Free software allows accurate prediction of maximum temperature and temperature differential

Source: CONCRETE CONSTRUCTION MAGAZINE
Publication date: 2006-11-01

By Jon Poole, Kyle Riding, Ralph Browne, and Anton Schindler

Thermal gradients in bridge elements were historically ignored in the United States, since the columns, footings, and bent caps were often relatively small. But as element size has increased for structural, traffic, and aesthetic reasons, thermal gradients and thermal cracking have become serious concerns for bridge engineers. Many mass concrete projects specify a 35° F maximum temperature differential, and some limit the maximum internal temperature (usually 160°F). Contractors and engineers on these projects are increasingly called upon to develop a temperature control plan to meet these specifications.

Unfortunately, even though the intent of the specification is well understood, the validity of the 35° F maximum temperature differential is questionable. Remember, the goal of these specifications is to minimize and control cracking. Thermal analysis of mass concrete elements alone will not directly control cracking risk. Clearly, a more unified approach to engineering mass concrete to prevent thermal cracking and improve long-term performance is needed. Research* suggests that the cracking risk of mass concrete can be lowered by a variety of methods, including:

• Reduction of the fresh concrete temperature
• Use of a larger maximum size aggregate
• Use of aggregate with a low coefficient of thermal expansion
• Use of crushed aggregate instead of smooth, round aggregate
• Replacement of cement with fly ash, slag, or other suitable supplementary cementitious materials (SCMs)
• Entrained air
• Reduction of cement content and paste content

*Springenshmid, R., and R. Breitenbücher, “Influence of Constituents, Mix Proportions and Temperature on Cracking Sensitivity of Concrete,” Prevention of Thermal Cracking in Concrete at Early Ages, Edited by R. Springenschmid, RILEM Report 15, EF Spon, London, 1998, pp. 40–50.

Unfortunately, there hasn't been a good way to quantify the effects of each of these methods. There is no easy answer when a contractor asks, “If I use a crushed limestone and reduced cement content in my mixture, will this meet the placement temperature specification?” Being able to answer such a question would improve the performance and economics of mass concrete.

In October 2001, the Texas Department of Transportation (TxDOT) was interviewed by CONCRETE CONSTRUCTION magazine for an article on mass concrete. After publication of that article, TxDOT took a critical look at both the existing mass concrete specification and ACI Committee 207's guidelines for mass concrete construction. TxDOT recognized the need for a tool to easily predict the performance of concrete mix designs for mass concrete applications.

Smaller “less massive” members can also be classified as mass concrete.

In September 2002, TxDOT initiated a research project to develop software to perform the temperature analysis of mass concrete elements. Within six weeks, the project grew to encompass a full-blown concrete mixture design, analysis, and performance prediction program named Concrete Works. The investment in the project grew from $ 400,000 to $1,300,000, and the duration from three years to five years. The goal of Concrete Works was to give laboratory technicians, engineers, and contractors one tool that combines concrete design, analysis, and performance predictions to improve and guide TxDOT to better designs. One additional outcome of the research may be a change to the 35° F maximum temperature differential criteria dictated in the specification. The change could be switching to a gradient in lieu of the differential restriction and relaxation of the differential or gradient due to various material influences on cracking. The results and suggested modification to the specification will be presented to the TxDOT specification committee at the completion of this research project.

The Concrete Durability Center located at the University of Texas at Austin (UT) submitted the winning proposal for this project. The research team is composed of Dr. Kevin Folliard (UT, principal investigator), Dr. Anton Schindler (Auburn University), Dr. Maria Juenger (UT), Dr. Mike Thomas (University of New Brunswick, Canada), and Dr. Loukas Kallivokas (UT).

Concrete Works models the three possible ways heat may be transferred—conduction, convection, and radiation. The conduction portion uses nonlinear material thermal properties that are dependant on the materials selected, time, and temperature. The convection portion accounts for the roughness of different surfaces (concrete, wood, steel, blankets) and the effect of wind speed. The radiation portion models the solar radiation, surface shading (for example, the bottom of a bent cap would be shaded), atmospheric radiation, ground-emitted radiation, and irradiation from the concrete member's surface.

Together, these modules allow the user to quickly and easily model different construction sequences and materials to ultimately build better, longer lasting, and hopefully more cost-effective concrete members. Mixture-specific heat of hydration values are used to accurately model the effect of various cementitious materials on the in-place concrete temperature distribution. The model has been calibrated with over 33,000 hours of temperature data collected from 12 concrete members instrumented in Texas.

Some of the unique tests performed as part of this project are semi-adiabatic calorimetry (250 tests performed), isothermal calorimetry (over 1000 tests performed), cracking frames (75 tests performed), rapid chloride permeability (ASTM C 1202, 300 tests performed), compressive strength (ASTM C 39), splitting tensile strength (ASTM C 496), modulus of elasticity (ASTM C 469), and time of set (ASTM C 403).

Estimating the risk of thermal cracking is not as simple as breaking a few cylinders, or performing a few calculations on the back of an envelope. Extensive testing and computational models are needed to assess this complex phenomenon.

In mass concrete, the tensile strength increases as the hydration of the cementitious system progresses and is strongly affected by the type of cementitious material, the water-to-cementitious materials ratio, the aggregate type, and the maturity level of the hardening concrete. The restraint that leads to cracking is primarily due to the thermal gradient that exists between the core and the surface of the structure. This thermal gradient causes differential thermal stress that may exceed the tensile strength capacity of the concrete.

