Probabilistic Analysis of an HSCT Modeled with an Equivalent Laminated Plate Wing
Mavris, Dimitri N.
Hayden, William T.
MetadataShow full item record
The High Speed Civil Transport (HSCT), a supersonic commercial transport currently under development, presents several challenges to traditional conceptual design. The current historical database used by many commercial transport design processes only include data for subsonic transports and therefore does not apply to innovative new configurations such as the HSCT. Therefore, physics-based, preliminary design tools must be used to model the characteristics of advanced aircraft in conceptual sizing routines. In addition, the evaluation of the aircraft design space often requires the analysis of many configurations in order to assess the impact of design constraints and determine the attainable range of system level metrics, a process which is very time consuming in both modeling and computer run time. To address these challenges, the equivalent plate structural analysis code ELAPS is used to model the wing structure of the HSCT and the Fast Probability Integration (FPI) technique is used to probabilistically assess the design space. After the ELAPS model is generated in a parametric manner, the structure is optimized to yield a weight for each component of the wing. A Response Surface Methodology approach is then implemented using Design of Experiments tables and an analysis of variance to generate response surface equations in terms of the most influential design variables for these wing component weights, as well as for the fuel volume available. These expressions are substituted into the sizing and synthesis code FLOPS in order to conduct system level design trade studies. FLOPS is subsequently enhanced with equations created from physics-based tools for the various disciplines to create a preliminary design synthesis tool. Ranges for the system level design variables are then introduced through the FPI technique, a probabilistic process which generates cumulative distributions for system level metrics such as take-off gross weight. This technique requires only 20 to 30 executions for FLOPS to generate these distributions, hence greatly reducing the time required to conduct the analysis. The results of this study indicate that the HSCT has only a 20% probability of achieving the system level design constraint of a 1,000,000 lb. take-off gross weight with the current level of technology and has no chance of achieving the desired goal of 750,000 lb. Through the use of new enabling technologies, however, these weight levels can be reduced to increase the probability of achieving technical feasibility and improve its economic viability. Future efforts will therefore focus on the evaluation of these technologies and their impact on system level performance and economics.