Next Article in Journal
Effect of Drilling Parameters on Machining Performance in Drilling Polytetrafluoroethylene
Next Article in Special Issue
The Experimental Investigation on Mechanics and Damage Characteristics of the Aeolian Sand Paste-like Backfill Materials Based on Acoustic Emission
Previous Article in Journal
Influence of Constant Magnetic Field upon Fatigue Life of Commercially Pure Titanium
Previous Article in Special Issue
A Decision-Making Algorithm for Concrete-Filled Steel Tubular Arch Bridge Maintenance Based on Structural Health Monitoring
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Data-Driven Techniques for Evaluating the Mechanical Strength and Raw Material Effects of Steel Fiber-Reinforced Concrete

by
Mohammed Najeeb Al-Hashem
1,
Muhammad Nasir Amin
1,*,
Waqas Ahmad
2,
Kaffayatullah Khan
1,
Ayaz Ahmad
3,
Saqib Ehsan
4,
Qasem M. S. Al-Ahmad
1 and
Muhammad Ghulam Qadir
5
1
Department of Civil and Environmental Engineering, College of Engineering, King Faisal University, Al-Ahsa 31982, Saudi Arabia
2
Department of Civil Engineering, COMSATS University Islamabad, Abbottabad 22060, Pakistan
3
MaREI Centre, Ryan Institute and School of Engineering, College of Science and Engineering, National University of Ireland Galway, H91 TK33 Galway, Ireland
4
Department of Civil Engineering, NFC Institute of Engineering and Fertilizer Research, Faisalabad 38090, Pakistan
5
Department of Environmental Sciences, Abbottabad Campus, COMSATS University Islamabad, Abbottabad 22060, Pakistan
*
Author to whom correspondence should be addressed.
Materials 2022, 15(19), 6928; https://doi.org/10.3390/ma15196928
Submission received: 14 September 2022 / Revised: 28 September 2022 / Accepted: 1 October 2022 / Published: 6 October 2022
(This article belongs to the Special Issue Testing of Materials and Elements in Civil Engineering (2nd Edition))

Abstract

:
Estimating concrete properties using soft computing techniques has been shown to be a time and cost-efficient method in the construction industry. Thus, for the prediction of steel fiber-reinforced concrete (SFRC) strength under compressive and flexural loads, the current research employed advanced and effective soft computing techniques. In the current study, a single machine learning method known as multiple-layer perceptron neural network (MLPNN) and ensembled machine learning models known as MLPNN-adaptive boosting and MLPNN-bagging are used for this purpose. Water; cement; fine aggregate (FA); coarse aggregate (CA); super-plasticizer (SP); silica fume; and steel fiber volume percent (Vf SF), length (mm), and diameter were the factors considered (mm). This study also employed statistical analysis such as determination coefficient (R2), root mean square error (RMSE), and mean absolute error (MAE) to assess the performance of the algorithms. It was determined that the MLPNN-AdaBoost method is suitable for forecasting SFRC compressive and flexural strengths. The MLPNN technique’s higher R2, i.e., 0.94 and 0.95 for flexural and compressive strength, respectively, and lower error values result in more precision than other methods with lower R2 values. SHAP analysis demonstrated that the volume of cement and steel fibers have the greatest feature values for SFRC’s compressive and flexural strengths, respectively.

1. Introduction

The simple production method for concrete and the abundant availability of its ingredients and several applications make it the most widely used construction material around the globe. The nature of concrete is conventionally brittle, having low strain capacity, toughness, and energy absorption capability. Accordingly, researchers are searching for ways to minimize the brittleness of typical concrete by enhancing its tensile strength. The dispersed incorporation of short-discrete fibers in conventional cementitious concrete is emerging as an effective method of enhancing concretes’ capacity for energy absorption [1,2,3,4]. Multiple researchers have explored the addition of steel/synthetic/natural fibers to concrete as reinforcement for improving characteristics like fatigue resistance, toughness, ductility, and resist propagation of cracks in concrete [5,6,7,8,9,10,11,12,13,14,15,16]. Steel fibers are incorporated into concrete to enhance its post-cracking phenomenon and toughness [17,18,19,20]. SFRC have multiple applications in different sections of the construction industry like building, pavements, rehabilitation, and repair. The enhanced mechanical properties of SFRC, as reported by some of the researchers, for different applications are summarized in Table 1.
Currently, the practice adopted for evaluating the mechanical properties of SFRC is the performance of the entire experimental program. A considerable amount of time and cost is involved in determining an accurate connection between properties of material and mix design through experimentation [26]. The variable SFRC parameters are the aggregates, cement, water, admixture/super-plasticizer, additive material and fiber (i.e., steel fibers) contents, and the admixture type. Despite the considerable experimental research in the literature, it is hard to forecast the characteristics of SFRC with different mixtures with the help of computational approaches. Hence, the current work is focused on estimating SFRC mechanical characteristics by employing a soft computational approach.
The employment of soft computational techniques may assist in resolving multiple complex problems in various fields of engineering [27,28,29]. ML techniques may be applied to forecast the ultimate outcome after incorporating a database as input parameters. Two ML approaches, a single model-based standalone method and ensemble Bagging and AdaBoost models, are employed in this research for the estimation of SFRC properties. Per the reported studies, ensemble modelling techniques are more effective than an individual model as shown in Table 2. Chaabene et al. [30] reported a detailed assessment of applying ML techniques for predicting the mechanical properties of concrete. Furthermore, multiple types of research have been conducted to estimate the mechanical properties of different concrete types like self-healing concrete [31], high-performance concrete (HPC) [32,33,34,35,36], phase change materials-integrated concrete [37], and recycled aggregate concrete (RAC) [38,39,40,41]. Han et al. [33] employed machine learning techniques for estimation of HPC compressive strength. The input parameters included age, water, cement, coarse aggregates, sand, fly-ash, and GGBFS, and five variable combinations were considered. The accurate compressive strength of HPC was obtained by the developed model. In this study, the SFRC compressive strength is predicted by applying soft computational approaches. The current research will provide a base for conserving cost and time of future researchers.
ML techniques have demonstrated possible prediction results with least difference in trials for various concrete types. For said purpose, experimental methods, including casting and testing procedures, consume considerable cost, effort, and time. Therefore, the current need is to develop data modeling-based algorithms in line with closely linked in-dependent parameter identification and the instant decrement in input matrix dimensions. The application of soft computational approaches is gaining more importance for predicting concrete materials’ behavior in civil engineering. ML is an alternative technique for predicting SFRC’s compressive and flexural strengths to conserve experimental cost and time. The current study involves applying an individual ML model and multiple ensembled ML techniques to predict SFRC compressive and flexural strengths. In addition, the influence of raw materials on mechanical properties remains largely unexplored in contemporary study and is still very limited. The integration of SHapley Additive exPlanations (SHAP) with ML algorithms is also performed in this paper, addressing a current research need. SHAP analysis is intended to provide in-depth knowledge of SFRC mix design in terms of its mechanical strength factors via complicated nonlinear behavior and the description of input factors’ contributions by assigning a weighting factor to each input component. MLPNN is taken as an individual ML model, while MLPNN-AdaBoost and MLPNN-Bagging are taken as ensembled machine learning algorithms. In addition, statistical analysis is performed for the evaluation of all the considered models, and all said ML algorithms are compared as well. Afterwards, based on numerous statistical factors’ performance, a superior model is proposed for predicting SFRC properties. As a whole, a correlation for valuable structure properties is established in this research by applying interpretable machine learning techniques through feature importance.

