Performance Development of Fluidic Oscillator Nozzle for Cleaning Autonomous-Driving Sensors

Abstract

Contaminated autonomous-driving sensors frequently malfunction, resulting in accidents; these sensors need regular cleaning. The autonomous-driving sensor-cleaning nozzle currently used is the windshield-washer nozzle; few studies have focused on the sensor-cleaning nozzle. We investigated the flow characteristics of the nozzle to improve its performance in cleaning the autonomous-driving sensor. The nozzle concept was based on the fluidic oscillator nozzle. Various performance parameters of the fluidic oscillator nozzle were selected and investigated. Transient fluid flow was simulated to determine the effect of the design parameters to maximize the oscillation flow phenomenon. Additionally, the spray angle and frequency were calculated. Analysis results showed that the change in flow speed affects the frequency, and the change in feedback-channel-inlet flow rate affects the angle change. To verify the simulation result, the three best models (R4+RC10, R6+RC11, R8+RC10) and the base model were manufactured and tested. The test results were consistent with the simulation results within a 6% error.

1. Introduction

The performance of vehicles is improving by combining various technologies. In the past, driving performance was the most important factor when judging a vehicle; nowadays, driving convenience options are important to customers [1,2,3]. Accordingly, automakers are developing various convenience options and technologies to increase vehicle sales. Autonomous driving is one of the options offering convenience that reduces driver fatigue. Thus, automakers have applied autonomous-driving technology to various vehicles. However, consumer complaints regarding frequent malfunctions of autonomous-driving sensors are increasing; these malfunctions often result in accidents. Various factors cause autonomous-driving malfunctions and accidents with the pollution of autonomous-driving sensors (Lidar, radar, cameras, ultrasonic sensors, etc.) contributing the most. Since the autonomous-driving sensor is exposed to the external environment, various pollutants (snow, ice, dead insects, dust, mud, etc.) deposit on the sensor surface, causing autonomous-driving malfunctions. Various studies have focused on eliminating this pollution. Won and Kang developed an algorithm that can determine the surface pollution of Lidar sensors in real time during vehicle operation and improve autonomous-driving accuracy by filtering pollution [4]. Ryu and Lee installed a sound wave generator on the back of the sensor glass and studied self-cleaning by removing pollutants using sound waves and vibration [5]. Kim et al. investigated the removal of fog, frost, snow, and water droplets by installing heating electrodes on autonomous-driving camera lenses [6]. Son et al. studied how various characteristics of pollutants such as blockage type, concentration, additives, and dryness affect blockage [7]. Figure 1 shows patents related to the cleaning of the autonomous-driving sensor [8,9,10,11,12]. Figure 1a–d show vehicle Lidar sensor-cleaning devices of the sliding, folding, up–down, and ultrasonic vibration wiper types, respectively, whereas Figure 1e shows the Lidar sensor-cleaning device. The common operating mechanism sequence of Figure 1a–d is largely divided into three steps. In the first step, cleaning liquid is sprayed through a cleaning nozzle installed at the top or side to initially remove pollutants attached to the sensor surface. In the second step, the ultrasonic sensor slides or folds to cover the autonomous-driving sensor and then generates ultrasonic vibration to remove pollutants from the sensor surface. The differences in the operation of ultrasonic sensors in Figure 1a–d are as follows. Figure 1a operates the ultrasonic sensor cover in the up and down direction through a gear attached to the side. Figure 1b folds up or down the ultrasonic sensor using an electric motor. Figure 1c moves the ultrasonic sensor in the up and down direction using electric motors on both sides. In Figure 1d, the ultrasonic sensor moves up and down through the lever attached to the side. In the third step, air is sprayed from an air nozzle to blow away pollutants and cleaning liquid from the surface of the autonomous-driving sensor and dry it. In Figure 1e, the cleaning liquid and air nozzles are arranged sequentially, and the surface of the autonomous-driving sensor is cleaned using a cleaning liquid and then dried with air.

