Numerical simulation of the airflow over and heat transfer through a vehicle windshield defrosting and demisting system A. Aroussi, A. Hassan, Y. S. Morsi Abstract A numerical model and technique are described to simulate the turbulent fluid flow over and heat transfer through a model of vehicle windshield defrosting and demisting systems. The simplified geometry and the dimensions of the numerical model are representative of vehicle system with accurate locations of the nozzles and outlet vents, including cabin features such as seating and the rear parcel shelf. The three-dimensional geometry of the numerical model is created in Auto Cad (Release 14). Surface meshing and a computational mesh of 750,000 (tetrahedral) fluid cells is generated in the pre-processor of the CFD software used for the simulation of the fluid flow. Turbulence is modelled by using the k-e turbulence equations together with the wall function method. This decision was made after comparing the k-e model’s performance with that of lower order models, and after considering the increased computer time requirements and decreased stability of more complex models, such as the Reynolds stress model. The numerical results of the study are very encouraging and compare fa- vourably with measurements obtained from the actual ve- hicle by Thermograph and Hot Bulb probe techniques. The findings highlight some of the drawbacks of the existing design of the windshield systems and show that the maxi- mum flow rates occur in the vicinity of the lower part of the windshield, progressing from the defroster nozzle in the dashboard. Keywords Defroster Nozzle, Windshield, Defrosting, Demisting, Numerical Simulation, Vehicles 1 Introduction To safeguard passenger safety, it is imperative that ade- quate visibility throughout vehicle windshield is main- tained at all times, particularly at very low temperatures when ice and/or mist layers tend to form on the windshield screen. One of the most important aspects of this re- quirement is the capacity of the windshield defrost/demist system to immediately and completely melt ice on the outer screen surface as well as eliminate the mist formed on the inner surface. However, due to the geometric complexity of the windshield and the defroster/demister system, the airflow produced by the defroster nozzles does not entirely cover the whole windshield area. Conse- quently, in most modern cars a critical dead zone occurs, most perpetually at the corners and at the upper regions of the windshield screen. Earlier investigators searched for ways of improving the design of windshield defroster/demister systems. They recognised the problem and applied recent advances in experimental diagnostics techniques and Computational Fluid Dynamics (CFD) to study the air flow. The work of Stouffer and Sharkitt (1987) was aimed at the development of a fluidic oscillator device to improve the airflow dis- tributions over the windshield. The device was also used as a windshield defrost/defog nozzle with some degree of success. Later, Dugand and Vitali (1990) carried out an experimental investigation where the Thermo-graphic technique was used to detect thermal fields on emitting surfaces. The authors proposed a specific combination of hardware/software for the processing of the images ob- tained and recommended various ways of improving windshield/defrosting systems. Carignano and Pippione (1990) used a computer assisted thermo-graphic technique to optimise the performance of windscreen defrosting for an industrial vehicle system. Lee et al. (1994) utilised a Computational Fluid Dynamics (CFD) code, namely ICEM-CFD, to simulate the mechanism of windshield de-icing. The complete vehicle configuration was trans- formed from CAD and the mesh was created and assembled using a multi-domain approach. The authors demonstrated the capability of the developed module in simulating cold room de-icing tests to supplement the experimental work. Recently, Brewster et al. (1997) used the CFD code STAR-CD to simulate mechanism of ice building on the windshield in three-dimensional form. The authors used a non-linear enthalpy-temperature relationship to simulate the ice/water layer. Melting contours were predicted every 5 minutes and the authors reported good agreement be- tween the numerical simulations and cold-room test data for the ice coverage contours. Abdul Nour (1998) con- ducted a similar study, also using the STAR-CD code. He examined the windshield flow fields and the vehicle Heat and Mass Transfer 39 (2003) 401–405 DOI 10.1007/s00231-002-0307-x Received: 6 August 2001 Published online: 1 June 2002 Ó Springer-Verlag 2002 A. Aroussi, A. Hassan School of Mechanical, Materials, Manufacturing, Engineering and Management Division of Mechanical Engineering University of Nottingham, University Park NG7 2RD, Nottingham, UK Y.S. Morsi (&) Modelling and Process Simulation Research Group Industrial Research Institute IRIS, Swinburne University of Technology Hawthorn, Australia 3122 E-mail: ymorsi@swin.edu.au 401