Temperature Development and Heat Control Analysis In Palu Bridge Pile Cap And Pillar

2.1. Project Overview

The Palu IV Bridge Ponulele (Palu Bridge), located in the Central Sulawesi Province of Indonesia, was significantly damaged by the earthquake and tsunami. The Palu Bridge rehabilitation program was launched to restore the city of Palu's accessibility and mobility, which was impacted by the earthquake on September 28, 2018. The new bridge was constructed using the balanced cantilever structure, which involved bridge pillars and pile cap structures. These components were built as mass concrete structures. Fig. (1) describes the dimensions of the pile cap and pillar. The concrete grade was fc'35 MPa; the pile cap size was 17m × 17m × 4m, with the concrete volume reaching 1156 m³. The pillar had the same concrete strength as the pile cap; it was 11 m in height and 11.90 m in width and had a hollow section in its core.

2.2. Temperature Control Scenario

Managing thermal behavior in mass concrete is essential for structural integrity, especially in large structures like bridge pillars and pile caps. Mass concrete creates temperature gradients between the core and surface, which can lead to cracking and weakened structural integrity if the gradient is too high. Research (e.g., by Wang et al. [2] and Hui et al. [19]) shows that cracking is likely when thermal stress exceeds the concrete’s tensile strength. Temperature control strategies such as pre-cooling, staged pouring, and cooling pipes are implemented to manage risks effectively. Pre-cooling involves lowering the initial temperature using cold water or ice. Staged pouring, which is utilized in this case, helps limit temperature rises by alternating layers of poured material with cooling periods in between. Cooling pipes offer a quicker and more direct solution by circulating cool water, allowing for a single-stage pour while effectively managing the core temperature. Studies by Lu et al. [11], Sun et al. [20], and Xie et al. [10] support these methods' effectiveness in reducing temperature differentials. Finite Element Analysis (FEA) is crucial for accurate temperature modeling in mass concrete. This study used MIDAS FEA software (MIDAS) to simulate temperature profiles for staged pouring and cooling pipe scenarios, considering factors like ambient conditions and concrete properties. The simulations provided valuable insights and confirmed compliance with ACI standards, including a maximum core temperature of 71°C and a differential limit of 21°C. Both scenarios met these thresholds, demonstrating the effectiveness of temperature control strategies in maintaining concrete integrity.

2.3. Concrete Mix Design

The raw materials and mixing ratio of fc'35 MPa concrete for the pile cap and Pillar of Palu Bridge are shown in Table 3. The cement used in this mix design is Portland Composite Cement (PCC) type. PCC is formed by grinding Portland cement slag and gypsum with one or more inorganic ingredients. The inorganic components include blast furnace slag, pozzolan, silicate compounds, and limestone. These minerals collectively comprise 6% to 35% of the mass of PCC [23]. Generally, PCC generates relatively low to moderate heat during hydration and is strongly resistant to sulfate attacks. A superplasticizer and water-reducing agent were used to increase the concrete's workability.

2.4. Calculation Parameters and Model in Scenario 1

This section will examine the heat generated during the hydration process in concrete, explicitly focusing on the construction phase of the pile cap and pillar structure. The MIDAS will be used to simulate the heat of hydration using the transient heat hydration method (Fig. 2a). In addition, hexahedral meshing models were employed to combine elements with a maximum size of 250 mm x 250 mm (Fig. 2b). The bottom side is a meshing model representing soil material and acts as a heat absorber through hydration caused by the mass concrete. The simulation of this mass concrete is carried out in multiple stages. The pile cap concrete is poured in four stages, while the pillar is poured in eight stages. Each layer of the pile cap has a thickness of 1 meter, with a three-day interval between castings. Whereas the pillar casting occurs in 1.5-meter-thick layers, also with a three-day interval between each session. The pile cap and concrete pillars are treated as a single continuous structure. Several measures have been implemented to ensure the integrity between the pile cap and pillars, including roughening the surface, applying a bonding agent, and installing shear connectors.

