A deep belief network to predict the hot deformation behavior of a Ni-based superalloy

Lin, Y.C.; Li, Jia; Chen, Mingsong; Liu, Yanxing; Liang, Yuqing

doi:10.1007/s00521-016-2635-7

Cited by 32 publications

(17 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Based on the training set, ANN models were developed, modeled, and trained. Structures and hyper‐parameters of a developed ANN model can be optimized to generate an excellent prediction accuracy; the numbers of hidden layers and hidden‐layer neurones are vital structural parameters for the model 28 . The numbers of hidden layers were defined as ranging from one to two, while the number of neurones per hidden layer was defined as ranging from 10 to 100 with a step size of 10.…”

Section: Resultsmentioning

confidence: 99%

“…Machine learning methods, which is the most important subfield of data‐driven science, has garnered considerable interest with the increase of experimental and computational data over the last few years. Most recently, these methods have evolved to predict the properties, 18‐22 damage, 23,24 fatigue, 25 and process‐related behaviors 26‐28 of materials. Machine learning is the science of programming computers that aims to extract knowledge and gain insight from large databases, wherein the program learns from previous computations to produce reliable, repeatable decisions and results.…”

Section: Introductionmentioning

confidence: 99%

“…Machine learning is the science of programming computers that aims to extract knowledge and gain insight from large databases, wherein the program learns from previous computations to produce reliable, repeatable decisions and results. Well‐trained machine learning models can provide alternative approaches toward experiments and simulations 18‐28 and can even guide the discovery and understanding of new materials with desired properties under the pressing need to shorten the development period 29‐31 . As a universal machine learning method, the artificial neural network (ANN) is widely used.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Rapid prediction and inverse design of distortion behaviors of composite materials using artificial neural networks

Luo

Zhang

et al. 2020

Polymers for Advanced Techs

View full text Add to dashboard Cite

If the process‐induced distortions (PIDs) of asymmetrical laminates can be predicted accurately and tailored at the early design stage, the production of curved panels from flat molds could be an attractive technique in a cost‐driven production environment. A data‐driven computational methodology which integrates the finite element method (FEM) and artificial neural network (ANN) is presented to rapidly predict the maximum PID and to perform high‐throughput screening of thermosetting‐matrix composites of an asymmetrical laminate for a targeted maximum PID. We performed a grid search on ANN architectures and hyper‐parameters using cross‐validation and obtained a well‐trained ANN model with high generalization performance. For the forward problem, the ANN model was adopted to predict the maximum PIDs of CYCOM X850 and CYCOM 977‐2 prepregs, which were subsequently verified experimentally. For inverse design, a large‐scale screening method based on the ANN model was utilized to determine the candidates for a targeted maximum PID, with an experimental demonstration using one of these candidates. The well‐trained ANN model provides an alternative approach to faster computation with high accuracy for the maximum PID prediction and further guides the discovery of materials with desired distortion behaviors.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Rapid prediction and inverse design of distortion behaviors of composite materials using artificial neural networks

Luo

Zhang

et al. 2020

Polymers for Advanced Techs

View full text Add to dashboard Cite

show abstract

“…At present, research on superalloys has focused on a variety of aspects. Some studies have investigated the microstructure analysis and hot deformation behavior of nickel-based superalloys [1,2,3,4,5,6,7,8,9,10,11,12,13]. In these studies, optical microscopy (OM), transmission electron microscopy (TEM), electron backscatter diffraction (EBSD), and other analytical techniques were used to study the grain morphology, grain boundary evolution, and dynamic recrystallization nucleation mechanism of nickel-based superalloys under different hot deformation conditions.…”

Section: Introductionmentioning

confidence: 99%

Study of the Dynamic Recrystallization Process of the Inconel625 Alloy at a High Strain Rate

Jia

Gao

Ji³

et al. 2019

Materials

View full text Add to dashboard Cite

High-temperature compression and electron backscatter diffraction (EBSD) techniques were used in a systematic investigation of the dynamic recrystallization (DRX) behavior and texture evolution of the Inconel625 alloy. The true stress–true strain curves and the constitutive equation of Inconel625 were obtained at temperatures ranging from 900 to 1200 °C and strain rates of 10, 1, 0.1, and 0.01 s−1. The adiabatic heating effect was observed during the hot compression process. At a high strain rate, as the temperature increased, the grains initially refined and then grew, and the proportion of high-angle grain boundaries increased. The volume fraction of the dynamic recrystallization increased. Most of the grains were randomly distributed and the proportion of recrystallized texture components first increased and then decreased. Complete dynamic recrystallization occurred at 1100 °C, where the recrystallized volume fraction and the random distribution ratios of grains reached a maximum. This study indicated that the dynamic recrystallization mechanism of the Inconel625 alloy at a high strain rate included continuous dynamic recrystallization with subgrain merging and rotation, and discontinuous dynamic recrystallization with bulging grain boundary induced by twinning. The latter mechanism was less dominant.

show abstract

“…Compared with other models such as Zerilli-Armstrong, the stain-compensated Arrhenius-type constitutive model can more accurately describe the flow behavior of alloys [1,13]. Recent researches combine with some typical intelligent method such as artificial neural network (ANN), support vector regression model, and deep belief networks, which are also popular in predicting deformation behavior [8,[18][19][20][21].…”

Section: Introductionmentioning

confidence: 99%

A Strain-Compensated Constitutive Model for Describing the Hot Compressive Deformation Behaviors of an Aged Inconel 718 Superalloy

Zhou

Chen

Qin

2018

High Temperature Materials and Processes

View full text Add to dashboard Cite

The hot deformation behaviors of an aged Inconel 718 superalloy are investigated by isothermal compression experiments at four strain rates and five deformation temperatures on a Gleeble-3500 thermo-mechanical simulator. The experimental results show that the true stresses are obviously affected by strain rate and deformation temperature. The true stress increases rapidly at the beginning of hot compressive deformation, which ascribes to the intense work hardening. The true stresses at high deformation temperatures are lower than those at lower deformation temperatures. The dynamic softening induced by DRX is weak at a relatively low deformation temperature. A strain-compensated Arrhenius-type constitutive equation linked with true stress, strain rate and deformation temperature is developed for the studied superalloy. The material constants (\alpha , n, Q and A) in the developed model are expressed as the functions of true strain. The flow stresses calculated by the developed constitutive equation are nicely consistent with the experimental ones, which confirms that the developed constitutive equation can accurately describe the hot deformation behaviors of the studied superalloy.

show abstract

A deep belief network to predict the hot deformation behavior of a Ni-based superalloy

Cited by 32 publications

References 57 publications

Rapid prediction and inverse design of distortion behaviors of composite materials using artificial neural networks

Rapid prediction and inverse design of distortion behaviors of composite materials using artificial neural networks

Study of the Dynamic Recrystallization Process of the Inconel625 Alloy at a High Strain Rate

A Strain-Compensated Constitutive Model for Describing the Hot Compressive Deformation Behaviors of an Aged Inconel 718 Superalloy

Contact Info

Product

Resources

About