Cjce-2023-0466 Proof Hi

Canadian Journal of Civil Engineering
DEVELOPMENT OF NUMERICAL MODELS TO PREDICT THE

COMPRESSIVE STRENGTH OF HSCs BASED ON MATURITY
METHOD
Journal: Canadian Journal of Civil Engineering
Manuscript ID cjce-2023-0466
Manuscript Type: Article
Date Submitted by the

20-Nov-2023
Author:
Complete List of Authors: Demir, Tuba; Firat University, Civil Engineering Department
ALYAMAÇ, Kürşat; Firat University, Civil Engineering Department
Is the manuscript for

Dr
consideration in a Special Not applicable (regular submission)

Issue or Collection?:
High strength concrete, compressive strength, maturity, silica fume,

Keyword:
aft
marble powder
© The Author(s) or their Institution(s)

Page 1 of 38 Canadian Journal of Civil Engineering
1 DEVELOPMENT OF NUMERICAL MODELS TO PREDICT THE COMPRESSIVE
2 STRENGTH OF HSCs BASED ON MATURITY METHOD
3 Tuba DEMIR1*, Kursat Esat ALYAMAC2
4 1Civil Engineering Department, Technology Faculty, Firat University, 23119 Elazig, Turkey
5 2Civil Engineering Department, Engineering Faculty, Firat University, 23119 Elazig, Turkey
6 Corresponding Author: Tuba DEMIR
7 E-mail: [email protected]
8 Abstract
9 This study investigated the quickly determine of in-situ compressive strength of high-strength concretes
10 (HSCs). The experimental tests included existing maturity methods. In this direction, 9 concrete series
11 with different concrete mixing ratios with a binder dosage of 400-450-500 kg/m3 were prepared. The
12 data obtained from the experimental test results were analyzed and a separate numerical models were
Dr
13 suggested for each maturity function applied. When the experimental values and predicted values were
14 compared, it is seen that the absolute relative deviation (ARD) values of the developed models are below
aft
15 10%. When the concrete compressive strength estimation results of the maturity methods were
16 examined, it was determined that the Arrhenius method gave higher accuracy results than the Nurse-
17 Saul method. It is envisaged that the study will contribute to the literature by determining the conditions
18 under which each of the models obtained can be used effectively in practice. In addition it will allow the
19 project duration to be completed at the desired time.
20 Keywords: High strength concrete, compressive strength, maturity methods, numerical models.
21 Authors Information:
22 Corresponding Address : Tuba Demir (corresponding author).
23 Affiliations: 1-Civil Engineering Department, Technology Faculty, Firat University, 23119 Elazig,
24 Turkey, [email protected]
25 Affiliations: 2-Civil Engineering Department, Engineering Faculty, Firat University, 23119Elazig,
27

Canadian Journal of Civil Engineering Page 2 of 38
28 1. Introduction
29 In the concrete industry, determining the compressive strength at the construction site with non-
30 destructive tests is of great importance in terms of knowing the project durations. For example, in order
31 to remove the moulding as soon as possible after pouring the concrete in place, it is of great benefit to
32 take advantage of knowing the strength gain time of the concrete depending on the temperature increase
33 in the concrete. The maturity method, which is one of the non-destructive testing methods, is also a
34 method that helps the progress of constructions to be made safely. This method provides a simple
35 approach to the determination of concrete compressive strength [1], [2]. The maturity method is related
36 to the temperature sensitivity of early strength development and there is no single maturity function
37 applicable to all concrete mixes [3]. The applicable maturity function for a given concrete can be
38 obtained by measuring the variation of the rate constant with the curing temperature [4].
39 Frequently used functions to calculate the maturity index of concrete from the measured
40 temperature history are described in ASTM C1074 [5]. ASTM C1074 includes two different maturity
Dr
41 functions: The first is the Nurse-Saul equation, which estimates the compressive strength of concrete
aft
42 depending on the temperature change [6]. The second one is the Arrhenius (Freiesleben Hansen and
43 Pedersen) equation developed depending on the activation energy [7]. Apart from these equations,
44 different maturity equations have also been proposed, but they are not explained in detail due to the lack
45 of in-depth analysis of their reliability and limitations [8].
46 Although the maturity method is generally used in conventional concretes, no study has been
47 found in the literature regarding its use in HSCs. In this study, the applicability of maturity functions in
48 determining the compressive strength of HSCs was investigated and numerical models were developed
49 to predict concrete compressive strength. HSCs are special types of concrete with low water/cement
50 (W/C) ratio, extremely durable and advanced mechanical properties [9], [10]. Many HSC applications
51 in the United States, Germany, Canada, France and Turkey have demonstrated the advantages of this
52 concrete type [11]–[13]. However, one of the most important limiting conditions when constructing
53 these structures is the duration of the project. It is very important that the project is completed in the
54 targeted time before the start of production. When the project period is exceeded, significant additional
55 costs come and financial problems are experienced. One of the most important issues affecting this

56 period is the duration of the concrete to gain strength. The speed of the project depends on the
57 demoulding removal and moulding advancement speed. The parameter that directly affects this speed is
58 the concrete compressive strength. Methods for determining the compressive strength of concrete are
59 explained in the standards of many countries such as European Standard, ACI and TSE [14,15]. These
60 methods are destructive (DT) and non-destructive (NDT) test methods [16]–[18]. Evaluation with the
61 coring method in the DT method is quite expensive and time consuming. It is also difficult and
62 impractical in some cases [19]. The application of the NDT method provides many conveniences in
63 terms of time, cost and labor. The NDT method provides the opportunity to evaluate the durability,
64 homogeneity and internal structure of the concrete as well as the strength property of the concrete [20]–
65 [22].
66 Concrete is one of the most widely used building materials in the world [23]. Despite this
67 widespread use, there are great difficulties in accurately predicting the strength of concrete at various
68 stages [24], [25] . Accurate estimation of concrete strength is of great importance in determining
Dr
69 formwork removal time and construction project durations. Determining these periods in the early stages
aft
70 of construction is of great importance in preventing various negative effects on engineering projects.
71 Real-time evaluation of concrete compressive strength by maturity method, which is one of the non-
72 destructive test methods, is of great importance for the construction industry [26]. In recent years, many
73 studies have recommended the use of the maturity method to evaluate the early age strength
74 development of concrete. In the maturity method, the effect of time and temperature on the strength of
75 concrete is taken into account. This method is widely used to determine the optimal demoulding time
76 for concrete. The maturity method provides a simple and applicable approach to evaluate the
77 compressive strength of concrete [27]. Although this method is currently used in conventional concretes,
78 no study has been found in the literature regarding its application in HSCs .
79 The aim of this study is to develop numerical models that predict the compressive strength of
80 concrete using Nurse-Saul and Arrhenius (NS-Arr) maturity functions for HSCs. Accordingly, 9
81 concrete series with different concrete mixing ratios with cement dosage of 400-450-500 kg/m3 were
82 prepared. In the prepared concrete mixtures, silica fume was used by replacing the cement at 5-10-15%
83 by volume, and marble powder by replacing 8-10-12% by volume with fine aggregate. While the

84 produced samples were kept in the standard water cure, the temperature values were measured every 5
85 minutes at the end of the 1st, 3rd, 7th, 14th and 28th days to be used in the maturity test and the results were
86 transferred to the computer. A multi-channel temperature measuring device was used for temperature
87 measurement. Compressive strength test was also applied in the same time periods. The data obtained
88 from the test results were analyzed and a separate numerical model was developed for each maturity
89 function applied. As a result, by estimating the compressive strength of concrete with high accuracy
90 with the obtained numerical models, it was ensured that the production times were determined during
91 the design phase and accurate determinations were made during the manufacturing phase in special
92 projects using HSC. In Figure 1, the experimental procedure is presented schematically.
93
Dr
aft
94
95 Figure 1. Experiment Procedure
96
97 2. Material Method
98 Within the scope of the study, CEM I 42.5 R portland cement produced in accordance with TS
99 EN 197-1 standards and supplied from the Elazig Seza cement factory was used as the binder in the
100 production of HSC [28]. Andesite aggregate, silica fume from Antalya Eti metallurgy and marble
101 powder from Elazig organized industrial zone were used. The physical and chemical properties of the
102 HSC components are presented in Table 1.
103
104 Table 1. Physical and chemical properties of using materials (%).

