Degradation Behaviour of Deteriorated RC Beams StrengthenedwithCFRPunder theCompoundEffects of Acid-Salt Mist and Carbon Dioxide

-e reinforcement effect and durability of carbon fibre-reinforced polymer (CFRP) sheets are important indicators that affect its promotion and application. -is study conducted an experimental investigation of the degradation behavior of reinforced concrete (RC) beams strengthened with CFRP.-e RC beams were deteriorated under the compound effects of acid-salt mist and carbon dioxide for a time period t1i, then strengthened with one or two layers of CFRP sheets, and placed in the same deterioration environment for an additional time period t2i. -e failure modes, load-deflection curves, rigidity, and flexural carrying capacity changes were studied after the deteriorated RC beams were loaded to failure. -ree combinations of deterioration periods were considered in this study: different t1i and same t2i, same t1i and different t2i, and same total time (t1i+ t2i).-is study found that the U-shaped hoops and the side concrete peeled off gradually as the CFRP-strengthened RC beams were deteriorated again with time. Under the ultimate load, the strengthening layer in the tension zone stripped. -e strengthened layer, which consisted of CFRP, a binder, and the concrete cover, was stripped from the RC beam during loading. -e deflection of the strengthened layer behaved differently from the other part; this disharmony prevented the mobilization of the tensile advantage of CFRP. -e mechanism of RC beam’s mechanical behaviour was analyzed in terms of the degrees of deterioration of RC beams and CFRP and their coupled effects. -e conclusions of this study can be used as references in the prediction of strength changing and service life of strengthened RC beams.


Introduction
Reinforced concrete (RC) structures in coalmines deteriorate rapidly in their deteriorative environment. It is a common practice to strengthen RC structures to increase their reliability [1]. Fibre-reinforced polymers have been widely used to strengthen RC structures due to their lightweight, high tensile strength, good corrosion resistance, and easy construction [2]. In particular, the carbon fibrereinforced polymer (CFRP) is widely accepted in RC structure-strengthening projects for its ability to increase the range of elastic area of RC structures, particularly when the tensile steel bars have yielded [3,4]. ree failure modes can be observed when CFRPstrengthened RC beams are under their ultimate loads: the main failure mode is concrete crushed while the tensile steel bars are yielded and the CFRP strengthening layer is stripped [5]. As the RC beams deteriorate further, the other two failure modes can be observed: CFRP tensile fracture and bonding layer shear fracture [6,7].
Epoxy glue, the organic binder to adhere CFRP onto concrete, has several shortcomings: (a) poor thermal stability and longer-term chemistry stability. e typical vitrification temperature (T g ) of the epoxy glue is 60-82°C. e elastic modulus and shear modulus of the epoxy glue decrease by approximately 90% when the temperature is 10°C to 20°C out of the typical range of T g [8]. (b) Greater flatness and smoothness of the concrete interface are required when grooves exist on the surface of concrete. e effective contact area between the concrete and binder decreases due to shrinkage of the epoxy glue. (c) Reverse tension is produced on the contact surface of the concrete, resulting in more ageing shrinkage on the interface of the CFRP and the stripped concrete [3,9].
As confirmed above, strengthening with CFRP increases the mechanical performance of RC beams, although the performance degrades under environmental effects. e strengthening effect differs for RC beams with different degrees of deterioration. e degradation mechanism of CFRP-strengthened RC beam still needs further investigation. e deterioration mechanism relates to the deterioration of the RC beam and CFRP and particularly the coupling effect of the two, which remains unclear to researchers.
us, this research conducted an experimental investigation of the degradation behavior of reinforced concrete (RC) beams strengthened with CFRP. e RC beams were deteriorated under the compound effects of acid-salt mist and carbon dioxide for a time period t 1i , then strengthened with one or two layers of CFRP sheets, and placed in the same deterioration environment for an additional time period t 2i . e failure modes, load-deflection curves, rigidity, and flexural carrying capacity changes were studied after the deteriorated RC beams were loaded to failure. ree combinations of deterioration periods were considered in this study: different t 1i and same t 2i , same t 1i and different t 2i , and same total time (t 1i + t 2i ). e flexural failure modes, load-deflection curves, rigidity, and loading capacity of RC beams that deteriorated for 18 to 30 months in the compound effects of acid-salt mist and carbon dioxide then exposed in the same environment for another 15 to 21 months after being strengthened with CFRP are studied; therefore, the mechanical performance of the RC beams in the deteriorative environment before and after strengthening and factors that affect the performance are analyzed in this paper. e research results better reflect the actual situation of the damage and deterioration of colliery ground RC structures and provide effective guides to the strengthening of deteriorated RC structures.

