LaSDI: Parametric Latent Space Dynamics Identification

Jang

Engineering with Computers

et al. 2023

An improved nonintrusive parametric model order reduction (pMOR) approach is proposed for the flow field interpolation regarding fluid–structure interaction (FSI) objects. Flow field computation using computational fluid dynamics (CFD) requires excessive computational time and memory. Nonintrusive and data-driven MOR schemes have been proposed to overcome such limitations. The present methodology is implemented by both proper orthogonal decomposition (POD) and a modified Nouveau variational autoencoder (mNVAE). POD attempts to reduce the number of degrees of freedom (DOFs) on the precomputed series of the full-order model parametric result. The reduced DOF yields parametrically independent reduced bases and dependent coefficients. Then, mNVAE is employed for the interpolation of POD coefficients, which will be combined with POD modes for parametrically interpolated flow field generation. The present approach is assessed on the benchmark problem of a two-dimensional plunging airfoil and the highly nonlinear FSI phenomenon of the limit cycle oscillation. The comparison was executed against other POD-based generative neural network approaches. The proposed methodology demonstrates applicability on highly nonlinear FSI objects with improved accuracy and efficiency.

Section: Introductionmentioning

confidence: 99%

Parametric model order reduction by machine learning for fluid–structure interaction analysis

Jang

Engineering with Computers

et al. 2023

“…Among others, great success has been encountered by the physics informed neural networks (PINNs) [ 17 , 18 ], which have been applied in multiple contexts, including solid mechanics [ 19 ]. Other DL-based ROM techniques have come as an inspiration to handle the complex reduction process of dynamical systems, unveiling low-dimensional features from black-box data streams [ 20 , 21 , 22 , 23 ]. A peculiar perspective has been taken by Fresca et al.…”

Section: Introductionmentioning

confidence: 99%

Reduced Order Modeling of Nonlinear Vibrating Multiphysics Microstructures with Deep Learning-Based Approaches

Gobat

Fresca

Manzoni

et al. 2023

Sensors

Micro-electro-mechanical-systems are complex structures, often involving nonlinearites of geometric and multiphysics nature, that are used as sensors and actuators in countless applications. Starting from full-order representations, we apply deep learning techniques to generate accurate, efficient, and real-time reduced order models to be used for the simulation and optimization of higher-level complex systems. We extensively test the reliability of the proposed procedures on micromirrors, arches, and gyroscopes, as well as displaying intricate dynamical evolutions such as internal resonances. In particular, we discuss the accuracy of the deep learning technique and its ability to replicate and converge to the invariant manifolds predicted using the recently developed direct parametrization approach that allows the extraction of the nonlinear normal modes of large finite element models. Finally, by addressing an electromechanical gyroscope, we show that the non-intrusive deep learning approach generalizes easily to complex multiphysics problems.

“…Deep Learning autoencoders are one example, allowing one to learn coordinates on a curved manifold and hence providing excellent compression capabilities [11]. These autoenconders are made of an encoder which maps a high-dimensional state x to the latent state z [12][13][14][15]. The reverse operation which returns the estimate x is carried out by the decoder.…”

Section: Introductionmentioning

confidence: 99%

“…In such fashion, purely data-driven non-intrusive ROMs are obtained. Several examples of developments on this area are found in the literature, i.e., [7,[13][14][15]20,21]. Some have already been applied in the emulation of injectors flows.…”

Section: Introductionmentioning

confidence: 99%

Data Driven Models for the Design of Rocket Injector Elements

et al. 2022

Improving the predictive capabilities of reduced-order models for the design of injector and chamber elements of rocket engines could greatly improve the quality of early rocket chamber designs. In the present work, we propose an innovative methodology that uses high-fidelity numerical simulations of turbulent reactive flows and artificial intelligence for the generation of surrogate models. The surrogate models that were generated and analyzed are deep learning networks trained on a dataset of 100 large eddy simulations of a single-shear coaxial injector chamber. The design of experiments was created considering three design parameters: chamber diameter, recess length, and oxidizer–fuel ratio. The paper presents the methodology developed for training and optimizing the data-driven models. Fully connected neural networks (FCNNs) and U-Nets were utilized as surrogate-modeling technology. Eventually, the surrogate models for the global quantity, average, and root mean square fields were used in order to analyze the impact of the length of the post’s recess on the performances obtained and the behavior of the flow.