When cleaning autonomous-driving sensors, nozzle performance is important because the sprayed cleaning solution physically cleans pollutants. In this study, the flow characteristics of the nozzle were studied to improve the cleaning performance of the nozzle in cleaning the autonomous-driving sensor. Currently, windshield-washer nozzles are used for cleaning autonomous-driving sensors, and few studies focused on nozzles for cleaning autonomous-driving sensors. Because the nozzle size is small, the error rate during precision machining is high, and the nozzle is sensitive to even minute changes in size. In this study, a fluidic oscillator nozzle shown in Figure 2 was used to remove pollutants adhering to the surface of the autonomous-driving sensor during autonomous driving. Figure 2 shows a patent for “apparatus for oscillating fluid injection” developed in our prior research. It has the advantage of widening the fluid injection direction and injection angle without a special device; thus, it was selected for this study. To study the nozzle for cleaning autonomous-drive sensors, we referred to studies on fluidic oscillators, which have the same internal structure [13,14,15].

A fluidic oscillator is a device that generates a jet that vibrates left and right even without an internal driving part. This fluidic oscillator has a feedback passage with a symmetrical structure formed by an internal block, and the flow supplied from the inlet generates a recirculation flow inside and ejects a vibrating jet through the outlet. A systematic design theory to investigate this phenomenon has not yet been reported. Therefore, the property variation of the fluidic oscillator due to internal shape changes needs to be investigated. Jung investigated the effects of fluid oscillator parameters on jet speed ratio and pressure drop [16]. The optimal shape of a fluidic oscillator for high jet speed and low pressure loss was determined using an optimal design technique. Park experimentally observed the vibration characteristics of a feedback-channel fluidic oscillator operated under supersonic conditions by applying a high-speed schlieren flow visualization technique [17]. Although research has been conducted on performance improvement using parameters like this, it is insufficient to construct specific designs. Therefore, it is necessary to develop a theory that can be applied to all fluidic oscillators [18,19,20,21]. Moon analyzed the results according to the FBC straight length, inlet wedge width, and mixing chamber thickness to analyze design variables that affect the frequency and spray angle [14,15]. This study aimed to analyze the internal flow and performance of the nozzle for cleaning the autonomous-driving sensor and improve its performance. In addition, we analyzed the influence of the sweeping angle and frequency according to the design variables and used them as the basic fluidic oscillator nozzle design factors.

2. Theoretical Background and Analysis Method

2.1. Theoretical Background

The Coanda effect is a phenomenon in which fluid moving around a wall sticks to the wall and flows along the wall without any additional equipment. The surface pressure along the curved wall changes depending on the conditions of the curved surface, and under certain conditions, the flow separates from the wall. The formula that expresses the fluid pressure and particle behavior of the Coanda effect is derived from the Bernoulli equation, and under the conditions shown in Equation (1), the pressure of the inviscid fluid becomes lower than the surrounding pressure.

(1)$\frac{\rho U^{2}b}{a}\le p_{\mathrm{\infty }}$

Here, $\rho$ is the density of the ejected fluid, U is the average flow speed, a is the radius of rotation of the curved surface, $p_{\infty }$ is the atmospheric pressure, and b is the spacing of the ejection slot [19,20]. During fluid ejection along the wall, the average velocity decreases and the surface pressure increases until the wall surface pressure becomes equal to atmospheric pressure. When the wall surface pressure $p_0$ = $p_{\infty }$, the flow separates from the curved surface. The Coanda effect can be explained by several physical parameters. Equation (2) was proposed by Newman [22] as the equation for the flow along the cylinder under Coanda flow at high Reynolds numbers.

(2)${\theta }_{sep}=f[{\left(\frac{\left(p_0-p_{\mathrm{\infty }}\right)·b·a}{\rho ·v^2}\right)}^{\frac{1}{2}}]$

Here, ${\theta }_{sep}$ is the angle at which the jet flow separates, ($p_0-p_{\infty }$) is the difference between the pressure of the wall surface and atmospheric pressure, b is the width of the slot, and a is the radius of curvature. It means that the flow separation angle ${\theta }_{sep}$ is a function of pressure difference, cylinder geometry, and fluid characteristics [23,24]. The slot width b and curvature radius a are important performance variables that affect the flow separation angle ${\theta }_{sep}$. Since the pressure difference is generally a given operating condition, the main design variables of this study were selected as the slot width and radius of curvature near the outlet.

