Computer simulation of the inlet port is helping to provide improved fuel economy and emissions in up-coming Jaguar models. Jaguar engineers have long wanted to improve the inlet port design process by simulating flow in alternative inlet port designs. The engineers, however, experienced problems in modeling the complex internal contours of the inlet port and chamber. The dramatic improvement in modeling speed combined with the high accuracy of sin1.ulation has made it practical to integrate computational fluid dynamics into the inlet port and chamber design process. Jaguar engineers are now able to evaluate concept designs within one week compared to the month or more that was required in the past. Engineers have been able to make significant improvements in the performance of their most recent inlet port designs, increasing flow efficiency by an average of 10 percent, while maintaining turbulence at acceptable levels. These improvements will provide significant reductions in fuel consumption and emissions in future Jaguar engines.
Computer simulation of the inlet port is helping to provide improved fuel economy and emissions in upcoming Jaguar models. The flow coefficient and tumble ratio—the ratio describing the angular motion of the incoming air—of the inlet port are critical factors in determining how lean the engine can be run with good combustion stability. In the past, it took at least a month to prototype a single inlet port configuration and run a test that provided no information on internal flow patterns. Simulation let engineers move much faster to improve the efficiency of inlet ports in upcoming models.
Jaguar Cars is a subsidiary of Ford Motor Co. that produces luxury motor vehicles. Currently, Jaguar has two lines, the XK8 and XJ8, both powered by AJ-V8 engines, available in 3.2-liter, 4.0-liter, and 4.0-liter supercharged versions. According to Jaguar, the 4.0-liter version of the AJ-V8 is the most efficient engine in its class with the highest specific power (72.5 horsepower per liter) and highest specific torque at low engine speeds. A minimum 80 percent of peak torque is available between 1,400 and 6,400 rpm for quick acceleration across an unusually wide range of speeds. The AJ-V8 is also the lightest engine in its class, at 441 pounds. A variable inlet cam phasing system optimizes valve timing, increasing both low-speed torque and top-end power.
Engineers working on new engine variants were charged with the task of significantly reducing combustion emissions and improving fuel economy. One of the primary means by which they set out to accomplish this task was to improve the flow characteristics of the inlet port and chamber. Optimizing the port’s flow provides a more uniform mixture of fuel and air, which makes it possible to operate the engine with a leaner fuel/air mixture, reducing emissions and fuel consumption. At the same time, engineers must be concerned about turbulent energy that is released as the angular motion of the fuel/air mixture is broken down during the compression stroke. Turbulence tends to increase with engine speed. Designers need to maintain turbulence at high enough levels at low speeds to provide a stable idle and good acceleration, yet avoid excessive turbulence at high speeds that will be experienced as noise by the occupants of the vehicle.
The traditional method of resolving these tradeoffs in the design of the inlet port and chamber is to build a prototype of conceptual port designs and test them on a steady state rig. Through an iterative approach, the port is tuned to provide the required motion and flow characteristics. One problem with this approach is that it takes about a month to build and test a prototype of the inlet port and chamber. The cost and time involved in building the port places a limitation on the number of different designs that can be evaluated. Reduced time in the design cycle would lead engineers to settle for a compromised design. But the biggest problem is that even after the test is run, there is little or no understanding of why the design performed the way it did. In particular, testing is unable to detect details of recirculating areas, turbulence, or constrictions that adversely affect performance and pressure.
Jaguar engineers have long wanted to improve the inlet port design process by simulating flow in alternative inlet port designs. Computational fluid dynamics, the obvious analysis tool to address this problem, has been used at Jaguar since the late 1980s to model vehicle flows, exhaust catalyst systems, and other design issues. A CFD simulation provides fluid velocity and pressure values throughout the solution domain for time-dependent problems with complex geometries and boundary conditions. As part of the analysis, a designer may change the geometry of the system or the boundary conditions such as inlet velocity and flow rate, and view the effect on fluid flow patterns. CFD is an efficient and effective tool for generating detailed parametric studies, significantly reducing the amount of experimentation necessary to develop a device.
