Temperature Effect on Mechanical Properties and Damage Identification of Concrete Structure

Static and dynamic mechanical properties of concrete are affected by temperature effect in practice. Therefore, it is necessary to investigate the corresponding influence law and mechanism. This paper demonstrates the variation of mechanical properties of concrete at temperatures from −20C to 60C. Temperature effects on cube compressive strength, splitting tensile strength, prism compressive strength, modulus of elasticity, and frequency are conducted and discussed.The results indicate that static mechanical properties such as compressive strength (cube and prism), splitting tensile strength, and modulus of elasticity have highly linear negative correlation with temperature; this law is also applied to the first order frequency of concrete slab. The coupling effect of temperature and damage on change rate of frequency reveals that temperature effect cannot be ignored in damage identification of structure. Mechanism analysis shows that variation of elastic modulus of concrete caused by temperature is the primary reason for the change of frequency.


Introduction
As widely used construction material in civil engineering structures, the properties of concrete have direct relationship with structural safety.Standard concrete specimens are usually cured and tested under ideal conditions according to specifications.However, concrete structures are exposed to external environment in service; the mechanical (static and dynamic) properties of concrete are complicated as they are affected by environmental conditions such as temperature and moisture [1,2].
Numerous researches have been conducted to investigate the relationships between concrete performances and environmental conditions.Peng et al. [3] present environmental investigations to obtain the relationship between explosive spalling occurrence and residual mechanical properties of high-performance concrete and high temperature ranging from 200 to 800 ∘ C. Cülfik and Özturan [4] investigated the temperature effect on mechanical properties of both normal and high strength concrete.Residual compressive and splitting tensile strength, as well as the static modulus of elasticity of specimens, were obtained under elevated temperatures (50, 100, 150, 200, and 250 ∘ C).The residual mechanical properties started to decrease at 100 ∘ C for normal strength concrete and high strength concrete performed better under different temperature cycles.Fares et al. [5] conducted an experimental study on performance of self-consolidating concrete under high temperature effect.Mechanical properties such as compressive strength and flexural strength are tested at different temperatures (150, 300, 450, and 600 ∘ C).The performance of concrete under transient temperature conditions has been comprehensively investigated, particularly under actions such as high temperatures from fire.However, most of concrete structures are exposed to external temperature between −20 ∘ C and 60 ∘ C in practical situation; relative little literature is available for properties of concrete under this condition.Shoukry et al. [6] investigated static mechanical properties of concrete under temperature ranging from −20 ∘ C to 50 ∘ C and relative humidity between 40% and 60%.The results indicated that concrete strength in compression, tension, and corresponding modulus of elasticity decreased with temperature increasing.However, the dynamic properties of concrete structure are not investigated.
Dynamic-based methods have been widely applied for health monitoring and damage identification of structures  [7][8][9], because the dynamic properties (frequency, mode shape, and damping) are closely related to structural parameters such as stiffness and mass [10][11][12].At present, most researches are focused on the effect of damage on dynamic properties without considering temperature effect [13].However, if the effects are not rationally taken into account and removed, the damage identification results cannot be reliable.Farrar et al. [14] present the monitoring results of Alamosa Canyon Bridge.They found that the first eigenfrequency of the structure varies approximately 5% over a 24-hour time period.Askegaard and Mossing [15] found that the variation of frequency for a three-span footbridge was 10% over the year.Wahab and de Roeck [16] observed a change of 4%∼ 5% in natural frequencies of a prestressed concrete bridge in spring and winter from dynamic test.Moser and Moaveni [17] demonstrated that significant variability in natural frequencies was identified from a continuous monitoring system installed on Dowling Hall Footbridge, and these changes in natural frequencies are strongly correlated with temperature.
Extensive research results reveal that temperature will lead to frequencies changes of structure, but explicit investigation of the effect of temperature is not obtained in experiments.
In this paper, static mechanical properties such as cube compressive strength, split tensile strength, prism compressive strength, and modulus of elasticity of concrete are conducted under different temperatures from −20 ∘ C to 60 ∘ C. The relationships between properties and temperatures are determined.Meanwhile, the temperature effect on first order natural frequency of concrete slab is also observed and corresponding influence mechanism is discussed.

