A study on quality evaluation for bituminous mixture using X-ray CT

Abstract

The objective of this paper is to propose a new quality evaluation method for asphalt concrete mixture using X-ray CT scanner. To achieve this aim, asphalt mixtures should be subjected to the X-ray CT scanning and its characteristics should be clarified. The approach employed in this study was as follows: 1) Coarse aggregate, fine aggregate, filler and bitumen were prepared; 2) dense-graded, coarse-graded and porous asphalt mixtures were made; 3) materials and mixtures were subjected to the X-ray CT scanning; 4) frequency of CT-value, threshold value, average slice CT-value, average segment CT-value were computed. In the material examination, CT-value of aggregate becomes smaller in the order of coarse aggregate, fine aggregate and filler and CT image of bitumen was nearly homogeneous. In the mixture examination, histograms of CT-value and four segmentation images made from CT images expressed the material and mixture characterization such as particle size and the difference in bitumen content and mixture type visibly and the bitumen content varies with the threshold values. In addition, the average segment CT-value without threshold value by dividing the fine aggregate from the coarse aggregate and average CT-value of the coarse aggregate, especially is highly correlated with average CT-value of the bitumen.

Keywords

asphalt concrete mixture / aggregate / bitumen / bitumen content / quality evaluation / X-ray CT

Introduction

An X-ray computed tomography (CT) has become essential medical apparatus to obtain a clear image with high accuracy in a short time within the human body. In industrial applications, X-ray CT has been used as a non-destructive testing in the field of automotive, aerospace, and electronic industries, etc. In the field of civil engineering materials, many studies have used X-ray CT for quality evaluation and understand the failure mechanisms. For example, Otani et al. [1], Desrues [2], Alshibli et al. [3], and Bésuelle et al. [4] used X-ray CT applied for triaxial consolidation tests of geomaterials. Otani [5] simulated the soil behavior when a model pile received loads and Chevalier et al. [6] estimated tunnel face failure. Temmyo et al. [7-8] quantified the internal cement concrete by void, mortar, and coarse aggregate, and evaluated the water cement ratio of the cement concrete.

X-ray CT has the potential to quantify the internal asphalt mixture exactly as well as the cement concrete. Indeed, density of asphalt mixture is almost as same as cement concrete. Besides, cement concrete is consisted of coarse aggregate, fine aggregate, and cement. Similarly, asphalt mixture is consisted of coarse aggregate, fine aggregate, and bitumen.

In the field of asphalt pavement materials, studies on the application of X-ray CT for asphalt mixtures have also been presented since the research on the Strategic Highway Research Program [9]. These studies have mainly focused on void issues such as the distribution, permeability and cracking, and shape of the aggregate. Masad et al. [10] stated the relation between the void of asphalt mixtures and times of gyrations using X-ray CT, and found that an increased number of gyrations led to increased compaction of the middle of specimens. Kutay et al. [11] evaluated the internal pore structure of the asphalt mixture com-paring the prediction value using X-ray CT and image analysis and theoretical equations. Taniguchi et al. [12] suggested using X-ray CT and three-dimensional analysis for diagnosis. You et al. [13] and Zelelew et al. [14] conducted dynamical evaluation of asphalt mixtures using X-ray CT images.

Relative compaction, gradation (2.36 mm and 0.075 mm), and bitumen content is normally designated as inspection items of asphalt mixtures [15-16]. For instance, Japan Road Association [17] recommends that the bitumen content is based on the extraction test. However, this extraction test is time-consuming and underestimates fine aggregate [18].

The key issue in the quality assurance of asphalt concrete pavement is the development of the quality evaluation method quickly, accurately, and fairly. Therefore, and a quicker, easier and more effective evaluation method is required. X-ray CT has the potential for application in evaluating the bitumen content and gradation, as well the possibility of deterioration evaluation and designing appropriate mixes of recycled materials for asphalt.

Taniguchi et al. [19-21] performed a series of X-ray CT test using asphalt concrete mixtures to suggest new quality evaluation method using X-ray CT. Difference of bitumen content, bitumen type and mixture type was examined in the X-ray CT test. The interest conclusions were that X-ray CT image could visualize the internal of the asphalt mixture, shape of aggregates, and distribution of binder and voids, and CT indexes such as CT-value, threshold value, average slice CT-value, average segment CT-value evaluate characteristics of the mixture. Especially, CT-value is effective to evaluate quality of the asphalt mixture since CT-value is almost linear to the density of the material [22].