Table 2 displays the thermal and mechanical parameters of concrete for the pile cap and pillar. The standard value for the Portland cement concrete mix ratio is determined based on the mechanical parameters of concrete. The specific heat, convention coefficient, coefficient of thermal expansion, and maximum heat of cement hydration were studied [24]. The ambient temperature was chosen based on the concrete and casting temperatures. The initial selected ambient temperature was calculated to be 27°C. The temperature of the concrete entering the mold was temporarily established at 27°C for this calculation. Temperature control in pile cap and pillar construction was determined based on the actual casting temperature of concrete. Adiabatic behavior in concrete could be defined using the “Function” menu in MIDAS. The function determine the highest temperature of the concrete mixture based on the 2012 JSCE standard [25]. This study utilizes standard plastic constitutive equations, effectively determining stress increments through the elastic component of strain increments. When materials are subjected to temperature changes, they experience thermal expansion or contraction, which can induce stress. If these stresses exceed the material's yield strength, plastic deformation occurs, and constitutive equations are used to model the transition from elastic to plastic behavior. This is particularly important in coupled thermo-mechanical analyses, where thermal gradients and mechanical forces interact.

After modeling and defining the material properties, a temperature load is applied to the model. This load includes the prescribed temperature, the heat source, and convection. The prescribed temperature represents the constant and uniform temperature of the exposed soil model. The heat source reflects the hydration heat generated from the concrete core based on the initial temperature outlined in the plan. Convection refers to the external air temperature, which helps dissipate the hydration heat from the concrete core.

2.5. Calculation Parameters and Model in Scenario 2

In the second scenario, concrete is cast in a single stage, requiring pipe cooling to manage excess heat in the pile cap. Due to high installation costs and time, the pipe cooling system is implemented only in the larger pile cap, which is 4 meters thick and features four layers of cooling pipes spaced 0.8 meters apart (Fig. 3).

The water in the cooling pipe shall be regarded as cooled water; its temperature shall not exceed 15°C; during the heating phase, the flow shall be 600 m³/hr. To meet these requirements, the inlet flow rate and temperature must be modified during the temperature control process. The physical parameter of the concrete used is the same as in scenario 1 (Table 2), while the pipe cooling specifications are explained in Table 3.

3.1. Stage Construction Scenario 1

The analysis results will display the heat development obtained in the pile cap and pillar elements. These results are displayed as heat contours from simulating each pouring stage. Therefore, the heat hydration output can be received at each construction stage. The heat development at the 1^st—5^th pouring stage is shown in Figs. (4-8), respectively.

Figs. (4-8) show that the maximum temperature occurs 72 hours after concrete pouring for the first pouring stage, whereas the second to fifth stage occurs at 36 hours. The longer time in the first pouring stage occurs because the bottom part of the concrete is laid on the soil while others stage on the concrete; the coefficient of thermal conductivity for soil is lower than concrete. Thus, it takes a longer time to absorb the heat. For example, at the first pouring stage, after 9 hours, the concrete shows a temperature of 37.9 °C at the top side and 40.9 °C at the core. Thus, after 72 hours, the concrete temperature rose to 49.2 °C at the top side and 66.9 °C at the core (Fig. 4); furthermore, after 88 hours (Fig. 5a), the concrete temperature at the core decreased to 51.9 °C. Therefore, the following pouring stage should be conducted after 72 hours, when the concrete reaches its peaks and starts declining afterward.

Figs. (4-8) show that the concrete core temperature from the first pouring cooled longer than other stages. It takes 240 hours from the peak temperature of 66.9 °C to cool down to the normal concrete temperature of 31–35 °C, whereas the second to fifth pouring stage takes less time than other stages. The longer time in the first pouring stage occurs because it keeps receiving thermal heat from the layers above during the pouring sequence. Therefore, the heat at its core could not be released in the free air.

The temperature of the concrete will be checked at each simulated pouring stage. The maximum temperature taken is divided into three specific locations: inside the concrete core, the top surface, and the side surface of the concrete. The temperature development of each pouring stage is shown in Table 4. Stages 1–4 were shown as the pouring stage for the pile cap, whereas stages 5 were for the pillars. The differential temperatures were taken from adjacent layers' highest and lowest temperatures.

After recapitulating the maximum temperature in the pile cap and pillar in Table 4, we obtained a maximum temperature value of 68.81°C, which does not exceed the requirement of 71°C [1]. Moreover, the highest differential temperature is 20.93°C, which does not exceed the requirement of 21°C [1]. Therefore, the staging strategy in the first scenario succeeded in keeping the heat of hydration during concrete hardening below the heat limitation set by ACI 301.

The additional cost of implementing Scenario 1 arises from applying a bonding agent to the hardened concrete surface before the next concrete pour. This ensures proper integration between the old and new concrete. The unit cost of epoxy-based bonding agent application is $25 per m². For a pile cap surface of 289 m², with four stages of application, the total cost is $28,900.

Scenario 1, which employs staged pouring for temperature control, is a cost-effective method compared to alternatives like cooling pipe systems. This approach avoids the need for specialized equipment or materials, relying instead on careful scheduling and standard materials, making it suitable for budget-constrained projects. However, the method is time-intensive, requiring waiting periods between pours (e.g., 72 hours) to allow sufficient cooling of previous layers, which can prolong the construction timeline. It is particularly applicable in projects where extended timelines are acceptable and ambient conditions favor gradual cooling. Despite its cost benefits, the method has limitations, including the risk of thermal accumulation in base layers, complex scheduling requirements, and dependence on environmental factors such as temperature and wind. These factors may restrict its effectiveness in projects with tight deadlines or regions with high ambient temperatures that hinder efficient cooling. Overall, Scenario 1 offers a low-cost solution for managing thermal stress but is best suited for projects with flexible schedules and favorable external conditions.

3.2. Construction Stage Scenario 2

Scenario 2 was employed to accelerate the concrete casting process for the pile cap. This scenario introduces two primary changes compared to Scenario 1: adding a cooling pipe and modifying the casting process from four stages to a single stage. In scenario 2, the pile cap is cast in a single continuous operation to a height of four meters. This approach is expected to reduce the overall casting time.

Fig. (9) displays the temperature development of pile cap concrete pouring within 24 and 48 hours. The temperature development shows constant values at 24 and 48 hours because the casting stages were conducted in one stage. Moreover, the cooling pipe was installed inside the concrete, where the heat was absorbed by pumping the water inside the pipe at a constant rate. The maximum temperature at 48 hours is shown in Table 5. The maximum temperature value was obtained at 58.69°C, which does not exceed the requirements of 71°C [1]. At the maximum differential temperature, the value obtained is 13.16°C, which does not exceed the requirement of 21°C [1]. Therefore, the pipe cooling strategy in the second scenario succeeded in keeping the heat of hydration during concrete hardening below the heat limitation set by ACI 301.

The additional cost of implementing Scenario 2 is due to the application of a concrete cooling pipe system, which consists of steel pipes, a pump, and concrete grouting poured into the pipes after the process is complete. The 4-layer steel pipes have a total length of 544 m, with a unit cost of $30 per meter for the installation and materials of 2” SCH 40 steel pipes, resulting in a total piping cost of $16,320. The pump rental for 2 days is $1,500. The grouting required to fill the pipes amounts to 1.5 m³, with a unit cost of $250 per m³, resulting in a total grouting cost of $375. This brings the additional cost for scenario 2 to $18,195.

Scenario 2, utilizing cooling pipes, effectively controls heat in mass concrete structures, enabling faster single-stage casting and making it ideal for large, time-sensitive projects. It suits environments with extreme temperatures but requires skilled personnel, reliable energy, and significant upfront costs. While efficient, its higher expense and complexity may limit use in budget-constrained or smaller projects.

3.3. Concrete Physical Modeling

A small-scale physical model was constructed for this study to validate the FEA results. The physical model consisted of a concrete block measuring 1,5 meters in width, 1,5 meters in length, and 1,5 meters in height (Fig. 10). The concrete mix design followed the specifications in Table 1. Thermocouples were installed at 10 cm from the surface (top position) and 50 cm from the surface (core position). As shown in Fig. (10)., the platform height was set to 1.5 meters to accommodate the pouring stage in scenario 1, with concrete poured in 1-meter stages.

Temperature readings were recorded every 12 hours at both the top and core positions, and the results are presented in Table 6. The peak temperature observed in the physical model was 66.7°C, with a differential temperature of 16.5°C. These values are slightly lower than those obtained from the FEA, which showed a maximum temperature of 68.81°C and a differential temperature of 20.93°C. The slight differences can be attributed to the continuous pouring assumed in the FEA simulation, where the heat generated from the previous stage is cumulatively added to subsequent pours, leading to slightly higher predicted temperatures.

Although no large-scale physical prototype was constructed for this study, the methodology aligns with proven approaches validated in prior research, specifically studies by Lu et al. [11] and Sun et al. [20], which used thermocouples and field measurements during construction, supporting the use of FEA for accurate thermal predictions. Similarly, Tahersima and Tikalsky [21] demonstrated strong correlations between FEA results and onsite temperature readings, confirming the reliability of the numerical method.