In terms MBE, XGB achieved the minimum value of MBE, followed by ANN, SVR, and CNN. This algorithm attempts to determine the value of a new point by exploring a collection of training sets located nearby40. Mater. All three proposed ML algorithms demonstrate superior performance in predicting the correlation between the amount of fly-ash and the predicted CS of SFRC. So, more complex ML models such as KNN, SVR tree-based models, ANN, and CNN were proposed and implemented to study the CS of SFRC. Properties of steel fiber reinforced fly ash concrete. Google Scholar. The simplest and most commonly applied method of quality control for concrete pavements is to test compressive strength and then use this as an indirect measure of the flexural strength. Scientific Reports Adding hooked industrial steel fibers (ISF) to concrete boosts its tensile and flexural strength. 16, e01046 (2022). Mater. 12, the SP has a medium impact on the predicted CS of SFRC. Erdal, H. I. Two-level and hybrid ensembles of decision trees for high performance concrete compressive strength prediction. 232, 117266 (2020). For quality control purposes a reliable compressive strength to flexural strength conversion is required in order to ensure that the concrete satisfies the specification. Khademi, F., Akbari, M. & Jamal, S. M. Prediction of compressive strength of concrete by data-driven models. 118 (2021). Moreover, the CS of rubberized concrete was predicted using KNN algorithm by Hadzima-Nyarko et al.53, and it was reported that KNN might not be appropriate for estimating the CS of concrete containing waste rubber (RMSE=8.725, MAE=5.87). J. Comput. It is a measure of the maximum stress on the tension face of an unreinforced concrete beam or slab at the point of. Among different ML algorithms, convolutional neural network (CNN) with R2=0.928, RMSE=5.043, and MAE=3.833 shows higher accuracy. & Aluko, O. This is particularly common in the design and specification of concrete pavements where flexural strengths are critical while compressive strengths are often specified. PubMedGoogle Scholar. The raw data is also available from the corresponding author on reasonable request. . Han et al.11 reported that the length of the ISF (LISF) has an insignificant effect on the CS of SFRC. Privacy Policy | Terms of Use As is reported by Kang et al.18, among implemented tree-based models, XGB performed superiorly in predicting the CS of SFRC. Han, J., Zhao, M., Chen, J. A more useful correlations equation for the compressive and flexural strength of concrete is shown below. ADS : Investigation, Conceptualization, Methodology, Data Curation, Formal analysis, WritingOriginal Draft; N.R. & Xargay, H. An experimental study on the post-cracking behaviour of Hybrid Industrial/Recycled Steel Fibre-Reinforced Concrete. 23(1), 392399 (2009). 45(4), 609622 (2012). & Chen, X. Mater. This useful spreadsheet can be used to convert the results of the concrete cube test from compressive strength to . Anyone you share the following link with will be able to read this content: Sorry, a shareable link is not currently available for this article. Google Scholar. Table 3 shows the results of using a grid and a random search to tune the other hyperparameters. In the current study, the architecture used was made up of a one-dimensional convolutional layer, a one-dimensional maximum pooling layer, a one-dimensional average pooling layer, and a fully-connected layer. This property of concrete is commonly considered in structural design. The best-fitting line in SVR is a hyperplane with the greatest number of points. What factors affect the concrete strength? Kabiru, O. Moreover, in a study conducted by Awolusi et al.20 only 3 features (L/DISF as the fiber properties) were considered, and ANN and the genetic algorithm models were implemented to predict the CS of SFRC. Moreover, some others were omitted because of lacking the information of mixing components (such as FA, SP, etc.). J. Adhes. 147, 286295 (2017). Materials 13(5), 1072 (2020). Sanjeev, J. Infrastructure Research Institute | Infrastructure Research Institute Date:3/3/2023, Publication:Materials Journal This is much more difficult and less accurate than the equivalent concrete cube test, which is why it is common to test the compressive strength and then convert to flexural strength when checking the concrete's compliance with the specification. This method converts the compressive strength to the Mean Axial Tensile Strength, then converts this to flexural strength and includes an adjustment for the depth of the slab. To generate fiber-reinforced concrete (FRC), used fibers are typically short, discontinuous, and randomly dispersed throughout the concrete matrix8. Artif. The test jig used in this video has a scale on the receiver, and the distance between