A neural network construction method for surrogate modeling of physics-based analysis

An innovative approach for integrated airline network and aircraft family optimization

An efficient methodology for multi-disciplinary design and optimization of transport was elaborated and developed. The methodology was implemented in a commercial known optimization framework. Semi-empirical methods were employed for wing weight estimation; a multi-block full-potential code was used for drag calculation; Vortex Lattice method was implemented for spanwise lift distribution in order to compute de aircraft maximum-lift coefficient via critical section method; a calibrated single- point Breguet simplified equation was considered for aircraft performance calculation. The optimization design variables are related to the wing planform and airfoil geometry and cruise speed. The design constraints were the fuel tank capacity, flight quality of the aircraft, and takeoff field length. A simple stability augmentation control system was implemented in order to compute its effects on optimal configurations. Multi-objective optimization tasks were performed accomplishing minimization of the block time and block fuel for a specified mission.

/pdf/wing-and-airfoil-optimized-design-of-transport-aircraft-344q0n8962.pdf

Wing and Airfoil Optimized Design of Transport Aircraft

A review of Russian computer modeling and validation in aerospace applications

An efficient framework for multi-disciplinary design and optimization of transport aircraft was elaborated and developed. Semi-empirical methods were employed for wing weight estimation; a multi-block full-potential code was used for drag calculation and maximum-lift coefficient estimation; a calibrated single-point Breguet simplified equation was considered for aircraft performance calculation. The optimization design variables are related to the wing planform geometry and cruise speed. The design constraints were the fuel tank capacity, flight quality of the aircraft, and takeoff field length. A simple stability augmentation control system was implemented in order to compute its effects on the optimal configurations. Multi-objective wing optimization cases were performed accomplishing minimization of the block time and block fuel for a given mission.

Optimal Conceptual Design of Transport Aircraft

This work shows a method for predicting pressure distributions over a generic bitrapezoidal wing using neural networks. The MatLab® Neural Network Toolbox was used for implementation and the training set was obtained using the BLWF® (Boundary Layer Wing-Fuselage) a full potential CFD code with boundary layer correction. Initially, a neural network considering only Mach speed, angle of attack and plan form parameters as design variables, was trained. Promising results motivated a generation of a more generic neural network, considering also airfoil parameters. Preliminary results show that this methodology can be an interesting solution for aircraft design, since it allows fast prediction of pressure distributions.

Juliano Cavalcanti

Papers

Optimal Conceptual Design of Transport Aircraft

Prediction of Pressure Distribution on Wings Using Neural Networks