2.2. Analysis Method

The performance of the nozzle in cleaning autonomous-drive sensors is largely divided into two categories: sweeping angle and frequency. The sweeping angle indicates the spraying operating angle. The larger the sweeping angle, the wider the range of pollutants that can be removed. With the increase in frequency of the fluidic oscillator, there is an advantage of reduced flow rate. Therefore, the nozzle-performance development was mainly performed from this perspective, and in developing the model representing the new performance, we referred to studies that analyzed the nozzle flow characteristics [7,8,9,10,11,12,13,14,15]. Figure 3 shows the base model of the nozzle selected in this study to clean autonomous-driving sensors.

2.3. Design Parameter Setting

Basic research was referred to for performance comparison according to the design variables of the sensor-cleaning nozzle [14,15]. In this study, the radius of outlet (R) and the radius of outlet curvature (RC) were set as the design variables to analyze the angle and flow rate of the nozzle for cleaning autonomous-driving sensors. The applied design variables are shown in Figure 4, and the variable-setting values are summarized in Table 1. Figure 4a is the radius of outlet and is a design factor of the R model. In Figure 4b, the circle indicated by the red dotted line is the radius of outlet curvature and is a design factor of the RC model. The R+RC model applies both the radius of outlet and the radius of outlet curvature. The variable with large angle change was set as the main variable, and additional analysis was conducted using a model (R+RC) combining the two main variables. A total of 25 models including the base model were analyzed.

2.4. Analysis Condition and Calculation Grid

The nozzle for cleaning autonomous-driving sensors used in this study was in an unsteady state in which crossflow occurred due to turbulence, and the continuity equation, momentum equation, and standard k-ω turbulence model were used. The fluid applied in the analysis was assumed to be washer fluid and single-phase flow. Figure 5a shows the boundary conditions of the nozzle for cleaning autonomous-driving sensors. The place marked with a blue circle is the inlet, the five sides marked with orange represent the outlet, and all other grays were walled. The inside of the nozzle was assumed to be atmospheric pressure as an initial condition. In the case of boundary conditions, the inlet specified the pump condition in the base model, 1.5 to 5 bar range and applied at 0.5 bar intervals. The outlet was applied at atmospheric pressure. The computational grid for calculating the flow area was created using the mesher function in STAR-CCM+ Ver.11.04 [25]. Polyhedral was applied to the grid type, and the base size is 0.2 mm. Prism layer mesh was applied to accurately calculate the rotational flow near the wall. Prism layer mesh is advantageous for calculating the rotational flow of fluid because a thin grating is created on the wall. The Wall y+ value is over 30 and is distributed at the nozzle outlet. The calculated error rate of the inlet and outlet mass flows was within 1%. The calculation grid settings are shown in Table 2, and Figure 5b shows the calculation grid creation result for the base model. The number of calculation grids ranged from 66,000 to 73,000 depending on each model.

3. Research Results and Considerations

3.1. Analysis Base Model

In the nozzle for cleaning autonomous-drive sensors, the flow goes back and forth between walls periodically due to the Coanda effect. Part of the main flow moves to the feedback-channel inlet ((d) in Figure 6). As the fluid flows inside the channel, the pressure increases, affecting the flow in the inlet area ((a) in Figure 6). The flow entering the feedback channel inlet ((d) in Figure 6) is released at outlet((f) in Figure 6) meets the main flow and forms a vortex, forming a reaction force and causing oscillations (gray circle in Figure 6). The flow meets the main flow and forms a vortex, forming a reaction force and causing oscillations (gray circle in Figure 6). If this flow pattern occurs periodically, a sweeping angle and frequency are formed. The sweeping angle was measured from when the crossflow stabilized. For frequency, the difference in total pressure and static pressure between the feedback-channel inlet ((d) in Figure 6) and the feedback-channel outlet ((e) in Figure 6) was calculated to determine the change in dynamic pressure, and the period was measured using the time difference. All internal flow patterns in this study were similar to the one shown in Figure 6.