The objective of this paper is to propose a new quality evaluation method for asphalt concrete mixture using X-ray CT scanner. To achieve this aim, asphalt mixtures should be subjected to the X-ray CT scanning and its characteristics should be clarified. The approach employed in this study was as follows: 1) Coarse aggregate, fine aggregate, filler and bitumen constituting asphalt mixture were prepared. 2) Dense-graded, coarse-graded and porous asphalt mixtures were made. 3) Materials and mixtures were subjected to the X-ray CT scanning. 4) Frequency of CT-value, threshold value, average slice CT-value, average segment CT-value were computed.

Materials and methods

X-ray CT

X-ray CT scanner

Figure 1 shows the principle of the industrial X-ray CT scanner owned by Kumamoto University. The X-ray CT scanner penetrates the collimated X-ray through the entire circumference of the specimen by translating and rotating the specimen table of this apparatus. The image data processing device assembles the detected data, and reconstructs the cross sectional images [22].

Voxel

A three-dimensional “voxel” constructs the X-ray CT image as shown in Fig. 2. Therefore, the volume of X-ray attenuation refers to the density in each voxel [22]. Figure 3 shows the relation between aggregate or filler size, and voxel size. Coarse aggregate is larger, some fine aggregate is partially smaller, and filler is for the most part smaller than voxel. Therefore, CT image cannot express filler as a particle. In this case, the use of bitumen and filler as the binder is appropriate since bitumen and filler work as a binder in asphalt mixtures. [20]

CT-value

The X-ray attenuation coefficient is an index of the X-ray impermeability and is dependent on the density of the material. Specifically, the higher the density of the target material is, the higher the X-ray attenuation coefficient, and the lower the density is, the lower X-ray attenuation coefficient. [22]

Equation (1) indicates the “CT-value” in the image processing analysis.

(1)\mathit{C}{\mathit{T}}_{\mathbf{v}}=\mathit{\kappa }(\mathit{\mu }-{\mathit{\mu }}_{\mathbf{w}})/{\mathit{\mu }}_{\mathbf{w}}

where BoldItalic_v = CT-value; BoldItalic = constant (normally BoldItalic = 1000); BoldItalic = X-ray attenuation coefficient of the target material; BoldItalic_w = X-ray attenuation coefficient for water

Equation (1) states that the CT-value also depends on the density. Moreover, the CT-value contains information about the CT image contrast. The CT images express a shaded gray or black color for a low CT-value, and a light gray or white color for a high CT-value. [22]

Specification of X-ray CT scanner

Table 1 shows the specification of X-ray CT scanner used in this study. A voltage of X-ray tube is related to the intensity of X-ray beam into the specimen. To avoid causing X-ray beam hardening in the CT image, 300 kV was selected in this study. [23] The beam thick ness was collimated as 2 mm in the material scanning in order to grasp the CT-value properties of the materials broadly, and as 1 m in the mixture scanning to understand the CT-value properties of the mixture minutely. The diameter of the scan area was selected 150 m in diameter, and the voxel number in the image was 2048 × 2048 voxels. Hence, the dimension of one voxel was 0.073 m × 0.073 m × 1.0 m or 2.0 mm³ because what the diameter of 150 mm in scan area divided 2048 was 0.073 mm.

Quantifying segment of asphalt mixture

This image data processing device calculates the CT-value of asphalt mixtures based on the X-ray attenuation. The CT-value varies by the segment ratio of the void, binder (bitumen including filler), and aggregate. Specifically, the higher the amount of the aggregate is, the higher the CT-value, and the greater the void is, the lower the CT-value. The threshold CT-value needs to be set to quantify the segment of asphalt mixtures.

