Development of a Reed Valve Model for Engine Simulations for Two-Stroke Engines 2004-01-1455
Engine manufacturers and product designers are currently under significant pressure to reduce fuel consumption, emissions, and noise from two-stroke engines. Improved engine breathing goes a long way toward overcoming these challenges by providing better charge delivery, complete exhaust gas scavenging, and minimal fuel short-circuiting.
Reed valve motion and the resulting gas flow are highly coupled phenomena. Understanding and controlling these dynamics and their interactions is critical to effective intake breathing. A comprehensive effort to model reed valves at a level sufficient for engine design using engine simulations is still incomplete. This work is focused on the assessment, development and validation of a reed valve model suitable for engine simulations.
In this work, a representative two-stroke engine reed valve is modeled in a detailed fluid/structures, multi-physics CFD code which fully resolves both the flow-field and the details of the deflecting reed plate.
This same reed valve geometry is modeled in a 1D engine simulations code. In lieu of detailed experimental measurements, the CFD simulation results are used to provide the flow and lift data for various differential pressures. These are input into the 1D model as flow coefficients vs. lift.
A comparison of results, between the 1D engine simulation and the original fluid/structure CFD, shows excellent agreement for the maximum lift, the lift vs. time profile, and for the overall flow predictions.
In the 1D simulation code, the standard component model representation of a reed valve is as a spring-mass-damper physical model. It was found that this model, while not an exact representation of the bending reed, did provide adequate agreement.
Additionally, the predicted lifts were compared with uniformly loaded deflecting beam theory assuming the up-stream and down-steam pressures are uniformly applied on either side of the reed valve. It was found that the while this assumption is an adequate first level approximation for low lift conditions and suitable for preliminary design analysis, the CFD results indicate that it does not hold for high lift conditions.
In this work, a method was demonstrated which allows design engineers to investigate reed valve design options from the detailed CFD perspective and from the 1D engine simulations perspective, and thus to develop a good reed valve model for analysis prior to testing. This allows the engineer to screen potential designs with simulation before the more expensive task of building and testing prototypes. The validated reed valve model can be incorporated directly into a full engine simulation with confidence.