Prediction Of Bond Strength Research Articles

Ensemble Machine Learning-Based Approach for Predicting of FRP–Concrete Interfacial Bonding

Developments in fiber-reinforced polymer (FRP) composite materials have created a huge impact on civil engineering techniques. Bonding properties of FRP led to its wide usage with concrete structures for interfacial bonding. FRP materials show great promise for rehabilitation of existing infrastructure by strengthening concrete structures. Existing machine learning-based models for predicting the FRP–concrete bond strength have not attained maximum performance in evaluating the bond strength. This paper presents an ensemble machine learning approach capable of predicting the FRP–concrete interfacial bond strength. In this work, a dataset holding details of 855 single-lap shear tests on FRP–concrete interfacial bonds extracted from the literature is used to build a bond strength prediction model. Test results hold data of different material properties and geometrical parameters influencing the FRP–concrete interfacial bond. This study employs CatBoost algorithm, an improved ensemble machine learning approach used to accurately predict bond strength of FRP–concrete interface. The algorithm performance is compared with those of other ensemble methods (i.e., histogram gradient boosting algorithm, extreme gradient boosting algorithm, and random forest). The CatBoost algorithm outperforms other ensemble methods with various performance metrics (i.e., lower root mean square error (2.310), lower covariance (21.8%), lower integral absolute error (8.8%), and higher R-square (96.1%)). A comparative study is performed between the proposed model and best performing bond strength prediction models in the literature. The results show that FRP–concrete interfacial bonding can be effectively predicted using proposed ensemble method.

Predicting Bond Strength between Steel Rebar and Ultra-High Performance Concrete Using Random Forest

Predicting Bond Strength between Steel Rebar and Ultra-High Performance Concrete Using Random Forest

Local bond strength prediction of deformed reinforcing bars in concrete considering heterogeneity at interface regions

An analytical model is proposed to predict local bond strength (τf) by incorporating heterogeneity at interface regions for deformed reinforcing bars centrally anchored in concrete. The rib width on the bar surface is introduced as an interfacial characteristic parameter G in the proposed model; this accounts for the heterogeneity. Both τf and the local interfacial fracture energy (GIIf) of each specimen were found to be linked to G and can be determined analytically from the maximum pull-out loads (Fmax) from tests. It was found that the predicted τf was larger than the maximum average bond stress (τavg-max); the discrepancy between the two values reduced with an increase in L/G. Moreover, with an increase in L/G, the predicted τf showed a certain decrease, with the reduction decreasing with stronger interfacial homogeneity. The predicted GIIf was found to be significantly increased because of the weaker boundary effect. The validity of the proposed model was verified using comparisons of predicted Fmax (using the determined values of τf and GIIf) and the experimental Fmax, with the only failure mode being bar pull-out. Moreover, the model can be applied to steel or fibre-reinforced polymer bars and the concrete refers to all types of cementitious materials.

Impact of fiber distribution and cyclic loading on the bond behavior of steel-reinforced UHPC

Ultra-high performance concrete (UHPC) is a class of concrete materials that exhibits high compressive strength (over 150 MPa) and is seeing increasing applications around the world. Current reinforced UHPC (R/UHPC) bond studies mostly adopt non-flexural setups, i.e., pull-out tests and tension-lap-splice tests. The bond behavior of R/UHPC in flexure remains largely unknown, which is a fundamental aspect of designing R/UHPC flexural elements. Additionally, the impact of fiber distribution and cyclic loading on R/UHPC bond is little understood, which is important for guiding the casting and design of R/UHPC structures especially in earthquake zones. This study experimentally investigates the impact of fiber volume, cast flow direction, and cyclic loading on R/UHPC bond behavior under a simulated flexural stress state. Results from twenty-five beam-end tests are discussed. Specimens cast with flow perpendicular to the bar exhibited a 9%–26% higher bond strength than specimens cast with the flow parallel to the bar due to the resulting higher fiber-bridging capacity across the splitting crack plane. Compared to the monotonically-loaded specimens, cyclic loading minimally impacts R/UHPC bond strength and accelerates bond-degradation after bond softening occurs. Current R/UHPC bond strength prediction methods are evaluated on the test results from this study and the literature. An energy-based R/UHPC bond-slip model is developed. This model captures the cyclic bond degradation of R/UHPC and other types of high-performance fiber-reinforced cementitious composites (HPFRCC) with a mean absolute error under 13%.

