AbstractFor many engineering structures the dynamical response during operation is rarely stationary or regular. In fact, many studies on measured data reveal the irregular nature of dynamical system responses, for example in the field of research on brake squeal. Therefore, linear measures, such as amplitude and frequency spectra, cannot characterize the core dynamics contained in an irregular time series. Additionally, linear methods for damping extraction from measured signals tend to lose significance when it comes to irregular responses of self‐ excited systems. This work aims at highlighting the need for quantifiers from nonlinear time series analysis to adequately characterize real‐world vibration data. In a first step a minimal model is studied to relate nonlinear time series quantifiers to internal system states and dissipation considerations. The goal of this research is to develop an understanding for the information contained in time series data concerning dissipation mechanisms of complex friction‐excited systems. (© 2017 Wiley‐VCH Verlag GmbH & Co. KGaA, Weinheim)