1 Numerical simulation calculation

 

1 1 Basic equations and turbulence model

 

Automobile speed is generally much lower than the speed of sound. Automobile aerodynamics belong to low-speed aerodynamics. Therefore, the flow field around the automobile can be regarded as a three-dimensional incompressible viscous isothermal flow field. Due to the complex shape of the automobile, the surrounding airflow is likely to cause separation. Turbulent flow treatment. In this paper, the Relizab le k turbulence model is used for numerical simulation.

 

The basic governing equation for turbulence calculation is the three-dimensional incompressible Reynolds time-average Navier Stokes equation, and its governing equation is as follows:

 

(1) Continuous equation u i / x i = 0 (1)

 

(2) Momentum equation (uiuj) xi = xj eff uixj + ujxi-pxj (2) where: ui and uj are average velocity components, xi and xj are coordinate components, p is the pressure on the fluid micro-element body, and eff is Effective viscosity coefficient of turbulence.

 

In the Reynolds time averaging process for Navier Stokes equations, a new variable term u iu j (Reynolds stress term) is introduced to make certain assumptions about Reynolds stress in order to close the equations. The standard k model may lead to negative positive stress when the time-average strain rate is particularly large. In order to make the flow conform to the physical laws of turbulence, the positive stress is mathematically restricted. In the R elizable k model, the transport equation for k and is as follows.

 

Turbulence energy k equation: (k)t +(ku i)x i = x j + t

 

k x j + G k-(3) Energy dissipation rate equation of turbulent flow: ()t + (u i) x i = x j + t

 

xj + C 1 E-C 2 k + v (4) where: is the fluid density, G k is the production term of the turbulent energy k caused by the average velocity gradient, G k = tuixj + ujxiuixi (5) turbulent effective viscosity coefficient eff is calculated by the following formula: ef = + t; t = C k 2 /; C = 1 A 0 + A s U

 

k /; C 1 = max 0 43,

 

+ 5; = (2E ij E ij) 1 /2 k; E ij = 1 2 uixj + ujxi A s = 6cos; = 1 3 arccos (6W) W = E ij E jk E kj (E ij E ij) 1 /2; U

 

= E ij E ij +% ij% ij;% ij =% ij-2 ijk

 

k;% ij =% ij-ijk

 

In the formula k, %ij is the time-average rotation rate tensor observed from the reference frame of angular velocity k. Reference <6> recommends the following values:

 

k = 1 0, = 1 2, C 2 = 1 9, A 0 = 4 0.

 

In the wind resistance model, the near-wall surface deformation is very large. Many swirling flows are generated in the nearby area, and the pressure gradient deformation is large. Therefore, the steady state Relizable k. combined with the use of the wall function is more suitable for the numerical calculation of the automobile external flow field.

 

1 2 Model establishment

 

The basic model simulated is a certain foreign container semi-trailer truck. Due to the limitation of the current computer hardware, the model of the truck has been simplified appropriately, door handles, wipers, etc. are removed, and the bottom of the trailer is smoothed. The corresponding simplified model is used to simulate complex structures such as tire pattern, wheel hub, rear axle, frame, rearview mirror, etc., and a 13 scaled model is used for numerical simulation as a whole. The simplified model is as shown. According to the similar principle of fluid mechanics, the Reynolds number of the model in the numerical simulation should be equal to the Reynolds number of the actual situation. According to the experimental theory, the Reynolds number has a self-modeling zone. When the Reynolds number in the model test exceeds a certain value, the aerodynamic force remains basically unchanged. The Reynolds number of the automobile model test recommended by the Society of Automotive Engineers should be no less than 0 7 106, and the self-modeling Reynolds number recommended by Japanese experts is 0 5 10 6. The simulated wind speed used in this article is 30m/s, and the characteristic length is the larger wheelbase. , The Reynolds number of the model is 7 9 106. Therefore, the scaled model in this paper can be used to simulate the real car.

