Effects of insulation materials on mass concrete with pozzolans

Construction and Building Materials 137 (2017) 261–271

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Effects of insulation materials on mass concrete with pozzolans Yuan-Yuan Chen a,⇑, Ssu-Yu Chen b, Chien-Jou Yang a, Hei-Tao Chen c a

Department of Architecture, Hwa Hsia University of Technology, New Taipei City, Taiwan, ROC Department of Civil and Construction Engineering, National Taiwan University of Science and Technology, Taipei City, Taiwan, ROC c Registered Professional Civil Engineer of Taiwan, General Manager of HWUSU Engineering Consultants Co., Ltd., Taipei City, Taiwan, ROC b

h i g h l i g h t s
The maximum temperature rise and maximum temperature differential are examined.
Insulation materials: included XPS, EPS, and PE boards.
The IMTC (Ko), temperature–time gradient (k), and R-value are examined.
The correlations of Ko vs. k, Ko vs. R-value, and Ko vs. R-value are analysed.
A 1-in. XPS board enhances efficiency on mass concrete for top and side isolation.

a r t i c l e

i n f o

Article history: Received 31 August 2016 Received in revised form 28 December 2016 Accepted 18 January 2017

Keywords: Mass concrete Insulation Temperature differential Insulation material temperature conductivity (Ko) Temperature–time gradient (k) Thermal resistance (R-value) Extruded expanded polystyrene board (XPS)

a b s t r a c t A mix incorporates cement(II), fly ash, and slag to design mass concrete. A simplified box is set up to measure Ko. Two blocks are established to evaluate the insulation with thin sides (1/2-in.PE; top: none) and thick sides (1-in.XPS and 1-in.EPS; top: 1-in.XPS) and wrapped with canvas. Results for Ko were 1-in. XPS < 1-in.PE < 1-in.EPS < 1/2-in.PE. The maximum temperature differentials are less than 12 °C. The insulation effect of the face was 1/2-in.PE < 1-in.EPS < 1-in XPS, and the top 1-in.XPS enhanced insulation effectively. The values of R2 for Ko vs. k, k vs. R-value, Ko and vs. R-value were 0.995, 0.987, and 0.966 to assess appropriate insulation materials. Ó 2017 Elsevier Ltd. All rights reserved.

1. Insulation and materials ACI 301 in ACI 116R defines mass concrete as any volume of concrete with dimensions large enough (at least 4 in. (120 cm)), where measures are required to cope with the generation of heat from hydration of the cement (more than 660 lb/yd3 (390 kg/m3)) and attendant volume changes in order to minimise cracking [1–4]. ACI 207.2R [5] states ‘All concrete elements and structures are subject to volume change in varying degrees, depenAbbreviations: XPS, extruded expanded polystyrene board; DEF, delayed ettringite formation; FDOT, Florida Department of Transportation; GDOT, Georgia Department of Transportation; HRWRA, high-range water-reducing agent; IMTC, insulation material temperature conductivity; EPS, expanded polystyrene; PE, polyethylene; SCC, self-consolidating concrete. ⇑ Corresponding author at: Department of Architecture, Hwa Hsia University of Technology, Gongzhuan Rd., Zhonghe Dist., New Taipei City 235, Taiwan, ROC. E-mail address: [email protected] (Y.-Y. Chen). http://dx.doi.org/10.1016/j.conbuildmat.2017.01.059 0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.

dent upon the makeup, configuration, and environment of the concrete’. For sizes of 5 ft. (1.5 m) or greater and lateral form dimensions of around 8–10 ft. (2.4–3 m), the temperature rise is essentially adiabatic in the central part of the mass of fresh concrete. At exposed surfaces (formed or unformed), the heat generated is dissipated into the surrounding air at a rate dependent upon the temperature differential (gradient). Therefore, the net temperature rise in concrete adjacent to the surface (or forms) is less than that in the interior. While this results in a gradually increasing temperature gradient from the surface to the interior, little or no stress (or strain) is developed because the concrete is not yet elastic. ACI 207.4R-05 ‘Cooling and Insulating Systems for Mass Concrete’ [6] states that the need to control volume change induced primarily by temperature change in mass concrete often requires cooling and insulating systems. For mass concrete, to protect the temperature gradient of the surface through core temperature not exceeded too many to induce cracking [7,8].