The Percentile method is one of the easy methods to find the threshold [24]. This method is effective when the volume ratio of each segment is already known. Figure 4 shows the CT-value distribution of a typical dense-graded asphalt mixture [20-21]. The areas shown in Fig. 4 are as below; BoldItalic₁ denotes the area of the void, BoldItalic₂ denotes the area of the binder, BoldItalic₃ denotes the area of fine aggregate, BoldItalic₄ denotes the area of coarse aggregate, and BoldItalic expresses the total area (BoldItalic₁ + BoldItalic₂ + BoldItalic₃ + BoldItalic₄). The volume ratios of each segment are as below; BoldItalic_v denotes the void ratio, BoldItalic_a denotes volumetric binder ratio, BoldItalic_s denotes volumetric fine aggregate ratio, and BoldItalic_g denotes volumetric coarse aggregate ratio. The following shows computing procedure:

1) Calculate the area from minimum CT-value

2) Set threshold CT-value No.1 (BoldItalic₁) by dividing the void from the binder to the CT-value to yield BoldItalic_v = BoldItalic₁/BoldItalic

3) Set threshold CT-value No.2 (BoldItalic₂) by dividing the binder from the fine aggregate to the CT-value to yield BoldItalic_a = BoldItalic₂/BoldItalic

4) Set threshold CT-value No.3 (BoldItalic₃) by dividing the fine aggregate from the coarse aggregate to the CT-value to yield BoldItalic_s = BoldItalic₃/BoldItalic

Average slice and segment CT-value (P)

Equation (2) indicates the average slice CT-value of specimens. This value indicates the specimen’s density.

(2)\mathit{C}{\mathit{T}}_{ave}={\displaystyle \sum _{\mathit{C}{\mathit{T}}_{\mathbf{m}\mathbf{i}\mathbf{n}}}^{\mathit{C}{\mathit{T}}_{\mathbf{m}\mathbf{a}\mathbf{x}}}i\times {\mathit{n}}_{\mathit{i}}/\mathit{N}},

where BoldItalic_ave = average CT-value of the specimen; BoldItalic_min = minimum CT-value; BoldItalic_max = maximum CT-value; BoldItalic_BoldItalic = frequency when CT-value is BoldItalic; BoldItalic = total voxel number.

Equations from (3) to (6) indicate the average CT-value of four segments (void, binder, fine aggregate, and coarse aggregate). These values indicate each segment’s density in the mixtures.

(3)\mathit{C}{\mathit{T}}_{\mathbf{v}}={\displaystyle \sum _{\mathit{i}=\mathit{C}{\mathit{T}}_{\min }}^{{\mathit{T}}_{\mathrm{1}}-\mathrm{1}}\mathit{i}\times {\mathit{n}}_{\mathit{i}}/{\mathit{N}}_{\mathbf{v}}},

(4)\mathit{C}{\mathit{T}}_{\mathbf{a}}={\displaystyle \sum _{\mathit{i}={\mathit{T}}_{\mathrm{1}}}^{{\mathit{T}}_{\mathrm{2}}-\mathrm{1}}\mathit{i}\times {\mathit{n}}_{\mathit{i}}/{\mathit{N}}_{\mathbf{a}}},

(5)\mathit{C}{\mathit{T}}_{\mathbf{s}}={\displaystyle \sum _{\mathit{i}={\mathit{T}}_{\mathrm{2}}}^{{\mathit{T}}_{\mathrm{3}}-\mathrm{1}}\mathit{i}\times {\mathit{n}}_{\mathit{i}}/{\mathit{N}}_{\mathbf{s}}},

(6)\mathit{C}{\mathit{T}}_{\mathbf{g}}={\displaystyle \sum _{\mathit{i}={\mathit{T}}_{\mathrm{3}}}^{\mathit{C}{\mathit{T}}_{\max }}\mathit{i}\times {\mathit{n}}_{\mathit{i}}/{\mathit{N}}_{\mathbf{g}}},

where BoldItalic_v = average CT-value of the void; BoldItalic_a = average CT-value of the binder; BoldItalic_s = average CT-value of the fine aggregate; BoldItalic_g = average CT-value of the coarse aggregate; BoldItalic_BoldItalic = frequency when the CT-value is BoldItalic; BoldItalic_v = total voxel number of the void; BoldItalic_a = total voxel number of the binder; BoldItalic_s = total voxel number of the fine aggregate; BoldItalic_g = total voxel number of the coarse aggregate

Material and asphalt mixture

Material

Crushed stone was used as coarse aggregate, its range of grain size was from 2.5 mm to 20 mm. Screenings, crushed sand and fine sand were used as fine aggregate and its maximum grain size was 2.36 mm. Mineral powder was used as filler and its 0.075 mm sieve passage is 80.9%. Straight asphalt, polymer modified asphalt against rutting (PMA II), and high viscosity polymer modified asphalt for porous asphalt concrete (PMA H) were adopted. Table 2 shows physical properties of these bituminous binders.

