- Email: [email protected]
Investigation on fatigue damage of asphalt mixture with different air-voids using microstructural analysis
Construction and Building Materials 125 (2016) 936–945
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Investigation on fatigue damage of asphalt mixture with different air-voids using microstructural analysis Jing Hu a, Pengfei Liu a, Dawei Wang a,b,⇑, Markus Oeser a,b, Yiqiu Tan b,c a
Institute of Highway Engineering, RWTH Aachen University, Aachen, Germany Germany Sino-European Research Center for Advanced Transportation Infrastructure Technology, Würselen, Germany c School of Transportation Science and Engineering, Harbin Institute of Technology, Harbin, China b
h i g h l i g h t s The relationship between air-voids ratio and compactness is linear. Initial air-voids number and temperature have influence on cracks distress. The effect of air-voids complexity on fatigue damage is significant.
a r t i c l e
i n f o
Article history: Received 11 January 2016 Received in revised form 17 July 2016 Accepted 28 August 2016
Keywords: Asphalt mixture Fatigue damage Digital Image Processing Microstructure Air-voids
a b s t r a c t Fatigue damage caused by vehicle loads is the main asphalt pavement distress, and it deteriorates the serviceability and strength seriously. In order to evaluate the fatigue damage of asphalt mixture under repeated load, microstructures were detected to investigate the morphology change of internal structures using Digital Image Processing (DIP). Field pavement named as ISAC test track was built, and cores were drilled to slice into test specimens with different air-voids that can reflect the real internal state of asphalt pavement. Fatigue properties were measured under temperature 10 °C, 0 °C and 10 °C respectively, the frequency of sinusoid load was 0.1 Hz, the minimum value was 0.035 MPa and maximum value was 0.5 MPa. Internal structures of asphalt mixture were scanned by X-ray Computed Tomography (XCT) device before and after fatigue damage, thus the relationship between microstructures and fatigue damage can be conducted. The results show that microstructural analysis can effectively determine the internal structure change of asphalt mixture. Charging compaction causes different air-voids distributions and morphologies, which have obvious influence on failure state of asphalt mixture. The effect of temperature and initial air-voids on fatigue performance is significant, and fatigue damage presents a linear relation with the complexity of air-voids. Methodology established in this paper provides an effective method for fatigue damage assessment. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction Asphalt mixture is a composite material that includes aggregates, air-voids and asphalt mastic, each microstructure has obvious influence on the performance. Air-voids are main structure component of asphalt pavement, and some functions such as drainage and noise reduction relate with air-voids significantly. However, air-voids can aggravate the fatigue damage under repeated load and reduce the strength of asphalt mixture, causing macro crack and some typical failures appear in asphalt pavement, such ⇑ Corresponding author at: Institute of Highway Engineering, RWTH Aachen University, Aachen, Germany. E-mail address: [email protected] (D. Wang). http://dx.doi.org/10.1016/j.conbuildmat.2016.08.138 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
as rutting and bleeding. Therefore, air-void can be regarded as a defect of asphalt mixture. Although the gradation of asphalt mixture is a critical factor for air-voids morphology, the compaction affects the spatial distribution of air-voids in asphalt pavement as well. Due to the different compaction energies the asphalt pavement suffers, the air-voids must be changed along with the depth, result in different states in different parts of pavement. Thus the effect of air-voids on fatigue performance should be investigated to evaluate the properties of asphalt pavement. In recently years, microstructure has become a prevalent research objective in pavement engineering, Digital Image Processing (DIP) base on the X-ray Computed Tomography (XCT) is generally used to analyze the failure mechanism and influencing factors of asphalt mixture. The development of pavement distress can be
937
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
2. Objective The investigation was aimed at providing an effective method for a better understanding of the change of internal structure after fatigue damage and the influence of microstructure characteristics on fatigue performance. A comprehensive research was conducted in order to achieve the following objectives: (1) Determining the relationship between pavement compaction and air-voids, and reconstructing the real distribution of air-voids in the asphalt mixture by DIP technology. (2) Counting the damage state of asphalt mixture, such as crack quantity, crack sharp and crack area, before and after fatigue test, and evaluating the effect of different temperatures (10 °C, 0 °C and 10 °C) and air-voids complexity on the fatigue performance of asphalt mixture. 3. Materials and methods
aggregates. The SMA-11S mixture was prepared using a PG 50/70 binder and diabase aggregates, Marshell method was used to determine the gradation of SMA-11S, the test temperature was 170 °C, the gradation of the diabase aggregates are shown in Fig. 1 and critical material parameters are listed in Table 1. 3.2. Test track construction, fatigue test and XCT scanning In order to obtain the internal structure of real asphalt pavement, the field test track was built at Institute of Highway Engineering of RWTH Aachen University and named as ISAC test track. Diabase aggregates were mixed at temperature of 170 °C. A brief overview about the procedure of the test track construction, fatigue test and XCT scanning follows. (1) A miniature paver was used to prepare the asphalt mixture, a heating part within the machine can keep the paving temperature higher than 150 °C. The length, width and thickness of the test track were 26 m, 1.2 m and 0.3 m respectively, as shown in Fig. 2(a). (2) Dividing ISAC test track into three layers, each layer has the same depth of 0.1 m. Using a miniature roller compacted these layers respectively, as shown in Fig. 2(b). It must be note that the test track was composed of three layers, thickness of each layer was 0.1 m after compression, the lower layer of test track suffers more compaction energy, so compaction degree was greater than that of other layers. Test track determined in this paper can provide different asphalt mixtures which have different porosity rate, these asphalt mixture samples are the foundation for investigating the effect of compaction (air-voids) on the fatigue property. (3) In order to obtain the internal structure of asphalt pavement, cylindrical cores were drilled at different sections of test track after maintenance, as shown in Fig. 2(c). Fig. 2(d) illustrates the cylindrical cores, which size were 150 mm in diameter and 300 mm in height. (4) Each cylindrical core can be sliced into five test specimens which size were 40 mm in height and 100 mm in diameter according to the demand of FGSV-Nr.430 of Germany, and test specimens were numbered base on their corresponding initial depth in the test track, as shown in Fig. 2(e). (5) The Universal Test Machine (UTM) was used to measure the fatigue performance of test specimens at test temperature of 10 °C, 0 °C and 10 °C in Fig. 2(f). The loading followed sinusoid mode that frequency was 0.1 Hz and stress control
100 90 80
Passing percentage/%
detected using the microstructure reconstructed by XCT images [1–6], and the presented literature shows that air-voids have obvious influence on the performance and durability of asphalt mixture [7]. Some researchers have certified the significant effect of airvoids on the failure of asphalt mixture, especially the mechanisms that air-voids affect the mechanical behavior of asphalt pavement are commonly involved base on the XCT scanning technology [8–11]. Coleri used XCT apparatus to scan the internal structure of asphalt pavement before and after high temperature deformation, then the air-voids change were measured and the effect of air-voids on rutting distress was evaluated [12,13], Sefidmazgi also analyzed the influence of microstructure on anti-deformation performance of asphalt mixture using XCT images and DIP [14]. Khan developed a 3D microstructure numerical model according to XCT data, the result shows that air-voids affect stress concentration and aggravate the pavement failure [15], besides, moisture damage of asphalt mixture was investigated as well [16]. In fact, air-voids are extended to form cracks distress commonly because of stress concentration, and the conclusion of presented research shows that air-voids have adverse influence on performance of asphalt mixture [17]. Furthermore, the air-voids distribution is uneven because of the different compaction energies during pavement construction. The components of asphalt mixture, especially the content of fine aggregates, determine the air-voids distribution [18]. Some microstructural characteristics of air-voids, such as the homogeneous of distribution state and discrete degree of morphology, can be analyzed by the XCT images and DIP [19–21]. However, most researchers study the asphalt mixture molded in laboratory, it is hard to reflect the real state of internal structure, and the investigations about the relationship between air-voids and fatigue damage are rare. Through the XCT scanning, this paper establishes the relationship between compaction and air-voids of asphalt mixture drilled from pavement layers in different depths. Air-voids were extracted, and DIP was used to determine morphology and distribution. Meanwhile, the effect of air-voids on fatigue performance and change of internal microstructure after fatigue damage were detected as well.
Asphalt mixture design Contral design Lower limit Upper limit
70 60 50 40 30 20
3.1. Laboratory design for material Initially, a Stone Mastic Asphalt (SMA) mixture with a nominal maximum aggregate size (NMAS) of 12.5 mm and bitumen of 6.9% was prepared and named as SMA-11S. The NMAS is defined as one sieve size larger than the first sieve to retain more than 10% of the
10 0 0.063
2
5.6 8 11.2
Sieve size/mm Fig. 1. Gradation of diabase aggregates of SMA-11S.
938
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
Table 1 Critical material parameters of SMA-11S.