262

Y.-Y. Chen et al. / Construction and Building Materials 137 (2017) 261–271

Mass concrete has two major important thermal characteristics, the temperature rise and the temperature differential. An excessive temperature rise (over 160 °F (71.1 °C)) may induce thermal cracking and delayed ettringite formation (DEF) [9–13]. However, in the Florida Department of Transportation (FDOT) [14] and in the Georgia Department of Transportation (GDOT) [15], the maximum temperature rise for mass concrete is up to 160 °F (71.1 °C). Due to the thermal movement restraint of structural elements for mass concrete, the temperature differential between adjacent positions (a point at the centre of a section and the adjacent surface) should not be greater than 20 °C between the centre of the concrete section and the surface. Otherwise, it may be constrained to increase the interior, exterior, or both restraints so as to enhance the probability of cracking [7,8]. In the GDOT, the maximum temperature differential is limited to 36 °F (20 °C) [15]. The thermal characteristics of mass concrete will affect its hardened properties and its long-term durability (only if it cracks). Insulation or insulated formwork is often used to retain heat at a concrete surface and reduce the temperature differential, which in turn reduces the potential for thermal cracking. For most mass pours, surface insulation does not appreciably increase the maximum concrete temperature, but it can significantly decrease the rate of cooling. Insulation is inexpensive, but delays resulting from the reduced cooling rate can be costly. Insulation often has to remain in place for several weeks or longer. Removing it too soon can cause the surface to cool quickly (causing thermal shock) and crack. A thermal resistance (R-value) of 4.0 h ft2 °F/Btu (0.70 m2 K/W) has been found to be effective for moderate climates. This is provided by a 1-in.-thick expanded synthetic material such as polystyrene or urethane. According to ACI 207.4r-05, acceptable temperature gradients can be maintained during the winter season in moderate climates by the application of insulation with an R-value of 4.0 h ft2 °F/Btu (0.70 m2 K/W). In severe climates, insulation with a R-value of 10.0 (1.76 ft2 K/W) is recommended. Many types of insulation materials (such as black blankets, polyethylene (PE; 1-in. R-3.0) boards, expanded polystyrene (EPS; 1-in. R-4.0) boards, extruded expanded polystyrene (XPS; 1-in. R-5.0) boards, canvases, etc. are available, and insulation levels can be optimised to meet the required temperature differentials to maximise the rate of cooling and to improve the construction economy [6,16–20]. A maximum allowable temperature differential of 36 °F (20 °C) is often specified in specifications for mass concrete. This temperature difference is a general guideline based on experience with unreinforced mass concrete installed in Europe more than 75 years ago. In many situations, limiting the temperature differential to 36 °F (20 °C) is overly restrictive; thermal cracking may not occur even at a higher temperature differential. In other cases, significant thermal cracking may still occur even when the temperature differential is less than 36 °F (20 °C) [1,17]. The maximum allowable temperature differential is a function of the mechanical properties of the concrete, such as thermal expansion (coarse aggregate), tensile strength, and elastic modulus, as well as the size and restraints of the concrete element. Some have suggested that, for concrete with gravel, granite, and limestone coarse aggregates, the temperature difference limit should be 36 °F (20 °C), 45 °F (25 °C) and 56 °F (31 °C), respectively. The coarse aggregate should also be from a single source to limit temperature differentials in order to avoid cracking [19]. Options open to an engineer seeking to limit heat generation of concrete include: (a) the use of Type I, Type II (moderate heat, MH), Type IV (low heat, LH), and Type V Portland cement, with specific maximum heat options to limit hydration if necessary (as covered by ASTM C150); (b) the use of blended hydraulic cements (Types IS, I(PM), IP, P, and I(SM), as covered by ASTM C150, C595, and C1157), which exhibit favourable characteristics for heat of hydration, which may be more firmly achieved by imposing limits on

heat of hydration for the cement clinker; (c) the use of hydraulic cements (Types GU, MS, HS, MH, and LH, as covered by ASTM C1157), and (d) the reduction of cement content by using a pozzolanic material, either fly ash or slag or both, to provide a reduction in maximum temperatures produced without sacrificing the long-term strength development and durability. When cement is used with pozzolans or with other cements, the materials are batched separately at the mixing plant, and improved economy and low temperature rises are both achieved by limiting the total cement content to as small an amount as possible. Type I and GU cements are suitable for use in general construction. However, they are not recommended for use alone in mass concrete without other measures that help to control temperature problems, because of their substantially higher heat of hydration [5,6,21]. Structural concrete that uses ternary binders (cement, slag, and fly ash) is a more recent application in Taiwan. Internationally, most mass concrete uses binary cementitious materials, such as cement and fly ash or ground granulated blast-furnace slag. A large portion in the USA uses all three. For example, in Canada and some states of the USA, it has been suggested to adopt ternary binders, including cement, fly ash, and slag, for the mixture of mass concrete. A total cement substitution of 50% [22–24] slag and fly ash is licensed by the Departments of Transportation of Iowa, Kentucky, West Virginia, and Florida [14]. For mass concrete in Canada [25], in order to significantly reduce the heat of hydration, it is recommended to use a minimum of 50% fly ash Type F or CI, or a mixture of both, or a minimum of 65% fly ash Type CH or slag, or a mixture of both, in hot weather (above 81 °F (27 °C)). Fly ash Type F, CI, and CH have CaO contents of less than 8%, 8–20%, and greater than 20%, respectively. Lawrence et al. [26] used a 3-D finite element analysis to study the effect of early age strength on cracking in mass concrete containing different supplementary cementitious materials, including a mixture with cement, fly ash (20%), and slag (30%). This paper examines mass concrete used in a new large medical building located in the city of Kaohsiung in southern Taiwan (note that the weather here is similar to that in Florida and Hong Kong). The concrete structure had non-reinforced mass concrete and was designed temperature steel reinforcement. The thickness of the mass concrete for a wall structure serving as a proton radiation engineering barrier was between 5 and 12 ft., and temperature differential was limited to 25 °C (construction took place below 20 °C). Before construction at the job site, a test model (block 1 or 2) of the mass concrete was selected in order to find the most locally appropriate insulation materials that conform to the requirements of controlled temperature and prevention of cracks. 2. Materials and methods The cementitious materials included Type-II Portland cement (ASTM C150; not Type-II (MH) cement), ground granulated blastfurnace slag (Grade 100, ASTM C989), and Class-F fly ash (ASTM C618). The properties of the binders are as shown in Table 1. The coarse and fine aggregates comprised crushed stone and natural sand. The maximum aggregate size was 3/4 in., and the fineness modulus of the sand was 2.81. The properties comply with ASTM C33. The high-range water-reducing agent (HRWRA) was ASTM C494 Type G. The water used was potable water. 3. Mixture design and experiments 3.1. Mixture design The specifications for the mass concrete are as follows:
The water-to-cementitious ratio (W/CM) less than 0.45;

263

Y.-Y. Chen et al. / Construction and Building Materials 137 (2017) 261–271 Table 1 Cementitious material properties. Component

Portland cement(II)

Slag (Grade 100)

Class F Fly ash

SiO2 Al2O3 Fe2O3 CaO MgO SO3 K2O N2O Specific gravity (Gs)

21.5 4.8 3.1 62.4 2.9 2.1 0.7 0.2 3.14

34.0 14.7 0.3 42.0 6.3 0.4 0.3 0.2 2.87

51.3 24.3 6.2 6.3 1.6 0.6 – – 2.21


Cement (Type II not II(MH)) content of 305–338 lb/ft3 (180– 200 kg/m3);
Pozzolans-to-binder ratio (P/B) less than 50% (±5%), Class-F fly ash no more than 25%, slag (Grade 100) less than 30%;
Slump flow greater than 20–26 in. (500–650 mm) at 90 min;
The specified strength (f’c) is 4000 psi (28 MPa) at 28 days. In order to match these specifications, in the laboratory, trial mixes were made in various batches in order to find suitable proportions, as shown in Table 2. The mixture was then provided to a ready-mixed concrete plant and the mixture was tested repeatedly. In Taiwan, a mix with ternary binders has been used for structural concrete for several decades, but it has not been as commonly applied to mass concrete. Internationally, there has been little application in this area.


Temperature measuring sensors shall be installed at five points or more (two at the core, two on the top surface, and one ambient).
The pour height shall not exceed 8 ft. 3.2.1. Insulation box and materials In Taiwan, there is a lack of experience with regards to the construction specifications and insulation materials used for mass concrete. The site engineering staff use a simple experiment to assess the properties of the local insulation materials. Before the experiments, a simplified insulation box (2-ft.  2-ft.  2-ft.) was set up as shown in Fig. 1(a). Fig. 1(a) shows an insulation box with the interior space wrapped with aluminium foil (protected and isolated between the box and a 100 W bulb) and various locally appropriate insulating materials to be tested. Three thermometers were installed. One was passed through the top and fixed, another was placed on the insulation material surface, and the other was in the ambient environment. Fig. 2 shows photos for the insulation material temperature conductivity (IMTC; K0) test and various insulations. In Fig. 2(a), a 100W bulb is set inside the box. Insulation material was placed on the top and a thermometer reading of the surface (Co) was made. Then the light bulb was turned on and the time taken for the box’s core temperature to reach 75 °C was recorded. As heat passed through a thickness (t) of insulation material, the surface thermometer reading (C) and the ambient temperature were recorded. A watch was used to record the initial time (T0) and the time taken for the temperature to reach 75 °C (T). Fig. 2(c) shows the K0 experiment for 1in. XPS board. The simplified IMTC (K0) through the insulation material can be identified using the following formula:

3.2. Experiments The temperature control specifications for construction are shown below:
For the first five days, the maximum temperature differential between the centre of mass (core) and the centre of top (face) shall be less than 25 °C.

K0 ¼

C  C0 DC ¼ ðT  T 0 Þ  t DT  t

ð1Þ

Here, K 0 is the IMTC (°C/(minin), DC is the temperature differential (°C), DT is the time differential (min), and t is the thickness of insulation material (in.). Fig. 1(b) shows a simplified relationship between time and temperature, with a linear relationship between (C 0 ; T 0 ) (the initial

Table 2 Mix proportions.

* **

W/CM

Materials, lb/ft3 (kg/m3)*

0.42

Cement 330 (195)

Fly ash 135 (80)

Slag 195 (115)

Stone 1452 (858)

Sand 1620 (957)

Water** 276 (163)

HRWRA 7.9 (4.68)

1 lb/m3 = 0.591 kg/m3. Water included superplasticisers.

(a) A simplified insulation box

(b) A simplified time–temperature curve

Fig. 1. Schematic diagram of a (a) simplified insulation box for isolation and (b) simplified time–temperature curve for the temperature conductivity test.

264

Y.-Y. Chen et al. / Construction and Building Materials 137 (2017) 261–271

(a) Insulation box setup

(b) Covered with AF

(c) XPS board test

(d) Template (1/2 in.)

(e) EPS board (1 in.)

(f) PE board (1/2 in.)

(g) PE board (1 in.)

(h) XPS board (1 in.)

Fig. 2. Actual photos of the simplified insulation box setup and various insulation materials.

stage) and (C T ; T) (when the box’s core temperature reached 75 °C and time), as formula 1. This assumption is not the true behaviour, but it is close enough for practical application, because the maximum temperature rise of mass concrete is estimated to range from 70 °C to 80 °C. This is a simple way to find the approximate performance of a locally appropriate insulation material. On a job site, model blocks are tested with light and heavy insulation materials to find the insulation performance. The insulation materials for the experiment were a PE board (t = 1/2, 1 in.), EPS board (1 in.), XPS board (1 in.), template

(1/2 in.), and aluminium foil (6.5 lm), as shown in Fig. 2(d)–(h), respectively. 3.2.2. Experiment The experiment on the simplified temperature conductivity will yield a reasonable guideline for insulation materials. In order to select suitable insulation materials for construction work, two model blocks with dimensions of 6 ft.  6 ft.  6 ft. were established. One had light insulation, including a template (1/2 in.) and a PE board (1/2 in.), which were adjacent to the centre of the

(a) Assembly of thermocouple sensors (two top, four face, two centre).

(b) Set up of template and wrapped faces with PE board.

(c) Wrapping faces with 1-in. EPS and XPS boards.

(d) Wrapping the top with 1-in. XPS board (thick).

(e)Wrapping the whole blocks with canvas.

(f) Automatic thermocouple temperature recorder.

Fig. 3. Photos of experimental procedure and set-up for a model block.

Y.-Y. Chen et al. / Construction and Building Materials 137 (2017) 261–271

face; it was then paved with a wet non-woven fabric and the whole block was wrapped in canvas. The other used heavy insulation materials, including an EPS board (1 in.) and a XPS board (1 in.), which were adjacent to the centre of the face; it was then paved with a wet non-woven fabric, the top was wrapped with XPS board (1 in.), and the whole block was covered with canvas, as shown in Fig. 3. The temperature sensor was embedded 4 in. below (or within) the surface at the centre of the top face and side. An automatic recorder device (SUPCON-R3000) was connected to the eight sensors, as shown in Fig. 3(f), and was calibrated prior to use. Eight sensor locations (points 1–8) were arranged on each block, including the top (points 1 and 2, middle and quarter positions), centres of adjacent faces (point 3 and point 6 in one group, point 4 and point 5 in the other group), and the centre of mass (points 7 and 8, core and a quarter position corresponding to the top surface).

265

The ambient temperature sensor was allocated at point 9 of block 1 to replace point 3, as shown in Fig. 4(a). Fig. 4(b) shows the set-up and tests of both blocks being carried out at the same time, before installing the insulation materials. Prior to pouring concrete, each block was first assembled by a template and then the reinforcement and auxiliary steel were configured. Afterwards, the measuring sensors were installed (locations: centre of top, centre of mass, centre of faces and ambient) to connect the thermocouples to the automatic recorder. Finally, the temperature control device was checked and tested. Upon completion of the above process, the construction and measurement procedure required the concrete to be placed, covered with a wet non-woven fabric, and wrapped with a canvas as insulation before obtaining measurements from the temperature sensors (with a seven-day duration). The workability of slump and slump flow spread were tested according to ASTM C143.

(a) Temperature sensor locations

(b) The setup of various insulation materials Fig. 4. Diagram of the setup of temperature sensor locations and various insulation materials for model blocks 1 and 2.

266

Y.-Y. Chen et al. / Construction and Building Materials 137 (2017) 261–271

4. Results and discussion

100

Model I

point1

90

4.1. Insulation materials

80

The experimental results for the IMTC and price are shown in Table 3. The lower the IMTC (K0), the higher the effect of insulation. Table 3 shows the IMTC (K0) values of all insulation materials examined in this paper. The values from lowest to highest are in order of 1-in. XPS board, 1-in. PE board, EPS board, and 1/2-in. PE board. The 1-in XPS board appears to produce the best insulation. The 1-in. PE board and 1-in. EPS board produced almost identical insulation effects. The unit price of a 1-in. PE board was around twice that of a 1-in. XPS board, and the IMTC of the 1/2-in. PE board (1.803 °C/(minin.)) was five times higher than that of the 1-in. XPS board (0.330 °C/(minin.)). Although the 1-in. EPS (0.483 °C/(minin.)) and 1-in. PE (0.457 °C/(minin)) boards had similar IMTCs, a job site still must consider economy and environmental friendliness. Therefore, it appears that the 1-in. XPS board is the most suitable option (see Table 3, IMTC and price), but this is only a reference guide and still requires a further validation test. 4.2. Temperature rise The temperature rise is influenced by the cement type, mix design, pozzolans, pouring temperature, etc. The temperature rise is mainly dominated by the heat of hydration of cementitious materials, especially cement. The hydration curve of a concrete mixture with ternary binders, including Portland cement, slag, and Class-F fly ash, is similar to that of a binary system, but the peak temperature and occurrence time are different. The temperature–time curves of the mix are illustrated in Figs. 5 and 6. The core temperature (point 7) was the highest of all sensor locations. The maximum temperatures of blocks 1 (Model I) and 2 (Model II) were, respectively, 71.6 °C and 73.3 °C. The difference between these is 1.7 °C. The error in the maximum temperature rise of both blocks is 2.4%. These results show that the variance in the maximum temperature rise due to the insulations applied appear very limited. Mixture pouring was performed at the same time, but the insulation materials applied on the tops (points 1 and 2) and centres of faces (points 3 and 6) of the two blocks were different. The heavy insulation of block 2 had a higher insulating effect than the light insulation of block 1. In particular, the maximum temperature rise for the top of block 2 exceeded 72.5 °C, which was higher than that of block 1, 65.2 °C. For the centres of faces of the blocks, the differences between block 1 (66.1 °C for PE, 66.5 °C for Template) and block 2 (67.3 °C for EPS, 69.3 °C for XPS) were not significant compared to those on the top. The maximum temperature of both blocks exceeded 71.1 °C (160 °F), which is the greatest limitation of the international specifications for mass concrete. This may lead

Maximum temperature rise : 71.6 oC

point2

point4

point5

point6

point7

point8

point9

70 60 50 40 30 20 10 0 0

20

40

60

80 100 Time(hour)

120

140

160

180

Fig. 5. The temperature–time curves of block 1 for all of the various sensor locations. (The sudden decline in the curve is due to the power supply instability at the site in the summer).

100

Model II

point1

90 80 Maximum temperature rise : 73.3 oC

point2

point3

point4

point5

point6

point7

point8 point9

70 60 50 40 30 20 10 0 0

20

40

60

80 100 Time(hour)

120

140

160

180

Fig. 6. The temperature–time curves of block 2 for all of the various sensor locations. (The sudden decline in the curve is due to the power supply instability at the site in the summer). At 144 h (6 days), the sudden decline in the curve is due to removal of the insulation materials and contact with the atmosphere (similar to thermal shock). In this case, the monitoring duration was only for 5 days. The data after 5 days were used as a reference.

Table 3 Results of the simplified IMTC (K0) test and price for the locally insulation materials.

*

NO

Type*

To (min) °Co (°C)

T (min) °CT (°C)

DT = T  To (min) DC = °CT  °C0 (°C)

Ko** = DC/(DTt) (°C/minin)

Unit price*** US$ (NT$)

1

EPS (1 in.)

0.457

5.04 (161.1)

3

PE (1/2 in.)

1.803

2.60 (83.1)

4

XPS (1 in.)

13 6 11 5 13 11 14 4.5

3.47 (111.1)

PE (1 in.)

14:56 30 14:33 29 15:25 34 15:51 28.5

0.483

2

14:43 24 14:22 24 15:12 23 15:37 24

0.330

2.99 (95.6)

XPS: extruded expanded polystyrene; EPS: expanded polystyrene; PE: polyethylene. ** IMTC = DC/(DTt) (°C/(minin); DC: °CT  °C0; DT: T  To; t: thickness of insulation material; °CT (T) and °C0 (To): when box core temperature reached 75 °C, the thermometer at insulation material surface measured temperature (time(75 °C)) and initial temperature (time(0, ambient temp.)). *** Exchange rate: US$1 = NT$32.

267

Y.-Y. Chen et al. / Construction and Building Materials 137 (2017) 261–271 Table 4 Results of temperature control for blocks 1 and 2 with maximum temperature rise and time durations. Block 1

Temp. rise

Location

Top (B/2) Point 1

Top (B/4) Point 2

Template Point 3

Template Point 6

1/2-in. PE Point 4

1/2-in. PE Point 5

Core Point 7

Max. (°C) Time (h)

65.2 51.0

64.1 51.0

– –

66.5 52.0

63.3 51.0

66.1 52.0

71.6 52.0

Location

Top (B/2) Point 1

Top (B/4) Point 2

1-in. EPS Point 3

1-in. EPS Point 6

1-in. XPS Point 4

1-in. XPS Point 5

Core Point 7

Max. (°C) Time (h)

72.5 55.5

69.6 55.5

67.3 55.5

67.6 55.5

69.3 55.5

69.3 55.5

73.3 55.5

Block 2

Temp. rise

to the issue of delayed ettringite formation (DEF). However, it was still less than 180 °F (81.1 °C) [14]. This specification only provides the temperature differential without the maximum temperature. The behaviours (microscale and macroscale) above 160 °F of mass concrete with a higher dose of pozzolans (40–50% fly ash or 50–75% slag) require further study. The maximum temperatures and the times to reach them for blocks 1 and 2 are shown in Table 4. According to the table, the time for block 2 for all centres of tops and faces was 55.5 h, which was a little later than that for block 1, which ranged from 51 to 52 h for the various points. The insulating time for block 2 can therefore be sustainably extended a little longer than for block 1. Figs. 5 and 6 also shows the temperature–time curves of the temperature sensor readings for the centre of the top (points 1 and 2) and mass (points 7 and 8) of the two blocks. During the test period of 144 h, the trends of the block 1 curves were steeper than those of block 2. The faster and higher cooling rates imply that the insulation work is insufficient. In particular, the temperature dropped significantly (or at night). The steeper temperature–time curves (block 1) indicate that the maximum temperature differential between point 7 (core) and point 1 (top) was higher, and it clearly accompanies the change in the maximum temperature rise. Points 8 and 2 had the same trend for both blocks. In addition to the top temperature, the temperatures of side faces are the other important aspects of mass concrete. Cracks induced by excessive interior restraint (temperature differential) must be avoided. For block 1, point 4 and point 5 used 1/2-in. PE boards, and point 6 only adopted the template. The results of both temperature–time curves showed apparent differences in the use of 1/2-in. PE. For the shadowed back side (point 4) and the side facing the sun (point 6), the temperature rise was influenced by the sunshine, and the temperature insulation effect was slightly enhanced. The centres of adjacent faces for block 2 (point 3 and point 6 with EPS boards; point 4 and point 5 with XPS boards) had the same results with or without the effect of sunshine, and there was no observable difference. However, after 144 h, the temperature of block 2 dropped faster than that of block 1. This implies that the higher the insulation, the more serious the thermal shock problem might be. Insulation must be removed when the temperature differential for the hottest portion and surface cools to the ambient air temperature (20 °C) [1].

ida in the USA. During the experiment, from 2nd April to 10th April 2015, the weather was sunny all month and very hot, with an ambient temperature above 30 °C, even over 36 °C. The ambient temperatures and temperature differentials for blocks 1 and 2 are shown in Figs. 7 and 8. The temperature differentials between the centre of mass and the top or face locations were all below 60

Model I

point7-point1

point8-point2

50

point7-point4

point7-point5

45

point7-point6

55

point8-point6 point9

40 35 30 25 20 15 10 5 0 -5 0

20

40

60

80 100 Time(hour)

120

140

160

180

Fig. 7. The maximum temperature differentials and ambient temperatures for various locations of the block with light insulation. (The sudden decline in the curve is due to the power supply instability at the site in the summer).

60

Model II

point7-point1

point8-point2

50

point7-point3

point7-point6

45

point7-point4

point7-point5

55

point9

40 35 30 25 20

4.3. Temperature differential

15 10

For structural mass concrete, such as walls and columns, cracks due to interior restraints are very common in structural elements, and are harmful to the integrity and durability of concrete. The purpose of insulation is mainly to decrease the temperature differential between nearby locations (namely the centre of mass and the top or faces). This reduces cracks due to the excessive temperature differential. At the experimental site, located in Kaohsiung, a city in southern Taiwan, the weather is very similar to that of the state of Flor-

5 0 -5 0

20

40

60

80 100 Time(hour)

120

140

160

180

Fig. 8. The maximum temperature differentials and ambient temperatures for various locations of the block with heavy insulation block. (The sudden decline in the curve is due to the power supply instability at the site in the summer).

268

Y.-Y. Chen et al. / Construction and Building Materials 137 (2017) 261–271

Table 5 Results of temperature control for blocks 1 and 2 with maximum temperature differential and times taken. Block 1 Location Temp. differ.

Point 7 – 1

Point 7 – 3

Point 7 – 6

Point 7 – 4

Point 7 – 5

Max. (°C) Time (h)

10.2(None) 39.5

– –

5.7(WS) 68

9.8(PE) 67.5

6.7(PE) 70

Point 7 – 1

Point 7 – 3

Point 7 – 6

Point 7 – 4

Point 7 – 5

Max. (°C) Time (h)

11.9(PS) 25.5

7.5(EPS) 94.5

7.3(EPS) 95.5

4.5(XPS) 35.5

4.8(XPS) 95

Block 2

Temp. differ.

12 °C, which matches the specification requirements, as shown in Table 5. Block 1, with a thin insulation of template and 1/2-in. PE board and the whole block wrapped with canvas, was sufficient to reduce the temperature difference (less than 20 °C) between the core, top, and faces. For block 2 with heavy (thick) insulation on the top, faces with 1-in XPS board and 1-in EPS board, and the whole block wrapped with canvas, there was sufficient thermal insulation. However, in cold seasons or on a rainy day, excessive temperature variations may overcome with the use of a 1-in. XPS board. These experiments showed that the effects of the insulation materials for the centres of faces were from lowest to highest in the order: 1/2-in. PE board (or 1/2-in. template), 1-in. EPS board, and 1-in. XPS board. The top portion indicated that a 1-in. XPS board (block 2) was better than nothing (block 1). Regardless of whether top or faces were examined, the polystyrene board was shown to be more suitable as insulation for mass concrete. The top surface covered with polystyrene board and a layer of canvas above might cause insulation similar to a partial semi-adiabatic effect. In Fig. 8, the maximum temperature differential of (points 7 - points 1) and (points 8 – points 2) during 48–144 h reached a value close to zero, indicating that there was no observable variance in the temperature differential between the centre of mass and the top location. The top layer (block 2) had a temperature differential almost equal to zero, or even negative. This effect was due to it being covered with thick polystyrene board. A full insulation effect has occurred. Table 5 also shows the times taken to reach the maximum temperature differentials at various locations. For the top location, block 2 with heavy insulation (25.5 h) was quicker than block 1 with light insulation (39.5 h). In the experiment, after the concrete was poured and covered, both blocks were covered with a wet and white of non-woven fabric for 16 h, block 2 was then paved with 1in XPS board on the top surface, and both blocks were tightly wrapped with canvas. As the experiment took place on hot and sunny days, heavy insulation (block 2) materials allowed insulation to occur effectively. Block 1 with light insulation and canvas had a similar status to block 2 in a moderate (warm) climate. For the centres of faces, the times for both blocks were still significantly different. Block 1 was faster than block 2, and the duration of insulation was shorter. Block 2 had a sustainable insulation function.

100

Temperature(oC)

Location

Model I

,point1

90

T(oC) = -0.199t(h) + 75.914 r 2 = 0.972 T(oC) = -0.190t(h) + 75.892 r 2 = 0.985 T(oC) = -0.187t(h) + 76.261 r 2 = 0.986

80

T(oC) = -0.204t(h) + 82.857 r 2 = 0.990

,point7

,point5 ,point6

70 60 50 40 30 40

50

60

70

80

90 100 Time(h)

110

120

130 140

150

Fig. 9. Regression analysis of temperature–time gradients of each measurement point for model (block) 1.

etc. The temperature–time gradient is a very useful tool for indicating the rate of temperature decrease and assessing insulation properties. For the two blocks, from the peak temperature to six days afterward, data were collected to analyse the regression results of the temperature–time gradient (k). The slope of the linear formula represents the temperature decrease per unit time (°C/h). For blocks 1 and 2, Figs. 9 and 10 show the results of regression analysis for the peak temperature to 144 h (6 days), for respectively point 1, point 5, point 6, and point 7, and for point 1, point 3, point 5, and point 7, shown in the following equations. For block 1, point 1, point 5, point 6, and point 7 are given by, respectively: Top: no insulation design.

Tð CÞ ¼ 0:199tðhÞ þ 75:914; r2 ¼ 0:972

ð2Þ

Face: wrapped with 1/2-in PE board. 

Tð CÞ ¼ 0:190tðhÞ þ 75:892; r2 ¼ 0:985

ð3Þ

Face: 1/2-in. template only. 

Tð CÞ ¼ 0:187tðhÞ þ 76:261; r2 ¼ 0:986

ð4Þ

Core: 4.4. Temperature–time gradient (k)

Tð CÞ ¼ 0:204tðhÞ þ 82:857; r2 ¼ 0:990

The value of k (rate of heat (temperature) loss) is evaluated as the gradient of the measured temperature–time curve when the heat generated by the concrete has become insignificant, and it has been found that the temperature drops for various insulation locations and material applications [17,18]. The relationships between temperature and time curves were shown in Figs. 5 and 6. After the peak temperature, the temperature dropped faster or slower according to the type of insulation, the mixture design,

For block 2, point 1, point 3, point 5, and point 7 are given by, respectively: Top: paved with 1-in. XPS board.



Tð CÞ ¼ 0:133tðhÞ þ 79:706; r2 ¼ 0:961

ð5Þ

ð6Þ

Face: wrapped with 1-in. EPS board. 

Tð CÞ ¼ 0:148tðhÞ þ 74:624; r2 ¼ 0:995

ð7Þ

269

Y.-Y. Chen et al. / Construction and Building Materials 137 (2017) 261–271

such phenomenon. In block 1, only the top and faces were wrapped with light insulation and canvas, so the insulation effect was not as high. Fig. 10 (block 2) shows point 7 and point 1 almost overlap with each other. According to Table 6, the value of k of point 7 of block 1 and block 2 dropped by 0.204 °C/h and 0.148 °C/h, respectively, which is a decrease of 27.5%, or over a quarter. For point 1(top), block 1 and block 2 dropped by respectively 0.199 °C/h and 0.133 °C/h, and k increased by 33.2%, approximately one third. For the centres of faces between both blocks, for 1/2-in. PE board (point 5) and 1/2-in. template (point 6), k dropped by 0.190 °C/h and 0.187 °C/h, and for 1-in. EPS (point 3) and 1-in. XPS board (point 5), k dropped by 0.148 °C/h and 0.139 °C/h, respectively. Using thin insulation with both a 1/2in. PE board and 1/2-in. template made little difference, and thick insulations (1-in. XPS and 1-in. EPS boards) appear to be a better isolation layer. From the analysis of the maximum temperature differential and temperature–time gradient, block 2 more greatly enhanced the efficiency of the thermal insulation than block 1.

100 Model II

T(oC)= -0.133t(h) + 79.706 r 2 = 0.961 ,point1 T(oC)= -0.148t(h) + 74.624 r 2 = 0.995 ,point3 T(oC)= -0.139t(h) + 76.471 r 2 = 0.991 ,point5 T(oC)= -0.148t(h) + 81.199 r 2 = 0.987 ,point7

90

Temperature( oC)

80 70 60 50 40 30 40

50

60

70

80

90 100 Time(h)

110

120

130

140

150

Fig. 10. Regression analysis of temperature–time gradients of each measurement point for model (block) 2.

4.5. Ko, k and R-value Once the experiments on the insulation of materials and blocks were completed, the IMTC (Ko) and temperature–time gradient (k) were obtained. Apart from the template, the insulation materials used in this test (face) were only 1/2-in. PE, 1-in. EPS, and 1-in. XPS board. According to [27], the insulation materials used in the experiments to find the thermal resistance (R-value) were 3 (1/2) = 1.5 h ft2 °F/Btu (1-in. PE), 4 h ft2 °F/Btu (1-in. EPS), and 5 h ft2 °F/Btu (1-in. XPS). The relationships between Ko, k, and the R-value are shown in Fig. 11. The linear regression results for k vs. Ko, k vs. R-value, and Ko, vs. R-value are shown in Formulas (10)–(12).

Face: wrapped with 1-in. XPS board.

Tð CÞ ¼ 0:139tðhÞ þ 76:471; r2 ¼ 0:991

ð8Þ

Core: 

Tð CÞ ¼ 0:148tðhÞ þ 81:199; r2 ¼ 0:987

ð9Þ

Correlations between each location all exceeded 0.960. These values have extremely high relevance. The regression equations of all locations were almost parallel to each other. Point 7 of block 1 was furthest away from point 1 and other points. Block 2 had no

Table 6 Results of regression analysis for the temperature–time gradient (k) for blocks 1 and 2. Block 1 Materials

1/2-in. PE(Face)

1/2-in. template (Face)

None (Top)

Core

k (°C/h)

0.190 (point 5)

0.187 (point 6)

0.199 (point 1)

0.204 (point 7)

Block 2 1-in. EPS (Face)

1-in. XPS (Face)

1-in. XPS (Top)

Core

k (°C/h)

0.148 (point 3)

0.139 (point 5)

0.133 (point 1)

0.148 (point 7)

Temperature-time gradient ( oC/h)

-0.13

2.0 1-in. XPS

-0.14

1-in. EPS

-0.15

1/2-in. PE

1-in. XPS

1.8 1.5

1-in. EPS

1.2 -0.16 1.0 -0.17 -0.18

Y = 0.015X - 0.212 r2 = 0.987

Y = -0.034X - 0.130 r2 = 0.995

1/2-in. PE

-0.19 0.0

0.5

1.0

1.5

Insulation material temperature conductivity (oC/min.in)

2

3

0.8

1-in. EPS

0.5

1-in. XPS

1/2-in. PE 2.0 1

Y = -0.441X + 2.417 r2 = 0.966

4

2 3 6 1 R-value (hr.ft2.deg F/Btu) 5

4

0.2 5

6

Insulation material temperature conductivity ( oC/min.in)

Materials

Fig. 11. Relationships between the IMTC (Ko), temperature–time gradient (k), and thermal resistance (R-value) of various insulation materials used in the experiments.

270

Y.-Y. Chen et al. / Construction and Building Materials 137 (2017) 261–271

k ¼ 0:034  K o  0:130; r2 ¼ 0:995

ð10Þ

k ¼ 0:015  R þ 0:212; r2 ¼ 0:987

ð11Þ

K o ¼ 0:441  R  2:417; r2 ¼ 0:966

ð12Þ

The correlations from the linear regressions of k vs. Ko, k vs. R-value, and Ko vs. R-value are r2 = 0.995, 0.987, and 0.966, respectively. Therefore, they have a very good relationship. For 1-in. XPS (thick) board, with a lower IMTC, the smaller temperature–time gradient and higher thermal resistance was easy to achieve. Thus, any two parameters (k, Ko, and R-value) might be adopted to assess appropriate insulation materials for mass concrete. In Fig. 11, the template (Wood form) was only used to build up the moulds. In this paper, it should not be regarded as a type of insulation material. 4.6. Workability The specifications only cover the measurement of slump flow spread instead of the slump of flowing concrete. The specifications for the workability of mass concrete state that it is high-flow concrete, not self-consolidating concrete (SCC). Between the 0–90 min, the slump flow spread was measured at 25–26 in., with little variation and without visible bleeding and segregation on hot days. This indicates the stability of the HRWRA and other materials acting on the fresh mass concrete. The slump over the period of measurement showed no significant change for 10–11 in., and this was very similar to the case for the slump flow spread. The flow time had a variance of 6–12 s. 5. Conclusions This paper used ternary cementitious binders to design a mass concrete mixture, and constructed two 6 ft.  6 ft.  6 ft. model blocks to evaluate appropriate insulation materials. Through a simplified insulation box (2 ft.  2 ft.  2 ft.) test, the two blocks were equipped with thin and thick insulations, and eight thermocouple sensors were set up on each block to measure the temperature– time curve to analyse the temperature characteristics and temperature–time gradient. The findings are as follows:
As shown in Table 3, test results for the insulation material temperature conductivity (Ko) indicated that the values could be arranged from lowest to highest in the order of 1-in. extruded expanded polystyrene (XPS) board, 1-in. polyethylene (PE) board, 1-in. expanded polystyrene (EPS) board, and 1/2-in. PE board.
The maximum temperature rises of block 1 and 2 were approximately 71.6 °C and 73.2 °C, respectively. The error ratio was 2.6%. This implied that the variance of the maximum temperature rise of the insulation materials for both blocks was small.
The maximum temperature differential at the centre of the top (point 7 to point 1 and point 8 to point 2) and faces (point 7 to point 3, point 4, point 5, and point 6) of block 1 and block 2 were all less than 12 °C, which fitted the specified requirements (less than 25 °C). The maximum temperature differential was significantly influenced by the insulation materials.
The top of block 2 was paved with 1-in. XPS board, and the whole block was covered with canvas. The experiment results (Fig. 10) appeared to show a partially semi-adiabatic temperature rise from the ambient temperature.
At the centres of faces, the temperature–time gradient, k, for the thin and thick insulation materials had the following values, from lowest to highest, 0.190 °C/h for 1/2-in. PE board, 0.148 °C/h for 1-in. EPS board, and 0.139 °C/h for 1-in. XPS

board. For the top surface, 1-in. XPS board (block 2, 0.133 °C/h) was much better than nothing (block 1, 0.199 °C/h). This indicated that the 1-in.XPS board suited the requirements of insulation and economy (see Table 3).
Comparing the k values for the centre of mass (point 7) of the two blocks, block 1 and block 2 dropped by 0.204 °C/h and 0.148 °C/h, respectively, which is an enhancement of 27.5%, or over a quarter.
The correlation coefficients for k vs. Ko, k vs. the R-value, and Ko vs. the R-value were respectively 0.995, 0.987, and 0.966, indicating good relationships with each other. Thus, any two parameters might be adopted to assess appropriate insulation materials for mass concrete. For the 1-in. XPS (thick) board, with a lower IMTC, the smaller temperature–time gradient and higher thermal resistance was easy to achieve the isolated effect.
In a hot climate and on sunny days, comparing the top and faces of the two blocks (with thin and thick insulation), all the maximum temperature differentials fitted the limits of the specifications. However, in cold weather or on rainy days, the temperature can change rapidly. The insulation design should adopt block 2 (heavy ones). Funding This work was supported by the Taiwan Professional Civil Engineering Association [Grant No. TPCEA-N-1537, 2015] and the Ministry of Science and Technology, ROC. [Grant No. NSC 102-2221-E146-010-, 2013]. The funding sources had no role in study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication. Conflicts of interest There are no conflicts of interest to declare. Acknowledgements The authors would like to thank Chung-Lin Engineering Co., Ltd. in Kaohsiung Construction Department gracious assistance and Goldsun Group in Kaohsiung Ready-Mixed Concrete plants helps. References [1] ACI Committee 301–10, Specification for structure Concrete, American Concrete Institute, Farmington Hills, MI, 2010. [2] ACI Committee 116R–05, Cement and Concrete Terminology, American Concrete Institute, Farmington Hills, MI, 2005. [3] P.K. Mehta, P.J.M. Monteiro, Concrete – Microstructure, Properties and Materials, third ed., The McGraw-Hill Companies Inc., New York, NY, 2006. [4] A.M. Neville, Properties of Concrete, fifth ed., Prentice Hall, San Francisco, 2012. [5] ACI Committee 207.2R-07, Report on thermal and volume change effects on cracking of mass concrete, American Concrete Institute, Farmington Hills, MI, 2007. [6] ACI Committee 207.4R-05, Cooling and Insulating Systems for Mass Concrete, American Concrete Institute, Farmington Hills, MI, 2005. [7] K.C. Tayade, N.V. Deshpande, A.D. Pofale, Experimental study of temperature rise of concrete and assessment of cracking due to internal restraint, Int. J. Civ. Struct. Eng. 4 (2014) 353–364. [8] S.J. Jeon, Advanced assessment of cracking due to heat of hydration, Int. Restraint ACI Mater. J 105 (2008) 325–333. [9] Y. Fu, J.J. Beauidoin, Mechanism of delayed ettringite formation in Portland cement system, ACI Mater. J. 93 (1996) 327–333. [10] S. Diamond, Delayed ettringite formation – process and problems, Cem. Concr. Compos. 18 (1996) 205–215. [11] R. Barbarulo, H. Peycelona, S. Prenéb, J. Marchand, Delayed ettringite formation symptoms on mortars induced by high temperature due to cement heat of hydration or late thermal cycle, Cem. Concr. Res. 35 (2005) 125–131. [12] L. Acquaye, Effect of High Curing Temperatures on the Strength, Durability and Potential of Delayed Ettringite Formation in Mass Concrete Structures, Doctor

Y.-Y. Chen et al. / Construction and Building Materials 137 (2017) 261–271

[13] [14]

[15]

[16] [17] [18] [19]

[20]

of Philosophy thesis, Department of Building Construction, University of Florida, 2006. H.F.W. Taylor, Cement Chemistry, Thomas Telford Publishing, Heron Quay, London, 1997. Florida DOT. Standard specifications for road and bridge construction (346–2.3 Pozzolans and slag (1) Mass concrete), Florida Department of Transportation, Tallahassee, FL, 2013. Georgia DOT. Standard specifications for road and bridge construction, Georgia Department of Transportation (SPECIAL PROVISION, Adding the following to Subsection 500.3.05: AM. Mass concrete), NW Atlanta GA, 2014. J. Gajda, M. VanGeem, Controlling temperature in mass concrete, Concr. Int. 24 (2002) 58–62. P.L. Ng, I.Y.T. Ng, A.K.H. Kwan, Heat loss compensation in semi-adiabatic curing test of concrete, ACI Mater. J. 105 (2008) 52–61. I.Y.T. Ng, P.L. Ng, A.K.H. Kwan, Effects of cement and water contents on adiabatic temperature rise of concrete, ACI Mater. J. 106 (2009) 42–49. C. Pongsak, S. Tangtermsirikul, Effect of aggregate type, casting, thickness and curing condition on restrained strain of mass concrete, Songklanakarin J. Sci. Technol. 32 (2010) 391–402. J. Gajda, M. Weber, I. Diaz-Loya, Rise mixture for mass concrete, Concr. Int. 8 (2014) 48–53.

271

[21] ACI Committee 207.1R-05, Guide to Mass Concrete, American Concrete Institute, Farmington Hills, MI, 2005. [22] J.J. Shaw, C.T. Jahren, K. Wang, J.L. Li, Iowa Mass Concrete for Bridge Foundations Study – Phase II. Report No. InTrans Project 10-384, Iowa State University, 2014. [23] M. Tia, A. Lawrence, C. Ferraro, S. Smith, E. Ochiai, Development of Design Parameters for Mass Concrete Using Finite Element Analysis. Final report, Department of Civil and Coastal Engineering, University of Florida, Gainesville, FL, 2010. [24] C.P. Bobkracino, R. Seracino, P. Zia, A. Edwards, Crack Free Mass Concrete Footings on Bridges in Coastal Environments, NCDOT Project 2012–09 (FHWA/ NC/2012-09), Department of Civil, Construction, and Environmental Engineering, North Carolina State University, 2014. [25] N. Bouzoubaâ, F. Simmon, Use of fly ash and slag in concrete. CAMENT Report MTL 2004–16 (TR-R), National Resources CANANA, Ottawa, 2005. [26] A.M. Lawrence, M. Tia, C. Ferraro, M. Bergin, Effect of early age strength on cracking in mass concrete containing differential supplementary cementitious materials: experiment and finite-element investigation, J. Mater. Civ. Eng. 24 (2012) 362–372. [27] R-value (insulation), https://en.wikipedia.org/wiki/R-value_(insulation), 2016 (accessed 24.10.16).