Search Results

More stringent heavy vehicle emissions legislation demands considerably higher performance for engine cooling systems. This paper presents a study of cooling airflow for a Freightliner Class 8 truck. The predicted radiator coolant inlet and charge-air-cooler outlet temperatures are in very good agreement with the measured data. The under hood flow behavior is described and potential areas of improvement leading to better cooling airflow performance are highlighted. The airflow simulation approach is based on the Lattice-Boltzmann Method (LBM) and is described in detail. It is shown that the presented simulation approach can provide accurate predictions of cooling airflow and coolant temperature across different fan speeds.

A validation study was performed comparing the simulation results of the Lattice-Boltzmann Equation (LBE) based flow solver, PowerFLOW®, to cooling cell measurements conducted at Volvo Trucks North America (VTNA). The experimental conditions were reproduced in the simulations including dynamometer cell geometry, fully detailed under-hood, and external tractor geometry. Interactions between the air flow and heat exchangers were modeled through a coupled simulation with the 1D-tool, PowerCOOL™, to solve for engine coolant and charge air temperatures. Predicted temperatures at the entry and exit plane of the radiator and charge-air-cooler were compared to thermocouple measurements. In addition, a detailed flow analysis was performed to highlight regions of fan shroud loss and cooling airflow recirculation. This information was then used to improve cooling performance in a knowledge-based incremental design process.

Presented are simulations of cooling airflow and external aerodynamics over Land Rover LR3 and Ford Mondeo cars under several driving conditions. The simulations include details of the external flow field together with the flow in the under-hood and underbody areas. Shown is the comparison between the predicted and measured coolant inlet temperature in the radiator, drag and lift coefficients, temperature distribution on the radiator front face, and wake total pressure distribution. Very good agreement is observed. In addition, shown is the complex evolution of the temperature field in the idle case with strong under-hood recirculation. It is shown that the presented Lattice-Boltzmann Method based approach can provide accurate predictions of both cooling airflow and external aerodynamics.

Air flow in the underhood area is the primary source of engine cooling. A quick look at the vehicle underhood reveals exceptionally complex geometry. In addition to the engine, there are fans, radiator, condenser, other heat exchangers and components. The air flow needs to have adequate access to all relevant parts that require cooling. Due to complex geometry, the task to ensure sufficient air cooling is not a simple one. The air flow entering from the front grille is affected by many components on its path through the underhood. Even small geometry details affect the flow direction and can easily cause recirculation regions which reduce the cooling efficiency. Therefore, air cooling flow analysis requires detailed treatment of the underhood geometry and at the same time accurate air flow modeling. Recent advances in the lattice-Boltzmann equation (LBE) modeling are allowing both.

In a tiered cooling pack, the airflow through the individual heat exchangers is determined by the package and aperture lay out. Each heat exchanger rejects heat as a function of the internal coolant flows, the cooling airflow and the air temperature. In a typical automotive cooling pack, the cooling airflow will be non-uniform in velocity and temperature due to fans, aperture geometry, exterior flows, heat exchangers and recirculation. In a drive cycle, these boundary conditions will change with vehicle operating conditions like vehicle speed, engine speed, ambient temperature, and altitude. These non-uniform conditions on the cooling pack can lead to significant errors when uniform boundary conditions are assumed in a transient simulation. This error is commonly corrected using vehicle test data. A predictive approach, which eliminates the need for correlation vehicle testing, is presented.