- Email: [email protected]
Measurement of pore water pressure in asphalt pavement and its effects on permeability
Accepted Manuscript Measurement of pore water pressure in asphalt pavement and its effects on permeability Junqi Gao, Chengcheng Guo, Yutao Liu PII: DOI: Reference:
S0263-2241(14)00561-2 http://dx.doi.org/10.1016/j.measurement.2014.11.013 MEASUR 3132
To appear in:
Measurement
Received Date: Revised Date: Accepted Date:
3 May 2013 13 November 2014 19 November 2014
Please cite this article as: J. Gao, C. Guo, Y. Liu, Measurement of pore water pressure in asphalt pavement and its effects on permeability, Measurement (2014), doi: http://dx.doi.org/10.1016/j.measurement.2014.11.013
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Measurement of pore water pressure in asphalt pavement and its effects on permeability Junqi Gao*, Chengcheng Guo, Yutao Liu Department of Civil Engineering, Nanjing University of Aeronautics and Astronautics,29 Yudao Street, Nanjing 210016, China
Abstract: On the effect of pore water pressure in a saturated asphalt pavement, fluid velocity and water infiltration rate will increase, which have a disadvantage influence on asphalt mixture strength. In order to measure the pore water pressure in pavement, a fiber optic hydraulic pressure sensor (FOHPS) is designed. The theoretical correlation between the applied pressure and the center wavelength shift of the FOHPS is derived by laboratory experiment. The pore water pressures in asphalt pavement at some running speeds were measured in situ. Furthermore, a falling head permeameter method was used to measure the permeability coefficients of asphalt mixture exposed to hydraulic pressures which were from 40kPa to 350kPa, and the correlation between permeability and hydraulic pressure was obtained. The experimental results showed that the pore water pressure would increase with the increasing car’s speed, but the lifetime of pore water pressure decreases with the increasing speed. The permeability coefficients of SMA-13 and AC-20 mixtures decrease with the increasing hydraulic pressure, but the water infiltration rate increases on an approximate linear curve as the hydraulic pressure increased from 40kPa to 350kPa.
Keywords: Pore water pressure, Fiber optic hydraulic pressure sensor, FBG, Asphalt pavement, *Corresponding author. Tel./fax: +86 25 84891754 E-mail address: [email protected]
Permeability coefficient
1. Introduction Asphalt concrete pavement is directly exposed to the natural environment, and traffic loads, environmental temperature and moisture are the major external factors which influence pavement performance. Some cracks will appear in the surface layer under the action of external loads. Water will infiltrate into pavement through connected pores and cracks. Under the repeated action of wheel loads, the pore water pressure in asphalt pavement is formed, which would make pavement surface layer spalling or loose, and finally cause pavement structural damage. Pore water pressure in pavement is considered to be one of the major causes of the asphalt pavement damages. There have been significant efforts in researching pore water pressure in asphalt pavement. Kutay et al. [1-3] computed dynamic water pressure gradient, pore water pressure and shearing stress at the solid–water interfaces by means of the lattice Boltzmann method. Masad [4,5] calculated the distribution of the pore water pressure gradient by means of the finite difference method. Kettil et al. [6] simulated wet asphalt pavement deformation and water flow. Zhou et al. [7] calculated the pore water pressure in saturated asphalt pavement and showed that when the permeability coefficient of asphalt mixture is 1×10-4cm/s and the load time is 0.005s, the pore water pressure is 0.57MPa. Dong et al. [8] also computed the pore water pressure in the asphalt pavement surface, whose value is 0.44MPa. Cui and Jin [9] studied the pore water pressure in asphalt pavement. Li and Deng [10] built a theory model in which the asphalt pavement was regarded as an axial symmetrical body of multilayered saturation elastic half space, and pore water pressure in asphalt pavement under mobile load was calculated. In addition, Gao et al. [11]
measured the dynamic water pressure in asphalt pavement surface. The results showed that when car speed is 80km/h, the water pressure is about 0.2MPa. Permeability of an asphalt pavement is one of the most important parameters that have a direct influence on its design life. Fixed head test and falling head test are often used to measure permeability coefficient of asphalt mixture [12]. Moreover, some models which are used to estimate asphalt mixture permeable coefficient have been suggested [13]. The Kozeny–Carman equation has been used over the years to approximate the permeability of granular materials. It was derived based on representing the air voids as capillary tubes and applying the hydraulic radius theory. Based on Kozeny-Carman equation, Masad presented a simple equation for approximating the permeability of asphalt mixes [14]. The permeability coefficient is required to be provided to calculate the pore water pressure by finite element model, and it is generally considered as constant. This assumption would be appropriate at low pressure head and not appropriate for higher hydrodynamic pressure head. At present, field data of pore water pressure in asphalt pavement is still relatively few. In order to verify some established theory models, it is required to obtain a lot of field measurements of the pore water pressure. Fiber Bragg grating (FBG) sensing technique can overcome the weak of conventional testing method, and the pore water pressure can be measured accurately by FBG sensing. In this paper, a FOHPS is designed, and using this fiber sensor to measure the pore water pressure in asphalt pavement is studied. After calibrating the fiber sensor, a field pressure testing was conducted. In addition, the permeabilities of different asphalt specimens exposed to pore water pressures are studied using water and a falling head approach.
2. Optical fiber sensor design
The FOHPS is designed as shown in Fig.1. The appearance of the sensor is like a cylinder, which is composed of an aluminum enclosure, a circular diaphragm, a fiber grating and seals. When the water in asphalt pavement flow into the sensor and act on the diaphragm, the diaphragm will produce deformation that transfers to the fiber grating pasted on back surface centre of the diaphragm. Then axial strain will occur in the grating. By detecting the wavelength shift of the grating, we can obtain the pore water pressure that exists in saturated asphalt pavement under loads.