2. Soft Computing Techniques

2.1. Multilayer Perceptron Neural Network (MLPNN)

The ANN model is among the most effective ML models. Its potential to resolve nonlinear issues has made it widely applied in hydrological and environmental engineering areas. Among multiple ANN models, the multilayer perceptron ANN (MLPNN) is the one that is used most frequently. The MLPNN model’s architecture comprises three layers: an input one followed by one or more hidden ones, and the output one. The three conventional functions of activation are; logsig, tansig, and purelin. Activations, weights, and bias functions are among the most important parts in both the output and hidden layers. The training of the model governs the parameters or weights of the model. The hidden layers employ the function of tansig activation; however, purelin is used for the output layer. The best structure is extracted by fivefold cross-validation. The top ANN model came out with three layers that are hidden (i.e., 9, 3, and 2) having optimal numbers of neurons against every hidden layer [48]. A typical/conventional neural network is shown in Figure 1. The composition of these networks is at three stages in a way that the input is transmitted by forward-pass, weight is multiplied by it, and the prediction of model output is made. The predicted results are then compared with the considered inputs. The input factors are considered for the model prediction outcomes. Based upon the objectives and performance, various loss functions are employed. The partial derivatives for cost function, linked with individual factors back in operation, are generated by backward propagation. Gradient descent is used for back loss propagation and model weight updating during this method.

2.2. MLPNN Bagging and MLPNN AdaBoost Techniques

The accuracy of prediction and recognition of ML can be improved by using ensemble approaches. These approaches usually help resolve problems by aggregating and integrating various algorithms having weaker predictions. A smart learner can be made by intelligently developing different sub-models (i.e., A, B, … N) along with the alteration of data for training and the merging of average and votes of combination measures, to obtain the correct result of projecting sub-models, for making an ideal model. The most frequently adopted ensembled modelling approach is Bagging, which implies the resampling bootstrap technique for calculating benefits and gathering the data. During this method, the first set of training with fractional algorithms is substituted for the actual algorithm. Some samples of data may seem in different algorithms, whereas few of them do not even appear in any model product. The average from all component models’ output is taken to calculate the final outcome model.
The Bagging approach, like boosting, creates a cumulative model that constructs multiple more precise components compared with non-ensemble models. Furthermore, the weighted averages are used in the Boosting process based on sub-models to determine their inclusion in the final model. Based upon MLPNN-like individual learners, the current work predicts the flexural and compressive strengths of SFRC using Bagging and Adaptive Boosting (AdaBoost) approaches.

3. Dataset

The literature-based dataset that was used comprises 150 mix ratios with 9 contribution parameters [50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66]. All the samples were water cured at 28 days. Figure 2 and Figure 3 exhibit the dataset that was employed to estimate SFRC strengths. These include water (kg/m3), cement (kg/m3), coarse aggregate (CA) (kg/m3), fine aggregate (FA) (kg/m3), superplasticizer (SP) (%), silica fume (%), volume fraction of steel fiber (Vf SF) (%), SF Length (mm), and SF diameter (mm). The prediction variables for output parameters (i.e., compressive and flexural strength) relied on the above-mentioned input parameters. Python scripting in Spyder Anaconda software was used to predict compressive and flexural strength.

4. Results and Discussion

4.1. Multiple-Layer Perceptron Neural Network (MLPNN)

The comparison of the MLPNN algorithm projected and experimental values for SFRC compressive strength are shown in Figure 4. MLPNN shows a reasonably estimated outcome with minimal variation in SFRC compressive strength. The appropriateness of the MLPNN model is represented by an acceptable R2 of 0.79. The error distribution of MLPNN predicted and experimental values for SFRC compressive strength are illustrated in Figure 5. The average error values for SFRC compressive strength are 8.69 MPa: 46% of the error values are below 5 MPa, 29% of these values range between 5 and 10 MPa, and 25% are more than 10 MPa.
The MLPNN projected and investigational results for SFRC flexural strength are presented in Figure 6. The R2 of 0.81 reveals the less appropriate outcome. Similarly, the projected results for the flexural strength of SFRC with the help of MLPNN lie in suitable array. The distribution of error for MLPNN projected, and investigational flexural strength of SFRC is shown in Figure 7. Nearly one third of values, 29%, are below 1 MPa, 66% are in the range of 1 to 5 MPa, and the remaining 5% are above 5 MPa.

4.2. MLPNN-AdaBoost

Figure 8 shows the predicted MLPNN-AdaBoost algorithm and investigational results for compressive strength of SFRC. The R2 of 0.95 for MLPNN-AdaBoost depicts a higher accuracy of outcomes than that of the MLPNN algorithm. Figure 9 represents the distribution of error for MLPNN-AdaBoost estimated and investigational results for the compressive strength of SFRC. It may be seen that 62% of values are below 5 MPa, 29% of these values range from 5 and 10 MPa, and 9% of values are above 10 MPa. The higher R2 and lower error values show more precision of the MLPNN-AdaBoost model than MLPNN.
The MLPNN-AdaBoost model’s estimated and investigational results were compared for SFRC flexural strength (Figure 10). MLPNN-AdaBoost depicts reduced variation in error for SFRC flexural strength and highly precise predicted results. The adequacy of the MLPNN-AdaBoost model is represented by an acceptable R2 of 0.94. The error distribution of MLPNN-AdaBoost predicted and experimental SFRC flexural strength is illustrated in Figure 11. The average error value for SFRC flexural strength is 1.57 MPa: 47% of total error values are below 1 MPa, 53% of these values are between 1 and 5 MPa, and no value is more than 5 MPa.

