Abstract:Generative models offer design diversity but tend to be computationally expensive, while non-generative models are computationally cost-effective but produce less diverse and often invalid designs. However, the limitations of non-generative models can be overcome with the introduction of augmented shape signature vectors (SSVs) to represent both geometric and physical information. This recent advancement has inspired a systematic comparison of the effectiveness and efficiency of generative and non-generative models in constructing design spaces for novel and efficient design exploration and shape optimization, which is demonstrated in this work. These models are showcased in airfoil/hydrofoil design, and a comparison of the resulting design spaces is conducted in this work. A conventional generative adversarial network (GAN) and a state-of-the-art generative model, the performance-augmented diverse generative adversarial network (PaDGAN), are juxtaposed with a linear non-generative model based on the coupling of the Karhunen–Loève Expansion and a physics-informed shape signature vector (SSV-KLE). The comparison demonstrates that, with an appropriate shape encoding and a physics-augmented design space, non-generative models have the potential to cost-effectively generate high-performing valid designs with enhanced coverage of the design space. In this work, both approaches were applied to two large foil profile datasets comprising real-world and artificial designs generated through either a profile-generating parametric model or a deep-learning approach. These datasets were further enriched with integral properties of their members' shapes, as well as physics-informed parameters. The obtained results illustrate that the design spaces constructed by the non-generative model outperform the generative model in terms of design validity, generating robust latent spaces with no or significantly fewer invalid designs when compared to generative models. The performance and diversity of the generated designs were compared to provide further insights about the quality of the resulting spaces. These findings can aid the engineering design community in making informed decisions when constructing design spaces for shape optimization, as it has been demonstrated that, under certain conditions, computationally inexpensive approaches can closely match or even outperform state-of-the art generative models.

What problem does this paper attempt to address?

The paper primarily explores the application and comparative effectiveness of generative and non-generative models in engineering shape optimization. The core issue of the study is to evaluate the efficiency of these two types of models in constructing design spaces to achieve efficient and innovative design exploration and shape optimization. Specifically, the paper focuses on the following points: 1. **Model Background**: Introduces the basic concepts of generative models (such as Generative Adversarial Networks [GANs]) and non-generative models (such as methods based on Karhunen–Loève Expansion [KLE]). Generative models can produce diverse designs but at a higher computational cost; whereas non-generative models, although computationally efficient, may generate less diverse and sometimes ineffective designs. 2. **Enhanced Non-Generative Method**: Proposes an enhanced non-generative method by introducing enhanced Shape Signature Vectors (SSVs) to simultaneously represent geometric and physical information. This method can overcome some limitations of traditional non-generative models. 3. **Model Comparison**: The paper compares a traditional Generative Adversarial Network (GAN) model, an advanced generative model—Performance Augmented Diverse Generative Adversarial Network (PaDGAN), and a linear non-generative model (based on KLE and physically enhanced SSV). These models are applied to airfoil/hydrofoil design, and the resulting design spaces are comparatively analyzed. 4. **Experimental Setup**: Two datasets are used for the experiments, including real airfoil data obtained from the UIUC airfoil database and artificial airfoil data generated through parametric models or deep learning methods. 5. **Research Findings**: The results show that, with appropriate shape encoding and physically enhanced design spaces, non-generative models can effectively generate high-performance valid designs at a lower cost and outperform generative models in terms of design validity and design space coverage. In summary, this paper aims to provide guidance to the engineering design community by systematically comparing the performance of generative and non-generative models in shape optimization tasks, particularly on how to choose the appropriate model based on specific circumstances when constructing design spaces.

Generative vs. Non-Generative Models in Engineering Shape Optimization

Generative VS non-Generative Models in Engineering Shape Optimization

Generative Creativity: Adversarial Learning For Bionic Design

Airfoil GAN: Encoding and Synthesizing Airfoils for Aerodynamic Shape Optimization

Generative Models for Anomaly Detection and Design-Space Dimensionality Reduction in Shape Optimization

Physics-guided training of GAN to improve accuracy in airfoil design synthesis

Airfoil Design Parameterization and Optimization using Bézier Generative Adversarial Networks

Diffusion Models Beat GANs on Topology Optimization

3D Design Using Generative Adversarial Networks and Physics-Based Validation

A generative design technique for exploring shape variations

Shape related constraints aware generation of Mechanical Designs through Deep Convolutional GAN

Deep Generative Design: Integration of Topology Optimization and Generative Models

Towards Goal, Feasibility, and Diversity-Oriented Deep Generative Models in Design

Design representation for performance evaluation of 3D shapes in structure-aware generative design

Deep Generative Model for Efficient 3D Airfoil Parameterization and Generation

Progress and recent trends in generative design

Deep Generative Models in Engineering Design: A Review

Aerodynamic shape optimization using graph variational autoencoders and genetic algorithms

MO-PaDGAN: Reparameterizing Engineering Designs for augmented multi-objective optimization

Generative Aerodynamic Design with Diffusion Probabilistic Models