Parameters sensitivity analysis in a solenoid common-rail injector model

Abstract

Parameter identification techniques are applied to find suitable values of unknown key input factors into a mathematical model. Usually, these factors are determined through optimization algorithms. In this case, the establishment of consistent initial values, and their corresponding searching ranges (which are unknown), is quite important to provide reliable results after the identification process. In this work, a sensitivity analysis is carried out to evaluate the influence of the main parameters involved in a numerical model used to simulate a common-rail solenoid injector. Experimental measurements for several operating conditions were done, and their results were compared with the model outputs to establish an objective function from which to perform the assessment. As a result of this study, it was established the importance of each parameter included in the model: firstly, those parameters with low variability were selected to be removed from the identification procedure, assuming a value for them that do not affect the model results; finally, the rest of parameters remained in the algorithm. After applying the aforementioned identification process, some accurate results are provided and discussed in order to propose improvements for future works.