Experimental Details
2.1.Materials and Mixtures.PO 42.5 type Portland cement confirming the requirements of GB175-2007 [18] is used in this study.Crushed stone with diameters ranging from 5 mm to 20 mm and natural sand with fineness modulus of 2.7 are adopted as coarse and fine aggregates, respectively.The mixture proportions of concrete are listed in Table 1.These ingredients are mixed for about 4 min.Slump results of concrete are tested with 38 mm, which indicate that the mixture is with favorable cohesiveness and meets the requirements of GB/T 50081-2002 [19] and GB 50164-2011 [20].
2.2.Specimens.Concrete specimens are 150 mm cubes (produced for testing cube compressive strength and splitting tensile strength) and 150 mm × 150 mm × 300 mm prisms (produced for testing prism compressive strength and  modulus of elasticity) which are casted horizontally in steel molds and compacted by vibrating machine.All specimens are removed from the molds 24 h after casting and cured in a curing room (20 ± 3 ∘ C and 95% RH (relative humidity)) for 28 days (GB/T 50082-2009) [21].
Slabs with length 60 cm, width 15 cm, and height 5 cm are also prefabricated, and they are also cured in a curing room (20 ± 3 ∘ C and 95% RH (relative humidity)) for 28 days.

Testing Machines.
Two different compression testing machines are used to obtain the static properties of concrete.Both cube and prism compressive strength and modulus of elasticity are tested through machine 1 (shown in Figure 1) with maximum load 2000 kN.Splitting tensile strength is measured by machine 2 (shown in Figure 2) with maximum load 1000 kN, which can realize the dynamic monitoring of force and output the force-displacement curves.
The temperatures are jointly controlled by refrigerator and oven.The former is used for −50 ∘ C-0 ∘ C, while the latter is for 20 ∘ C-300 ∘ C.
Piezoelectric acceleration sensor is applied for measuring acceleration response signals for slabs, and DH 5920 type dynamic signal measurement system is used to obtain the natural frequency from the measured signals.
As for the damage of slab, it is achieved by cutting machine (shown in Figure 3).

Testing Procedures.
The cube compressive strength, splitting tensile strength, prism compressive strength, modulus of elasticity, and frequency of concrete structures are tested under temperatures −20 ∘ C, 0 ∘ C, 20 ∘ C, 40 ∘ C, and 60 ∘ C. In order to eliminate the influence of moisture, the concrete specimens are wrapped by preservative film to maintain humidity stability.Each specimen is placed in temperature control apparatus for 4 h to make the core temperature of concrete consistent with environmental temperature before testing.

Cube Compressive Strength.
Totally fifteen 150 mm × 150 mm × 150 mm specimens are tested for calculating the cube compressive strength (i.e., three specimens at each temperature).In the process of cube compressive strength testing, machine 1 imposes loads at the speed of 0.5-0.8MPa/s [19].The loads are recorded until the failure of specimens.The cube compressive strength of each specimen can be calculated by where   cu is cube compressive strength of the th concrete specimen,   cu is the corresponding failure load under compression,   cu is bearing area of specimens, and  = 1, 2, 3 in this paper.
The final cube compressive strength  cu of concrete can be determined by the following regulations [19].
(1) The average value of cube compressive strength for three specimens is treated as final compressive strength.
(2) If one of maximum and minimum value of cube compressive strength for three specimens exceeds the medium one by 15%, the medium value of cube compressive strength is used as final cube compressive strength.
(3) If all of maximum and minimum values of cube compressive strength for three specimens exceed the medium one by 15%, the results are considered to be invalid and another group of specimens should be retested.

Splitting Tensile Strength.
Another fifteen 150 mm × 150 mm × 150 mm specimens are measured for calculating the splitting tensile strength (i.e., three specimens at each temperature).In the process of splitting tensile strength testing, machine 2 applied loads at the speed of 0.05-0.08MPa/s [19] until the failure of specimens.The splitting tensile strength of concrete can be obtained by where   ts is splitting tensile strength of the th concrete specimen,   ts is the corresponding failure load under splitting, and   ts is splitting area of specimens.Final splitting tensile strength  ts can be calculated according to the same rules of getting  cu .

