27th Structures, Structural Dynamics and Materials Conference · 1986 · 36 citations · 1 references
EngineeringMechanical EngineeringStructural AnalysisStructural DynamicsVibrationDynamic Response AnalysisNonlinear MechanicsNonlinear Structural AnalysisNonlinear Structural DynamicsStructural Response AnalysisSystem IdentificationMemory EffectsNonlinear ModelingNonlinear System IdentificationNonlinear System ModelingNonlinearity CharacterizationState Vector FormulationState Vector RepresentationStructural State Estimation
The force‑state map, which plots transmitted force versus mechanical state including position and rate effects, displays distinct, superposable features that uniquely identify common structural nonlinearities even when hysteresis appears in force‑stroke plots. The authors present a procedure to identify potentially strong nonlinear properties of structural members, such as joints, by expressing the transmitted force as a function of the member’s mechanical state. The method analyzes the influence of memory effects, transient response, and measurement uncertainty on precision, and is tested on three joint models with incomplete state measurements typical of real testing environments. Simulated data confirm the algorithm’s accuracy, and tests on real joint models demonstrate the procedure’s ability to estimate the complete state vector and reconstruct the measured nonlinear characteristics.
A procedure is presented for identifying the potentially strong nonlinear properties of structural members, such as joints, by expressing the force transmitted by the member as a function of its mechanical state. By explicitly including position and rate dependent effects, the surface of transmitted force versus state, the force-state map, has distinct, unique, superposable features for common structural nonlinearities, even those which appear to indicate hysteresis on a force-stroke presentation. An analysis is performed on the influence of true memory effects, transient response, and uncertainty in the measurements and system mass on the precision of the procedure. The successful identification of simulated data verifies the accuracy of the identification algorithm. Tests are then conducted on three actual joint models, with incomplete state measurements typical of an actual testing environment. The ability of the procedure to estimate the complete state vector and to analyze and reconstruct the measured nonlinear characteristics is demonstrated.
1