In the base model, we confirmed how pressure conditions affect the sweeping angle and frequency. As a result of the analysis, it was observed that the frequency increased with the increase in inlet pressure. As the pressure inside the nozzle increases, the inlet fluid speed increases. When the fluid inlet speed increases, the pressure decreases according to Bernoulli’s theorem and flows into the feedback-channel inlet at a high speed. As the pressure inside the nozzle increases, the crossflow in the feedback channel proceeds faster and the frequency increases. The sweeping angle and frequency for each variable are shown in Table 3. To measure the sweeping angle, the velocity distribution according to the change in inlet pressure was measured as shown in Figure 7. As the nozzle internal pressure increases, there is little change in angle, but the frequency increases (Figure 8). In the base model, the frequency doubled as the inlet pressure increased from 1.5 to 5.0 bar. When the inlet pressure is 2.5 bar or less, the sweeping angle tends to increase. However, for 3 bar or more, the sweeping angle is fixed at 54°. However, the pressure variation has a minor effect on the sweeping angle. The inlet pressure is considered a major factor affecting the frequency.

3.2. Analysis for Radius of Outlet Model

Analysis of the main variable, radius of outlet, was performed based on the pressure condition of 5.0 bar, which showed the highest frequency in the base model. Table 4 shows the sweeping angle and frequency for each design variable model. The result of the analysis showed that the sweeping angle increased as the radius of the outlet decreased. This means that as the outlet area increases, the outlet velocity decreases compared to the base model, and the outlet pressure increases according to Bernoulli’s theorem. Therefore, an increase in the outlet pressure causes the sweeping angle to increase because it increases the amount of inflow flowing into the feedback-channel inlet. Figure 9 shows the speed distribution used to measure the sweeping angle of the autonomous-driving sensor-cleaning nozzle. It was observed that the frequency was constant despite changes in the radius of outlet. Figure 10 shows the comparison of the sweeping angle and frequency for each model. As the outlet area increases, the sweeping angle decreases, and the frequency tends to increase. The highest sweeping angle is 70° for R6, and the highest frequency is 333 Hz for R8. Therefore, it was difficult to analyze the effect because the frequency performance improvement of the radius of outlet models compared to that of the base model was less than 3%. Excellent performance improvement was obtained at an approximately 30% increase in the sweeping angle of the R6 model. It is believed that the radius variable in these outlets has a major effect on the sweeping angle.

3.3. Analysis for Radius of Outlet Curvature Model

Based on the pressure condition of 5.0 bar in the base model, the main variable radius of outlet curvature was analyzed as a design variable. Table 5 shows the sweeping angle and frequency results for each model. The sweeping angles for each radius of outlet curvature model were compared. It was found that as the radius of outlet curvature decreased, the flow rate at the outlet increased and the sweeping angle rapidly increased. This is because as the flow rate increases, the flow coming out of the feedback-channel outlet meets the main flow and generates more vortices than those of the base model. Due to these vortices, the reaction force pushing the main flow in the opposite direction increases, causing the sweeping angle to increase. Additionally, when the flow resistance pushes in the opposite direction, the flow is sprayed along the outlet wall due to the Coanda effect, which affects the increase in the sweeping angle. Therefore, the increase in the radius of outlet curvature has a major effect on the sweeping angle. As a result, the frequency was over 313 Hz. Figure 11 shows the sweeping-angle measurements. As the radius of outlet curvature increases, the sweeping angle decreases and the frequency increases. The sweeping angle and frequency for each model are compared in Figure 12. The highest sweeping angle was 63.5° in RC10, and the frequency was almost constant. Therefore, as the radius variable at the outlet decreases, the sweeping angle tends to increase. Compared to the base model, the sweeping angle of RC10 increased by 15%, showing excellent performance improvement. It is believed that the radius variable in this outlet curvature has a major effect on the sweeping angle.

