Damage Identification in Offshore Platforms Using Frequency Response Functions

Abstract

The damage identification problem in offshore platforms, specifically in jacket platforms, has been studied since the seventies. In spite of the effort dispended, there is still no spread out methodology that can be applied in this kind of structure, due its specific characteristics (structural complexity, load and operational conditions, etc.). The development of such methods would be extremely useful to preview structural failures, once this kind of structures is almost completely under water. Most of the developed methods for damage identification use modal parameters, by comparisons of different structural time conditions, and the damage may be expressed through the observed discrepancy. This paper presents a methodology to locate damage based on an iterative method. It was developed an optimization software which uses the Goal Programming Technique to choose the best set of variables that minimizes the difference between experimental and numerical Frequency Response Functions (FRFs), which can better represent the structure situation. Some numerical simulations were performed on a scaled fixed oil platform model to verify the efficiency of the method. The results show that such method is efficient when damages are restricted to a set of optimization variables.