Analysis of the aerodynamic performance of truck semi-trailers

In the last time, trucks have shown a very high interest in terms of their technical and fuel consumption performance. Prioritizing the phenomenon of having trucks that transport as much, as quickly and as economically as possible, this represent a new challenge for industry and engineering, which are in continuous development. The attempt to create a bridge as stable as possible between technical and consumer performance is supported by the multitude of areas that, based on studies and experimental trials, allow optimization without interfering, substantially, with the actual transport of the goods (loading capacities and transport). The main objective of this paper is to present the possibility of improving the aerodynamic performance of trucks, focusing on semi-trailers. For this study, a truck modelled at one to one scale was used, whose aerodynamic performance was verified in the virtual environment (CFD–Computational Fluid Dynamics); this environment has the possibility to test several geometric models without requiring the layout of a vehicle in physical form. To check the performance, a basic model was created, a conventional truck (tractor unit with semi-trailer), to which various semi-trailer modifications were made, such as: total and partial coverage of the side areas, idea to create a more aerodynamic profile of the semi-trailer in the back and the addition of fairings under the platform (underbody of the semi-trailer). The results describe the impact these elements have and allow the visualization of the possible future profile that is as aerodynamic as possible for semi-trailers (therefore also for trucks) from both theoretical and technical point of view. Through the proposed improvements the capacity and the transport of the goods is not affected, in a significant way so it can be said that there is the possibility of implementation in the industry.


Introduction
Transport of goods on the roads is mostly done with truck assembly of tractor unit with semi-trailer.This is due to the fact that the transport is done between longer distances, which involves the use of highways or another's high-speed roads; therefore a relatively high speed for trucks and this implies a higher fuel consumption.To prevent, or better said to minimize this negative impact which has implications in the economic part, over the time more and more possibilities have been studied to improve the aerodynamic performance of truck semi-trailers.Given the precise shape for this type of structure, rectangular one, the areas where optimizations can be made are quite limited.Zones with the biggest potential were and remain: the wheel housing, rear, side areas and the roof.Most of the cases for the rear part, on5 the ceiling, the changes are made by bevelling the last third in order to have the possibility of generating a good wake.But this modification is constrained by the way of loading the goods, which, most of the time, in done through the rear.In Figure 1 are presented two types of semi-trailers that mainly have the same areas covered: sides and improvements in the rear part.Image from Figure 1a has a more aerodynamic profile so a lower drag resistance; the second one, from Figure 1b, has a profile adapted to the current transport status.As was mentioned previously, Figure 2 shows a semi-trailer with aerodynamic improvements on the roof.In this case, although it seems that the rear area can reduce the access for loading goods, an alternative can be considered for the lateral side.In the actual research, the possibilities of improving the aerodynamic performance of semi-trailers will be analysed by working on the side and rear side of the roof.