Cement Silica Fume Marble Powder

Chemical Properties
(C) (SF) (MP)
CaO 63.18 0.40 40.45
SiO2 18.61 94.10 28.35
Fe2O3 3.7 1.50 9.70
Al2O3 4.81 0.90 0.17
SiO3 2.91 94.10 0.02
Na2O 0.39 0.40 0.05
K2O 0.61 0.90 0.01
MgO 1.82 0.10 16.25
Cl 0.01 - -
Insoluble residue 0.54 - -
Loss of ignition 3.42 - 4.84
Physical Properties
Specific surface cm2/g 3838 3920
Specific gravity g/cm3 3.13 2.20 2.71
Initial setting time (min) 135 - -
Final setting time(min) 215 - -
Total volume exp. (mm) 1 - -
105
106 The suitability of andesite aggregate for use in HSC was determined by performing aggregate
107 tests at the first stage (Table 2) [29]. Andesite aggregate is a material containing 52-63% quartz [30]. It
Dr
108 is suitable for use in high-strength concretes due to its dark-colored, non-water-absorbing, water-
109 dispersible and dense texture [31],[32].

aft
110
111 Table 2. Los Angeles experiment results
Number of Interval Load of Initial Weight after 500 Weight loss after
marbles class marbles Weight cycles 500 cycles
12 8-16 mm 5030 gr 5000 gr 4569.5 430.5

Aggregate Abrasion Resistance = LA = 8.61
112
113 The maximum aggregate particle diameter used in the experimental study was 16 mm, and the
114 aggregates were divided into 4 different groups as 0-2 mm, 2-4 mm, 4-8 mm, 8-16 mm (Table 3). The
115 granulometry curve of the aggregate is presented in Figure 2.
116 Table 3. Properties of the aggregate (Saturated dry surface, gr/cm3)
Mixing ratio
Aggregate size Andesite
(%)
0-2 mm 2.74 30
2-4 mm 2.71 20
4-8 mm 2.69 20
8-16 mm 2.69 30

117
118
119
120
100
Passing through the sieve ( %)
80
60
40
20 B16
A16
Mixture
0
0 2 4 6 8 10 12 14 16
Sieve opening (mm)

120
Dr
121 Figure 2. Aggregate granulometry curve

122
aft
123 After the aggregate tests, a pozzolanic activity test was performed for the mineral additive micro
124 silica (Table 4). Pozzolan is a material that does not have a binding property on its own, but gains binding
125 properties when it is finely ground and combined with calcium hydroxide in an aqueous medium [33],
126 [34]. Micro silica is one of the most commonly used mineral additives with fine grain properties in
127 materials science [35], [36]. In order to determine the adequate pozzolanic activity of micro silica, which
128 is a pozzolanic material, pozzolanic activity tests are carried out in accordance with ASTM C618 and
129 TS EN 196-2 standards [37]. In the same way, pozzolanic activity test was performed for silica fume
130 and the pozzolanic activity index was determined with the help of the equation given in Equation 1 [38].
131 The P.A.I values obtained as a result of the experiment provided the relevant standard.
132
A
133 Pozzolanic Activity Index (P.A.I) = ( B )*100 (1)
134 A : Average compressive strength of mortar samples with pozzolan

135 B : Average compressive strength of control mortar samples
136
137 Table 4. Pozzolanic activity test results
Control Sample (CS) Pozzolan Sample (PS)

fc(7) fc(28) fc(7) fc(28)

CS-1 27.6 40.6 PS-1 29.3 39.2
CS-2 28.1 37.1 PS-2 27.2 45.5
CS-3 25.8 39.4 PS-3 25.9 46.4
CS-ORT 27.2 39.0 PS-ORT 27.5 43.7
138
139 The marble powder used in the concrete mixture was used to create a filling effect. In order to
140 determine the usage rate of marble powder in concrete, many trial mixtures were poured and the
141 optimum usage rate was determined [39], [40].
142
143 2.1. Preparation of Test Samples
144
145 Within the scope of the study, a wide range of trial mixes were poured in the first stage. As a
146 result of these castings, 9 reference concrete mixtures with W/C ratios of 0.20-0.25-0.30 and binder
147 material (cement+micro silica) dosage of 400-450-500 kg/m3 were prepared (Table 5). In the series
148 names in the table, the M value represents maturity and HSC400 represents binder material dosage.
Dr
149 Micro silica cement and marble powder were used as mineral additives in the prepared concrete mixtures
150 by volume substituting with fine aggregate. The cement was replaced with micro silica. Marble powder
aft
151 was replaced with aggregate. The total amount of binder was kept constant. In all mixtures, CHRYSO
152 Fluid 518 hyperplasticizer was used as a chemical additive at the rate of 3% of the cement amount. Many
153 trial mixtures have been made in concrete mixtures. In these trial mixtures, different ratios of chemical
154 additives were used and the additive ratio that ensures appropriate workability was determined. After
155 the trial mixtures made in this way, it was seen that the most suitable workability for HSCs was 3%.
156 In addition since the W/C ratio is low in high-strength concrete, many trial mixtures were made and the
157 mixtures were repeated until appropriate workability was achieved. In this way, S2 and S3 slump classes
158 were obtained [39] .
159 Table 5. Mixture amounts of M-HSC (kg/m3)

Fine Fine Medium Coarse
Micro Marble
Serial Name Cement Water Aggregate Aggregate Aggregate Aggregate
Silica Powder
(0-2) mm (2-4) mm (4-8) mm (8-16) mm
M-HSC400-1 400 80 28.1 103 559 368 406 609
M-HSC400-2 400 100 14.1 101 550 362 400 599
M-HSC400-3 400 120 14.1 79 547 360 389 583
M-HSC450-1 450 90 47.4 78 544 358 387 580
M-HSC450-2 450 113 47.4 114 504 332 374 562
M-HSC450-3 450 135 31.6 74 516 340 367 550
M-HSC500-1 500 100 17.6 116 513 338 381 572
M-HSC500-2 500 125 35.1 95 517 341 376 564
M-HSC500-3 500 150 52.7 87 473 312 344 516

160
161 2.2. Application of Maturity Tests
162
163 This study was carried out to develop numerical models that predict concrete compressive
164 strength with high accuracy by using Nurse-Saul and Arrhenius maturity functions separately [41]. With
165 these maturity functions, the compressive strength of early age concretes can be estimated and the
166 demoulding time can be determined [42]. Thus, advantages are provided in terms of project duration,
167 time and cost [43]. In the study, firstly, the Nurse-Saul method given in Equation 2, in which the
168 compressive strength of the concrete is estimated by looking at the temperature history, was used.
169
t
170 M = ∑0(T ― T0) ∗ ∆t (2)
171
172 Here;
173 M = Maturity index (°C-hours or °C-days)

Dr
174 T = Average concrete temperature (°C)

aft
175 T0 = Reference temperature
176 t = elapsed time (day or hour)
177 ∆t = Time range (day or hour)
178
179 Copeland et al. in 1960 suggested that the effects of cement hydration on the strength gain rate
180 of concrete can be described by the Arrhenius equation. In 1977, Freisleben Hansen and Pedersen
181 proposed the following equation for equivalent age based on the Arrhenius equation (Equation 3) [7].
182 The application and numerical values related to the experiment are explained in ASTM C1074 [5].
183
184
𝐄 𝟏 𝟏
t ― ( ― )
185 te = ∑0𝐞 𝐑 𝟐𝟕𝟑 + 𝐓 𝟐𝟕𝟑 + 𝐓𝐫 ∆t (3)
186
187 Here;
188 te = Equivalent age at reference temperature

189 E = Apparent activation energy
190 R = Universal gas constant
191 T = Average absolute temperature of concrete
192 Tr = Reference temperature
193 A multi-channel temperature measuring device was used to determine the concrete temperatures
194 in the maturity calculations. The device was set up to take measurements every 5 minutes after the
195 concrete was poured and placed in the mold and placed in the samples (Figure 3).
196
197
Dr
aft
198
199
200 Figure 3. Placement of the temperature measuring device
201
202 2.3. Measurements on Concrete Samples
203 Temperatures were measured from the beginning of the concrete casting and the temperature results
204 taken every 5 minutes at the end of the 1st, 3rd, 7th, 14th and 28th days within the scope of the C1074
205 standard were transferred to the computer. The time-dependent graph of these temperatures is given in
206 Figure 4. Figure 4 shows the time-dependent variation of the temperatures forty-eight, ninety-six hours
207 after the concrete is placed in the mold, and the average temperatures of one hundred and ninety-two,
208 three hundred and sixty-six hours after it is placed in the mold.

30
HSC400-25
HSC400-30
25
Temperature (°C)
20
15
10
0 10000 20000 30000 40000 50000
209 Time (min)

210 (a)
211
30
HSC450-25
Dr
HSC450-30
25
aft
Temperature (°C)
20
15
10
0 10000 20000 30000 40000 50000
212 Time (min)

213 (b)
214

30
HSC500-25
HSC500-25
25
Temperature (°C)
20
15
10
0 10000 20000 30000 40000 50000
215 Time (min)

216
Dr
(c)
217
218 Figure 4. (a) Temperature values of 400-dose series (b) Temperature values of 450-dose series (c) Temperature
aft
219 values of 500-dose series

220
221 When Figure 3 is examined, the temperature values of the series in which the W/C ratio is 0.25
222 and 0.30 are given. Since there is not enough water in the environment as specified in the C1074 standard
223 in the series with a W/C ratio of 0.20 and proper hydration has not occurred, the data of these series
224 were not evaluated.
225 3. Evaluation of Experimental Results
226 3.1. Concrete Temperature Values and Evaluation
227 Concrete temperatures were obtained using a multi-channel temperature measuring device. This
228 temperature measurement started from the placement of the concrete in the mold (Figure 5.) When
229 Figure 5 is examined, it is seen that the temperature change varies according to the usage rate of the
230 materials in the concrete composition. For example, it is clearly seen that the temperature value of the
231 series with a binder dosage of 500 and a W/C ratio of 25% is higher than the temperature value of a
232 series with a binder dosage of 400 and a W/C ratio of 25%.

