Concentration‐Specific Constitutive Modeling of Gelatin Based on Artificial Neural Networks

Abdolazizi, Kian P.; Linka, Kevin; Sprenger, Johanna; Neidhardt, Maximilian; Schlaefer, Alexander; Cyron, Christian J.

doi:10.1002/pamm.202000284

Cited by 2 publications

(3 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Based on [165], Abdolazizi et al [166] proposed a viscoelastic constitutive artificial neural networks (vCANNs) for anisotropic nonlinear viscoelasticity in the large-strain regime. By incorporating nonlinear strain coefficients and relaxation times as neural networks within the framework of the generalized Maxwell model, the vCANNs can adapt automatically during training.…”

Section: The Role Of ML Inferring Constitutive Relationsmentioning

confidence: 99%

Insight into constitutive theories of 4D printed polymer materials: a review

Rodriguez-Morales,

Duan,

et al. 2024

Smart Mater. Struct.

View full text Add to dashboard Cite

Four-dimensional (4D) printing has emerged as a branch of Additive Manufacturing (AM) that utilizes stimuli-responsive materials to generate three-dimensional (3D) structures with functional features. In this context, constitutive models play a paramount role in designing engineering structures and devices using 4D printing, as they help understand mechanical behavior and material responses to external stimuli, providing a theoretical framework for predicting and analyzing their deformation and shape-shifting capabilities. This article thoroughly discusses available constitutive models for single-printed and multi-printed materials. Later, we explore the role of machine learning algorithms in inferring constitutive relations, particularly in viscoelastic problems and, more recently, in shape memory polymers. Moreover, challenges and opportunities presented by both approaches for predicting the mechanical behavior of 4D printed polymer materials are examined. Finally, we concluded our discussion with a summary and some future perspectives expected in this field. This review aims to open a dialogue among the mechanics community to assess the limitations of analytical models and encourage the responsible use of emerging techniques, such as machine learning. By clarifying these aspects, we intend to advance the understanding and application of constitutive models in the rapidly growing field of 4D printing.

show abstract

Section: The Role Of ML Inferring Constitutive Relationsmentioning

confidence: 99%

Insight into constitutive theories of 4D printed polymer materials: a review

Rodriguez-Morales,

Duan,

et al. 2024

Smart Mater. Struct.

View full text Add to dashboard Cite

show abstract

“…In a similar approach, finite viscoelasticity is modeled by replacing the Helmholtz free energy function and dissipation potential with data-driven functions that a priori satisfy the second law of thermodynamics, using neural ordinary differential equations (NODEs). 54 Finally, very recently, a viscoelastic model called vCANN (viscoelastic constitutive artificial neural networks), which is directly building up on the concept of generalized Maxwell models enhanced with nonlinear strain (rate)-dependent properties, has been introduced by Abdolazizi et al 55 After the brief overview given above, it can be summarized that there are a variety of NN-based approaches to model inelasticity, with very different levels of incorporated physics. Most approaches were applied exclusively for describing one specific material class, elastoplasticity or viscoelasticity.…”

Section: Introductionmentioning

confidence: 99%

“…In a similar approach, finite viscoelasticity is modeled by replacing the Helmholtz free energy function and dissipation potential with data‐driven functions that a priori satisfy the second law of thermodynamics, using neural ordinary differential equations (NODEs) 54 . Finally, very recently, a viscoelastic model called vCANN (viscoelastic constitutive artificial neural networks), which is directly building up on the concept of generalized Maxwell models enhanced with nonlinear strain (rate)‐dependent properties, has been introduced by Abdolazizi et al 55 …”

Section: Introductionmentioning

confidence: 99%

A comparative study on different neural network architectures to model inelasticity

Rosenkranz

Kalina

Brummund

et al. 2023

Numerical Meth Engineering

View full text Add to dashboard Cite

The mathematical formulation of constitutive models to describe the path‐dependent, that is, inelastic, behavior of materials is a challenging task and has been a focus in mechanics research for several decades. There have been increased efforts to facilitate or automate this task through data‐driven techniques, impelled in particular by the recent revival of neural networks (NNs) in computational mechanics. However, it seems questionable to simply not consider fundamental findings of constitutive modeling originating from the last decades research within NN‐based approaches. Herein, we propose a comparative study on different feedforward and recurrent neural network architectures to model 1D small strain inelasticity. Within this study, we divide the models into three basic classes: black box NNs, NNs enforcing physics in a weak form, and NNs enforcing physics in a strong form. Thereby, the first class of networks can learn constitutive relations from data while the underlying physics are completely ignored, whereas the latter two are constructed such that they can account for fundamental physics, where special attention is paid to the second law of thermodynamics in this work. Conventional linear and nonlinear viscoelastic as well as elastoplastic models are used for training data generation and, later on, as reference. After training with random walk time sequences containing information on stress, strain, and—for some models—internal variables, the NN‐based models are compared to the reference solution, whereby interpolation and extrapolation are considered. Besides the quality of the stress prediction, the related free energy and dissipation rate are analyzed to evaluate the models. Overall, the presented study enables a clear recording of the advantages and disadvantages of different NN architectures to model inelasticity and gives guidance on how to train and apply these models.

show abstract

Concentration‐Specific Constitutive Modeling of Gelatin Based on Artificial Neural Networks

Cited by 2 publications

References 5 publications

Insight into constitutive theories of 4D printed polymer materials: a review

Insight into constitutive theories of 4D printed polymer materials: a review

A comparative study on different neural network architectures to model inelasticity

Contact Info

Product

Resources

About