A finite element approach for the implementation of magnetostrictive material terfenol-D in automotive CNG fuel injection actuation
Choadhury, Habibullah Amin
(2008)
A finite element approach for the implementation of magnetostrictive material terfenol-D in automotive CNG fuel injection actuation.
Master of Engineering thesis, Dublin City University.
Magnetostriction is the deformation that spontaneously occurs in ferromagnetic materials when an external magnetic field is applied. In applications broadly defined for actuation, magnetostrictive material Terfenol-D possesses intrinsic rapid response times while providing small and accurate displacements and high-energy efficiency, which are some of the essential parameters required for fast control of fuel injector valves for decreased engine emissions and lower fuel consumption compared with the traditional solenoid fuel injection system.
A prototype CNG fuel injector assembly was designed, which primarily included magnetostrictive material Terfenol-D as the actuator material, 1020 Steel having soft magnetic properties as the injector housing material, AWG copper wire as the coil material and 316 Stainless Steel having non-magnetic properties as the plunger material. A 2D cross-sectional geometry including the injector housing, coil, Terfenol-D shaft, and plunger, was modeled in both Finite Element Method Magnetics (FEMM) and ANSYS for 2D axisymmetric magnetic simulation. The magnetic simulations were performed in order to determine the coil-circuit parameters and the magnetic field strength to achieve the required magnetostrictive strain, and consequently, the injector needle lift. The FEMM magnetic simulations were carried out with four different types of AWG coil wires and four different injector coil thicknesses in order to evaluate the relationship between the different coil types and thicknesses against the achieved strain or injector lift. Eventually, the optimized parameter obtained from FEMM results analysis was verified against ANSYS electromagnetic simulation.
Subsequently, a three dimensional replica of the CNG flow conduit was modelled in GAMBIT with the resultant injector lift. The meshed conduit was then simulated in FLUENT using the 3D time independent segregated solver with standard k-ε, realizable k-ε and RSM turbulent models to predict the mass flow rate of CNG to be injected. Eventually, the simulated flow rates were verified against mathematically derived static flow rate required for a standard automotive fuel injector considering standard horsepower, BSFC and injector duty cycle.