Experimental study of strain prediction on wave induced structures using modal decomposition and quasi static Ritz vectors

TY - JOUR

T1 - Experimental study of strain prediction on wave induced structures using modal decomposition and quasi static Ritz vectors

AU - Skafte, Anders

AU - Kristoffersen, Julie Carøe

AU - Vestermark, Jonas

AU - Tyge Tygesen, Ulf

AU - Brincker, Rune

PY - 2017

Y1 - 2017

N2 - Offshore structures are continuously subjected to dynamic loading from wind and waves which makes fatigue an important parameter for the structures expected lifetime. Monitoring the vibrations of the structure using real time operating data enables an assessment of the general health state of the structure. This paper proposes a method for an accurate full-field prediction of the strain history. Experimental mode shapes are found by the use of operational modal analysis and expanded to strain modes using a well correlated finite element model. The measured response from the structure is divided into two parts using complementary filters: Low frequency response caused by the quasi-static effect of the waves acting on the structure, and the high frequency response given by the modal properties of the structure. The high frequency response is then decomposed into modal coordinates using the experimental mode shapes. Strain histories are predicted by multiplying the modal coordinates with the expanded strain mode shapes. The low frequency response is decomposed using Ritz-vectors corresponding to the shapes that the structure vibrates with due to the wave loading. Strain Ritz-vectors are then extracted from the finite element model by applying a load corresponding to a representative wave and the strain history for the low frequency response is found by multiplying the decomposed signal with the strain Ritz-vectors. Finally the combined strain history is found by adding the strain histories from the low and high frequency responses. To validate the theory tests were performed on a scaled model of an offshore structure where the strain history was predicted using only the response from the accelerometers.

AB - Offshore structures are continuously subjected to dynamic loading from wind and waves which makes fatigue an important parameter for the structures expected lifetime. Monitoring the vibrations of the structure using real time operating data enables an assessment of the general health state of the structure. This paper proposes a method for an accurate full-field prediction of the strain history. Experimental mode shapes are found by the use of operational modal analysis and expanded to strain modes using a well correlated finite element model. The measured response from the structure is divided into two parts using complementary filters: Low frequency response caused by the quasi-static effect of the waves acting on the structure, and the high frequency response given by the modal properties of the structure. The high frequency response is then decomposed into modal coordinates using the experimental mode shapes. Strain histories are predicted by multiplying the modal coordinates with the expanded strain mode shapes. The low frequency response is decomposed using Ritz-vectors corresponding to the shapes that the structure vibrates with due to the wave loading. Strain Ritz-vectors are then extracted from the finite element model by applying a load corresponding to a representative wave and the strain history for the low frequency response is found by multiplying the decomposed signal with the strain Ritz-vectors. Finally the combined strain history is found by adding the strain histories from the low and high frequency responses. To validate the theory tests were performed on a scaled model of an offshore structure where the strain history was predicted using only the response from the accelerometers.

KW - Modal expansion

KW - Monitoring

KW - Offshores structures

KW - Ritz-vectors

KW - Strain estimation

KW - Wave load

UR - http://www.scopus.com/inward/record.url?scp=85009999968&partnerID=8YFLogxK

U2 - 10.1016/j.engstruct.2017.01.014

DO - 10.1016/j.engstruct.2017.01.014

M3 - Journal article

SN - 0141-0296

VL - 136

SP - 261

EP - 276

JO - Engineering Structures

JF - Engineering Structures

ER -