233 The temperature values in Figure 3 and the maturity functions in Equation 2 and Equation 3
234 were calculated separately and the Nurse-Saul-time and Arrhenius-time graphs given in Figure 5 were
235 obtained.
236
8000
Nurse-Saul (°C-min)
7000
6000 HSC500-30
5000 HSC500-25
4000 HSC450-30
3000 HSC450-25
2000 HSC400-30
1000 HSC400-25
0
0
1255
2510
3765
5020
6275
7530
8785
10040
11295
12550
13805
15060
16315
17570
18825
20080
21335
22590
23845
25100
26355
27610
28865
30120
31375
32630
33885
35140
36395
37650
38905
40160
41415
Time (min)
237
238 (a)
239
8000
7000
Arrhenius (min)
6000 HSC500-30
Dr
5000 HSC500-25
4000 HSC450-30
3000 HSC450-25
aft
2000 HSC400-30
1000 HSC400-25
0
0
1235
2470
3705
4940
6175
7410
8645
9880
11115
12350
13585
14820
16055
17290
18525
19760
20995
22230
23465
24700
25935
27170
28405
29640
30875
32110
33345
34580
35815
37050
38285
39520
40755
Time (min)
240
241 (b)
242
243 Figure 5. (a) Nurse-Saul values-Time (b) Arrhenius values-Time comparisons
244 In the Figure 5 showing the time-dependent variation of the maturity function, it is seen that the
245 binder dosage and maturity values are directly proportional. As the dosage value increases, the hydration
246 increases, so the maturity value also increases. When evaluated in terms of W/C ratio, the maturity value
247 was found to be higher in the 400 and 450-dose series in which the W/C ratio was 30%. This indicates
248 that sufficient water for hydration is found in series with a 30% W/C ratio. In the 500-dose series, the
249 maturity value of the series in which the W/C ratio was 25% was high.
250 The relationship between maturity values calculated by Nurse-Saul and Arrhenius equations and
251 compressive strength is shown in Figure 6.
252

Nurse-Saul HSC400-25 HSC400-30 HSC450-25

HSC450-30 HSC500-25 HSC500-30
Compressive Strength (MPa)
85
75
65
55
45
35
400 2400 4400 6400 8400 10400
Maturity (°C-hour)
253
254 (a)
Arrhenius
HSC400-25 HSC400-30 HSC450-25
HSC450-30 HSC500-25 HSC500-30
90
85
80
75
70
65
60
55
50
45
40
1 3 Maturity (°C-hour) 5 7
Dr
255
256 (b)
aft
257 Figure 6. Comparison of (a) Compressive strength-Nurse-Saul values (b) Compressive strength-Arrhenius
258 maturity values
259
260 In Figure 6, it is seen that there is a linear relationship between maturity and compressive
261 strength depending on the change in concrete temperature. In accordance with the C1074 standard, the
262 compressive strength value of the concrete increases with the increase of the maturity value. The
263 numerical models estimating the compressive strength according to the NS and Arr maturity methods
264 for each dosage and W/C value from the compressive strength maturity graph are given in Table 6.
265 While determining the models, maturity value was determined as input and compressive strength of
266 concrete value was determined as output.
267
268 Table 6. Numerical models developed with NS and Arr maturity methods
Serial Name NS Arr
M-HSC400-25 fc = 0.0027M + 48.131 fc = 6.7943M + 33.573

M-HSC400-30 fc = 0.0027M + 45.805 fc = 7.0137M + 30.937
M-HSC450-25 fc = 0.0027M + 51.143 fc = 6.8513M + 36.05
M-HSC450-30 fc = 0.0028M + 46.995 fc = 7.0219M + 31.914
M-HSC500-25 fc = 0.0032M + 54.705 fc = 6.9625M + 37.284
M-HSC500-30 fc = 0.0029M + 52.118 fc = 7.4902M + 35.333

269
270 In Figure 7, the estimation results and ARD values obtained by the numerical models developed
271 with the experimentally calculated compressive strength values are given. When the experimental values
272 and predicted values are compared, it is seen that the ARD values of the developed models are below
273 10%. It was noteworthy that all these values were below 5%, especially for early age strength. When
274 the ARD values of the Arrhenius maturity function were examined, it was noted that it was the maturity
275 function that predicted fc with the highest accuracy. Especially in fc-1 daily strength values, highly
276 accurate predictions were made and ARD values were below 1% for almost all other days.
277 Although the ARD results estimated by the Nurse-Saul equation are higher than the ARD results
278 obtained by the Arrhenius equation, the compressive strength values obtained with this function also
279 have high accuracy.
ARD value-Nurse-Saul ARD value-Arrhenius

25 10
M-HSC400-25 M-HSC400-25
Dr
15 5 M-HSC400-30
M-HSC400-30
5 M-HSC450-25 M-HSC450-25
0 M-HSC450-30
-5 M-HSC450-30
aft
M-HSC500-25 M-HSC500-25
-15 -5
M-HSC500-30 M-HSC500-30
-25 -10
280
M -HSC400-25 M -HSC400-30 M -HSC450-25 M -HSC450-30 M -HSC500-25 M -HSC500-30
Days fc fc-NS fc-Arr fc fc-NS fc-Arr fc fc-NS fc-Arr fc fc-NS fc-Arr fc fc-NS fc-Arr fc fc-NS fc-Arr
1. 50 48.4 49.6 47 48.4 47.2 54 53.8 52.4 48 49.7 48.6 61 58.4 57.3 57 55.2 53.5
3. 57 52.9 56.1 55 50.6 54.0 59 56.0 59.2 56 52.0 55.2 64 62.2 67.7 61 57.4 60.6
7. 63 57.2 62.5 61 54.8 60.6 68 60.2 65.6 63 56.7 61.9 73 67.2 74.2 67 61.9 67.6
14. 67 64.2 68.6 65 61.9 66.8 69 67.3 71.8 67 63.7 68.0 81 75.5 80.4 76 69.5 75.3
28. 75 78.4 74.5 73 76.0 72.1 78 81.5 77.6 74 78.4 74.1 88 92.6 86.8 80 84.8 81.2
281
282 Figure 7. Experimental and predicted results
283
284 Conclusions
285 In this study, numerical models were developed to predict the compressive strength of HSCs using
286 NS and Arr maturity functions. In particular, to verify the validity of the maturity functions proposed so
287 far, compressive strength-Nurse-Saul maturity function and compressive strength-Arrhenius maturity
288 functions were compared and their suitability was determined. As a result of the study, the following
289 conclusions were drawn:

290 1- The strength-maturity curve in each series of concrete mixtures followed the literature. As the
291 strength value increased, the maturity value also tended to increase.
292 2- It has been determined that the maturity method can be used for HSCs, and it has been observed
293 that the Arr method gives higher accuracy results in estimating the compressive strength of
294 HSCs. The ARD values of the compressive strength values obtained with the Arr method were
295 lower.
296 3- The average relative error percentage in estimating the compressive strengths of concrete varies
297 between 0.1% and 10.1%. Mixtures producing the highest errors were generally found in the
298 calculation made by the Nurse-Saul method. The compressive strength values obtained by the
299 Arhenius method were closer to the experimental results.
300 4- In series with low W/C ratio, strength losses occurred due to insufficient hydration. Therefore,
301 it is recommended to be more careful in the estimations to be made with the maturity method
302 as the W/C ratio decreases in practice.

Dr
303 5- By using the maturity method and determining the strength gaining time of the concrete, the
aft
304 demoulding times can be determined. Thus, the problem of getting molds early is eliminated
305 and an advantage is provided in determining the project durations.
306 6- In the current maturity functions, the concrete temperature is taken according to the standards.
307 Concrete compressive strength can also be determined more accurately and with higher
308 accuracy by obtaining the temperatures with high accuracy by using a temperature measuring
309 device in concrete.
310 7- Verification of proposed modifications for in situ concrete strength estimation should also be
311 done, especially for concretes made with different types of cement and additives.
312
313 As a result, the places where HSCs are used in practice are special productions. Therefore, the duration
314 of the project is of great importance in such productions. The most important parameter for the
315 construction to progress in accordance with the project period is the molding time. Since this period
316 cannot be calculated correctly, the maturity method has emerged as a sensitive, fast and practical
317 method. The use of this method will ensure that the concrete compressive strength is correctly estimated

318 and the construction will continue in a healthy way in terms of strength and will proceed without loss of
319 time in the project-work plan.
320
321 Acknowledgments
322 This study is supported by the Scientific Research Project Fund of Fırat University under the
323 project number MF.21.51
324
325 References
326 [1] I. Belykh, V. Sopov, L. Butska, L. Pershina, and O. Makarenko, “Predicting the strength and
327 maturity of hardening concrete,” in MATEC Web of Conferences, 2018, vol. 230, p. 3001.
328 [2] F. A. Oluokun, E. G. Burdette, and J. H. Deatheridge, “Early-age concrete strength prediction
329 by maturity--another look,” Mater. J., vol. 87, no. 6, pp. 565–572, 1990.
330 [3] M. Soutsos, F. Kanavaris, and A. Hatzitheodorou, “Critical analysis of strength estimates from
Dr
331 maturity functions,” Case Stud. Constr. Mater., vol. 9, p. e00183, 2018.
332 [4] D. Miller, N.-M. Ho, and N. Talebian, “Monitoring of in-place strength in concrete structures
aft
333 using maturity method–An overview,” in Structures, 2022, vol. 44, pp. 1081–1104.
334 [5] C. ASTM, “1074-11.(2011).“,” Stand. Pract. Estim. Concr. Strength by Matur. Method, Am.
335 Soc. Test. Mater.
336 [6] J.-S. Jung and Y.-H. Cho, “Comparison between nurse-saul and arrhenius equations,” Can. J.
337 Civ. Eng., vol. 36, no. 3, pp. 514–518, 2009.
338 [7] D. A. Wood, “Re-establishing the merits of thermal maturity and petroleum generation multi-
339 dimensional modeling with an Arrhenius Equation using a single activation energy,” J. Earth
340 Sci., vol. 28, no. 5, pp. 804–834, 2017.
341 [8] G. Chengju, “Maturity of concrete: method for predicting early-stage strength,” Mater. J., vol.
342 86, no. 4, pp. 341–353, 1989.
343 [9] T. Demir, M. Ulucan, and K. E. Alyamac, “Determination of Early Age Strength of High
344 Strength Concretes Using RSM Method,” Fırat Univ. J. Eng. Sci., vol. 34, no. 1, pp. 105–114,
345 2022.

