Publications

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Structural dynamic models of mechanical, aerospace, and civil structures often involve connections of multiple subcomponents with rivets, bolts, press fits, or other joining processes. Recent model order reduction advances have been made for jointed structures using appropriately defined whole joint models in combination with linear substructuring techniques. A whole joint model condenses the interface nodes onto a single node with multi-point constraints resulting in drastic increases in computational speeds to predict transient responses. One drawback to this strategy is that the whole joint models are empirical and require calibration with test or high-fidelity model data. A new framework is proposed to calibrate whole joint models by computing global responses from high-fidelity finite element models and utilizing global optimization to determine the optimal joint parameters. The method matches the amplitude dependent damping and natural frequencies predicted for each vibration mode using quasi-static modal analysis.

Complex mechanical structures are often subjected to random vibration environments. One strategy to analyze these nonlinear structures numerically is to use finite element analysis with an explicit solver to resolve interactions in the time domain. However, this approach is impractical because the solver is conditionally stable and requires thousands of iterations to resolve the contact algorithms. As a result, only short runs can be performed practically because of the extremely long runtime needed to obtain sufficient sampling for long-time statistics. The proposed approach uses a machine learning algorithm known as the Long Short-Term Memory (LSTM) network to model the response of the nonlinear system to random input. The LSTM extends the capability of the explicit solver approach by taking short samples and extending them to arbitrarily long signals. The efficient LSTM algorithm enables the capability to perform Monte Carlo simulations to quantify model-form and aleatoric uncertainty due to the random input.

Structural dynamic models of mechanical, aerospace, and civil structures often involve connections of multiple subcomponents with rivets, bolts, press fits, or other joining processes. Recent model order reduction advances have been made for jointed structures using appropriately defined whole joint models in combination with linear substructuring techniques. A whole joint model condenses the interface nodes onto a single node with multi-point constraints resulting in drastic increases in computational speeds to predict transient responses. One drawback to this strategy is that the whole joint models are empirical and require calibration with test or high-fidelity model data. A new framework is proposed to calibrate whole joint models by computing global responses from high-fidelity finite element models and utilizing global optimization to determine the optimal joint parameters. The method matches the amplitude dependent damping and natural frequencies predicted for each vibration mode using quasi-static modal analysis.

An assessment of two methodologies used at Sandia National Laboratories to model mechanical interfaces is performed on the Ministack finite element model. One method uses solid mechanics models to model contacting surfaces with Coulomb frictional contact to capture the physics. The other, termed the structural dynamics reduced order model, models the interface with a simplified whole joint model using four-parameter Iwan elements. The solid mechanics model resolves local kinematics at the interface while the simplified structural dynamics model is significantly faster to simulate. One of the current challenges to using the whole joint model is that it requires calibration to data. A novel approach is developed to calibrate the reduced structural dynamics model using data from the solid mechanics model to match the global dynamics of the system. This is achieved by calibrating to amplitude dependent frequency and damping of the system modes, which are estimated using three different approaches.

Abstract not provided.

Nonlinear joints and interfaces modeled with a discrete four-parameter Iwan element are defined by parameters that are often unknown a priori or require calibration to get better agreement with test data. While this constitutive model has been validated experimentally, its drawback lies in the difficulty of identifying the correct coefficients. This work proposes a parameter estimation approach using a genetic algorithm to minimize the residual between experimental and model data. Global optimization schemes have the ability to find global minima/maxima of a broad parameter space but require a very large number of function evaluations. This research focuses on decreasing the computational cost of the optimization scheme by developing a simplified model of the structure of interest and defining the objective function with amplitude dependent frequencies and damping ratios. A recently developed quasi-static modal analysis technique is used to determine these amplitude dependent properties of the model at a significantly reduced cost in comparison to solutions obtained with numerical time integration. This technique is demonstrated on a structure termed the Ministack which contains a foam-to-metal interface held together with a press fit joint.

Abstract not provided.