Experimental study and field application of calcium sulfoaluminate cement for rapid repair of concrete pavements

Abstract

The fast-track repair of deteriorated concrete pavement requires materials that can be placed, cured, and opened to the traffic in a short period. Type III cement and Calcium Sulfoaluminate (CSA) cement are the most commonly used fast-setting hydraulic cement (FSHC). In this study, the properties of Type III and CSA cement concrete, including compressive strength, coefficient of thermal expansion (CTE) and shrinkage were evaluated. The test results indicate that compressive strength of FSHC concrete increased rapidly at the early age. CSA cement concrete had higher early-age and long term strength. The shrinkage of CSA cement concrete was lower than that of Type III cement concrete. Both CSA and Type III cement concrete had similar CTE values. Based on the laboratory results, the CSA cement was selected as the partial-depth rapid repair material for a distressed continuously reinforced concrete pavement. The data collected during and after the repair show that the CSA cement concrete had good short-term and long-term performances and, therefore, was suitable for the rapid repair of concrete pavement.

Keywords

Calcium Sulfoaluminate (CSA) cement / Type III cement / coefficient of thermal expansion (CTE) / shrinkage / rapid repair

Introduction

Spalling is one of the most common distresses of concrete pavement and could cause ride as well as noise problems for the traveling public. It can occur as early as within two years after pavement construction and tends to propagate under repeated thermal stress and/or traffic loading. Also, horizontal delaminations are often observed at spalled areas [1]. If left unrepaired, these distresses would lead to the eventual dislodging of concrete and become hazardous to the traveling public [1]. Partial-depth repair (PDR) is a common and effective method for rehabilitating localized spalls, restoring functional performance of pavement, deterring further deterioration, and extending pavement life. As a preventive maintenance technique, PDR also offers advantages of less construction cost and time, shorter lane closure and less damage to the existing pavement [2,3]. The effectiveness of PDR depends largely on the properties of repair materials. Desirable properties for patching materials include high workability, rapid strength development, short curing time, strong bond strength, long-term durability, thermal compatibility with existing concrete, and reasonable cost [1].

Nowadays, with the increasing traffic volume, delays during pavement repair are becoming less tolerated while many numerical analysis on pavement have been conducted [4,5]. Therefore, high early strength, which is required for early opening of pavement to traffic, becomes a key factor for patching material [6]. There are several fast-setting hydraulic cements (FSHC) available, including Calcium sulfoaluminate (CSA) cement and Type III cement. Previous studies show that CSA, compared with ordinary Portland cement (OPC), has advantages of high early-age strengths [7–9], short setting times, impermeability [10–12], high sulfate and chloride corrosion resistance and low alkalinity. Moreover less CO₂ will be released during the production of CSA [13–17]. Type III Portland cement is designed to develop early strength more quickly than Type I cement. Both CSA and Type III cement concrete could develop sufficient strength to carry traffic within 6–24 h after finishing [6].

Except for the high early strength development, the volume change induced by shrinkage and thermal contraction is also very important from the point of view that it could cause debonding between the repair material and existing pavement and accelerate the failure of the patch [18–20]. However, limited studies were performed to evaluate the shrinkage and thermal properties of FSHC concrete. Caltran had conducted a detailed study about using FSHC concrete for jointed plain concrete pavement. The strength and shrinkage properties of CSA cement were fully studied in the laboratory and filed. The field test showed that cracking could occur due to the combined effect of drying shrinkage and nigh time temperature gradients [21–24].

In this study, the CSA and Type III cements were first evaluated in the laboratory. The cement with better properties was then selected and applied for the rapid repair of spalls of continuously reinforced concrete pavement. In order to evaluate the performance of the FSHC concrete patch in the field, the temperature and strain sensors were embedded to monitor the development of pavement temperature and stresses at different locations.

Experimental program

Experimental materials

Type III and CSA cements were selected for this study. The major chemical compositions are presented in Table 1. The CSA cement contains 13.7% Al₂O₃ and 12.5% SO₃. These two values are higher than that of Type III cement, indicating faster strength gain and high durability .