*
Asphalt mixture
Virgin materials
Bitumen content (%)
Max. density (g/cm3)
Ave. density (g/cm3)
SMA-11S
Bitumen binder: Pen Bitumen 50/70 Fibers: Cellulose Filler (0/0.09 mm): Limestone Sand (0/2 mm): Diabase Coarse aggregates: Diabase
6.9
2.488
2.415
*
Void ratio (%) 2.090
Average value of 44 specimens.
Fig. 2. The illustration of test track construction, fatigue test and XCT scanning. (a) Paving ISAC test track. (b) Compacting ISAC test track. (c) Drilling cylindrical cores. (d) Cylindrical cores. (e) Test specimens used for fatigue test. (f) Device of Universal Test Machine. (g) Loading apparatus of fatigue test. (h) XCT scanning before and after fatigue test. (i) Sequence gray images (six gray images were selected as illustration).
mode was utilized in test, the range of loading was 0.035 MPa to 0.5 MPa to ensure the peak value of horizontal strain of test specimen from 0.05‰ to 0.3‰. Furthermore, test time was set as 2 h to assure that each test specimen bore the same energy, thus the impact of different internal structures on fatigue damage can be evaluated. The loading device is shown in Fig. 2(g). (6) Y.CT Precision S mode X-CT device was used to scan the test specimens and the internal structures can be obtained, as shown in Fig. 2(h). Scanning interval of 0.1 mm was set to assure the precision of gray image, the resolution of each
gray image was 1024 1024 Pixel2, and the size of pixel was 80 lm. The gray images were presented in Fig. 2(i). 3.3. Fracture mechanism of fatigue damage The test specimens were obtained at the different depths of asphalt pavement; air-voids ratio and distribution were changed due to the different compaction energies. In this paper, three specimens obtained from the same pavement depth were tested at the same temperature. At low temperature, assuming the fatigue damage is mainly caused by cracks distress, namely cracks deteriorate
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
to the results of fatigue test (an example illustrated in Fig. 3), assuming the asphalt mixtures are within micro-damage stage after fatigue test except the complete damage one, so the development of micro damage in asphalt mixture with different air-void characteristics can be evaluated.
15150
Modulus/MPa
15100 Test specimen #1 (30mm) Test specimen #2 (90mm) Test specimen #3 (150mm) Test specimen #4 (210mm) Test specimen #5 (270mm)
15050
15000 Fatigue Damage Point Cycle number: 28530 Modules: 15087MPa
14950 0
10000
20000
30000
40000
50000
60000
70000
Cycle number Fig. 3. Modulus state of asphalt mixture at test temperature of 10 °C.
n*modulus
Beginning Macro-cracks
0
Beginning Micro-cracks
Completed Fracture
0
939
Loading cycles:n
N
Fig. 4. Damage accumulation stage of asphalt mixture.
4. Digital Image Processing The internal structures of asphalt mixture are detected using XCT gray images, gray value is from 0 to 255. The gray value is determined by XCT device according to material density. As for asphalt mixture, the gray of aggregates have the maximum value and gray of air-voids have the minimum value. The gray images of asphalt mixture in different depths of pavement are shown in Fig. 5. In order to analyze the structure change of asphalt mixture, microstructure must be extracted by DIP at first. The main functions of DIP technology used in gray images are extraction and analysis, and the critical step of the whole research system is identifying the air-voids accurately. The Artificial Neural Network (ANN) has been developed to extract microstructure [22,23], besides, gray images of asphalt mixture can be converted into binary images to represent the different microstructures, and a technique named as Otsu has been proved to be an effective binary method for gray images [24,25]. However, the Otsu cannot solve the problem that uneven gray distribution of XCT image scanned from asphalt mixture. In this research, an improved Otsu method was developed to extract the binary image of air-voids [26], and then, the microstructure characteristics were analyzed by Matlab software, as shown in Fig. 6. Using the binary images of air-voids, 3D morphology models can be created, the distributions of air-voids are presented in Fig. 7. Fig. 7 shows the spatial air-voids state in different depths of pavement, air-voids of depth 30 mm and 90 mm are great, and decreased rapidly at the range deeper than 90 mm. In addition, individual air-voids shown in Fig. 7(a)–(c) have a large volume, and exhibit more complex shape than the air-voids shown in Fig. 7(e) and (f). 5. Discussing and analysis
the strength and durability of asphalt mixture. Modulus was used as an index to indicate the property of test specimen, and results of five typical specimens drilled from different pavement depth were selected to show the test situation, as shown in Fig. 3. Fig. 3 presents the modulus curves of five test specimens, which located at different depths of pavement. The test specimen drilled from depth of 30 mm was fractured completely during the fatigue test because of the large air-voids ratio, and the modulus is also less compared with other test specimens. The results indicate that the influence of air-voids on fatigue properties is significant, and fatigue can be regarded as a damage accumulation stage for asphalt mixture. Micro cracks are appeared initially, and they gradually develop into macro cracks, then, enough macro cracks will cause complete damage of asphalt mixture. Fig. 4 presents the damage accumulation stage of asphalt mixture introduced from regulation of FGSV-Nr.430, it reflects the development of internal structures with the increasing load, and three critical points can be determined through curve slope to divide the curve into three damage stage. The linear section is non-damage stage, which means there is not damage appeared in asphalt mixture. The first curve section is micro-damage stage, micro cracks appear after Beginning Micro-cracks point, and then, micro cracks are extended to form macro cracks after Beginning Macro-cracks point and asphalt mixture is fractured completely, this curve section represents the macro-damage stage. According
5.1. The feature of air-voids distribution Binary image is composed of pixels which size is 80 lm, therefore, the air-voids area at different depths of pavement can be calculated through the amount of pixel belonged to air-voids, and range of air-voids ratio are inferred, as shown in Fig. 8. Fig. 8 indicates the range of air-voids ratio along with the pavement depth. Due to different compaction energies, air-voids ratio decreases along with the depth of asphalt pavement, air-voids ratios are decreased significantly during the range 0 mm– 150 mm, especially in the range from 90 mm to 150 mm, all compaction degree of asphalt mixtures are increased obviously. The curves show that discrepancies of air-voids ratio nearby the pavement surface are great, such as the air-voids ratio at depth of 30 mm and 90 mm. However, air-voids ratios appear similar value at depth of 150 mm, this means that asphalt mixture below depth of 150 mm bears the similar compaction energies. Compaction can be calculated by density and theoretical maximum density, it is a quantitative indicator that represents the compaction energy. The relationship between compaction and airvoids should be determined, and then, microstructural index can be established to evaluate the compaction state. Fig. 9 presents the compaction degree of fifteen test specimens with respect to corresponding air-voids ratio.
940
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
Fig. 5. XCT gray images in different depths of test track. (a) 30 mm. (b) 90 mm. (c) 150 mm. (d) 210 mm. (e) 270 mm.
Fig. 6. Extracting air-voids from XCT gray image. (a) Origin gray image. (b) Binary image of aggregates. (c) Binary image of air-voids.
Fig. 7. 3D air-voids distributions in different depths. (a) 30 mm. (b) 90 mm. (c) 150 mm. (d) 210 mm. (e) 270 mm.
941
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
0.5
1.0
1.5
2.0
2.5
0
2.5
Air-voids ratio/%
Surface 50
Depth/mm
90mm
100 150 200
Temperature -10°C:Change of air-voids ratio Temperature 0°C:Change of air-voids ratio Temperature 10°C:Change of air-voids ratio Temperature -10°C:Before fatigue damge Temperature -10°C:After fatigue damge Temperature 0°C:Before fatigue damge Temperature 0°C:After fatigue damge Temperature 10°C:Before fatigue damge Temperature 10°C:After fatigue damge
3.0
150mm
Air-voids ratio - Lower limit Air-voids ratio - Average Air-voids ratio - Upper limit
2.0
Bottom Fig. 8. Air-voids ratio states of ISAC Test Track.
100
Compaction degree/%
99
Relationship between compaction degree and air-void ratio Fitting curve
98
0.3
0.2 1.0 0.1
0.5
0.0 30
300
0.4
1.5
0.0
250
0.5
Change of air-voids ratio/%
3.0
Air-voids ratio/% 0.0
90
150
210
270
Depth/mm Fig. 10. Air-voids ratio state of ISAC test track.
fatigue test, three test specimens (30 mm) were measured at 10 °C and other three (30 mm) were measured at 0 °C, the test results show that all of six test samples are totally damage during the fatigue test. Both great porosity and high temperature cause adverse effects on the fatigue performance of asphalt mixture.
