TY - JOUR
T1 - Fishbone model-based inversion to estimate physical parameters of elastic structures under earthquake excitations
AU - Kajiwara, Koichi
AU - Kishida, Akiko
AU - Fujiwara, Jun
AU - Enokida, Ryuta
N1 - Publisher Copyright:
Copyright © 2023 Kajiwara, Kishida, Fujiwara and Enokida.
PY - 2023
Y1 - 2023
N2 - This study established an inversion based on a fishbone model to estimate physical parameters from the responses of elastic building structures subjected to an earthquake. A fishbone model, which has rotational springs and dashpots in addition to the elements in a lumped mass model, is effective for demonstrating structural rotations that happen at the connections of columns and beams. This model is commonly applied to computational calculations of seismic responses of structures and is classified into a forward problem obtaining responses from known systems and excitations. Although its effectiveness for the forward problem has been well demonstrated, it has rarely been applied to the inverse problem, where structural properties are estimated from known responses and excitations. First, this study inverted multi/single-mass-system fishbone models. Then, the inversion was applied to an elastic fishbone model of a 3-mass system, which was built based on an E-Defense shaking table experiment, and its structural responses were numerically simulated. This numerical simulation demonstrated its effectiveness for accurately estimating parameters in the fishbone model of the 3-mass system, especially when its structural responses are not contaminated by noises. Lastly, it was applied to responses containing some noise to examine its influence on the estimation accuracy. The estimation accuracy of damping elements was found to be sensitive to noise, whereas that of stiffness was more insensitive than the damping elements. The proposed inversion is particularly suitable for estimating rotational stiffness, which is not obtainable from the inversion of lumped mass systems.
AB - This study established an inversion based on a fishbone model to estimate physical parameters from the responses of elastic building structures subjected to an earthquake. A fishbone model, which has rotational springs and dashpots in addition to the elements in a lumped mass model, is effective for demonstrating structural rotations that happen at the connections of columns and beams. This model is commonly applied to computational calculations of seismic responses of structures and is classified into a forward problem obtaining responses from known systems and excitations. Although its effectiveness for the forward problem has been well demonstrated, it has rarely been applied to the inverse problem, where structural properties are estimated from known responses and excitations. First, this study inverted multi/single-mass-system fishbone models. Then, the inversion was applied to an elastic fishbone model of a 3-mass system, which was built based on an E-Defense shaking table experiment, and its structural responses were numerically simulated. This numerical simulation demonstrated its effectiveness for accurately estimating parameters in the fishbone model of the 3-mass system, especially when its structural responses are not contaminated by noises. Lastly, it was applied to responses containing some noise to examine its influence on the estimation accuracy. The estimation accuracy of damping elements was found to be sensitive to noise, whereas that of stiffness was more insensitive than the damping elements. The proposed inversion is particularly suitable for estimating rotational stiffness, which is not obtainable from the inversion of lumped mass systems.
KW - fishbone model
KW - physical parameter estimation
KW - structural health monitoring
KW - system identification
KW - time domain inversion
UR - http://www.scopus.com/inward/record.url?scp=85167345265&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85167345265&partnerID=8YFLogxK
U2 - 10.3389/fbuil.2023.1201048
DO - 10.3389/fbuil.2023.1201048
M3 - Article
AN - SCOPUS:85167345265
SN - 2297-3362
VL - 9
JO - Frontiers in Built Environment
JF - Frontiers in Built Environment
M1 - 1201048
ER -