the external fulcrums (distance between the two outer fulcrums . RF consists of many parallel decision trees and calculates the average of fitted models on different subsets of the dataset to enhance the prediction accuracy6. Eng. Constr. Build. CNN model is a new architecture for DL which is comprised of several layers that process and transform an input to produce an output. The CS of SFRC was predicted through various ML techniques as is described in section "Implemented algorithms". Appl. Chen, H., Yang, J. The Offices 2 Building, One Central Mechanical and fracture properties of concrete reinforced with recycled and industrial steel fibers using Digital Image Correlation technique and X-ray micro computed tomography. Moreover, Nguyen-Sy et al.56 and Rathakrishnan et al.57, after implementing the XGB, noted that the XGB was the best model for predicting the CS of NC. In this paper, two factors of width-to-height ratio and span-to-height ratio are considered and 10 side-pressure laminated bamboo beams are prepared and tested for flexural capacity to study the flexural performance when they are used as structural members. In SVR, \(\{ x_{i} ,y_{i} \} ,i = 1,2,,k\) is the training set, where \(x_{i}\) and \(y_{i}\) are the input and output values, respectively. Eur. Kang et al.18 observed that KNN predicted the CS of SFRC with a great difference between actual and predicted values. Article Lee, S.-C., Oh, J.-H. & Cho, J.-Y. Intell. To develop this composite, sugarcane bagasse ash (SA), glass . Hence, the presented study aims to compare various ML algorithms for CS prediction of SFRC based on all the influential parameters. Ati, C. D. & Karahan, O. Normalization is a data preparation technique that converts the values in the dataset into a standard scale. Comput. Mater. 94, 290298 (2015). Mech. Al-Baghdadi, H. M., Al-Merib, F. H., Ibrahim, A. Note that for some low strength units the characteristic compressive strength of the masonry can be slightly higher than the unit strength. Skaryski, & Suchorzewski, J. The ideal ratio of 20% HS, 2% steel . The KNN method is a simple supervised ML technique that can be utilized in order to solve both classification and regression problems. Compressive strength of fly-ash-based geopolymer concrete by gene expression programming and random forest. Also, a specific type of cross-validation (CV) algorithm named LOOCV (Fig. 3) was used to validate the data and adjust the hyperparameters. In this regard, developing the data-driven models to predict the CS of SFRC is a comparatively novel approach. Use of this design tool implies acceptance of the terms of use. However, it is depicted that the weak correlation between the amount of ISF in the SFRC mix and the predicted CS. Some of the mixes were eliminated due to comprising recycled steel fibers or the other types of ISFs (such as smooth and wavy). Jamshidi Avanaki, M., Abedi, M., Hoseini, A. 49, 20812089 (2022). Adding hooked industrial steel fibers (ISF) to concrete boosts its tensile and flexural strength. J. Enterp. Article where \(x_{i} ,w_{ij} ,net_{j} ,\) and \(b\) are the input values, the weight of each signal, the weighted sum of the \(j{\text{th}}\) neuron, and bias, respectively18. Build. Eng. 6(4) (2009). The main focus of this study is the development of a sustainable geomaterial composite with higher strength capabilities (compressive and flexural). To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. & Gupta, R. Machine learning-based prediction for compressive and flexural strengths of steel fiber-reinforced concrete. Polymers 14(15), 3065 (2022). In many cases it is necessary to complete a compressive strength to flexural strength conversion. Mater. Finally, the model is created by assigning the new data points to the category with the most neighbors. Chou, J.-S., Tsai, C.-F., Pham, A.-D. & Lu, Y.-H. Machine learning in concrete strength simulations: Multi-nation data analytics. MathSciNet Constr. Buildings 11(4), 158 (2021). Ly, H.-B., Nguyen, T.-A. Struct. As with any general correlations this should be used with caution. Google Scholar. Mater. Hence, After each model training session, hold-out sample generalization may be poor, which reduces the R2 on the validation set 6. Zhu, H., Li, C., Gao, D., Yang, L. & Cheng, S. Study on mechanical properties and strength relation between cube and cylinder specimens of steel fiber reinforced concrete. Depending on the mix (especially the water-cement ratio) and time and quality of the curing, compressive strength of concrete can be obtained up to 14,000 psi or more. The alkali activated mortar based on the ultrafine particle of GPOFA produced a maximum compressive strength (57.5 MPa), flexural strength (10.9 MPa), porosity (13.1%), water absorption (6.2% . & Gao, L. Influence of tire-recycled steel