Prism Compressive Strength.
Totally fifteen 150 mm × 150 mm × 300 mm prisms are tested for measuring the prism compressive strength of concrete.In the process of prism compressive strength testing, machine 1 imposes loads at the speed of 0.5-0.8MPa/s [19].The loads are recorded until the failure of specimens.The prism compressive strength of each specimen can be calculated by where   cp is prism compressive strength of the th concrete specimen,   cp is the corresponding failure load under compression,   cp is bearing area of specimens, and  = 1, 2, 3 in this paper.
Final prism compressive strength  cp can be calculated according to the same rules of getting  cu .

Modulus of Elasticity.
Totally fifteen 150 mm × 150 mm × 300 mm prisms are tested for calculating the modulus of elasticity; it can be calculated by where    is modulus of elasticity of the th concrete specimen;    and   0 are loads corresponding to the stress of one-third prism compressive strength and 0.5 MPa, respectively;  is gauge length of specimen,  = 150 mm in this paper; Δ  is average value of deformation for concrete from   to   , and it can be obtained by where    is the average value of deformation for concrete specimens under    and   0 is the average value of deformation for concrete specimens under   0 .   and   0 are tested through dial gauges.
Final modulus of elasticity   can be calculated according to GB/T 50081-2002 [19].Firstly, the slabs are measured under intact condition and temperature effect on frequency for undamaged slab is discussed.Secondly, the midspan sections of slabs are cut by special machine to simulate damage.Damage severity is represented by cutting depth; it can be calculated by where  is damage severity and   and  are widths of damaged and intact midspan sections, respectively.2. The relationship between cube compressive strength, splitting tensile strength, and temperatures is shown in Figures 6 and 7.

Static Mechanical Properties
Linear formulas are applied to simulate the relationship between cube compressive strength, splitting tensile strength, and temperature.They are listed in the following equations, respectively:  As can be seen from Figures 6 and 7 and Table 2, the cube compressive strength and splitting tensile strength both decrease with the increasing of temperature.The cube compressive strength improves by 100.3% from 60 ∘ C to −20 ∘ C, while it is 132.9% for splitting tensile strength.

Prism Compressive Strength and Modulus of Elasticity.
Prism compressive strengths of specimens under temperatures −20 ∘ C, 0 ∘ C, 20 ∘ C, 40 ∘ C, and 60 ∘ C are measured and calculated.The results are listed in Table 3 and Figure 8.
The relationship between prism compressive strength and temperature is listed in The relationship reveals that prism compressive strength decreases with the increasing of temperature and presents   The measurement results and temperature effects on modulus of elasticity are listed in Table 4 and Figure 10.
Regression analysis is conducted through linear formula to represent the relationship between modulus of elasticity and temperature; it is listed in From Table 4 and Figure 10, modulus of elasticity increases with the decreasing of temperature, and a good linear relationship is presented.The modulus of elasticity improves by 48.3% from 60 ∘ C to −20 ∘ C.   11.Testing results of first order natural frequency for three slabs (, , and ) are listed in Table 5.

Advances in Materials Science and Engineering
Temperature effect on frequency of slabs is also demonstrated in Figure 12.It can be concluded that the influence of temperature on first order frequency is obvious.The frequency decreases with the increasing of temperature, and it shows clear negative correlation.

Temperature Effect on Damage
Identification.The results of slab  are used as examples to investigate the variation of frequency under the integrated effect of temperature and damage.Five damage severities including 0% (undamaged), 10%, 20%, 30%, and 40% are simulated, and the measurement results are listed in Table 6 and Figure 13.
As can be seen from Table 6 and Figure 13, the first order natural frequency decreases with the increasing of damage severity under the same temperature, while it decreases with the increasing of temperature under the same damage severity.
In damage identification of structures, change rate of natural frequency is an effective damage indicator.According to dynamic theory, damage of structure will lead to the reduction of frequencies.In order to illustrate the effect of temperature in damage identification of concrete slab, the baseline natural frequency of undamaged structure is assumed to be measured under temperature 40 ∘ C, and the damaged ones are tested under temperatures −20 ∘ C, 0 ∘ C, 20 ∘ C, 40 ∘ C, and 60 ∘ C. The corresponding change rates of first order frequency are calculated by (10), and the results are listed in Table 7 and Figure 14.Consider where  is the measured first order natural frequency,  is the change rate of frequency,  is damage severity ( = 0%, 10%, 20%, 30%, and 40%), and  is testing temperature ( = −20 ∘ C, 0 ∘ C, 20 ∘ C, 40 ∘ C, and 60 ∘ C).
In theory, damage of structure will lead to the reduction of frequency.Therefore, the change rate of frequency is a negative value.However, the change rates of frequency for damage severity 10% at temperatures −20 ∘ C and 0 ∘ C are positive.Reasons for this phenomenon lie in the fact that frequencies increase from temperature 40 ∘ C to −20 ∘ C and 0 ∘ C, and the recruitments due to temperatures are greater than the decreasing value caused by damage.It reveals that frequency changes caused by damage have been submerged by temperature effect, and damage identification results will be unreliable in practice.The range analysis is also conducted and   listed in Table 7.As can be seen from this table, the maximum range caused by damage is 15.368%, while it is 12.720% for temperature.The temperature effect is not negligible.