Geometrical models and set-up
The geometric models used in this paper have a real scale and it was tried to be build up as close as possible to reality: were modelled with storage boxes, axles, shock bar and other elements that could influence the air flow in any way.Also the geometric model was inspired based on the global observation of European trucks, in circulation on public roads, without following a specific model.The areas where optimization was attempted are those mentioned in the rows above.The base geometry with which the simulations improvements will be compared is represented by a truck assembly consisting of a tractor unit and a semi-trailer without aerodynamic elements.In Figure 3 it is presented the base model without any aerodynamic elements on it.Based on it, several types of configurations will be tried in which different areas will be closed or optimized to obtain the best aerodynamic performance.Based on this model a virtual wind tunnel(test chamber) was built and tested in Ansys software with the following dimensions: length of 25738mm, height of 6744mm and width of 7600 mm.The dimensions were chosen to have fastest results after the simulations; in addition, an attempt was made to avoid influence of the air flow with the test chamber walls in order not to have erroneous results.As input data for the velocity of the air, the value introduced was of 150 km/h and the walls have condition of free slip walls.The speed value of 150 km/h was chosen only to have a better and clearer picture of the airflow.Exit parameter from the virtual wind tunnel is of 1 atm and air temperature it is considered at 298,15 K and density at 1,185 kg/m 3 .In terms of meshing, due to the fact that the simulations were performed on real scale models value for elements was of 90 mm.Regarding the power of calculation, the values for parameters are: 12 GB RAM, 2.50 GHz CPU, 2GB GPU.In Figure 4 it can be seen the test chamber where the base model was positioned and also the discretization (image from Figure 4b it is as a presentation model, the size of elements for meshing is 700 mm because the actual discretization of the model with 90 mm would be hard to be understandable).To characterize the flow of the fluid, in this case air, the Reynolds number can be used.In this way, it can be found out if the airflow is laminar, transient or turbulent.
As input parameters for calculation we have: airflow velocity v=41,66 m/s, length of the assembly L=16,238 m, dynamic air viscosity µ=1,837x10 5 kg/m•s, air density ρ=1.185 kg/m 3 correspondent to temperature of T=298,15 K. Based on formula (1) this number can be calculated and as a result it can into the following regimes: if Re<5x10 5 describes a laminar air flow, if Re is between 5x10 5 <Re<10 7 describes a transient air flow and if Re>10 7 describes a turbulent air flow.So, based on the value obtained from formula (1) the Reynolds number has the value of 4,36x10 7 which describes a turbulent airflow.[3], [4].In the following lines the configurations with the proposals for optimization are presented.The first aspect that was emphasized is the rear part of the roof of the semi-trailer.As can be seen in Figure 5, three configurations are presented: a -base model, b -starting from the base model, the roof was chamfered starting from the last axle of the semi-trailer, c -staring, also, from the base model, the roof was chamfered starting from the first axle of the semi-trailer.Considering these changes, the impact will be checked for configurations "b" and "c".In these cases, given the chamfered areas at the back, the two configurations have the potential to have much better aerodynamic performance than the base model.In addition, between the 3 iterations there are no changes for other areas of the semitrailer, so the impact will be specific to the roof.In the following lines the configurations with the proposals for optimization are presented.The first aspect that was emphasized is the rear part of the roof of the semi-trailer.As can be seen in Figure 5, three configurations are presented: a -base model, b -starting from the base model, the roof was chamfered starting from the last axle of the semi-trailer, c -staring, also, from the base model, the roof was chamfered starting from the first axle of the semitrailer.Considering these changes, the impact will be checked for configurations b and c.In these cases, given the chamfered areas at the back, the two configurations have the potential to have much better aerodynamic performance than the base model.In addition, between the 3 iterations there are no changes for other areas of the semi-trailer, so the impact will be specific to the roof.[5] 1) 2) 3) 4) Figure 6.Configurations for the sides: 1) full cover for sides, 2) partial cover for sides (open on wheel areas and straight profile in the back), 3) partial cover for sides and incline profile in the back, 4) partial cover o sides and straight partial cover in the back Figure 6 shows the configurations for the optimization possibilities on the side areas.As follows, a presentation of each configuration will be made: 1-starting from the base model "a", the side areas were completely closed and for the front part of the side panels an inclination (angle of 20 deg compared to X axis) was made to allow air a smoother attachment on the surface of the panels; 2 -starting from "1" configuration, the area that cover wheels was eliminated, and the rear part of the panels stopped near the shock bar; 3-starting from "2" configuration the only change made is in the rear area of the panels where an inclination in Y axis (to the interior of the semi-trailer with an angle of 10 deg compared to X axis) was added; 4-starting from configuration "2" the rear area of the panels was limited to cover only the storage boxes.All seven configurations will be checked both from the point of view of the air flow and by comparing the force on X axis.For the four simulations in Figure 6, the back part of the roof was considered standard, so the square shape without chamfering.[6]

Results
Based on the CFD simulations for the configurations presented in the previous chapter, an analysis will be made to see which iteration offers the most possibilities for improvement.As will be presented in the following lines, the analysis of the air flow on the 7 configurations will be done in 2 stages: the analysis for "a", "b" and "c" to observe the impact for the ceiling chamfer and the analysis for "1", "2", "3" and "4" to observe the different air flows for the side areas.

