|Benchmarks in Multidisciplinary Optimization and Design for Affordable Military Vehicles
|Applied Vehicle Technology|
Benchmarks, Multidisciplinary Design Optimization, Multiobjective and multiconstraint optimization, Vehicle Design, Verification and Validation
Implementation of rapid analytical design development methodologies can lead to reduced technology risk, reduce discovery of late defects (and hence reduce the need to correct or compensate for late defects), in addition to the reduced cycle time and reduced cost of development. An automated multidisciplinary approach also permits a more in depth consideration of operational and maintenance issues in the design process which in turn will reduce operational and maintenance costs. Verification and validation (V&V) are key tools to developing successful rapid analytical design methodologies. The range of available methodologies and processes need to be evaluated and benchmarked.
The primary purpose of this task group is to identify and collect enduring benchmark problems in Multidisciplinary Optimization and Design (MDO) that represent a variety of multidisciplinary optimization and design issues for military vehicles. The purpose of the benchmark problems is to aid the development, assessment and promotion of multidisciplinary optimization and design methods. The availability of benchmark problems provide a means of verifying and validating new methods, define a process for multidisciplinary optimization and design, show the potential along with the limitations, and provide confidence in analytical results.
Of primary interest to this activity is establishing standards and a repository for multidisciplinary optimization and design for land, sea and air military vehicles. The standards for bench mark problems and resulting MDO design studies will define the information needed to sufficiently describe the bench mark and its area of applicability including its limits of applicability. Each problem in the repository of bench mark problems will adhere to the standards set forth by the group.
The topics of interest to this working group fall into three categories: Optimization and design methods, analytical methods, and physical testing. Problems of interest in these categories address land, sea and air vehicles.
Bench marks for optimization and design methods address techniques such as multi-objective and multi-level optimization, the applicability of the methods, optimization and design for robustness, and the processes associated with each method.
Bench marking computational methods represent to some degree the degenerative case of optimization and design, where the design space has been reduced to a single point. However, it is still important to understand these problems, especially the applicability and limits of the methods for that defines the part of the design space in which reasonable results can be expected. Computational methods issues of interest include analysis tool limits, tool fidelity interaction, discipline coupling, virtual representation of manufacturing quality to address scale-up issues, and uncertainty quantification.
Physical bench marks are used to validate point designs within the design space to increase confidence in the computational results. Of primary interest with the physical bench marks is to obtain as much information as possible regarding the test to increase reusability of test results.