Improvement of a tail-plane strucural model using vibration test data.

To achieve the best structural model improvement using vibration test data, a major effort has been made to identify poor modelling regions as a guideline for subsequent model updating. The method presented and used in this paper is the energy error estimation method. In the method the difference between analytical and test data based energies at element scale is estimated to indicate any poor structural mass and stiffness modelling. As a result, poor modelling regions can be distinguished from those modelled correctly and the improvement of the original structural model can be carried out effectively and accurately. To demonstrate the application of this method, a full-scale tail-plane structure has been studied by using simulated "test" modes as a simulated case and using measured modes as a practical case. In both cases poor modelling regions of the original structural model have been accurately located. Subsequently, a significant improvement of the structural model with a reduction of average frequency error from original 2·2% down to 0·1% for the simulated case and from 4·6 to 1·8% for the practical case has been achie