IEEE TRANSACTIONS ON MAGNETICS, VOL. 41, NO. 3, MARCH 2005 1163 Design Features for Enhancing the Performance of Electromagnetic Valve Actuation Systems Richard E. Clark, Geraint W. Jewell, Stephen J. Forrest, Jan Rens, and Christophe Maerky Abstract—The paper describes a type of variable air-gap reluc- tance actuator that offers potential for enhancing the dynamic per- formance of electromagnetic valve actuation systems for internal combustion engines. In both the stator and armature, the actu- ator incorporates design features that allow the force-displacement characteristic to be tailored to meet operational requirements. The paper demonstrates the considerable scope for varying actuator characteristics by means of detailed two- and three-dimensional finite-element modeling. The key findings from the finite-element modeling are validated by experimental measurements on a proto- type actuator. Index Terms—Electromagnetic actuators, finite-element analysis. I. INTRODUCTION T HE benefits of employing electromagnetic valve actua- tion in internal combustion engines are well recognized. Studies of engines equipped with prototype valve actuation sys- tems indicate that potential reductions in fuel consumption of around 17% could be achieved [1]–[5]. However, this applica- tion has very demanding performance specifications, with typ- ical strokes of 8 mm and opening and closing event durations of 3 ms. The opening of a valve consists of three distinct phases, that is, a release phase in which the valve-lash is taken up in a controlled manner with a low impact velocity between the stems of the armature and the valve, a central portion over which little control is exercised over the armature motion (this being deter- mined predominantly by combination of the moving mass and the net spring stiffness) and a landing phase in which the ar- mature impacts onto the stator. In the case of valve closure, no control is required in the release phase, since the armature re- mains in contact with the valve stem in the open position, but it is necessary to precisely control both the landing of the valve in its seat and the subsequent landing of the armature on the upper stator. The armature landing and release phases have a signifi- cant bearing on both acoustic noise emissions and component wear, and are hence subject to very demanding limits, with im- Manuscript received August 3, 2004; revised December 20, 2004. This work was supported by the Fifth Framework Programme of the European Union under Project ELVAS G3RD-CT-2000-00363, by The Royal Academy of Engineering, and by the UK Engineering and Physical Sciences Research Council. R. E. Clark, G. W. Jewell, S. J. Forrest, and J. Rens are with the Department of Electronic and Electrical Engineering, The University of Sheffield, Sheffield S1 3JD, U.K. (e-mail: r.e.clark@sheffield.ac.uk; G.Jewell@sheffield.ac.uk; s.j. forrest@sheffield.ac.uk; J.Rens@sheffield.ac.uk). C. Maerky is with Johnson Controls, 95526 Cergy-Pontoise Cedex, France (e-mail: christophe.maerky@jci.com). Digital Object Identifier 10.1109/TMAG.2004.843342 Fig. 1. Typical actuator configuration for an electromagnetic valve actuation system. pact velocities of 0.05 ms at low engine speed being a typ- ical specification. The difficulties associated with meeting these performance specifications are compounded by the very limited space avail- able on a modern cylinder head to accommodate a series of ac- tuators. For the actuators considered in this paper, the maximum width is limited to 33 mm. Fig. 1 shows a schematic cross section through one type of ac- tuator that has been employed in several published valve actua- tion systems [3], [6]–[8]. It consists of a pair of laminated E-core stators, a rectangular steel armature and a pair of relatively stiff springs (typical values of stiffness being 50 and 80 N/mm). For the particular example shown in Fig. 1, the actuator spring is located between the actuator housing and the valve spring, al- though there are also several designs in which this spring is lo- cated on top of the actuator housing. The devices shown in Fig. 1 are essentially resonant oscil- lating devices in which energy is interchanged between the two springs as the valve moves from its fully open to closed posi- tion and vice-versa. The E-core reluctance actuators compen- sate for the losses in the system, latch the armature in one of two fixed positions thus allowing the instant of armature release to be precisely controlled, and facilitate closed-loop control of landing and departure velocities. The vast majority of valve actuation systems employ conventional laminated E-core sta- tors and carbon-steel rectangular armatures of the general type shown in the cross section of Fig. 2. This type of reluctance actuator is reliant on a normal compo- nent of variable-reluctance force, and hence the force-displace- 0018-9464/$20.00 © 2005 IEEE