Experimental characterization of frictional hysteresis for nonlinear dynamic analysis of jointed structures

Abstract

Modern high-value jointed structures such as aero-engines are carefully designed and optimized to prevent failure and maximise durability. During operation, these structures are subjected to high-frequency vibrations that, uncontrolled, might lead to loss of performance, higher maintenance and, in the worst-case scenario, catastrophic failures due to high cycle fatigue. Accurate dynamics predictions are therefore needed to achieve safer designs. Often, challenges in obtaining such predictions are introduced by the presence of friction between the thousands of components in contact. Despite the considerable research efforts over the past decades, the effects of friction on the dynamics are not yet fully understood, due to the complex mechanisms that govern the interface behaviour. As a result, fully validated and predictive modelling approaches are not available and expensive physical testing is still required to assess the performance and reliability of new designs.

The aims of this PhD research are thus to i) provide a better understanding of the friction mechanisms, ii) quantify their effects on the dynamic response of jointed structures and iii) validate/upgrade state-of-the-art measuring and modelling approaches for nonlinear dynamic analysis. For those purposes, an extensive round robin test was performed on two high-frequency friction rigs of different institutions to measure a wide range of hysteresis loops. Novel experiment-based contact laws were discovered to be used in simulations. Simultaneously with the hysteresis measurements, ultrasonic measurements were performed to monitor the vibrating contact interfaces in real time, providing unprecedented insights into the friction mechanisms. These experimental methods were also employed to monitor the dynamic behaviour of the friction rigs, leading to an improved understanding of the effects of friction on the dynamics. The knowledge thus gained was used to propose upgrades for state-of-the-art modelling approaches, which were then validated by comparing new simulations with experimental measurements.