Projects

Introduction

Structural Health Monitoring (SHM) methods are of increasing interest for the evaluation of huge structures such as aging bridges, buildings in earthquake areas or temporarily inaccessible offshore wind energy plants. Varying dynamics due to changing wind excitation and temperatures makes monitoring of these kinds of structures a challenging task. Fraunhofer LBF develops SHM methods focusing particularly on the applicability to real world structures. The model of a wind turbine is used to demonstrate an autonomously operating SHM system. Sensors applied to the wind turbine tower measure the wind induced vibration. The vibration data is next processed to extract the natural frequencies and mode shapes of the structure.

Concept and realization

Accelerometers applied along the tower continuously transfer time domain data to a sensor node. The data is preprocessed on the sensor node by the means of the Random Decrement (RD) method. Signals surpassing a certain trigger level are logged for a predefined duration. Averaging a huge number of logged signals for each accelerometer yields the so called RD functions. The RD functions correspond to the correlation functions except for a scaling factor. Preprocessing the data on the sensor node significantly reduces the data to be transferred to a central processing unit. This is an essential requirement particularly for the realization of wireless sensor networks. For data analysis on the central processing unit the RD functions are transformed to frequency range. A Frequency Domain Decomposition is next conducted to calculate the structure mode shapes.

Conclusion

Laboratory and field tests revealed that the calculated mode shapes are reproducible – even in changing temperature and wind conditions. Structure modifications, such as damage induced stiffness losses or mass variations are reflected by the mode shapes. Thus information about the presence and location of damage can be drawn from analyzing the calculated mode shapes.