Isolating and Quantifying Uncertainties in the Vibration Iso-lation Round-Robin Challenge

In the field of structural dynamics, mathematical models generate predictions and inform decisions about complex systems.For example, models may define a bridge's maximum allowable load, chronicle the dynamics of a jointed structure, or describe a driving car's vertical mechanical behavior.But significant discrepancies between observations and predictions often result from model-form uncertainty (e.g., missing dependencies or simplified forces).The Uncertainty Quantification in Vibration Isolation round-robin challenge considers modeling a vibration isolation experiment and the uncertainty that results from a simplified representation of the experimental configuration.As a result of model-form uncertainty, discrepancies emerge between experimental data and model predictions from a linear two-mass oscillator.Last year, we embedded a nonlinear, stochastic enrichment operator into the two-mass oscillator to correct the active damping force.This improved the time-domain amplitude of the interior mass's acceleration, but discrepancies in its time-domain frequency remained.This year, we focus on isolating and quantifying the measurement uncertainty in the experiment, uncertainties in the initial condition, and parameter uncertainties in the linear two-mass oscillator.After Bayesian calibration, the linear two-mass oscillator accurately predicts the resonant frequency of the active vibration isolation experiment.