International Journal of the Physical Sciences Vol. 7(29), pp. 5193-5201, 16 August, 2012 Available online at http://www.academicjournals.

org/IJPS DOI: 10.5897/IJPS12.155 ISSN 1992-1950 2012 Academic Journals

Full Length Research Paper

Prediction of concrete compressive strength in buildings that would be reinforced by fuzzy logic
Metin Uzunolu, Ercan zgan* and Tuncay Kap
Construction Department, Faculty of Technical Education, Duzce University, Konuralp, Dzce, Turkey.
Accepted 9 July, 2012

In this study, core samples were taken from column or reinforced wall in order to make reinforcement or restoration. The effect of the element height and volume of voids of these samples to the compressive strength were investigated. C25 type ready to use concrete was used and a column without steel was produced. During the production process, 15 cm reference cube samples were taken from the fresh concrete. After 28 days, 8 cores from reference cubes and 128 samples from column were taken and their compressive strengths were determined. As a result, the average compressive strength of the reference core samples was 36.95 Mpa and according to the height compressive strengths of other samples was ranging between 37.3 and 43.0 Mpa. The ratio of the increasing of the compressive strength changes between 0.95 and 16.37% according to the reference sample. Based on the experimental results, fuzzy Logic method was used for modelling of the experimental results. The paper demonstrates that fuzzy logic can be successfully used in modelling the compressive strength of concrete for different height and volume of voids. This model enables us to easily and reliably estimate the compressive strength of concrete. Key words: Concrete, reinforcement, compressive strength, prediction model, fuzzy logic.

INTRODUCTION A necessary prerequisite for devising a restoration or reinforcement scheme for a deteriorating structure is the acquisition of relevant in situ data to be used for evaluating structural adequacy as well as defining the scope of restoration work. The methodology for conducting a successful structural in situ investigation includes: strategic planning, preliminary visual inspection, loading assessment, material property evaluation and load testing techniques. The single and the most important after material property used for the evaluation of residual load capacity and structural adequacy of deteriorated concrete structures is the in situ concrete strength. The most direct method of obtaining this data is to test core specimens retrieved from the structure. It is not, however, always feasible to core for specimens and even if it is, the number of cores that can be taken is often limited. The solution is therefore to use indirect non-destructive and partially destructive techniques which measure concrete properties other than strength in combination with testing core specimens. Several of those techniques especially suited for in situ investigation are reviewed. Prior to using any of the indirect methods, a correlation needs to be established with the in situ concrete strength (Tay and Tam, 1996). The compressive strength of concrete is used as the most basic and important material property when reinforced concrete structures are designed (Yi et al., 2006). The common way of determining in-situ strength of concrete is to drill and test cores (Troxell et al., 1968; Neville, 1995; Mindess et al., 1981; Arioglu and Arioglu, 1998; Erdoan, 2003; Sullivan, 1991; Haque and AlKhaiat, 1997; Miao et al., 1993; Price et al., 1996; Khayat et al., 1997). Although the method consists of expensive and time consuming operations, cores give reliable and useful results since they are mechanically tested to be destructive (Neville, 1995). However, the test results should be carefully interpreted because core strengths are affected by a number of factors such as diameter, h/d ratio and moisture condition of the core specimen, the direction of drilling, presence of reinforcement steel bars

*Corresponding author. E-mail: ercanozgan@gmail.com.

5194

Int. J. Phys. Sci.

Table 1. Concrete mix design parameters class of the concrete C 20/25.

Dmax Aggregate 0-5K 15-25 5-15 Stone dust CEM2 42.5R Super plasticizers Water

22 Amount (kg) 546 470 486 347 290 2.32 183 2.324

Consistency 8.3 Goal of the Moisture content (%) prescription (kg) 2.730 5.6 2.350 1.0 2.430 1.0 1.735 5.5 9.245 288 kg 1.450 11.6 915 2.377 3 Unit weight (for each m )

Water/Cement Goal of the system (kg) 2.880 2.372 2.452 1.828 9.532 1.450 11.6 626 2.088 2.293