346 [10] Y. B. Guo, G. F. Gao, L. Jing, and V. P. W. Shim, “Response of high-strength concrete to
347 dynamic compressive loading,” Int. J. Impact Eng., vol. 108, pp. 114–135, 2017.
348 [11] T. Demir and K. E. Alyamaç, “Investigation of the Use of Marble Powder in Production of
349 High Strength Concretes,” Open J. Nano, vol. 7, no. 1, pp. 18–25, 2022.
350 [12] A. Kandiri, E. M. Golafshani, and A. Behnood, “Estimation of the compressive strength of
351 concretes containing ground granulated blast furnace slag using hybridized multi-objective
352 ANN and salp swarm algorithm,” Constr. Build. Mater., vol. 248, p. 118676, 2020.
353 [13] V. M. Sounthararajan and A. Sivakumar, “Effect of the lime content in marble powder for
354 producing high strength concrete,” ARPN J. Eng. Appl. Sci., vol. 8, no. 4, pp. 260–264, 2013.
355 [14] T. S. EN, “TS EN 12504-2,” Test. Concr. Struct. part-2 non-destructive testing, Determ.
356 rebound number, Stand., 2004.
357 [15] B. Standard, “Structural Use of Concrete: Code of Practice for Design and Construction, Part 1,
358 BS 8110.” British Standard Institution, UK, 1997.

Dr
359 [16] V. M. Malhotra, “Testing hardened concrete: nondestructive methods,” 1976.

aft
360 [17] R. L. Wankhade and A. B. Landage, “Non-destructive testing of concrete structures in Karad
361 region,” Procedia Eng., vol. 51, pp. 8–18, 2013.
362 [18] A. G. Davis et al., “Nondestructive test methods for evaluation of concrete in structures,” Am.
363 Concr. Institute, ACI, vol. 228, no. 4, 1998.
364 [19] J. Malek and M. Kaouther, “Destructive and non-destructive testing of concrete structures,”
365 Jordan J. Civ. Eng., vol. 8, no. 4, pp. 432–441, 2014.
366 [20] O. Tsioulou, A. Lampropoulos, and S. Paschalis, “Combined non-destructive testing (NDT)
367 method for the evaluation of the mechanical characteristics of ultra high performance fibre
368 reinforced concrete (UHPFRC),” Constr. Build. Mater., vol. 131, pp. 66–77, 2017.
369 [21] A. A. Shah and Y. Ribakov, “Non-destructive evaluation of concrete in damaged and
370 undamaged states,” Mater. Des., vol. 30, no. 9, pp. 3504–3511, 2009.
371 [22] T. Demir, “Development of Combined Methods to Estimate the in-Place Strength of High
372 Strength Concrete Using Non-Destructive Testing Methods,”,” Fırat University, 2022.
373 [23] N. Mylvaganam and Y. Elakneswaran, “A Systematic Review and Assessment of Concrete

374 Strength Prediction Models,” Case Stud. Constr. Mater., p. e01830, 2023.
375 [24] L. Wang, H. Zhou, J. Zhang, Z. Wang, L. Zhang, and M. L. Nehdi, “Prediction of concrete
376 strength considering thermal damage using a modified strength-maturity model,” Constr. Build.
377 Mater., vol. 400, p. 132779, 2023.
378 [25] Y. Murad, “Compressive strength prediction for concrete modified with nanomaterials,” Case
379 Stud. Constr. Mater., vol. 15, p. e00660, 2021.
380 [26] S. N. Bengal, L. S. Pammar, and C. B. Nayak, “Engineering application of organic materials
381 with concrete: A review,” Mater. Today Proc., vol. 56, pp. 581–586, 2022.
382 [27] M. Soutsos and F. Kanavaris, “Applicability of the Modified Nurse-Saul (MNS) maturity
383 function for estimating the effect of temperature on the compressive strength of GGBS
384 concretes,” Constr. Build. Mater., vol. 381, p. 131250, 2023.
385 [28] P. and C. C. Cement - Part 1: General Cements, Composition, “TS EN 197-1,” Turkey, 2012.
386 [29] T. S. I. TS, “TS EN 706-Aggregate for concretes.” Ankara, 1980.

Dr
387 [30] O. Soykan, Ö. Cengiz, and Ö. Cenk, “Investigation of the Usability of Slate and Andesite as
aft
388 Concrete Aggregate,” J. Suleyman Demirel Univ. Grad. Sch. Nat. Appl. Sci., vol. 19, no. 1,
389 2015.
390 [31] M. SÜMER and B. SÖYLER, “yüKSEK DAYANIMLI BETON ÜRETİMİNDE ÇİMENTO
391 ve SUPERAKIŞKANLAŞTIRICI BETON KATKI MADDELERİNİN ETKİNLiG i,” Sak.
392 Üniversitesi Fen Bilim. Enstitüsü Derg., vol. 6, no. 3, pp. 3–15, 2002.
393 [32] T. S. ENSTİTÜSÜ, “TS 706 EN 12620+ A1 Concrete Aggregates,” TSE, Ankara, Türkiye,
394 2009.
395 [33] A. Shayan and A. Xu, “Performance of glass powder as a pozzolanic material in concrete: A
396 field trial on concrete slabs,” Cem. Concr. Res., vol. 36, no. 3, pp. 457–468, 2006.
397 [34] N. Flores Medina, G. Barluenga, and F. Hernández-Olivares, “Combined effect of
398 Polypropylene fibers and Silica Fume to improve the durability of concrete with natural
399 Pozzolans blended cement,” Constr. Build. Mater., vol. 96, 2015.
400 [35] R. Yu, P. Spiesz, and H. J. H. Brouwers, “Development of an eco-friendly Ultra-High
401 Performance Concrete (UHPC) with efficient cement and mineral admixtures uses,” Cem.

402 Concr. Compos., vol. 55, 2015.
403 [36] M. Vigneshwari, K. Arunachalam, and A. Angayarkanni, “Replacement of silica fume with
404 thermally treated rice husk ash in Reactive Powder Concrete,” J. Clean. Prod., vol. 188, 2018.
405 [37] T. Standard, “TS EN 196–2,” Methods Test. Cem., vol. 2.
406 [38] requirements and conformity criteria Natural pozzolan (Trass) for use in cement and concrete -
407 Definitions, “TS 25,” Ankara,Turkey, 2015.
408 [39] T. DEMİR, M. ULUCAN, and K. E. ALYAMAÇ, “Yüksek Dayanımlı Betonların Erken Yaş
409 Dayanımının RSM Metodu Kullanılarak Belirlenmesi,” Fırat Üniversitesi Mühendislik Bilim.
410 Derg., vol. 34, no. 1, pp. 105–114.
411 [40] T. Demir, B. Demirel, and M. Öztürk, “An Evaluation of the Effect of Waste Aluminum
412 Sawdust on the Carbonation of Concrete,” Bitlis Eren Univ. J. Sci., vol. 11, no. 4, pp. 993–999.
413 [41] M. T. de Grazia, L. F. M. Sanchez, D. J. De Souza, L. Ismail, M. Noel, and S. Decarufel,
414 “Evaluation of electrical resistivity and maturity for estimating the early-age properties of pre-
Dr
415 packaged concrete,” Can. J. Civ. Eng., vol. 49, no. 7, pp. 1231–1238, 2022.
aft
416 [42] C. J. Willis and J. H. Rankin, “Demonstrating a linkage between construction industry maturity
417 and performance: a case study of Guyana and New Brunswick,” Can. J. Civ. Eng., vol. 39, no.
418 5, pp. 565–578, 2012.
419 [43] D. Breysse, “Nondestructive evaluation of concrete strength: An historical review and a new
420 perspective by combining NDT methods,” Constr. Build. Mater., vol. 33, pp. 139–163, 2012.
421

1 DEVELOPMENT OF NUMERICAL MODELS TO PREDICT THE COMPRESSIVE
2 STRENGTH OF HSCs BASED ON MATURITY METHOD
3 Tuba DEMIR1*, Kursat Esat ALYAMAC2
4 1Civil Engineering Department, Technology Faculty, Firat University, 23119 Elazig, Turkey
5 2Civil Engineering Department, Engineering Faculty, Firat University, 23119 Elazig, Turkey
6 Corresponding Author: Tuba DEMIR
7 E-mail: [email protected]
8 Abstract
9 This study investigated the quickly determine of in-situ compressive strength of high-strength concretes
10 (HSCs). The experimental tests included existing maturity methods. In this direction, 9 concrete series
11 with different concrete mixing ratios with a binder dosage of 400-450-500 kg/m3 were prepared. The
12 data obtained from the experimental test results were analyzed and a separate numerical models were
Dr
13 suggested for each maturity function applied. When the experimental values and predicted values were
14 compared, it is seen that the absolute relative deviation (ARD) values of the developed models are below
aft
15 10%. When the concrete compressive strength estimation results of the maturity methods were
16 examined, it was determined that the Arrhenius method gave higher accuracy results than the Nurse-
17 Saul method. It is envisaged that the study will contribute to the literature by determining the conditions
18 under which each of the models obtained can be used effectively in practice. In addition it will allow the
19 project duration to be completed at the desired time.
20 Keywords: High strength concrete, compressive strength, maturity methods, numerical models.
21
22 Authors Information:
23 Corresponding Address : Tuba Demir (corresponding author).
24 Affiliations: 1-Civil Engineering Department, Technology Faculty, Firat University, 23119 Elazig,
26 Affiliations: 2-Civil Engineering Department, Engineering Faculty, Firat University, 23119Elazig,

