3-D Multiphase Flow Simulation of Coolant Filling and Deaeration Processes in an Engine Coolant System

Abstract

<div class="section abstract"><div class="htmlview paragraph">The thermal performance of an engine coolant system is efficient when the engine head temperature is maintained within its optimum working range. For this, it is desired that air should not be entrapped in the coolant system which can lead to localized hot spots at critical locations. However, it is difficult to eliminate the trapped air pockets completely. So, the target is to minimize the entrapped air as much as possible during the coolant filling and deaeration processes, especially in major components such as the radiator, engine head, pump etc. The filling processes and duration are typically optimized in an engine test stand along with design changes for augmenting the coolant filling efficiency. However, it is expensive and time consuming to identify air entrapped locations in tests, decide on the filling strategy and make the design changes in the piping accordingly.</div><div class="htmlview paragraph">In the current effort, a simulation-based testing method for coolant filling and deaeration processes is developed for an engine coolant system using an advanced 3D Computational Fluid Dynamics (CFD) tool, Simerics-MP+. The multiphase flow of coolant and air in a 38-litre coolant system for Cummins OFF-highways Tier 4 engine are modeled and analyzed in detail to identify the air entrapped locations and predict the coolant filling efficiency. Explicit Volume of Fluid (VOF) approach with high resolution interface tracking scheme is used to capture the sharp coolant-air interface for better conservation of mass for both the phases. The simulation is performed for gravity filling followed by stabilization, deaeration at different pump speeds, and a drawdown process to understand the effect of leakages. The computed coolant filling efficiency matches within 5% with the test data at the end of gravity filling and deaeration processes. The complete simulation process is automated through a Python script which helps to reduce the run time up to 30% without any manual intervention.</div></div>