1. Introduction

Environmental pollution is being discussed anew in Korea due to the increase in dirt on roads and fine dust in densely populated residential and commercial areas. The Korean Ministry of the Environment and related agencies are paying keen attention to the status of research and solutions to environmental pollution. Therefore, the need to develop a road-cleaning vehicle that can effectively remove fine dust and various dirt pollutants at low cost is urgent.^[1] Small vehicles with a small turning radius can effectively clean narrow roads and residential areas with the same performance as existing cleaning vehicles. The main frame of the vehicle developed in this study was designed as a pivot-type for smooth operation. The center of the main frame with a pivot-type connection is stressed by the bending moments due to the vehicle weight and the frame weight. In particular, the strength and durability of the pin, a key component that connects the front and rear parts of the frame, is critical for maintaining the performance and functionality of the main structure. Although traditional frame structural analysis and experiments have been focused on the overall deformation of the vehicle^[2], few studies have been conducted to examine the overall stress states caused by loads applied to vehicles in parallel with interpreting the experimental results. Accordingly, the intention of the present study was to conduct durability testing to verify the reliability of the vehicle frame. In this study, the frame stiffness and durability test conditions in consideration of the vehicle's weight and load were determined via testing. During the experiments, the strain on the main frame under applied load was checked via installed strain gauges, and the reliability was evaluated by comparison with the structural analysis results using the obtained strain data.

2. Experimental products and testing processes

2.1 Small E-sweeper main frame design

Figure 1 shows the small E-sweeper being developed in this research project. A key component, the main frame, was designed as a pivot-type with two pins in the structure to reduce the turning radius (Figure 1(a)). Figure 1(c) and 1(d) show the upper part and sides of the body frame. Most vehicle frames are constructed by welding, and the center of the frame in Figure 1(a) was designed to be 2 mm thicker than the other parts to reinforce vulnerable areas.

2.2 Testing the characteristics and durability of the main frame

Durability testing of the main frame was carried out by loading at three locations (front, center, and rear) using three linear hydraulic testing machines (MTS). As shown in Figure 2, durability testing was conducted by separating the load application point because the frame load varies depending on the road surface, the progress of the vehicle, and the reaction force of the pivot structure when the load is applied to the front, center, and rear of the frame.

Three types of loading conditions were applied to examine the structural characteristics of the main frame, as shown in Figure 3. Experimental and structural analysis was performed depending on the vertical load direction; this can be affected by roll/yaw motions (at a speed of 4 km/h) of the small electric vehicle, and thus it is important to ensure the reliability of the vehicle frame according to the vertical load.

The load directional component was defined as shown in Figure 3, with the compression load being set to (-). Load case I (Figure 3(a)) was used to assess the strength of the pivot part, as well as to examine the strain and deformation caused by the reaction force on the fixed part by applying loads to the front, center, and rear parts of 2,500, 10,000, 4,500 N, respectively. Load case II (Figure 3(b)) was used to inspect the stiffness and deformation of the main frame by applying a force of 10,000 [N] to the front load points. Last, load case III (Figure 3(c)) was used to evaluate the stiffness and deformation of the main frame by applying 10,000 N to the rear load points. The characteristics of the main frame were tested under each set of load conditions, after which the results were compared with the structural analysis results.^[3], ^[4], ^[5], ^[6]

Durability testing to ensure the reliability of the main frame was carried out with the same load points as in the characteristic testing (Figure 3) but with simultaneous loading because proceeding with the same load conditions (Figure 4) as the characteristic tests would have been time-consuming and applying the loads at the same time was harsher. The tests were conducted using three linear hydrolytic test machines and a data logger. Figure 5 presents the hydraulic equipment used during the testing; the data logger for the strain measurements; and load conditions of 10,000, 9,000, and 8,500 N at the front, rear, and center, respectively, at 0.5 Hz.

2.3 Main frame structural analysis

Figure 6 shows the model used in the simulation process. Simulation model optimization and simplification were accomplished by removing unnecessary parts from the initial design frame model followed by remodeling using CATIA for modeling modification and ANSYS for the structural analysis.^[7], ^[8], ^[9], ^[10]

The load conditions were the same as those shown in Figure 3; the loading points and anchor points are shown in Figure 6. Furthermore, part-specific contact was established under the fully bound condition 6DOF(Degrees Of Freedom) fixed. The overall mesh state of the analytical model is shown in Figure 7(a), along with those for the rear (Figure 7(b)), center (Figure 7(c)), and front of the frame (7(d)). A shell model was utilized to create the mesh, and time in the analysis was saved by saving the mesh structures. In addition, parts with relatively complex structural features were partitioned using mesh and size features with no feature distortion. In particular, the parts that endure high strain and stress due to the load were designed with a slightly tighter mesh to obtain accurate values. The detailed nodes were 469,250 and the elements were 292,099. The material properties used in the structural analysis were STKR41 (the same material as the actual product), the material properties for which are summarized in Table 1.