3.4. Analysis for Radius of Outlet and Radius Outlet Curvature Models

The analysis of the main variables revealed that the models with the most increased sweeping angles were the R4 and R6 models of the radius of outlet and the RC9, RC10, and RC11 models of the radius of outlet curvature. Therefore, additional analysis was conducted by combining the two design variable models. The analysis was based on the pressure condition of 5.0 bar, which had the best frequency performance in the base model, and the sweeping angle and frequency of the analysis results for each model are shown in Table 6. Results show that the sweeping angle increased in the model with two design variables compared to the model with one. This further improved performance by increasing the amount of inflow into the feedback channel through the two variables, radius of outlet and radius of outlet curvature. The sweeping angle and frequency for each variable are shown in Table 6. The velocity distribution used to measure the sweeping angle for each model is shown in Figure 13. Figure 14 shows that the sweeping angle of the model with two design variables was larger than that of the model with one design variable. The highest sweeping angle was 75° for R8+RC10, and the frequency was almost constant. Compared to the base model, the sweeping angle of R8+RC10 showed excellent performance improvement with an increase rate of 39%. It is believed that the radius variable in this outlet curvature has an excellent effect on sweeping angle change.

3.5. Verification of Analysis Results through Experiments

The three models with the best sweeping angles for each design variable and the base model were selected and tested. The models with the best sweeping angles are R4+RC10, R6+RC11, and R8+RC10. Therefore, three models and the base model were produced in aluminum. The produced models are shown in Figure 15. The experimental device is shown in Figure 16a, and the experiment was performed under a pressure of 5 bar. The experimental device is shown in Figure 16b.

The spray experiment was conducted five times, and the average was calculated excluding the maximum and minimum values. An image of the experiment results is shown in Figure 17. In Figure 17, the sweeping angle was measured using a protractor and arrows. The average sweeping angle measured through the experiment is as follows. The base model was measured at 54°, R4+RC10 was measured at 70.6°, R6+RC11 was measured at 70.3°, and R8+RC10 was measured at 73.3°. When comparing spray angles by model, the R8+RC10 model had the widest spray angle of 73.3°. The measured sweeping angle and frequency are shown in Figure 18, and are compared with the simulation results. The comparison of the experimental and computational results shows similar results with the sweeping angle error rate being within approximately 6% and the frequency error rate being approximately 15%. This is believed to be because the design factors (radius of outlet and radius curvature of outlet) were manufactured differently from the design values due to the surface treatment and processing errors when manufacturing the prototype nozzles. Therefore, when manufacturing nozzles, the sweeping angle and frequency may differ from those of the computational models due to slight manufacturing errors. Therefore, autonomous-driving sensor-cleaning nozzles must be manufactured with high accuracy.

4. Conclusions

This study determined the main design factors of the sweeping angle and frequency of the autonomous-driving sensor-cleaning nozzle through unsteady-state flow analysis. The sweeping angle and frequency were compared based on design factors. We optimized the sweeping angle and frequency through simulation and experiment by combining models that showed excellent performance. The following conclusions were drawn.

• Results of analyzing the pressure condition as a variable in the base model showed that the frequency doubled as the inlet pressure increased from 1.5 to 5.0 bar. However, the pressure variation had a minor effect on the sweeping angle. The inlet pressure was considered as the major factor affecting the frequency.

• The highest sweeping angle was 70° for R6, and the highest frequency was 333 Hz for R8. Therefore, it was difficult to analyze the effect because the frequency performance improvement of the radius of outlet models compared to the base model was less than 3%. Excellent performance improvement was obtained at approximately 30% increment in the sweeping angle of the R6 model. It is believed that the radius variable in these outlets has a major effect on the sweeping angle.

• Analysis of giving the radius variable to the outlet curvature showed that the highest sweeping angle was 63.5° in RC10, and the frequency was almost constant. Therefore, as the radius variable at the outlet decreases, the sweeping angle tends to increase. Compared to that of the base model, the sweeping angle of RC10 increased by 15%, showing excellent performance improvement. It is believed that the radius variable in this outlet curvature has a major effect on the sweeping angle.