97 96
5.2. The microstructural characteristics of air-voids
95
The state of internal structures of asphalt mixture can be evaluated by DIP. Initial air-voids are changed after damage, meanwhile, cracks appear in asphalt mastic and the interface between aggregate particles and asphalt mastic as well [27]. In this research, assuming the fatigue damage of asphalt mixture is caused by cracks, so, determining the extended air-void state and new cracks in each cross section of asphalt mixture can provide information for damage evaluation, as shown in Fig. 11. In order to determine the cross section pair that located at the same position of test specimen before and after fatigue damage, comparing the areas of aggregate particles and air-voids to select the appropriate cross sections. Two hundred cross section pairs in the range of 20 mm of test specimen center along with the thickness direction were selected as research objectives, and the change of internal structures (mainly includes air-voids and cracks) before and after fatigue damage was measured through comparing the gray images, as shown in Fig. 12. Fig. 12 presents the change of internal structures at different temperatures, the curves of internal structures in the same test specimen presents the similar trend before and after fatigue damage. According to the curves of area change, negative values are happened because some air-voids are compressed by repeated load, but the majority of air-voids are extended. In addition, the influence of test temperature on internal structures is significant,
94 93 92 91 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6
Air-void ratio/% Fig. 9. The relationship between compaction degree and air-voids ratio.
According to the data shown in Fig. 9, the compaction degree linear decrease with the increase of air-voids ratio. The increasing compaction degree means greater density, therefore, the air-voids are compressed because of compaction energy. It can be assumed that the effect of compaction on performance of asphalt mixture is mainly caused by the air-voids structure, and air-voids play a significant role in asphalt pavement. Based on the binary image of air-voids before and after fatigue damage, the area can be compared so the change states of air-voids ratio are shown in Fig. 10. It can be concluded from Fig. 10 that the air-voids ratio of different test specimens are increased after fatigue damage. The change rate decreases along with the depth of pavement at the same test temperature, because increasing the compaction degree of asphalt mixture can reduce the air-voids ratio, so the anti-fatigue performance is enhanced. Furthermore, the changes of air-voids ratio show that air-voids at high test temperature are enlarged or cause new cracks easily compare with the change at low test temperature. Therefore, rising temperature has an adverse influence on the fatigue performance of asphalt mixture during the low temperature range 10 °C to 10 °C. Meanwhile, it should be note that the initial air-voids ratio of test specimens drilled from 30 mm have the maximum value, it means that these specimens suffer damage easily compared with other one, especially at high temperature, because the high temperature can decrease the asphalt mixture’s strength. During the
Interface crack
Air-void extension
Fig. 11. Damage state of asphalt mixture.
942
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
300
350
Test temperature: -10°C
Test temperature: 0°C Internal structure area after fatigue damage
Internal structure area after fatigue damage Change of the area
Internal structure area/mm2
Internal structure area/mm2
300
Internal structure area before fatigue damage
250 200 150 100 50
Internal structure area before fatigue damage Change of the area
250 200 150 100 50
0
0
-50
-50 10
15
20
25
30
10
15
20
Distance/mm
Distance/mm
(a)
(b) 350
25
30
Test temperature: 10°C Internal structure area after fatigue damage
Internal structure area/mm2
300
Internal structure area before fatigue damage Change of the area
250 200 150 100 50 0 -50 10
15
20
25
30
Distance/mm
(c) Fig. 12. Change of internal structures area. (a) Test temperature: 10 °C. (b) Test temperature: 0 °C. (c) Test temperature: 10 °C.
it can be found that the air-voids regions are extended or cracks are caused are more great at high temperature. Cracks and air-voids are main structures in asphalt mixture, and the damage state can be evaluated by cracks that appeared after fatigue test, therefore, identifying the cracks can investigate the effect of air-voids on fatigue damage. Shape discrepancies are used to distinguish cracks and air-voids, Shape Index (SI) is defined to represent the morphology of different internal structure:
SI ¼
lmmd lmnd
ð1Þ
where, lmmd represents the maximum distance of cracks or air-voids in length direction, and lmnd represents the minimum distance of cracks or air-voids in width direction. Due to the pixel size is 80 lm, the cracks which width is greater than 80 lm are identified. Internal structures, which SI was more than 10 meanwhile the area was less than 5 mm2, were assumed as cracks in this research, therefore, SI and area of internal structures can be used to distinguish cracks and air-voids after fatigue damage. Three test specimens obtained from the depth of 90 mm were selected as research objectives, the statistical data about internal structures before and after fatigue test at different temperatures are illustrated in Fig. 13.
Fig. 13 shows that there are a few of cracks in asphalt mixture initially but cracks are increased obviously after fatigue test. With the enlargement of average SI and average air-voids area, indicating the asphalt mastic and interface between aggregates and asphalt mastic have been fractured due to the fatigue damage. It can be concluded from Fig. 13 that internal structures of asphalt mixture can be divided into three groups, namely large air-voids, small air-voids and cracks, most air-voids are small air-voids which SI less than 10 and area less than 5 mm2. In addition, the average SI and average area of air-voids are all increased after fatigue test, and the phenomenon indicates that fatigue damage affects the morphology and area of air-voids simultaneously. Air-void is regard as a defect in asphalt mixture because it has adverse influence on structural load-carrying capacity, so, the more air-voids in asphalt mixture, the serious damage it will be suffered under loading. The amount of air-voids in different depths of pavement is changed due to the fatigue damage, the relationship between amount of air-voids and cracks are shown in Fig. 14. As illustrated in Fig. 14, the number of cracks increase with the increase of initial air-voids, proving the inference that air-void is a defect for asphalt mixture and cause cracks distress easily. Meanwhile, temperature has significant influence on cracks, the test
943
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
60
Cracks
50
Before fatigue damage Average SI:2.61 Average air-void area:2.39mm2 After fatigue damage Average SI:2.74 Average air-void area:2.44mm2
Cracks 50
Shape Index
40
Shape Index
60
Before fatigue damage Average SI:2.37 Average air-void area:2.29mm2 After fatigue damage Average SI:2.44 Average air-void area:2.34mm2
30
20
40
30
20
Small Air-voids
Small Air-voids 10
10
Large Air-voids
Large Air-voids 0
0 -5
0
5
10
15
20
25
30
35
40
45
50
-5
0
5
10
15
Air-void area/mm2
20
25
30
35
40
45
50
2
Air-void area/mm
(a)
(b) 60
Before fatigue damage Average SI:2.31 Average air-void area:2.63mm2 After fatigue damage Average SI:2.62 Average air-void area:2.78mm2
Cracks
Shape Index
50
40
30
20
Small Air-voids 10
Large Air-voids 0 -5
0
5
10
15
20
25
30
35
40
45
50
2
Air-void area/mm
(c) Fig. 13. Number of internal structure before and after fatigue damage. (a) Test temperature: 10 °C. (b) Test temperature: 0 °C. (c) Test temperature: 10 °C.
40
Test temperature: -10 Fitting curve of test temperature: -10 Test temperature: 0 Fitting curve of test temperature: 0 Test temperature: 10 Fitting curve of test temperature: 10
36
Number of cracks
32 28
y=-2.548*10-4x2+0.199x-0.655 R2=0.979
24 20
y=-1.341*10-4x2+0.122x-0.741 R2=0.992
16 12
y=-6.658*10-5x2+0.078x-1.025 R2=0.987
8 4 0 -50
0
50
100
150
200
250
300
350
400
Number of initial air-voids Fig. 14. Number of cracks in asphalt mixture with different initial air-voids.
specimen measured at 10 °C appears more cracks compared with that at 0 °C and 10 °C due to the strength of asphalt mastic and bonding capability between aggregates and asphalt mastic are decreased at 10 °C. Besides, the increasing rate of cracks are
decreased gradually, the phenomenon can be explained that although a large number of air-voids reduce the cracking resistance of asphalt mixture, the volume of asphalt mastic and contact area between aggregate particles and asphalt mastic are decreased as well, this can decrease the probability of cracks initiation and propagation. 5.3. Influence of air-voids morphology on fatigue damage The morphology or shape of air-voids is very complicated, thus traditional geometrical method cannot describe the air-void’s complexity effectively. Fractal dimension provides a quantitative value to represent the complexity of irregular structure, the presented research show that plane morphology of aggregate particle has been evaluated by fractal dimension [28], and crack distress of cement concrete was investigated by fractal dimension as well [29,30]. In order to determine the influence of air-voids on fatigue damage, Damage Ratio (DR) is defined as below:
DR ¼
Sca þ Seaa Siaa
ð2Þ
where, Sca and Seaa are cracks area and extended air-voids area appeared after fatigue test, Siaa is the initial air-voids area.
944
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
0.6
0.5 Asphalt mixture at test temperature: -10°C Fitting Curve
Asphalt mixture at test temperature: 0°C Fitting Curve
0.5
0.4
DR/%
DR/%
0.4
0.3
0.3
0.2
0.2
y=0.977x+-0.841 R2=0.998
0.1
y=0.687x-0.556 R2=0.887
0.1
0.0
0.0 0.8
0.9
1.0
1.1
1.2
0.8
1.3
0.9
1.0
1.1
1.2
Fractal Dimension
Fractal Dimension
(a)
(b) 0.5 Asphalt mixture at test temperature: 10°C Fitting Curve
0.4
DR/%
0.3
0.2
y=0.579x-0.403 R2=0.989 0.1
0.0 0.8
0.9
1.0
1.1
1.2
1.3
1.4
Fractal Dimension
(c) Fig. 15. Fractal dimension of air-voids and DR. (a) Test temperature: 10 °C. (b) Test temperature: 0 °C. (c) Test temperature: 10 °C.
