Search Results

Automotive thermal systems are becoming complicated each year. The powertrain efficiency improvement initiatives are driving transmission and engine oil heaters into coolant network design alternatives. The initiatives of electrified and autonomous vehicles are making coolant networks even more complex. The coolant networks these days have many heat exchangers, electric water pumps and valves, apart from typical radiators, thermostat and heater core. Some of these heat exchangers, including cabin heaters deal with very small amount of coolant flow rates at different ambient conditions. This paper describes how viscosity can be a major reason for simulation inaccuracy, and how to deal with it for each component in the coolant network. Both experimental and computational aspects have been considered in this paper with wide range of ambient temperatures.

A process to create a vehicle resistance curve based on airflow predictions using Computational Fluid Dynamics (CFD) simulation technique is presented. 1-dimensional engine cooling system simulation tool KULI is used to compute the coefficients of vehicle resistance curve. A full factorial Design of Experiment (DOE) established the relationship between the coefficients and the sum of absolute difference between KULI and CFD predictions. The NLPQL optimization routine is used to accurately predict the coefficients so that sum of absolute difference between KULI and CFD predictions is minimized.

This paper explores the trade-off between reliability-based design and robustness for an automotive under-hood thermal system using the iSIGHT-FD environment. The interaction between the engine cooling system and the heating, ventilating, and air-conditioning (HVAC) system is described. The engine cooling system performance is modeled using Flowmaster and a metamodel is developed in iSIGHT. The actual HVAC system performance is characterized using test bench data. A design of experiment procedure determines the dominant factors and the statistics of the HVAC performance is obtained using Monte Carlo simulation (MCS). The MCS results are used to build an overall system response metamodel in order to reduce the computational effort. A multi-objective optimization in iSIGHT maximizes the system mean performance and simultaneously minimizes its standard deviation subject to probabilistic constraints.

This paper describes the use of robust engineering in engine cooling system design. 1-D thermal-fluid network simulation software has been used in conjunction with statistical variation analysis software. It shows how we can increase the reliability of an automotive engine cooling system by considering the variations of all factors of design. Design of experiment (DOE) and Monte Carlo simulation techniques have been used to optimize the engine cooling system design.