Mechanism Analysis.
For a simply supported uniform beam with length , height ℎ, density , and modulus of Temperature effect on modulus of elasticity has been tested and listed in Table 8.Theoretical calculation results of the first order frequency can be obtained based on (11); they are shown in Table 8 and Figure 15.As can be seen from Table 8 and Figure 15, the calculated frequencies based on measured modulus of elasticity and dynamic theory are consistent with the measured ones.The results indicate that temperature effect on frequency is mainly caused by the influence of temperature on modulus of  elasticity.It can provide reference for damage identification of structures.

Conclusions
Temperature effects on cube compressive strength, splitting tensile strength, prism compressive strength, modulus of elasticity, and frequency are conducted and the following conclusions can be drawn.
(1) Compressive strength (cube and prism), splitting tensile strength, and modulus of elasticity change obviously under temperature effect.These static mechanical properties all decrease with the increasing of temperature, and linear formulas can effectively simulate their relationships.The compressive strength improves by 100.3% from 60 ∘ C to −20 ∘ C; it is 132.9% and 48.3% for splitting tensile strength and modulus of elasticity, respectively.(2) Influence of temperature on the first order frequency of concrete slab is significant and the frequency decreases with the increasing of temperature; a good negative relationship is present.Under the coupling effect of damage and temperature, the frequency change caused by damage would be submerged by temperature effect.Therefore, it is necessary to consider the influence of temperature in damage identification of structure.The range analysis reveals that the temperature effect cannot be neglected in practice.
Mechanism analysis of temperature effect on frequency is also conducted.The results demonstrated that it is mainly caused by the influence of temperature on modulus of elasticity.

Figure 3 :
Figure 3: Damage simulation of slab using cutting machine.

2. 4 . 5 .
Frequency and Damage Identification.Three concrete slabs are prefabricated and tested for frequency under temperatures −20 ∘ C, 0 ∘ C, 20 ∘ C, 40 ∘ C, and 60 ∘ C. The concrete slabs are also wrapped by preservative film to maintain humidity stability to eliminate the influence of moisture.

3. 1 .
Cube Compressive and Splitting Tensile Strength.The testing arrangements of cube compressive and splitting tensile strength are shown in Figures4 and 5, respectively.The testing results of concrete under different temperatures are listed in Table

Figure 6 :Figure 7 :
Figure 6: Temperature effect on cube compressive strength of concrete.

Figure 9 :
Figure 9: Testing arrangement of modulus of elasticity.

Figure 10 :
Figure 10: Temperature effect on modulus of elasticity.

Figure 13 :
Figure 13: Integrated effect of temperature and damage on the first order natural frequency.

Figure 14 :
Figure 14: Integrated effect of temperature and damage on change rate of natural frequency.

Figure 15 :
Figure 15: Relationship between theoretical and measured frequency of concrete slab.

Table 2 :
Results of cube compressive strength and splitting tensile strength for concrete under different temperatures.

Table 3 :
Results of prism compressive strength for concrete under different temperatures.

Table 4 :
Modulus of elasticity for concrete under different temperatures.

Table 5 :
First order natural frequency for intact concrete slab.
4.1.Temperature Effect on Frequency.The test process is illustrated in Figure

Table 6 :
The first order natural frequencies under coupling effect of temperature and damage (Hz).

Table 7 :
Change rate of the first order natural frequency (%).

Table 8 :
Theoretical and measured results of the first order frequency for concrete slab.