Fig. 1 Fiber optic hydraulic pressure sensor
The radius of the designed sensor is 27mm with a height of 26mm, and the diaphragm radius is 15 mm. The fiber grating is attached to the back surface of the diaphragm in a two-step process. First, the diaphragm is cleaned with alcohol until the alcohol evaporates. Second, a tension is applied to the fiber grating and then the grating is bonded on the diaphragm with an epoxy resin, which can also protect the sensor from damage. Under the action of uniform pressure p [11], the radial strain on the point with a distance r from the diaphragm center can be expressed as
εr =
3p (1 − µ 2 )( R 2 − 3r 2 ) 2 8h E
(1)
Where E is the Young’s modulus of the metal diaphragm, µ is Poisson’s ratio, h is the thickness of the diaphragm, and R is the radius of the metal diaphragm. From Eq. (1), it can be seen that the grating strain εr varies according to different distance r, and at the center of the round diaphragm its strain εr is maximum. In the designed sensor the length of
the grating is about 10mm, and the diaphragm radius R is 15 mm. By using these data, the FBG average strain can be calculated as
εr =
2.89 p (1 − µ 2 ) R 2 2 8h E
(2)
By using Eq. (2), the linear relationship between strain εr detected by FBG and pressure p can be observed when other variables does not change.
3. Calibration tests Three fiber optic hydraulic pressure sensors (S1, S2 and S3) were designed. The peak wavelengths of S1, S2 and S3 are 1558.365nm, 1561.447nm and 1555.552nm respectively. Diaphragm materials used in these sensors are stainless steel, whose elastic modulus is 200GPa, Poisson's ratio is 0.3, and thickness is 0.3mm. For sensors S1, S2 and S3, the pressure loads were 0.05MPa, 0.1MPa, 0.14MPa, 0.2MPa, 0.24MPa, 0.3MPa, 0.35MPa, respectively. During the load procedure, the pressure load was applied by using the pneumatic pump and maintained for 5 minutes to measure the strain of diaphragm. The outputs of these sensors were monitored by a si425-500 optical sensing interrogator (Micron Optics, Inc., USA). Furthermore, the si425-500 communicated with the control terminal, a notebook computer, via an Ethernet cable for the transmission of the enormous monitoring data. Before the field measurement, a calibration test for sensors S1, S2 and S3 was implemented to evaluate their respective behavior of linearity. The measurement results here and subsequent have been corrected for temperature compensation using a free FBG. Fig. 2 demonstrated the correlation between the center wavelength shift of the FBG and the applied pressure, and a good linear agreement, with a correlation coefficient of 0.998, can be well observed for S1, S2 and S3. For S1, S2 and S3, the linear regression slope coefficients are 4.0nm/MPa,4.9nm/MPa, and
4.5nm/MPa respectively.
Fig. 2 Relationship between applied pressure and center wavelength shift of the sensor
4. Field measurement of pore water pressure In order to verify the feasibility of using FOHPS to measure the pore water pressure outdoors, sensors S1, S2 and S3 were installed in the pavement as shown in Fig. 3. The test site was instrumented during roadway construction to prevent unnecessary disturbance to the permeability of asphalt concrete. Before the lower asphaltic concrete layer was in place, a core specimen with a diameter of 150mm was drilled out from the old asphalt pavement by using a coring machine. Then, sensor S1 was installed horizontally in the drill hole C and subsequently covered by coarse gravel. Third, the drill hole C was strengthened and covered by polyester fiberglass cloth. Finally, the lower asphaltic concrete layer was constructed over the drill hole C. Sensor S2 was installed in the same approach. After the asphalt surface was in place, sensor S3 was installed horizontally in the drill hole A. In order to prevent the wheel load directly applying on the sensor, the upper part of sensor S3 was adjusted with 2cm distance to the pavement surface. In order to measure the pore water pressure under the traveling vehicle load, the drill holes B and C were filled with water using a fine steel pipe and then sealed. For drill hole A, using a roadway flusher to sprinkle water, so we can ensure that the pavement were in state of water flowing and like on a rainy day. While measuring the pore water pressure, a car was driven to pass over the holes A, B and C at the speeds of 20 km /h, 40 km /h, 60 km /h, 80 km /h and 100 km /h. Sometimes the car maybe
can’t pass just over the sensors, so at each running speeds the dynamic hydraulic pressures in asphalt pavement were detected several times. The outputs of these sensors were monitored by a si425-500 optical sensing interrogator with the acquisition speed 250Hz. The weight of the car is 1.4 tons, and its axle base is 2.578m. The car’s tire specifications is 205/55 R16, and tire pressure is 0.26MPa. Fig. 4 shows the pore water pressures measured by sensor S3 at pavement surface. In Fig. 4, it can be seen that the pore water pressures grows with the increment of speeds. When the vehicle speeds are 20km/h, 40km/h, 60km/h, 80km/h and 100km/h, the value of the pore water pressure at the pavement surface are 31.0kPa, 45.2kPa, 80.8kPa, 90.9kPa and 170.2kPa respectively. The observed results are expected. The theoretical analysis of Li et al [15] on the pore water pressures also showed an increase in pore water pressure on pavement surface with increasing vehicle speed, namely, the pore water pressure on pavement surface is in proportion to the squared traveling speed [15]. Additionally, Li etal also measured the pore water pressures on pavement surface by means of electromagnetic pressure sensor in situ [15]. For a given vehicle speed the pore water pressures measured by FOHPS in this study is different from that measured by electromagnetic pressure sensor. This difference was attributed to the different measurement conditions, such as the different kind of asphalt pavement and different tire pressure. However, the pore water pressure on pavement surface measured by FOHPS appears to be more close to the theoretical value compared to the pore water pressure measured by electromagnetic pressure sensor. So it can be confirmed that the fiber optic hydraulic pressure sensor designed is excellent and accurate. Furthermore, the lifetime of pore water pressure is directly related to the load duration and the load duration is inversely proportional to the vehicle speed, so the lifetime of pore water pressure decreases with
the increasing speed. The authors have previously described this phenomenon [11]. Pore water pressure is so high that it will inevitably lead to substantial increase in the permeability of the asphalt concrete. Further, a small amount of water would infiltrate into asphalt pavement under the pore water pressure and gathered, which may damage the pavement seriously. Fig. 5 illustrates the pore water pressures measured by sensor S1 and sensor S2 in pavement. In Fig. 5, it can also be seen that the pore water pressures grows with the increment of speeds. For vehicle speeds of 20km/h, 40km/h, 60km/h, 80km/h and 100km/h, the pore water pressures measured by sensor S2 in pavement at depth of 4cm are 48.7kPa, 99.2kPa, 109.2kPa, 125.1kPa and 144.8kPa respectively. In addition, the pore water pressures measured by sensor S1 in pavement at depth of 10cm are 46.8kPa, 54.2kPa, 70.0kPa, 100.2kPa and 103.4kPa respectively. For the same vehicle speed the pore water pressures in pavement are different at different depths. The pore water pressure in pavement at depth of 4cm is higher than it at depth of 10cm.