The coarse aggregates used for both CSA and Type III cement concrete mixes were Pea gravel. The maximum aggregate size was 2 inches. The gradation is shown in Table 2. The natural river sand was used as fine aggregate. The physical properties of coarse and fine aggregates are listed in Table 3. The citric acid retarder and high range water reducer (HRWR) were used to adjust the workability and setting time of concrete.

Mix design

The concrete mix proportions were determined in accordance with ACI 211.1-91 [25]. The dosages of retarder and HRWR of each mix were adjusted to keep concrete slump between 4 and 6 inches (101 to 152 mm). As listed in Table 4, for CSA cement concrete, three different cement/water ratios, 0.45, 0.50, and 0.55, and three levels of cement content, 6, 6.5 and 7 sacks, were employed. The mix proportions of concrete with 6 and 6.5 sacks of Type III cement and W/C ratio of 0.45 are shown in Table 5.

Laboratory experiments

The compressive strength development was investigated according to ASTM C39-03 (2003) [26]. Immediately following the mixing, the concrete was placed into the 4 inch×8 inch cylinders and then cured in the standard curing room with a constant temperature around 73.5±3.5°F and 98% relative humidity (RH). The compressive strengths were tested at 3 hours, 4 hours, 6 hours, 24 hours , 3 days, 7 days, and 28 days.

The shrinkage and CTE of FSHC concrete were measured to evaluate the volume change property. For concrete shrinkage, test was conducted based on ASTM C157-08 (2005) [27]. The 3 inch×3 inch×11.25 inch prism specimens were cast and kept in the environmental chamber with 73°F temperature and 50% relative humidity (RH). The length changes of the specimens with time were measured with digital length comparator. Concrete CTE test was carried out according to AASHTO TP-60 (2000) [28] with one exception. Instead of being placed in the water bath, specimens were sealed with aluminum foil and put in the chamber with controlled temperature of 73°F (23°C) and 50% RH. During the testing, chamber temperature changed cyclically from 73°F (23°C) to 127°F (53°C). The specimens’ temperatures were measured by the thermocouples and length changes were monitored by crack meters (Fig. 1). The CTE was calculated based on the temperature and length change.

Laboratory testing results and discussion

Compressive strength development

Figure 2 shows the compressive strength development of FSHC concrete with a constant W/C ratio of 0.45. For the CSA cement concrete, the compressive strengths increased quickly at very early age and reached over 70% of the 28-day strength at only one day. After three days, the compressive strength development slowed down and became stable. The 4-hour compressive strengths of 6 and 7 sacks CSA cement concrete were more than 3000 psi (13.79 MPa), which meet the requirement for pavement opening to traffic. The compressive strength of Type III cement concrete increased slower than that of CSA cement concrete. Since the tests on the Type III cement concrete cylinders before 1 day failed, the compressive strengths had been illustrated from 1 day in Fig. 2. Concrete gained about 70 percent of the 28-day compressive strength at three days.

From the standpoint of cement content, the highest compressive strength of CSA cement concrete belonged to six sacks cement content, while specimen of Type III cement concrete showed a higher result when the cement content was 6.5 sacks. These results show the cement content affects the strength of the concrete, but there is not a consistent relationship between the strength value and cement content. More research is required.

The effect of W/C ratio on the compressive strength of concrete with 6 sack CSA cement is shown in Fig. 3. There was a consistent relationship between the strength and the W/C ratio, which was that the lower the W/C ratio, the higher the compressive strength was. The results were consistent with Abrams’ W/C ratio law [29] “For the given materials, the strength of the concrete depends on a sole factor- W/C ratio”.

Shrinkage

Figure 4 shows the shrinkage of CSA and Type III cement concrete with W/C ratio of 0.45. Obviously, the shrinkage of CSA cement concrete was much lower than that of Type III cement concrete, indicating lower risk of cracking in the field. At 56 day, the shrinkage of CSA cement concrete was less than 38% of that of the Type III cement concrete. Meanwhile, the concrete shrinkage increased with the increase of cement content. However, the difference in shrinkage caused by cement content was much smaller than that caused by cement type.

