In the automotive lighting industry, car manufacturers (OEMs) require that lamps pass vibration resonance test. Failure to pass this test means that the lamp cannot be delivered to the customer. Resonant frequency is a direct cause of lamp failure due to large deformations of lamp components at a specific frequency.
Resonance tests can be estimated experimentally using a vibration machine or numerically utilizing finite element analysis software. These methods estimate the resonance of the lamp after the lamp is fabricated or the complete lamp model is created on the CAD software. The objective of this paper is to teach the development of a prediction tool to estimate the resonant frequency of lamps before fabricating the lamp or creating the model on the CAD system.
Design of computer experimentation (DOCE) methodology has been utilized to identify the significance level of the parameters involved to estimate the resonant frequency of a lighting system. The Maximin approach has been used to optimize the sampling selection process for the runs to conduct the analysis.
Consequently, a second order non-linear response function is generated to quantify the non-linear behavior of resonance. The model is a quadratic regression equation that includes the significant parameters, their interactions and the significant second order parameters with 95% confidence level. The proposed approach has been validated by comparing the results with finite element model simulation.