In order to reduce costs for development and production, the objective in gearbox development and design is to predict running and noise behavior of a gearbox without manufacturing a prototype and running expensive experimental investigations. To achieve this objective, powerful simulation models have to be set up in a first step. Afterwards, those models have to be qualified and compared to experimental investigations. During the investigation procedure of gearboxes, there are two possibilities to evaluate the running and noise behavior: quasi-static and dynamic investigations. In times of engine downsizing, e-mobility and lightweight design, the dynamic excitation behavior is becoming increasingly important. Opposed to the quasi-static excitation, the dynamic excitation behavior of the gearbox regards increased load at resonance points caused by the dynamic behavior of the complete drive train.

The dynamic behavior of one-stage gearboxes has been the subject of scientific research for many years. In contrast, there have been few reports which investigate the running behavior of multi-stage gearboxes in the last few years. Most of them are based on computational investigations. However, to utilize the results of those simulation models for design purposes, it is indispensable to compare simulation results with experimental results.

As a first step, test rig setups have to be designed to analyze the significant parameters for dynamic excitation behavior of multi-stage gearboxes. Those parameters are, for example, the macro geometry of gears (helix angle, number of teeth, etc.), the micro geometry of the tooth flanks (profile and lead crowning, etc.), and the stiffness of the intermediate shafts between the gear meshes. Furthermore, the difference between the beginning of the gear meshes, also called phase shift, has to be regarded. An extensive research concerning varying parameters and their dynamic interaction has not been focused on in existing publications and is part of this paper.

Additionally, the objective of this paper is the analysis of the excitation behavior and the development of a simulation model for a two-stage gearbox. Therefore, a single-stage simulation model is validated by experimental results. This single-stage model is enhanced to a two-stage model. Regarding the experimental investigations, the design of a new test rig which allows analyzing the dynamic behavior of a two-stage gearbox is presented. Based on a reference test variant, several gear parameters, such as helix angle and number of teeth, are varied in both stages by changing the gears and are investigated. Subsequently, a simulation model is presented which is able to calculate the dynamic excitation behavior of the two-stage gearbox tests.