28
29 1. Introduction
30 In the concrete industry, determining the compressive strength at the construction site with non-
31 destructive tests is of great importance in terms of knowing the project durations. For example, in order
32 to remove the moulding as soon as possible after pouring the concrete in place, it is of great benefit to
33 take advantage of knowing the strength gain time of the concrete depending on the temperature increase
34 in the concrete. The maturity method, which is one of the non-destructive testing methods, is also a
35 method that helps the progress of constructions to be made safely. This method provides a simple
36 approach to the determination of concrete compressive strength [1], [2]. The maturity method is related
37 to the temperature sensitivity of early strength development and there is no single maturity function
38 applicable to all concrete mixes [3]. The applicable maturity function for a given concrete can be
39 obtained by measuring the variation of the rate constant with the curing temperature [4].
40 Frequently used functions to calculate the maturity index of concrete from the measured
Dr
41 temperature history are described in ASTM C1074 [5]. ASTM C1074 includes two different maturity
aft
42 functions: The first is the Nurse-Saul equation, which estimates the compressive strength of concrete
43 depending on the temperature change [6]. The second one is the Arrhenius (Freiesleben Hansen and
44 Pedersen) equation developed depending on the activation energy [7]. Apart from these equations,
45 different maturity equations have also been proposed, but they are not explained in detail due to the lack
46 of in-depth analysis of their reliability and limitations [8].
47 Although the maturity method is generally used in conventional concretes, no study has been
48 found in the literature regarding its use in HSCs. In this study, the applicability of maturity functions in
49 determining the compressive strength of HSCs was investigated and numerical models were developed
50 to predict concrete compressive strength. HSCs are special types of concrete with low water/cement
51 (W/C) ratio, extremely durable and advanced mechanical properties [9], [10]. Many HSC applications
52 in the United States, Germany, Canada, France and Turkey have demonstrated the advantages of this
53 concrete type [11]–[13]. However, one of the most important limiting conditions when constructing
54 these structures is the duration of the project. It is very important that the project is completed in the

55 targeted time before the start of production. When the project period is exceeded, significant additional
56 costs come and financial problems are experienced. One of the most important issues affecting this
57 period is the duration of the concrete to gain strength. The speed of the project depends on the
58 demoulding removal and moulding advancement speed. The parameter that directly affects this speed is
59 the concrete compressive strength. Methods for determining the compressive strength of concrete are
60 explained in the standards of many countries such as European Standard, ACI and TSE [14,15]. These
61 methods are destructive (DT) and non-destructive (NDT) test methods [16]–[18]. Evaluation with the
62 coring method in the DT method is quite expensive and time consuming. It is also difficult and
63 impractical in some cases [19]. The application of the NDT method provides many conveniences in
64 terms of time, cost and labor. The NDT method provides the opportunity to evaluate the durability,
65 homogeneity and internal structure of the concrete as well as the strength property of the concrete [20]–
66 [22].
67 Concrete is one of the most widely used building materials in the world [23]. Despite this
Dr
68 widespread use, there are great difficulties in accurately predicting the strength of concrete at various
aft
69 stages [24], [25] . Accurate estimation of concrete strength is of great importance in determining
70 formwork removal time and construction project durations. Determining these periods in the early stages
71 of construction is of great importance in preventing various negative effects on engineering projects.
72 Real-time evaluation of concrete compressive strength by maturity method, which is one of the non-
73 destructive test methods, is of great importance for the construction industry [26]. In recent years, many
74 studies have recommended the use of the maturity method to evaluate the early age strength
75 development of concrete. In the maturity method, the effect of time and temperature on the strength of
76 concrete is taken into account. This method is widely used to determine the optimal demoulding time
77 for concrete. The maturity method provides a simple and applicable approach to evaluate the
78 compressive strength of concrete [27]. Although this method is currently used in conventional concretes,
79 no study has been found in the literature regarding its application in HSCs .
80 The aim of this study is to develop numerical models that predict the compressive strength of
81 concrete using Nurse-Saul and Arrhenius (NS-Arr) maturity functions for HSCs. Accordingly, 9
82 concrete series with different concrete mixing ratios with cement dosage of 400-450-500 kg/m3 were

83 prepared. In the prepared concrete mixtures, silica fume was used by replacing the cement at 5-10-15%
84 by volume, and marble powder by replacing 8-10-12% by volume with fine aggregate. While the
85 produced samples were kept in the standard water cure, the temperature values were measured every 5
86 minutes at the end of the 1st, 3rd, 7th, 14th and 28th days to be used in the maturity test and the results were
87 transferred to the computer. A multi-channel temperature measuring device was used for temperature
88 measurement. Compressive strength test was also applied in the same time periods. The data obtained
89 from the test results were analyzed and a separate numerical model was developed for each maturity
90 function applied. As a result, by estimating the compressive strength of concrete with high accuracy
91 with the obtained numerical models, it was ensured that the production times were determined during
92 the design phase and accurate determinations were made during the manufacturing phase in special
93 projects using HSC. In Figure 1, the experimental procedure is presented schematically.
94
Dr
aft
95
96 Figure 1. Experiment Procedure
97
98 2. Material Method
99 Within the scope of the study, CEM I 42.5 R portland cement produced in accordance with TS
100 EN 197-1 standards and supplied from the Elazig Seza cement factory was used as the binder in the
101 production of HSC [28]. Andesite aggregate, silica fume from Antalya Eti metallurgy and marble
102 powder from Elazig organized industrial zone were used. The physical and chemical properties of the
103 HSC components are presented in Table 1.

104
105 Table 1. Physical and chemical properties of using materials (%).

Cement Silica Fume Marble Powder
Chemical Properties
(C) (SF) (MP)
CaO 63.18 0.40 40.45
SiO2 18.61 94.10 28.35
Fe2O3 3.7 1.50 9.70
Al2O3 4.81 0.90 0.17
SiO3 2.91 94.10 0.02
Na2O 0.39 0.40 0.05
K2O 0.61 0.90 0.01
MgO 1.82 0.10 16.25
Cl 0.01 - -
Insoluble residue 0.54 - -
Loss of ignition 3.42 - 4.84
Physical Properties
Specific surface cm2/g 3838 3920
Specific gravity g/cm3 3.13 2.20 2.71
Initial setting time (min) 135 - -
Final setting time(min) 215 - -
Total volume exp. (mm) 1 - -
106
107 The suitability of andesite aggregate for use in HSC was determined by performing aggregate
Dr
108 tests at the first stage (Table 2) [29]. Andesite aggregate is a material containing 52-63% quartz [30]. It
109 is suitable for use in high-strength concretes due to its dark-colored, non-water-absorbing, water-
aft
110 dispersible and dense texture [31],[32].
111
112 Table 2. Los Angeles experiment results
Number of Interval Load of Initial Weight after 500 Weight loss after
marbles class marbles Weight cycles 500 cycles
12 8-16 mm 5030 gr 5000 gr 4569.5 430.5

Aggregate Abrasion Resistance = LA = 8.61
113
114 The maximum aggregate particle diameter used in the experimental study was 16 mm, and the
115 aggregates were divided into 4 different groups as 0-2 mm, 2-4 mm, 4-8 mm, 8-16 mm (Table 3). The
116 granulometry curve of the aggregate is presented in Figure 2.
117 Table 3. Properties of the aggregate (Saturated dry surface, gr/cm3)
Mixing ratio
Aggregate size Andesite
(%)
0-2 mm 2.74 30

2-4 mm 2.71 20
4-8 mm 2.69 20
8-16 mm 2.69 30
118
119
120
120
100
Passing through the sieve ( %)
80
60
40
20 B16
A16
Mixture
0
Dr
0 2 4 6 8 10 12 14 16
Sieve opening (mm)

121
122
aft
Figure 2. Aggregate granulometry curve

123
124 After the aggregate tests, a pozzolanic activity test was performed for the mineral additive micro
125 silica (Table 4). Pozzolan is a material that does not have a binding property on its own, but gains binding
126 properties when it is finely ground and combined with calcium hydroxide in an aqueous medium [33],
127 [34]. Micro silica is one of the most commonly used mineral additives with fine grain properties in
128 materials science [35], [36]. In order to determine the adequate pozzolanic activity of micro silica, which
129 is a pozzolanic material, pozzolanic activity tests are carried out in accordance with ASTM C618 and
130 TS EN 196-2 standards [37]. In the same way, pozzolanic activity test was performed for silica fume
131 and the pozzolanic activity index was determined with the help of the equation given in Equation 1 [38].
132 The P.A.I values obtained as a result of the experiment provided the relevant standard.
133
A
134 Pozzolanic Activity Index (P.A.I) = ( B )*100 (1)
135 A : Average compressive strength of mortar samples with pozzolan

