A methodology for structural technology performance characterization to enable reduction of structural uncertainty
Corman, Jason A.
MetadataShow full item record
Government programs have been established to identify solutions that will significantly reduce the impact of aviation on the environment in the upcoming generations. Airframe structural technologies were identified as a category of potential solutions to meet environmental goals in the N+2 timeframe, and these technologies are assessed and compared by their ability to reduce airframe structural weight. The benchmark approach for structural technology weight reduction, i.e. performance, characterization uses a medium-to-high fidelity physics-based structural weight estimation approach. However, this approach considered only a single conceptual design point, or outer mold line (OML), and single structural layout design when comparing the structural technology to baseline, state of the art structure. It was hypothesized that treating weight reduction performance as a scalar and neglecting its functional relationship with the OML and structural layout was a significant source of epistemic uncertainty. This uncertainty would introduce risk in technology selection and implementation in conceptual design as well as experiment design for structural technology development and demonstration. A significant effort is required to estimate structural technology performance as a function of design spaces rather than a single design point. This thesis presents a repeatable, traceable methodology to characterize the functional performance relationship within a tollgate framework to mitigate expended efforts. Experiments were performed to demonstrate the approach on a test case, the Pultruded Rod Stitched Efficient Unitized Structure (PRSEUS) technology. These experiments examined performance at the technology level, structural layout level, and OML level design spaces. The impact of the performance relationship on 1) technology selection, 2) technology implementation in conceptual design, and 3) experiment design for technology development was also comparatively assessed to the benchmark scalar approach. It was shown that the ability to quantify the performance function using this methodology for the PRSEUS test case presents a significant advantage over the benchmark for these applications. Error distributions for treating weight reduction as a scalar rather than a function were on the same order as uncertainty distributions representative of a TRL 3 structural technology.