 

1 3 Mesh division and boundary condition setting

 

The outer contour of the calculation domain is a cuboid, and the car model is in a certain area of ​​the cuboid. The entrance is 3 times the length of the vehicle from the front end of the model, and the exit is 6 times the length of the vehicle from the rear of the model. The total height is 5 times the vehicle height and the total width is 7 times the vehicle width. The OCTREE method is used to generate an unstructured spatial grid in the entire calculation domain, and a triangular prism grid parallel to the surface of the car body is stretched, and a density box is used in the calculation sensitive area to achieve the purpose of local refinement and improve calculation accuracy. At the same time, in order to avoid the influence of grid changes caused by local changes of the model, the same grid size is set in the same area of ​​the model in different situations. The total number of grids generated in each simulation is about 2.8 million.

 

In the simulation process, set the uniform flow velocity at the inlet boundary u = 30 m/s, and the turbulence intensity at 0 5%; outlet pressure p = 0 (relative to atmospheric pressure); considering the influence of the bottom of the car on the airflow, set the floor as a slippery ground, The speed is the same as the speed of the incoming flow, u = 30m/s. The car body is set as a solid-wall non-slip wall boundary, and the left and right and upper surfaces of the calculation domain are set as a sliding wall boundary.

 

2 Analysis of CFD simulation results

 

Solving by the F luent software, the aerodynamic drag coefficient Cd of the vehicle can be obtained as 0 812 5. From the 2 body surface pressure cloud diagram, it can be seen that there is a large positive pressure area on the front face of the vehicle, and the front of the container is higher than the front of the vehicle. There is also a positive pressure zone on the steps of the car. Due to the high flow velocity on the lower part of the front face and the side of the car, the airflow separates, resulting in a negative pressure zone, and the pressure gradient is larger. It can be seen from the velocity streamline and velocity cloud diagram at the front of the vehicle that the airflow velocity is low at the gap between the rear of the cab and the container, and a large-scale vortex is generated behind the cab, where the airflow loses more energy. Since the top of the container is directly exposed to the incoming air, there is also air separation, and there is a backflow not far from the corner of the step. It can be seen from the velocity streamline and velocity cloud diagram at the rear of the vehicle that the airflow at the bottom is higher than the pressure at the top, and the airflow rushing from the bottom rolls upwards under the action of the pressure difference, forming a large-scale vortex.

 

3 Improvement of aerodynamic resistance characteristics

 

It can be seen from the above analysis that the aerodynamic drag characteristics of the car can be improved by installing additional devices. The paper analyzes the influence of structural parameters such as the fillet radius R of the deflector and the distance between the cab and the container L on the aerodynamic drag coefficient, and finds the R and L corresponding to the small aerodynamic drag, and then considers the best R and L. Combined with the influence on the aerodynamic drag coefficient.

 

3 1 R's improvement on aerodynamic drag coefficient

 

Schematic diagram of the corner radius of the diversion cover. Shows the law that the coefficient of aerodynamic drag changes with the radius R of the corner of the diversion cover. It can be seen from the curve in Figure 6 that as the fillet radius increases, the drag coefficient decreases gradually. When R = 500mm, the decrease is large; R decreases from 1 000 2 500mm is not obvious; R decreases from 3 00 03 500mm to a small amount.

 

And respectively are the pressure cloud diagrams of the body surface and the symmetry plane after installing the air deflector. It can be seen that due to the existence of the deflector, the positive pressure zone at the upper front of the original container has obviously disappeared, and the airflow does not separate here, but a local high pressure zone is generated at the right corner of the cargo box, and the pressure gradient is small. 9 is the velocity streamline and velocity cloud diagram of the symmetry plane at the front of the vehicle body.

 

It can be seen that the airflow flows smoothly from the upper part of the air deflector to the rear of the car, and the wake is formed by the shear of the airflow at the gap. However, compared with the original model (Figure 3), the scale of the vortex is smaller, and the center of the vortex is smaller than the original. When the model moves up, the energy loss is also small. Therefore, the pressure at the gap is relatively small, which effectively reduces the differential pressure resistance, and finally reduces the aerodynamic drag coefficient.

 

0 is the velocity streamline and velocity cloud diagram of the symmetry plane at the rear of the vehicle body.

 

It can be seen from 0 that after the installation of the deflector, the change of the wake vortex is not obvious. This is because the deflector is far away from the tail and has little effect on the flow field at the rear of the car.