Four typical test specimens were investigated, for each test specimen, four hundred gray images were conducted to calculate the average fractal dimension and DR value, the relationships between fractal dimension and DR are shown in Fig. 15. Fig. 15 shows fractal dimension of air-voids versus DR, it can be find that the DR with respect to fractal dimension of air-voids, and the fitting line is linear at different temperatures. Therefore, the air-voids’ morphology or shape have significant influence on fatigue damage, namely air-voids with complicated shape will cause serious damage, the reason can be deduced that stress concentration is appeared easily due to the effect of complicated air-voids and more cracks are formed in asphalt mastic and interface between aggregate particle and asphalt mastic. 6. Conclusion This paper investigates the air-voids change of asphalt mixture before and after fatigue damage, XCT device was used to obtain the gray images of internal structure, the relationship between airvoids characteristics and fatigue damage was evaluated by DIP. Results of this study can be summarized as follows.
(2) The air-voids ratio of asphalt mixtures are increased after fatigue damage at different test temperatures, and research shows that rising temperature is harm to the fatigue performance during the low temperature range 10 °C to 10 °C. (3) Utilizing the shape index can identify the cracks and determine air-voids change before and after fatigue test, initial air-voids number and temperature have obvious influence on cracks appeared after fatigue damage. (4) The DR of asphalt mixture can be evaluated by the fractal dimension of air-voids, the relationship between them is linear, the result indicates that decreasing the complexity of air-voids can effectively reduce the fatigue damage of asphalt mixture. The above analysis reflects that increasing compaction degree of asphalt mixture and reducing the complexity of air-voids can improve the fatigue performance of asphalt mixture. Besides the construction technology, the aggregate particles also affect the morphology of air-voids, and the relevant research will be launched in next step. Acknowledgments
(1) The air-voids ratio decreases along with the depth of asphalt pavement, especially at the range from 90 mm to 150 mm, and the relationship between air-voids ratio and compaction degree is almost linear.
The work underlying this project was carried out under the Research grant number FOR 2089, on behalf of the grant sponsor, the German Research Foundation (DFG).
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
References [1] A.K. Apeagyei, J.R.A. Grenfell, G.D. Airey, Observation of reversible moisture damage in asphalt mixtures, Constr. Build. Mater. 60 (2014) 73–80. [2] D.L. Presti, R. Khan, N.A. Hassan, G. Airey, A. Collop, Laboratory mix design of asphalt mixture containing reclaimed material, Adv. Mater. Sci. Eng. 2014 (2014) 1–11. [3] M. Mohajeri, A.A.A. Molenaar, M.F.C. Van de Ven, Experimental study into the fundamental understanding of blending between reclaimed asphalt binder and virgin bitumen using nanoindentation and nano-computed tomography, Road Mater. Pavement Des. 15 (2) (2014) 372–384. [4] Z.A. Arega, A. Bhasin, T.D. Kesel, Influence of extended aging on the properties of asphalt composites produced using hot and warm mix methods, Constr. Build. Mater. 44 (2013) 168–174. [5] D. Braz, R.C. Barroso, R.T. Lopes, L.M.G. Motta, Crack detection in asphaltic mixtures by computed tomography, NDT&E Int. 44 (2011) 195–201. [6] A. Moriyoshi, N. Takahashi, O. Ikeda, M. Kawashima, T. Akabane, Strain distribution in asphalt mixtures during the wheel tracking test at high temperatures, Constr. Build. Mater. 2013 (40) (2013) 1128–1135. [7] D. Castillo, S. Caro, M. Darabi, E. Masad, Studying the effect of microstructural properties on the mechanical degradation of asphalt mixtures, Constr. Build. Mater. 93 (2015) 70–83. [8] M.R.M. Aliha, H. Fazaeli, S. Aghajani, F.M. Nejad, Effect of temperature and air void on mixed mode fracture toughness of modified asphalt mixtures, Constr. Build. Mater. 95 (2015) 545–555. [9] K. Yan, D.D. Ge, L.Y. You, X.L. Wang, Laboratory investigation of the characteristics of SMA mixtures under freeze–thaw cycles, Cold Reg. Sci. Technol. 119 (2015) 68–74. [10] H. Khosravi, S.M. Abtahi, B. Koosha, M. Manian, An analytical–empirical investigation of the bleeding mechanism of asphalt mixes, Constr. Build. Mater. 45 (2013) 138–144. [11] S. Kodippily, T.F.P. Henning, J.M. Ingham, G. Holleran, Computed tomography scanning for quantifying chipseal material volumetrics, J. Comput. Civ. Eng. 28 (3) (2014) 1–10. [12] E. Coleri, J.T. Harvey, K. Yang, J.M. Boone, Micromechanical investigation of open-graded asphalt friction courses’ rutting mechanisms, Constr. Build. Mater. 44 (2013) 25–34. [13] E. Coleri, J.T. Harvey, K. Yang, J.M. Boone, A micromechanical approach to investigate asphalt concrete rutting mechanisms, Constr. Build. Mater. 30 (2012) 36–49. [14] N.R. Sefidmazgi, L. Tashman, H. Bahia, Internal structure characterization of asphalt mixtures for rutting performance using imaging analysis, Road Mater. Pavement Des. 13 (S1) (2012) 21–37.
945
[15] R. Khan, J. Grenfell, A. Collop, Micromechanical modelling of typical Dense Bitumen Macadam (DBM) from laboratory testing and X-ray Computed Tomography (CT), Constr. Build. Mater. 77 (2015) 1–6. [16] R. Khan, J. Grenfell, A. Collop, G. Airey, H. Gregory, Moisture damage in asphalt mixtures using the modified SATS test and image analysis, Constr. Build. Mater. 2013 (43) (2013) 165–173. [17] Y.Q. Zhang, X. Luo, R. Luo, R.L. Lytton, Crack initiation in asphalt mixtures under external compressive loads, Constr. Build. Mater. 72 (2014) 94–103. [18] L. Tashman, B. Pearson, Characterisation of stone matrix asphalt mixtures, Int. J. Pavement Eng. 13 (4) (2012) 297–309. [19] D.L. Presti, N.A. Hassan, R. Khan, G. Airey, Reclaimed asphalt test specimen preparation assisted by image analysis, J. Mater. Civ. Eng. 27 (8) (2014) 1–5. C4014005. [20] A.A. Rose, E. Arambula, G. Liu, C.J. Glover, Use of a total air void diffusion depth to improve pavement oxidation modeling with field validation, Pet. Sci. Technol. 33 (2015) 1198–1207. [21] N.A. Hassan, G.D. Airey, N.I.M. Yusoff, M.R. Hainin, R. Putrajaya, M.E. Abdullah, M.M.A. Aziz, Microstructural characterisation of dry mixed rubberised asphalt mixtures, Constr. Build. Mater. 82 (2015) 173–183. [22] H.M. Zelelew, A.T. Papagiannakis, A volumetrics thresholding algorithm for processing asphalt concrete X-ray CT images, Int. J. Pavement Eng. 12 (6) (2011) 543–551. [23] M.E. Kutay, E. Arambula, N. Gibson, J. Youtcheff, Three-dimensional image processing methods to identify and characterise aggregates in compacted asphalt mixtures, Int. J. Pavement Eng. 11 (6) (2010) 511–528. [24] X.Y. Xu, S.Z. Xu, L.H. Jin, E. Song, Characteristic analysis of Otsu threshold and its applications, Pattern Recogn. Lett. 32 (2011) 956–961. [25] A. Mizushima, R. Lu, An image segmentation method for apple sorting and grading using support vector machine and Otsu’s method, Comput. Electr. Agric. 94 (2013) 29–37. [26] J. Hu, Z.D. Qian, D.W. Wang, M. Oeser, Influence of aggregate particles on mastic and air-voids in asphalt concrete, Constr. Build. Mater. 93 (2015) 1–9. [27] N.A. Hassan, G.D. Airey, M.R. Hainin, Characterisation of micro-structural damage in asphalt mixtures using image analysis, Constr. Build. Mater. 54 (2014) 27–38. [28] S. Arasan, E. Yener, F. Hattatoglu, S. Akbulut, S. Hinislioglu, The relationship between the fractal dimension and mechanical properties of asphalt concrete, Int. J. Civ. Struct. Eng. 1 (2) (2010) 165–170. [29] J.U. Baer, T.F. Kent, S.H. Anderson, Image analysis and fractal geometry to characterize soil desiccation cracks, Geoderma 154 (2009) 153–163. [30] A. Farhidzadeh, E. Dehghan-Niri, A. Moustafa, S. Salamone, A. Whittaker, Damage assessment of reinforced concrete structures using fractal analysis of residual crack patterns, Exp. Mech. 53 (2013) 1607–1619.