4.3. MLPNN-Bagging

Figure 12 shows the projected and investigational outputs in case of MLPNN-Bagging for SFRC compressive strength. The R2 of 0.89 for this model depicts comparatively less appropriate results than the above-mentioned ensembled MLPNN-AdaBoost model. The estimated SFRC compressive strength outcomes for MLPNN-Bagging are superior to the individual MLPNN model. Figure 13 depicts the distribution of error for MLPNN-Bagging projected and investigational results for SFRC compressive strength: 49% of values are below 5 MPa, 28% are from 5 to 10 MPa, and the remaining 22% of these values are higher than 10 MPa. The R2 and error values for SFRC compressive strength in the case of MLPNN are more precise than the MLPNN-Bagging model. At the same time, the MLPNN ensembled machine learning algorithms error and R2 values are satisfactory. Therefore, this result shows the higher accuracy of estimation outcomes of MLPNN compared to other considered models.
The estimated MLPNN-AdaBoost and investigational results for SFRC flexural strength are presented in Figure 14. The R2 of 0.92 for MLPNN-AdaBoost displays less accurate outcomes compared with MLPNN-AdaBoost. The distribution of error for the MLPNN-AdaBoost estimated and investigational results for SFRC flexural strength are presented in Figure 15. It is assessed that 33% of values are below 1 MPa, 62% lie in the 1 to 5 MPa range, and 4% are above 5 MPa. The higher R2 and lower error values demonstrate the higher accuracy of MLPNN-AdaBoost compared with MLPNN. In contrast, the attained R2 and error values for MLPNN-Bagging ensembled machine learning algorithms are suitable: this result showed the most accuracy for estimation outputs of MLPNN-AdaBoost compared with the other considered algorithms.

4.4. Comparison of All Models

The k-fold technique was adopted for cross-validation in order to assess the performance of model while implementation. The performance of a model is determined by employment of statistical checks [67,68,69,70]. Normally, in the said k-fold process, there is splitting data in 10 clusters for random spreading by repeating this process 10 times for attaining suitable results. Table 3 provides the employed statistical checks. The compressive strength R2 values for MLPNN, MLPNN-Bagging, and MLPNN-AdaBoost models were 0.79, 0.89, and 0.95, respectively, as presented in Figure 16a–c. In the case of flexural strength, the R2 values for MLPNN, MLPNN-Bagging, and MLPNN-AdaBoost model were 0.81, 0.92, and 0.94, respectively, as presented in Figure 17a–c. It is observed that the R2 for MLPNN-AdaBoost is higher than those of the other considered algorithms, having lower error values for the SFRC compressive and flexural strengths.
The comparison of current models with the models in the literature is shown in Table 3. SFRC compressive strength is estimated by applying ensembled ML techniques in the current study, which intends to offer reliable and efficient results as compared to the other studies in the literature. The R2 of 0.95 for MLPNN-AdaBoost outcomes provides a more precise estimation of SFRC compressive strength. The ensembled MLPNN-AdaBoost ML models perform better in predicting compressive strength by utilizing an optimized model extracted from 20 sub-models, as presented in Figure 18a,b. It can be observed that ensembled MLPNN-AdaBoost models depict more accuracy and lower error than other models as well as the models reported in the literature. Despite this, SFRC flexural strength is estimated by applying ensembled ML techniques in the current study, which intends to offer reliable and efficient outcomes. The R2 of 0.94 in the case of MLPNN-AdaBoost results provides a more precise estimation for the compressive strength of SFRC. Out of 20 sub-models, an optimized model is used to estimate SFRC flexural strength in the case of ensembled MLPNN-AdaBoost ML models that perform better (Figure 19a,b). In comparison with other models, the ensembled MLPNN-AdaBoost models show higher accuracy and lower error.

5. Feature Importance of ML Models for Compressive and Flexural Strength

A thorough explanation is also given in the current research of the ML algorithm and interactions of considered input features. Different feature importance correlation for compressive strength of SFRC is shown in Figure 20. It can be observed that the feature value of cement feature is highest, i.e., 0.46, for SFRC compressive strength estimation. The cement feature has a positive influence, which means that by enhancing the cement content, the SFRC compressive strength increases. The SHAP plot (Figure 21) also shows that cement has the highest impact on SFRC compressive strength. The water feature has the second highest feature value of 0.26 for SFRC compressive strength; however, it influences negatively. Increasing the amount of water will reduce the compressive strength (Figure 21). Thirdly, the main factor for SFRC is silica fume, and this feature has approximately 0.1 feature value (Figure 20 and Figure 21). Further, the content of silica fume as a feature is positively influencing the SFRC compressive strength. It means that enhancement in its content turns into more compressive strength of SFRC. Coarse aggregate is next in line with a feature value of almost 0.8, but, in this case, the increase in coarse aggregate content up to optimum content only results in enhanced compressive strength. Beyond this optimum content of coarse aggregates, the SFRC compressive strength decreases. This behavior shows coarse aggregates’ positive and negative influence on SFRC compressive strength. Similarly, the feature value for sand, super-plasticizer is next, followed by steel fiber length, volume and diameter. All these considered features have more or less the same feature values near zero, showing their minimal impact on compressive strength of SFRC.
Similarly, Figure 22 and Figure 23 present the features’ importance correlations and features and SHAP plot for SFRC flexural strength. In this scenario, it is indicated in Figure 22 that the volume content of steel fiber has the highest feature value of 0.24 for flexural strength prediction of SFRC. It may also be observed from Figure 23 that the enhancing content of steel fiber volumes is increasing the SFRC flexural strength and vice versa. The second highest feature value of 0.22 is for coarse aggregates feature in the case of SFRC flexural strength. At the third level, the water has a feature value of 0.18, but with a negative influence, which means the enhancement in water content causes a reduction in flexural strength (Figure 23). Figure 22 depicts that the cement feature has a feature value up to 0.14 and positively influences the flexural strength of SFRC. The higher the cement content, the more the SFRC flexural strength (Figure 23). Afterwards, the silica fume, an important feature of SFRC, has a feature value of 0.11 for SFRC flexural strength, which is almost the same as for compressive strength of SFRC. The enhancing silica fume results in more SFRC flexural strength (Figure 23). Subsequently, the fine aggregates feature has a feature value of almost 0.07, followed by the feature values of super-plasticizer, steel fiber diameter, and length. The same feature values for all these features are nearly zero, depicting their lesser influence on SFRC flexural strength. The database used in the current study is the base of this prediction, and highly accurate results can be achieved with added data points.
This study assessed the compressive and flexural strength of 150 mixture proportions using 9 input factors with satisfactory performance. A substantially more accurate model might be generated by increasing the number of datasheets, importing a significantly larger number of mixes, and taking into account a greater number of input parameters. In order to increase the quantity of data points and outcomes in future research, it is recommended that experimental work, field testing, and numerical analysis utilizing a variety of techniques be employed (e.g., Monte Carlo simulation, among others). To improve the models’ performance, environmental conditions (such as high temperatures and humidity) might be incorporated in the input parameters along with a full explanation of the raw materials. The detailed limitations of machine learning models to estimate the strength properties of concrete is already reported in the literature [76].

