in the fluid domain

Pressure Coefficient distribution along the domain length

Pressure coefficients can be found experimentally using wind tunnels which can then be verified with the flow simulations to validate the CFD analysis.

6. Streamlines -

Streamlines are used to visualise the flow and help in analysing flow recirculation and flow through critical sections of the geometry.

Streamline Videos -

CASE - 2: Only Tractor

In this case, only the front tractor was simulated and the trailer components were not used. All other parameters such as WT size, physics continua were kept same. Only the mesh values were reduced to create a finer mesh.

Surface Wrapped geometry

Mesh

Same physics models were selected with inlet velocity of 33.33m/s.

The simulation was run for 1000 iterations and same plots were set up as before.

Results -

1. Residuals plot -

2. Scalar Plots -

Velocity and Pressure Contours

Pressure distribution along the tractor body

3. Drag Coefficient -

Cd = 0.76770

4. Lift Coefficient -

Cl = 0.03899

5. Pressure Coefficient -

6. Streamlines -

Streamline Videos -

CONCLUSION:

• The residuals were stabilised and steady-state was reached in both the cases.
• Drag Coefficient, Cd = 0.99446, was found to be very high in the first case. The drag coefficient value depends on a lot of factors such as wind tunnel size, flow physics.etc. In this case, the truck was taking a turn which increases the drag area, thus increasing the drag coefficient. In the second case, the Cd = 0.76770 was very low as compared to that of the first case. This is due to the fact that the trailer was removed, which reduced the surface area reducing the drag.
• Lift Coefficient was found to be Cl = 0.25862 whereas in the second case it dropped down to 0.03899. A low lift coefficient is always desirable in terms of automobiles as it increases the downforce which in turn increases the traction.
• Velocity streamlines were created to better visualise the airflow around the truck.
• Pressure coefficient function was created along with the use of point probes to capture the values.

Google Drive link for simulation files -

https://drive.google.com/open?id=1qq6quQYLFyU-xsrMJ-0IW-Z8aML6rg3J