Model Update of a Micro Air Vehicle (MAV) Flexible Wing Frame with Uncertainty Quantification

This paper describes a procedure to update parameters in the finite element model of a Micro Air Vehicle (MAV) to improve displacement predictions under aerodynamics loads. Because of fabrication, materials, and geometric uncertainties, a statistical approach combined with Multidisciplinary Design Optimization (MDO) is used to modify key model parameters. Static test data collected using photogrammetry are used to correlate with model predictions. Results show significant improvements in model predictions after parameters are updated; however, computed probabilities values indicate low confidence in updated values and/or model structure errors. Lessons learned in the areas of wing design, test procedures, modeling approaches with geometric nonlinearities, and uncertainties quantification are all documented. INTRODUCTION Micro air vehicles (MAV), because of their small size, weight and flexibility, can provide unprecedented new capabilities to many sectors in our community. For example, firefighters can deploy an MAV in a burning building to search for survivors. Another example is surveillance of urban areas using cameras on-board the MAV. To develop these systems for commercial use, government, industry, and several universities partners are joining forces to mature the technology. By combining biologically inspired concepts with modern composite materials new innovative vehicles are emerging. For example, the University of Florida has a vehicle design using a truss-like structure constructed of graphite/epoxy material covered with a thin transparent monofilm membrane. In this case, flexible wings allow the vehicle to naturally adapt to aerodynamic changes. This adaptability results in smooth flight characteristics and unprecedented agility even under gusty conditions. Although this class of new vehicles is quite different from conventional vehicles, flight controls, will likely be developed using traditional methods. This involves wind tunnel test and model validation under aerodynamic loads to predict the vehicle behavior. Unfortunately, because of the low cost associated with these vehicles, analytical models are rarely pursued. This is the area this work is focused on. In structural modeling and validation, a significant part of this effort deals with the problem of parameter updates to reconcile differences between test and analysis. This