6. Conclusions

The main aim of the current study is to determine the precision level of soft computational techniques to estimate SFRC compressive and flexural strengths. The considered input parameters for said prediction are cement, water, fine aggregate (FA), coarse aggregate (CA), super-plasticizer (SP), silica fume, the volume fraction of steel fiber (Vf SF), SF length (mm), and SF diameter (mm). The conclusions are as follows:
  • As demonstrated by the R2 of 0.95, the MLPNN-AdaBoost technique may be applied for precise estimation of SFRC compressive strength from its actual dataset. In contrast, individual ML MLPNN and ensembled MLPNN-Bagging ML models have R2 values of 0.79 and 0.89, respectively, providing satisfactory results for SFRC compressive strength.
  • The predicted compressive strength of SFRC is optimized by employing 20 sub-models from 10 to 200 estimators. SFRC compressive strength is more effectively predicted by an ensembled model MLPNN than other models. K-fold validation outcomes show that MLPNN models have lower MAE and RMSE with higher R2 for SFRC compressive strength than other considered models. The model for having the best prediction for SFRC compressive strength is MLPNN.
  • Statistical checks like RMSE and MAE are used to evaluate the model’s performance. However, the superiority of MLPNN is demonstrated by its having a higher determination coefficient and fewer error values for SFRC compressive strength. The MLPNN is the most effective soft computational technique for predicting SFRC compressive strength.
  • The cement content has the highest influence on compressive strength prediction of SFRC, followed by the contents of water, silica fume, and coarse aggregates, as revealed from SHAP analysis. The diameter of steel fibers has the least influence on SFRC compressive strength. The SHAP plot shows that the cement and silica fume content positively influence the compressive strength of SFRC.
  • SFRC flexural strength is accurately predicted from its actual data by the MLPNN-AdaBoost technique as evident from the R2 of 0.94. However, the R2 of 0.81 and 0.92 in the case of individual MLPNN and ensembled MLPNN-Bagging ML models, respectively, estimated suitable results for SFRC flexural strength.
  • The predicted flexural strength of SFRC is augmented by employing 20 sub models from 10 to 200 estimators. The more precise estimation of SFRC flexural strength is come out in case of an ensembled MLPNN model compared to other models. After applying the k-fold checks, the MLPNN algorithms are come out with higher R2 values and lower RMSE and MAE values for SFRC flexural strength than other models.
  • MLPNN is come out with the best prediction for SFRC flexural strength. RMSE and MAE statistical checks are applied to evaluate the performance of the model. Similarly, the higher determination coefficient with lower values of error show the superiority of MLPNN for the prediction of SFRC flexural strength. Among soft computational techniques, MLPNN emerged as the most effective technique for the estimation of SFRC flexural strength.
  • It is revealed from SHAP analysis that the volume of steel fiber significantly influenced the predicted SFRC flexural strength, followed by contents of coarse aggregates, water, cement, and silica fume. However, the SFRC flexural strength is least influenced by steel fiber length. The SHAP plot shows that the steel fiber volume positively influences the flexural strength of SFRC.

Author Contributions

M.N.A.-H.: Funding Acquisition, Resources, Visualization, Writing—reviewing and editing. M.N.A.: Conceptualization, Funding acquisition, Resources, Project administration, Supervision, Writing-Reviewing and Editing. W.A.: Conceptualization, Data curation, Software, Methodology, Investigation, Validation, Writing—original draft. K.K.: Methodology, Investigation, Writing—Reviewing and Editing. A.A.: Resources, Visualization, Writing-Reviewing and Editing. S.E.: Visualization, Validation. Q.M.S.A.-A.: Visualization, Resources. M.G.Q.: Funding acquisition, Writing—Reviewing and Editing. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the Deanship of Scientific Research, Vice Presidency for Graduate Studies and Scientific Research, King Faisal University, Saudi Arabia (Project No. GRANT1406), through its KFU Research Summer Initiative. The APC was also funded by Project No. GRANT1406.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The data used in this research have been properly cited and reported in the main text.