The development of in-place stresses are influenced by the effective modulus of elasticity (a time-dependent measure of stiffness development), the coefficient of thermal expansion, setting characteristics, restraint conditions, and temperature history of the hardening concrete. Trying to quantify these variables is complicated at early ages, and many have complex interactions.

For the TxDOT research project, an early-age cracking frame test is being used to evaluate the 35° F maximum temperature differential criteria. The susceptibility of hardening concrete to cracking can be assessed by this cracking frame that restrains a concrete specimen against movement. These frames are designed to allow fresh concrete to be cast into molds within each frame, which enable the study of very early-age behavior of concrete mixtures. Shrinkage effects can be assessed in actual concrete specimens, and the information will be used in ConcreteWorks to quantify the cracking sensitivity of various mixtures used in mass concrete applications.

This cracking frame test measures the cracking susceptibility of the concrete mix.

ConcreteWorks Version 1.0 is a stand-alone Microsoft Windows-based software package released in January 2005. The current version is free and can be downloaded at www.texasconcreteworks.com. This user-friendly software has been designed to run on a desktop computer. The version currently available performs the following functions:

• Predicts the in-place temperature history of mass concrete members. Geometric configurations that are typically used by TxDOT are used as default shapes with adjustable dimensions. The effect of an array of different combinations of cement, ground-granulated blast-furnace slag, fly ash, silica fume, and chemical admixtures can be evaluated.
• Estimates the in-place strength development through the use of the maturity method (ASTM C 1074).
• Estimates the effects of formwork type, curing methods, insulating blankets, liquid nitrogen, and ice addition.
• Estimates the effects of weather. The National Oceanic & Atmospheric Administration's 30-year average ambient weather conditions are embedded for weather centers in all 50 states. This data is used to determine hourly climatic data for any time of placement year-round.
• Proportions the concrete mixture in accordance with the procedures of ACI 211.
• Checks TxDOT specifications for: Maximum concrete temperature difference between the central core of the placement and the exposed concrete surface, which may not exceed 35° F.; Alkali silica reaction (ASR): Checks that the total alkali contribution from the cement does not exceed 4.0 pounds/yd3 for plain portland cement mixtures. Seven other ASR mitigation options are provided when different SCMs and chemical admixtures are used; Delayed ettringite formation (DEF) (ettringite is a byproduct of the cement hydration reaction): Checks the maximum in-place temperature. For mass placements, the temperature at the central core of the placement may not exceed 160° F. For precast members, the concrete temperature may not exceed 150º F and 170° F for plain portland cement mixtures and mixtures that contain SCMs, respectively.

The final version of ConcreteWorks, due in September 2008, will perform the following additional functions:

• Predict the in-place temperature of prestressed concrete members, bridge decks, and concrete pavements.
• Integrate models for corrosion prediction.
• Classify (low, moderate, or high) the cracking tendency in mass concrete applications.
• Make the program compatible with TxDOT project management software, SiteManager (an AASHTO product).

On the first screen of Concrete-Works (see page 48), the general inputs screen, the user starts by choosing the closest geographic location available to the placement location, the date of placement, and the time of placement. On the next screen, typical structural shapes are presented and the user selects the shape that most closely matches the application and the boundary conditions (for example, is there water, soil, or air on the outside of the element?).

The user picks the mixture proportions on the next series of screens. Several tabs allow the concrete to be designed according to ACI 211. The chemistry of the cement, and type of coarse and fine aggregate are defined on these screens. Choose a default average or check the box to modify cement composition according to your mill sheet

Next, the user defines the placement temperatures, ice addition, formwork type, formwork color, formwork stripping times, and other construction-related inputs. Wood, steel, insulated, and lined formwork are all possible formwork types that the user can select.

On the next few screens, the user defines the environmental conditions associated with the concrete placement: air temperatures, cloud cover, relative humidity, and solar radiation values are required for accurate predictions. To start, historical average data for the time and day of placement will be generated, but the user can override these values should more specific data be available. The user can view graphs of each of the environmental variables.

The next screen displays checks of the most important input values. Input values are compared to practical ranges and unreasonable values are flagged in red, while normal inputs are flagged in green. This user-friendly feature ensures that no mistakes or unreasonable assumptions have been made with the inputs.

The next screen produces the results in a multitude of formats: text, graph, or animated. The maximum concrete temperature, the minimum concrete temperature, the ambient temperature, and the maximum temperature difference are all displayed. The user can get estimates of the strength development over time if the parameters that uniquely define the strength-maturity relationship are input. The temperature distribution in the cross section can also be animated versus time.

Currently, ConcreteWorks is being beta tested and specified in some TxDOT offices. The results from several large projects under construction in central Texas have shown that the software can predict temperatures within 10% of the measured field results. Updates to the software will be added as work progresses and will be posted at the Web site, www.texasconcreteworks.com. Contact Jennifer Moore (EIT) at Jmoore5@dot.state.tx.us for additional information.

<i>— Jon Poole and Kyle Riding are civil engineering Ph.D. candidates at the University of Texas at Austin. Ralph A. Browne, P.E. is the North Tarrant Area engineer (North Fort Worth) with the Texas Department of Transportation; and Dr. Anton Schindler, P.E., Ph.D. is the Gottlieb assistant professor of civil engineering at Auburn University, Ala.</i>