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Investigation on fatigue damage of asphalt mixture with different air-voids using microstructural analysis Jing Hu a, Pengfei Liu a, Dawei Wang a,b,⇑, Markus Oeser a,b, Yiqiu Tan b,c a
Institute of Highway Engineering, RWTH Aachen University, Aachen, Germany Germany Sino-European Research Center for Advanced Transportation Infrastructure Technology, Würselen, Germany c School of Transportation Science and Engineering, Harbin Institute of Technology, Harbin, China b
h i g h l i g h t s The relationship between air-voids ratio and compactness is linear. Initial air-voids number and temperature have influence on cracks distress. The effect of air-voids complexity on fatigue damage is significant.
a r t i c l e
i n f o
Article history: Received 11 January 2016 Received in revised form 17 July 2016 Accepted 28 August 2016
Keywords: Asphalt mixture Fatigue damage Digital Image Processing Microstructure Air-voids
a b s t r a c t Fatigue damage caused by vehicle loads is the main asphalt pavement distress, and it deteriorates the serviceability and strength seriously. In order to evaluate the fatigue damage of asphalt mixture under repeated load, microstructures were detected to investigate the morphology change of internal structures using Digital Image Processing (DIP). Field pavement named as ISAC test track was built, and cores were drilled to slice into test specimens with different air-voids that can reflect the real internal state of asphalt pavement. Fatigue properties were measured under temperature 10 °C, 0 °C and 10 °C respectively, the frequency of sinusoid load was 0.1 Hz, the minimum value was 0.035 MPa and maximum value was 0.5 MPa. Internal structures of asphalt mixture were scanned by X-ray Computed Tomography (XCT) device before and after fatigue damage, thus the relationship between microstructures and fatigue damage can be conducted. The results show that microstructural analysis can effectively determine the internal structure change of asphalt mixture. Charging compaction causes different air-voids distributions and morphologies, which have obvious influence on failure state of asphalt mixture. The effect of temperature and initial air-voids on fatigue performance is significant, and fatigue damage presents a linear relation with the complexity of air-voids. Methodology established in this paper provides an effective method for fatigue damage assessment. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction Asphalt mixture is a composite material that includes aggregates, air-voids and asphalt mastic, each microstructure has obvious influence on the performance. Air-voids are main structure component of asphalt pavement, and some functions such as drainage and noise reduction relate with air-voids significantly. However, air-voids can aggravate the fatigue damage under repeated load and reduce the strength of asphalt mixture, causing macro crack and some typical failures appear in asphalt pavement, such ⇑ Corresponding author at: Institute of Highway Engineering, RWTH Aachen University, Aachen, Germany. E-mail address: [email protected] (D. Wang). http://dx.doi.org/10.1016/j.conbuildmat.2016.08.138 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
as rutting and bleeding. Therefore, air-void can be regarded as a defect of asphalt mixture. Although the gradation of asphalt mixture is a critical factor for air-voids morphology, the compaction affects the spatial distribution of air-voids in asphalt pavement as well. Due to the different compaction energies the asphalt pavement suffers, the air-voids must be changed along with the depth, result in different states in different parts of pavement. Thus the effect of air-voids on fatigue performance should be investigated to evaluate the properties of asphalt pavement. In recently years, microstructure has become a prevalent research objective in pavement engineering, Digital Image Processing (DIP) base on the X-ray Computed Tomography (XCT) is generally used to analyze the failure mechanism and influencing factors of asphalt mixture. The development of pavement distress can be
937
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
2. Objective The investigation was aimed at providing an effective method for a better understanding of the change of internal structure after fatigue damage and the influence of microstructure characteristics on fatigue performance. A comprehensive research was conducted in order to achieve the following objectives: (1) Determining the relationship between pavement compaction and air-voids, and reconstructing the real distribution of air-voids in the asphalt mixture by DIP technology. (2) Counting the damage state of asphalt mixture, such as crack quantity, crack sharp and crack area, before and after fatigue test, and evaluating the effect of different temperatures (10 °C, 0 °C and 10 °C) and air-voids complexity on the fatigue performance of asphalt mixture. 3. Materials and methods
aggregates. The SMA-11S mixture was prepared using a PG 50/70 binder and diabase aggregates, Marshell method was used to determine the gradation of SMA-11S, the test temperature was 170 °C, the gradation of the diabase aggregates are shown in Fig. 1 and critical material parameters are listed in Table 1. 3.2. Test track construction, fatigue test and XCT scanning In order to obtain the internal structure of real asphalt pavement, the field test track was built at Institute of Highway Engineering of RWTH Aachen University and named as ISAC test track. Diabase aggregates were mixed at temperature of 170 °C. A brief overview about the procedure of the test track construction, fatigue test and XCT scanning follows. (1) A miniature paver was used to prepare the asphalt mixture, a heating part within the machine can keep the paving temperature higher than 150 °C. The length, width and thickness of the test track were 26 m, 1.2 m and 0.3 m respectively, as shown in Fig. 2(a). (2) Dividing ISAC test track into three layers, each layer has the same depth of 0.1 m. Using a miniature roller compacted these layers respectively, as shown in Fig. 2(b). It must be note that the test track was composed of three layers, thickness of each layer was 0.1 m after compression, the lower layer of test track suffers more compaction energy, so compaction degree was greater than that of other layers. Test track determined in this paper can provide different asphalt mixtures which have different porosity rate, these asphalt mixture samples are the foundation for investigating the effect of compaction (air-voids) on the fatigue property. (3) In order to obtain the internal structure of asphalt pavement, cylindrical cores were drilled at different sections of test track after maintenance, as shown in Fig. 2(c). Fig. 2(d) illustrates the cylindrical cores, which size were 150 mm in diameter and 300 mm in height. (4) Each cylindrical core can be sliced into five test specimens which size were 40 mm in height and 100 mm in diameter according to the demand of FGSV-Nr.430 of Germany, and test specimens were numbered base on their corresponding initial depth in the test track, as shown in Fig. 2(e). (5) The Universal Test Machine (UTM) was used to measure the fatigue performance of test specimens at test temperature of 10 °C, 0 °C and 10 °C in Fig. 2(f). The loading followed sinusoid mode that frequency was 0.1 Hz and stress control
100 90 80
Passing percentage/%
detected using the microstructure reconstructed by XCT images [1–6], and the presented literature shows that air-voids have obvious influence on the performance and durability of asphalt mixture [7]. Some researchers have certified the significant effect of airvoids on the failure of asphalt mixture, especially the mechanisms that air-voids affect the mechanical behavior of asphalt pavement are commonly involved base on the XCT scanning technology [8–11]. Coleri used XCT apparatus to scan the internal structure of asphalt pavement before and after high temperature deformation, then the air-voids change were measured and the effect of air-voids on rutting distress was evaluated [12,13], Sefidmazgi also analyzed the influence of microstructure on anti-deformation performance of asphalt mixture using XCT images and DIP [14]. Khan developed a 3D microstructure numerical model according to XCT data, the result shows that air-voids affect stress concentration and aggravate the pavement failure [15], besides, moisture damage of asphalt mixture was investigated as well [16]. In fact, air-voids are extended to form cracks distress commonly because of stress concentration, and the conclusion of presented research shows that air-voids have adverse influence on performance of asphalt mixture [17]. Furthermore, the air-voids distribution is uneven because of the different compaction energies during pavement construction. The components of asphalt mixture, especially the content of fine aggregates, determine the air-voids distribution [18]. Some microstructural characteristics of air-voids, such as the homogeneous of distribution state and discrete degree of morphology, can be analyzed by the XCT images and DIP [19–21]. However, most researchers study the asphalt mixture molded in laboratory, it is hard to reflect the real state of internal structure, and the investigations about the relationship between air-voids and fatigue damage are rare. Through the XCT scanning, this paper establishes the relationship between compaction and air-voids of asphalt mixture drilled from pavement layers in different depths. Air-voids were extracted, and DIP was used to determine morphology and distribution. Meanwhile, the effect of air-voids on fatigue performance and change of internal microstructure after fatigue damage were detected as well.
Asphalt mixture design Contral design Lower limit Upper limit
70 60 50 40 30 20
3.1. Laboratory design for material Initially, a Stone Mastic Asphalt (SMA) mixture with a nominal maximum aggregate size (NMAS) of 12.5 mm and bitumen of 6.9% was prepared and named as SMA-11S. The NMAS is defined as one sieve size larger than the first sieve to retain more than 10% of the
10 0 0.063
2
5.6 8 11.2
Sieve size/mm Fig. 1. Gradation of diabase aggregates of SMA-11S.
938
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
Table 1 Critical material parameters of SMA-11S.
*
Asphalt mixture
Virgin materials
Bitumen content (%)
Max. density (g/cm3)
Ave. density (g/cm3)
SMA-11S
Bitumen binder: Pen Bitumen 50/70 Fibers: Cellulose Filler (0/0.09 mm): Limestone Sand (0/2 mm): Diabase Coarse aggregates: Diabase
6.9
2.488
2.415
*
Void ratio (%) 2.090
Average value of 44 specimens.
