TY - JOUR
T1 - Simulation of dynamics of beam structures with bolted joints using adjusted Iwan beam elements
AU - Song, Y.
AU - Hartwigsen, C. J.
AU - McFarland, D. M.
AU - Vakakis, A. F.
AU - Bergman, L. A.
N1 - Funding Information:
This work was partially supported by Sandia National Laboratories through Contract Number DOE SNL BF-0162, Dr. Daniel Segalman, Project Director, and by the Office of Naval Research through Contract Number N00014-00-1-10187, Dr. Luise Couchman, Project Director.
PY - 2004/5/21
Y1 - 2004/5/21
N2 - Mechanical joints often affect structural response, causing localized non-linear stiffness and damping changes. As many structures are assemblies, incorporating the effects of joints is necessary to produce predictive finite element models. In this paper, we present an adjusted Iwan beam element (AIBE) for dynamic response analysis of beam structures containing joints. The adjusted Iwan model consists of a combination of springs and frictional sliders that exhibits non-linear behavior due to the stick-slip characteristic of the latter. The beam element developed is two-dimensional and consists of two adjusted Iwan models and maintains the usual complement of degrees of freedom: transverse displacement and rotation at each of the two nodes. The resulting element includes six parameters, which must be determined. To circumvent the difficulty arising from the non-linear nature of the inverse problem, a multi-layer feed-forward neural network (MLFF) is employed to extract joint parameters from measured structural acceleration responses. A parameter identification procedure is implemented on a beam structure with a bolted joint. In this procedure, acceleration responses at one location on the beam structure due to one known impulsive forcing function are simulated for sets of combinations of varying joint parameters. A MLFF is developed and trained using the patterns of envelope data corresponding to these acceleration histories. The joint parameters are identified through the trained MLFF applied to the measured acceleration response. Then, using the identified joint parameters, acceleration responses of the jointed beam due to a different impulsive forcing function are predicted. The validity of the identified joint parameters is assessed by comparing simulated acceleration responses with experimental measurements. The capability of the AIBE to capture the effects of bolted joints on the dynamic responses of beam structures, and the efficacy of the MLFF parameter identification procedure, are demonstrated.
AB - Mechanical joints often affect structural response, causing localized non-linear stiffness and damping changes. As many structures are assemblies, incorporating the effects of joints is necessary to produce predictive finite element models. In this paper, we present an adjusted Iwan beam element (AIBE) for dynamic response analysis of beam structures containing joints. The adjusted Iwan model consists of a combination of springs and frictional sliders that exhibits non-linear behavior due to the stick-slip characteristic of the latter. The beam element developed is two-dimensional and consists of two adjusted Iwan models and maintains the usual complement of degrees of freedom: transverse displacement and rotation at each of the two nodes. The resulting element includes six parameters, which must be determined. To circumvent the difficulty arising from the non-linear nature of the inverse problem, a multi-layer feed-forward neural network (MLFF) is employed to extract joint parameters from measured structural acceleration responses. A parameter identification procedure is implemented on a beam structure with a bolted joint. In this procedure, acceleration responses at one location on the beam structure due to one known impulsive forcing function are simulated for sets of combinations of varying joint parameters. A MLFF is developed and trained using the patterns of envelope data corresponding to these acceleration histories. The joint parameters are identified through the trained MLFF applied to the measured acceleration response. Then, using the identified joint parameters, acceleration responses of the jointed beam due to a different impulsive forcing function are predicted. The validity of the identified joint parameters is assessed by comparing simulated acceleration responses with experimental measurements. The capability of the AIBE to capture the effects of bolted joints on the dynamic responses of beam structures, and the efficacy of the MLFF parameter identification procedure, are demonstrated.
UR - http://www.scopus.com/inward/record.url?scp=2342486078&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=2342486078&partnerID=8YFLogxK
U2 - 10.1016/S0022-460X(03)00499-1
DO - 10.1016/S0022-460X(03)00499-1
M3 - Article
AN - SCOPUS:2342486078
SN - 0022-460X
VL - 273
SP - 249
EP - 276
JO - Journal of Sound and Vibration
JF - Journal of Sound and Vibration
IS - 1-2
ER -