0.63 Weighted quantity (kg) 2.861 2.251 2.417 1.851 1.448 11.497 623 Total (kg) 2.861 2.251 2.417 1.851 9.380 1.448 11.497 623 2.082 11.462

in the specimen and even the strength level of the concrete (Bungey, 1979; Bartlett and MacGregor, 1994; Bartlett and MacGregor, 1993; Bartlett, 1997). The diameter of the core plays an important role in the evaluation of core strength results. Both ASTM and British Standards (BS) specify a minimum core diameter of 100 mm providing that the diameter of the core is at least three times larger than the maximum aggregate size in concrete mixture ASTM C 42-90, (1994), and BS 1881: Part 120 (1983). It was indicated that core strength was affected by both specimen and aggregate sizes. Therefore, it was proposed that the size of specimen and aggregate, type of aggregate, age of the specimen should be taken into consideration for the interpretation of the results used to convert the strengths of cores to those of standard cylinders or cubes (Tuncana et al., 2008). Turkish Standard TS EN 12504-1 allows the use of 50 mm diameter cores TS EN 12504-1 (2002). However, no correction factors to convert the strength of 50 mm cores to those of cores having larger diameter are given in the standard. By using cylinder, cube, and prism specimens, the effect of specimen sizes, specimen shapes, and placement directions on compressive strength of concrete specimens were experimentally investigated based on fracture mechanics. The test results are curvefitted using least square method (LSM) to obtain the new parameters for the modified size effect law (MSEL). The analysis results show that the effect of specimen sizes, specimen shapes, and placement directions on ultimate strength is present. In addition, correlations between compressive strengths with size, shape, and placement direction of the specimen are investigated (Seong-Tae et al., 2006). However, by using different techniques, different prediction models were developed and used by researchers in different studies to predict of experimental or observational results (Jafri et al., 2012a; Jafri et al., 2012b; Jafri et al., 2012c; Jafri et al., 2012d). The purpose of this study is to experimentally

investigate the axial compressive strength of concrete core samples retrieved from column and suggest a model for the prediction of compressive strengths with height of column and volume of voids in core samples based on the fracture mechanics. Based on the experimental results, a numerical model was established within the Mat-lab programming environment and fuzzy Logic method was used for this purpose. The paper demonstrates that fuzzy logic can be successfully used in modelling the compressive strength of concrete for different height and volume of voids. This model enables us to easily and reliably estimate the compressive strength of concrete by using only the height and volume of voids parameters of concrete.

EXPERIMENTAL DETAILS Formwork In this study, a column block in a cross-section with 100 25 cm and 250 cm height was produced. While producing the block, 3 cm wide metal framed wooden panel with 50 100 2.5 cm dimensions was used on the front and back surfaces. 25 250 2.5 cm pine timber was used on the sides. After this process, sika concentrated block liquid was applied on the interior surfaces of the block and the formwork was prepared for pouring the concrete. Concrete C25 ready mixed concrete was poured into the formwork as to prevent segregation. The mixture design of the concrete is given in Table 1. Concrete was poured in the form of 50 cm layers. Each layer was vibrated by a 3 cm diameter vibrator from 4 different points. During this process, in order to use as reference, 8 cube samples in 15 cm dimension were taken according to Turkish Standards TS EN 12350-1 and they were kept in cure tank for 28 days.

Taking of the core samples 3 days after pouring the concrete, the formworks were left out. Wet

Uzunolu et al.

5195

weights in water were measured. Samples were kept in drying oven in 1052C till they have constant weights.

Determination of the compressive strength of samples In order to conduct compressive strength test, top and bottom surface of the samples were covered by 70% sulphur and 30% graphite mixture. After keeping the samples for 24 h in the laboratory setting, compressive strength test was applied according to the Turkish Standard TS EN 123904. The compressive strength values of the reference samples and the physical properties of others were given in Table 2. APPLICATION OF FUZZY LOGIC METHOD The effects of the structural element height and volume of voids in concrete on the compressive strength of the concrete core samples was tried to modelling by using the fuzz logic method. Using the fuzzy logic, the structural element height and volume of voids of the concrete on the compressive strength for the concrete core samples with reasonable precisions are all systematically varied to identify the effects of each combination for finding the compressive strength of the concrete in laboratory or especially in field. It is likely impossible to find the different compressive strength of the concrete without using an efficient prediction method that satisfies the constraints. Fuzzy logic was implemented to determine a set of unknown parameters which best matched the compressive strength model predictions with experimental data. This modelling procedure can be used as a guideline for experiments to determine the compressive strength of the concrete. In the presented study, based on the experimental results, a numerical model was established within the Mat-lab programming environment. Fuzzy Logic method was used this purpose. The paper demonstrates that fuzzy logic can be successfully used in modelling the compressive strength of concrete for different height and volume of voids. This model enables us to easily and reliably estimate the compressive strength of concrete by using only the height and volume of voids parameters of concrete. Fuzzy inference systems (FIS) The process of fuzzy inference involves all of the pieces that are membership functions, fuzzy logic operators, and if-then rules [17]. There are two types of fuzzy inference systems, they are; Mamdanitype and Sugeno-type. In this study, Mamdani fuzzy inference system was used for modelling the stability of asphalt concrete [18], [19], [20]. While the height and volume of voids were used as inputs, the compressive strength of concrete core was used as output. To observe the relationship between concrete properties and compressive strength correlation analysis was conducted and its results were shown in Table 3. From Table 3, it can be seen that the highest correlation coefficient between the height of the column and compressive strength was 0.916. However, the correlation coefficient between compressive strength and the ratio of absorption of water and volume of voids are nearly equal which values are -0.5. The other correlation coefficient between other properties and compressive strength are so poor. Because of the high correlation coefficient between compressive strength, height of column and volume of voids were used as inputs for fuzz logic. Fuzzy inference system was showed as schematically (Figure 3). Membership functions (MF) The degree of memberships was shown for height of structural