Fig. 2. The illustration of test track construction, fatigue test and XCT scanning. (a) Paving ISAC test track. (b) Compacting ISAC test track. (c) Drilling cylindrical cores. (d) Cylindrical cores. (e) Test specimens used for fatigue test. (f) Device of Universal Test Machine. (g) Loading apparatus of fatigue test. (h) XCT scanning before and after fatigue test. (i) Sequence gray images (six gray images were selected as illustration).
mode was utilized in test, the range of loading was 0.035 MPa to 0.5 MPa to ensure the peak value of horizontal strain of test specimen from 0.05‰ to 0.3‰. Furthermore, test time was set as 2 h to assure that each test specimen bore the same energy, thus the impact of different internal structures on fatigue damage can be evaluated. The loading device is shown in Fig. 2(g). (6) Y.CT Precision S mode X-CT device was used to scan the test specimens and the internal structures can be obtained, as shown in Fig. 2(h). Scanning interval of 0.1 mm was set to assure the precision of gray image, the resolution of each
gray image was 1024 1024 Pixel2, and the size of pixel was 80 lm. The gray images were presented in Fig. 2(i). 3.3. Fracture mechanism of fatigue damage The test specimens were obtained at the different depths of asphalt pavement; air-voids ratio and distribution were changed due to the different compaction energies. In this paper, three specimens obtained from the same pavement depth were tested at the same temperature. At low temperature, assuming the fatigue damage is mainly caused by cracks distress, namely cracks deteriorate
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
to the results of fatigue test (an example illustrated in Fig. 3), assuming the asphalt mixtures are within micro-damage stage after fatigue test except the complete damage one, so the development of micro damage in asphalt mixture with different air-void characteristics can be evaluated.
15150
Modulus/MPa
15100 Test specimen #1 (30mm) Test specimen #2 (90mm) Test specimen #3 (150mm) Test specimen #4 (210mm) Test specimen #5 (270mm)
15050
15000 Fatigue Damage Point Cycle number: 28530 Modules: 15087MPa
14950 0
10000
20000
30000
40000
50000
60000
70000
Cycle number Fig. 3. Modulus state of asphalt mixture at test temperature of 10 °C.
n*modulus
Beginning Macro-cracks
0
Beginning Micro-cracks
Completed Fracture
0
939
Loading cycles:n
N
Fig. 4. Damage accumulation stage of asphalt mixture.
4. Digital Image Processing The internal structures of asphalt mixture are detected using XCT gray images, gray value is from 0 to 255. The gray value is determined by XCT device according to material density. As for asphalt mixture, the gray of aggregates have the maximum value and gray of air-voids have the minimum value. The gray images of asphalt mixture in different depths of pavement are shown in Fig. 5. In order to analyze the structure change of asphalt mixture, microstructure must be extracted by DIP at first. The main functions of DIP technology used in gray images are extraction and analysis, and the critical step of the whole research system is identifying the air-voids accurately. The Artificial Neural Network (ANN) has been developed to extract microstructure [22,23], besides, gray images of asphalt mixture can be converted into binary images to represent the different microstructures, and a technique named as Otsu has been proved to be an effective binary method for gray images [24,25]. However, the Otsu cannot solve the problem that uneven gray distribution of XCT image scanned from asphalt mixture. In this research, an improved Otsu method was developed to extract the binary image of air-voids [26], and then, the microstructure characteristics were analyzed by Matlab software, as shown in Fig. 6. Using the binary images of air-voids, 3D morphology models can be created, the distributions of air-voids are presented in Fig. 7. Fig. 7 shows the spatial air-voids state in different depths of pavement, air-voids of depth 30 mm and 90 mm are great, and decreased rapidly at the range deeper than 90 mm. In addition, individual air-voids shown in Fig. 7(a)–(c) have a large volume, and exhibit more complex shape than the air-voids shown in Fig. 7(e) and (f). 5. Discussing and analysis
the strength and durability of asphalt mixture. Modulus was used as an index to indicate the property of test specimen, and results of five typical specimens drilled from different pavement depth were selected to show the test situation, as shown in Fig. 3. Fig. 3 presents the modulus curves of five test specimens, which located at different depths of pavement. The test specimen drilled from depth of 30 mm was fractured completely during the fatigue test because of the large air-voids ratio, and the modulus is also less compared with other test specimens. The results indicate that the influence of air-voids on fatigue properties is significant, and fatigue can be regarded as a damage accumulation stage for asphalt mixture. Micro cracks are appeared initially, and they gradually develop into macro cracks, then, enough macro cracks will cause complete damage of asphalt mixture. Fig. 4 presents the damage accumulation stage of asphalt mixture introduced from regulation of FGSV-Nr.430, it reflects the development of internal structures with the increasing load, and three critical points can be determined through curve slope to divide the curve into three damage stage. The linear section is non-damage stage, which means there is not damage appeared in asphalt mixture. The first curve section is micro-damage stage, micro cracks appear after Beginning Micro-cracks point, and then, micro cracks are extended to form macro cracks after Beginning Macro-cracks point and asphalt mixture is fractured completely, this curve section represents the macro-damage stage. According
5.1. The feature of air-voids distribution Binary image is composed of pixels which size is 80 lm, therefore, the air-voids area at different depths of pavement can be calculated through the amount of pixel belonged to air-voids, and range of air-voids ratio are inferred, as shown in Fig. 8. Fig. 8 indicates the range of air-voids ratio along with the pavement depth. Due to different compaction energies, air-voids ratio decreases along with the depth of asphalt pavement, air-voids ratios are decreased significantly during the range 0 mm– 150 mm, especially in the range from 90 mm to 150 mm, all compaction degree of asphalt mixtures are increased obviously. The curves show that discrepancies of air-voids ratio nearby the pavement surface are great, such as the air-voids ratio at depth of 30 mm and 90 mm. However, air-voids ratios appear similar value at depth of 150 mm, this means that asphalt mixture below depth of 150 mm bears the similar compaction energies. Compaction can be calculated by density and theoretical maximum density, it is a quantitative indicator that represents the compaction energy. The relationship between compaction and airvoids should be determined, and then, microstructural index can be established to evaluate the compaction state. Fig. 9 presents the compaction degree of fifteen test specimens with respect to corresponding air-voids ratio.
940
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
Fig. 5. XCT gray images in different depths of test track. (a) 30 mm. (b) 90 mm. (c) 150 mm. (d) 210 mm. (e) 270 mm.
Fig. 6. Extracting air-voids from XCT gray image. (a) Origin gray image. (b) Binary image of aggregates. (c) Binary image of air-voids.
Fig. 7. 3D air-voids distributions in different depths. (a) 30 mm. (b) 90 mm. (c) 150 mm. (d) 210 mm. (e) 270 mm.
941
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
0.5
1.0
1.5
2.0
2.5
0
2.5
Air-voids ratio/%
Surface 50
Depth/mm
90mm
100 150 200
Temperature -10°C:Change of air-voids ratio Temperature 0°C:Change of air-voids ratio Temperature 10°C:Change of air-voids ratio Temperature -10°C:Before fatigue damge Temperature -10°C:After fatigue damge Temperature 0°C:Before fatigue damge Temperature 0°C:After fatigue damge Temperature 10°C:Before fatigue damge Temperature 10°C:After fatigue damge
3.0
150mm
Air-voids ratio - Lower limit Air-voids ratio - Average Air-voids ratio - Upper limit
2.0
Bottom Fig. 8. Air-voids ratio states of ISAC Test Track.
100
Compaction degree/%
99
Relationship between compaction degree and air-void ratio Fitting curve
98
0.3
0.2 1.0 0.1
0.5
0.0 30
300
0.4
1.5
0.0
250
0.5
Change of air-voids ratio/%
3.0
Air-voids ratio/% 0.0
90
150
210
270
Depth/mm Fig. 10. Air-voids ratio state of ISAC test track.
fatigue test, three test specimens (30 mm) were measured at 10 °C and other three (30 mm) were measured at 0 °C, the test results show that all of six test samples are totally damage during the fatigue test. Both great porosity and high temperature cause adverse effects on the fatigue performance of asphalt mixture.
97 96
5.2. The microstructural characteristics of air-voids
95
The state of internal structures of asphalt mixture can be evaluated by DIP. Initial air-voids are changed after damage, meanwhile, cracks appear in asphalt mastic and the interface between aggregate particles and asphalt mastic as well [27]. In this research, assuming the fatigue damage of asphalt mixture is caused by cracks, so, determining the extended air-void state and new cracks in each cross section of asphalt mixture can provide information for damage evaluation, as shown in Fig. 11. In order to determine the cross section pair that located at the same position of test specimen before and after fatigue damage, comparing the areas of aggregate particles and air-voids to select the appropriate cross sections. Two hundred cross section pairs in the range of 20 mm of test specimen center along with the thickness direction were selected as research objectives, and the change of internal structures (mainly includes air-voids and cracks) before and after fatigue damage was measured through comparing the gray images, as shown in Fig. 12. Fig. 12 presents the change of internal structures at different temperatures, the curves of internal structures in the same test specimen presents the similar trend before and after fatigue damage. According to the curves of area change, negative values are happened because some air-voids are compressed by repeated load, but the majority of air-voids are extended. In addition, the influence of test temperature on internal structures is significant,
94 93 92 91 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6
Air-void ratio/% Fig. 9. The relationship between compaction degree and air-voids ratio.
According to the data shown in Fig. 9, the compaction degree linear decrease with the increase of air-voids ratio. The increasing compaction degree means greater density, therefore, the air-voids are compressed because of compaction energy. It can be assumed that the effect of compaction on performance of asphalt mixture is mainly caused by the air-voids structure, and air-voids play a significant role in asphalt pavement. Based on the binary image of air-voids before and after fatigue damage, the area can be compared so the change states of air-voids ratio are shown in Fig. 10. It can be concluded from Fig. 10 that the air-voids ratio of different test specimens are increased after fatigue damage. The change rate decreases along with the depth of pavement at the same test temperature, because increasing the compaction degree of asphalt mixture can reduce the air-voids ratio, so the anti-fatigue performance is enhanced. Furthermore, the changes of air-voids ratio show that air-voids at high test temperature are enlarged or cause new cracks easily compare with the change at low test temperature. Therefore, rising temperature has an adverse influence on the fatigue performance of asphalt mixture during the low temperature range 10 °C to 10 °C. Meanwhile, it should be note that the initial air-voids ratio of test specimens drilled from 30 mm have the maximum value, it means that these specimens suffer damage easily compared with other one, especially at high temperature, because the high temperature can decrease the asphalt mixture’s strength. During the
Interface crack
Air-void extension
Fig. 11. Damage state of asphalt mixture.
942
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
300
350
Test temperature: -10°C
Test temperature: 0°C Internal structure area after fatigue damage
Internal structure area after fatigue damage Change of the area
Internal structure area/mm2
Internal structure area/mm2
300
Internal structure area before fatigue damage
250 200 150 100 50
Internal structure area before fatigue damage Change of the area
250 200 150 100 50
0
0
-50
-50 10
15
20
25
30
10
15
20
Distance/mm
Distance/mm
(a)
(b) 350
25
30
Test temperature: 10°C Internal structure area after fatigue damage
Internal structure area/mm2
300
Internal structure area before fatigue damage Change of the area
250 200 150 100 50 0 -50 10
15
20
25
30
Distance/mm
(c) Fig. 12. Change of internal structures area. (a) Test temperature: 10 °C. (b) Test temperature: 0 °C. (c) Test temperature: 10 °C.
it can be found that the air-voids regions are extended or cracks are caused are more great at high temperature. Cracks and air-voids are main structures in asphalt mixture, and the damage state can be evaluated by cracks that appeared after fatigue test, therefore, identifying the cracks can investigate the effect of air-voids on fatigue damage. Shape discrepancies are used to distinguish cracks and air-voids, Shape Index (SI) is defined to represent the morphology of different internal structure:
SI ¼
lmmd lmnd
ð1Þ
where, lmmd represents the maximum distance of cracks or air-voids in length direction, and lmnd represents the minimum distance of cracks or air-voids in width direction. Due to the pixel size is 80 lm, the cracks which width is greater than 80 lm are identified. Internal structures, which SI was more than 10 meanwhile the area was less than 5 mm2, were assumed as cracks in this research, therefore, SI and area of internal structures can be used to distinguish cracks and air-voids after fatigue damage. Three test specimens obtained from the depth of 90 mm were selected as research objectives, the statistical data about internal structures before and after fatigue test at different temperatures are illustrated in Fig. 13.
Fig. 13 shows that there are a few of cracks in asphalt mixture initially but cracks are increased obviously after fatigue test. With the enlargement of average SI and average air-voids area, indicating the asphalt mastic and interface between aggregates and asphalt mastic have been fractured due to the fatigue damage. It can be concluded from Fig. 13 that internal structures of asphalt mixture can be divided into three groups, namely large air-voids, small air-voids and cracks, most air-voids are small air-voids which SI less than 10 and area less than 5 mm2. In addition, the average SI and average area of air-voids are all increased after fatigue test, and the phenomenon indicates that fatigue damage affects the morphology and area of air-voids simultaneously. Air-void is regard as a defect in asphalt mixture because it has adverse influence on structural load-carrying capacity, so, the more air-voids in asphalt mixture, the serious damage it will be suffered under loading. The amount of air-voids in different depths of pavement is changed due to the fatigue damage, the relationship between amount of air-voids and cracks are shown in Fig. 14. As illustrated in Fig. 14, the number of cracks increase with the increase of initial air-voids, proving the inference that air-void is a defect for asphalt mixture and cause cracks distress easily. Meanwhile, temperature has significant influence on cracks, the test
943
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
60
Cracks
50
Before fatigue damage Average SI:2.61 Average air-void area:2.39mm2 After fatigue damage Average SI:2.74 Average air-void area:2.44mm2
Cracks 50
Shape Index
40
Shape Index
60
Before fatigue damage Average SI:2.37 Average air-void area:2.29mm2 After fatigue damage Average SI:2.44 Average air-void area:2.34mm2
30
20
40
30
20
Small Air-voids
Small Air-voids 10
10
Large Air-voids
Large Air-voids 0
0 -5
0
5
10
15
20
25
30
35
40
45
50
-5
0
5
10
15
Air-void area/mm2
20
25
30
35
40
45
50
2
Air-void area/mm
(a)
(b) 60
Before fatigue damage Average SI:2.31 Average air-void area:2.63mm2 After fatigue damage Average SI:2.62 Average air-void area:2.78mm2
Cracks
Shape Index
50
40
30
20
Small Air-voids 10
Large Air-voids 0 -5
0
5
10
15
20
25
30
35
40
45
50
2
Air-void area/mm
(c) Fig. 13. Number of internal structure before and after fatigue damage. (a) Test temperature: 10 °C. (b) Test temperature: 0 °C. (c) Test temperature: 10 °C.
40
Test temperature: -10 Fitting curve of test temperature: -10 Test temperature: 0 Fitting curve of test temperature: 0 Test temperature: 10 Fitting curve of test temperature: 10
36
Number of cracks
32 28
y=-2.548*10-4x2+0.199x-0.655 R2=0.979
24 20
y=-1.341*10-4x2+0.122x-0.741 R2=0.992
16 12
y=-6.658*10-5x2+0.078x-1.025 R2=0.987
8 4 0 -50
0
50
100
150
200
250
300
350
400
Number of initial air-voids Fig. 14. Number of cracks in asphalt mixture with different initial air-voids.
specimen measured at 10 °C appears more cracks compared with that at 0 °C and 10 °C due to the strength of asphalt mastic and bonding capability between aggregates and asphalt mastic are decreased at 10 °C. Besides, the increasing rate of cracks are
decreased gradually, the phenomenon can be explained that although a large number of air-voids reduce the cracking resistance of asphalt mixture, the volume of asphalt mastic and contact area between aggregate particles and asphalt mastic are decreased as well, this can decrease the probability of cracks initiation and propagation. 5.3. Influence of air-voids morphology on fatigue damage The morphology or shape of air-voids is very complicated, thus traditional geometrical method cannot describe the air-void’s complexity effectively. Fractal dimension provides a quantitative value to represent the complexity of irregular structure, the presented research show that plane morphology of aggregate particle has been evaluated by fractal dimension [28], and crack distress of cement concrete was investigated by fractal dimension as well [29,30]. In order to determine the influence of air-voids on fatigue damage, Damage Ratio (DR) is defined as below:
DR ¼
Sca þ Seaa Siaa
ð2Þ
where, Sca and Seaa are cracks area and extended air-voids area appeared after fatigue test, Siaa is the initial air-voids area.
944
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
0.6
0.5 Asphalt mixture at test temperature: -10°C Fitting Curve
Asphalt mixture at test temperature: 0°C Fitting Curve
0.5
0.4
DR/%
DR/%
0.4
0.3
0.3
0.2
0.2
y=0.977x+-0.841 R2=0.998
0.1
y=0.687x-0.556 R2=0.887
0.1
0.0
0.0 0.8
0.9
1.0
1.1
1.2
0.8
1.3
0.9
1.0
1.1
1.2
Fractal Dimension
Fractal Dimension
(a)
(b) 0.5 Asphalt mixture at test temperature: 10°C Fitting Curve
0.4
DR/%
0.3
0.2
y=0.579x-0.403 R2=0.989 0.1
0.0 0.8
0.9
1.0
1.1
1.2
1.3
1.4
Fractal Dimension
(c) Fig. 15. Fractal dimension of air-voids and DR. (a) Test temperature: 10 °C. (b) Test temperature: 0 °C. (c) Test temperature: 10 °C.
