View Video Presentation: https://doi.org/10.2514/6.2021-3059.vid A new approach of unsupervised neural network-based kriging is proposed to capture non-stationary responses under uncertainty of a complex and high-dimensional system, such as a conceptual hypersonic vehicle. It is well known that typical kriging formulated with the assumption of stationarity of data variations fails to capture non-stationary system behaviors and having a large number of training samples causes numerical difficulties in kriging modeling, especially for high-dimensional aerospace systems. To address the computational challenges, the proposed approach uses unsupervised learning to cluster samples by localized spatial correlations of data variations. For each cluster, a kriging model is constructed by training a neural network model of correlation hyperparameters to capture non-stationary data variations. The predictions from the multiple kriging models are aggregated into a consensus prediction by using non-deterministic gaussian regression. The proposed method would enable us to 1) capture non-stationary system behaviors, 2) divide one large data set into multiple clusters with manageable sizes to avoid computational issues, 3) use parallel modeling of localized krigings for identified clusters, and 4) update local models adaptively and efficiently, instead of retraining the entire model. The proposed method is demonstrated with fundamental analytical examples, as well as a representative hypersonic vehicle design problem.
Current fatigue life modeling with respect to defects is only dependent on the defect size and the applied cyclic stress for a given material. This paper augments the process to include defect location into the model analysis for a more precise prediction of the number of cycles to failure and predict where finial failure could occur within a component. The focus is a turbine blade structure using nickel-based superalloy 718 subjected to a pure vibration environment. The augmented model predicts component life using a stress map from the frequency analysis of the developed Finite Element Models (FEMs) and measured or predicted defect sizes and locations. Printed test specimen are evaluated to experimentally validate the capabilities of the augmented model to predict fatigue life and crack initiation regions.
A computational capability is developed for the optimum design of radial drilling machine structure to satisfy static rigidity and natural frequency requirements using finite element idealization. The radial drilling machine structure is idealized with frame elements and is analyzed by using different combinations of cross sectional shapes for the radial arm and the column. From the results obtained, the best combination of cross sectional shapes is suggested for the structure. With this combination of cross sectional shapes, mathematical programming techniques are used to find the minimum weight design of the radial drilling machine structure. A sensitivity analysis is conducted about the optimum point to find the effects of changes in design variables on the structural weight and the response quantities.
The authors discuss the framework of a knowledge-based expert system for studying the survivability of the aerospace structures exposed to high energy lasers using VAASEL (vulnerability analysis of aerospace structures exposed to lasers) software. VAASEL is a synthesis tool built around NASTRAN and ASTROS programs. The knowledge base involves threat characterization, temperature distribution, failure prediction, linear and nonlinear statics, air loads and aeroelastic disciplines. A description of the VAASEL expert system modules, the graphics interface and the input data generator is included.< >
Advances in hypersonic vehicle technologies, including high-temperature materials, lightweight structures, and high thrust-to-weight ratio engines, indicate that a strategic vehicle at higher Mach numbers (approximately 6) is feasible. An efficient performance analysis method is developed along with a control strategy to evaluate potential hydrogen-fueled turbojet/ramjet propulsion systems for advanced technology hypersonic cruise vehicles. A conceptual strategic reconnaissance mission is evaluated for tradeoffs between throttle and angle of attack control in minimizing fuel consumption/maximizing range. Many mission, flight, and vehicle-related requirements and constraints are satisfied in the design process. In addition, powered hypersonic flight produces unique performance characteristics not encountered at subsonic speeds.
Given experimental data measured from an engineering system, response predictions by a stochastic simulation model involve both parametric uncertainty and random errors. Also, model form uncertainty arises when two or more simulation models predict the responses of an engineering system because it is beyond capability to identify the best approximating model among the considered model set. In this research, a methodology is developed to quantify model probability using measured deviations between experimental data and model predictions of the data under a Bayesian statistical framework. Model averaging is used to combine the predictions of a system response by a model set into a single prediction. Then, a nonlinear spring-mass system is used to demonstrate the process for implementing model averaging. Finally, the methodology is applied to the engineering benefits of a laser peening process, and a confidence band for a residual stress field is established to indicate the reliability of the composite prediction of the stress field.
This paper presents an optimization routine to solve a Modified Multi- Vehicle Dubins Traveling Salesman Problem with time window constraints, non-constant distances between points, and delivery item constraints. The problem represents the delivery of items via fixed wing aircraft to multiple delivery points with arrival time and item weight constraints. The problem was solved via a genetic algorithm utilizing modified crossover and mutation functions. The modified crossover function generates an estimate of the minimum and maximum arrival times at a point to determine whether aircraft speed or point order should be adjusted. The modified mutation function affects either point order or arrival order. Non-integer variables were optimized during a subroutine of the genetic algorithm. A feasible, near-optimal solution was able to be calculated by the algorithm.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)