fibers on strength and flexural behavior of reinforced concrete. Midwest, Feedback via Email Machine learning-based compressive strength modelling of concrete incorporating waste marble powder. Bending occurs due to development of tensile force on tension side of the structure. Khan et al.55 also reported that RF (R2=0.96, RMSE=3.1) showed more acceptable outcomes than XGB and GB with, an R2 of 0.9 and 0.95 in the prediction CS of SFRC, respectively. Commercial production of concrete with ordinary . Khan, M. A. et al. The linear relationship between compressive strength and flexural strength can be better expressed by the cubic curve model, and the correlation coefficient was 0.842. I Manag. Res. Li, Y. et al. Therefore, the data needs to be normalized to avoid the dominance effect caused by magnitude differences among input parameters34. Evidently, SFRC comprises a bigger number of components than NC including LISF, L/DISF, fiber type, diameter of ISF (DISF) and the tensile strength of ISFs. 103, 120 (2018). This web applet, based on various established correlation equations, allows you to quickly convert between compressive strength, flexural strength, split tensile strength, and modulus of elasticity of concrete. MLR predicts the value of the dependent variable (\(y\)) based on the value of the independent variable (\(x\)) by establishing the linear relationship between inputs (independent parameters) and output (dependent parameter) based on Eq. By submitting a comment you agree to abide by our Terms and Community Guidelines. Date:11/1/2022, Publication:Structural Journal Case Stud. Since the specified strength is flexural strength, a conversion factor must be used to obtain an approximate compressive strength in order to use the water-cement ratio vs. compressive strength table. Therefore, based on expert opinion and primary sensitivity analysis, two features (length and tensile strength of ISF) were omitted and only nine features were left for training the models. The experimental results show that in the case of [0/90/0] 2 ply, the bending strength of the structure increases by 2.79% in the forming embedding mode, while it decreases by 9.81% in the cutting embedding mode. Date:9/30/2022, Publication:Materials Journal Statistical characteristics of input parameters, including the minimum, maximum, average, and standard deviation (SD) values of each parameter, can be observed in Table 1. It uses two commonly used general correlations to convert concrete compressive and flexural strength. Figure10 also illustrates the normal distribution of the residual error of the suggested models for the prediction CS of SFRC. J. Terms of Use The user accepts ALL responsibility for decisions made as a result of the use of this design tool. Comparison of various machine learning algorithms used for compressive strength prediction of steel fiber-reinforced concrete, $$R_{XY} = \frac{{COV_{XY} }}{{\sigma_{X} \sigma_{Y} }}$$, $$x_{norm} = \frac{{x - x_{\min } }}{{x_{\max } - x_{\min } }}$$, $$\hat{y} = \alpha_{0} + \alpha_{1} x_{1} + \alpha_{2} x_{2} + \cdots + \alpha_{n} x_{n}$$, \(y = \left\langle {\alpha ,x} \right\rangle + \beta\), $$net_{j} = \sum\limits_{i = 1}^{n} {w_{ij} } x_{i} + b$$, https://doi.org/10.1038/s41598-023-30606-y. 230, 117021 (2020). Date:1/1/2023, Publication:Materials Journal Average 28-day flexural strength of at least 4.5 MPa (650 psi) Coarse aggregate: . INTRODUCTION The strength characteristic and economic advantages of fiber reinforced concrete far more appreciable compared to plain concrete. Investigation of mechanical characteristics and specimen size effect of steel fibers reinforced concrete. In addition, Fig. Al-Abdaly et al.50 reported that MLR algorithm (with R2=0.64, RMSE=8.68, MAE=5.66) performed poorly in predicting the CS behavior of SFRC. Constr. PMLR (2015). The loss surfaces of multilayer networks. The CivilWeb Flexural Strength of Concrete suite of spreadsheets includes the two methods described above, as well as the modulus of elasticity to flexural strength converter. To try out a fully functional free trail version of this software, please enter your email address below to sign up to our newsletter. 3-point bending strength test for fine ceramics that partially complies with JIS R1601 (2008) [Testing method for flexural strength of fine ceramics at room temperature] (corresponding part only). In contrast, the XGB and KNN had the most considerable fluctuation rate. 175, 562569 (2018). However, the understanding of ISFs influence on the compressive strength (CS) behavior of concrete is still questioned by the scientific society. MAPE is a scale-independent measure that is used to evaluate the accuracy of algorithms.
Federal Bureau Of Prisons Hiring Process,
Medtronic Pacemaker Monitor Flashing Lights,
Articles F