Friday, April 13, 2012
3:00 pm
Guggenheim 133 (Lees-Kubota Lecture Hall) – Guggenheim Aeronautical Laboratory

GALCIT Colloquium

Recent Experience Applying Advanced Modeling and Uncertainty Quantification to Spacecraft Integrated Modeling
Lee Peterson, JPL
Flight system certification using an integrated system model is increasingly becoming important to the success of space missions. As science missions continue to push the edge of what can be tested fully before flight, models are becoming more central to validating engineering performance. However, the fidelity afforded by conventional tools can be too limiting to adequately anticipate complex system interactions that otherwise would be detected by testing. High fidelity models including coupled, nonlinear system effects, along with rigorous model verification, validation, and uncertainty quantification are necessary.

This talk will discuss these challenges and attempt to define some needed technology advancements. The focus of the talk will summarize recent experience applying advanced modeling and simulation tools developed at Sandia National Laboratories to notional space flight systems. For this study, the Sandia-developed Sierra thermal tool Aria and the solid mechanics tool Adagio were coupled to the JPL MACOS optics analysis tool. Unlike conventional tools, where results would be passed from a separate thermal to structural to optical code, this toolset creates a truly coupled multiphysics simulation. This allows iterators such as DAKOTA to apply efficient, directed search algorithms for uncertainty quantification. Moreover, the thermal, structural and optical meshes can be different, allowing independent convergence of the meshes, so that numerical convergence error can be estimated and optimized.

A major conclusion from this study was these advanced tools offer a quantum leap in analysis capability for spacecraft development. Future research needs will also be discussed, including methods for using these techniques to design model validation test campaigns that result in an acceptable level of system performance uncertainty and margin.

Contact Xin Ning at 626-395-3073
Add this event to my calendar