Specimen was made by filling coarse aggregate, fine aggregate, filler and straight asphalt to 55 mm height in 50 mm diameter polycarbonate container (Fig. 5). The density of these materials is shown as Table 3.

Asphalt mixture

First, five Dense-graded asphalt mixtures using straight asphalt were prepared and their bitumen content was set to 4.5%, 5.0%, 5.5%, 6.0%, and 6.5% since the optimum bitumen content was 5.5%. Then, dense-graded using PMA II, coarse-graded using straight asphalt, and porous asphalt mixture using PMA H were added in the second stage and their bitumen content was set to 5.0% since the optimum bitumen content of coarse-graded and porous was 5.0%. So, eight specimens were made according to the experimental design table as shown in Table 4. Specimens were 101.6 mm in diameter and 68.7 mm in height for the dense and coarse-graded and 61.2 mm in height for the porous. The combined grading of each mixture type is shown in Fig. 6. The percentage mass of the aggregate is shown in Table 5. The volume ratio of the asphalt mixture is shown in Table 6. The density of the specimen is shown in Table 7.

X-ray CT scanning

Figure 7 shows the irradiation positions set to 10.0, 25.0, 40.0 mm from the bottom for the materials specimen of 55 mm in height, to 10.0, 21.5, 34.3, 46.4 and 58.7 mm from the bottom for the dense and coarse-graded asphalt mixture specimen of 68.7 mm in height, and 10.0, 20.3, 30.6, 40.9 and 51.2 mm for the porous asphalt mixture specimen of 61.2 mm in height.

X-ray CT scanning was performed divided into three stages. First, materials were scanned. Second specimens with different bitumen content were scanned. Third, different bitumen and mixture type of specimens were scanned.

Result and discussion

Material

Figure 8 shows the CT images of the coarse aggregate, fine aggregate, filler, and straight asphalt at 25 mm in the irradiating height.

Coarse aggregate

CT image of coarse aggregate clearly expressed the shape of aggregate and the void as shown in Fig. 8(a). The CT-value histogram of coarse aggregate showed bimodal as shown in Fig. 9. This is because the grain size of the coarse aggregate was larger than the voxel size (0.073 mm × 0.073 mm × 2.0 mm) as shown in Fig. 3, and no material fill the void of coarse aggregate. The peak CT-value of about -1000 represents the void and about 1600 represents the aggregate. The CT-value frequency was about 100 in the range from -1000 to 1600 since the aggregate and void were mixed in the voxel.

Fine Aggregate

CT image of fine aggregate indicated that the small particle filled the void with the large particle as shown in Fig. 8(b). The CT-value histogram of fine aggregate showed unimodal as shown in Fig. 10. This is because almost grain size was partially smaller than the voxel size as shown in Fig. 3, and both void and aggregate existed in the voxel. Hence, the peak CT-value of fine aggregate is smaller than coarse aggregate.

Filler

Figure 8(c) shows that filler is impossible to recognize as a particle. Figure 11 indicates that the peak CT-value of filler is smaller than coarse and fine aggregate although the density of filler is larger than the aggregates. This is because both the void and the aggregate existed in the voxel since the size of filler is almost smaller than the voxel size as shown in Fig. 3.

Bitumen

Figure 8(d) shows CT image of straight asphalt is nearly homogeneous. Figure 12 indicates that three lines overlapped in the histogram. However, many dark parts, that is to say void of bitumen, were found in the CT image of bitumen. Filler fill these voids and improve the stability and quality [25]. Therefore bitumen and fill were treated as a binder segment in this study.

Asphalt mixture

Figure 13 shows the CT images of eight asphalt mixtures at the center of the specimen. These figures indicate the shape of the aggregates and the position of the void.

Difference in bitumen content

Figure 14 shows the histogram of specimen from No. 1 to No. 5. The X-axis indicates the CT-value, and the Y-axis indicates voxel number when the CT-value is BoldItalic.

