Mitsubishi Motors Corporation

Over the years, Mitsubishi Motors Corporation (MMC), in Japan, has successfully utilized CFD technology to greatly improve their heavy-duty production truck design. MMC simulations typically include flow in the engine compartment, the HVAC ductwork, the fan, and the outside truck body. Recently, significant enhancements in both computer technology and CFD technology have allowed MMC to develop more accurate models. This has resulted in significant reduction of costs that were previously spent on physical testing.

In particular, CFD has become a powerful tool for engine compartment applications. Engine compartments and fans tend to be enclosed for exterior noise reduction, so they are not easy to access to obtain detailed experimental flow data. Therefore, establishment of prediction methodology using CFD is important.

Flow pathlines inside the engine compartment

Simplified models, such as a fan model based on fan performance curve data, were traditionally used to predict the flow field around a cooling fan. However, such simplification causes problems in terms of accuracy in the vicinity of the fan blade, where swirling flow and turbulent energy distribution effects are significant.

MMC developed a model using FLUENT’s multiple reference frames, or MRF model, in which stationary and rotating fluid zones were calculated at the same time. To validate the model, experimental data were taken using test equipment that was comprised of an inter-cooler, a radiator, a shroud, and a cooling fan. A Laser Doppler Velocimeter (LDV), located on the rear surface of the hub, was used to measure the flow data. Grid adaption and a standard turbulent model were used in the prediction model. The results of the CFD predictions showed good agreement with the experimental data to within 5 percent accuracy. At high-pressure loss locations, some reverse flow was observed and fluctuation of flow rate distribution was predicted because of swirling flow.

MMC performed internal flow analyses of the engine compartment, including the cooling fan engine, the chassis, and the auxiliaries, using Fluent’s unstructured CFD code. The detailed shape of the cooling fan was modeled based on the 3D CAD data, which were imported into the preprocessors, ICEM/CFD™ and TGrid™, to develop the approximately 1,700,000 tetrahedral unstructured mesh model. An MRF model was used for the cooling fan and the running speed was set to 30 km/hr in this calculation. The predicted flow rate agreed to within 5 percent of the experiment in the vicinity of the engine compartment, the exhaust manifold, and the chassis.

FluentNEWS