Conceptual design

Aerospace engineering

Structural engineering

Aerodynamics

Engineering

Artificial intelligence

Stiffness

Computer science

Wing

Generative grammar

Mechanical engineering

Robot

Journal of Computational Design and Engineering

Gliding is generally one of the most efficient modes of flight in natural fliers that can be further emphasized in the aircraft industry to reduce emissions and facilitate endured flights. Natural wings being fundamentally responsible for this phenomenon are developed over millions of years of evolution. Artificial wings, on the other hand, are limited to the human-proposed conceptual design phase often leading to sub-optimal results. However, the novel Generative Design (GD) method claims to produce mechanically improved solutions based on robust and rigorous models of design conditions and performance criteria. This study investigates the potential applications of this Computer-Associated Design (CAsD) technology to generate novel micro aerial vehicle wing concepts that are structurally more stable and efficient. Multiple performance-driven solutions (wings) with high-level goals are generated by an infinite scale cloud computing solution executing a machine learning-based GD algorithm. Ultimately, the highest performing CAsD concepts are numerically analysed, fabricated, and mechanically tested according to our previous study, and the results are compared to the literature for qualitative as well as quantitative analysis and validations. It was concluded that the GD-based tandem wings’ (forewing and hindwing) ability to withstand fracture failure without compromising structural rigidity was optimized by 78% compared to its peer models. However, the weight was slightly increased by 11% with 14% drop in stiffness when compared to our models from previous study.

Generative design and fabrication of a locust-inspired gliding wing prototype for micro aerial robots