136 B : Average compressive strength of control mortar samples

137
138 Table 4. Pozzolanic activity test results
Control Sample (CS) Pozzolan Sample (PS)
fc(7) fc(28) fc(7) fc(28)
CS-1 27.6 40.6 PS-1 29.3 39.2
CS-2 28.1 37.1 PS-2 27.2 45.5
CS-3 25.8 39.4 PS-3 25.9 46.4
CS-ORT 27.2 39.0 PS-ORT 27.5 43.7
139
140 The marble powder used in the concrete mixture was used to create a filling effect. In order to
141 determine the usage rate of marble powder in concrete, many trial mixtures were poured and the
142 optimum usage rate was determined [39], [40].
143
144 2.1. Preparation of Test Samples
145
146 Within the scope of the study, a wide range of trial mixes were poured in the first stage. As a
147 result of these castings, 9 reference concrete mixtures with W/C ratios of 0.20-0.25-0.30 and binder
148 material (cement+micro silica) dosage of 400-450-500 kg/m3 were prepared (Table 5). In the series
Dr
149 names in the table, the M value represents maturity and HSC400 represents binder material dosage.
aft
150 Micro silica cement and marble powder were used as mineral additives in the prepared concrete mixtures
151 by volume substituting with fine aggregate. The cement was replaced with micro silica. Marble powder
152 was replaced with aggregate. The total amount of binder was kept constant. In all mixtures, CHRYSO
153 Fluid 518 hyperplasticizer was used as a chemical additive at the rate of 3% of the cement amount. Many
154 trial mixtures have been made in concrete mixtures. In these trial mixtures, different ratios of chemical
155 additives were used and the additive ratio that ensures appropriate workability was determined. After
156 the trial mixtures made in this way, it was seen that the most suitable workability for HSCs was 3%.
157 In addition since the W/C ratio is low in high-strength concrete, many trial mixtures were made and the
158 mixtures were repeated until appropriate workability was achieved. In this way, S2 and S3 slump classes
159 were obtained [39] .
160 Table 5. Mixture amounts of M-HSC (kg/m3)

Fine Fine Medium Coarse
Micro Marble
Serial Name Cement Water Aggregate Aggregate Aggregate Aggregate
Silica Powder
(0-2) mm (2-4) mm (4-8) mm (8-16) mm
M-HSC400-1 400 80 28.1 103 559 368 406 609
M-HSC400-2 400 100 14.1 101 550 362 400 599
M-HSC400-3 400 120 14.1 79 547 360 389 583
M-HSC450-1 450 90 47.4 78 544 358 387 580
M-HSC450-2 450 113 47.4 114 504 332 374 562

M-HSC450-3 450 135 31.6 74 516 340 367 550

M-HSC500-1 500 100 17.6 116 513 338 381 572
M-HSC500-2 500 125 35.1 95 517 341 376 564
M-HSC500-3 500 150 52.7 87 473 312 344 516
161
162 2.2. Application of Maturity Tests
163
164 This study was carried out to develop numerical models that predict concrete compressive
165 strength with high accuracy by using Nurse-Saul and Arrhenius maturity functions separately [41]. With
166 these maturity functions, the compressive strength of early age concretes can be estimated and the
167 demoulding time can be determined [42]. Thus, advantages are provided in terms of project duration,
168 time and cost [43]. In the study, firstly, the Nurse-Saul method given in Equation 2, in which the
169 compressive strength of the concrete is estimated by looking at the temperature history, was used.
170
t
171 M = ∑0(T ― T0) ∗ ∆t (2)
172
Dr
173 Here;
174 M = Maturity index (°C-hours or °C-days)

aft
175 T = Average concrete temperature (°C)
176 T0 = Reference temperature
177 t = elapsed time (day or hour)
178 ∆t = Time range (day or hour)
179
180 Copeland et al. in 1960 suggested that the effects of cement hydration on the strength gain rate
181 of concrete can be described by the Arrhenius equation. In 1977, Freisleben Hansen and Pedersen
182 proposed the following equation for equivalent age based on the Arrhenius equation (Equation 3) [7].
183 The application and numerical values related to the experiment are explained in ASTM C1074 [5].
184
185
𝐄 𝟏 𝟏
t ― ( ― )
186 te = ∑0𝐞 𝐑 𝟐𝟕𝟑 + 𝐓 𝟐𝟕𝟑 + 𝐓𝐫 ∆t (3)
187
188 Here;

189 te = Equivalent age at reference temperature
190 E = Apparent activation energy
191 R = Universal gas constant
192 T = Average absolute temperature of concrete
193 Tr = Reference temperature
194 A multi-channel temperature measuring device was used to determine the concrete temperatures
195 in the maturity calculations. The device was set up to take measurements every 5 minutes after the
196 concrete was poured and placed in the mold and placed in the samples (Figure 3).
197
Dr
198
aft
199
200
201 Figure 3. Placement of the temperature measuring device
202
203 2.3. Measurements on Concrete Samples
204 Temperatures were measured from the beginning of the concrete casting and the temperature results
205 taken every 5 minutes at the end of the 1st, 3rd, 7th, 14th and 28th days within the scope of the C1074
206 standard were transferred to the computer. The time-dependent graph of these temperatures is given in
207 Figure 4. Figure 4 shows the time-dependent variation of the temperatures forty-eight, ninety-six hours

208 after the concrete is placed in the mold, and the average temperatures of one hundred and ninety-two,
209 three hundred and sixty-six hours after it is placed in the mold.
30
HSC400-25
HSC400-30
25
Temperature (°C)
20
15
10
0 10000 20000 30000 40000 50000
210 Time (min)

Dr
211 (a)
212
30
aft
HSC450-25
HSC450-30
25
Temperature (°C)
20
15
10
0 10000 20000 30000 40000 50000
213 Time (min)

214 (b)
215

30
HSC500-25
HSC500-25
25
Temperature (°C)
20
15
10
0 10000 20000 30000 40000 50000
216 Time (min)

217
Dr
(c)
218
219 Figure 4. (a) Temperature values of 400-dose series (b) Temperature values of 450-dose series (c) Temperature
aft
220 values of 500-dose series

221
222 When Figure 3 is examined, the temperature values of the series in which the W/C ratio is 0.25
223 and 0.30 are given. Since there is not enough water in the environment as specified in the C1074 standard
224 in the series with a W/C ratio of 0.20 and proper hydration has not occurred, the data of these series
225 were not evaluated.
226 3. Evaluation of Experimental Results
227 3.1. Concrete Temperature Values and Evaluation
228 Concrete temperatures were obtained using a multi-channel temperature measuring device. This
229 temperature measurement started from the placement of the concrete in the mold (Figure 5.) When
230 Figure 5 is examined, it is seen that the temperature change varies according to the usage rate of the
231 materials in the concrete composition. For example, it is clearly seen that the temperature value of the
232 series with a binder dosage of 500 and a W/C ratio of 25% is higher than the temperature value of a
233 series with a binder dosage of 400 and a W/C ratio of 25%.

234 The temperature values in Figure 3 and the maturity functions in Equation 2 and Equation 3
235 were calculated separately and the Nurse-Saul-time and Arrhenius-time graphs given in Figure 5 were
236 obtained.
237
8000
Nurse-Saul (°C-min)
7000
6000 HSC500-30
5000 HSC500-25
4000 HSC450-30
3000 HSC450-25
2000 HSC400-30
1000 HSC400-25
0
0
1255
2510
3765
5020
6275
7530
8785
10040
11295
12550
13805
15060
16315
17570
18825
20080
21335
22590
23845
25100
26355
27610
28865
30120
31375
32630
33885
35140
36395
37650
38905
40160
41415
Time (min)
238
239 (a)
240
8000
7000
Arrhenius (min)
6000 HSC500-30
Dr
5000 HSC500-25
4000 HSC450-30
3000 HSC450-25
aft
2000 HSC400-30
1000 HSC400-25
0
0
1235
2470
3705
4940
6175
7410
8645
9880
11115
12350
13585
14820
16055
17290
18525
19760
20995
22230
23465
24700
25935
27170
28405
29640
30875
32110
33345
34580
35815
37050
38285
39520
40755
Time (min)
241
242 (b)
243
244 Figure 5. (a) Nurse-Saul values-Time (b) Arrhenius values-Time comparisons
245 In the Figure 5 showing the time-dependent variation of the maturity function, it is seen that the
246 binder dosage and maturity values are directly proportional. As the dosage value increases, the hydration
247 increases, so the maturity value also increases. When evaluated in terms of W/C ratio, the maturity value
248 was found to be higher in the 400 and 450-dose series in which the W/C ratio was 30%. This indicates
249 that sufficient water for hydration is found in series with a 30% W/C ratio. In the 500-dose series, the
250 maturity value of the series in which the W/C ratio was 25% was high.
251 The relationship between maturity values calculated by Nurse-Saul and Arrhenius equations and
252 compressive strength is shown in Figure 6.
253