. Figure 1. Schematic showing of the column sample.

Figure 2. Core samples were taken from column.

cure was applied for 28 days in the morning, in the afternoon and in the evening as in the construction site. After 28 days of pouring the concrete, in order to take the core sample, the position of the column was changed into horizontal from vertical. 16 from the height and 8 from the width in total 128 core samples were taken. Diameter of the core was 100 mm and h/d=1.5 (Figures 1 and 2). The core samples were taken from the reference cube samples in the same way.

Determination of the physical properties of the samples All the samples taken were kept for 48 h in the laboratory and their air dry weight were measured with 0.1 g sensitivity. Afterwards these samples were kept in 202C water till their weight become constant and saturated to water. Saturated weights and their

5196

Int. J. Phys. Sci.

Table 2. The compressive strength, physical properties and height of the concrete core samples.

Core row numbers for core samples on the column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Reference

Natural unit volume weight 3 (g/cm ) 2.552 2.596 2.550 2.899 2.808 2.590 2.604 2.606 2.635 2.585 2.573 2.587 2.538 2.573 2.554 2.547 2.322

Unit volume weight for dry 3 air (g/cm ) 2.486 2.512 2.489 2.541 2.513 2.517 2.524 2.528 2.545 2.516 2.505 2.516 2.491 2.512 2.500 2.493 2.343

Saturated unit volume 3 weight (g/cm ) 2.574 2.618 2.570 2.655 2.608 2.614 2.625 2.627 2.658 2.606 2.594 2.609 2.557 2.592 2.573 2.566 2.368

Amount of the water absorption (%) 3.514 4.151 3.245 4.491 3.760 3.830 3.979 3.890 4.409 3.582 3.546 3.662 2.648 3.179 2.892 2.893 1.074

Volume of voids 3 (cm ) 8.030 9.365 7.437 10.213 8.595 8.744 9.092 8.912 10.000 8.229 8.151 8.386 6.177 7.379 6.717 6.713 2.518

Height (cm) 235 224 213 191 180 147 136 114 103 92 70 59 48 35 23 11 -

Compressive strength (Mpa) 37.30 38.40 37.70 38.60 39.80 40.20 39.60 41.00 40.80 39.90 40.30 41.00 42.00 41.90 42.50 43.00 36.95

Table 3. Result of correlation analysis and correlation coefficients between some physical properties.

Correlation coefficient Compressive strength (Mpa) Natural unit volume weight (g/cm3) Unit volume weight for dry air (g/cm3) Saturated unit volume weight (g/cm3) Amount of the water absorption (%) Volume of voids (cm3) Height (cm)

Compressive Strength (Mpa) 1.00 -0.273 0.001 -0.219 -0.504 -0.491 0.916

Saturated Natural unit Unit volume volume weight weight for dry air unit volume (g/cm3) (g/cm3) weight (g/cm3) -0.273 0.001 -0.219 1.00 0.605 0.638 0.605 1.00 0.966 0.638 0.966 1.00 0.621 0.813 0.935 0.628 0.825 0.942 -0.327 0.020 -0.211

Amount of the water absorption (%) -0.504 0.621 0.813 0.935 1.00 1.000 -0.508

Volume of voids (cm3) -0.491 0.628 0.825 0.942 1.000 1.00 -0.492

Height (cm) 0.916 -0.327 0.020 -0.211 -0.508 -0.492 1.00

Figure 3. Fuzzy inference system for the compressive strength of concrete.