• Additional analysis was performed using the radius variables of the outlet and outlet curvature with the highest sweeping angle for each design variable. The highest sweeping angle was 75° for R8+RC10, and the frequency was almost constant. Compared to that of the base model, the angle of R8+RC10 increased by 39%, confirming excellent performance improvement. It is believed that the radius variable in this outlet curvature has an excellent effect on the sweeping angle.

• To improve the accuracy of the analysis and confirm the physical phenomenon, models combining the radius variables of the outlet and outlet curvature were produced and tested. Comparing the experimental and analysis results, the error rate for the sweeping angle was within about 6%, which was close to the analysis; however, the error rate for the frequency was approximately 15%. In the case of frequency error, the values were different due to surface treatment and processing errors. Therefore, autonomous-driving sensor-cleaning nozzles must be manufactured with high accuracy.

Author Contributions

Software, J.-H.C.; Validation, C.-H.K.; Investigation, C.-H.K.; Writing—original draft, C.-H.K. and J.-H.C.; Writing—review & editing, S.-Y.P.; Supervision, S.-Y.P.; Project administration, S.-Y.P. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The data that support the findings of this study will be available from the corresponding author upon reasonable request. The data are not publicly available due to privacy.

Conflicts of Interest

Author Ji-Hyun Choi was employed by the company Soosan CSM. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1. Structures of patented autonomous-driving sensor-cleaning devices. (a) Lidar sensor-cleaning device for vehicle [8], (b) Lidar sensor-cleaning device for vehicle [9], (c) Lidar sensor-cleaning device for vehicle [10], (d) Lidar sensor-cleaning device for vehicle [11], and (e) Apparatus for cleaning Lidar sensor [12].

Figure 2. Structure of patented “apparatus for oscillating fluid injection”.

Figure 3. Three-dimensional (3D) model of nozzle for cleaning autonomous-driving sensors. (a) Side view. (b) Isometric view. (c) Top view. (d) Design variable name.

Figure 4. Design parameter. (a) Radius of outlet. (b) Radius of outlet curvature.

Figure 5. Boundary condition and calculation grid of base model. (a) Boundary condition. (b) Calculation grid.

Figure 6. Internal flow of the base model.

Figure 7. Sweeping angle with varying pressure.

Figure 8. Frequency and sweeping angle with pressure change in the base model.

Figure 9. Sweeping angle (radius of outlet models).

Figure 10. Sweeping angle and frequency (radius of outlet models).

Figure 11. Sweeping angle (radius of outlet curvature).

Figure 12. Sweeping angle and frequency (radius of outlet curvature).

Figure 13. Sweeping angle (radius of outlet and radius of outlet curvature models).

Figure 14. Sweeping angle and frequency (radius of outlet and radius of outlet curvature).

Figure 15. Nozzle test specimen. (a) Base. (b) R4+RC10. (c) R6+R11. (d) R8+RC10.

Figure 16. (a) Experimental equipment. (b) Experiment schematic.

Figure 17. Experimental sweeping angle. (a) Base. (b) R4+RC10. (c) R6+R11. (d) R8+RC10.

Figure 18. Error rate of simulation and experiment. (a) Frequency. (b) Sweeping angle.

Table 1

Design parameter value.

Model	Radius of Outlet(mm)	Radius of Outlet Curvature(mm)	Model	Radius of Outlet(mm)	Radius of Outlet Curvature(mm)
Base	0.2	1.5	RC12	0.2	1.2
R2	0.02	1.5	RC13	0.2	1.3
R4	0.04	1.5	RC14	0.2	1.4
R6	0.06	1.5	R4+RC9	0.04	0.9
R8	0.08	1.5	R4+RC10	0.04	1.0
R10	0.1	1.5	R4+RC11	0.04	1.1
R12	0.12	1.5	R6+RC9	0.06	0.9
R14	0.14	1.5	R6+RC10	0.06	1.0
R16	0.16	1.5	R6+RC11	0.06	1.1
R18	0.18	1.5	R8+RC9	0.08	0.9
RC9	0.2	0.9	R8+RC10	0.08	1.0
RC10	0.2	1.0	R8+RC11	0.08	1.1
RC11	0.2	1.1

Table 2

Calculation grid setting.

