Abstract: The modal parameters of gearbox housing of high speed train under assembly condition are identified by a specially designed test rig. The gearbox is assembled in a wheelset system and its actual constraints in the bogie frame are considered in the test. By using multi-band stationary random excitation, the recognition of the first six-order modal parameters of the gearbox housing including modal frequency, damping ratio, vibration mode is completed with PolyMAX method. The identification results agree well with the simulation, indicating that the parameter identification is reliable. For the false mode appearing in the parameter recognition process, the responses of the shaft, gearbox, axlebox and other components are obtained by selecting different excitation points through the hammer test. The main vibration source of the false mode is determined by analyzing the transfer function of the subsystems such as axle-gearbox, axle box-gearbox and axle-axle box. The influence of axle and axle box on the vibration of gearbox housing in the test rig is explored. The combination of continuous wavelet transform and singular value decomposition is used to weaken the false modal with low energy value to further improve the accuracy of the in-situ modal parameter recognition of gearbox housing.

Key words: assembly condition, gearbox housing, high speed train, modal identification