Liu, Yu, Wu and Yuan 1 MECHANICAL PROPERTIES OF A WATERPROOFING ADHESIVE LAYER USED ON CONCRETE BRIDGES UNDER HEAVY TRAFFIC AND TEMPERATURE LOADING Yun Liu, Ph.D, Lecturer (Corresponding Author) College of Civil and Transportation Engineering, Hohai University 1, Xikang Road, Nanjing 210098, China Tel: (86)025-83787353 Fax: (86)025-83786633 Email: [email protected]Xin Yu, Ph.D, Associate Professor College of Civil and Transportation Engineering, Hohai University 1, Xikang Road, Nanjing 210098, China Tel: (86)025-83787353 Fax: (86)025-83786633 Email: [email protected]Jiantao Wu, Ph.D, Lecturer College of Civil and Transportation Engineering, Hohai University 1, Xikang Road, Nanjing 210098, China Tel: (86)025-83787353 Fax: (86)025-83786633 Email: [email protected]Yuan Yuan, Master, Lecturer International Cooperation School, Nanjing Institute of Railway Technology 65, Jianning Road, Nanjing, 210015, China Tel: (86)025-85838538 Fax: (86)025-85838652 Email: [email protected]Word count: 3164+ 17⋅250 = 7414 Submitted on July 27, 2011 TRB 2012 Annual Meeting Original paper submittal - not revised by author.
Transcript
Liu, Yu, Wu and Yuan 1
MECHANICAL PROPERTIES OF A WATERPROOFING ADHESIVE
LAYER USED ON CONCRETE BRIDGES UNDER HEAVY TRAFFIC
AND TEMPERATURE LOADING
Yun Liu, Ph.D, Lecturer (Corresponding Author)
College of Civil and Transportation Engineering, Hohai University 1, Xikang Road, Nanjing 210098, China
Word count: 3164+ 17⋅250 = 7414 Submitted on July 27, 2011
TRB 2012 Annual Meeting Original paper submittal - not revised by author.
Liu, Yu, Wu and Yuan 2
ABSTRACT: Based on the data collected from a concrete box girder bridge in Shanghai-Hangzhou Expressway Widening and Rebuilding Project, the on-site temperature monitoring system of pavement, finite element method and laboratory direct-shear and pull-off tests were designed to study the adhesive behavior of a waterproof layer used between a concrete-bridge deck and an asphalt mixture pavement. Firstly, the sensors were applied to monitor the temperature gradient distribution of pavement in consecutive days in winter and summer. Secondly, a three-dimensional, finite-element model was developed to analyze the interfacial shear stress and tensile stress in response to vehicle and temperature loading. Lastly, the shear strength and tensile strength of SBS modified emulsified asphalt, SBS modified asphalt, asphalt-rubber SAMI and FYT waterproof coating respectively used as a waterproof adhesive layer between SMA (Stone mastic asphalt) pavement and bridge deck were tested and compared. Results indicate that the max tensile stress appears when the loads move on the pavement surface above the quarter-span, and the max shear stress appears when the loads move on the pavement surface at the center of a span. By comparing the computational results with test results, the tensile strength of the four alternative waterproofing adhesive layer materials can always be greater than FE analysis results, and the SBS modified asphalt and rubber asphalt SAMI can be primarily considered for the material of waterproof adhesive layer.
TRB 2012 Annual Meeting Original paper submittal - not revised by author.
Liu, Yu, Wu and Yuan 3
0 INTRODUCTION In the transportation industry, asphalt mixture pavement is commonly used as a wearing course constructed on concrete-bridge decks. Pavement debonding occurs when the shear stress and/or normal tensile stress exceed the interfacial shear strength and/or pull-off strength. To prevent this problem, a waterproofing adhesive layer (WAL) can be placed as an interlayer between the bridge deck and the asphalt mixture pavement to prevent permeation of water and enhance interface adhesion.
Laboratory and field tests for evaluating the engineering properties of WALs have been investigated, including the ones conducted by the National Cooperative Highway Research Program (NCHRP) in the United States (1, 2). The engineering properties tested for WALs include tensile strength, durability, toughness, elasticity, water impermeability, puncture resistance, temperature susceptibility, etc. Laboratory test methods for evaluating material properties and repairing techniques in field construction were comprehensively investigated in UK (3-5). However, the primary performance criterion to assess the benefits of WALs is the interface adhesive strength. Accordingly, laboratory and field test equipments and methods have been investigated (6).
Less attention has been paid to the research of structural modeling and stress analysis of WALs. Due to the complicated multi-state loading conditions applied on bridges and the membrane structure of WALs as an interlay, it is very difficult to effectively measure the interface shear and normal tensile stresses between WAL and bridge deck or pavement. Therefore, it would be very meaningful to capture these critical stresses by using numerical modeling.
Meanwhile, the adhesive capability of WALs is highly affected by many field conditions, which include construction conditions, material properties, vehicle loading, bridge structure and environmental temperature. Therefore, studies on the influences of these critical factors on the mechanical behaviors of WALs are considered essential in order to design more reliable materials and structures.
Accordingly, this paper aims to describe the adhesive behavior of WALs used on the concrete bridge deck. Kinds of WAL were investigated to study the influences of some critical factors that affect the adhesive behavior of WAL system, using field monitoring, finite-element (FE) modeling technique and laboratory testing methods. Designs of materials and structures are also discussed in terms of research results. 1 TEMPERATURE MONITORING Asphalt mixture is a thermosensitive pavement material with lower tensile strength. When the bridge pavement structure is in service, continuous monitoring of temperature will provide information about the temperature load at different depths of the pavement. 1.1 Description of Structure and Materials A simply supported prestressed concrete box-girder bridge included in Shanghai-Hangzhou Expressway Widening and Rebuilding Project in China is taken as the engineering background. The bridge with each standard span of 30m has a nine-cell cross-section. Transverse diaphragms are located at each of the two ending supports of each span. The pavement structure under investigation consists of an 8cm-thick C40 steel reinforced concrete bridge deck placed on the bridge box-girder, a 3 mm-thick WAL bonded to bridge deck, and a 10 cm-thick SMA 13 asphalt mixture pavement placed on the WAL, as displayed in Figure 1.
TRB 2012 Annual Meeting Original paper submittal - not revised by author.
Liu, Yu, Wu and Yuan 4
C40 steel reinforced concrete bridge deck
C50 Concrete bridge girder
SMA 13 pavement
WAL
FIGURE 1 Pavement structures on concrete bridge.
1.2 Monitoring Equipment and Plan TP-K01 surface thermocouple and TES1310 digital temperature gauge from TES Electrical Electronic Corp were applied to monitor and read the temperature of pavement, as shown in Figure 2. The performance index of surface thermocouple and digital temperature gauge were listed in Table 1 and 2. Firstly, pavement coring and drilling were performed. Then, the thermocouple was adhered to the holes at different depths of pavement sample. The locations of thermocouples are shown in Figure 3. The temperature of different depths of pavement was recorded per hour in a couple of days in winter (December) and summer (August).
(a) TP-K01 surface thermocouple.
(b) Digital temperature gauge.
FIGURE 2 Monitoring equipment. TABLE 1 Performance Index of TP-K01 Surface Thermocouple
Measuring range ()
Measuring accuracy ()
Zero resistance (Ω )
Resistance coefficient
( 1−Ω ) Insulation resistance
( ΩM )
-50～200 ±0.3 46.60 5 ≥50 TABLE 2 Performance Index of TES1310 Digital Temperature Gauge
Measuring range ()
Measuring accuracy
() Resolution
Sampling rate (times per second)
Working temperature ()
-50～1300 ±（0.2%+1） 0.1 2.5 0～50
1.5cm4cm
7cm
10cm
SurfaceMonitoring point 1Monitoring point 2Monitoring point 3
Monitoring point 4
Monitoring point 5
FIGURE 3 Monitoring points of temperature sensors in pavement.
TRB 2012 Annual Meeting Original paper submittal - not revised by author.
Liu, Yu, Wu and Yuan 5
1.3 Monitoring Results Because the pavement temperature variation in consecutive days is almost daily periodicity, the temperature variations of pavement at different depths on November 17 and August 3 were shown in Figure 4. It can be seen that the pavement temperature is lower than air temperature form 6:00 to 15:00 in november 17 in winter and the pavement temperature is higher than air temperature from 8:00 to 18:00 in august 3 in summer.
0
2
4
6
8
10
12
14
0 2 4 6 8 10 12 14 16 18 20 22 24Time (h)
Tem
pera
ture
()
air temperature
0cm
1.5cm
4cm
7cm
10cm
(a) November 17
202530354045505560
0 2 4 6 8 10 12 14 16 18 20 22 24Time (h)
Tem
pera
ture
()
air temperature
0cm
1.5cm
4cm
7cm
10cm
(b) August 3
FIGURE 4 Temperature distribution of pavement. 2 FE MODELING 2.1 Analysis Method The thermal strain is given by
( )[ ]T00 000111TT −=αε （1）
Where α is thermal expansion coefficient (1/), Φ0 is the initial temperature, and Φ is the steady temperature or transient temperature. Φ can be obtained through interpolation based on Φi which can be got by thermal analysis. The equation can be expressed as follows:
（2） ( ) ei
n
ii zyxN
e
Nφ== ∑=
φφ ,,1
TRB 2012 Annual Meeting Original paper submittal - not revised by author.
Liu, Yu, Wu and Yuan 6
Considering thermal strain, the stress-strain relationship can be determined as follows:
( )0εεDσ −= （3）
We shall consider an axisymmetrical elastic layered half-space problem with the material characteristics depending on the reference temperature. The equations of equilibrium in the cylindrical coordinates system are (7)
0
0
=+∂∂
+∂∂
=−
+∂∂
+∂∂
rrz
rzrzrzrz
rzrr
ττσ
σστσ θ
（4）
When considering the elastic modulus, Poisson's ratio and thermal expansion coefficient as function of temperature, the stress and strain relations can be written as:
⎟⎠⎞
⎜⎝⎛
∂∂
+∂∂
=
+⎟⎠⎞
⎜⎝⎛ +∂∂
+∂∂
=
+⎟⎠⎞
⎜⎝⎛
∂∂
+∂∂
+=
+⎟⎠⎞
⎜⎝⎛
∂∂
++∂∂
=
rw
zud
cru
rub
zwa
czw
rub
rua
czw
rub
rua
zr
z
r
τ
φσ
φσ
φσ
θ （5）
In Eq. (4) and (5), u、w are horizontal and vertical displacement respectively, σr, σθ, σz are r, θ, z-direction stress respectively and τzr is the shear stress.
( ) ( )[ ]( )[ ] ( )[ ]( ) ( )( )[ ] ( )[ ]
( ) ( )( )( )( )[ ]φμ
φφμφαφ
φμφμφμφ
φμφμφμφ
+=
−=
−+=
−+−
=
12
21
211
2111
Ed
Ec
Eb
Ea
；
Where E(Φ) is elastic modulus, μ(Φ) is poisson ratio, and α(Φ) is thermal expansion coefficient. The following equation can be obtained by applying equation (3) into the formula of principle of virtual displacement:
( )∫ ∫ =−−V S
SVσ
δδδ 0dd TTT Tufuσε （6）
The functional equation of principle of minimum potential which was used to solve the problem of thermal stress can be obtained as follows:
( ) ΓΩΓΩ
dd21 TT
0TT
p TufuDεεDεεu ∫∫ −⎟⎠⎞
⎜⎝⎛ −−=∏
σ
（7）
TRB 2012 Annual Meeting Original paper submittal - not revised by author.
Liu, Yu, Wu and Yuan 7
By applying principle of minimum of potential energy, finite element method solves the following equations:
PKa = （8） The P that includes temperature load vector can be expressed as follows:
0ε
PPP += f （9）
Where Pz0 is loading items caused by temperature strain which can be calculated by Eq. (10), and Pf is the other load items mainly caused by traffic.
∑∫Ω=e
T
e
Ωd00DεBPε （10）
2.2 Structure of The Model The ADINA program was used to build a full-scaled, three-dimensional, finite-element model (3D-FE) to simulate the bridge, as shown in Figure 5. An eight-node solid element (3D SOLID) was used to model the steel reinforced concrete bridge deck, concrete box-girder and SMA pavement. A four-node membrane element (SHELL) was used to model the WAL. In the FE model, it is assumed that the interface between the WAL and bridge deck is considered fully bonded when the project has just completed. The node at the bottom of box-girder connected with pier is fixed in the FE model.
FIGURE 5 FE model of pavement system on concrete girder bridge.
2.3 Material model Concrete is a brittle material and performs in an elastic way at a small stress level that does not reach its ultimate strength. C40 steel reinforced concrete and C50 concrete are assumed to be linear elastic model. The WAL primarily performs elastic behavior before breakage. Therefore, the WAL is assumed to be an elastic material in the FE model. The determined material parameters are summarized in Table 3 (8-10), the modulus value of SMA13 pavement under 20 is 1200MPa according to previous research results (11).
Asphalt mixture is a viscoelastic-plastic material, and its mechanical behavior is highly dependent on time and temperature. Generally, when it is subjected to a small strain level, asphalt mixture exhibits
TRB 2012 Annual Meeting Original paper submittal - not revised by author.
Liu, Yu, Wu and Yuan 8
linear-viscoelastic behavior without noticeable deterioration. Witczak model (12) was used to predict the dynamic modulus of SMA pavement versus temperature. Figure 6 give the dynamic modulus of SMA pavement on November 17 and August 3.
5600
5800
6000
6200
6400
6600
6800
7000
7200
7400
0 2 4 6 8 10 12 14 16 18 20 22 24Time (h)
Dyn
amic
mod
ulus
(MPa
0
2
4
6
8
10
12
14
Pave
men
t tem
pera
ture
()
Dynamic modulus of upper layer Dynamic modulus of lower layerTemperature of upper layer Temperature of lower layer
(a) November 17
200
400
600
800
1000
1200
1400
1600
1800
2000
0 2 4 6 8 10 12 14 16 18 20 22 24Time (h)
Dyn
amic
mod
ulus
(MPa
0
10
20
30
40
50
60
70
Pave
men
t tem
pera
ture
()
Dynamic modulus of upper layer Dynamic modulus of lower layerTemperature of upper layer Temperature of lower layer
(b) August 3
FIGURE 6 Dynamic modulus. 2.4 Load Model As stress concentration was mainly caused by wheel load, a double-rectangle uniformly distribution pressure
TRB 2012 Annual Meeting Original paper submittal - not revised by author.
Liu, Yu, Wu and Yuan 9
was used to simulate the single rear axle of Grade I highway load in the FE model according to General Code for Design of Highway Bridges and Culverts (JTG D60-2004) (10), as shown in Figure 7 and 8. The value of standard tyre-ground contact pressure in China is 0.707MPa.
The contact areas of the load on pavement were simulated by two rectangles with a space of 10cm between two tires, and each rectangle has a dimension of 20cm (width) by 23cm (length), as shown in Figure 10 (13-15). The tire-pavement friction coefficient can reach 0.5 at the vehicle acceleration or braking (16). In order to account for the tire friction force, a truck is assumed to brake suddenly, which results in an additional tangible force applied on the surface of the asphalt mixture pavement.
30kN 120kN120kN 140kN 140kN
3.0m 1.4m 1.4m7.0m
(a) Side view
(b) Plan form FIGURE 7 Layout of Grade I highway load.
230mm 100mm 230mm
200m
m
Double wheel load
FIGURE 8 Uniform load area of single rear axle.
In order to discuss the influence of traffic load position on the stress of pavement, the transversal and longitudinal load positions of tire contact areas are shown in Figure 9. In the FE model, wheel load moves both longitudinally and transversely at different bridge positions in order to determine the most unfavorable loading positions.
Load position 4Load position 3
Load position 2Load position 1
(a) Transversal Load position
TRB 2012 Annual Meeting Original paper submittal - not revised by author.
Liu, Yu, Wu and Yuan 10
pier
pier
pier
midspan
midspan
Load position c
Load position bLoad position a
(b) Longitudinal load position
FIGURE 9 Load position. Modeling results show that the maximum tensile stress in the WAL occurred at transversal load
position 3 and longitudinal load position b. Meanwhile, the maximum interface shear stress appeared under transversal load position 1 and longitudinal load position a. In Figure 10, the “1, 2, 3, 4” on abscissa axis represent transversal load position and the mark “a, b, c” in datagraph represent longitudinal load position. In order to obtain the mechanical index, the traffic load and temperature load in a whole day are applied on the pavement structure simultaneously in FE analysis.
0
0.01
0.02
0.03
0.04
0.05
1 2 3 4
Max
imum
tens
ile st
ress
/MPa
a b c
(a) Tensile stress
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
1 2 3 4
Max
imum
shea
r stre
ss /M
Pa a b c
(b) Shear stress FIGURE 10 Peak value of mechanical response under different loading positions.
TRB 2012 Annual Meeting Original paper submittal - not revised by author.
Liu, Yu, Wu and Yuan 11
3 EXPERIMENTS OF SYNERGISTIC EFFECT OF PAVEMENT The alternatives of WAL used in this system include SBS modified emulsified asphalt, SBS modified asphalt, asphalt-rubber SAMI and FYT waterproof coating. In this research the laboratory tests, including the direct-shear and pull-off tests, were designed to measure the interfacial adhesive strengths of WALs used on the concrete-bridge deck. 3.1 Skew-shear test The skew-shear test was designed to determine the interfacial shear strength using UTM (17), as shown in Figure 11. In this test, a concrete slab and asphalt mixture slab both measuring 70mm (width) by 50mm (length) by 50mm (height) were presented. A 3-mm WAL was heated and adhered to the top of the PCC slab. Likewise, the asphalt mixture slab was heated and then compacted on the WAL. Consequently, the asphalt mixture slab was fixed by a steel box, to which a compressive force was applied on the top and a shear force was applied on the side, as shown in Figure 2. The shear speed of 50 mm/min was applied (18). 3.2 Pull-off test The pull-off test was designed to determine the interface pull-off strength, as depicted in Figure 12. The test samples included a concrete cylinder and asphalt mixture cylinder with a diameter of 100mm and a height of 80mm. The concrete and asphalt mixture cylinder were bonded with a WAL as an interlay, using the same method as explained for the shear adhesion test. The drawing speed of 100~200N/s was applied (6).
FIGURE 11 Skew-shear adhesion test.
FIGURE 12 Pull-off adhesion test.
4 FINITE ELEMENT ANALYSIS AND EXPERIMENTAL RESULTS The shear/tensile strength (the average value of corresponding three test results) and calculated maximum shear/tensile stress of the waterproof adhesive layer under temperature and traffic load are shown in Table 4 and 5. The shear strength of the first two waterproof adhesive layer material meets the requirement of the FE analysis results. As for SBS modified asphalt, the ratio of test shear strength to FE analysis stress are 12.6, 2.9, and 1.4 on the condition that temperature is 0 , 20 , and 60 , respectively. As for Asphalt-rubber SAMI, the value ratio of test shear strength to FE analysis stress are 12.7, 3.1, and 1.7 on the condition that temperature is 0 , 20 , and 60 , respectively. The tensile strength of the four waterproof adhesive layer material meets the requirement of the FE analysis results, and the ratio of test tensile strength to FE analysis stress are all above 1.
In conclusion, whether under high or low temperature conditions, tensile strength is basically larger than the FE analysis results. According to the shear and pull-out test results, SBS modified asphalt and asphalt-rubber SAMI are the best for waterproof adhesive layer material.
TRB 2012 Annual Meeting Original paper submittal - not revised by author.
5 CONCLUSIONS The adhesive behavior of WALs used on concrete bridges under traffic and temperature is studied in this paper, and the shear/tensile strength is tested to verify the FE analysis. The results from the testing and the modeling support the following conclusions:
1. Based on the temperature monitoring results, the pavement temperature is basically lower than air temperature in winter and the pavement temperature is basically higher than air temperature in summer.
2. The maximum tensile stress appears when the load move on pavement surface above the quarter-span, and the single rear axle is above one end of trapezoidal box in the transversal direction.
3. The maximum shear stress appears when the loads move on the pavement surface at the center of a span, and the single rear axle is above wet joint in the transversal direction.
4. According to the shear/tensile strength under different temperature, the pavement debonding occurs easily in summer due to the repeat-action of traffic and temperature loading.
5. As the shear/tensile strength can meet the requirement of FE analysis results for design，SBS modified asphalt and asphalt-rubber SAMI are recommended for waterproof adhesive layer material which was used on concrete bridges.
ACKNOWLEDGMENT This research has been funded by Research Project of the Shanghai Urban Construction and Communications Commission of China (No. 2009-003-006). REFERENCES 1. Van Til CJ, Carr BJ, Vallerga BA. Waterproof membranes for protection of concrete bridge
decks-laboratory phase. NCHRP report 165, National Research Council, Washington, DC, 1976. 2. Manning DG. Waterproofing membranes for concrete bridge decks. Washington, DC: National Academy
TRB 2012 Annual Meeting Original paper submittal - not revised by author.
Liu, Yu, Wu and Yuan 13
Press, 1995. 3. Price AR. Laboratory tests on waterproofing systems for concrete bridge decks. Report 248, TRL,
Crowthorne, United Kingdom, 1990. 4. Price AR. Waterproofing of concrete bridge decks: site practice and failures. Report 317, TRL, Crowthorne,
United Kingdom, 1991. 5. Stevenson A, Evans W. The adhesion of bridge deck waterproofing materials. Report 325, TRL,
Crowthorne, United Kingdom, 1992. 6. Wei Huang. Theory and method of deck paving design for long-span bridges. Beijing: China Constr Ind
Press, 2006. 7. Litao Geng, Yang Zhong, Na Qiao. Thermal Stresses of Asphalt Pavement with Material Property
Dependent on Temperature. Journal of Highway and Transportation Research and Development, Vol. 26, No. 11, 2009, pp. 16-20.
8. Qinwu Xu, Qinghua Zhou, Medina Cesar. Experimental and numerical analysis of a waterproofing adhesive layer used on concrete-bridge decks. International Journal of Adhesion and Adhesives, Vol. 29, No. 5, 2009, pp. 525-534.
9. Qinwu Xu, Zengzhi Sun, Hu Wang. Laboratory Testing Material Property and FE Modeling Structural Response of PAM-Modified Concrete Overlay on Bridges. Journal of Bridge Engineering, Vol. 14, No. 1, 2009, pp. 26-35.
10. JTG D60-2004. General specifications for design of highway bridges and culverts. Beijing: Communication Ministry Press of China, 2004.
11. Guanghui Wang, Chenglong Wei, Bin Li, et al. The emulation analysis to the temperature effect of the bridge deck pavement of long-span bridge. Journal of Hunan Institute of Science and Technology (Natural Sciences), Vol. 20, No. 4, 2007, pp. 83-87.
12. Sheng Zeng, Hui Wu, Qi Xu. The Characteristic Analysis of Dynamic Modulus in Pavement. Journal of Changsha Communications University, Vol. 20, No. 2, 2004, pp. 34-37.
13. Zhendong Qian, Wei Huang, and Quan Mao. Mechanical Analysis of Nanjing Second Yangtze River Bridge Deck Pavement. Journal of Highway and Transportation Research and Development, Vol. 18, No. 6, 2001, pp. 43-46.
14. Xingyu Gu. Research on Mechanical Analysis and Structural Design of Suspension Bridge Asphalt Pavement, Ph.D. Thesis, Southeast University, Nanjing, China, 2002.
15. Zhenqing Liu. Research on Key Technology of Theory for Long-span Steel Bridges Deck Surfacing Design, Ph.D. Thesis, Southeast University, Nanjing, China, 2004.
16. Wei Xu, Haitao Bai, Xiaoning Zhang. Numerical simulation and analysis of bridge deck overlay of long span cable-stayed bridge. Journal of Harbin Institute of Technology, Vol. 35, No. 6, 2003, pp. 750-754.
17. Yiming Wu. Mechanics Analysis about Asphalt Paving on Long-span Steel Bridge, Master Thesis, Southeast University, Nanjing, China, 2005.
18. Youhua Dai. Research on waterproof and cohesive layer of concrete bridge deck asphalt pavement, Master Thesis, Southeast University, Nanjing, China, 2007.
TRB 2012 Annual Meeting Original paper submittal - not revised by author.