Item	Value	Unit
Mesh type	Polyhedral	(-)
Base size	0.2	(mm)
Number of prism layers	2	(-)
Prism layers thickness	0.05	(mm)
Prism layers stretching	1.5	(-)
Minimum surface size	0.04	(mm)

Table 3

Analysis results for base pressure parameter.

Pressure [bar]	Sweeping Angle [°]	Frequency [Hz]
1.5	55	161
2.0	56	188
2.5	57	212
3.0	54	238
3.5	54	256
4.0	54	277
4.5	54	303
5.0	54	322

Table 4

Analysis results for the radius of outlet parameter.

Model	R2	R4	R6	R8	R10	R12	R14	R16	R18
Sweeping Angle(°)	66	66	70	66	66	66.5	66	58	61
Frequency(Hz)	323	323	323	333	323	294	323	323	323

Table 5

Analysis results for radius of outlet curvature parameter.

Model	RC9	RC10	RC11	RC12	RC13	RC14
Sweeping Angle (°)	57	63.5	61	57.5	58	54
Frequency (Hz)	323	313	313	323	313	323

Table 6

Analysis results for radius of outlet and radius outlet curvature parameter.

Model	R4+RC9	R4+RC10	R4+RC11	R6+RC9	R6+RC10	R6+RC11	R8+RC9	R8+RC10	R8+RC11
Sweeping Angle(°)	68	73	72.5	67.5	72.5	73.5	70.5	75	73.5
Frequency(Hz)	333	323	323	323	313	323	323	313	323

References

1. Sjaanie, K.; Judith, C.; Brian, F.; Michael, F. How important is vehicle safety in the new vehicle purchase process?. Accid. Anal. Prev.; 2008; 40, pp. 994-1004. [DOI: https://dx.doi.org/10.1016/j.aap.2007.11.006]

2. Muhammad, W.; Petros, A.I. Automatic Vehicle Following Under Safety, Comfort, and Road Geometry Constraints. IEEE Trans. Intell. Veh.; 2023; 8, pp. 531-546. [DOI: https://dx.doi.org/10.1109/TIV.2022.3177176]

3. Shengnan, S.; Manisa, P.; Saifur, R. Grid Integration of Electric Vehicles and Demand Response With Customer Choice. IEEE Trans. Smart Grid; 2012; 3, pp. 543-550. [DOI: https://dx.doi.org/10.1109/TSG.2011.2164949]

4. Won, J.J.; Kang, Y.S.; You, B.Y. Real-time Contaminated Surface Determination Algorithm for Automotive LiDAR Sensors. Trans. Korean Soc. Automot. Eng.; 2023; 31, pp. 531-545. [DOI: https://dx.doi.org/10.7467/KSAE.2023.31.7.531]

5. Ryu, T.S.; Lee, S.M.; Lee, K.Y.; Chung, S.K. Self-Cleaning Drop Free Glass by Multiple Piezoactuators for improving Optical Sensor Visibility. Proceedings of the KSVI, Spring Academic Conference; Daejeon, Republic of Korea, 20 May 2021.

6. Kim, Y.K.; Lee, J.M.; Lee, S.M.; Chung, S.K. Thermally-driven Self-Cleaning Drop Free Glass for Automotive Cameras. Proceedings of the KSVI, Fall Academic Conference; Busan, Republic of Korea, 25 November 2021.