Nurse-Saul HSC400-25 HSC400-30 HSC450-25

HSC450-30 HSC500-25 HSC500-30
85
75
65
55
45
35
400 2400 4400 6400 8400 10400
Maturity (°C-hour)
254
255 (a)
Arrhenius
HSC400-25 HSC400-30 HSC450-25
HSC450-30 HSC500-25 HSC500-30
90
85
80
75
70
65
60
55
50
45
40
1 3 Maturity (°C-hour) 5 7
Dr
256
257 (b)
aft
258 Figure 6. Comparison of (a) Compressive strength-Nurse-Saul values (b) Compressive strength-Arrhenius
259 maturity values
260
261 In Figure 6, it is seen that there is a linear relationship between maturity and compressive
262 strength depending on the change in concrete temperature. In accordance with the C1074 standard, the
263 compressive strength value of the concrete increases with the increase of the maturity value. The
264 numerical models estimating the compressive strength according to the NS and Arr maturity methods
265 for each dosage and W/C value from the compressive strength maturity graph are given in Table 6.
266 While determining the models, maturity value was determined as input and compressive strength of
267 concrete value was determined as output.
268
269 Table 6. Numerical models developed with NS and Arr maturity methods
Serial Name NS Arr
M-HSC400-25 fc = 0.0027M + 48.131 fc = 6.7943M + 33.573

M-HSC400-30 fc = 0.0027M + 45.805 fc = 7.0137M + 30.937
M-HSC450-25 fc = 0.0027M + 51.143 fc = 6.8513M + 36.05
M-HSC450-30 fc = 0.0028M + 46.995 fc = 7.0219M + 31.914
M-HSC500-25 fc = 0.0032M + 54.705 fc = 6.9625M + 37.284
M-HSC500-30 fc = 0.0029M + 52.118 fc = 7.4902M + 35.333

270
271 In Figure 7, the estimation results and ARD values obtained by the numerical models developed
272 with the experimentally calculated compressive strength values are given. When the experimental values
273 and predicted values are compared, it is seen that the ARD values of the developed models are below
274 10%. It was noteworthy that all these values were below 5%, especially for early age strength. When
275 the ARD values of the Arrhenius maturity function were examined, it was noted that it was the maturity
276 function that predicted fc with the highest accuracy. Especially in fc-1 daily strength values, highly
277 accurate predictions were made and ARD values were below 1% for almost all other days.
278 Although the ARD results estimated by the Nurse-Saul equation are higher than the ARD results
279 obtained by the Arrhenius equation, the compressive strength values obtained with this function also
280 have high accuracy.
ARD value-Nurse-Saul ARD value-Arrhenius

25 10
M-HSC400-25 M-HSC400-25
Dr
15 5 M-HSC400-30
M-HSC400-30
5 M-HSC450-25 M-HSC450-25
0 M-HSC450-30
-5 M-HSC450-30
aft
M-HSC500-25 M-HSC500-25
-15 -5
M-HSC500-30 M-HSC500-30
-25 -10
281
M -HSC400-25 M -HSC400-30 M -HSC450-25 M -HSC450-30 M -HSC500-25 M -HSC500-30
Days fc fc-NS fc-Arr fc fc-NS fc-Arr fc fc-NS fc-Arr fc fc-NS fc-Arr fc fc-NS fc-Arr fc fc-NS fc-Arr
1. 50 48.4 49.6 47 48.4 47.2 54 53.8 52.4 48 49.7 48.6 61 58.4 57.3 57 55.2 53.5
3. 57 52.9 56.1 55 50.6 54.0 59 56.0 59.2 56 52.0 55.2 64 62.2 67.7 61 57.4 60.6
7. 63 57.2 62.5 61 54.8 60.6 68 60.2 65.6 63 56.7 61.9 73 67.2 74.2 67 61.9 67.6
14. 67 64.2 68.6 65 61.9 66.8 69 67.3 71.8 67 63.7 68.0 81 75.5 80.4 76 69.5 75.3
28. 75 78.4 74.5 73 76.0 72.1 78 81.5 77.6 74 78.4 74.1 88 92.6 86.8 80 84.8 81.2
282
283 Figure 7. Experimental and predicted results
284
285 Conclusions
286 In this study, numerical models were developed to predict the compressive strength of HSCs using
287 NS and Arr maturity functions. In particular, to verify the validity of the maturity functions proposed so
288 far, compressive strength-Nurse-Saul maturity function and compressive strength-Arrhenius maturity
289 functions were compared and their suitability was determined. As a result of the study, the following
290 conclusions were drawn:

291 1- The strength-maturity curve in each series of concrete mixtures followed the literature. As the
292 strength value increased, the maturity value also tended to increase.
293 2- It has been determined that the maturity method can be used for HSCs, and it has been observed
294 that the Arr method gives higher accuracy results in estimating the compressive strength of
295 HSCs. The ARD values of the compressive strength values obtained with the Arr method were
296 lower.
297 3- The average relative error percentage in estimating the compressive strengths of concrete varies
298 between 0.1% and 10.1%. Mixtures producing the highest errors were generally found in the
299 calculation made by the Nurse-Saul method. The compressive strength values obtained by the
300 Arhenius method were closer to the experimental results.
301 4- In series with low W/C ratio, strength losses occurred due to insufficient hydration. Therefore,
302 it is recommended to be more careful in the estimations to be made with the maturity method
303 as the W/C ratio decreases in practice.

Dr
304 5- By using the maturity method and determining the strength gaining time of the concrete, the
aft
305 demoulding times can be determined. Thus, the problem of getting molds early is eliminated
306 and an advantage is provided in determining the project durations.
307 6- In the current maturity functions, the concrete temperature is taken according to the standards.
308 Concrete compressive strength can also be determined more accurately and with higher
309 accuracy by obtaining the temperatures with high accuracy by using a temperature measuring
310 device in concrete.
311 7- Verification of proposed modifications for in situ concrete strength estimation should also be
312 done, especially for concretes made with different types of cement and additives.
313
314 As a result, the places where HSCs are used in practice are special productions. Therefore, the duration
315 of the project is of great importance in such productions. The most important parameter for the
316 construction to progress in accordance with the project period is the molding time. Since this period
317 cannot be calculated correctly, the maturity method has emerged as a sensitive, fast and practical
318 method. The use of this method will ensure that the concrete compressive strength is correctly estimated

319 and the construction will continue in a healthy way in terms of strength and will proceed without loss of
320 time in the project-work plan.
321
322 Acknowledgments
323 This study is supported by the Scientific Research Project Fund of Fırat University under the
324 project number MF.21.51
325
326 References
327 [1] I. Belykh, V. Sopov, L. Butska, L. Pershina, and O. Makarenko, “Predicting the strength and
328 maturity of hardening concrete,” in MATEC Web of Conferences, 2018, vol. 230, p. 3001.
329 [2] F. A. Oluokun, E. G. Burdette, and J. H. Deatheridge, “Early-age concrete strength prediction
330 by maturity--another look,” Mater. J., vol. 87, no. 6, pp. 565–572, 1990.
331 [3] M. Soutsos, F. Kanavaris, and A. Hatzitheodorou, “Critical analysis of strength estimates from
Dr
332 maturity functions,” Case Stud. Constr. Mater., vol. 9, p. e00183, 2018.
333 [4] D. Miller, N.-M. Ho, and N. Talebian, “Monitoring of in-place strength in concrete structures
aft
334 using maturity method–An overview,” in Structures, 2022, vol. 44, pp. 1081–1104.
335 [5] C. ASTM, “1074-11.(2011).“,” Stand. Pract. Estim. Concr. Strength by Matur. Method, Am.
336 Soc. Test. Mater.
337 [6] J.-S. Jung and Y.-H. Cho, “Comparison between nurse-saul and arrhenius equations,” Can. J.
338 Civ. Eng., vol. 36, no. 3, pp. 514–518, 2009.
339 [7] D. A. Wood, “Re-establishing the merits of thermal maturity and petroleum generation multi-
340 dimensional modeling with an Arrhenius Equation using a single activation energy,” J. Earth
341 Sci., vol. 28, no. 5, pp. 804–834, 2017.
342 [8] G. Chengju, “Maturity of concrete: method for predicting early-stage strength,” Mater. J., vol.
343 86, no. 4, pp. 341–353, 1989.
344 [9] T. Demir, M. Ulucan, and K. E. Alyamac, “Determination of Early Age Strength of High
345 Strength Concretes Using RSM Method,” Fırat Univ. J. Eng. Sci., vol. 34, no. 1, pp. 105–114,
346 2022.

347 [10] Y. B. Guo, G. F. Gao, L. Jing, and V. P. W. Shim, “Response of high-strength concrete to
348 dynamic compressive loading,” Int. J. Impact Eng., vol. 108, pp. 114–135, 2017.
349 [11] T. Demir and K. E. Alyamaç, “Investigation of the Use of Marble Powder in Production of
350 High Strength Concretes,” Open J. Nano, vol. 7, no. 1, pp. 18–25, 2022.
351 [12] A. Kandiri, E. M. Golafshani, and A. Behnood, “Estimation of the compressive strength of
352 concretes containing ground granulated blast furnace slag using hybridized multi-objective
353 ANN and salp swarm algorithm,” Constr. Build. Mater., vol. 248, p. 118676, 2020.
354 [13] V. M. Sounthararajan and A. Sivakumar, “Effect of the lime content in marble powder for
355 producing high strength concrete,” ARPN J. Eng. Appl. Sci., vol. 8, no. 4, pp. 260–264, 2013.
356 [14] T. S. EN, “TS EN 12504-2,” Test. Concr. Struct. part-2 non-destructive testing, Determ.
357 rebound number, Stand., 2004.
358 [15] B. Standard, “Structural Use of Concrete: Code of Practice for Design and Construction, Part 1,
359 BS 8110.” British Standard Institution, UK, 1997.