Predictive Models for Bond Strength of Reinforced Concrete with the Application of ANN

The bond strength of grip b/w steel and concrete can be defined as the resistant to separating concrete or mortar from the reinforced bar. This bond strength is the most critical characteristic of reinforced-cement concrete. Structural performance depends upon this characteristic, especially in the failure phase. Bond strength is primarily dependable on many variables that affect this attribute. These variables include the diameter of the reinforced steel bar, bond extent, length to diameter ratio, cube compressive strength, concrete cover, cover to dimeter ratio, volume fraction and most importantly, different temperatures. Up to 150°C, there is no such change in bond strength of reinforcement concrete, but when the temperature rises beyond 150°C, it starts to decreasegradually. We have collected experimental data from the internationally published record. This study will see the change in bond strength at these temperature variations i.e., 200°C, 400°C, and 600°C. This observational study will represent a soft computing tool, i.e., an Artificial Neural network (ANN), to predict and measure the grip strength between concrete and steel bar at elevated temperatures. The bond strength of reinforced concrete has been predicted by using ANN Models. Data set based upon the different factor that affects the bond strength has been used as input for generating ANN model & ultimate bond strength of reinforced concrete has been used as output during the development of the ANN model. This model was then prepared to predict bond strength and affected by many input features and recorded a linear regression analysis. The predicted result then confirmed the accuracy and high estimation capability of the model.

Bond behavior between corroded steel bar and concrete under sustained load

The degradation of bond behavior between corroded steel bars and concrete under sustained load is among the main issue for the durability of structures. In this study, the impressed current method was used to achieve target corrosion levels of tensile steel bar (0%, 1%, 3%, and 5%), while specimens bearing the sustained load (0%, 25%, 45%, and 65% ultimate load). The sustained load lasted for 60 days, during which the instantaneous slip and time-dependent slip were recorded. Finally, the bond strength of all kinds of specimens was measured by the flexural-bond test. The test results show that corrosion and load significantly affect the slip and bond strength between steel bar and concrete. Slight corrosion (less than 1%) leads to higher bond strength, up to 1.22 times that of the control group, while severe corrosion (more than 2.42%) leads to lower. Additionally, the strength reduction caused by the coupling of load and corrosion is more obvious than the single, by up to 45.66%. In addition, based on the test results, the time-dependent slip and bond strength prediction model were proposed, which can be used for predicting the service life of structures in the erosive environment.

An experimental and numerical investigation into the effects of bond line thickness variations on the strength of composite bonded joints

The structural integrity of adhesive joints used in the attachment of laminated composite panels has received considerable attention over the past few decades. Of particular concern and the topic of this paper is the effect that a nonuniform bond line thickness has on the load carrying capacity of a bonded joint. To this end, carbon fiber adherends were joined in a single lap joint configuration using a two-part epoxy adhesive. The joints were initially fabricated and tested with various uniform bond line thickness in a single lap joint configuration to establish a benchmark. Bonded single lap joints with linearly varying bond line thickness were then investigated. A finite element study was performed to predict bond strength of the uniform and linearly varying bond lines using the Critical Zone method. The failure strength of the uniform bond line thickness specimens was found to decrease with an increase in the bond line thickness. In the case of linearly varying bond line thickness specimens, the bond strength decreased as the bond thickness at one end of the bonded joint overlap deviated from its intended value. The finite element model using the Critical Zone method was able to predict the failure strength of the joints in both uniform and linearly varying bond thickness configurations. The reduction in failure strength was attributed to an increase in the stresses in the adherend ply adjacent to the bond line as determined through the finite element stress analysis.

Predicting Bond Strength of Steel Reinforcement in Self-Compacting Concrete (SCC) Using Adaptive Neuro-Fuzzy Inference System (ANFIS)

The advantages of using self-compacting concrete (SCC) are reducing the time of construction and the number of employments, reducing noise that can disturb the surrounding environment, and increasing the density of hardened concrete structural elements, automatically affecting bond strength reinforcement in SCC. The bond strength is a parameter as an essential factor affecting the behavior of reinforced concrete. In this manuscript, the Adaptive Neuro-Fuzzy Inference System (ANFIS) model was built to predict the bond strength in SCC. For showing the performance of the ANFIS model, the level of accuracy-based correlation coefficient (R²) and Root Mean Square Error (RMSE) were determined. Learning process data consists of input and output. The input in this study includes compressive strength of concrete (f'_c), the diameter of steel reinforcement (d_b), and development of length (L_d), while the output bond strength (<img src=image/14824241_01.gif>). The results of the proposed model were in good agreement with the experimental results, as evidenced by an R² of 0.71 and an RMSE of 3.31 MPa in the testing data, indicating that the proposed ANFIS model is capable of accurately predicting steel reinforcement bond strength in SCC.

Bond durability and degradation mechanism of GFRP bars in seawater sea-sand concrete under the coupling effect of seawater immersion and sustained load

In this paper, the bond durability of glass fiber-reinforced polymer (GFRP) bars in seawater sea-sand concrete (SWSSC) under the coupling effect of seawater immersion and sustained load was experimentally investigated. The test parameters included bar diameter (12 and 16 mm), immersion time (90, 180, and 270 days), and sustained load condition (loaded and unloaded). Test results showed that for the specimens with a small bar diameter, the bond strength had a slight growth at the initial stage of exposure mainly due to the water swelling of the bar, and then decreased due to the bond degradation. For the specimens with a large bar diameter, the bond strength increased with the increase of immersion time, which was determined by the increasing compressive strength of SWSSC. The sustained load reduced the bond strength of the specimens with a small bar diameter due to its acceleration of bond degradation, but had little effect on the bond strength of the specimens with a large bar diameter. The degradation mechanism was the rib deterioration of the bars caused by seawater immersion. Moreover, the sustained load could accelerate the degradation. The predicted bond strength retention of the specimen with a bar diameter of 12 mm under the coupling effect of the marine environment (moisture saturated and the mean annual temperature of 25–35 °C) and sustained load is 52% after 50 years of service.

Prediction of bond strength in the fused shirt composites

Purpose Bond strength is an indicator of the quality of the fusing process. The study's primary purpose is to predict the bond strength using easily measurable variables. This study focuses on shirting fabrics fused with woven interlinings and changes in bond strength before and after washing. Design/methodology/approach This study attempts to model and predict bond strength of fused shirt composites using an initial screening design followed by full factorial design of experiments. After screening out, those found significantly affecting the bond strength are fabric fiber content, interlining areal weight and fusing temperature. This study proposes the regression models explaining the effect of the three variables on bond strength before and after washing the fused composites. Findings This study found that heavy interlinings (250 g/sq.m) require higher fusing temperatures than the lighter interlining (225 g/sq.m). After washing, the bond strength of samples fused at high temperatures reduced in some instances. Maintaining a high temperature without considering interlining weight can lower the bond strength after washing the fused composites. Originality/value A combination of screening and full factorial DOE is used to analyze and predict bond strength of composites comprising medium-weight shirting fabrics and heavier weight interlinings used in shirts.

Parameter Study on Friction Surfacing of AISI316Ti Stainless Steel over EN8 Carbon Steel and Its Effect on Coating Dimensions and Bond Strength.

Friction surfacing is a solid-state coating process that uses plastic deformation to improve the efficiency of the core metallic pattern, resulting in fine-grained coatings with superior wear and corrosion properties. This article focuses on the development of inherently homogeneous, non-diluted coating of AISI316Ti stainless steel above EN8 and also encloses the empirical relationship for the prediction of bond strength (Bs), coating thickness (Ct), and coating width (Cw). The key individualities for bonding geometry were believed to be the process parameters such as rotational speed (rpm), traverse speed (mm/s), and axial load (kN). The effect of input parameters on the bond’s external dimensions and strength was investigated using a multi-objective optimization approach through experimentation. The bond’s strength improved as the coating thickness was reduced and the coating width was increased. The grain-refined coatings superimposing martensitic microstructure with no deposition of carbide particles added value to the metallurgical study using the scanning electron microscope.

Development of data-driven prediction model for CFRP-steel bond strength by implementing ensemble learning algorithms

Bonding carbon fiber reinforced polymer (CFRP) laminates has been broadly utilized in steel structure rehabilitation. As for the final strengthened capacity, the bond strength between CFRP and steel usually plays a dominant role instead of the CFRP’s mechanical property. However, the bond behavior of the CFRP-steel (CS) interface is very complicated with various failure modes and consequently the bond strength is hard to estimate leading to the CFRP strengthened steel structure insecure. Under this circumstance, in order to accurately predict the bond strength of CS, efficient data-driven models were developed through implementing ensemble learning (EL) algorithms named by “gradient boosting decision tree (GBDT)” and “random forest (RF)” as two representative ones on a collected CS single-shear test database. These models’ performances on bond strength prediction were compared and also three representative machine learning algorithms “artificial neural network (ANN)”, “support vector machine (SVM)” and “classification and regression tree (CART)” are utilized for validating the necessity. The comparison results indicate that the model generated by the GBDT algorithm attains the best accuracy for CS interfacial bond strength prediction (R2 = 0.98) among the ensemble and machine learning algorithms. Through model explaning analysis, the mechanism behind the GBDT based prediction model was also carefully verified. After these tests and analyses, the GBDT based model was proved to have the potential to facilitate the design and evaluation of CFRP strengthened steel structures.

Local bond strength prediction of deformed reinforcing bars in concrete by considering heterogeneity at interface regions

An analytical model is proposed in this paper to predict the local bond strength τf by incorporating the heterogeneity at interface regions for deformed reinforcing bars centrally anchored in concrete. The rib width on bar surface is introduced as an interfacial characteristic parameter G in the proposed model accounting for the heterogeneity. Both the τf and local interfacial fracture energy GIIf from each specimen are linked to the G and determined analytically if the maximum pull-out loads Fmax are given from the test. It is found that the predicted τf is larger than the maximum average bond stress τavg-max. The discrepancy between the two values is reduced with the increasing of L/G. Moreover, as the L/G increases, the predicted τf shows a certain decrease and the reduction becomes smaller with stronger interfacial homogeneity. The predicted GIIf is significantly increased because of the weaker boundary effect. The validity of the proposed model is verified upon the comparisons between the predicted Fmax using the determined τf and GIIf and the experimental ones with the only failure mode of bar pull-out. Moreover, the bars here can be steel or FRP (fiber-reinforced polymer) bars, and the concrete refers to all types of cementitious materials.

Ensemble learning based approach for FRP-concrete bond strength prediction

Nowadays, externally bonding fiber reinforced polymer (FRP) plates or sheets have become a major maintenance approach for aged reinforced concrete flexure structures. However, the capacity of strengthend structure cannot be precisely estimated as a result of the critical FRP-concrete interfacial (FCI) bond strength unpredictable. In order to solve this issue, many experimental studies have been carried out with corresponding emipirical models proposed. Due to limited experiment samples, these models were found more or less lacking the generalization ability. Under this circumstance, in this study, an ensemble learning algorithm “gradient boosted regression trees” (GBRT) was employed to develop a prediction model for FCI bond strength prediction based on a collected comprehensive database containing 520 tested samples. The model’s performance has been thoroughly compared with the representative empirical models and the common utilized machine learning algorithms. The rationality of this model has also been discussed through feature importance analysis. The results showed that the model in this study exhibits the highest accuracy and is proven to be feasible for predicting FCI bond strength in actual practice.

Bond durability of basalt‐fiber‐reinforced‐polymer bars embedded in lightweight aggregate concrete subjected to freeze–thaw cycles

AbstractIn recent years, basalt‐fiber‐reinforced‐polymer (BFRP) bars have become popular to replace steels. However, little is known about the performance of BFRP bars inserted in lightweight aggregate concrete (LWAC), especially the bonding performance under freeze–thaw (FT) cycles. In this study, a bond strength test was conducted for BFRP bars embedded in LWAC after FT cycles. BFRP bars, epoxy‐coated bars, and steels were tested for comparison purposes. The test parameters are the bar type, compressive strength of the LWAC, and FT cycles; 108 specimens were subjected to 0, 15, 30, and 50 FT cycles. Moreover, the bond strength models of the three types of bars were fitted to LWAC. The test results prove that the number of FT cycles significantly affects the bonding performance of the bar embedded in LWAC. The results also show that the bond strength prediction of the fitting model considering the FT cycle effect is consistent with the experimental results.

Bond strength degradation in concrete cracked by expansion agent filled pipes

AbstractIn this study, the bond behavior of corroded reinforced concrete was investigated by conducting pull‐out tests on concrete specimens with variable induced crack widths and different splitting modes. To simulate cracks due to corrosion and focus on their effect on bond strength degradation, a novel method using expansion agent filled pipes is proposed. With the increase of the crack width over elapsed time from filling the aluminum pipe with an expansion agent, a target crack width was obtained. The results confirm that the type of splitting greatly influences the reduction in bond strength with deterioration of bond strength more severe in “Side‐split” than in “Single‐split.” Furthermore, using the surface crack width as a variable, empirical formulas to predict bond strength degradation due to corrosion of rebar are proposed. These prediction models give a good correlation when compared to the available literature.

Experimental Investigation and Artificial Neural Network Based Prediction of Bond Strength in Self-Compacting Geopolymer Concrete Reinforced with Basalt FRP Bars

The current research on concrete and cementitious materials focuses on finding sustainable solutions to address critical issues, such as increased carbon emissions, or corrosion attack associated with reinforced concrete structures. Geopolymer concrete is considered to be an eco-friendly alternative due to its superior properties in terms of reduced carbon emissions and durability. Similarly, the use of fibre-reinforced polymer (FRP) bars to address corrosion attack in steel-reinforced structures is also gaining momentum. This paper investigates the bond performance of a newly developed self-compacting geopolymer concrete (SCGC) reinforced with basalt FRP (BFRP) bars. This study examines the bond behaviour of BFRP-reinforced SCGC specimens with variables such as bar diameter (6 mm and 10 mm) and embedment lengths. The embedment lengths adopted are 5, 10, and 15 times the bar diameter (db), and are denoted as 5 db, 10 db, and 15 db throughout the study. A total of 21 specimens, inclusive of the variable parameters, are subjected to direct pull-out tests in order to assess the bond between the rebar and the concrete. The result is then compared with the SCGC reinforced with traditional steel bars, in accordance with the ACI 440.3R-04 and CAN/CSA-S806-02 guidelines. A prediction model for bond strength has been proposed using artificial neural network (ANN) tools, which contributes to the new knowledge on the use of Basalt FRP bars as internal reinforcement in an ambient-cured self-compacting geopolymer concrete.

Bond strength prediction of concrete-encased steel structures using hybrid machine learning method

The mechanical behavior of concrete-encased steel (CES) structures is crucially linked to the bond strength between steel section and concrete which should be predicted with ease and accuracy. In this paper, an efficient and robust soft computing strategy was proposed, where the artificial neural networks (ANN) are hybridized with genetic algorithm (GA) or particle swarm optimization (PSO) to predict the bond strength in CES structures. Seven features were extracted from push-out tests in the available literature, and a database containing 191 records was established for model training and testing. Then, the performance of three machine learning models ANN, GA-ANN, and PSO-ANN was thoroughly compared. The results showed that the developed GA-ANN and PSO-ANN models exhibit superior performance to both conventional ANN model and existing empirical equations. The PSO-ANN outperforms the GA-ANN in terms of convergence speed and prediction error owing to its unique information-sharing mechanism. Further, sensitivity analysis of main contributing factors was conducted on PSO-ANN model. It is quantitatively confirmed that the relative concrete cover has the most significant effect on bond strength while the influence of relative bonded length is relatively minimal. Eventually, an explicit formula was directly derived from the PSO-ANN model and a practical tool with a graphical user interface was created for design practice. The outcome of this study could be employed to intelligently estimate the bond strength in CES without costly and time-consuming tests.

Bond Behavior of Steel-Recycled Aggregate Concrete Interface After High Temperatures and Spraying Water Cooling

Using recycled aggregate concrete (RAC) in steel-reinforced concrete structure is an effective way to eliminate the adverse effects of recycled aggregate, which has an excellent application prospect. Fire has a great destructiveness to steel-reinforced recycled aggregate concrete (SRRAC) structure; hence, the bond performance of SRRAC after high temperature, as the prerequisite for the composite between steel and RAC, is the key problem for structural safety and the corresponding safety assessment after fire. In this article, the residual bond behaviors of steel-recycled aggregate concrete interface after different high temperatures and spraying water cooling were studied through the push-out test. The failure modes and load–slip curves were examined. The ultimate bond strength, residual bond strength, and elastic bond shear stiffness of specimens after high temperature and cooling for the regime of spraying water were evaluated and compared to that of natural air. A parametric analysis of temperature, replacement percentage, and studs was conducted. A calculation approach for the ultimate bond strength and residual bond strength of SRRAC after high temperature was developed based on the sensitive analysis of gray system theory and regression analysis. Results showed that the bond properties of SRRAC specimens after high temperature were decreased as exposure temperature increased. The cooling regime of spraying water has a more significant influence on the ultimate strength and residual strength than that of natural air. The specimens with studs on both flange and web have the highest bond properties after high temperature. In all the factors, the number of studs showed the largest gray relational degree to the bond strength of SRRAC. The developed approach provided a reliable prediction of bond strength for SRRAC after high temperature.

Bond strength prediction of spliced GFRP bars in concrete beams using soft computing methods

The bond between the concrete and bar is a main factor affecting the performance of the reinforced concrete (RC) members, and since the steel corrosion reduces the bond strength, studying the bond behavior of concrete and GFRP bars is quite necessary. In this research, a database including 112 concrete beam test specimens reinforced with spliced GFRP bars in the splitting failure mode has been collected and used to estimate the concrete-GFRP bar bond strength. This paper aims to accurately estimate the bond strength of spliced GFRP bars in concrete beams by applying three soft computing models including multivariate adaptive regression spline (MARS), Kriging, and M5 model tree. Since the selection of regularization parameters greatly affects the fitting of MARS, Kriging, and M5 models, the regularization parameters have been so optimized as to maximize the training data convergence coefficient. Three hybrid model coupling soft computing methods and genetic algorithm is proposed to automatically perform the trial and error process for finding appropriate modeling regularization parameters. Results have shown that proposed models have significantly increased the prediction accuracy compared to previous models. The proposed MARS, Kriging, and M5 models have improved the convergence coefficient by about 65, 63 and 49%, respectively, compared to the best previous model.