Uzunolu et al.

5197

Figure 4. Membership functions for height of structural element.

Figure 5. Membership functions for volume of voids.

Figure 6. Membership functions for compressive strength.

element, volume of voids and compressive strength (Figures 4, 5 and 6). On the membership function graphics, y axis value is maximum (=1) and minimum (=0) for height of structural element,volume of voids and compressive strength. There are Height1, Height2, Height3 and Height4 on the membership functions graphics for height. The meanings of the symbols were explained linguistically, Height1 represents minimum height, Height2 represents middle height, Height3 represents high

and Height4 represents the highest. There are Vo1, Vo2, Vo3, and Vo4 on the membership functions graphics for volume of voids. The meanings of the symbols were explained linguistically; Vo1 represents minimum volume of voids, Vo2 represents middle volume of voids, Vo3 represents high volume of voids, and Vo4 represents the highest volume of voids. There are Comp1, Comp2, Comp3 and Comp4 on the membership functions graphics for compressive strength of the concrete. The meanings of the symbols

5198

Int. J. Phys. Sci.

Figure 7. A part of the rule viewer used for modelling the compressive strength of the concrete core samples.

were explained linguistically such as; Comp1 represents minimum compressive strength, Comp2 represents middle compressive strength, Comp3 represents high compressive strength, and Comp4 represents the highest compressive strength.

RESULTS AND DISCUSSION In this study, core samples were taken from structure elements such as column, reinforced wall in order to make reinforcement or restoration. The effect of the element height and volume of voids of these samples to the compressive strength were investigated. The results are stated below. The average compressive strength of the referenced samples was 36.95 Mpa and the compressive strength of other samples was ranging between 37.3 and 43.0 Mpa. While the height increases, compressive strength of concrete increases as well. This increase ratio according to the reference sample is; 0.95% for 0.15 m, 3.38% for 0.5 m, 8.53% for 1.0 m, 8.25% for 1.5 m, 13.13%, for 2 m, 15.83% for 2.35 m and 16.37% for 2.39 m. It is determined that; natural unit volume weight of core 3 samples change between 2.54 and 2.90 g/cm , dry unit 3 volume weight between 2.49 and 2.54 g/cm , saturated 3 unit volume weight between 2.56 and 2.66 g/cm , absorption amount between 2.65 and 4.49, volume of 3 voids between 6.18 and 10.21 cm . The correlation coefficient between the compressive strength and saturated unit volume weight of the core sample was 0.273, it was 0.01 at dry unit volume weight, 0.219 at saturated unit volume weight, -0.504 at water absorption

Rule editors, rule viewer and surface viewer Fuzzy sets and fuzzy operators are the subjects and verbs of fuzzy logic. These if-then rule statements are used to formulate the conditional statements that comprise fuzzy logic. A fuzzy if-then rule could be assumed for this study as; If height is Height1 and volume of voids is Vo4 then compressive strength is Comp1. Where, height and volume of voids are linguistic values defined for fuzzy sets on the ranges (universes of discourse) Height and Volume of voids. The if-part of the rule height and volume of voids are called antecedent or premise, while the then part of the rule Compressive strength is called the consequent or conclusion. The rule editor was edited the list of the rules that defined the behaviour of the system. Figure 6 shows a part of formed rules in rule editor. Which rules are active or how individual membership function shapes influence the results could be seen on the viewers which are seen in Figures 7 and 8. Also, the surface viewer or surface map was used to display the dependency of the compressive strength on the height and volume of voids. The surface viewer or surface map shows the behaviour of the entire system in Figure 9. The relationship between prediction values with fuzzy logic and experimental results for the compressive strength of concrete core samples was shown in Figure 10.

Uzunolu et al.

5199

Figure 8. A part of the formed rule editor.

Figure 9. The surface viewer or surface map for the entire system.

amount, -0.491 at volume of voids and 0.916 at height of the concrete. According to this situation, the relationship between height and the compressive strength is quite

strong, on the other hand, the relationship between compressive strength, water absorption amount and volume of voids is nearly same or middle level but

5200

Int. J. Phys. Sci.