Four typical test specimens were investigated, for each test specimen, four hundred gray images were conducted to calculate the average fractal dimension and DR value, the relationships between fractal dimension and DR are shown in Fig. 15. Fig. 15 shows fractal dimension of air-voids versus DR, it can be find that the DR with respect to fractal dimension of air-voids, and the fitting line is linear at different temperatures. Therefore, the air-voids’ morphology or shape have significant influence on fatigue damage, namely air-voids with complicated shape will cause serious damage, the reason can be deduced that stress concentration is appeared easily due to the effect of complicated air-voids and more cracks are formed in asphalt mastic and interface between aggregate particle and asphalt mastic. 6. Conclusion This paper investigates the air-voids change of asphalt mixture before and after fatigue damage, XCT device was used to obtain the gray images of internal structure, the relationship between airvoids characteristics and fatigue damage was evaluated by DIP. Results of this study can be summarized as follows.
(2) The air-voids ratio of asphalt mixtures are increased after fatigue damage at different test temperatures, and research shows that rising temperature is harm to the fatigue performance during the low temperature range 10 °C to 10 °C. (3) Utilizing the shape index can identify the cracks and determine air-voids change before and after fatigue test, initial air-voids number and temperature have obvious influence on cracks appeared after fatigue damage. (4) The DR of asphalt mixture can be evaluated by the fractal dimension of air-voids, the relationship between them is linear, the result indicates that decreasing the complexity of air-voids can effectively reduce the fatigue damage of asphalt mixture. The above analysis reflects that increasing compaction degree of asphalt mixture and reducing the complexity of air-voids can improve the fatigue performance of asphalt mixture. Besides the construction technology, the aggregate particles also affect the morphology of air-voids, and the relevant research will be launched in next step. Acknowledgments
(1) The air-voids ratio decreases along with the depth of asphalt pavement, especially at the range from 90 mm to 150 mm, and the relationship between air-voids ratio and compaction degree is almost linear.
The work underlying this project was carried out under the Research grant number FOR 2089, on behalf of the grant sponsor, the German Research Foundation (DFG).
J. Hu et al. / Construction and Building Materials 125 (2016) 936–945
References [1] A.K. Apeagyei, J.R.A. Grenfell, G.D. Airey, Observation of reversible moisture damage in asphalt mixtures, Constr. Build. Mater. 60 (2014) 73–80. [2] D.L. Presti, R. Khan, N.A. Hassan, G. Airey, A. Collop, Laboratory mix design of asphalt mixture containing reclaimed material, Adv. Mater. Sci. Eng. 2014 (2014) 1–11. [3] M. Mohajeri, A.A.A. Molenaar, M.F.C. Van de Ven, Experimental study into the fundamental understanding of blending between reclaimed asphalt binder and virgin bitumen using nanoindentation and nano-computed tomography, Road Mater. Pavement Des. 15 (2) (2014) 372–384. [4] Z.A. Arega, A. Bhasin, T.D. Kesel, Influence of extended aging on the properties of asphalt composites produced using hot and warm mix methods, Constr. Build. Mater. 44 (2013) 168–174. [5] D. Braz, R.C. Barroso, R.T. Lopes, L.M.G. Motta, Crack detection in asphaltic mixtures by computed tomography, NDT&E Int. 44 (2011) 195–201. [6] A. Moriyoshi, N. Takahashi, O. Ikeda, M. Kawashima, T. Akabane, Strain distribution in asphalt mixtures during the wheel tracking test at high temperatures, Constr. Build. Mater. 2013 (40) (2013) 1128–1135. [7] D. Castillo, S. Caro, M. Darabi, E. Masad, Studying the effect of microstructural properties on the mechanical degradation of asphalt mixtures, Constr. Build. Mater. 93 (2015) 70–83. [8] M.R.M. Aliha, H. Fazaeli, S. Aghajani, F.M. Nejad, Effect of temperature and air void on mixed mode fracture toughness of modified asphalt mixtures, Constr. Build. Mater. 95 (2015) 545–555. [9] K. Yan, D.D. Ge, L.Y. You, X.L. Wang, Laboratory investigation of the characteristics of SMA mixtures under freeze–thaw cycles, Cold Reg. Sci. Technol. 119 (2015) 68–74. [10] H. Khosravi, S.M. Abtahi, B. Koosha, M. Manian, An analytical–empirical investigation of the bleeding mechanism of asphalt mixes, Constr. Build. Mater. 45 (2013) 138–144. [11] S. Kodippily, T.F.P. Henning, J.M. Ingham, G. Holleran, Computed tomography scanning for quantifying chipseal material volumetrics, J. Comput. Civ. Eng. 28 (3) (2014) 1–10. [12] E. Coleri, J.T. Harvey, K. Yang, J.M. Boone, Micromechanical investigation of open-graded asphalt friction courses’ rutting mechanisms, Constr. Build. Mater. 44 (2013) 25–34. [13] E. Coleri, J.T. Harvey, K. Yang, J.M. Boone, A micromechanical approach to investigate asphalt concrete rutting mechanisms, Constr. Build. Mater. 30 (2012) 36–49. [14] N.R. Sefidmazgi, L. Tashman, H. Bahia, Internal structure characterization of asphalt mixtures for rutting performance using imaging analysis, Road Mater. Pavement Des. 13 (S1) (2012) 21–37.
945
[15] R. Khan, J. Grenfell, A. Collop, Micromechanical modelling of typical Dense Bitumen Macadam (DBM) from laboratory testing and X-ray Computed Tomography (CT), Constr. Build. Mater. 77 (2015) 1–6. [16] R. Khan, J. Grenfell, A. Collop, G. Airey, H. Gregory, Moisture damage in asphalt mixtures using the modified SATS test and image analysis, Constr. Build. Mater. 2013 (43) (2013) 165–173. [17] Y.Q. Zhang, X. Luo, R. Luo, R.L. Lytton, Crack initiation in asphalt mixtures under external compressive loads, Constr. Build. Mater. 72 (2014) 94–103. [18] L. Tashman, B. Pearson, Characterisation of stone matrix asphalt mixtures, Int. J. Pavement Eng. 13 (4) (2012) 297–309. [19] D.L. Presti, N.A. Hassan, R. Khan, G. Airey, Reclaimed asphalt test specimen preparation assisted by image analysis, J. Mater. Civ. Eng. 27 (8) (2014) 1–5. C4014005. [20] A.A. Rose, E. Arambula, G. Liu, C.J. Glover, Use of a total air void diffusion depth to improve pavement oxidation modeling with field validation, Pet. Sci. Technol. 33 (2015) 1198–1207. [21] N.A. Hassan, G.D. Airey, N.I.M. Yusoff, M.R. Hainin, R. Putrajaya, M.E. Abdullah, M.M.A. Aziz, Microstructural characterisation of dry mixed rubberised asphalt mixtures, Constr. Build. Mater. 82 (2015) 173–183. [22] H.M. Zelelew, A.T. Papagiannakis, A volumetrics thresholding algorithm for processing asphalt concrete X-ray CT images, Int. J. Pavement Eng. 12 (6) (2011) 543–551. [23] M.E. Kutay, E. Arambula, N. Gibson, J. Youtcheff, Three-dimensional image processing methods to identify and characterise aggregates in compacted asphalt mixtures, Int. J. Pavement Eng. 11 (6) (2010) 511–528. [24] X.Y. Xu, S.Z. Xu, L.H. Jin, E. Song, Characteristic analysis of Otsu threshold and its applications, Pattern Recogn. Lett. 32 (2011) 956–961. [25] A. Mizushima, R. Lu, An image segmentation method for apple sorting and grading using support vector machine and Otsu’s method, Comput. Electr. Agric. 94 (2013) 29–37. [26] J. Hu, Z.D. Qian, D.W. Wang, M. Oeser, Influence of aggregate particles on mastic and air-voids in asphalt concrete, Constr. Build. Mater. 93 (2015) 1–9. [27] N.A. Hassan, G.D. Airey, M.R. Hainin, Characterisation of micro-structural damage in asphalt mixtures using image analysis, Constr. Build. Mater. 54 (2014) 27–38. [28] S. Arasan, E. Yener, F. Hattatoglu, S. Akbulut, S. Hinislioglu, The relationship between the fractal dimension and mechanical properties of asphalt concrete, Int. J. Civ. Struct. Eng. 1 (2) (2010) 165–170. [29] J.U. Baer, T.F. Kent, S.H. Anderson, Image analysis and fractal geometry to characterize soil desiccation cracks, Geoderma 154 (2009) 153–163. [30] A. Farhidzadeh, E. Dehghan-Niri, A. Moustafa, S. Salamone, A. Whittaker, Damage assessment of reinforced concrete structures using fractal analysis of residual crack patterns, Exp. Mech. 53 (2013) 1607–1619.