Artificial boundary condition approach for structural identification: A laboratory perspective

The artificial boundary method (ABM) provides a possibility to generate additional modal frequency information for a test structure under different support conditions (artificial boundary condition or ABC) without the need to physically impose such additional supports during a vibration test. Therefore, it has the potential to considerably enrich the response dataset for the purpose of structural identification and finite element model updating. The general implementation and effectiveness of the approach has been demonstrated extensively in a recent study by the authors with numerical examples. However, as the method is based on the premise that the required frequency response functions can be measured with good accuracy, the practical aspects of the implementation in an actual measurement environment need to be investigated. This paper presents a preliminary experimental study of the method with the laboratory test of a steel beam model. The issues associated with the inversion calculation of the FRF function matrix are highlighted, and possible improvements are discussed. The feasibility of using only the lowest few ABC frequencies in conjunction with a variety of ABC configurations is also examined with a numerical example.