 

3 2 L improvement of aerodynamic drag coefficient

 

1 and 2 are the schematic diagram of the gap between the driver's cab and the container and the influence of the gap change on the aerodynamic drag coefficient. It can be seen from 2 that as the gap gradually increases, the drag coefficient gradually increases. When L increases from 1 076mm to 1 126mm, this increasing trend is particularly obvious.

 

1 Schematic diagram of the gap between the cab and the container 2 The effect of the change in the distance between the cab and the container on the drag coefficient 3 and 4 are the pressure cloud diagrams of the body surface pressure and the symmetry plane respectively. It can be seen that, compared with the original model, the pressure distribution does not change significantly, and the reduction in spacing does not effectively reduce the pressure difference. 15 is the velocity streamline and velocity cloud diagram of the symmetry plane at the front of the vehicle body.

 

It can be seen from the figure that the characteristics of the flow field at the gap between the cab and the container are significantly improved. The airflow flows from the top of the cab. Due to the influence of the steps of the container above the cab, part of the airflow crosses the steps and flows backwards. The flow rate is too high. A strong air separation is formed, and a positive pressure zone is formed on the windward side of the container. At the same time, a backflow occurs not far from the corner of the steps, forming a vortex; another part of the airflow flows down the gap due to the blocking of the container. The reduction of the spacing causes the original larger shear vortex to decompose into two smaller vortices. Appeared above and below the rear of the cab respectively. Makes the airflow energy loss relatively reduced. Therefore, the drag coefficient is reduced to a certain extent. 5 The velocity streamline and velocity cloud diagram of the symmetrical surface of the front of the car body (L = 876mm) 6 is the velocity streamline and velocity cloud diagram of the symmetrical surface of the vehicle body rear.

 

It can be seen from the figure that compared with the original model, the tail flow field does not change significantly. The reason is the same as above. Since the gap between the cab and the container is far from the tail, the change of the gap has little effect on the wake.

 

3 3 The combination of R and L improves aerodynamic drag coefficient

 

6 The velocity streamline and velocity cloud diagram (L = 876mm) of the symmetrical surface at the rear of the car body are taken from the above simulation results to obtain the small drag coefficient R (500mm) and L (876mm) values ​​for comprehensive numerical simulation, and the drag coefficient is relative to the original The change amount C d of the model is -20 59%, which is significantly better than the effect of individual changes of each parameter.

 

7 is the surface pressure cloud map of the vehicle body. It can be seen from the figure that similar to Section 31, the positive pressure zone at the front of the container has obviously disappeared, effectively reducing the differential pressure resistance.

 

7 Body surface pressure cloud map (R = 500mm, L = 876mm) 8 Speed ​​streamline and speed cloud map of the symmetrical surface of the car body (R = 500mm, L = 876mm) 8 Speed ​​streamline and speed cloud of the symmetrical surface of the car body.

 

It can be seen from the figure that the front flow smoothly bypasses the deflector and flows to the rear. Due to the small distance, air flow backflow occurs between the cab and the container due to the shearing effect of the air flow, and it flows back without forming a complete vortex. The disappearance of the vortex reduces the energy loss of the airflow and the drag coefficient is further reduced.

 

9 is the velocity streamline and velocity cloud diagram of the symmetry plane of the rear of the vehicle body.

 

It can be seen from the figure that the tail flow field has not changed significantly compared to the original model. Because the deflector and the gap are far from the tail, the changes of various parameters have little effect on the tail flow field.

 

9 The speed streamline and speed cloud diagram of the symmetry plane of the rear of the car body (R = 500mm, L = 876mm)

 

4 Conclusion

 

(1) A strong blocking area is formed in front of the original model container.

 

It has a greater impact on the aerodynamic resistance of the vehicle.

 

(2) After installing the diversion cover, the aerodynamic resistance decreases as the radius R of the diversion cover becomes smaller, and the largest decrease is 19 34%.

 

(3) As the distance between the cab and the container decreases, the aerodynamic resistance becomes smaller. The big drop was 4 66%.

 

(4) The effect of installing a baffle on the improvement of aerodynamic resistance is significantly better than the effect of changing the distance between the cab and the container.

 

(5) Considering the good fillet radius of the air deflector and the good distance between the cab and the container, it is better than the influence of a single parameter change on the aerodynamic resistance, with a decrease of 20 59%.