The engineers, however, experienced problems in modeling the complex internal contours of the inlet port and chamber. Producing a model suitable for CFD simulation using conventional analysis tools takes nearly as long as building a prototype. Conventional CFD codes use a structured mesh approach requiring that a hexahedral block structure be defined by hand before a volume mesh can be generated. It takes weeks or months to produce a grid using this approach. By the time the analysis is complete, many other changes usually have been made to the parts and the results are obsolete.
In 1996, Jaguar engineers began working with Fluent/UNS, an unstructured mesh CFD software package from Fluent Inc. of Lebanon, N.H. The software greatly simplifies the modeling of complex flow geometries. The program itself generates an unstructured tetrahedral grid based on a geometry that is created from scratch or imported by the user from a CAD program, making it possible to analyze a new port design within a week and provide complete information on flow velocity and pressure throughout the model.
Jaguar engineers typically import geometry from Structured Dynamics Research Corp. of Milford, Ohio, whose I-DEAS Master Series CAD program they use for mechanical design, into Fluent’s volume mesh software, TGRID. They use TGRID to automatically generate a tetrahedral mesh. They then use the software to adaptively refine the mesh so that important flow features are resolved locally, while a more cost-effective coarse mesh represents regions of smooth flow. Once a surface grid has been generated, the entire process of volume meshing the component takes under 60 minutes.
A critical step in the implementation process was validating that CFD could accurately simulate the performance of the inlet port and chamber. Engineers used Fluent/UNS to simulate the performance of the AJ-V8 inlet ports on a steady state rig for two inlet conditions, a medium valve lift and a high valve lift. The predicted data were compared with detailed flow field information, gathered using sophisticated laser techniques. The boundary conditions were chosen to represent rig operating conditions. The inlet and outlet boundaries were modeled as pressure boundaries and the port, chamber, valves, and tube were modeled as a nonslip wall surface. A surface roughness of 25 microns was applied to the ports. Turbulence was modeled using the K-epsilon model. The near wall region was modeled using the standard wall function.
A converged solution was achieved in approximately 400 iterations, taking 4.5 hours on a Power Challenge workstation from Silicon Graphics Inc. of Mountain View, Calif., with four R10000 microprocessors. The analysis results gave Jaguar engineers, for the first time, a clear picture of the airflow within the inlet port and chamber. For the high lift case, the majority of the flow entered over the top of the valve and was deflected across the top of the chamber, forming a tumbling motion. The remaining flow entered below the valve, forming a secondary, counter-rotating motion. A third region of flow was evident beneath the valve seat. For the medium lift case, the flow was more evenly distributed over the top and bottom of the valve. The two tumbling motions were still apparent but weaker, while the flow remained attached in the valve seat region.
The simulation correlated well with the experimental data. Flow coefficients were predicted within 5 percent and the tumble ratio between 10 percent and 40 percent.
Perhaps most important of all, comparisons of simulation results for different port designs were consistently accurate in predicting the direction of change in critical output parameters. Some of the differences between the simulation and experimental data were easy to explain. The simulation results were symmetrical across one particular cross-section, while the experimental data was not. Since the design itself was symmetrical, this discrepancy is thought to have resulted from small geometric differences between the ports, which occur during manufacturing.
The dramatic improvement in modeling speed combined with the high accuracy of simulation has made it practical to integrate CFD into the inlet port and chamber design process. Jaguar engineers are now able to evaluate concept designs within one week compared to the month or more that was required in the past. In addition, simulation provides far more understanding of the reason why a concept design performs the way it does, making it possible to iterate toward an optimum design that much more quickly.
As a result, engineers have been able to make significant improvements in the performance of their most recent inlet port designs, increasing flow efficiency by an average of 10 percent, while maintaining turbulence at acceptable levels. These improvements will provide significant reductions in fuel consumption and emissions in future Jaguar engines.