7. Son, S.H.; Lee, W.S.; Jung, H.G.; Lee, J.K.; Kim, C.R.; Lee, H.W.; Cho, S.W.; Jang, J.A.; Lee, M.; Ryu, H.C. Experimental Analysis of Various Blockage Performance for LiDAR Sensor Cleaning Evaluation. Sensors; 2023; 23, 2752. [DOI: https://dx.doi.org/10.3390/s23052752] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36904952]

8. Park, S.Y. Lidar Sensor Cleaning Device for Vehicle. Korea Patent; 10-2021-0184082, 21 December 2021.

9. Park, S.Y. Lidar Sensor Cleaning Device for Vehicle. Korea Patent; 10-2021-0187393, 24 December 2021.

10. Park, S.Y. Lidar Sensor Cleaning Device for Vehicle. Korea Patent; 10-2021-0184083, 21 December 2021.

11. Park, S.Y. Lidar Sensor Cleaning Device for Vehicle. Korea Patent; 10-2021-0187394, 24 December 2021.

12. Park, S.Y. Apparatus for Cleaning Lidar Sensor. Korea Patent; 10-2021-0192664, 30 December 2021.

13. Park, S.Y. Apparatus for Oscillating Fluid Injection. Korea Patent; 10-2273574, 30 June 2021.

14. Moon, H.S. Unsteady Computational Fluid Analysis of Fluid Vibration Characteristics of Windshield Spray Nozzle. Master’s Thesis; Kongju National University: Gongju, Republic of Korea, 2020.

15. Moon, H.S.; Park, S.H.; Kim, C.H.; Park, S.Y. Correlation Among the Flow Characteristics of Windshield Spray Nozzle through Unsteady Computational Fluid Analysis. Asia-Pac. J. Converg. Res. Interchange; 2023; 9, pp. 53-63. [DOI: https://dx.doi.org/10.47116/apjcri.2023.09.05]

16. Jung, H.S. Design Optimization of a Fluidic Oscillator Using 3-D Flow Analysis. Master’s Thesis; Inha University: Incheon, Republic of Korea, 2017.

17. Park, S.H.; Kang, M.J.; Lee, Y. Experimental Investigation of the Characteristics of a Supersonic Fluidic Oscillator with Feedback Channels. Trans. Korean Soc. Mech. Eng.; 2018; 42, pp. 541-549. [DOI: https://dx.doi.org/10.3795/KSME-B.2018.42.8.541]

18. Lee, S.Y.; Park, S.H.; Kang, M.J.; Lee, Y. Influences of the Internal Geometry of the Fluidic Oscillator on the Characteristics of Supersonic Sweeping Jet. Trans. Korean Soc. Mech. Eng.; 2019; 43, pp. 99-107. [DOI: https://dx.doi.org/10.3795/KSME-B.2019.43.2.099]

19. Jeong, H.S.; Kim, K.Y. Effects of Fluidic Oscillator Geometry on Performance. Korean Soc. Comput. Fluids Eng.; 2016; 21, pp. 77-88. [DOI: https://dx.doi.org/10.6112/kscfe.2016.21.3.077]

20. Chelsea, V. Numerical Analysis of the Internal and External Dynamics of a Fluidic Oscillator. Master’s Thesis; Iowa State University: Ames, IA, USA, 2017.

21. Kim, S.H. Effects of Installation Conditions of Fluidic Oscillators on Flow Separation Control. Master’s Thesis; Inha University: Incheon, Republic of Korea, 2019.

22. Newman, B.G. The deflexion of plane jets by adjacent boundaries in Coanda effect. Boundary Layer and Flow Control; Lachmann, G.V. Pergamon Press: Oxford, UK, 1961; pp. 232-264.

23. Kim, Y.S.; Seok, Y.S.; Shin, H.J.; Ko, S.C. A Study on Improving Filter Performance Using Coanda Effect. J. Korean Soc. Manuf. Process. Eng.; 2020; 19, pp. 99-104. [DOI: https://dx.doi.org/10.14775/ksmpe.2020.19.03.099]

24. Jung, Y.S.; Kim, Y.S.; Shin, H.J.; Ko, S.C. A Study on Rain Gutters with Coanda Effect. J. Korean Soc. Manuf. Process Eng.; 2020; 19, pp. 58-64. [DOI: https://dx.doi.org/10.14775/ksmpe.2020.19.04.058]

25. Siemens Digital Industries Software. 2016. Simcenter STAR-CCM+ User Guide version 11.04. Available online: https://support.industrysoftware.automation.siemens.com/general/documentation.shtml (accessed on 12 May 2023).