Acknowledgments

The authors acknowledge the Deanship of Scientific Research, Vice Presidency for Graduate Studies and Scientific Research, King Faisal University, Saudi Arabia (Project No. GRANT1406), through its KFU Research Summer Initiative. The authors extend their appreciation for the financial support that made this study possible.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Naaman, A.E. High Performance Fiber Reinforced Cement Composites. In High-Performance Construction Materials: Science and Applications; Shi, C., Mo, Y.L., Eds.; World Scientific Publishing: Singapore, 2008; pp. 91–153. [Google Scholar]
  2. Imam, M.; Vandewalle, L.; Mortelmans, F.; Van Gemert, D. Shear domain of fibre-reinforced high-strength concrete beams. Eng. Struct. 1997, 19, 738–747. [Google Scholar] [CrossRef]
  3. Furlan, S., Jr.; de Hanai, J.B. Shear behaviour of fiber reinforced concrete beams. Cem. Concr. Compos. 1997, 19, 359–366. [Google Scholar] [CrossRef]
  4. Kene, K.S.; Vairagade, V.S.; Sathawane, S. Experimental study on behavior of steel and glass fiber reinforced concrete composites. Bonfring Int. J. Ind. Eng. Manag. Sci. 2012, 2, 125–130. [Google Scholar] [CrossRef] [Green Version]
  5. Cao, M.; Mao, Y.; Khan, M.; Si, W.; Shen, S. Different testing methods for assessing the synthetic fiber distribution in cement-based composites. Constr. Build. Mater. 2018, 184, 128–142. [Google Scholar] [CrossRef]
  6. Khan, M.; Cao, M.; Hussain, A.; Chu, S. Effect of silica-fume content on performance of CaCO3 whisker and basalt fiber at matrix interface in cement-based composites. Constr. Build. Mater. 2021, 300, 124046. [Google Scholar] [CrossRef]
  7. Arshad, S.; Sharif, M.B.; Irfan-ul-Hassan, M.; Khan, M.; Zhang, J.-L. Efficiency of supplementary cementitious materials and natural fiber on mechanical performance of concrete. Arab. J. Sci. Eng. 2020, 45, 8577–8589. [Google Scholar] [CrossRef]
  8. Xie, C.; Cao, M.; Guan, J.; Liu, Z.; Khan, M. Improvement of boundary effect model in multi-scale hybrid fibers reinforced cementitious composite and prediction of its structural failure behavior. Compos. Part B Eng. 2021, 224, 109219. [Google Scholar] [CrossRef]
  9. Cao, M.; Khan, M. Effectiveness of multiscale hybrid fiber reinforced cementitious composites under single degree of freedom hydraulic shaking table. Struct. Concr. 2021, 22, 535–549. [Google Scholar] [CrossRef]
  10. Khan, U.A.; Jahanzaib, H.M.; Khan, M.; Ali, M. Improving the tensile energy absorption of high strength natural fiber reinforced concrete with fly-ash for bridge girders. Key Eng. Mater. 2018, 765, 335–342. [Google Scholar] [CrossRef]
  11. Khan, M.; Cao, M.; Ai, H.; Hussain, A. Basalt Fibers in Modified Whisker Reinforced Cementitious Composites. Period. Polytech. Civ. Eng. 2022, 66, 344–354. [Google Scholar] [CrossRef]
  12. Zhang, N.; Yan, C.; Li, L.; Khan, M. Assessment of fiber factor for the fracture toughness of polyethylene fiber reinforced geopolymer. Constr. Build. Mater. 2022, 319, 126130. [Google Scholar] [CrossRef]
  13. Khan, M.; Ali, M. Improvement in concrete behavior with fly ash, silica-fume and coconut fibres. Constr. Build. Mater. 2019, 203, 174–187. [Google Scholar] [CrossRef]
  14. Ramakrishnan, V.; Wu, G.Y.; Hosalli, G. Flexural fatigue strength, endurance limit and impact strength of fiber reinforced concretes. Transp. Res. Rec. 1989, 1226, 17–24. [Google Scholar]
  15. Gupta, S.; Rao, V.K.; Sengupta, J. Evaluation of polyester fiber reinforced concrete for use in cement concrete pavement works. Road Mater. Pavement Des. 2008, 9, 441–461. [Google Scholar] [CrossRef]
  16. Sinha, D.; Mishra, C.; Solanki, R. Comparison of normal concrete pavement with steel fiber reinforced concrete pavement. Indian J. Appl. Res. 2014, 4, 233–235. [Google Scholar] [CrossRef]
  17. Khan, M.; Cao, M.; Chu, S.; Ali, M. Properties of hybrid steel-basalt fiber reinforced concrete exposed to different surrounding conditions. Constr. Build. Mater. 2022, 322, 126340. [Google Scholar] [CrossRef]
  18. Li, L.; Khan, M.; Bai, C.; Shi, K. Uniaxial tensile behavior, flexural properties, empirical calculation and microstructure of multi-scale fiber reinforced cement-based material at elevated temperature. Materials 2021, 14, 1827. [Google Scholar] [CrossRef]
  19. Khan, M.; Cao, M.; Xie, C.; Ali, M. Hybrid fiber concrete with different basalt fiber length and content. Struct. Concr. 2022, 23, 346–364. [Google Scholar] [CrossRef]
  20. Khan, M.; Cao, M.; Xie, C.; Ali, M. Effectiveness of hybrid steel-basalt fiber reinforced concrete under compression. Case Stud. Constr. Mater. 2022, 16, e00941. [Google Scholar] [CrossRef]
  21. Purkiss, J. Steel fibre reinforced concrete at elevated temperatures. Int. J. Cem. Compos. Lightweight Concr. 1984, 6, 179–184. [Google Scholar] [CrossRef]
  22. Patil, S.P.; Sangle, K.K. Tests of steel fibre reinforced concrete beams under predominant torsion. J. Build. Eng. 2016, 6, 157–162. [Google Scholar] [CrossRef]
  23. Noaman, A.T.; Bakar, B.A.; Akil, H.M.; Alani, A. Fracture characteristics of plain and steel fibre reinforced rubberized concrete. Constr. Build. Mater. 2017, 152, 414–423. [Google Scholar] [CrossRef]
  24. Boulekbache, B.; Hamrat, M.; Chemrouk, M.; Amziane, S. Flexural behaviour of steel fibre-reinforced concrete under cyclic loading. Constr. Build. Mater. 2016, 126, 253–262. [Google Scholar] [CrossRef]
  25. Gholamhoseini, A.; Khanlou, A.; MacRae, G.; Scott, A.; Hicks, S.; Leon, R. An experimental study on strength and serviceability of reinforced and steel fibre reinforced concrete (SFRC) continuous composite slabs. Eng. Struct. 2016, 114, 171–180. [Google Scholar] [CrossRef]
  26. Xu, M.; Bao, Y.; Wu, K.; Xia, T.; Clack, H.L.; Shi, H.; Li, V.C. Influence of TiO2 incorporation methods on NOx abatement in Engineered Cementitious Composites. Constr. Build. Mater. 2019, 221, 375–383. [Google Scholar] [CrossRef]
  27. Shang, M.; Li, H.; Ahmad, A.; Ahmad, W.; Ostrowski, K.A.; Aslam, F.; Joyklad, P.; Majka, T.M. Predicting the Mechanical Properties of RCA-Based Concrete Using Supervised Machine Learning Algorithms. Materials 2022, 15, 647. [Google Scholar] [CrossRef]
  28. Ahmad, A.; Ahmad, W.; Chaiyasarn, K.; Ostrowski, K.A.; Aslam, F.; Zajdel, P.; Joyklad, P. Prediction of Geopolymer Concrete Compressive Strength Using Novel Machine Learning Algorithms. Polymers 2021, 13, 3389. [Google Scholar] [CrossRef]
