Predicting the Performance of a Cleaning Nozzle : A Cost-Effective CFD Modeling Approach Compared Against Experimental Data

Abstract: With the increasing reliance and use of sensors to improve the capabilities of modern road vehicles, ensuring their functionality during different weather conditions has become an important design consideration. The sensing equipment is utilized for a multitude of tasks, ranging from aiding the user with parking to avoiding potentially dangerous situations. If the sensory equipment's functionality is compromised, these additional features might fail. Various solutions are used to ensure the integrity of the sensing equipment, primarily the location of the sensor is chosen to minimize contamination. However, in certain situations, exposure to contaminants cannot be avoided, and active cleaning systems are utilized to mitigate the issue. These systems often contain some sort of cleaning nozzle that sprays liquid onto the sensor in question.  At Volvo Cars Corporation (VCC), the group of Contamination & Core CFD is responsible for assessing the vehicle performance when exposed to contamination, such as rain, snow, or dust. This task is accomplished by employing tools such as computational fluid dynamics (CFD), and experiments. In the first part of this thesis, physical tests were conducted in the VCC wind tunnel to assess the performance of a cleaning nozzle mounted on a side-view mirror at different vehicle speeds. With the knowledge gained during the experiments, the second part of the thesis focused on the development of a cost-effective computational method that could be used to evaluate the cleaning performance virtually. The novel computational method utilized a hybrid approach, including volume of fluid simulations, discrete particle modeling, and Eulerian wall film modeling. Different analysis tools and approaches were used to extract quantitative information from the inherently qualitative data obtained from the experiments, which was then compared against the results from the computational method. The computational method was evaluated for different incoming wind velocities, and three metrics were compared to investigate if any could predict the cleaning performance seen in the experiments. At low incoming wind velocities, an accumulated film mass metric exhibited a good correlation with the experimental data. However, at higher wind speeds, an accumulated film momentum metric showed the closest correlation.