The effect of W/C on the shrinkage of six sacks of CSA cement content is shown in Fig. 5. As the W/C increased, higher shrinkage was observed. The effect of W/C was also less significant than that of cement type.

Coefficient of thermal expansion

CTE of concrete is an important parameter in analyzing thermally induced stresses in pavement during the first 72-hours after paving and over the design life. High CTE of repair material could cause large deformation under temperature change and result in debonding issue between the repair material and surrounding concrete [29]. The variations of temperature and strain during the CTE testing were plotted in Fig. 6. The CTE value, which is the slope, was determined by performing the linear regression. The CTE values of the CSA cement and Type III cement concrete were 6.77 and 6.64 microstrain per °F, respectively, which were in the range of suggested values [30]. Since coarse aggregate makes up the bulk of the volume of concrete, the most influential factor in the CTE of the concrete is the CTE of the coarse aggregate [30,31]. Both concrete contained the same type and amount of coarse aggregate, therefore, the CTE values were very close. This is consistent with existing findings [32].

Application of CSA cement forCRCP repair

Based on the laboratory testing results, the CSA concrete was more suitable for rapid repair of spalls. Since the field environment is more complicated than controlled laboratory conditions, the performance of in-situ CSA concrete has to be evaluated. A distressed CRCP section, located at US 290 in Hempstead, Texas, was selected for the field test (Fig. 7). The visualized and sonic inspection showed that it would be a partial-depth distress. To repair the spall, researchers first determined the repair boundaries, and then cut and removed the concrete to the depth of the longitudinal steel.

Surface preparation and gage installation

Most problems associated with failure of a PDR are not because the material does not have sufficient strength. It is usually a failure of the bond between the repair material and the surrounding concrete. In order to improve the bonding condition at the bottom surface, four pieces of #5 steel bars were drilled into the existing pavement (Fig. 8). Two of the bars were hook-shaped, and the other two were straight. The centers of the 7-inch bars were located right at the interface, with three steel strain gages attached on each of them. Three thermocouples were placed at 0.5 inches, 2.5 inches, and 4.5 inches from the pavement surface to monitor the slab temperature distribution. The arrangement of sensors was illustrated in Fig. 9. The CR1000 data logger was used to collect strain and temperature data for 14 days.

The 6 sacks CSA cement concrete with W/C of 0.5 was used for the spall repair. Once the gages and thermocouples were installed, the CSA cement concrete was mixed and cast on site. The placement started at 0:15 pm and finished at 1:48 pm. At the same time, 4 inch×8 inch specimens were prepared for the compressive strength, which was tested 1 hour and 16 minutes later (3:04 pm) at the Houston district laboratory. The compressive of repair CSA cement concrete was 3560 psi, which exceeded the minimum compressive strength requiement for pavement openning to traffic.

Figure 10 shows the temperature variaton at different depth with time. During the first seveal hours after placement, the concrete temperature increased rapidly, indicating quick hydration of the CSA cement concete. After that, the temperature decreased and cycled daily. The diurnal temperature variation was the greatest near pavement surface due to the quick heat exchange with surroudings. The temperature change at bottom was the smaller. The temperature gradients along the depth that can lead to curling and warping movements and stresses.

Field performance of the repaired pavement

Based on the steel strain data investigated, the steel stress of the bars was calculated by following equation and presented in Fig. 11.

(1){\sigma }_{\mathrm{bar}}=E{\epsilon }_{\mathrm{bar}}.