Fig. 3 Roadway cross section showing sensors placement Fig. 4 Relationship between vehicle speed and pore water pressure at pavement surface Fig. 5 Relationship between vehicle speed and pore water pressure in pavement
5. Permeability of asphalt mixtures Permeability is an important performance of asphalt concrete. Permeability is related to the air void and pore structure of asphalt concrete. On the other hand, it is also affected by the external pressure. Permeability of asphalt concrete will be enhanced with the increment of air void or interconnected air void. Moreover, the higher the external water pressures the higher permeability
of asphalt concrete. A highly permeable asphalt concrete helps to deposit excessive water underneath it. Excessive water can lead to excessive deflection, cracking, fracture, and reduction in pavement support, and thereby it endangers the base. High permeability can also cause stripping of asphalt coatings from the aggregate’s surface. Many permeability tests so far were always carried out at low pressure and low velocity conditions, and the pressure was about tens of centimeters high water column. In fact, the pore water pressures in pavement were much higher than tens of centimeters of water column under the vehicle load if there was excessive water in pavement. In our paper, the pore water pressure measured by FOHPS in pavement is all above 30.0kPa. So a permeability test method under high pressure state was proposed and the experimental results of SMA-13 and AC-20 mixes were analyzed. According to AASHTO T215-70 and ASTM D2434-74, the fluid flow will be in laminar conditions when the hydraulic gradient is from 0.2 to 0.3 for low compactness material or from 0.3 to 0.5 for high compactness material. The permeability coefficient will decrease with the increment of water head difference if it is higher, and the permeability coefficient is in a nonlinear trend which is the characteristics of turbulence. But the low hydraulic gradient or water head difference will make the data error enlargement. According to the practical pore water pressure in pavement, the high pressure head was chosen in our test. The permeability coefficient can be calculated as
k=
h ln 1 C Ft h2
αL
(3)
Where α is internal cross-sectional area of the graduated cylinder (cm2), L is the average thickness or length of the sample (cm), F is average cross-sectional area of the test sample (cm2), t
is elapsed time between h1 and h2 (s), h1 is the initial head, h2 is the final head, and C is temperature correction for viscosity, where a temperature of 20°C is used as a standard. The permeability test system we designed is shown in Fig. 6. The pore water pressure at the asphalt pavement surface was calculated as 0.44MPa [8]. In addition, the pore water pressure measured at the pavement surface in our paper is over 0.3MPa in high-speed conditions. So the maximum water pressure applied in test is 0.35 MPa, which was generated by air pump. The vacuum saturated core specimen with a diameter of 150mm was installed in mould and then the sides of it were sealed with the sealing material. The core specimens were all drilled from the test site where the pore water pressures were measured.
Fig. 6 Schematic of laboratory setup for permeability test Fig. 7 Relationship between permeability and hydraulic pressure
The permeability coefficients of core samples in different water pressure were listed in Table 1. The test results of specimens CS1 and CS2 were shown in Fig. 7. From Fig. 7, it can be shown that the permeability coefficient decreased with the water pressure being higher. This indicated that when the pressure is from 40kPa to 350kPa, the water flow in asphalt mixture is in turbulent. The measurement results of Xu et al [16 ]and Ma et al [17 ] on permeability coefficient of porous asphalt mixture also showed a decrease in permeability with increasing water head (i.e., increasing hydraulic gradient ). When the water head imposed to asphalt mixture is lower than 0.2 cm, the permeability coefficient of porous asphalt mixture tends to be constant. In this study, the water head is higher than 4 m. So observing the phenomenon that the permeability coefficient of asphalt
mixture decreases with increasing hydraulic pressure is reasonable. Though the permeability coefficient of asphalt mixture at high water pressure is slightly lower than the one at low water pressure, the water infiltration rate of asphalt mixture at high water pressure is much larger than that at low water pressure. The relationship between the water infiltration rate and water pressure manifests as an approximate linear curve as shown in Fig.7. For CS1, the water infiltration rate is 364.2ml/min under 40kPa, but when the water pressure increases to 350kPa, the water infiltration rate is almost 1241.2 ml/min, with an increase of 3.4 times. For CS2, the water infiltration rate is 103.3ml/min under 40kPa, but when the water pressure increases to 350kPa, the water infiltration rate is almost 282 ml/min, with an increase of 2.7 times. The permeability of asphalt mixture which is usually regarded as impermeable maybe greatly increased and up to 1241.2ml/min if the vehicle moved with high speed. Excessive water will penetrate deep into the pavement and be deposited on the base, which may directly induce the moisture damage in asphalt pavement.
Fig.8 Relationship between pore water pressure and permeability coefficient
In addition, from Table 1 it can be seen that the permeability coefficient decreased gradually with the increase of the water pressure applied on the samples. In order to analyze the relationship between the permeability coefficient and the applied water pressure, the permeability coefficient of asphalt mixture at pressure of 40kPa is regarded as the baseline, and the normalized regression analysis is shown in Fig. 8. The permeability coefficient of the asphalt mixture at different pressures can be expressed as
k = 5.6408 × k0 × p -0.47362
(4)
Where k is the permeability coefficient of asphalt mixture (10-4 cm / s), p is hydraulic pressure (kPa), k0 is the permeability coefficient at the pressure of 40 kPa (10-4 cm / s). According to Darcy’s law, the water infiltration into asphalt mixture can be given as
Q = kF
∆h L
(5)
Where Q is water infiltration rate (cm3/s), k is the permeability coefficient (cm/s) L is the average thickness or length of the sample (cm), F is average cross-sectional area of the test sample (cm2), ∆h is the water head (cm). When the water pressure p was applied to the upper surface of the sample, the water head ∆h can be expressed as
∆h =
100 × p 9.8
(6)
Accounting for Eq. (4), Eq. (5) can be written as follows:
Q = 57.56
k0 F × p 0.52638 L
(7)
Where Q is water infiltration rate (10-4 cm3 /s), p is water pressure (kPa), k0 is the permeability coefficient at the pressure of 40 kPa (10-4 cm / s). For a given core sample, its thickness L, cross-sectional area F and the permeability coefficient k0 are constant. From Eq. (7), it can be seen that the water infiltration rate Q is a function of water pressure p. The water infiltration rate increases on an approximate linear curve as the hydraulic pressure increased from 40kPa to 350kPa, as shown in Fig.7.
6. Summary and conclusions Infiltration of water into the pavement can affect the durability of that pavement. Water infiltration also endangers the base course by developing excess water underneath the asphalt mix layer. This
in turn develops pore pressure leading to excessive deflection, cracking, and reduction in load carrying capacity of the pavement. An effective way to study its hazards is to measure the value of the pore water pressure accurately. By using of the fiber grating and some suitable materials, a fiber optic hydraulic pressure sensor is designed. Firstly, the theoretical correlation between the applied pressure and the center wavelength shift of the FOHPS is derived by laboratory experiment. Secondly, a calibration test for FOHPS was implemented to obtain the parameters of the calibration curve. In order to verify the feasibility of using FOHPS to measure the pore water pressure outdoors, three FOHPS were installed in the pavement, and the pore water pressures in asphalt pavement at some running speeds were measured in situ. Moreover, a falling head permeameter method was used to measure the permeability coefficients of asphalt mixture exposed to hydraulic pressures which were from 40kPa to 350kPa, and the correlation between permeability and hydraulic pressure was obtained. The experimental results showed that the pore water pressure would increase with the increasing vehicle speed, but the lifetime of pore water pressure decreases with the increasing speed, which is consistent with the findings presented by previous researchers. The permeability coefficients of SMA-13 and AC-20 mixtures decrease with the increasing hydraulic pressure, but the water infiltration rate increases on an approximate linear curve as the hydraulic pressure increased from 40kPa to 350kPa.
Acknowledgments This work is supported by “the Fundamental Research Funds for the Central Universities”，NO.
NS2013012.
References
[1] Kutay ME, Aydilek A H. Dynamic Effects on Moisture Transport in Asphalt Concrete. Journal of Transportation Engineering 2007; 133(7): 406-414. [2] Kutay ME, Aydilek AH, Masad E. Laboratory validation of lattice Boltzmann method for modeling pore-scale flow in granular materials. Computers and Geotechnics 2006; 33(8), 381-395. [3] Kutay ME, Aydilek AH. Pore Pressure and Viscous Shear Stress Distribution Due to Water Flow within Asphalt Pore Structure. Computer-Aided Civil and Infrastructure Engineering 2009; 24(3): 212-224. [4] Masad E, Birgisson B, Al-Omari A, Cooley A. Analytical derivation of permeability and numerical simulation of fluid flow in hot-mix asphalt. Journal of Materials in Civil Engineering 2004; 16(5): 487-496. [5] Al-Omari A, Masad E. Three dimensional simulation of fluid flow in x-ray CT images of porous media. International Journal for Numerical and Analytical Methods in Geomechanics 2004; 28(13), 1327-1360. [6] Kettil P, Engstrom G, Wiberg N-E. Coupled hydro-mechanical wave propagation in road structures. Computers and structures 2005; 83(21-22): 1719-1729. [7] Zhou CH, Wang ZR, Chen JY, Qiao YJ. Numerical computation and analysis on dynamic pore water pressure in asphalt pavement. International Conference on Transportation Engineering 2007, American Society of Civil Engineers, 2007: 2981-2986.
[8] Dong ZJ, Tan YQ, Cao LP, Zhong Y. Research on pore pressure within asphalt pavement under the coupled moisture- loading action. Journal of Harbin Institute of Technology 2007; 39(10): 1614-1617. [9] Cui XZ, Jin Q. The dynamic response of saturated asphalt pavement under wheel loads. Journal of Shandong University 2008; 38(5):19-24. [10] Li ZG, Deng XY. Axial symmetric elastic solution of pore water pressure in asphalt pavement under mobile load. Journal of Southeast University 2008; 38(5): 804-810. [11] Gao JQ, Chen H, Ji TJ, Liu HY. Measurement of dynamic hydraulic pressure in asphalt pavement using fiber Bragg grating. Transducer and Microsystem Technologies 2009; 28(9):59-61. [12] Kanitpong K, Benson CH, Bahia HU. Hydraulic conductivity (permeability) of laboratory compacted asphalt mixtures. Proceedings, 80th Annual Meeting of the Transportation Research Board, Washington, D.C. 2001, 1767: 25-32. [13] Tarefder RA, White L, Zaman M. Neural network model for asphalt concrete permeability. Journal of Materials in Civil Engineering 2005; 17(1), 19-27. [14] Ernest O Doebelin. Measurement systems: application and design. Boston: McGraw-Hill; 2004. [15] Li SB, Zhang HC, Sun LJ. Development and simulation measurement of dynamic hydraulic pressure. Journal of Tongji University 2004; 35(7), 915-918. [16] Xu H, Ni FJ , Liu QQ, Shen H, Chen RS. Research on hydraulic conductivity of porous asphalt mixture [J]. China Journal of Highway and Transport 2004; 17(3), 1-5.
[17] Ma X, Ni FJ, Wang Y, Chen RS. Test and Analysis on Permeability of Porous Asphalt Mixture. Journal of Building Materials 2009; 12(2), 168-172.
Table1 Permeability coefficient of asphalt mixtures Permeability coefficient (10-4 cm/s) Core Gradation
40
60
kPa
kPa
80
100
150
200
250
300
350
kPa
kPa
kPa
kPa
kPa
kPa
kPa
specimen
CS1
SMA-13
2.79
2.18
1.74
1.64
1.44
1.24
1.13
1.07
1.15
CS2
SMA-13
1.05
0.77
0.66
0.61
0.49
0.41
0.38
0.36
0.35
CS3
AC-20
5.27
4.41
4.05
3.70
2.88
2.57
2.35
2.21
2.08
CS4
AC-20
0.28
0.22
0.20
0.18
0.15
0.12
0.11
0.11
0.10
CS5
AC-20
4.91
4.25
3.57
3.32
2.68
2.27
2.15
2.07
1.85
We propose a fiber optic hydraulic pressure sensor composed of FBG and other parts. The pore water pressure in asphalt pavement was measured by using this sensor in situ. The pore water pressure in asphalt pavement would increase with the increasing vehicle speed. The pore water pressure in pavement at depth of 4cm is higher than it at depth of 10cm. The permeability of asphalt mixture at different hydraulic pressures was obtained.
S0263-2241(14)00561-2 http://dx.doi.org/10.1016/j.measurement.2014.11.013 MEASUR 3132
To appear in:
Measurement
Received Date: Revised Date: Accepted Date:
3 May 2013 13 November 2014 19 November 2014
Please cite this article as: J. Gao, C. Guo, Y. Liu, Measurement of pore water pressure in asphalt pavement and its effects on permeability, Measurement (2014), doi: http://dx.doi.org/10.1016/j.measurement.2014.11.013
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Measurement of pore water pressure in asphalt pavement and its effects on permeability Junqi Gao*, Chengcheng Guo, Yutao Liu Department of Civil Engineering, Nanjing University of Aeronautics and Astronautics,29 Yudao Street, Nanjing 210016, China
Abstract: On the effect of pore water pressure in a saturated asphalt pavement, fluid velocity and water infiltration rate will increase, which have a disadvantage influence on asphalt mixture strength. In order to measure the pore water pressure in pavement, a fiber optic hydraulic pressure sensor (FOHPS) is designed. The theoretical correlation between the applied pressure and the center wavelength shift of the FOHPS is derived by laboratory experiment. The pore water pressures in asphalt pavement at some running speeds were measured in situ. Furthermore, a falling head permeameter method was used to measure the permeability coefficients of asphalt mixture exposed to hydraulic pressures which were from 40kPa to 350kPa, and the correlation between permeability and hydraulic pressure was obtained. The experimental results showed that the pore water pressure would increase with the increasing car’s speed, but the lifetime of pore water pressure decreases with the increasing speed. The permeability coefficients of SMA-13 and AC-20 mixtures decrease with the increasing hydraulic pressure, but the water infiltration rate increases on an approximate linear curve as the hydraulic pressure increased from 40kPa to 350kPa.
Keywords: Pore water pressure, Fiber optic hydraulic pressure sensor, FBG, Asphalt pavement, *Corresponding author. Tel./fax: +86 25 84891754 E-mail address: [email protected]
Permeability coefficient
1. Introduction Asphalt concrete pavement is directly exposed to the natural environment, and traffic loads, environmental temperature and moisture are the major external factors which influence pavement performance. Some cracks will appear in the surface layer under the action of external loads. Water will infiltrate into pavement through connected pores and cracks. Under the repeated action of wheel loads, the pore water pressure in asphalt pavement is formed, which would make pavement surface layer spalling or loose, and finally cause pavement structural damage. Pore water pressure in pavement is considered to be one of the major causes of the asphalt pavement damages. There have been significant efforts in researching pore water pressure in asphalt pavement. Kutay et al. [1-3] computed dynamic water pressure gradient, pore water pressure and shearing stress at the solid–water interfaces by means of the lattice Boltzmann method. Masad [4,5] calculated the distribution of the pore water pressure gradient by means of the finite difference method. Kettil et al. [6] simulated wet asphalt pavement deformation and water flow. Zhou et al. [7] calculated the pore water pressure in saturated asphalt pavement and showed that when the permeability coefficient of asphalt mixture is 1×10-4cm/s and the load time is 0.005s, the pore water pressure is 0.57MPa. Dong et al. [8] also computed the pore water pressure in the asphalt pavement surface, whose value is 0.44MPa. Cui and Jin [9] studied the pore water pressure in asphalt pavement. Li and Deng [10] built a theory model in which the asphalt pavement was regarded as an axial symmetrical body of multilayered saturation elastic half space, and pore water pressure in asphalt pavement under mobile load was calculated. In addition, Gao et al. [11]
measured the dynamic water pressure in asphalt pavement surface. The results showed that when car speed is 80km/h, the water pressure is about 0.2MPa. Permeability of an asphalt pavement is one of the most important parameters that have a direct influence on its design life. Fixed head test and falling head test are often used to measure permeability coefficient of asphalt mixture [12]. Moreover, some models which are used to estimate asphalt mixture permeable coefficient have been suggested [13]. The Kozeny–Carman equation has been used over the years to approximate the permeability of granular materials. It was derived based on representing the air voids as capillary tubes and applying the hydraulic radius theory. Based on Kozeny-Carman equation, Masad presented a simple equation for approximating the permeability of asphalt mixes [14]. The permeability coefficient is required to be provided to calculate the pore water pressure by finite element model, and it is generally considered as constant. This assumption would be appropriate at low pressure head and not appropriate for higher hydrodynamic pressure head. At present, field data of pore water pressure in asphalt pavement is still relatively few. In order to verify some established theory models, it is required to obtain a lot of field measurements of the pore water pressure. Fiber Bragg grating (FBG) sensing technique can overcome the weak of conventional testing method, and the pore water pressure can be measured accurately by FBG sensing. In this paper, a FOHPS is designed, and using this fiber sensor to measure the pore water pressure in asphalt pavement is studied. After calibrating the fiber sensor, a field pressure testing was conducted. In addition, the permeabilities of different asphalt specimens exposed to pore water pressures are studied using water and a falling head approach.
2. Optical fiber sensor design
The FOHPS is designed as shown in Fig.1. The appearance of the sensor is like a cylinder, which is composed of an aluminum enclosure, a circular diaphragm, a fiber grating and seals. When the water in asphalt pavement flow into the sensor and act on the diaphragm, the diaphragm will produce deformation that transfers to the fiber grating pasted on back surface centre of the diaphragm. Then axial strain will occur in the grating. By detecting the wavelength shift of the grating, we can obtain the pore water pressure that exists in saturated asphalt pavement under loads.
Fig. 1 Fiber optic hydraulic pressure sensor
The radius of the designed sensor is 27mm with a height of 26mm, and the diaphragm radius is 15 mm. The fiber grating is attached to the back surface of the diaphragm in a two-step process. First, the diaphragm is cleaned with alcohol until the alcohol evaporates. Second, a tension is applied to the fiber grating and then the grating is bonded on the diaphragm with an epoxy resin, which can also protect the sensor from damage. Under the action of uniform pressure p [11], the radial strain on the point with a distance r from the diaphragm center can be expressed as
εr =
3p (1 − µ 2 )( R 2 − 3r 2 ) 2 8h E
(1)
Where E is the Young’s modulus of the metal diaphragm, µ is Poisson’s ratio, h is the thickness of the diaphragm, and R is the radius of the metal diaphragm. From Eq. (1), it can be seen that the grating strain εr varies according to different distance r, and at the center of the round diaphragm its strain εr is maximum. In the designed sensor the length of
the grating is about 10mm, and the diaphragm radius R is 15 mm. By using these data, the FBG average strain can be calculated as
εr =
2.89 p (1 − µ 2 ) R 2 2 8h E
(2)
By using Eq. (2), the linear relationship between strain εr detected by FBG and pressure p can be observed when other variables does not change.
3. Calibration tests Three fiber optic hydraulic pressure sensors (S1, S2 and S3) were designed. The peak wavelengths of S1, S2 and S3 are 1558.365nm, 1561.447nm and 1555.552nm respectively. Diaphragm materials used in these sensors are stainless steel, whose elastic modulus is 200GPa, Poisson's ratio is 0.3, and thickness is 0.3mm. For sensors S1, S2 and S3, the pressure loads were 0.05MPa, 0.1MPa, 0.14MPa, 0.2MPa, 0.24MPa, 0.3MPa, 0.35MPa, respectively. During the load procedure, the pressure load was applied by using the pneumatic pump and maintained for 5 minutes to measure the strain of diaphragm. The outputs of these sensors were monitored by a si425-500 optical sensing interrogator (Micron Optics, Inc., USA). Furthermore, the si425-500 communicated with the control terminal, a notebook computer, via an Ethernet cable for the transmission of the enormous monitoring data. Before the field measurement, a calibration test for sensors S1, S2 and S3 was implemented to evaluate their respective behavior of linearity. The measurement results here and subsequent have been corrected for temperature compensation using a free FBG. Fig. 2 demonstrated the correlation between the center wavelength shift of the FBG and the applied pressure, and a good linear agreement, with a correlation coefficient of 0.998, can be well observed for S1, S2 and S3. For S1, S2 and S3, the linear regression slope coefficients are 4.0nm/MPa,4.9nm/MPa, and
4.5nm/MPa respectively.
Fig. 2 Relationship between applied pressure and center wavelength shift of the sensor
4. Field measurement of pore water pressure In order to verify the feasibility of using FOHPS to measure the pore water pressure outdoors, sensors S1, S2 and S3 were installed in the pavement as shown in Fig. 3. The test site was instrumented during roadway construction to prevent unnecessary disturbance to the permeability of asphalt concrete. Before the lower asphaltic concrete layer was in place, a core specimen with a diameter of 150mm was drilled out from the old asphalt pavement by using a coring machine. Then, sensor S1 was installed horizontally in the drill hole C and subsequently covered by coarse gravel. Third, the drill hole C was strengthened and covered by polyester fiberglass cloth. Finally, the lower asphaltic concrete layer was constructed over the drill hole C. Sensor S2 was installed in the same approach. After the asphalt surface was in place, sensor S3 was installed horizontally in the drill hole A. In order to prevent the wheel load directly applying on the sensor, the upper part of sensor S3 was adjusted with 2cm distance to the pavement surface. In order to measure the pore water pressure under the traveling vehicle load, the drill holes B and C were filled with water using a fine steel pipe and then sealed. For drill hole A, using a roadway flusher to sprinkle water, so we can ensure that the pavement were in state of water flowing and like on a rainy day. While measuring the pore water pressure, a car was driven to pass over the holes A, B and C at the speeds of 20 km /h, 40 km /h, 60 km /h, 80 km /h and 100 km /h. Sometimes the car maybe
can’t pass just over the sensors, so at each running speeds the dynamic hydraulic pressures in asphalt pavement were detected several times. The outputs of these sensors were monitored by a si425-500 optical sensing interrogator with the acquisition speed 250Hz. The weight of the car is 1.4 tons, and its axle base is 2.578m. The car’s tire specifications is 205/55 R16, and tire pressure is 0.26MPa. Fig. 4 shows the pore water pressures measured by sensor S3 at pavement surface. In Fig. 4, it can be seen that the pore water pressures grows with the increment of speeds. When the vehicle speeds are 20km/h, 40km/h, 60km/h, 80km/h and 100km/h, the value of the pore water pressure at the pavement surface are 31.0kPa, 45.2kPa, 80.8kPa, 90.9kPa and 170.2kPa respectively. The observed results are expected. The theoretical analysis of Li et al [15] on the pore water pressures also showed an increase in pore water pressure on pavement surface with increasing vehicle speed, namely, the pore water pressure on pavement surface is in proportion to the squared traveling speed [15]. Additionally, Li etal also measured the pore water pressures on pavement surface by means of electromagnetic pressure sensor in situ [15]. For a given vehicle speed the pore water pressures measured by FOHPS in this study is different from that measured by electromagnetic pressure sensor. This difference was attributed to the different measurement conditions, such as the different kind of asphalt pavement and different tire pressure. However, the pore water pressure on pavement surface measured by FOHPS appears to be more close to the theoretical value compared to the pore water pressure measured by electromagnetic pressure sensor. So it can be confirmed that the fiber optic hydraulic pressure sensor designed is excellent and accurate. Furthermore, the lifetime of pore water pressure is directly related to the load duration and the load duration is inversely proportional to the vehicle speed, so the lifetime of pore water pressure decreases with
the increasing speed. The authors have previously described this phenomenon [11]. Pore water pressure is so high that it will inevitably lead to substantial increase in the permeability of the asphalt concrete. Further, a small amount of water would infiltrate into asphalt pavement under the pore water pressure and gathered, which may damage the pavement seriously. Fig. 5 illustrates the pore water pressures measured by sensor S1 and sensor S2 in pavement. In Fig. 5, it can also be seen that the pore water pressures grows with the increment of speeds. For vehicle speeds of 20km/h, 40km/h, 60km/h, 80km/h and 100km/h, the pore water pressures measured by sensor S2 in pavement at depth of 4cm are 48.7kPa, 99.2kPa, 109.2kPa, 125.1kPa and 144.8kPa respectively. In addition, the pore water pressures measured by sensor S1 in pavement at depth of 10cm are 46.8kPa, 54.2kPa, 70.0kPa, 100.2kPa and 103.4kPa respectively. For the same vehicle speed the pore water pressures in pavement are different at different depths. The pore water pressure in pavement at depth of 4cm is higher than it at depth of 10cm.
Fig. 3 Roadway cross section showing sensors placement Fig. 4 Relationship between vehicle speed and pore water pressure at pavement surface Fig. 5 Relationship between vehicle speed and pore water pressure in pavement
5. Permeability of asphalt mixtures Permeability is an important performance of asphalt concrete. Permeability is related to the air void and pore structure of asphalt concrete. On the other hand, it is also affected by the external pressure. Permeability of asphalt concrete will be enhanced with the increment of air void or interconnected air void. Moreover, the higher the external water pressures the higher permeability
of asphalt concrete. A highly permeable asphalt concrete helps to deposit excessive water underneath it. Excessive water can lead to excessive deflection, cracking, fracture, and reduction in pavement support, and thereby it endangers the base. High permeability can also cause stripping of asphalt coatings from the aggregate’s surface. Many permeability tests so far were always carried out at low pressure and low velocity conditions, and the pressure was about tens of centimeters high water column. In fact, the pore water pressures in pavement were much higher than tens of centimeters of water column under the vehicle load if there was excessive water in pavement. In our paper, the pore water pressure measured by FOHPS in pavement is all above 30.0kPa. So a permeability test method under high pressure state was proposed and the experimental results of SMA-13 and AC-20 mixes were analyzed. According to AASHTO T215-70 and ASTM D2434-74, the fluid flow will be in laminar conditions when the hydraulic gradient is from 0.2 to 0.3 for low compactness material or from 0.3 to 0.5 for high compactness material. The permeability coefficient will decrease with the increment of water head difference if it is higher, and the permeability coefficient is in a nonlinear trend which is the characteristics of turbulence. But the low hydraulic gradient or water head difference will make the data error enlargement. According to the practical pore water pressure in pavement, the high pressure head was chosen in our test. The permeability coefficient can be calculated as
k=
h ln 1 C Ft h2
αL
(3)
Where α is internal cross-sectional area of the graduated cylinder (cm2), L is the average thickness or length of the sample (cm), F is average cross-sectional area of the test sample (cm2), t
is elapsed time between h1 and h2 (s), h1 is the initial head, h2 is the final head, and C is temperature correction for viscosity, where a temperature of 20°C is used as a standard. The permeability test system we designed is shown in Fig. 6. The pore water pressure at the asphalt pavement surface was calculated as 0.44MPa [8]. In addition, the pore water pressure measured at the pavement surface in our paper is over 0.3MPa in high-speed conditions. So the maximum water pressure applied in test is 0.35 MPa, which was generated by air pump. The vacuum saturated core specimen with a diameter of 150mm was installed in mould and then the sides of it were sealed with the sealing material. The core specimens were all drilled from the test site where the pore water pressures were measured.
Fig. 6 Schematic of laboratory setup for permeability test Fig. 7 Relationship between permeability and hydraulic pressure
The permeability coefficients of core samples in different water pressure were listed in Table 1. The test results of specimens CS1 and CS2 were shown in Fig. 7. From Fig. 7, it can be shown that the permeability coefficient decreased with the water pressure being higher. This indicated that when the pressure is from 40kPa to 350kPa, the water flow in asphalt mixture is in turbulent. The measurement results of Xu et al [16 ]and Ma et al [17 ] on permeability coefficient of porous asphalt mixture also showed a decrease in permeability with increasing water head (i.e., increasing hydraulic gradient ). When the water head imposed to asphalt mixture is lower than 0.2 cm, the permeability coefficient of porous asphalt mixture tends to be constant. In this study, the water head is higher than 4 m. So observing the phenomenon that the permeability coefficient of asphalt
mixture decreases with increasing hydraulic pressure is reasonable. Though the permeability coefficient of asphalt mixture at high water pressure is slightly lower than the one at low water pressure, the water infiltration rate of asphalt mixture at high water pressure is much larger than that at low water pressure. The relationship between the water infiltration rate and water pressure manifests as an approximate linear curve as shown in Fig.7. For CS1, the water infiltration rate is 364.2ml/min under 40kPa, but when the water pressure increases to 350kPa, the water infiltration rate is almost 1241.2 ml/min, with an increase of 3.4 times. For CS2, the water infiltration rate is 103.3ml/min under 40kPa, but when the water pressure increases to 350kPa, the water infiltration rate is almost 282 ml/min, with an increase of 2.7 times. The permeability of asphalt mixture which is usually regarded as impermeable maybe greatly increased and up to 1241.2ml/min if the vehicle moved with high speed. Excessive water will penetrate deep into the pavement and be deposited on the base, which may directly induce the moisture damage in asphalt pavement.
Fig.8 Relationship between pore water pressure and permeability coefficient
In addition, from Table 1 it can be seen that the permeability coefficient decreased gradually with the increase of the water pressure applied on the samples. In order to analyze the relationship between the permeability coefficient and the applied water pressure, the permeability coefficient of asphalt mixture at pressure of 40kPa is regarded as the baseline, and the normalized regression analysis is shown in Fig. 8. The permeability coefficient of the asphalt mixture at different pressures can be expressed as
k = 5.6408 × k0 × p -0.47362
(4)
Where k is the permeability coefficient of asphalt mixture (10-4 cm / s), p is hydraulic pressure (kPa), k0 is the permeability coefficient at the pressure of 40 kPa (10-4 cm / s). According to Darcy’s law, the water infiltration into asphalt mixture can be given as
Q = kF
∆h L
(5)
Where Q is water infiltration rate (cm3/s), k is the permeability coefficient (cm/s) L is the average thickness or length of the sample (cm), F is average cross-sectional area of the test sample (cm2), ∆h is the water head (cm). When the water pressure p was applied to the upper surface of the sample, the water head ∆h can be expressed as
∆h =
100 × p 9.8
(6)
Accounting for Eq. (4), Eq. (5) can be written as follows:
Q = 57.56
k0 F × p 0.52638 L
(7)
Where Q is water infiltration rate (10-4 cm3 /s), p is water pressure (kPa), k0 is the permeability coefficient at the pressure of 40 kPa (10-4 cm / s). For a given core sample, its thickness L, cross-sectional area F and the permeability coefficient k0 are constant. From Eq. (7), it can be seen that the water infiltration rate Q is a function of water pressure p. The water infiltration rate increases on an approximate linear curve as the hydraulic pressure increased from 40kPa to 350kPa, as shown in Fig.7.
6. Summary and conclusions Infiltration of water into the pavement can affect the durability of that pavement. Water infiltration also endangers the base course by developing excess water underneath the asphalt mix layer. This
in turn develops pore pressure leading to excessive deflection, cracking, and reduction in load carrying capacity of the pavement. An effective way to study its hazards is to measure the value of the pore water pressure accurately. By using of the fiber grating and some suitable materials, a fiber optic hydraulic pressure sensor is designed. Firstly, the theoretical correlation between the applied pressure and the center wavelength shift of the FOHPS is derived by laboratory experiment. Secondly, a calibration test for FOHPS was implemented to obtain the parameters of the calibration curve. In order to verify the feasibility of using FOHPS to measure the pore water pressure outdoors, three FOHPS were installed in the pavement, and the pore water pressures in asphalt pavement at some running speeds were measured in situ. Moreover, a falling head permeameter method was used to measure the permeability coefficients of asphalt mixture exposed to hydraulic pressures which were from 40kPa to 350kPa, and the correlation between permeability and hydraulic pressure was obtained. The experimental results showed that the pore water pressure would increase with the increasing vehicle speed, but the lifetime of pore water pressure decreases with the increasing speed, which is consistent with the findings presented by previous researchers. The permeability coefficients of SMA-13 and AC-20 mixtures decrease with the increasing hydraulic pressure, but the water infiltration rate increases on an approximate linear curve as the hydraulic pressure increased from 40kPa to 350kPa.
Acknowledgments This work is supported by “the Fundamental Research Funds for the Central Universities”，NO.
NS2013012.
References
[1] Kutay ME, Aydilek A H. Dynamic Effects on Moisture Transport in Asphalt Concrete. Journal of Transportation Engineering 2007; 133(7): 406-414. [2] Kutay ME, Aydilek AH, Masad E. Laboratory validation of lattice Boltzmann method for modeling pore-scale flow in granular materials. Computers and Geotechnics 2006; 33(8), 381-395. [3] Kutay ME, Aydilek AH. Pore Pressure and Viscous Shear Stress Distribution Due to Water Flow within Asphalt Pore Structure. Computer-Aided Civil and Infrastructure Engineering 2009; 24(3): 212-224. [4] Masad E, Birgisson B, Al-Omari A, Cooley A. Analytical derivation of permeability and numerical simulation of fluid flow in hot-mix asphalt. Journal of Materials in Civil Engineering 2004; 16(5): 487-496. [5] Al-Omari A, Masad E. Three dimensional simulation of fluid flow in x-ray CT images of porous media. International Journal for Numerical and Analytical Methods in Geomechanics 2004; 28(13), 1327-1360. [6] Kettil P, Engstrom G, Wiberg N-E. Coupled hydro-mechanical wave propagation in road structures. Computers and structures 2005; 83(21-22): 1719-1729. [7] Zhou CH, Wang ZR, Chen JY, Qiao YJ. Numerical computation and analysis on dynamic pore water pressure in asphalt pavement. International Conference on Transportation Engineering 2007, American Society of Civil Engineers, 2007: 2981-2986.
[8] Dong ZJ, Tan YQ, Cao LP, Zhong Y. Research on pore pressure within asphalt pavement under the coupled moisture- loading action. Journal of Harbin Institute of Technology 2007; 39(10): 1614-1617. [9] Cui XZ, Jin Q. The dynamic response of saturated asphalt pavement under wheel loads. Journal of Shandong University 2008; 38(5):19-24. [10] Li ZG, Deng XY. Axial symmetric elastic solution of pore water pressure in asphalt pavement under mobile load. Journal of Southeast University 2008; 38(5): 804-810. [11] Gao JQ, Chen H, Ji TJ, Liu HY. Measurement of dynamic hydraulic pressure in asphalt pavement using fiber Bragg grating. Transducer and Microsystem Technologies 2009; 28(9):59-61. [12] Kanitpong K, Benson CH, Bahia HU. Hydraulic conductivity (permeability) of laboratory compacted asphalt mixtures. Proceedings, 80th Annual Meeting of the Transportation Research Board, Washington, D.C. 2001, 1767: 25-32. [13] Tarefder RA, White L, Zaman M. Neural network model for asphalt concrete permeability. Journal of Materials in Civil Engineering 2005; 17(1), 19-27. [14] Ernest O Doebelin. Measurement systems: application and design. Boston: McGraw-Hill; 2004. [15] Li SB, Zhang HC, Sun LJ. Development and simulation measurement of dynamic hydraulic pressure. Journal of Tongji University 2004; 35(7), 915-918. [16] Xu H, Ni FJ , Liu QQ, Shen H, Chen RS. Research on hydraulic conductivity of porous asphalt mixture [J]. China Journal of Highway and Transport 2004; 17(3), 1-5.
[17] Ma X, Ni FJ, Wang Y, Chen RS. Test and Analysis on Permeability of Porous Asphalt Mixture. Journal of Building Materials 2009; 12(2), 168-172.
Table1 Permeability coefficient of asphalt mixtures Permeability coefficient (10-4 cm/s) Core Gradation
40
60
kPa
kPa
80
100
150
200
250
300
350
kPa
kPa
kPa
kPa
kPa
kPa
kPa
specimen
CS1
SMA-13
2.79
2.18
1.74
1.64
1.44
1.24
1.13
1.07
1.15
CS2
SMA-13
1.05
0.77
0.66
0.61
0.49
0.41
0.38
0.36
0.35
CS3
AC-20
5.27
4.41
4.05
3.70
2.88
2.57
2.35
2.21
2.08
CS4
AC-20
0.28
0.22
0.20
0.18
0.15
0.12
0.11
0.11
0.10
CS5
AC-20
4.91
4.25
3.57
3.32
2.68
2.27
2.15
2.07
1.85
We propose a fiber optic hydraulic pressure sensor composed of FBG and other parts. The pore water pressure in asphalt pavement was measured by using this sensor in situ. The pore water pressure in asphalt pavement would increase with the increasing vehicle speed. The pore water pressure in pavement at depth of 4cm is higher than it at depth of 10cm. The permeability of asphalt mixture at different hydraulic pressures was obtained.