Artificial neural network approach to predict the fracture parameters of the size effect model for concrete

Yan, Yun; Ren, Quan; Xia, Ning; Shen, Luming; Gu, Jiming

doi:10.1111/ffe.12309

Cited by 19 publications

(8 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The complex constitution of concrete results in the sophisticated damage experience in the material under loading. Specifically, the arbitrary distribution of numerous initial defects in the material causes the localization of stresses, and the localized stresses produce the complex process of damage in an increasingly complicated way during load process. Additionally, experimental studies had been conducted to understand the damage mechanism in concrete.…”

Section: Introductionmentioning

confidence: 99%

A simple damage model for concrete considering irreversible mode‐II microcracks

Shan

Ouyang

et al. 2016

Fatigue Fract Eng Mat Struct

View full text Add to dashboard Cite

A B S T R A C T A simple damage model with the concept of mode-II microcracks on crack wall contributing to the irreversible strains for concrete is developed. By applying the micromechanics method, a microcell-based damage model is introduced to understand the damage behaviour. Further, by introducing the physical interpretation of the damage variable using thermodynamic method, a novel damage variable (irreversible-damage variable) is proposed to describe the irrecoverable strains generated by both mode-II microcracks and irreversible-frictional sliding. With this methodology, a simple continuum damage mechanics model is developed in which both elastic and irreversible damages are considered. As demonstrated by the comparison with experimental results, the proposed model is characterized by accuracy of solutions, sufficiency of physical sense and convenience of implementation. CDM = continuum damage mechanics FPZ = fracture process zone RVE = representative volume element STD = standard deviation SIF = stress intensity factor D = damage variable tensor D e , ω = elastic-damage variable tensor D i = irreversible-damage variable tensor I = four order identity tensor, p i ± = fitting parameters related to damage properties D ± = scalar damage for tensile/compressive uniaxial case D i ± = scalar irreversible-damage for uniaxial tensile/compressive case E = nominal Young's modulus E 0 = initial Young's modulus E = effective Young's modulus E d = degradation of Young's modulus K I = stress intensity factor K IC = critical stress intensity factor 2a = length of crack b = crack opening due to mode-II microcracks b′ = crack opening due to irreversible-frictional sliding da = increment of crack length on crack tip i = number of microcell r = stress ratio α = parameter considering the bia-compressive effects ε 0,δ , ε de,δ , ε di,δ = small additive strain Correspondence: Z. Shan.

show abstract

Section: Introductionmentioning

confidence: 99%

A simple damage model for concrete considering irreversible mode‐II microcracks

Shan

Ouyang

et al. 2016

Fatigue Fract Eng Mat Struct

View full text Add to dashboard Cite

show abstract

“…Based on the previous studies [ 28 , 37 ], one hidden layer is applied due to the limited input and output variables. The network with too many hidden layers or neurons can be easily over-trained and will not sufficiently predict the new data.…”

Section: Methodsmentioning

confidence: 99%

“…Yan et al [ 28 ] reported the following formulation to estimate the neuron number in the single hidden layer:

where

and

the neuron numbers of hidden, input and output layers, respectively.

is a constant, ranging from 1 to 10.…”

Section: Methodsmentioning

confidence: 99%

“…Liu and Athanasiou [ 23 ] successfully captured the complex relationship among the plane–strain stress intensity factor, the indentation load, and the specimen dimensions through ANN. Yan et al [ 28 ] proved that ANN models are viable for predicting fracture parameters with high accuracy by using only 77 sets of experimental data. For predicting the rock strength, the previous ANN models primarily take the inherent physical properties of rock materials into account, e.g., porosity [ 29 ], P wave velocity [ 29 ], unit weight [ 30 ] and point load index [ 29 , 30 ].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Using Artificial Neural Networks to Predict Influences of Heterogeneity on Rock Strength at Different Strain Rates

2021

View full text Add to dashboard Cite

Pre-existing cracks and associated filling materials cause the significant heterogeneity of natural rocks and rock masses. The induced heterogeneity changes the rock properties. This paper targets the gap in the existing literature regarding the adopting of artificial neural network approaches to efficiently and accurately predict the influences of heterogeneity on the strength of 3D-printed rocks at different strain rates. Herein, rock heterogeneity is reflected by different pre-existing crack and filling material configurations, quantitatively defined by the crack number, initial crack orientation with loading axis, crack tip distance, and crack offset distance. The artificial neural network model can be trained, validated, and tested by finite 42 quasi-static and 42 dynamic Brazilian disc experimental tests to establish the relationship between the rock strength and heterogeneous parameters at different strain rates. The artificial neural network architecture, including the hidden layer number and transfer functions, is optimized by the corresponding parametric study. Once trained, the proposed artificial neural network model generates an excellent prediction accuracy for influences of high dimensional heterogeneous parameters and strain rate on rock strength. The sensitivity analysis indicates that strain rate is the most important physical quantity affecting the strength of heterogeneous rock.

show abstract

“…Some of these studies on various concrete types are as follows. Yan et al [1] presumed the fracture parameters of the size effect model for concrete, Chen et al [2] predicted the concrete properties, Wang et al [3] estimated the expansion behavior of self-stressing concrete, Imam et al [4] modeled residual strength of corroded reinforced concrete Beams. Kostic and Vasovic [5] built a model for compressive strength of basic concrete.…”

Section: Introductionmentioning

confidence: 99%

A Neural Network Model for Ultrasonic Pulse Velocity and the Modulus of Elasticity of Polymer Concrete

2016

IJACEBS

View full text Add to dashboard Cite

Abstract-In this study, polymer concrete properties such as ultrasonic pulse velocity and the modulus of elasticity were modeled by artificial neural network. Seven polymers were selected as high density polyethylene, low density polyethylene, polypropylene, thermoplastic elastomer, dimethyl terephthalate, polyethylene terephthalate, polyethylene naphthalate and twenty-seven experiments was used to train the network. Two type of network training such as cascade forward back prop and feed forward back prop containing one hidden layer which has ten neurons was compared to predict the ultrasonic pulse velocity and the modulus of elasticity. Moreover, five experiments were used for comparing of the network models. The result show that feed forward back prop type training is more effective to predict the ultrasonic pulse velocity and the modulus of elasticity of polymer concrete comparing the cascade forward back prop training type.Keywords-A Neural Network Model, Polymer Concrete, Ultrasonic Pulse Velocity and the Modulus of Elasticity.

show abstract

Artificial neural network approach to predict the fracture parameters of the size effect model for concrete

Cited by 19 publications

References 32 publications

A simple damage model for concrete considering irreversible mode‐II microcracks

A simple damage model for concrete considering irreversible mode‐II microcracks

Using Artificial Neural Networks to Predict Influences of Heterogeneity on Rock Strength at Different Strain Rates

A Neural Network Model for Ultrasonic Pulse Velocity and the Modulus of Elasticity of Polymer Concrete

Contact Info

Product

Resources

About