Figure 7. Comparison of airflow between configurations a, b and c
Figure 7 shows the air flow in vector form on a plane at the middle of the truck on the 3 types of semi-trailers.The significant impact of this chamfer in the back of the truck is visible.It can be considered that configuration "a" has the worst wake (big in X and Z axis), therefore the worst aerodynamic performance.As for iterations "b" and "c", it can be seen that the wake it is improved and also although "c" has a larger chamfer of the roof, it can be observed that the air detaches before the end of the semi-trailer and this leads to a slight imbalance of the wake.Between iteration "b" and "c" the wake for "c" it is smaller in Z axis, concluding that this has the best aerodynamic performance.Also based on the forces on X axis values shows the same idea: a -15046 N, b -14545 N, c -14574 N. Regarding the air flow in the sides of the truck, in Figure 8 are presented the 4 configurations with side panels in which the air flow is presented under form of vector on a plane positioned at the middle of the wheels.It can be observed that in all cases the wake is different because of airflow complexity.From a theoretical point of view, the best performance should be given by the situation in which the sides are completely closed, so the configuration "1".Of all the iterations, the smallest wake on X axis is given by the configuration "3" and this is because there is an inclined panel in the back, at the shock bumper.Although "3" can be considered the iteration with the best performance, to have even more robust image, it is necessary to consult the air flows in a plane on Y axis at the middle of the truck.Therefore, in Figure 9, based on the analysis of the wake, it can be observed that, indeed, the best performance is given by the configuration "1".In addition to this, the configuration "3", although it seemed to have a small wake, is quite unbalanced.Concluding from the air flows presented in Figure 8 and Figure 9, we can say that the best performance is given by configuration "1", where the wake is balanced in Y and also in Z. Also, configurations "2" and "4" are approximately at the same level of performance, but the "2" version has a more balanced wake.Configuration in which we have side panels inclined in the back ("3") describes an unbalanced wake on Y axis, which leads to the degradation of the aerodynamic performance.Regarding the forces on the X axis, the values are: d -14143 N, e -14274 N, f -14400 N, g -14287 N. Based on the numerical values, it is emphasized that the best aerodynamic performance is given by the configuration "1" and the worst one given by the configuration "3".Comparing with the truck without improvements (configuration a), configuration "1" has an aerodynamic performance with 6% better than it.

Conclusion
As was presented, current studies to improve aerodynamic performance are successfully achieving their goal.The areas with the biggest potential for improving aerodynamic performance are the rear (roof) and the side areas of the semi-trailer.Regarding the chamfering of the roof of semi-trailer, it was observed that the generation of a geometric profile that is too aerodynamic will bring a degradation in performance.That is why it is necessary to establish the air detachment area very good, in order to create the best wake and at the same time, the chamfer (height) must be correlated with the loading capabilities of the semi-trailer.In addition to this aspect, closing the side areas completely generates a very good aerodynamic performance.Considering the above, it can be concluded that in order to have a small forward resistance as possible in the case of trucks, the analysis on semi-trailers must be done very well and specifically on certain areas.

Figure 4 .
Figure 4. a) Test chamber, b) Meshing of the test chamber and the base model

Figure 5 .
Figure 5. Configurations for the roof: a) base model, b) chamfered starting from the last axle, c) chamfered starting from the first axle

Figure 9 .
Figure 9.Comparison of airflow between configurations 1, 2, 3 and 4 [1] https://www.betterflow.com/futureflow/,accessed in June 2023 [2] https://www.cargobull.com/ro/presa/2021/ecogeneration,accessed in June 2023 [3] Scurtu I L and Gheres M I 2021 Numerical evaluation of vehicles aerodynamics in platoon using CFD simulation, Department of Automotive Engineering and Transports, Technical University of Cluj-Napoca, Romania [4] Harun C, Hazim M, Abdulkadir A, Iftekhar K, Firoz A, Simon W 2013 A study on aerodynamic drag of a semi-trailer truck, 5 th BSME International Conference on Thermal Engineering, Melbourne [5] Bukreev K R 2023 Shape optimization of teardrop trailers to minimize aerodynamic drag in articulated lorries, School of Mechanical Engineering Sciences, Faculty of Engineering and Physical Science, University of Surrey, Stag Hill, Guilford, UK, Vol 18, 2023 [6] Chilbule C, Upadhyay A and Mukkamala Y 2014 Analyzing the profile modification of truck-trailer to prune the aerodynamic drag and its repercussion on fuel consumption, Automotive Engineering, SMBS, VIT University Vellore, India [7] Cavusoglu O F 2017 Aerodynamic around wheels and wheel-houses, Chalmers University of Technology, Gothenburg, Sweden