4. Test and structural analysis results

4.1 Characteristic and durability test results

The experiments were conducted according to the set load command. In load case I, loads of 10, 2.5, and 4.5 kN were applied to the center, front, and rear of the frame, respectively. Moreover, a 10 kN load was applied to the front and rear part of the frame in load cases II and III, respectively. Figure 8 (a–c) graphically shows the load according to time recorded by the load cells during the testing of the main frame characteristics for the three different load cases.

Figure 9(a–c) shows the strain according to time measured using a strain gauge installed at each load point location for the three test types as the characteristic tests. The maximum strain in Figure 9(a) for test types I, II, and III were approximately 200, 600, and 200 με, respectively, indicating a change in strain with time depending on the location of the load points. It can also be seen that the strain values varied depending on the location of the installation, thereby indicating differences in the stress according to load in each part. The attachment locations of the strain gauges are shown in Figures 10(a) and 10(b). They were installed in areas with relatively large deformations when the loads were applied. When strain gauges can be installed on pins, which are important parts of the frame, the pin strength comparison evaluation is the most accurate, but this could not be carried out in the present study due to the locations of the other components. For this reason, the strain gauges were installed around the pin (Figure 10(a)).

Figure 11 shows the peak-valley load history in the endurance test. The maximum load forces of linear hydraulic test machines 1, 2, and 3 were 10, 9, and 9.5 kN, respectively. Tests were carried out with a load close to the initial set point and after the durability and terminated without cracks or fractures.

4.2 Structural analysis results

As shown in Figure 12, the distribution plot of stress varies depending on the load conditions. The stress pattern for load case I in Figure 12(a) shows the distribution of stress throughout the frame by applying a load to the center and both ends of the frame. The stress was concentrated at the anchorages where reaction forces were applied according to the authorized load. Figure 13(a) shows the stress pattern for the pin in load case I; since the peak stress was 458 MPa and the yield stress was more than 410 MPa, it was necessary to improve the stress concentration by changing the pin or structure. For load case II in Figure 12(b), the stress was concentrated on the front of the frame and peak stress was generated at the anchorage points; structural analysis of the stress results for the pin in Figure 13(b) concerning peak stress in the z-direction signify that strength improvement was required and led to a change of material for the pin. The stress pattern for load case III in Figure 12(c) shows that there was no stress concentration at the front and center parts but significant stress in the rear of the frame; the stress was concentrated as the center part gets closer due to the generation of momentum according to the authorized load at the rear of the frame. In Figure 13(c), peak stress on the pin was around 260 MPa, which was less than the yield stress; this needed to be mitigated by improving the shape of the rear of the frame rather than the strength of the pin.

To examine the feasibility of this structural analysis, the strain measured using the strain gauges was compared to that measured in the structural analysis. Figure 14(a) shows the structural analysis and experimental results for the strain according to time. The strain characteristics were different depending on the load conditions, as shown in Figure 14(a–c). Moreover, it was necessary to determine the cause of the difference between the structural analysis and the experimental results as they were notably inconsistent. For load case II (Figure 14(b)), a comparison between the experimental values and the numerical analysis results shows a large difference, although the fitted lines follow a relatively similar trend.

In contrast to load case II, the numerical analysis results for load cases I and III were lower than the experimental values. The reason for the different trends in load case II is that the stress was concentrated in the front of the frame (Figure 12) and there were different constraints between the experiment and numerical analysis. Hence, subsequently varying the constraints and thickness of the frame to make it lighter is required. Although the numerical analysis could not derive the same strain values as the experiment, the approach is still of great help for predicting the stress state of the structure.

5. Conclusions

For the first time, analysis of the main frame characteristics and durability testing to evaluate the body reliability of a small E-sweeper were conducted with authorized loads. Structural analysis was conducted to assess the effect of stress on a product that is difficult to obtain experimentally, the results of which are as follows.