Y = 0.86x + 5.55 R 2 = 0.99

Figure 10. Relationship between experimental results and prediction values with fuzzy logic.

negative. The relationship between unit volume weight and the compressive strength of the concrete is not significant. Particularly core samples are taken from buildings that would be reinforced in order to determine the compressive strength of concrete. In this study which aims to determine whether the compressive strength of the samples differed according to the height or not, the compressive strength of the samples according to height changed between 0.95 and 16.37%. For this reason, the compressive strength of the core samples can be used if they are converted into reference compressive strength according to the suggested coefficient below. Thus in order to convert the compressive strength of the concrete into reference value, the height must be measured while taking the core samples. Besides, the compressive 2 strength of the concrete could be predicted with R =0.99 using obtained in equation in Figure 10. It can be suggested that according to the volume of voids and height of the core samples, Fuzzy logic method can be used for predicting the compressive strength. In case of using different class, mixture design of concrete and different types and amounts of plasticizers, investigating the effects of height in structure elements such as column, reinforced wall on the compressive strength of hardened concrete can be useful.
REFERENCES ASTM C 42-90 (1994). Test for obtaining and testing drilled cores and

sawed beams of concrete, Annual Book of ASTM Standards. Arioglu E, Arioglu N (1998). Testing of concrete core samples and evaluations. Evrim Publisher. BS 1881, Part 120 (1983). Method for determination of the compressive strength of concrete cores, British Standards. Bartlett FM. MacGregor JG (1993). Effect of moisture condition on concrete core strengths. ACI Mater. J. 91(3):227236. Bartlett FM, MacGregor JG (1994). Effect of core diameter on concrete core strengths, ACI Mater. J. 91(5):460470. Bartlett FM (1997). Precision of in-place concrete strengths predicted using core strength correction factors obtained by weighed regression analysis. Struct Saf. 19(4):397-410. Bungey JH (1979). Determining concrete strength by using small diameter cores. Mag. Concr. Res. 31(107):9198. Erdoan TY (2003). Concrete. METU Publisher. pp.567-568. Haque MN, Al-Khaiat H (1997). Carbonation of concrete structures in hot dry coastal regions. Cement and Concrete Comp. 19(2):123-129. Jafri YZ, Sami M, Waseem A, Murtaza G, Akbar S (2012a). Stochastic approaches for time series forecasting of rate of dust fall: a case study of North West of Balochistan, Pakistan. IJPS 7(4):676686. Jafri YZ, Kamal L, Raza SM, Shah SH, Sami M, Waseem A (2012b). Chaotic Time Series Prediction and Mackey-Glass Simulation with Fuzzy Logic. IJPS 7(17):25962606. Jafri YZ, Kamal L, Waseem A, Raza SM (2012c) Rule Based Fuzzy Logic (FL) Time Series Prediction. IJPS 7(11):18621873. Jafri YZ, Waseem A, Kamal L, Raza SM, Sami M (2012d). Recent Trends on Artificial Neural Networks for Prediction of Wind Energy. SRE 7(14):1521-1526. Khayat KH, Manai K, Trudel A (1997). In-situ mechanical properties of wall elements cast using self-consolidating concrete. ACI Mater. J. 94(6):491500. Miao B, Aitcin PC, Cook WD, Mitchell D (1993). Influence of concrete strength on in situ properties of large columns. ACI Mater. J. 90(3):214219. Mindess S, Young JF (1981). Concrete. Prentice-Hall. Inc. New Jersey. Neville A. M. 1996. Properties of Concrete. 4th Ed., ELBS, Addison Wesley Longman Ltd., Harlow, England, ISBN 0582 279380. Seong-Tae Y, Eun-Ik Y, Joong-Cheol C (2006). Effect of specimen

Uzunolu et al.

5201

sizes, specimen shapes, and placement directions on compressive strength of concrete. Nuc. Eng. Des. 236:115-127. Price WF, Hynes JP (1996). In-situ strength testing of high strength concrete. Mag. Concr. Res. 48(176):189197. Sullivan PJE (1991). Testing and evaluation of concrete strength in structures. ACI Mater. J. 88(5):530535. Tay DCK. Tam CT (1996). In situ investigation of the strength of deteriorated concrete. Constr. Build. Mater. 10(1):17-26. Troxell GE, Davis HE, Kelly JW (1968). Composition and properties of concrete. McGraw-Hill Book Company.

TS EN 12504-1. (2002). Testing concrete in structures-part 1: cored specimens-taking, examining and testing in compression. Turkish Standards Institute. pp. 1-5. Tuncana M, Arioza O, Ramyarb K, Karasuc B (2008). Assessing concrete strength by means of small diameter cores. Constr. Build. Mater. 22(5): 981-988.