Since the strain gage at the interface location of S1 was not functioning properly, only the tresses measured at 0.75 inch and 1.5 inches from the interface were presented in Fig. 11(a). During the early hours of CSA cement hydration, although the concrete temperature increased a lot, there was no stress developed because concrete at this stage was still plastic and allow deformation. After concrete hardened, the compressive stress occurred in the steel due to the negative temperature gradient. The periodic variation of temperature will cause concrete expand and contract cyclically, which induced cyclic change of the steel bar stress. As shown in the figure, as temperature decreased, the stress increased and vice versa. What should be noticed is that stress of bars at 0.75 inch to the interface was larger than that of bars at 1.5 inches to the interface. This indicates that the interface was in good bonding condition during the testing period. Figure 11(b) includes three steel bar stress at loaction S2. The stress change at S2 had the similar trend as that at S1. The stress distribution along the depth was not linear and decreased from the interface to the top of pavement surface.

Since the data collected from H1 is odd and discrete due to the malfunction of the installed strain/temperature sensors, the collected data is not presented in this paper. The data of H2 show that the compressive stresses in the hook bar changed with temperature. The maximum values decrease from the interface to the top surface of the repair slab. Fluctuations in the measured stress were more pronounced in the hook bar data than in the straight bar data. This phenomena can’t be explained by the data of this testing. More tests are required to explore the mechanism.

After two years, there were no distresses found in the patch based on the periodic observation, indicating that the CSA cement concrete is the appropriate material for the rapid concrete pavement repair.

Conclusions

This paper studied the strength and thermal properties of CSA and Type III cement concrete. Based on the laboratory tests, CSA cement concrete was selected for the spall repair in the field. The stress and temperature inside the patch were monitored. The primary findings and recommendations are summarized as follows:

1) Compared with Type III cement concrete, CSA cement concrete had higher early age and long-term compressive strength. The early age strength gain of CSA cement was faster than that of Type III cement concrete. For the CSA cement concrete, the strength is strongly related to the W/C ratio. The higher the W/C ratio, the lower the compressive strength was.

2) The shrinkage of CSA cement concrete was lower than that of Type III cement concrete.

3) Both CSA and Type III cement concrete had similar CTE values due to the same type and amount of coarse aggregate used in both concrete. Therefore, compared with Type III cement concrete, CSA cement concrete will not cause more displacement and stress in the pavement.

4) The application of the CSA cement concrete in the rapid repair for spalls was proved to be feasible. The collected data show that the existing pavement and repair material were in good bonding. However, the performance of the straight and hook bars needs more study in the future.

References

[1] Chen D, Lin  H, Sun R . Field performance evaluations of partial-depth repairs. Construction & Building Materials, 2011, 25(3): 1369–1378

[2] CalTran CDOT. Maintenance Technical Advisory Guide Volume II- Rigid pavement preservation (2nd Edition). Caltrans Division of Maintenance, 2007

[3] ACPA. Guidelines for partial-depth spall repair. Concrete Paving Technology. American Concrete Pavement Association, 2007, 15p

[4] Hou Y, Sun  F, Sun W ,  Guo M, Xing  C, Wu J . Quasi-brittle fracture modeling of preflawed bitumen using a diffuse interface model. Advances in Material Science and Engineering, 2016, 2016(6): 1–7

[5] Hou Y, Wang  L, Yue P ,  Pauli T ,  Sun W. Modeling mode I cracking failure in asphalt binder by using nonconserved phase-field model. Journal of Materials in Civil Engineering, 2014, 26(4): 684–691

[6] Buch N, Van Dam  T J, Peterson  K, Sutter L . Evaluation of high-early strength PCC mixtures used in full depth repairs. Construction & Building Materials, 2008, 22(3): 162–174

[7] Quillin K. Performance of belite–sulfoaluminate cements. Cement and Concrete Research, 2001, 31(9): 1341–1349

[8] Glasser F P, Zhang  L. High-performance cement matrices based on calcium sulfoaluminate–belite compositions. Cement and Concrete Research, 2001, 31(12): 1881–1886

[9] Odler I. Special Inorganic Cements. CRC Press, 2000

[10] García-Maté M ,  De la Torre A G ,  León-Reina L ,  Aranda M A G ,  Santacruz I . Hydration studies of calcium sulfoaluminate cements blended with fly ash. Cement and Concrete Research, 2013, 54: 12–20

[11] Cau Dit Coumes C ,  Courtois S ,  Peysson S ,  Ambroise J ,  Pera J. Calcium sulfoaluminate cement blended with OPC: a potential binder to encapsulate low-level radioactive slurries of complex chemistry. Cement and Concrete Research, 2009, 39(9): 740–747 doi:10.1016/j.cemconres.2009.05.016

[12] Zhou Q, Milestone  N B, Hayes  M. An alternative to Portland Cement for waste encapsulation—The calcium sulfoaluminate cement system. Journal of Hazardous Materials, 2006, 136(1): 120–129

[13] Aranda M, De la Torre  A G. Sulfoaluminate cement. Eco-Efficient Concrete, 2013, 488–522

[14] Alaoui A, Feraille  A, Steckmeyer A ,  Roy R L . New Cements for Sustainable Development. 12th International Congress on the Chemistry of Cement, 2007, 1–8

[15] Gartner E. Industrially interesting approaches to “low-CO₂” cements. Cement and Concrete Research, 2004, 34(9): 1489–1498

[16] Popescu C D, Muntean  M, Sharp J H . Industrial trial production of low energy belite cement. Cement and Concrete Composites, 2003, 25(7): 689–693

[17] Sharp J H, Lawrence  C D, Yang  R. Calcium sulfoaluminate cements-low-energy cements, special cements or what? Advances in Cement Research, 1999, 11(1): 3–13 doi:10.1680/adcr.1999.11.1.3

[18] Hou Y, Wang  L, Yue P ,  Sun W. Fracture failure in crack interaction of asphalt binder by using a phase field approach. Materials and Structures, 2015, 48(9): 2997–3008

[19] Sun R J, Won  H, Won M C . The application and early-age behaviors of continuously reinforced bonded concrete overlay of distressed jointed concrete pavements. Journal of Testing and Evaluation, 2011, 39(5): 208–215

[20] Thomas J J. The Science of Concrete. Report submitted to the Infrastructure Technology Institute for TEA-21, 2010

[21] Hou Y, Wang  L, Pauli T ,  Sun W. Investigation of the asphalt self-healing mechanism using a phase-field model. Journal of Materials in Civil Engineering, 2015, 27: 040141183

[22] Heath A C, Roesler  J R, Harvey  J T. Modeling longitudinal, corner and transverse cracking in jointed concrete pavements. International Journal of Pavement Engineering, 2003, 4(1): 51–58 doi:10.1080/102984303100016073393

[23] Heath A C, Roesler  J R. Top-down cracking of rigid pavements constructed with fast-setting hydraulic cement concrete. Transportation Research Record, 2000,  1712: 3‒12 

[24] Heath A C, Roesler  J R. Shrinkage and Thermal Cracking of Fast Setting Hydraulic Cement Concrete Pavements in Palmdale, California. Report for California Department of Transportation, 1999

[25] Committee A. Standard Practice for Selecting Proportions for Normal, Heavyweight and Mass Concrete. Reported by ACI Committee 211, 1991, 1–91

[26] ASTM. Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens. ASTM C39-03. West Conshohocken, 2003

[27] ASTM. Standard Test Method for Linear Shrinkage and Coefficient Ecpansion of Chemical-resistant Mortars, Grouts, Monolithic Surfacings, and Polymer Concrete. ASTM C531-05. West Conshohocken: ASTM international, 2005

[28] AASHTO: Standard Method for the Coefficient of Thermal Expansion of Hydraulic Cement Concrete. AASHTO TP60-00, 2000

[29] Won M. Improvements of testing procedures for concrete coefficient of thermal expansion. Transportation Research Record, 2005, 1919(1): 23–28

[30] Won M C, Kim  S M, Merritt  D, Mccullough B F . Horizontal Cracking and Pavement Distress in Portland Cement Concrete Pavement. International Air Transport Conference, 2002

[31] Abrams D A. Design of Concrete Mixtures, Vol. 1: Structural Materials Research Laboratory. Lewis Institute, 1919

[32] Yang H H. Pavements Analysis and Design. Prentice Hall, 2004

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg