DETERMINATION OF REQUIRED INSULATION FOR PREVENTING EARLY-AGE CRACKING IN MASS CONCRETE FOOTINGS

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin 0 0 0 0 DETERMINATION OF REQUIRED INSULATION FOR PREVENTING EARLY-AGE CRACKING IN MASS CONCRETE FOOTINGS Tu A. Do Post-Doctoral Associate Dept. of Civil and Coastal Engineering University of Florida P.O. Box 0, Weil Hall Gainesville, FL (Corresponding author) Tel: -0-0 E-mail: doanhtu@ufl.edu Adrian M. Lawrence Post-Doctoral Associate Dept. of Civil and Coastal Engineering University of Florida P.O. Box 0, Weil Hall Gainesville, FL Tel: -- E-mail: alawrence@ce.ufl.edu Mang Tia Professor and Graduate Coordinator Dept. of Civil and Coastal Engineering University of Florida P.O. Box 0, Weil Hall Gainesville, FL Tel: -- X E-mail: tia@ce.ufl.edu Michael J. Bergin State Structural Materials Engineer FDOT Materials Office 00 NE th Avenue,Gainesville, FL 0 Phone: -- E-Mail: michael.bergin@dot.myflorida.com Submitted: November 0 Word count:, words, Tables, Figures (, equivalent words) TRB 0 Annual Meeting

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin 0 ABSTRACT One of the methods for controlling heat of hydration of mass concrete structures is to insulate the poured concrete. Currently, no method can provide adequate insulation to prevent early-age cracking for mass concrete footings. This study developed a method for determining the required insulation for rectangular footings. The study included isothermal calorimetry testing of cementitious materials, field monitoring of temperature in footings, and finite element modeling. A fully insulated bridge pier footing constructed in the field in Florida was monitored for temperature development and for assessing the efficiency of insulation used. A parametric study consisting of finite element analyses was conducted on different footing shapes consisting of cubic shape, length:width:depth ratio of ::, and length:width:depth ratio of ::, to determine required insulation for footings with volume-to-surface area ratio ranging from. ft to. ft. From the obtained results, it was found that the Styrofoam insulation used for the specific monitored footing might be excessive, thus the actual insulation thickness of inches should be reduced to inch in order to reduce the construction cost. In general, with a volumeto-surface area ratio of less than.0 ft, under the same insulation condition, and using the same concrete mix, larger footings require a greater thickness of insulation. However, with a volumeto-surface area ratio of.0 ft or greater, larger footings only require a similar thickness of insulation to prevent cracking. The developed method for determining the required insulation for footings presented in this paper would be practice-ready for implementing in the field. TRB 0 Annual Meeting

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin 0 0 0 0 INTRODUCTION When Portland cement is mixed with water, heat is liberated. This heat is called the heat of hydration, the result of the exothermic chemical reaction between cement and water. The heat generated by the cement hydration causes a temperature rise in concrete. The exterior surface of concrete cools faster than the interior portion, resulting in temperature differential. Through thermal expansion and contraction, the temperature differential induces thermal (tensile) stresses at the surface. If the tensile stresses exceed the early age tensile strength of the concrete, cracking will occur (), (). Mass concrete is defined by the American Concrete Institute (ACI) as any volume of concrete with dimensions large enough to require that measures be taken to cope with generation of heat from hydration of the cement and attendant volume change, to minimize cracking (). In practice, mass concrete applies not only to dams but also to large foundations, columns, bridge piers, etc. The Florida Department of Transportation (FDOT) Structures Manual provides criteria for denoting mass concrete for all bridge components except drilled shafts and segmental superstructure pier and expansion joint segments as: When the minimum dimension of the concrete exceeds feet and the ratio of volume of concrete to the surface area is greater than foot, provide for mass concrete. (The surface area for this ratio includes the summation of all the surface areas of the concrete component being considered, including the full underside (bottom) surface of footings, caps, construction joints, etc.) (), (). To minimize thermal cracking problem in mass concrete, the Florida Department of Transportation Standard Specifications for Road and Bridge Construction requires that the maximum allowable temperature is 0 F ( C) and temperature differential between the concrete core and the exterior surface does not exceed F (0 C) (). In order to reduce the temperature differentials, and thus prevent cracking in mass concrete, some of the common methods such as precooling of concrete, post cooling of concrete with cooling pipes, and insulation or insulated formwork should be used. The thermal stresses in large mass concrete elements were effectively reduced with the use of thick layers of insulating polystyrene foam. This method is advantageous because the polystyrene foam, if removed carefully, can be reused often making it relatively inexpensive when compared to other single use methods such as cooling pipes and liquid nitrogen. Adequate insulation should be used in conjunction with the usual formwork material to reduce the temperature differentials during the early age hydration of massive concrete (). Do et al. () also showed that insulation should be used at the bottom of mass concrete footings in high ground water, and. in. ( mm) of insulation layer would be adequate to prevent early-age thermal cracking in a footing with dimensions of ft ft ft in. However, no recommended insulation thicknesses were made for footings with different dimensions. This paper presents follow-up research of the previous work () that addresses a method to provide adequate insulation for concrete footings to prevent cracking. The study included isothermal calorimetry testing of cementitious materials, field monitoring of temperature in footings, and finite element analyses. A parametric study consisting of finite element analyses was conducted to determine adequate insulation for rectangular footings of various dimensions. TRB 0 Annual Meeting

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin 0 TEMPERATURE MONITORING OF BRIDGE PIER FOOTINGS A ft ft ft-in.- bridge pier footing (Footing ) located at the S.R. (or Palmetto Expressway) and S.R. (or Dolphin Expressway) interchange in Miami, Florida was monitored for temperature developments in March 0 (). Temperatures in Footing were measured using 0 temperature sensors. (These temperature sensors had internal memories that can store the measured temperatures and did not require any external power source.) The predicted temperatures from the previously developed finite element model showed close agreement with those measured in the field (). Another bridge pier footing (Footing ) located at I- US- Braided Ramp, Orlando, Florida was also monitored for temperatures in July 0, as shown in Figure. Part of this field testing was to determine whether or not the insulation thickness used for the footing was adequate. The footing had dimensions of ft 0 ft ft- in. The concrete mix design used in Footing was a Class IV concrete mix with a total cementitious content of 00 lbs/yd, of which.% was Portland cement,.% Type F fly ash, and.% Boral Micron Ultra-Fine Fly Ash. The water-to-cementitious content ratio of this concrete mixture was 0.. 0 FIGURE Footing at I- US- Braided Ramp, Orlando, FL. The temperature sensors were placed at three locations: inches below the top surface, at the center, and inches above the bottom surface of Footing (on the vertical axis of symmetry). The reason the temperatures were monitored along the vertical axis is that the direction of fastest heat flow from the concrete center to the atmosphere will occur in the least dimension of the structure, which in this case is the vertical plane. Footing was fully insulated with inches of Styrofoam layer at its bottom, top, and sides. Styrofoam used for this footing had an R-value (thermal resistance) of.0 per inch. The concrete had a placement temperature of 0 F (. C). The average ambient temperature during monitoring period was. F (0. C). The temperatures recorded in the sensors are presented in Figure. The peak temperature in the pier TRB 0 Annual Meeting

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin footing was measured as F ( C) at the center of the footing hours after concrete placement. The maximum temperature differential recorded was. F (0. C) between the middle sensor and the top sensor. This temperature difference is significantly less than the allowable maximum temperature differential of F (0 C) required by FDOT, thus suggesting that the insulation used might be excessive. 0 Temperature ( F) 0 0 0 00 Bottom Sensor Middle Sensor Top Sensor 0 0 0 0 0 0 0 0 00 0 0 0 0 Time (hours) FIGURE Measured temperatures at top, middle, and bottom of Footing. FINITE ELEMENT MODEL The finite element model in this study was developed using the TNO DIANA.. software (). The finite element analysis utilizes the DIANA s Heatflow-Stress Staggered D feature, in which the thermal analysis is combined with a subsequent structural analysis. The model comprises two domains: one for the thermal flow analysis and one for the structural analysis (), (), (0). The finite element model developed in this study consists of a rectangular mass concrete footing lying on a soil layer. The concrete is insulated at the top, bottom, and all the sides. Based on the double symmetry of the rectangular footing, only one-quarter of the whole structure was to be modeled to reduce the computation time and the output file size from the DIANA software. The finite element mesh of one-quarter of the footing is illustrated in Figure. The modeled concrete had the full depth but half length and half width of the actual concrete of the footing. The modeled soil layer beneath the footing was. ft deep and extended. ft wider on each side of the footing in order to ensure adequate medium for heat transfer from the concrete. Although the actual reinforcing steel in the concrete can hold potential cracks caused by thermal contraction, and conducts heat rapidly, it was not modeled due to its complexity of geometry. Steel formwork was also not modeled since it has little protection against heat loss from the concrete. Shrinkage of concrete was not considered in this study. TRB 0 Annual Meeting

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin 0 0 FIGURE Finite element mesh of one-quarter of footing. The boundary conditions imposed for the thermal analysis consist of the initial temperature of the model and the external temperature. The initial temperature is the concrete temperature at the time of concrete placement, whereas the external temperature is set at the average ambient temperature. Figure illustrates the boundary temperatures imposed on the thermal finite element model. The external temperature is applied to the surfaces of the structure that are exposed to the environment including the outer surfaces of the concrete-insulation structure and the top surface of the soil layer. The fixed temperature is applied to the bottom and the sides of the soil layer. The fixed temperature is the same in magnitude as the external temperature; the only difference between these two loads is that there is air convection at the surface where the external temperature is applied while there is no convection at the surface where the fixed temperature is imposed. The boundary conditions imposed on the structural finite element model of the quarter footing consisted of the restriction of displacements against the symmetry planes. Since the depth of footing was relatively thick, it was assumed that there was no curling in the footing. Therefore, the base of the footing was modeled as being constrained against displacements along the z direction, but was free in the x and y directions as friction between the base of footing and soil was neglected. Isothermal calorimetry testing was conducted on the cementitious materials to obtain the energy released during hydration. The adiabatic energy rise was then converted to temperature rise that would be used as inputs for the finite element model (), (). TRB 0 Annual Meeting

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin 0 FIGURE External temperatures imposed on finite element model. PARAMETRIC STUDY A parametric study consisting of finite element analyses was conducted. Three different shapes of rectangular footings were considered consisting of cubic shape, length:width:depth ratio of ::, and length:width:depth ratio of ::, as shown in Figure. Each footing shape included footings with volume-to-surface area ratio (V/A) ranging from.0 ft to. ft. For instance, Footing had volume of, ft and surface area of 0 ft, thus it had a V/A of. ft. 0 Cubic Footing :: Footing :: Footing FIGURE Footing shapes. The modeled footing was fully insulated at its top, sides, and bottom with Styrofoam that had an R-value of.0 per inch. The modeled insulation thicknesses were 0. in., in., and. in. The thermal conductivity of the insulating material was determined from its R-value by the following equation: d k R () where k is thermal conductivity (BTU/hr-ft- F), d is thickness (ft), and R is thermal resistance (hr-ft - F/BTU). TRB 0 Annual Meeting

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin 0 0 The concrete properties used were the same as those of Mix that had water-tocementitious content ratio of 0. and the cementitious material consisted of 00% Type I Portland cement (0). The mechanical properties of the concrete were obtained from laboratory testing as shown in Table. The concrete had density of 0 pcf, Poisson s ratio of 0., and coefficient of thermal expansion of.0 0 - in./in.- F (). TABLE Mechanical Properties of Concrete Time (day) Tensile Strength (psi) Young s Modulus (ksi).,0 0.,0.,0 0.0,0 To assess the cracking potential of the modeled concrete, the following function called the crack index was used: ft ( t) Icr( t) () I ( t) where I cr is crack index, f t is tensile strength of concrete, σ I is maximum tensile stress, and t is time. If the crack index falls below.0, cracking is likely to initiate. Discussion of Results Tables through show dimensions, V/As, maximum temperature differences, and crack indices in Cubic Footings insulated with 0.-in Styrofoam, -in Styrofoam, and.-in Styrofoam, respectively. Those parameters were also obtained for the :: Footings and :: Footings using the finite element analyses. TABLE Temperatures and Crack Indices in Cubic Footings Insulated with 0.-in Styrofoam Length Width Depth Volume:Surface Max. Temperature Crack (ft) (ft) (ft) Area Ratio Difference Index (ft) ( F)....0....... 0...... 0...... 0...... 0.0.....0 0...... 0. TRB 0 Annual Meeting

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin 0 TABLE Temperatures and Crack Indices in Cubic Footings Insulated with -in Styrofoam Length Width Depth Volume:Surface Max. Temperature Crack (ft) (ft) (ft) Area Ratio Difference Index (ft) ( F)....0.0........... 0..0..... 0...... 0...... 0..... 0. 0. TABLE Temperatures and Crack Indices in Cubic Footings Insulated with.-in Styrofoam Length Width Depth Volume:Surface Max. Temperature Crack (ft) (ft) (ft) Area Ratio Difference Index (ft) ( F)....0..0.................................... The maximum temperature differences and crack indices in all the modeled footings were then compared together for each insulation thickness level. Figures through present the maximum temperature differences in the different footings fully insulated with 0. in.,.0 in., and. in. of Styrofoam, respectively. For each insulation thickness level, the maximum temperature differences in footings that had the same V/A were very close to each other regardless of their shapes. Figures through show the crack indices in the different footings for the cases of insulation with 0. in.,.0 in., and. in. of Styrofoam, respectively. Again, for each insulation thickness level, the crack indices in footings that had the same V/A were very similar regardless of their shapes. TRB 0 Annual Meeting

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin 0 Max. Temp. Difference ( F) 0 0 0 0 Cubic :: :: Volume:Surface Area Ratio (ft) FIGURE Maximum temperature difference in footings insulated with 0.-in Styrofoam. Max. Temp. Difference ( F) 0 0 Volume:Surface Area Ratio (ft) Cubic :: :: FIGURE Maximum temperature difference in footings insulated with -in Styrofoam. TRB 0 Annual Meeting

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin 0 Max. Temp. Difference ( F) 0 Cubic :: :: Volume:Surface Area Ratio (ft) FIGURE Maximum temperature difference in footings insulated with.-in Styrofoam. Crack Index. 0. Cubic :: :: 0 Volume:Surface Area Ratio (ft) FIGURE Crack Indices in footings insulated with 0.-in Styrofoam. TRB 0 Annual Meeting

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin Crack Index. 0. Cubic :: :: 0 Volume:Surface Area Ratio (ft) FIGURE 0 Crack Indices in footings insulated with -in Styrofoam. Crack Index.. Cubic :: :: 0. 0 0 Volume:Surface Area Ratio (ft) FIGURE Crack Indices in footings insulated with.-in Styrofoam. DETERMINATION OF REQUIRED INSULATION THICKNESS In the analyses presented above, the insulation thickness was increased from 0. in. until it was adequate to prevent cracking in the modeled concrete. Figure shows the crack indices in all the footings insulated with 0. inch of Styrofoam. Beginning from a value slightly above.0 at a V/A of.0 ft, the crack index dropped sharply to below.0 as the V/A increased. Therefore, 0. inch of Styrofoam did not provide adequate insulation for footings with a V/A of greater than.0 ft to prevent cracking. As shown in Figure 0, the crack index in a footing insulated with.0 in. of Styrofoam was greater than.0 at a V/A in the range of.0 ft and.0 ft, and was slightly below.0 as the TRB 0 Annual Meeting

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin 0 0 0 V/A became larger. Hence, one inch of Styrofoam was adequate for footings with a V/A of.0 ft or smaller to insure no occurrence of cracking in the concrete. Figure shows the crack index in footings insulated with. inches of insulation. There was a sharp drop in the crack index as the V/A increased from.0 ft to. ft. The crack index then remained almost constant at a value of. when the V/A became larger than.0 ft. Therefore, an insulation thickness of. inches was adequate for footings with a V/A up to.0 ft to prevent cracking in the concrete induced by thermal contraction. Table shows the required insulation thickness to prevent cracking and the maximum temperature differentials for the footings (using concrete of Mix ) with V/As ranging from. to. ft. It was found that the maximum temperature differentials varied according to the size of the mass concrete footings especially for V/As less than. ft. TABLE Required insulation thickness and maximum temperature differential for different V/As V/A R-value per in. Required Insulation Thickness Maximum Temperature Differential (ft) (in.) ( F)..0 0....0 0....0.0 0...0....0....0....0.. SUMMARY OF FINDINGS The main findings from the parametric study can be summarized as follows: The temperature differential in a concrete footing decreased with the increase in insulation thickness. Rectangular footings that had the same V/A but different shapes (dimensional proportions) would develop a similar maximum temperature difference and a similar crack index under the same insulation condition. With a V/A of less than.0 ft, under the same insulation condition, and using the same concrete mix, larger footings require a greater thickness of insulation, however, with a V/A of.0 ft or greater, larger footings only require a similar thickness of insulation to prevent cracking. Smaller footings allow slightly higher maximum temperature differential to prevent cracking by thermal contraction. When the concrete of Mix is used, footings with a V/A from. ft to. ft require a maximum allowable temperature differential of. F, footings with a V/A from. ft to. ft require a maximum allowable temperature differential of. F, and footings with a V/A from. ft to. ft require a maximum allowable temperature differential of. F. When Styrofoam with an R-value of.0 per inch and concrete of Mix were used, 0. inch would provide adequate insulation for a footing with a V/A of around.0 ft,.0 inch would provide adequate insulation for a footing with a V/A less than.0 ft,. inches would provide TRB 0 Annual Meeting

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin 0 0 0 0 adequate insulation for a footing with a V/A up to.0 ft. If another type of insulating material is used, an equivalent insulation thickness can be determined from the material s R-value. Since Footing had a V/A of. ft, inch of Styrofoam might be an adequate insulation thickness to prevent early-age cracking. CONCLUSIONS A method for determining the required insulation for mass concrete footings of various dimensions was developed using the finite element analyses. Based upon the results, it can be concluded that the Styrofoam insulation used for Footing might be excessive, thus the actual insulation thickness of inches should be reduced to inch. For insulation that cannot be removed and re-used such as bottom insulation of footing, using thinner and adequate insulation would reduce construction cost. It is recommended that less insulation should be used for smaller footings and the required insulation thickness should be determined using this method for concrete footings constructed in the field. It is also recommended that a database of adiabatic temperature rise of different cement blends used in typical FDOT mass concrete mix designs should be developed so that it can be used conveniently for prediction of temperature rises and needed insulation for mass concrete structures using the developed thermal and structural analysis software. ACKNOWLEDGMENTS The authors would like to acknowledge the tremendous efforts put forward by those who contributed to this project. Special acknowledgment is given to Dr. Yu Chen for her assistance in the testing of heat of hydration of the cementitious material. The study was made possible by funding provided by the Florida Department of Transportation. REFERENCES. ACI Committee 0. 0.R-0: Guide to Mass Concrete, Farmington Hills, MI USA, 00.. ACI Committee 0. 0.R-0: Report on Thermal and Volume Change Effects on Cracking of Mass Concrete, Farmington Hills, MI USA, 00.. ACI Committee. R-00: Cement and Concrete Terminology, Farmington Hills, MI USA, 000.. Florida Department of Transportation. FDOT Structures Manual. Tallahassee, FL, 00.. Florida Department of Transportation. Structures Design Guidelines: FDOT Structures Manual Volume. Tallahassee, FL, 0.. Florida Department of Transportation. Standard specifications for road and bridge construction. Tallahassee, FL, 00.. Tia, M., A. Lawrence, C. Ferraro, S. Smith, and E. Ochiai. Development of Design Parameters for Mass Concrete Using Finite Element Analysis. Final Report, Department of Civil & Coastal Engineering, University of Florida, Gainesville, Florida, 00.. Do, A. T., A. M. Lawrence, M. Tia, and M. J. Bergin. Importance of Insulation at the Bottom of Mass Concrete Placed on Soil with High Ground Water. In Transportation Research Record: Journal of the Transportation Research Board, Transportation Research Board of the National Academies, Washington, D.C., 0 (in press). TRB 0 Annual Meeting

Tu A. Do, Adrian M. Lawrence, Mang Tia, and Michael J. Bergin. DIANA: Displacement Analyzer Version... [Computer software]. TNO DIANA BV, Delft, Netherlands, 0. 0. Lawrence, A. M., M. Tia, C. Ferraro, and M. Bergin. Effect of Early Age Strength on Cracking in Mass Concrete Containing Different Supplementary Cementitious Materials: Experimental and Finite-Element Investigation. Journal of Materials in Civil Engineering, (), 0, pp... Ferraro C. C. Determination of Test Methods for the Prediction of the Behavior of Mass Concrete. Ph.D. Dissertation, University of Florida, Gainesville, FL, 00. TRB 0 Annual Meeting