Dr
360 [16] V. M. Malhotra, “Testing hardened concrete: nondestructive methods,” 1976.

aft
361 [17] R. L. Wankhade and A. B. Landage, “Non-destructive testing of concrete structures in Karad
362 region,” Procedia Eng., vol. 51, pp. 8–18, 2013.
363 [18] A. G. Davis et al., “Nondestructive test methods for evaluation of concrete in structures,” Am.
364 Concr. Institute, ACI, vol. 228, no. 4, 1998.
365 [19] J. Malek and M. Kaouther, “Destructive and non-destructive testing of concrete structures,”
366 Jordan J. Civ. Eng., vol. 8, no. 4, pp. 432–441, 2014.
367 [20] O. Tsioulou, A. Lampropoulos, and S. Paschalis, “Combined non-destructive testing (NDT)
368 method for the evaluation of the mechanical characteristics of ultra high performance fibre
369 reinforced concrete (UHPFRC),” Constr. Build. Mater., vol. 131, pp. 66–77, 2017.
370 [21] A. A. Shah and Y. Ribakov, “Non-destructive evaluation of concrete in damaged and
371 undamaged states,” Mater. Des., vol. 30, no. 9, pp. 3504–3511, 2009.
372 [22] T. Demir, “Development of Combined Methods to Estimate the in-Place Strength of High
373 Strength Concrete Using Non-Destructive Testing Methods,”,” Fırat University, 2022.
374 [23] N. Mylvaganam and Y. Elakneswaran, “A Systematic Review and Assessment of Concrete

375 Strength Prediction Models,” Case Stud. Constr. Mater., p. e01830, 2023.
376 [24] L. Wang, H. Zhou, J. Zhang, Z. Wang, L. Zhang, and M. L. Nehdi, “Prediction of concrete
377 strength considering thermal damage using a modified strength-maturity model,” Constr. Build.
378 Mater., vol. 400, p. 132779, 2023.
379 [25] Y. Murad, “Compressive strength prediction for concrete modified with nanomaterials,” Case
380 Stud. Constr. Mater., vol. 15, p. e00660, 2021.
381 [26] S. N. Bengal, L. S. Pammar, and C. B. Nayak, “Engineering application of organic materials
382 with concrete: A review,” Mater. Today Proc., vol. 56, pp. 581–586, 2022.
383 [27] M. Soutsos and F. Kanavaris, “Applicability of the Modified Nurse-Saul (MNS) maturity
384 function for estimating the effect of temperature on the compressive strength of GGBS
385 concretes,” Constr. Build. Mater., vol. 381, p. 131250, 2023.
386 [28] P. and C. C. Cement - Part 1: General Cements, Composition, “TS EN 197-1,” Turkey, 2012.
387 [29] T. S. I. TS, “TS EN 706-Aggregate for concretes.” Ankara, 1980.

Dr
388 [30] O. Soykan, Ö. Cengiz, and Ö. Cenk, “Investigation of the Usability of Slate and Andesite as
aft
389 Concrete Aggregate,” J. Suleyman Demirel Univ. Grad. Sch. Nat. Appl. Sci., vol. 19, no. 1,
390 2015.
391 [31] M. SÜMER and B. SÖYLER, “yüKSEK DAYANIMLI BETON ÜRETİMİNDE ÇİMENTO
392 ve SUPERAKIŞKANLAŞTIRICI BETON KATKI MADDELERİNİN ETKİNLiG i,” Sak.
393 Üniversitesi Fen Bilim. Enstitüsü Derg., vol. 6, no. 3, pp. 3–15, 2002.
394 [32] T. S. ENSTİTÜSÜ, “TS 706 EN 12620+ A1 Concrete Aggregates,” TSE, Ankara, Türkiye,
395 2009.
396 [33] A. Shayan and A. Xu, “Performance of glass powder as a pozzolanic material in concrete: A
397 field trial on concrete slabs,” Cem. Concr. Res., vol. 36, no. 3, pp. 457–468, 2006.
398 [34] N. Flores Medina, G. Barluenga, and F. Hernández-Olivares, “Combined effect of
399 Polypropylene fibers and Silica Fume to improve the durability of concrete with natural
400 Pozzolans blended cement,” Constr. Build. Mater., vol. 96, 2015.
401 [35] R. Yu, P. Spiesz, and H. J. H. Brouwers, “Development of an eco-friendly Ultra-High
402 Performance Concrete (UHPC) with efficient cement and mineral admixtures uses,” Cem.

403 Concr. Compos., vol. 55, 2015.
404 [36] M. Vigneshwari, K. Arunachalam, and A. Angayarkanni, “Replacement of silica fume with
405 thermally treated rice husk ash in Reactive Powder Concrete,” J. Clean. Prod., vol. 188, 2018.
406 [37] T. Standard, “TS EN 196–2,” Methods Test. Cem., vol. 2.
407 [38] requirements and conformity criteria Natural pozzolan (Trass) for use in cement and concrete -
408 Definitions, “TS 25,” Ankara,Turkey, 2015.
409 [39] T. DEMİR, M. ULUCAN, and K. E. ALYAMAÇ, “Yüksek Dayanımlı Betonların Erken Yaş
410 Dayanımının RSM Metodu Kullanılarak Belirlenmesi,” Fırat Üniversitesi Mühendislik Bilim.
411 Derg., vol. 34, no. 1, pp. 105–114.
412 [40] T. Demir, B. Demirel, and M. Öztürk, “An Evaluation of the Effect of Waste Aluminum
413 Sawdust on the Carbonation of Concrete,” Bitlis Eren Univ. J. Sci., vol. 11, no. 4, pp. 993–999.
414 [41] M. T. de Grazia, L. F. M. Sanchez, D. J. De Souza, L. Ismail, M. Noel, and S. Decarufel,
415 “Evaluation of electrical resistivity and maturity for estimating the early-age properties of pre-
Dr
416 packaged concrete,” Can. J. Civ. Eng., vol. 49, no. 7, pp. 1231–1238, 2022.
aft
417 [42] C. J. Willis and J. H. Rankin, “Demonstrating a linkage between construction industry maturity
418 and performance: a case study of Guyana and New Brunswick,” Can. J. Civ. Eng., vol. 39, no.
419 5, pp. 565–578, 2012.
420 [43] D. Breysse, “Nondestructive evaluation of concrete strength: An historical review and a new
421 perspective by combining NDT methods,” Constr. Build. Mater., vol. 33, pp. 139–163, 2012.
422

Cjce-2023-0466 Proof Hi

Uploaded by

Copyright:

Available Formats

Cjce-2023-0466 Proof Hi

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cjce-2023-0466 Proof Hi

Uploaded by

Copyright:

Available Formats

Canadian Journal of Civil Engineering

DEVELOPMENT OF NUMERICAL MODELS TO PREDICT THE

Journal: Canadian Journal of Civil Engineering

Manuscript Type: Article

Date Submitted by the

Is the manuscript for

consideration in a Special Not applicable (regular submission)

High strength concrete, compressive strength, maturity, silica fume,

© The Author(s) or their Institution(s)

1 DEVELOPMENT OF NUMERICAL MODELS TO PREDICT THE COMPRESSIVE

2 STRENGTH OF HSCs BASED ON MATURITY METHOD

3 Tuba DEMIR1*, Kursat Esat ALYAMAC2

6 Corresponding Author: Tuba DEMIR

19 project duration to be completed at the desired time.

22 Corresponding Address : Tuba Demir (corresponding author).

25 Affiliations: 2-Civil Engineering Department, Engineering Faculty, Firat University, 23119Elazig,

© The Author(s) or their Institution(s)

45 of in-depth analysis of their reliability and limitations [8].

© The Author(s) or their Institution(s)

70 of construction is of great importance in preventing various negative effects on engineering projects.

© The Author(s) or their Institution(s)

92 projects using HSC. In Figure 1, the experimental procedure is presented schematically.

102 HSC components are presented in Table 1.

104 Table 1. Physical and chemical properties of using materials (%).

© The Author(s) or their Institution(s)

Cement Silica Fume Marble Powder

109 dispersible and dense texture [31],[32].

111 Table 2. Los Angeles experiment results

marbles class marbles Weight cycles 500 cycles

12 8-16 mm 5030 gr 5000 gr 4569.5 430.5

115 granulometry curve of the aggregate is presented in Figure 2.

116 Table 3. Properties of the aggregate (Saturated dry surface, gr/cm3)

© The Author(s) or their Institution(s)

Sieve opening (mm)

121 Figure 2. Aggregate granulometry curve

134 A : Average compressive strength of mortar samples with pozzolan

© The Author(s) or their Institution(s)

fc(7) fc(28) fc(7) fc(28)

141 optimum usage rate was determined [39], [40].

158 were obtained [39] .

159 Table 5. Mixture amounts of M-HSC (kg/m3)

© The Author(s) or their Institution(s)

173 M = Maturity index (°C-hours or °C-days)

174 T = Average concrete temperature (°C)

175 T0 = Reference temperature

176 t = elapsed time (day or hour)

177 ∆t = Time range (day or hour)

188 te = Equivalent age at reference temperature

© The Author(s) or their Institution(s)

189 E = Apparent activation energy

190 R = Universal gas constant

191 T = Average absolute temperature of concrete

192 Tr = Reference temperature

202 2.3. Measurements on Concrete Samples

© The Author(s) or their Institution(s)

209 Time (min)

212 Time (min)

© The Author(s) or their Institution(s)