This histogram shows that the increase also reduced the voxel number of the CT-value in the case the CT-value is more than 500 and less than 1000, and increased in the case the CT-value is more than 1000.

In addition, this histogram indicates bimodal distribution. The lower peak can be regarded as fine aggregate and higher peak can be regarded as coarse aggregate. This is because the lower peak is present between BoldItalic₂ and BoldItalic₃ and the CT-value of coarse aggregate is larger as BoldItalic₃ as shown in Fig. 15. Besides, coarse aggregate is larger, some fine aggregate is partially smaller, and filler is for the most part smaller than voxel as shown in Fig. 3 and the peak CT-value of coarse aggregate is larger than fine aggregate as shown in Fig. 9 and 10.

Figure 16 shows the relation between bitumen content and three threshold values, T₁ dividing the void from the binder indicated a tendency to increase when the bitumen content was less than 5%. This is because bitumen got into the void under low bitumen content. However, T₁ indicated a tendency to decrease when the bitumen content was higher than 5%. This is because as the voxel number of the void decreases the bitumen content excessively increases. Therefore, T₁ becomes small.

BoldItalic₂ dividing the binder from the fine aggregate indicated a tendency to become large as the bitumen content increased. This is because the size of fine aggregate was partially less than the voxel size (0.073 mm × 0.073 mm × 1.0 mm) shown in Fig. 3, and the bitumen got into the voxel.

BoldItalic₃ dividing the fine aggregate from the coarse aggregate did not vary much by the bitumen content. This is because the percent mass of the coarse aggregate was constant.

Four-segmentation images shown in Fig. 17 were drawn based on the threshold value. These figures clearly indicate that the void shown in white decreases and the binder shown in black increases due to the increase in the bitumen content.

Figure 18 shows the relation between the average slice CT-value and density of specimens. The correlation was high in both, and the linear nature was formed. The linear approximate expression became Eq. (7), and the coefficient of correlation was 0.970.

(7)\mathit{d}=\mathrm{8.847}\times {\mathrm{10}}^{-\mathrm{4}}C{T}_{\mathrm{a}\mathrm{v}\mathrm{e}}+\mathrm{1.356},

where BoldItalic = density of specimen (g/cm³); BoldItalic_ave = average slice CT-value of specimen

Figure 19 shows the relation between the bitumen content and average CT-value of four segments.

The average CT-value of the void (BoldItalic_v) indicated a tendency of raising the CT-value when the bitumen content was less than 6.0%. Conversely, the BoldItalic_v indicated a tendency to reduce when the bitumen content was greater than 6.0%. This was exactly the same tendency as threshold value No.1 (BoldItalic₁). This is because bitumen gets into the void under low bitumen content and the voxel number of the void decreases at high bitumen content.

The average CT-value of the binder (BoldItalic_a) was highly correlated with the bitumen content (BoldItalic). Equation (8) expresses the linear approximate expression, and the coefficient of correlation was 0.982.

(8)C{T}_{\mathrm{a}}=\mathrm{65.4}\mathrm{x}{\mathrm{10}}^{-\mathrm{4}}\mathit{A}\mathit{S}+\mathrm{515.1}

The average CT-value of the fine aggregate (BoldItalic_s) slightly increases with the increase of the bitumen content. This tendency is the same as with threshold value No.2 (BoldItalic₂) since the aggregate size is less than the voxel size shown in Fig. 3.

The average CT-value of the coarse aggregate (BoldItalic_g) was mostly constant, and independent of the bitumen content.

Difference in bitumen type

Figure 20 shows histogram of the CT-value of specimen No. 2 and 6. This figure indicates that the frequency of the specimen No. 6 (PMA II) is larger at the peak of fine aggregate and smaller at the peak of coarse aggregate than the specimen No. 2 (straight asphalt). However, the tendency of the distribution is almost unchanged between specimen No. 2 and specimen No. 6. This is because the percentage volume of aggregate was slightly changed.

Figure 21 shows threshold values of specimen No. 2 and 6. This figure indicates that BoldItalic₁ of the specimen No. 6 is slightly smaller than No. 2. This is because the bitumen void ratio of specimen No. 6 is smaller than No. 2. Therefore, increase of bitumen content raises BoldItalic₁. For BoldItalic₂ and BoldItalic₃, almost same values were obtained since the volume ratio of the fine and coarse aggregate is the same between specimen No. 2 and 6.

Figure 22 shows four-segmentation images of specimen No. 2 and 6. From this figure, similar and fine distribution of four segments appeared.

Accordingly, the effect of bitumen type on CT images is much smaller than other factors.

Difference in mixture type

Figure 23 shows histogram of the CT-value of specimen No. 2, 7 and 8. This histogram indicates that the peak appeared in specimen No. 7 (coarse-graded asphalt mixture) and 8 (porous asphalt mixture) at the range of CT-value from -1000 to -500. The peak frequency is about 60 in the specimen coarse-graded when the CT-value is about -900 and about 500 in the porous when the CT-value is about -850. (Normally, the CT-value of a void is -1000 for reference.) Many coarse aggregates are thought to cause uneven placement of materials while dense-graded asphalt mixture are thought to give an even distribution of materials.

On the contrary, the peak disappeared in the coarse-graded and porous asphalt mixture in the range of CT-value from 1000 to 1300. This is because the volume of fine aggregate decreases.

As shown in Fig. 24, the 3 value of dense graded asphalt mixtures is the highest and coarse-graded is the lowest focusing on BoldItalic₁. The volume ratio of the void in porous asphalt mixtures is much higher than coarse-graded. However, the BoldItalic₁ of the porous mixture is much higher than that of the coarse-graded mixture since the concentration to the peak occurred and the non-peak value became extremely small. For BoldItalic₂, almost the same values were obtained since the volume ratio of the fine aggregate is the same between specimen No. 2, 7 and 8. BoldItalic₃ increased in the order of dense-graded, coarse-aggregate and porous asphalt mixture. This is because void got into the coarse aggregate.

Four-segmentation images shown in Fig. 25 clearly indicate that the area of the void increase and the fine aggregate decrease due to the increase of the coarse aggregate and decrease of the fine aggregate. It is also found that the distribution of the coarse-graded asphalt mixture is rougher than the dense-graded and porous mixtures.

Conclusions

This paper presents the results of X-ray CT test for aggregate, filler, bitumen and binder. The following is a summary of conclusions:

1) CT-value of aggregate becomes smaller in the order of coarse aggregate, fine aggregate and filler and CT image of bitumen was nearly homogeneous;

2) Histograms of CT-value and 4 segmentation images made from CT images expressed the material and mixture characterization such as particle size and the difference in bitumen content and mixture type visibly.

3) The bitumen content varies with the threshold values, and the average segment CT-value without BoldItalic₃ and BoldItalic_BoldItalic, especially is highly correlated with BoldItalic_a.

Since CT-value characteristics necessary to evaluate the quality of asphalt mixture has been grasped, proposal of the quality evaluation method and mechanism of damage using X-ray CT will be conducted for future study.

References

[1] Otani J, Mukunoki T, Obara Y. Characterization of failure in sand under triaxial compression using an industrial X-ray CT scanner. International Journal of Physical Modelling in Geotechnics, 2002, 2(1): 15–22

[2] Desrues J. Tracking strain localization in geomaterials using computed tomography. In: Proceeding of the 1st International Workshop on X-Ray CT for Geomaterials (GEOX2003). Kumamoto, Japan, 27–41

[3] Alshibli K A, Batiste S N, Sture S, Lankton M. Micro-characterization of shearing in granular materials using computed tomography. In: Proceeding of the 2nd International Workshop on X-Ray CT for Geomaterials (GeoX2006). Grenoble, France, 2006, 17–34

[4] Bésuelle P, Viggiani G, Lenoir N, Desrues J, Bornert M. X-ray micro CT for studying strain localization in clay rocks under triaxial compression. In: Proceeding of the 2nd International Workshop on X-Ray CT for Geomaterials (GeoX2006). Grenoble, France, 2006, 35–52

[5] Otani J. X-ray computed tomography for geotechnical engineering. In: Proceeding of the 2nd International Workshop on X-Ray CT for Geomaterials (GeoX2006). Grenoble, France, 2006, 95–115

[6] Chevalier B, Takano D, Otani J. Comparison of X-ray CT and discrete element method in the evaluation of tunnel face failure. In: Proceeding of the 3rd International Workshop on X-Ray CT for Geomaterials (GeoX2010). New Orleans, the United States, 2010, 397–405

[7] Temmyo T, Murakami Y, Obara Y. Estimation of structural characteristics of RCD by an X-ray CT method. In: Proceeding of the 1st International Workshop on X-Ray CT for Geomaterials (GEOX2003). Kumamoto, Japan, 199–205

[8] Temmyo T, Murakami Y, Obara Y. Evaluation of water cement ratio of hardened cement concrete by X-ray CT method. In: Proceeding of the 2nd International Workshop on X-Ray CT for Geomaterials (GeoX2006). Grenoble, France, 2006, 443–449

[9] Synolakis C E, Leahy R M, Singh M B, Zhou Z, Song S M, Shannon D S. Development of an asphalt core tomographer: SHRP-A-656. National Research Council, 1993

[10] Masad E, Muhunthan B, Shashidhar N, Harman T. Internal structure of characterization of asphalt concrete using image analysis. Journal of Computing in Civil Engineering, 1999, 13(2): 88–95

[11] Kutay M E, Aydilek A H. Accuracy of the two common semi-analytical equations in predicting asphalt permeability. In: Proceeding of the 2nd International Workshop on X-Ray CT for Geomaterials (GeoX2006). Grenoble, France, 2006, 301–306

[12] Taniguchi S, Nishizaki I, Moriyoshi A. Three-dimensional diagnosis of pavement damage using CT scanner. In: Proceeding of 6th International Conference on Maintenance and Rehabilitation of Pavements and Technological Control (MAIREPAV6). Torino, Italy, 2009, 935–944

[13] You Z, Adhikari S, Kutay M E. Dinamic modulus simulation of the asphalt concrete using the X-ray computed tomography images. Materials and Structures, 2009, 42(5): 617–630

[14] Zelelew H M, Papagiabbakis A T. DEM simulation of asphalt concrete uniaxial creep. In: Proceeding of 6th International Conference on Maintenance and Rehabilitation of Pavements and Technological Control (MAIREPAV6). Torino, Italy, 2009, 242– 251

[15] Carifornia Department of Transportation. Quality Control Manual for Hot Mix Asphalt, 2009

[16] Japan Road Association. Pavement Design and Construction Guideline, 2006 (in Japanese)

[17] Japan Road Association. Pavement Surveying and Testing Handbook, 2007(in Japanese)

[18] Akiba S, Kanou Y. A study on the test for extraction of asphalt from asphalt mixture of the environment balance type using sub-critical water. In: Journal of Pavement Engineering, Japan Society of Civil Engineering, 2010, 15: 23–29 (in Japanese)

[19] Taniguchi S, Otani J, Nishizaki I. New evaluation index of asphalt concrete mixture using X-ray CT method. In: Proceedings of the 6th China-Japan Workshop on Pavement Technologies. Naha, Japan, 2011, 271–279

[20] Taniguchi S, Nishizaki I, Ogawa K, Otani J. A feasibility study on the quality inspection method of asphalt pavement using X-ray CT. In: Proceedings of the 7th International Conference on Maintenance and Rehabilitation of Pavements and Technological Control (MAIREPAV7). Auckland, New Zealand, 2012, No. 53

[21] Taniguchi S, Nishizaki I, Ogawa K, Otani J. New X-ray CT evaluation method of engineering characteristics of asphalt mixture. In: Proceedings of the 2nd International Conference on Transportation Geotechnics (ICTG). Sapporo, Japan, 2012, No. 57

[22] Otani J. State of the art report on geotechnical X-ray CT research at Kumamoto University. In: Proceedings of the 1st International Workshop on X-Ray CT for Geomaterials (GEOX2003). Kumamoto, Japan, 2003, 43–77

[23] Mukunoki T, Kumano N, Otani J. Image analysis of soil failure on defective underground pipe due to cyclic water supply and drainage using X-ray CT. Frontier of Structural and Civil Engineering, 2012, 6(2): 85–100

[24] Parker J R. Algorithms for Image Processing and Computer Vision. Indianapolis: Wiley Computer Publishing, 1997

[25] Japan Road Association. Manual for Asphalt Pavement. Japan Road Association. 1989

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg