Abstract
Modern common rail injection systems allow the engine developer to flexibly form the injection rate by using multiple injection patterns. Five injections per combustions cycle is state of the art and this number trends to increase. Due to the fast opening and closing of the injectors pressure waves are generated on the connection pipe between injector and common rail. These oscillations influence the amount and progression of the injected fuel mass of subsequent injections and have relevant influence on the exhaust gas emissions of the engine. In state-of-the-art solutions the variation of the injected fuel masses are measured on a test bench over the whole system working range. The results are identified as parametric curves and the curve parameters are stored in look-up tables. As amplitude, frequency, and damping of the fuel mass variations depend on many system parameters, this approach is not very accurate in all working points. Online-knowledge of the real fuel pressure close to the injector could improve calculation of the interference of two adjacent injections and therefore enhance controlling of the injection rate. In earlier publications a new sensor concept was introduced [1]. It is based on the magneto-elastic effect and allows non-invasive pressure measurements with low sensor costs. The sensor basically consists of a coil which is mounted on the injection pipe. The injection pipe is made of steel which changes its magnetic properties under high mechanical stress. This leads to a time-varying magnetic field which induces a voltage in the coil around the pipe. This signal is amplified and can be matched to the signal of a reference sensor by appropriate scaling. With this sensor a new strategy of correcting fuel mass variations with increased accuracy is possible. Hydraulic simulations of a common rail injection system are done. The results show how the amount of injected fuel depends on phase and amplitude of the pressure waves on the pipe. A strategy for compensating the influence of the pressure waves, by use of the measured pressure signal and the results of the hydraulic simulations, is outlined. The results are validated with measurement data. An estimate of the enhanced accuracy of the injected fuel mass by use of the magneto-elastic pressure sensor is given.
Keywords: Diesel Fuel Injection, Fuel Pressure, Pressure Sensor, Injection Accuracy