SCALING OF OPERATIONAL MODE SHAPES USING SENSITIVITY ANALYSIS

Recently, identification of dynamic properties of structures such as bridges, towers and building, which are not easy to excite, is possible by measuring only the output responses. Due to measuring only the responses when the structure is excited only by ambient or operational forces, these methods...

Full description

Bibliographic Details
Main Authors: محمد مهدی Khatibi, محمد رضا Ashory, علیرضا Albooyeh
Format: Article
Language:fas
Published: Semnan University 2009-12-01
Series:مجله مدل سازی در مهندسی
Subjects:
Online Access:https://modelling.semnan.ac.ir/article_1542_98111020360c3296050d823117809f9d.pdf
Description
Summary:Recently, identification of dynamic properties of structures such as bridges, towers and building, which are not easy to excite, is possible by measuring only the output responses. Due to measuring only the responses when the structure is excited only by ambient or operational forces, these methods are referred as output-only modal analysis or operational modal analysis (OMA). One drawback of OMA is that the mode shapes are not scaled. However the scaled mode shapes are required for some important applications such as damage detection or model updating. In this paper, a procedure is proposed for precise scaling of the operational mode shapes. The method is based on the sensitivity analysis of the effective parameters in the accuracy of scaling. The finite element model of a ten storey building is used to present the capability of the method. The natural frequencies and unscaled mode shapes have been estimated using FDD method and the mass change method is applied for scaling the mode shapes. The sensitivity of the results to the amount, location and the number of added masses and the number of modes under investigation have been evaluated. It is shown that the amount of added masses can be estimated by try and error of the FE model of structure and the masses should be distributed all over the structure to decrease the scaling error.
ISSN:2008-4854
2783-2538