  29. Zheng, D.; Wu, R.; Sufian, M.; Kahla, N.B.; Atig, M.; Deifalla, A.F.; Accouche, O.; Azab, M. Flexural Strength Prediction of Steel Fiber-Reinforced Concrete Using Artificial Intelligence. Materials 2022, 15, 5194. [Google Scholar] [CrossRef]
  30. Chaabene, W.B.; Flah, M.; Nehdi, M.L. Machine learning prediction of mechanical properties of concrete: Critical review. Constr. Build. Mater. 2020, 260, 119889. [Google Scholar] [CrossRef]
  31. Ramadan Suleiman, A.; Nehdi, M.L. Modeling self-healing of concrete using hybrid genetic algorithm–artificial neural network. Materials 2017, 10, 135. [Google Scholar] [CrossRef]
  32. Xu, Y.; Ahmad, W.; Ahmad, A.; Ostrowski, K.A.; Dudek, M.; Aslam, F.; Joyklad, P. Computation of High-Performance Concrete Compressive Strength Using Standalone and Ensembled Machine Learning Techniques. Materials 2021, 14, 7034. [Google Scholar] [CrossRef]
  33. Han, Q.; Gui, C.; Xu, J.; Lacidogna, G. A generalized method to predict the compressive strength of high-performance concrete by improved random forest algorithm. Constr. Build. Mater. 2019, 226, 734–742. [Google Scholar] [CrossRef]
  34. Al-Shamiri, A.K.; Yuan, T.-F.; Kim, J.H. Non-tuned machine learning approach for predicting the compressive strength of high-performance concrete. Materials 2020, 13, 1023. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  35. Dingqiang, F.; Rui, Y.; Zhonghe, S.; Chunfeng, W.; Jinnan, W.; Qiqi, S. A novel approach for developing a green Ultra-High Performance Concrete (UHPC) with advanced particles packing meso-structure. Constr. Build. Mater. 2020, 265, 120339. [Google Scholar] [CrossRef]
  36. Fan, D.; Yu, R.; Shui, Z.; Wu, C.; Song, Q.; Liu, Z.; Sun, Y.; Gao, X.; He, Y. A new design approach of steel fibre reinforced ultra-high performance concrete composites: Experiments and modeling. Cem. Concr. Compos. 2020, 110, 103597. [Google Scholar] [CrossRef]
  37. Marani, A.; Nehdi, M.L. Machine learning prediction of compressive strength for phase change materials integrated cementitious composites. Constr. Build. Mater. 2020, 265, 120286. [Google Scholar] [CrossRef]
  38. Deng, F.; He, Y.; Zhou, S.; Yu, Y.; Cheng, H.; Wu, X. Compressive strength prediction of recycled concrete based on deep learning. Constr. Build. Mater. 2018, 175, 562–569. [Google Scholar] [CrossRef]
  39. Zhang, J.; Huang, Y.; Aslani, F.; Ma, G.; Nener, B. A hybrid intelligent system for designing optimal proportions of recycled aggregate concrete. J. Clean. Prod. 2020, 273, 122922. [Google Scholar] [CrossRef]
  40. Han, T.; Siddique, A.; Khayat, K.; Huang, J.; Kumar, A. An ensemble machine learning approach for prediction and optimization of modulus of elasticity of recycled aggregate concrete. Constr. Build. Mater. 2020, 244, 118271. [Google Scholar] [CrossRef]
  41. Behnood, A.; Golafshani, E.M. Machine learning study of the mechanical properties of concretes containing waste foundry sand. Constr. Build. Mater. 2020, 243, 118152. [Google Scholar] [CrossRef]
  42. Javed, M.F.; Farooq, F.; Memon, S.A.; Akbar, A.; Khan, M.A.; Aslam, F.; Alyousef, R.; Alabduljabbar, H.; Rehman, S.K. New Prediction Model for the Ultimate Axial Capacity of Concrete-Filled Steel Tubes: An Evolutionary Approach. Crystals 2020, 10, 741. [Google Scholar] [CrossRef]
  43. Zhu, Y.; Ahmad, A.; Ahmad, W.; Vatin, N.I.; Mohamed, A.M.; Fathi, D. Predicting the Splitting Tensile Strength of Recycled Aggregate Concrete Using Individual and Ensemble Machine Learning Approaches. Crystals 2022, 12, 569. [Google Scholar] [CrossRef]
  44. Iftikhar, B.; Alih, S.C.; Vafaei, M.; Elkotb, M.A.; Shutaywi, M.; Javed, M.F.; Deebani, W.; Khan, M.I.; Aslam, F. Predictive modeling of compressive strength of sustainable rice husk ash concrete: Ensemble learner optimization and comparison. J. Clean. Prod. 2022, 348, 131285. [Google Scholar] [CrossRef]
  45. Ahmad, A.; Ahmad, W.; Aslam, F.; Joyklad, P. Compressive strength prediction of fly ash-based geopolymer concrete via advanced machine learning techniques. Case Stud. Constr. Mater. 2022, 16, e00840. [Google Scholar] [CrossRef]
  46. Song, H.; Ahmad, A.; Farooq, F.; Ostrowski, K.A.; Maślak, M.; Czarnecki, S.; Aslam, F. Predicting the compressive strength of concrete with fly ash admixture using machine learning algorithms. Constr. Build. Mater. 2021, 308, 125021. [Google Scholar] [CrossRef]
  47. Ahmad, A.; Farooq, F.; Niewiadomski, P.; Ostrowski, K.; Akbar, A.; Aslam, F.; Alyousef, R. Prediction of compressive strength of fly ash based concrete using individual and ensemble algorithm. Materials 2021, 14, 794. [Google Scholar] [CrossRef] [PubMed]
  48. Hadzima-Nyarko, M.; Nyarko, E.K.; Lu, H.; Zhu, S. Machine learning approaches for estimation of compressive strength of concrete. Eur. Phys. J. Plus 2020, 135, 682. [Google Scholar] [CrossRef]
  49. Nafees, A.; Khan, S.; Javed, M.F.; Alrowais, R.; Mohamed, A.M.; Mohamed, A.; Vatin, N.I. Forecasting the Mechanical Properties of Plastic Concrete Employing Experimental Data Using Machine Learning Algorithms: DT, MLPNN, SVM, and RF. Polymers 2022, 14, 1583. [Google Scholar] [CrossRef]
  50. Soulioti, D.; Barkoula, N.; Paipetis, A.; Matikas, T. Effects of fibre geometry and volume fraction on the flexural behaviour of steel-fibre reinforced concrete. Strain 2011, 47, e535–e541. [Google Scholar] [CrossRef]
  51. Yoo, D.-Y.; Yoon, Y.-S.; Banthia, N. Flexural response of steel-fiber-reinforced concrete beams: Effects of strength, fiber content, and strain-rate. Cem. Concr. Compos. 2015, 64, 84–92. [Google Scholar] [CrossRef]
  52. Lee, J.-H.; Cho, B.; Choi, E. Flexural capacity of fiber reinforced concrete with a consideration of concrete strength and fiber content. Constr. Build. Mater. 2017, 138, 222–231. [Google Scholar] [CrossRef]
  53. Köksal, F.; Altun, F.; Yiğit, İ.; Şahin, Y. Combined effect of silica fume and steel fiber on the mechanical properties of high strength concretes. Constr. Build. Mater. 2008, 22, 1874–1880. [Google Scholar] [CrossRef]
  54. Yoon, E.-S.; Park, S.-B. An experimental study on the mechanical properties and long-term deformations of high-strength steel fiber reinforced concrete. KSCE J. Civ. Environ. Eng. Res. 2006, 26, 401–409. [Google Scholar]
  55. Abbass, W.; Khan, M.I.; Mourad, S. Evaluation of mechanical properties of steel fiber reinforced concrete with different strengths of concrete. Constr. Build. Mater. 2018, 168, 556–569. [Google Scholar] [CrossRef]
  56. Yoo, D.-Y.; Yoon, Y.-S.; Banthia, N. Predicting the post-cracking behavior of normal-and high-strength steel-fiber-reinforced concrete beams. Constr. Build. Mater. 2015, 93, 477–485. [Google Scholar] [CrossRef]
  57. Lee, H.-H.; Lee, H.-J. Characteristic strength and deformation of SFRC considering steel fiber factor and volume fraction. J. Korea Concr. Inst. 2004, 16, 759–766. [Google Scholar] [CrossRef]
  58. Oh, Y.-H. Evaluation of flexural strength for normal and high strength concrete with hooked steel fibers. J. Korea Concr. Inst. 2008, 20, 531–539. [Google Scholar]
  59. Song, P.; Hwang, S. Mechanical properties of high-strength steel fiber-reinforced concrete. Constr. Build. Mater. 2004, 18, 669–673. [Google Scholar] [CrossRef]
  60. Jang, S.-J.; Yun, H.-D. Combined effects of steel fiber and coarse aggregate size on the compressive and flexural toughness of high-strength concrete. Compos. Struct. 2018, 185, 203–211. [Google Scholar] [CrossRef]
  61. Aldossari, K.; Elsaigh, W.; Shannag, M. Effect of steel fibers on flexural behavior of normal and high strength concrete. Int. J. Civ. Environ. Eng. 2014, 8, 22–26. [Google Scholar]
  62. Dinh, N.H.; Park, S.-H.; Choi, K.-K. Effect of dispersed micro-fibers on tensile behavior of uncoated carbon textile-reinforced cementitious mortar after high-temperature exposure. Cem. Concr. Compos. 2021, 118, 103949. [Google Scholar] [CrossRef]
  63. Thomas, J.; Ramaswamy, A. Mechanical properties of steel fiber-reinforced concrete. J. Mater. Civ. Eng. 2007, 19, 385–392. [Google Scholar] [CrossRef]
  64. Sivakumar, A.; Santhanam, M. Mechanical properties of high strength concrete reinforced with metallic and non-metallic fibres. Cem. Concr. Compos. 2007, 29, 603–608. [Google Scholar] [CrossRef]
  65. Afroughsabet, V.; Ozbakkaloglu, T. Mechanical and durability properties of high-strength concrete containing steel and polypropylene fibers. Constr. Build. Mater. 2015, 94, 73–82. [Google Scholar] [CrossRef]
  66. Atiş, C.D.; Karahan, O. Properties of steel fiber reinforced fly ash concrete. Constr. Build. Mater. 2009, 23, 392–399. [Google Scholar] [CrossRef]
  67. Farooq, F.; Ahmed, W.; Akbar, A.; Aslam, F.; Alyousef, R. Predictive modeling for sustainable high-performance concrete from industrial wastes: A comparison and optimization of models using ensemble learners. J. Clean. Prod. 2021, 292, 126032. [Google Scholar] [CrossRef]
  68. Aslam, F.; Farooq, F.; Amin, M.N.; Khan, K.; Waheed, A.; Akbar, A.; Javed, M.F.; Alyousef, R.; Alabdulijabbar, H. Applications of gene expression programming for estimating compressive strength of high-strength concrete. Adv. Civ. Eng. 2020, 2020, 8850535. [Google Scholar]
  69. Ahmad, W.; Ahmad, A.; Ostrowski, K.A.; Aslam, F.; Joyklad, P.; Zajdel, P. Application of Advanced Machine Learning Approaches to Predict the Compressive Strength of Concrete Containing Supplementary Cementitious Materials. Materials 2021, 14, 5762. [Google Scholar] [CrossRef]
  70. Khan, M.; Cao, M.; Ali, M. Cracking behaviour and constitutive modelling of hybrid fibre reinforced concrete. J. Build. Eng. 2020, 30, 101272. [Google Scholar] [CrossRef]
  71. Amin, M.N.; Ahmad, W.; Khan, K.; Ahmad, A.; Nazar, S.; Alabdullah, A.A. Use of Artificial Intelligence for Predicting Parameters of Sustainable Concrete and Raw Ingredient Effects and Interactions. Materials 2022, 15, 5207. [Google Scholar] [CrossRef]
  72. Amin, M.N.; Khan, K.; Ahmad, W.; Javed, M.F.; Qureshi, H.J.; Saleem, M.U.; Qadir, M.G.; Faraz, M.I. Compressive Strength Estimation of Geopolymer Composites through Novel Computational Approaches. Polymers 2022, 14, 2128. [Google Scholar] [CrossRef]
  73. Khan, K.; Ahmad, W.; Amin, M.N.; Ahmad, A.; Nazar, S.; Alabdullah, A.A.; Arab, A.M.A. Exploring the Use of Waste Marble Powder in Concrete and Predicting Its Strength with Different Advanced Algorithms. Materials 2022, 15, 4108. [Google Scholar] [CrossRef] [PubMed]
  74. Khan, K.; Ahmad, A.; Amin, M.N.; Ahmad, W.; Nazar, S.; Arab, A.M.A. Comparative Study of Experimental and Modeling of Fly Ash-Based Concrete. Materials 2022, 15, 3762. [Google Scholar] [CrossRef] [PubMed]
  75. Zou, Y.; Zheng, C.; Alzahrani, A.M.; Ahmad, W.; Ahmad, A.; Mohamed, A.M.; Khallaf, R.; Elattar, S. Evaluation of Artificial Intelligence Methods to Estimate the Compressive Strength of Geopolymers. Gels 2022, 8, 271. [Google Scholar] [CrossRef] [PubMed]
  76. Khan, M.; Lao, J.; Dai, J.-G. Comparative study of advanced computational techniques for estimating the compressive strength of UHPC. J. Asian Concr. Fed. 2022, 8, 51–68. [Google Scholar] [CrossRef]
Figure 1. Architecture of a typical neural network [49].
Figure 1. Architecture of a typical neural network [49].
Materials 15 06928 g001
Figure 2. Data description of compressive parameters.
Figure 2. Data description of compressive parameters.
Materials 15 06928 g002
Figure 3. Data description of flexural parameters.
Figure 3. Data description of flexural parameters.
Materials 15 06928 g003
Figure 4. Experimental and MLPNN predicted results for compressive strength.
Figure 4. Experimental and MLPNN predicted results for compressive strength.
Materials 15 06928 g004
Figure 5. Experimental and MLPNN predicted values with errors for compressive strength.
Figure 5. Experimental and MLPNN predicted values with errors for compressive strength.
Materials 15 06928 g005
Figure 6. Experimental and MLPNN predicted results for flexural strength.
Figure 6. Experimental and MLPNN predicted results for flexural strength.
Materials 15 06928 g006
Figure 7. Experimental and MLPNN predicted values with errors for flexural strength.
Figure 7. Experimental and MLPNN predicted values with errors for flexural strength.
Materials 15 06928 g007
Figure 8. Experimental and MLPNN-AdaBoost predicted results for compressive strength.
Figure 8. Experimental and MLPNN-AdaBoost predicted results for compressive strength.
Materials 15 06928 g008
Figure 9. Experimental and MLPNN-AdaBoost predicted values with errors for compressive strength.
Figure 9. Experimental and MLPNN-AdaBoost predicted values with errors for compressive strength.
Materials 15 06928 g009
Figure 10. Experimental and MLPNN-AdaBoost predicted results for flexural strength.
Figure 10. Experimental and MLPNN-AdaBoost predicted results for flexural strength.
Materials 15 06928 g010
