Numerical and Experimental Investigation of Thermal Conditions Inside the Engine Compartment of Snowmobiles

• Published in: SAE International Journal of Commercial Vehicles 2015-05, Vol.8 (1), p.225-235, Article 2015-01-9017
• Main Authors: Wurm, Johannes, Fitl, Matthias, Gumpesberger, Michael, Väisänen, Esa, Hochenauer, Christoph
• Format: Article
• Language: English
• ISSN: 1946-391X; 1946-3928; 1946-3928
• DOI: 10.4271/2015-01-9017

Nowadays, investigating underhood airflow by using numerical simulation is a standard task in the development process of passenger cars and commercial vehicles. Numerous publications exist which deal with simulating the airflow through the engine compartment of road vehicles. However, hardly anything can be found which deals with off-road vehicles and nothing exists which focuses on snowmobiles. In the presented paper the airflow and the thermal conditions inside the engine compartment of a snowmobile are investigated by the usage of computational fluid dynamics (CFD) as well as experimental methods. Field tests at arctic conditions have been conducted on a serial snowmobile to measure temperatures inside the compartment and to gain realistic boundary conditions for the numerical simulation. Thermocouples (type K) were attached under the hood to measure exhaust, air, coolant and surface temperatures of several components at previously defined load cases. Wind tunnel tests have been performed to examine the external airflow. Hence a direct comparison between simulation and experiment is possible and excellent agreement can be achieved. State-of-the-art CFD methods for simulating underhood airflow are applied in this investigation. The modelling process and the various numerical methods which are necessary to take all effects of underhood airflow into account are described in detail. The structure of the airflow and the component temperatures are analyzed to identify hot spots and recirculation zones influencing the system's cooling performance and bear the risk of thermal failure. Proposals to decrease the time-consuming modelling process and improvements for fast simulation set-up are made.