Improvement of analytical dynamic models using vibration test data.
Generally speaking, difficulties encountered during the improvement of an Analytical
Dynamic Model (ADM) using vibration test data come from both the Spatial Coordinate
Incompatibility (SCI) and especially the Modal Coordinate Incompatibility (MCI)
between the ADM and the test data. Efforts were therefore made in this project to cope
with these two problems by extending some of the existing methods and also by
developing new methods with consideration of their feasibility, efficiency and accuracy.
A general description of this part of the project and the literature survey of this study
area are presented in part 1 of the thesis.
In part 2, in order to solve the SCI problem, a new extended Complete Modal
Expansion (CME) and a Branch Modal Expansion (BME) method were proposed
especially for the case when using a branch mode method to produce the ADM.
Application of these two methods and the existing physical expansion method were
demonstrated in a beam example in this part and were also used in some of the examples
later in this project.
In part 3, efforts were made to extend the existing Direct Matrix Updating (DMU) and
the Direct Parameter Identification (DPI) methods for solving the MCI problem using a
direct approach. Firstly a new Direct Modal Extension (DME) method was proposed
and compared with the DMU method when they were used to improve a reduced-size
ADM. Secondly, in order to overcome the main limitation of the existing DPI methods
in their practical use, an extended Corrected Modal Constraint (CMC) method was
In part 4, in order to achieve the feasibility and accuracy of ADM improvement, efforts
were then made in the study of the indirect approach. Firstly a procedure using a new
Orthogonality Sensitivity Method (OSM) working together with a model reduction
method was proposed. Secondly, a new Energy Error Estimation (EEE) method was
also presented. The original contribution of the EEE method is that the poorly modelled
stiffness and mass elements of an ADM can be identified and corrected accurately and
Applications of these new proposed methods were demonstrated by taking beam
examples. Further application of the EEE method was examined in a full-scale aircraft
tail plane example. A general discussion, conclusion and recommendation for further
study of these methods are presented in the fmal part 5.
Based on the study of this project, it is concluded that the feasibility and accuracy of the
direct methods described in part 3 of this thesis are at a low level for practical use.
Therefore, the main efforts and contributions in this project were made in the study of
the indirect methods described in part 4 of this thesis. It is concluded that both of the
new proposed OSM and EEE methods provide feasible tools for ADM improvement and
possess a high level of accuracy.