Figure 11. Experimental and MLPNN-AdaBoost predicted values with errors for flexural strength.
Figure 11. Experimental and MLPNN-AdaBoost predicted values with errors for flexural strength.
Materials 15 06928 g011
Figure 12. Experimental and MLPNN-Bagging predicted results for compressive strength.
Figure 12. Experimental and MLPNN-Bagging predicted results for compressive strength.
Materials 15 06928 g012
Figure 13. Distribution of experimental and MLPNN-Bagging predicted values with errors for compressive strength.
Figure 13. Distribution of experimental and MLPNN-Bagging predicted values with errors for compressive strength.
Materials 15 06928 g013
Figure 14. Experimental and MLPNN-Bagging predicted results for flexural strength.
Figure 14. Experimental and MLPNN-Bagging predicted results for flexural strength.
Materials 15 06928 g014
Figure 15. Distribution of experimental and MLPNN-Bagging predicted values with errors for flexural strength.
Figure 15. Distribution of experimental and MLPNN-Bagging predicted values with errors for flexural strength.
Materials 15 06928 g015
Figure 16. Compressive strength statistical representation: (a) MLPNN; (b) MLPNN-AdaBoost; (c) MLPNN-Bagging.
Figure 16. Compressive strength statistical representation: (a) MLPNN; (b) MLPNN-AdaBoost; (c) MLPNN-Bagging.
Materials 15 06928 g016aMaterials 15 06928 g016b
Figure 17. Flexural strength statistical representation: (a) MLPNN; (b) MLPNN-AdaBoost; (c) MLPNN-Bagging.
Figure 17. Flexural strength statistical representation: (a) MLPNN; (b) MLPNN-AdaBoost; (c) MLPNN-Bagging.
Materials 15 06928 g017aMaterials 15 06928 g017b
Figure 18. Compressive strength sub-models results: (a) MLPNN-AdaBoost; (b) MLPNN-Bagging.
Figure 18. Compressive strength sub-models results: (a) MLPNN-AdaBoost; (b) MLPNN-Bagging.
Materials 15 06928 g018
Figure 19. Flexural strength sub-models results: (a) MLPNN-AdaBoost; (b) MLPNN-Bagging.
Figure 19. Flexural strength sub-models results: (a) MLPNN-AdaBoost; (b) MLPNN-Bagging.
Materials 15 06928 g019
Figure 20. Compressive strength feature importance.
Figure 20. Compressive strength feature importance.
Materials 15 06928 g020
Figure 21. Compressive strength SHAP plot.
Figure 21. Compressive strength SHAP plot.
Materials 15 06928 g021
Figure 22. Flexural strength feature importance.
Figure 22. Flexural strength feature importance.
Materials 15 06928 g022
Figure 23. Flexural strength SHAP plot.
Figure 23. Flexural strength SHAP plot.
Materials 15 06928 g023
Table 1. Reported applications and properties of SFRC.
Table 1. Reported applications and properties of SFRC.
Study FindingsConducted Studies
Purkiss [21]Patil and Sangle [22]Noaman, et al. [23]Boulekbache, et al. [24]Gholamhoseini, et al. [25]
Studied Properties of SFRCResidual Flexural Strength
Residual Compressive Strength
Compressive Strength
Flexural Strength
Compressive Strength
Flexural Strength
Compressive Strength
Flexural Strength
Compressive Strength
Flexural Strength
Reported Outcomes78–91% loss@800 °C
57–74% loss@800 °C
45.7 MPa
5.3 MPa
47 MPa
3.0–3.9 MPa
27–28.5 MPa
3.5–5.5 MPa
33.8–36.2 MPa
3.9–4.3 MPa
Considered ApplicationsRefractory MaterialBeams-Structural BeamsSlabs
Table 2. ML techniques used in the previous studies.
Table 2. ML techniques used in the previous studies.
Ref.Material TypeProperties PredictedML Techniques EmployedNo. of Input ParametersData PointsBest ML Technique Recommended
[42]Concrete-Filled Steel TubesUltimate axial capacityGene expression programming6227-
[43]Recycled aggregate concreteSplit-tensile strengthGene expression programming, artificial neural network, and bagging regressor9166Bagging regressor
[44]Rice husk ash concreteCompressive strengthGene expression programming and random forest6192Gene expression programming
[45]Geopolymer concreteCompressive strengthDecision tree, bagging regressor, and AdaBoost9154Bagging regressor
[46]Fly ash-based concreteCompressive strengthGene expression programming, artificial neural network, decision tree, and bagging regressor798Bagging regressor
[47]Fly ash-based concreteCompressive strengthGene expression programming, decision tree, and bagging regressor8270Bagging regressor
Table 3. Statistical checks of comparison of this study with the literature.
Table 3. Statistical checks of comparison of this study with the literature.
Material TypeParametersTechniquesMAE (MPa)RMSE (MPa)R2References
SFRCCompressive strengthMLPNN8.712.30.79This study
SFRCMLPNN-AdaBoost4.55.80.95This study
SFRCMLPNN-Bagging6.68.80.89This study
SFRCFlexural strengthMLPNN2.02.60.81This study
SFRCMLPNN-AdaBoost1.62.00.94This study
SFRCMLPNN-Bagging1.82.30.92This study
Recycled coarse aggregate concrete (RCAC)Compressive strengthSVM-AdaBoost7.79.50.94Amin, et al. [71]
Geopolymer concreteCompressive strengthMLPNN5.87.40.81Amin, et al. [72]
Geopolymer concreteCompressive strengthSupport vector machine6.78.10.78Amin, et al. [72]
Waste marble powder Concrete (WMC)Compressive strengthDT-AdaBoost3.97.90.91Khan, et al. [73]
Fly ash concreteCompressive strengthDecision Tree--0.88Khan, et al. [74]
Fly ash concreteCompressive strengthMLP--0.90Khan, et al. [74]
Fly ash concreteCompressive strengthBagging--0.93Khan, et al. [74]
Geopolymer concreteCompressive strengthDecision Tree4.16.20.88Zou, et al. [75]
Recycled coarse aggregate concrete (RCAC)Compressive strengthDT-XGBoost7.710.50.94Amin, et al. [71]
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Al-Hashem, M.N.; Amin, M.N.; Ahmad, W.; Khan, K.; Ahmad, A.; Ehsan, S.; Al-Ahmad, Q.M.S.; Qadir, M.G. Data-Driven Techniques for Evaluating the Mechanical Strength and Raw Material Effects of Steel Fiber-Reinforced Concrete. Materials 2022, 15, 6928. https://doi.org/10.3390/ma15196928

AMA Style

Al-Hashem MN, Amin MN, Ahmad W, Khan K, Ahmad A, Ehsan S, Al-Ahmad QMS, Qadir MG. Data-Driven Techniques for Evaluating the Mechanical Strength and Raw Material Effects of Steel Fiber-Reinforced Concrete. Materials. 2022; 15(19):6928. https://doi.org/10.3390/ma15196928

Chicago/Turabian Style

Al-Hashem, Mohammed Najeeb, Muhammad Nasir Amin, Waqas Ahmad, Kaffayatullah Khan, Ayaz Ahmad, Saqib Ehsan, Qasem M. S. Al-Ahmad, and Muhammad Ghulam Qadir. 2022. "Data-Driven Techniques for Evaluating the Mechanical Strength and Raw Material Effects of Steel Fiber-Reinforced Concrete" Materials 15, no. 19: 6928. https://doi.org/10.3390/ma15196928

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop