Fig 2 - uploaded by Majid Al-gburi
Content may be subject to copyright.
Source publication
The use of carbon fiber reinforced composite materials is an accepted technology that
is being used in practice to strengthen existing reinforced concrete (R/C) elements. An
artificial neural network (ANN) model was developed using past experimental data on
flexural failure of R/C beams strengthened by carbon FRP. The input parameters cover the
car...
Context in source publication
Context 1
... concrete cover up to 38 mm, the ultimate load capacity decrease with increasing concrete strength, the relationship is reverberate for concrete cover more than 38 mm. Fig (12) shows the ANN prediction of variation of load capacity and concrete cover (cov) for a rectangular beam of different values of carbon length. The ultimate loads are increase with increasing the concrete cover. ...
Similar publications
This paper presents an experimental study on shear strengthening of rectangular reinforced concrete (RC) beams with advanced composite materials. Key parameters of this study include: (a) the strengthening system, namely textile-reinforced mortar (TRM) jacketing and fiber-reinforced polymer (FRP) jacketing, (b) the strengthening configuration, name...
Fabric reinforced cementitious matrix (FRCM) systems are promising composite materials which are increasingly used for the strengthening of reinforced concrete components. The interaction of high performance, fine-grained concretes and non-corrosive, high strength textile reinforcements combines many advantages of shotcrete linings and externally b...
The present investigation deals with carbon-fiber-reinforced polymer (CFRP) rebar manufacturing, experimental flexural and shear response of near-surface-mounted (NSM) CFRP-strengthened reinforced concrete (RC) beams, and detailed parametric study of NSM-strengthened RC beams using Abaqus. The CFRP bars were manufactured locally using the twisting...
This research is a summary of previous researchers work done in this area. This paper provides a critical review of recent studies on strengthening of reinforced concrete and unreinforced masonry (URM) structures by fiber reinforced polymers (FRP) through near-surface mounting (NSM) method. The use of NSM-FRP has been on the rise, mainly due to com...
Steel fiber reinforced polymer (SRP) composite materials, which consist of continuous unidirectional steel wires (cords) embedded in a polymeric matrix, have recently emerged as an effective solution for strengthening of reinforced concrete (RC) structures. SRP is bonded to the surface of RC structures by the same matrix to provide external reinfor...
Citations
... Investigations have, however, revealed certain difficulties, such as the composites' tendency to debond from concrete surfaces before they achieve the laminates' rupture strength. As a solution to these issues, research into ANN models for structural behavior prediction has garnered interest [17] . The goal of this study is to predict the ultimate load capacity of reinforced concrete beams strengthened with FRP using multilayer perceptron neural networks and backpropagation algorithms. ...
... This study suggests a novel approach to fill this gap by estimating the bending moments of reinforced beams using ANN models. In civil engineering applications, ANNs are frequently employed as a tool for a variety of operations, including structural assessments, and shear strength prediction in FRP-strengthened RC elements [13,17,21,22] . The implementation of AI-based predictive maintenance systems for civil infrastructure is a possible area of future research. ...
... The model's parameters are 1 , 2 , 3 , 4 , 5 , 6 , 7 . Equation (17) can be used to expand Equation (16) since all factors can be altered linearly. Mu may be influenced by a variety of elements that engage in conversation with one another. ...
This study's objective is to overcome limitations in current design recommendations by exploring the application of machine learning to predict the flexural behavior of fiber-reinforced polymer (FRP)-strengthened reinforced concrete beams. Although FRP composites have completely changed structural strengthening, it might be challenging to predict bending moments with precision. This work fills the theoretical and experimental findings gaps by utilizing Artificial Neural Network (ANN) models in conjunction with computational techniques and statistical analysis. It includes gathering data, conducting a thorough literature review, and developing three models: Artificial neural network (ANN), Non-linear Regression (NLR), and Linear Regression (LR). Despite other models, the ANN model stands out for its superior performance and accurate predictions. Understanding material characteristics, FRP properties, and beam dimensions is critical in predicting flexural strength. The most significant parameter studied in this research is the overall depth of the beam (h), followed by the variation in bottom flexural reinforcement (ρs). Additionally, the FRP ratio (ρf) and beam width (b), which are both regarded as significant attributes, influence the flexural capacity of FRP-strengthened beams. The ultimate moment (Mu) may be predicted by the ANN model with an error range of-20% to +15%, indicating a significant advancement in strengthening approach optimization. This development could reduce the requirement for expensive experimental testing during construction, thereby enhancing the predictive capacity of structural engineering procedures. Furthermore, the design of flexurally strengthened RC beams with FRP may be made possible by depending on this model, specifically the ANN, without the need for experimental effort. https://creativecommons.org/licenses/by-nc/4.0/
... Civil engineers have used ANN to produce dynamic models that evaluate the response variables, and assess the relationships of the input parameters via parametric studies [3][4][5][6][7][8][9][10]. A variety of ANN models were employed to numerically investigate the nonlinear behavior of concrete structures [1][2][3][4][5][6][7]. Recent studies were conducted to numerically investigate the usage of ANN modeling in accurately predicting the shear capacity of RC beams and in performing sensitivity analyses; that would help in measuring the correlations between the independent parameters and the response variable [11][12][13]. ...
This study aims at the utilization of machine learning techniques in investigating the effect of measured geometric and mechanical properties of shear-strengthened reinforced concrete (RC) beams on the shear capacity of fiber reinforced polymers (FRP). Two complementary machine learning techniques were used; artificial neural network (ANN) and neural interpretation diagram (NID). The input parameters obtained from an experimental database were used to construct an ANN model that was programmatically validated. The validated ANN model was used to generate a NID that visually identifies the input parameters which have a direct association with the shear capacity of FRP. Moreover, two ANN models were developed, the first model consisted of all the independent parameters and the second model contained only the selected independent parameters. As a result, the ANN model with the selected independent parameters yielded predictions that are in close agreement with the experimental results compared to the ANN model with all the independent parameters. Thus, the implementation of machine learning techniques has proven to be an adaptive tool that can be fully expanded to other areas in structural engineering.
... A number of learning rules are available. The back propagation learning algorithm is used in this study (Salim & Majid, 2010). In is a simplest type of neural networks, A single node in the neural network model is called perceptron, a model consist of a number of these perceptrons arranged in layers that is why sometimes called multi-layer perceptron or MLP, starting from input nodes to the hidden layer nodes and finally to the output layer nodes, The cyclic path of the data is only one direction (forward), it is not going back to the input nodes or hidden layer nodes, that is why called by (feed-forward). ...
: Shear strength of ultra high performance reinforced concrete deep beams without stirrups
predicted by neural network models. The neural network model based on 233 beams from literatures
considering different parameters such as span to depth ratio, shear span to depth ratio, concrete
compressive strength, amount of longitudinal reinforcement,…etc. Neural network can be used as an
effective tool for predicting the shear capacity of normal & high strength concrete deep beams.
Prediction shear strength by neural network very close to the experimental results with correlation
coefficient of 0.836, while for ACIdesign eq., proposed eq. by Aziz & Zsutty where 0.394, 0.5624, and
0.488 respectively. The predicted shear strength model by neural network compared with ACI Code,
Aziz and Zsutty equations, the results show that the Neural Network approach adequately captured the
influence of concrete compressive strength on the shear capacity of reinforced concrete deep beams
without shear reinforcement.
... One form of artificial intelligence is the ANN, which attempts to mimic the function of the human brain and nerve system, but a simple unit of a neural network is much simpler than the biological cell [11]. A typical structure of ANN consists of a number of processing elements (PEs), or neurons, that usually are arranged in an input layer, an output layer and one or more hidden layers between, see figure 3 [12]. ...
... The LM algorithm is known to be significantly faster than the more traditional gradient descent type algorithms for training neural networks. It is, in fact, mentioned as the fastest method for training moderately sized feed-forward neural network [11]. While each iteration of the LM algorithm tends to take longer time than each repetition of the other gradient descent algorithms, the LM algorithm yields far better results using little iteration, leading to a net saving in computer processor time. ...
... It should be noted that any new input data should be scaled before being presented to the network, and the corresponding predicted values should be un-scaled before use, [11] and [13]. 7 ...
Restraint Formulation for Wall on Slab At Early Age Concrete Structures By Using ANN
... One form of artificial intelligence is the ANN, which attempts to mimic the function of the human brain and nerve system, but a simple unit of a neural network is much simpler than the biological cell [12]. ...
... It should be noted that any new input data should be scaled before being presented to the network, and the corresponding predicted values should be un-scaled before use, [12] and [14]. ...
Existing restraint curves have been applied to the method of artificial neural networks (ANN) to model restraint in the wall for the typical structure wall-on-slab. It has been proven that ANN is capable of modeling the restraint with good accuracy. The usage of the neural network has been demonstrated to give a clear picture of the relative importance of the input parameters. Further, it is shown that the results from the neural network can be represented by a series of basic weight and response functions. Thus, the results can easily be made available to any engineer without use of complicated software.
In addition to cement, sand, gravel, and water, the current investigation of the influence of additives on the compressive strength of concrete at 28 days includes fly ash, silica fume, and slag. 315 concrete compositions with various amounts of additives are trained and tested using an artificial neural network. Concrete strength is largely affected by the specific gravity of cement and the specific gravity of fine and coarse particles, according to the studies. For greater compressive strength, it is preferable to use materials with a higher specific gravity. Compressive strength has grown as the amount of silica fumes has increased. Increased amounts of slag or superplasticizer resulted in the same behavior. When the amount of fly ash was increased, the compressive strength of the material decreased.
This paper presents the use Machine Learning (ML) techniques to study the behavior of shear-deficient reinforced concrete (RC) beams strengthened in shear with side-bonded and U-wrapped fiber-reinforced polymers (FRP) laminates. An extensive database consisting of 120 tested specimen and 15 parameters was collected. The resilient back-propagating neural network (RBPNN) was used as a regression tool and the recursive feature elimination (RFE) algorithm and neural interpretation diagram (NID) were employed within the validated RBPNN to identify the parameters that greatly influence the prediction of FRP shear capacity. The results indicated that the RBPNN with the selected parameters was capable of predicting the FRP shear capacity more accurately (r² = 0.885; RMSE = 8.1 kN) than that of the RBPNN with the original 15 parameters (r² = 0.668; RMSE = 16.6 kN). The model also outperformed previously established standard predictions of ACI 440.R-17, fib14 and CNRDT200. A comprehensive parametric study was conducted and it concluded that the implementation of RBPNN with RFE and NID, separately, is a viable tool for assessing the strength and behavior of FRP in shear strengthened beams.
One of the widespread issues in concrete structures is cracks occurring at early age. Cracks that
appear in the young concrete may cause early start of corrosion of rebars or early penetration of
harmful liquids or gases into the concrete body. These situations could result in reduced service
life and in significantly increased maintenance cost of structures. Therefore it is important for
construction companies to avoid these cracks.
Volumetric deformations in early age concrete are caused by changes in temperature and/or the
moisture state. If such movements are restrained, stresses will occur. If the tensile stresses are
high enough, there will be a damage failure in tension and visible cracks arise. These stresses
are always resulting from a self-balancing of forces, either within the young concrete body
alone, i.e. without structural joints to other structures, or from the young concrete in
combination with adjacent structures through structural joints.
The decisive situation within a young concrete body alone is typically high stresses at the
surface when the temperature is near the peak temperature within the body. This situation occur
rather early for ordinary structures, say within a few days after casting for structures up to about
some meters thickness, but for very massive structures like large concrete dams, it might take
months and even years to reach the maximum tensile stresses at the surface. Usually this type of
cracks is denoted "surface cracks", and in some cases only a temperature calculation may give a
good perception to make decisions of the risk of surface cracking.
On the other hand, the decisive situation within a young concrete body connected to adjacent
structures, might include both risk of surface cracking at some distance away from the structural
joint and risk of through cracking starting in the neighborhood of the structural joint. If the
young concrete body is small in accordance to the adjacent structure, or, in other words, if there
is an overall high restraint situation in the young concrete, the risk of early surface cracking
might be out of question. So, restraint from adjacent structures represents one of the main
sources of thermal and shrinkage stresses in a young concrete body.
This study is mainly concentrated on establishing the restraint inside the young concrete body
counteracted by adjacent structures, and how to estimate the risk of through cracking based on
such restraint distributions. The restraint values in the young concrete are calculated with use of
the finite element method, FEM. Any spatial structure may be analyzed with respect to the level
of restraint. Calculations of risk of cracking are demonstrated with use of existing compensation
plane methods, and a novel method denoted equivalent restraint method, ERM, is developed for
the use of restraint curves. ERM enables the use of both heating of the adjacent structure and/or
cooling of the young concrete, which are the most common measures used on site to reduce the
risk of early cracking.
In a design situation many parameters are to be considered, like type of cement, different
concrete mixes, temperature in the fresh concrete, surrounding temperatures, temperature in the
adjacent structure, measures on site (heating/cooling/insulation), sequence order of casting.
Therefore, in general a lot of estimations concerning risks of cracking are to be performed. The
main objective with the present study is to develop methods speeding up and shorten the design
process.
Furthermore, established restraint curves have been applied to the method of artificial neural
networks (ANN) to model restraint in the slab, wall, and roof for the typical structures wall-onslab
and tunnel. It has been shown that ANN is capable of modeling the restraint with good
accuracy. The usage of the neural network has been demonstrated to give a clear picture of the
relative importance of the input parameters. Further, results from the neural network can be represented by a series of basic weight and response functions, which enables that the restraint
curves easily can be made available to any engineer without use of complicated software.
A new casting technique is proposed to reduce restraint in the newly cast concrete with a new
arrangement of the structural joint to the existing old concrete. The proposed technique is valid
for the typical structure wall-on-slab using one structural joint. This casting method means that
the lower part of the wall is cast together with the slab, and that part is called a kicker. It has
been proven by the beam theory and demonstrated by numerical calculations that there is a clear
reduction in the restraint from the slab to the wall using kickers.
Restraint is affected by casting sequence as well as boundary conditions and joint position
between old and new concrete elements. This study discusses the influence of different possible
casting sequences for the typical structure wall-on-slab and slab-on-ground. The aim is to
identify the sequence with the lowest restraint to reduce the risk of cracking.
In the last years, a great number of experimental tests have been performed to determine the ultimate strength of reinforced concrete (RC) beams retrofitted in flexure by means of externally bonded carbon fiber-reinforced polymers (CFRP). Most of design proposals for flexural strengthening are based on a regression analysis from experimental data corresponding to specific configurations which makes it very difficult to capture the real interrelation among the involved parameters. To avoid this, an intelligent predicting system such as artificial neural network (ANN) has been developed to predict the flexural capacity of concrete beams reinforced with this method. An artificial neural network model was developed using past experimental data on flexural failure of RC beams strengthened by CFRP laminates. Fourteen input parameters cover the CFRP properties, beam geometrical properties and reinforcement properties; the corresponding output is the ultimate load capacity. The proposed ANN model considers the effect of these parameters which are not generally account together in the current existing design codes with the purpose of reaching more reliable designs. This paper presents a short review of the well-known American building code provisions (ACI 440.2R-08) for the flexural strengthening of RC beams using FRP laminates. The accuracy of the code in predicting the flexural capacity of strengthened beams was also examined with comparable way by using same test data. The study shows that the ANN model gives reasonable predictions of the ultimate flexural strength of the strengthened RC beams. Moreover, the study concludes that the ANN model predicts the flexural strength of FRPstrengthened beams better than the design formulas provided by ACI 440
One of the widespread issues in concrete structures is cracks occurring at early age. Cracks that
appear in the young concrete may cause early start of corrosion of rebars or early penetration of
harmful liquids or gases into the concrete body. These situations could result in reduced service
life and in significantly increased maintenance cost of structures. Therefore it is important for
construction companies to avoid these cracks. Restraint represents one of the main sources of
thermal and shrinkage stresses at early age concrete.
Paper I, deals with both the compensation plane method, CPM, and local restraint method,
LRM, as alternative methods studying crack risks for early age concrete. It is shown that CPM
can be used both for cooling and heating, but basic LRM cannot be applied to heating. This
paper presents an improved equivalent restraint method, ERM, which easily can be applied both
for usage of heating and cooling for general structures. Restraint curves are given for two
different infrastructures, one founded on frictional materials and another on rock. Such curves
might be directly applied in design using LRM and ERM.
In Paper II, existing restraint curves have been applied to the method of artificial neural
networks (ANN) to model restraint in the wall for the typical structure wall-on-slab. It has been
proven that ANN is capable of modeling the restraint with good accuracy. The usage of the
neural network has been demonstrated to give a clear picture of the relative importance of the
input parameters. Further, it is shown that the results from the neural network can be represented
by a series of basic weight and response functions. Thus, the results can easily be made
available to any engineer without use of complicated software.
Paper III, discusses the influence of five casting sequences for the typical structure slab-onground.
The aim is to map restraints from adjacent structures for a number of possible casting
sequences, and to identify the sequence with the lowest restraint. The paper covers both
continuous and jumped casting sequences, which include one, two and three contact edges. The
result shows that the best casting sequence is the continuous technique with one contact edge.