Materials and Beam
Specimen. Ordinary 32.5-grade Portland cement with a specific gravity of 3.18 and a specific surface area of 350 m 2 /kg was used in this investigation. e coarse aggregate was crushed stone with a maximum size of 16 mm. Natural river sand, medium size with a fineness modulus of 2.42, was used as the fine aggregate in this study. e ratio of fine aggregate to coarse aggregate was a fixed number of 35%. Concrete mixtures were designed according to the Chinese Standard JGJ55 [10] and GB50367 [11]. e mixture's weight proportion was water: cement: fine aggregate: coarse aggregate � 192 : 343 : 576 : 1063, with a 28day cubic compressive strength of 35.2 MPa. e properties of the CFRP and binder used in the experiment are shown in Table 1. e beam specimens had a length of 1,500 mm and a rectangular cross section of 100 mm × 200 mm (width × depth). e dimensions of the beam specimens are shown in Figure 1(a). Two plain round bars with diameters of 8 mm were placed at the top of the beam as hangers, and two 12 mm diameter ribbed bars were placed at the bottom of the beam as tensile rebar. e concrete cover was 15 mm thick. e yield and ultimate rebar strengths were 380 and 552 MPa, respectively. After a designed period of deterioration, the RC beams were strengthened with CFRP, as shown in Figure 1(b). Two layers of CFRP with the same width of the beam were adhered on the bottom of the RC beam, and three U-shaped hoops with a width of 50 mm were adhered to the sides and bottom of the beam in the shear section on each ends of the beam. e first hoop to the end of the beam was placed at 150 mm from the end. e spacing between each of the four hoops was 100 mm. To keep the reinforcing effect, two CFRP sheets with the width of 5 cm are used to strengthen both sides of the compression zone.

Test Design.
Twelve beam specimens, designated as L0 through L11, were made for this study. L0 and L1 were the control specimens, and L2 through L11 were placed in a simulated deteriorative environment. In the simulated environment, salt mist consisting of NaCl, MgCl 2 , and Na 2 SO 4 was sprayed on the beam specimens. Alternating spraying, drying, and freeze-thaw cycles were applied. e spray contents and simulation environment are listed in Table 2.
As shown in Table 3, the eleven specimens, L1 through L11, were first placed in the simulation environment for a period of t 1i and then strengthened with CFRP. e strengthened specimens were placed back to the same environment for a second period of t 2i , where i stands for the specimen Li. Based on the research results of [1,12], the investigation time was determined with 3 months per cycle. ree combinations of deterioration periods were considered in this study: (a) t 1i � 18 to 30 months and t 2i � 15 months; (b) t 1i � 21 months and t 2i � 15 to 21 months; (c) t 1i + t 2i � 42 months. All eleven test specimens were cast by the same batch. To accelerate deterioration, sodium chloride with 10% of water weight was added to beams L2, L4, and L7 during concrete mixing. After curing for 28 days and then sitting in the regular laboratory room condition for 7 months, the beam specimens were placed in the simulation environment.
e deterioration conditions of the beam specimens were examined every three months.
After the two deterioration periods, the test beams were ready to be tested. As shown in Figure 2, the specimens were simply supported and subjected to the four-point bending test. Each specimen had a clear span of 1,200 mm and shear spans of 400 mm. Bending was constant in the central zone with a length of 400 mm. A hydraulic jack was used to apply the static load through a steel spreader beam. Transducers were used to collect the displacement data at the midspan, loading points, and supports. A schematic of the transducer arrangement is shown in Figure 2.
After the bending test, concrete cores were drilled from the ends of the beams and the corroded rebars were obtained from the specimens. A 1% phenolphthalein alcohol solution was sprayed on the surface of the core as a pH neutralization indicator. e maximum and minimum pH neutralization depths as well as various median values were recorded, and the average value was taken as the neutralization depth. e concrete cores were axially loaded to obtain their compressive strengths. e rebars were loaded for tensile strengths after the corrosive ratios were checked.

Deterioration and Failure Mode Characteristics.
Cracks appeared on the top of the beams along the hanger bars, except in beams L10 and L11. e U-shaped hoops of strengthened beams L6, L8, and L9 peeled off when they were re-deteriorated for more than 18 months. Cracks also appeared on the sides of the RC beams along the direction of the tensile steel bars, with the maximum crack width exceeding 6 mm. e reinforcement layer, including the    Advances in Civil Engineering concrete cover, noticeably stripped from the RC beam but did not peel off with the effect of the U-shaped hoop. e deterioration characteristics of the RC beams and flexural loads are presented in Table 4, where f cu is the concrete strength after the test, X(t) is the neutralization depth, η is the rebar mass loss ratio, η is the weight loss ratio of the rebar, F u is the ultimate flexural load, Δ y and Δ u are the midspan deflection of the beam specimens under the yield and ultimate flexural load, respectively, and μ Δ is the ductility of RC beams. Figure 3 illustrates the deterioration before the test and failure modes of the test beam specimens after the test.
Under the ultimate load, all RC beams except beam L8 failed with the reinforcement layers peeling off together with the concrete cover. In the loading process, a noticeable sound was noticed when the CFRP wires pulled off, and the reinforcement layers broke suddenly as brittle failure occurred. e U-shaped hoops of beam L8 peeled off when the concrete was compressed; thus, the compressive area decreased heavily, and the concrete was compressively crushed.

Carrying Capacity.
e ultimate carrying capacity of the test beams is illustrated in Figure 4.
Except for L2, L7, L10, and L11, the carrying capacity of the re-deteriorated beams decreased considerably compared with that of beams L0 and L1, with a maximum reduction of 46%. As the protective covers of beams L10 and L11 were recasted before strengthening, their carrying capacities increased to close to or higher than those of beams L0, L1, and L2 and other beams. Several trends were observed. (a) With the increase of the deterioration time ti1, the carrying capacity of the strengthened beam decreased with the same redeterioration time ti2, the carrying strength of beam L4, L5, and L6 was less than that of beam L2 due to the effect of different former deterioration time ti1. (b) If the total deterioration time ti1 + ti2 was the same, the capacity of the strengthened beam decreased as the time ti1 increased, such as beams L6 and L7. e same phenomenon was observed in beams L10 and L11. (c) e sodium chloride introduced in the concrete mix speeded up the deterioration process which was observed by comparing L8 with other beams.

Load-Deflection Curves.
e load-deflection curves of test beams are shown in Figure 5.
As shown in Figure 5, the inflection points corresponding to the cracking loads cannot be determined on the load-deflection curves of the re-deteriorated beams. When the beams yielded, they maintained large rigidities till brittle failure.
is is mainly due to the mechanical character of CFRP, not to that of the steel bar. e elastic modulus of CFRP is greater than that of the steel bar, and thus, the strain of CFRP is smaller than that of the steel bar under the same load. e brittleness of the reinforcement layer was greatly enhanced as two layers of CFRP were attached to it. When the load reached the ultimate value, brittle fracture of CFRP occurred and the beam failed with minor noticeable deformation. e yield platform was relatively noticeable for the deteriorated beams because those beams were seriously deteriorated and entered the yield stage prematurely.

Concrete Strain in the Midspan.
Strain gauges were set up on the side and bottom of the test beams to collect the strain data during the static loading test. e collected data are shown in Figure 6.
Before the beam yielded, the concrete strain in the midspan of the RC beams increased linearly with the distance from the centre axis, which indicated that the plane section assumption was suitable for the carrying capacity calculation of the strengthened beams. As the load increased, particularly when the load exceeded the yielding load, the tensile strain increased continuously, whereas the compressive strain remained relatively constant until the beams failed. is is due to the cracks caused by the corrosion of the hanger steel bars, which led the concrete in the compression zone to be gradually peeled off with load increasing. As the strain gauges were destroyed by the cracks, the measured value corresponding to the ultimate load decreased to zero, as shown in Figures 6(c) and 6(e).

Tensile Strain of CFRP.
Five resistance strain gauges with a length of 50 mm were longitudinally attached to the pure bent section of the test beams to record the CFRP strain. e test results are shown in Figure 7. At the initial loading stage, the strain of CFRP increased with the load increasing, and the values were relatively similar at each measuring locations. When the beam cracked, the strain increased dramatically.
is increase became particularly large when the beam entered the yielding state. Under the ultimate load, large gaps appeared between the CFRP strains at different locations and the maximum strain value was not necessarily in the middle of the beam. e beams strengthened with CFRP had good durability to the compound effects of acid-salt mist and carbon dioxide. Due to the serious deterioration of the RC beam, concrete at some locations cracked and peeled off during the loading process. e CFRP strains at these locations were inconsistent with those at the other locations, resulting in local stress concentration and large strains. e CFRP strengthening layer, consisted of CFRP, a binder, and the concrete cover, was too brittle to bear the flexural load; therefore, the tensile strain of CFRP was not effectively utilized.

Experimental Analysis
Five failure modes of the RC beams strengthened with CFRP under the ultimate load have been reported in various literatures [12][13][14]. e five failure modes are as follows. (I) Concrete in the compression zone was crushed, while the tensile steel bars did not reach their yielding strengths, and the strain of CFRP did not reach its ultimate value. (II) Concrete in the compression zone was crushed, and the tensile steel bars reached their yield strengths, and the strain of CFRP did not reach its ultimate value. (III) e tensile steel bars yielded, and the strain of CFRP reached its ultimate value, while the concrete in the compression zone was not crushed. (IV) e strengthening layer, which consisted of a concrete protective cover, a binder, and CFRP, was sheared to peel off. (V) CFRP separated from the concrete cover of the RC beam. Failure modes (I) and (IV) occurred in this study.
As failure mode (I) occurred, several assumptions are adopted to calculate the carrying capacity of the strengthened RC beams.  Figure 8. e height of compressive zone x is less than the ultimate value x cr , and the strain of tensile steel bar ε s is less than the yield value ε y ; it is taken that ε s � ε y . e concrete strain ε c exceeds the ultimate value ε cu , and then, ε c � ε cu . e tensile strain of CFRP ε cf is less than the design value [ε cf ], which is equal to two thirds of the ultimate strain of CFRP, which has a value of 0.01. en, according to the assumptions, ε cf � [ε cf ], and the flexural carrying capacity of the strengthened RC beam can be calculated as follows: and x cr can be obtained by the following equation: where M u is the ultimate bending moment of the strengthened RC beam, A s is the cross-section area of the tensile steel bar, h 0 ′ is the effective cross-section height of the deteriorated RC beam, h′ is the cross-section height of the deteriorated RC beam strengthened with CFRP, b′ is the effective cross-section width of the deteriorated RC beam, E cf is the elastic modulus of CFRP, A cf is the sectional area of CFRP, φ cf is the effective utilization rate of CFRP, f yc ′ and f y ′ are the yield strength of the compressive and tensile steel bar, respectively, A s ′ is the cross-section area of the compressive (c) As the concrete cover of the beam L10 and L11 were lost, the grouting material was used to re-cast the protective layer firstly, and then, the beams were strengthened with CFRP like the others.  Advances in Civil Engineering steel bar, ε cf0 is the lagged strain of CFRP, and α 1 and β 1 can be obtained from the Chinese standard GB50367. As failure mode (IV) occurred, the CFRP was sheared off and the RC beam failed immediately. e strain of concrete in compression zone εc ≤ εcu, and the steel bar tensile strain εs ≥ εy, whereas the CFRP tensile strain examined that εcf was less than the ultimate value of [εcf]. CFRP shear failure typically occurred on the side of the U-shaped hoop close to the pure flexural zone. e carrying capacity of the strengthened RC beam can be calculated by the following formula:

Advances in Civil
where G c and G cf are the shear modulus of concrete and the CFRP strengthening layer, respectively, τ and τ cf are the shear strain of concrete and the CFRP strengthening layer, respectively, f v and f v ′ are the shear capacity of the tensile and compressive steel bar, respectively, and F S is the shear bearing capacity of the strengthened RC beam.
Based on equations (1) and (5), the carrying capacity of the strengthened RC beam is affected by the deterioration of the RC beam and the CFRP strengthening layer. us, the mechanism should be analyzed according to these two factors.

Deterioration of RC Beams.
e following have been observed as the RC beam deteriorated [12]:(a) the compressive strength of the deteriorated concrete decreased; (b) the local sectional geometry of the RC beam decreased as the steel bars corroded, leading to the cracked and peeling of the concrete; (c) the cross-sectional area of the corroded steel bars decreased, and the tensile strength decreased; and, (d) the bond performance between the deteriorated concrete and corroded steel bars degenerated.
As the concrete deteriorated, the compressive strength decreased and the failure mode of the RC beam changed. With the decrease of the concrete tensile strength, the confining force to the steel bars decreased. Furthermore, as the steel bar corroded and expanded, the bonding concrete on the steel bar was under radical compression. e width of microcracks in the concrete was widened, and the concrete cover cracked [13][14][15][16]. In the corrosive environment, the concrete deterioration and steel bar corrosion occurred at the same time and affected each other [12,17]. Tepfer [18] stated that the concrete around the steel bar should be divided into two parts:(a) a cracked inner part with an external diameter R i and an internal diameter R 0 , where R 0 is the initial diameter of the deformed steel bar; and, (b) an elastic outer part with an external diameter R c and internal diameter R i . Furthermore, R c + R 0 � c. As the concrete compressive strength reduction was coupled with the steel bar corrosion, R 0 increased, while R c decreased to zero. en, the bond between the concrete and steel bar decreased to a nonbond. As a result, the effective sectional height and rigidity of the RC beam decreased.

Deterioration of the CFRP Strengthening
Layer. Due to the poor infiltration of the epoxy resin to concrete, the CFRP, epoxy resin, and concrete surface layer produced a thin reinforcement layer. In the corrosive environment, the epoxy resin aged and changed the strengthening layer into a brittle sheet. Under the effect of moisture and temperature, the strengthening layer was stripped with the concrete of the RC beam [19][20][21]. Figure 9 illustrates the deflections of the strengthening layer and the steel bar under the ultimate load.
Ideally, the deflection of the strengthening layer should be equal to that of the steel bar, i.e., δ 1 � δ 2 and θ 1 � θ 2 � θ 3 � θ 4 . As the strengthening layer became brittle, the actual state was δ 1 > δ 2 , θ 1 > θ 3 , θ 2 > θ 4 , and θ 1 ≠ θ 2 . en, shear force was produced due to the difference between the deflections of the strengthening layer and steel bar. Under the constraint of the U-shaped hoop, the deflections of the strengthening layer and steel bar were the same in the position of the U-shaped hoop. us, reverse shear force was produced. erefore, shear failure occurred under the interaction between the two shear forces, and the normal tensile properties of CFRP were not utilized.    As the deterioration time increased, the RC beam and CFRP were further deteriorated, and the bond between them decreased continually. e displacement ductility coefficient of the strengthened beam became smaller as the RC beam deteriorated heavier during the time ti1.

Conclusions
(a) Under the joint effects of acid-salt mist and carbon dioxide, the deteriorated RC beam strengthened with CFRP was deteriorated again. Surface concrete was peeled off by tensile force generated by the U-shaped hoops. Stripping failure occurred in the strengthened layer in the tensile zone under the ultimate load. (b) As the deterioration time increased, the carrying capacity of the re-deteriorated beams decreased considerably compared with that of beams L0 and L1, with a maximum reduction of 46%. With the increase of the deterioration time t i1 , the carrying capacity of the strengthened beam decreased even though they had the same strengthening mode and re-deterioration time t i2 .With the same deterioration time t i1 , as the re-deterioration time t i2 increased, the carrying capacity of the beam decreased considerably with a value of 33-43%. If the total deterioration time t i1 + t i2 was the same, the capacity of the strengthened beam decreased as the time t i1 increased. (c) e plane section assumption was satisfied for the carrying capacity in the investigation of the beam specimens until they yielded. (d) e strengthened layer, which consisted of CFRP, a binder, and the concrete cover, was stripped from the RC beam during loading. e deflection of the strengthened layer behaved differently from the other part, and this disharmony prevented the mobilization of the tensile advantage of CFRP. (e) e conclusions of this study can be used as references in the prediction of strength changing and service life of strengthened RC beams.

Data Availability
e data used to support the findings of the study are included with the article.

Conflicts of Interest
e authors declare that they have no conflicts of interest.