Construction Quality Control for Rutting Resistance of Asphalt Pavement Using BIM Technology

Abstract

During the course of building asphalt pavement, a lack of quality control will lead to the abandonment of the asphalt mixtures. One of the most common problems with asphalt pavement is rutting. Improving the construction’s quality is an important measure to reduce rutting. The purpose is to ensure the high-temperature durability of asphalt mixtures during the construction workflow to reduce the waste of asphalt mixtures, as well as to provide a methodology for the current monitoring of the quality based on the building information modeling (BIM). Rutting resistance was appraised utilizing the static uniaxial creep examination. Oblique photography technology was used to obtain terrain data. The software of Revit 2016 was used to build the spatial model of highways and bridges. The results show that the size distribution of particles, the asphalt proportion, and the forming specimen’s temperature are the vital elements influencing the high-temperature behavior. The gradation was identified as the most important factor. The second was the asphalt binder content. Gradation variation should be given more consideration during paving using asphalt mixtures. Furthermore, the developed BIM platform can also monitor rutting resistance to reduce rework during construction.

1. Introduction

A lack of construction quality control of asphalt layers will often lead to rework, waste materials, and increased cost [1,2,3,4,5]. This is not conducive to the realization of energy saving and emission reduction. On the other hand, rutting is common distressful for asphalt pavements. It does not only affect driving comfort but also threatens driving safety, especially in rain and snow. Researchers found that one of the major factors causing pavement rutting was the construction effects of asphalt mixtures. The leading goal of controlling the construction process is to verify that the quality of asphalt layers satisfies the design requirements. The construction’s quality can be well controlled through the detection of construction control indices. The asphalt layer construction’s process control is carried out in accordance with the asphalt pavement construction criteria [6,7]. However, the pavement construction criteria do not specify which construction quality assurance measures are associated with rutting resistance or the level of impact. According to the construction criteria, the asphalt binder content, raw material properties, rolling temperature, mixture gradation, compaction passes, asphalt layer thickness, compaction level, and evenness are among the construction control indices [6,7]. Through analysis, it can be found that the key influencing variables of construction quality and the volumetric parameters of asphalt concrete include the aggregate size distribution, asphalt amount, rolling passes, rolling temperature, and layer thickness. However, there is a lack of research on the asphalt mixtures’ performance at high temperatures in construction based on BIM technology control. Leahy [8] used repeated-load triaxial tests and static creep tests to examine the effects of test temperature, deviator stress, asphalt type, asphalt content, aggregate type, and compaction effort on rutting resistance. The factors that influence the formation of ruts are aggregate, binder, mixture, and test field conditions [9]. An innovative machine learning approach was applied to estimate the flow value of asphalt concrete [10]. The NCHRP report 704 lists several factors that impact the rutting of asphalt pavements [11]. Unfortunately, the report does not provide a clear grasp of how these elements influence the genesis of ruts. The sensitivity analysis is not focused on the asphalt layer construction because there are deviations in the values of various factors in the sensitivity analysis in the actual construction process. Moreover, these studies do not provide a comparison of the degree of influence of the aforementioned factors on rutting resistance. It is essential to identify the key indicators of construction control and analyze the impact of various construction control indicators on the asphalt mixtures’ performance at high temperatures to reduce raw material waste.

After determining which control indicators need to be monitored, it is also essential to handle the real-time detection of control indicators. Using digital imaging technology [12,13,14,15] or real-time monitoring technology [6,7], the particle size distribution and the asphalt percentage can be prognosticated promptly in the asphalt mixing plant. In addition, GPR can determine the air gap and asphalt surface thickness in actual time [16,17]. Furthermore, the conditions of the pavement can also be checked through visual analysis [18,19].

When the construction control indices can be monitored, the next question is how to better monitor the values of construction control indices. Currently, the level of informatization of asphalt mixture layer construction is relatively low in the process of construction [20,21]. Construction information is still collected manually and stored in paper files, which severely limits the speed and accuracy of construction quality control. The construction industry has employed Building Information Modeling (BIM) technology to surmount these challenges. To address the challenges of enormous amounts of data analysis, information protection, and cost in the use of the BIM model, Ding et al. [22] designed a cloud-based system for storing BIM concepts. Das et al. [23] proposed an ontological framework leveraging web services to solve the problem of integrating and managing building supply chain data. Lin et al. [24] used machine reading comprehension to solve the problem of quickly finding the required information, and advanced a clever data access technique for cloud-based BIM systems. Sattineni et al. [25] integrated the BIM model and RFID technology for monitoring the status of constructors, equipment, and materials in actual time. Unfortunately, the application of BIM to superintend the construction of asphalt layers has drawn relatively negligible regard. There are many types of BIM software 2016 [26,27]. The majority of these types of software are used for road and bridge modeling and design. Only a few BIM platforms provide for real-time monitoring of construction control indices. As a result, a BIM platform for displaying construction data in real-time must be developed.

The aim of this research is to recognize the basic control indicators for asphalt mixtures’ performance at high temperatures to reduce their waste during construction and to provide a method based on real-time monitoring of these indicators using the proposed BIM platform. In this study, the gradation, asphalt binder content, temperature control during rolling, segregation of gradation, and segregation of temperature were all taken as factors affecting the asphalt concrete’s behavior at elevated temperatures, and it was evaluated using the static uniaxial creep test. More importantly, the BIM system was built for instantaneous tracking of the construction quality with the purpose of reducing their rework. The significance of this study was to identify key indicators of rutting resistance of asphalt mixtures and monitor them using the BIM technology.

2. Materials and Methods

The key control indicators for rutting resistance during construction were determined and monitored in real time, as presented in Figure 1.

2.1. Raw Materials

2.1.1. Asphalt

It can be observed from

Table 1. Characteristics of the asphalt.

Item		Tested Result	Requirement	Test Methods [28]
Penetration (25 °C, 100 g, 5 s) (0.1 mm)		68	60~80	T 0604
Penetration index		0.32	≥−0.4	T 0604
Ductility (5 °C, 5 cm/min) (cm)		44	≥30	T 0605
Softening point (°C)		77.9	≥55	T 0606
Kinematical viscosity (135 °C) (Pa·s)		1.73	≤3	T 0625
Flashpoint (°C)		256	≥230	T 0611
Solubility in trichloroethylene (%)		99.64	≥99	T 0607
Elastic recovery (%)		85.8	≥65	T 0662
Density (25 °C) (g/cm3)		1.023	NA	T 0603
Storage stability (°C)		1.3	≤2.5	T 0661
Thin-film heating test (163 °C)	Mass loss (%)	−0.04919	−1.0~1.0	T 0609
	Residual penetration Ratio (25 °C) (%)	73.5	≥60	T 0604
	Residual ductility (5 °C) (cm)	26.37	≥20	T 0605

that the SBS altered asphalt (I-C) utilized in this research conforms to JTG F40-2004 [6].

2.1.2. Aggregates

For this research, AC-13 was prepared using the crushed basalt aggregates of 10–15 mm, 5–10 mm, 3–5 mm, and 0–3 mm. As exhibited in

Table 2. Characteristics of the aggregates.

Item			Test Result	Requirement	Test Method [29]
Coarse-sized particles	Crushing value (%)		16.9	≤26	T0316
	Loss of sample material owing to abrasion and impact (%)		17.5	≤28	T0317
	9.5~13.2 mm	Ratio of flattened and elongated particles (%)	6.9	≤12	T0312
	4.75~9.5 mm		6.3	≤18
	Polished stone value		46	≥42	T0321
Fine-sized particles	Sand equivalent (%)		68	≥60	T 0334
	Solidness (%)		16	≥12	T 0340

and

Table 3. Limestone filler characteristics.

Item		Test Result	Requirement	Test Method
Specific density		2.655	-	T0352
Water content (%)		0.4	≤1	T0332
Passing (%)	0.6 mm	100	100	T0351
	0.15 mm	91.6	90~100
	0.075 mm	79.1	75~100
Quantitative indication of hydrophilicity		0.3	<1	T0353
Plasticity index (%)		3	<4	T0354

, the attributes of the aggregates and fillers fulfill the technical prerequisites stipulated in [6].

2.2. Mixture Design

Figure 2 depicts the grading range of AC-13. The optimum asphalt binder content (OAR) was determined by applying the Marshall testing protocol. The samples were formed at a thermal level of 160 °C.

Table 4. Marshall test results.

OAR (%)	Air Voids (%)	Voids in Mineral Aggregate (%)	Voids Filled with Asphalt (%)	Marshall Flow (0.1 mm)	Marshall Stability (kN)
5.0	3.8	14.0	72.8	26.9	11.17

displays the outcomes of the mixing design.

2.3. Designing Variability in Construction Monitoring Metrics

2.3.1. Size Distribution, Asphalt Binder Content, and Rolling Thermal Level

According to the aforementioned analysis, the size distribution, asphalt binder content, rolling thermal level, number of passes during compaction, and layer depth all play a role in the high-temperature performance. The number of compaction repetitions can be precisely managed with GPS-RTK technology [30]. Consequently, the gradation, the asphalt proportion, and the rolling temperature were the primary considered factors evaluated in this study for rutting resistance. Figure 3 shows the permissible variation range of the size distribution in the asphalt mixing plant [6]. The maximum amount of variation in asphalt binder content that can be tolerated with ±0.3% [6]. The temperature for molding (or rolling) was kept within a design variation range of 15 °C. The appraisals of how these variances influence the high-temperature actions of asphalt mixtures were then evaluated.

When assessing the influence of grading fluctuation on asphalt concrete’s behavior, it is impractical to accurately attain the intended gradation employing four aggregate sizes. As a result, aggregates of various sizes were sieved into closer size fractions of 0 to 75 μm, 75~150 μm, 150~300 μm, 300~600 μm, 600~1180 μm, 1180~2360 μm, 2360~4750 μm, 4750~9500 μm, 9500~13200 μm, and 13,200~16,000 μm and stored separately. These sieved aggregates were then used to make the asphalt mixture. The asphalt mixtures were created using the sieved aggregates, and the porosity was measured, as shown in

Table 5. Results of Marshall compaction tests.

Number		Gradation	Asphalt Binder Content (%)	Molding Temperature (°C)	Air Voids (%)
		A	B	C
Orthogonal test	1	Upper limit of fluctuation	4.7	145	5.0
	2	Upper limit of fluctuation	5.0	160	3.8
	3	Upper limit of fluctuation	5.3	175	2.9
	4	Design gradation	4.7	160	4.4
	5	Design gradation	5.0	175	3.5
	6	Design gradation	5.3	145	3.0
	7	Lower limit of fluctuation	4.7	175	6.0
	8	Lower limit of fluctuation	5.0	145	5.8
	9	Lower limit of fluctuation	5.3	160	4.4
Control group		Design gradation	5.0	160	3.7

.

2.3.2. Segregation of Size Distribution and Temperature

The problems of segregation of gradation and temperature often occur in asphalt mixtures. Previous studies have shown that these two segregation problems reduce the performance of asphalt mixtures [31,32,33,34,35]. Most studies mainly model large areas of isolation. As shown in Figure 4a, the width of the segregation zone approaches or exceeds the rolling width of the roll. Large-area segregation, on the other hand, is uncommon during the regular building of an asphalt layer. In addition, the capability of the asphalt blend in the dissociated area is uncertain when the segregation region’s magnitude is smaller than the compactor’s flattening width, as exhibited in Figure 4b. Hence, this study chiefly concentrates on examining the effectiveness of limited-area segregation.

As demonstrated in Figure 5, four degrees of gradation dissociation were created. It is clear that employing the same asphalt binder percentage as in the control group for asphalt mixtures with gradation segregation is not a good idea. Directly taking cores from the segregation zone and determining the asphalt binder content is the most feasible way. This method, however, will obliterate the asphalt pavement’s structure. An alternative approach is to sort the unbound asphalt mixture by passing it through sieves of different sizes, such as 9.5 mm and 4.75 mm sieves. By combining these components, it is possible to produce various asphalt mixtures with different levels of aggregate separation and asphalt content. Unfortunately, the loose modified asphalt mixture tends to clump together, and it is hard to alter the ratio of coarse and fine particles. Furthermore, when it comes to altering the degree of gradation segregation, this method is blind.

The effective asphalt film thickness was seen as nearly matching that of the control test batches 1–4 since the graded mixes came from unsegregated asphalt. Employing the statistics in

Table 6. Results from testing the density of aggregates with different sizes.

Item	0~2.36 mm	2.36~4.75 mm	4.75~9.5 mm	9.5~13.2 mm	13.2~16 mm
Bulk specific gravity	2.782	2.819	2.758	2.797	2.763
Apparent specific gravity	2.816	2.915	2.835	2.909	2.881
Proportion (%)	33.1	18.2	28.5	17.9	2.3

and

Table 7. Data for determining the effective coating thickness of the control group.

Bulk Definite Density of the Combined Aggregate	Apparent Specific Gravity	Effective Specific Gravity	Effective Asphalt Content (%)	Surface Area per Unit Mass of the Mixed Aggregates (m2/kg)
2.784	2.857	2.834	4.1	5.21

, and Equation (1), the effectual layer width of the asphalt coat for the control set can be assessed as 7.59 μm. The functional asphalt ratio can be deduced via reverse calculation when the efficient thickness of the asphalt folio is specified. Equations (2) and (3) can then be employed to determine the asphalt content.

$$DA=\frac{P_{be}}{{\gamma }_b\times SA}\times 10$$

(1)

$$P_{ba}=\frac{{\gamma }_{se}-{\gamma }_b}{{\gamma }_{se}\times {\gamma }_{sb}}\times {\gamma }_b\times 100$$

(2)

$$P_b=P_{be}+\frac{P_{ba}}{100}\times P_s$$

(3)

where DA is the effective thickness of the asphalt film (μm); P_be is the effective asphalt content (%); γ_b is the specific gravity of the asphalt (25 °C/25 °C); SA is the specific surface area of the combined aggregate (m²/kg); P_ba is the proportion of asphalt absorbed into the aggregate (asphalt-aggregate ratio) (%); γ_se is the effective specific gravity of the combined aggregate; γ_sb is the bulk specific gravity of the combined aggregate; P_b is the asphalt content (%); and P_s is the ratio of the mass of the aggregates to the mass of the asphalt mixture (%).

According to Reference [36], the computed asphalt binder amount can be adjusted using the following equation:

$$y=1.0475x-0.0737$$

(4)

where y represents the revised asphalt binder content (%); and x represents the computed asphalt binder content (%).

Table 8. Computation of asphalt binder contents of TGs 1~4.

TG	SA (m2/kg)	Pbe (%)	γse	γsb	Calculated Asphalt Binder Content (%)	Corrected Asphalt Binder Content (%)
1	4.38	3.4	2.839	2.787	4.2	4.3
2	3.51	2.7	2.846	2.791	3.5	3.6
3	6.07	4.7	2.825	2.774	5.6	5.8
4	6.93	5.4	2.813	2.762	6.4	6.6

shows the asphalt binder contents of TGs 1~4.

This study examined two degrees of temperature separation, and the temperatures for molding were fixed at 145 °C and 130 °C.

As demonstrated in Table 43 shown in NCHRP report 441 [33], the degree of gradation segregation can be evaluated via the surface texture ratio. TG 1 was segregated at a low level, while TG 2 was segregated at a high level. TGs 3 and 4 had a relation with segregation at low and high levels, respectively. Table 5 of the NCHRP report 441 shows the range of temperature variances [33]. TGs 5 and 6 were categorized under low-tier and high-tier segregation, in that order.

2.3.3. Estimating the Void Content of Segregated Asphalt Pavement

To measure the void content of segregated zone, the Superpave compaction technique was applied. This study examined the impact of the non-segregation zone on the segregation zone. Figure 6 depicts the mesh mold. Before forming samples, first, it is necessary to measure the weight of both homogeneous and heterogeneous asphalt blends. The Superpave Gyratory Compactor’s records showed a correlation between the specimen’s height (or air spaces) and the gyrations number. The asphalt mixtures’ air voids after paving were about 10% [37] for the road construction project. To reach 10 percent and 3.7 percent air voids, the gyration number needed was 11 and 63, respectively. The height of the specimen for segregated zone with 11 gyrations should be equal to that of the non-segregated zone since both had the same thickness after paving. This is the basis for determining the mass of the segregated specimen. Following that, a preset mass of non-segregated and segregated asphalt pavements was deposited on opposite sides of the barrier. The molds for making specimens were designated to correspond to the partition’s position. When the asphalt mixture reaches the predetermined forming temperature, the separator can be removed and placed in the Superpave gyratory compactor for forming. The screen is removed after 63 revolutions. The sample was divided into two parts based on the marking. The cut specimens’ air voids were determined following T 0705−2011 [28], and

Table 9. Measured void content.

Gradation Segregation				Temperature Segregation
TG 1	TG 2	TG 3	TG 4	TG 5	TG 6
8.7%	10.6%	3.1%	2.5%	4.0%	5.3%

shows the results. The research specified a 1:1 ratio between the area of segregation and the area without segregation. By altering the width and position of the divider, different ratios can be achieved.

2.4. Method of Preparing and Testing Specimens

The static uniaxial creep tests were engaged to assess rutting resistance. The specimen preparation and testing procedure were conducted according to appendix C of NCHRP report 465 [38]. The specimen (150 mm × 165 mm) was molded based on Table 5 and Table 9. The flow time can be calculated as follows [39]:

$$\frac{d{\epsilon }_i}{dt}\cong \frac{{\epsilon }_{i+\Delta t}-{\epsilon }_{i-\Delta t}}{2\Delta t}$$

(5)

$$\frac{d{\epsilon }_i^{\ast }}{dt}=\frac{1}{5}\left(\frac{d{\epsilon }_{i-2\Delta t}}{dt}+\frac{d{\epsilon }_{i-\Delta t}}{dt}+\frac{d{\epsilon }_i}{dt}+\frac{d{\epsilon }_{i+\Delta t}}{dt}+\frac{d{\epsilon }_{i+2\Delta t}}{dt}\right)$$

(6)

where dε_i/dt is the creep rate at i sec; ε_{i − Δt} is the strain at (i − Δt) sec; ε_{i + Δt} is the strain at (i + Δt) sec; Δt is the sampling interval; $d{\epsilon }_i^{\ast }/dt$ is the smoothed creep rate at i sec; dε_{i − 2Δt}/dt is the creep rate at (i − 2Δt) sec; dε_{i − Δt}/dt is the creep rate at (i − Δt) sec; dε_{i + Δt}/dt is the creep rate at (i + Δt) sec; and dε_{i + 2Δt}/dt is the creep rate at (i + 2Δt) sec.

2.5. BIM System

The BIM system was established by the project team to enable the real-time and intuitive monitoring of the asphalt pavement’s construction quality. The platform consists of four components: 3D model of road and bridge, geographical data, tracking data, and system attributes. Among them, 3D models can display the construction information associated with roads and bridges. Topographic data can be used to obtain accurate and clear topographic maps. Construction monitoring indicators and construction control data are obtained through experiments and sensors, respectively. The platform’s function is to analyze and display construction information. Figure 7 illustrates the process of constructing the BIM framework.

2.5.1. Pavement Structure Composition

The 3D models were created including subgrade, base course of cement-stabilized macadam, and the surface of asphalt layer in Autodesk Revit 2016 software, and the pavement structure was integrated into the BIM platform. The division of components is based on the paving distance and position of a truck of mixture. The origin of the module is the preliminary paving site of asphalt combination on a lorry. Figure 8 depicts the BIM models for asphalt pavement and subgrade.

2.5.2. Developing BIM Framework

As illustrated in Figure 9, our team created a BIM platform that enabled the tracking of construction data. In the BIM platform, the terrain data can be obtained through oblique photography or existing map data, as shown in Figure 10. The control indicators were measured in the asphalt pavement construction process, and the detection data were uploaded to the data management system of the BIM platform through the 5G/4G network. The components were linked to the observed data. When the component was clicked, the detection data were displayed. In addition, the data from the detection can be displayed and analyzed. For instance, the variation in the asphalt binder content over time can be presented through the BIM platform. The system is accessible to software users from any place where there is a network connection. They can also view and modify models based on their permissions.

3. Results and Discussions

3.1. Effect of Factors Considered on Rutting Performance

Table 10. Results of orthogonal test for asphalt mixtures’ performance at high temperatures.

Number	Gradation	Asphalt Binder Content (%)	Molding Temperature (°C)	Flow Time (s)
	A	B	C
1	Upper fluctuation threshold	4.7	145	1397
2	Upper fluctuation threshold	5.0	160	2969
3	Upper fluctuation threshold	5.3	175	1673
4	Design gradation	4.7	160	1820
5	Design gradation	5.0	175	2316
6	Design gradation	5.3	145	1055
7	Lower fluctuation threshold	4.7	175	713
8	Lower fluctuation threshold	5.0	145	227
9	Lower fluctuation threshold	5.3	160	115
Ij	6039	3930	2679	-
IIj	5191	5512	4904	-
IIIj	1055	2843	4702	-
Ij/3	2013.0	1310.0	893.0	-
IIj/3	1730.3	1837.3	1634.7	-
IIIj/3	351.7	947.7	1567.3	-
Rj	1661.3	889.7	741.7	-

presents the orthogonal test results of the rutting performance.

3.1.1. Extremum Difference Analysis

The operation of the extremum difference analysis is easy and handy, though it is coarse. The outcomes of the extremum difference analysis are displayed in Table 10.

I_j is the summation of flow time values for level 1 in j (j = 1, 2, 3 or 4) column. II_j and III_j are similar to I_j. R_j is the difference between I_j/3, II_j/3, and III_j/3’s max and min values.

The range reflects the extent to which the influence of the factor level changes on the evaluation indices. The significance of the factors can be assessed based on the ranges provided in Table 10. The order of the factors that influence the outcome, from most to least important, is size distribution, percentage of asphalt, and production temperature. Consequently, the control of size distribution and asphalt binder content in the asphalt mixing plant should be paid more attention to with the purpose of obtaining sufficient anti-rutting performance.

Figure 11 presents the relationship between the factor levels and the flow time. For the defined aggregate size levels, the flow time reduced as the aggregate size level moved from the maximum fluctuation limit through the intended aggregate distribution to the minimum fluctuation limit. The coarser the gradation level, the smaller the particular surface area of the mineral particles. Furthermore, if the asphalt-stone ratio is at a high level, the amount of free asphalt will increase, leading to the decrease of the flow time. For the preset levels of percentage of asphalt, the flow time of the asphalt mixture first goes up and then goes down as the asphalt percentage increases. When the percentage of asphalt changes from 4.7% to 5.0%, the flow time increases accordingly. A higher asphalt binder content can facilitate the compaction effect, enhance the structural asphalt quantity, and improve the asphalt concrete’s adhesion. Raising the asphalt binder content from 5.0% to 5.3% results in a reduction in the flow time. If the asphalt binder content exceeds a certain value, the amount of free asphalt increases, resulting in a reduction in the flow time. When the temperature rises between 145 °C and 160 °C during molding, the flow time increases obviously. This is because increasing the molding temperature can lower the asphalt viscosity, contributing to achieving a higher degree of compaction. However, the flow time remains relatively constant with a temperature of 160–175 °C.

3.1.2. Variance Analysis

Table 11. Variance analysis in calculation process.

Test Number	Gradation	Asphalt Binder Content (%)	Molding Temperature (°C)	Flow Time (s)	Square
	A	B	C
1	Upper fluctuation threshold	4.7	145	1397	1,951,609
2	Upper fluctuation threshold	5.0	160	2969	8,814,961
3	Upper fluctuation threshold	5.3	175	1673	2,798,929
4	Design gradation	4.7	160	1820	3,312,400
5	Design gradation	5.0	175	2316	5,363,856
6	Design gradation	5.3	145	1055	1,113,025
7	Lower fluctuation threshold	4.7	175	713	508,369
8	Lower fluctuation threshold	5.0	145	227	51,529
9	Lower fluctuation threshold	5.3	160	115	13,225
Ij	6039	3930	2679	K = 12,285	W = 23,927,903
IIj	5191	5512	4904	-	-
IIIj	1055	2843	4702	-	-
U	21,509,675.7	17,969,897.7	17,778,353.7	P = 16,769,025	-
Q	4,740,650.7	1,200,872.7	1,009,328.7	-	-

and

Table 12. Results of variance analysis.

-	Dispersion	Degree of Freedom	Quadratic Mean Deviation	F
A	4,740,650.7	2	2,370,325.3	22.8
B	1,200,872.7	2	600,436.3	5.8
C	1,009,328.7	2	504,664.3	4.9
Error	208,026.0	2	104,013.0	-
Total	7,158,878.0	8	-	-

exhibit the results of variance analysis.

$$K={\displaystyle \sum _{\mathrm{i}=1}^9{Y}_i}$$

(7)

$$P=\frac{1}{9}{K}^2$$

(8)

$$W={\displaystyle \sum _{\mathrm{i}=1}^9{Y}_i^2}$$

(9)

UA = 1/3 (I₁)² + (II₁)² + (III₁)²]

(10)

UB = 1/3[(I₂)² + (II₂)² + (III₂)²]

(11)

UC = 1/3[(I₃)² + (II₃)² + (III₃)²]

(12)

UD = 1/3[(I₄)² + (II₄)² + (III₄)²]

(13)

QA = UA − P

(14)

QB = UB−P

(15)

QC = UC − P

(16)

QD = UD − P

(17)

QT = W − P

(18)

where QT is the total sum of the deviations’ squares.

For α = 5%, Fα(2, 2) is 19. Hence, F_A was higher than Fα(2, 2), while F_B and F_C were both smaller than Fα(2, 2). Factor A had a significant impact on the flow time; factors B and C have no significant effect on the evaluation index.

3.2. Effect of Segregation on Flow Time

The specimens of the segregated asphalt mixture were molded according to Table 9. Figure 12 displays the outcomes of the static uniaxial creep test for the segregated asphalt mixture.

As depicted in Figure 12, when the asphalt concrete’s gradation grows coarser as a result of segregation, the flow time of the asphalt concrete at first undergoes a minor increase followed by a decrease. This is due to the fact that as the ratio of coarse aggregates increases, the aggregate’s role as a skeletal framework is reinforced. But if the fraction of coarse aggregates exceeds a specific boundary, the asphalt mortar percentage decreases and the bonding capability of the asphalt mixture declines. Conversely, when the size distribution turns finer due to segregation, there is a reduction in the asphalt mixtures’ flow time. This is due to the declining fraction of coarse aggregates and rising proportion of asphalt mortar. The skeleton function of the aggregates will be rapidly weakened. In addition, when temperature segregation occurs, the rolling temperature decreases. The asphalt layer’s compaction degree is reduced, and the skeleton function of the aggregates is significantly affected.

3.3. Comprehensive Comparison of the Extent of Impact of Various Factors on Rutting Resistance

To examine the effect of the fluctuation in individual factors on the anti-rutting of the asphalt mix during the construction phase, it is necessary to incorporate two additional TGs, as depicted in

Table 13. Mixture percentages of supplementary TGs.

Supplementary TG	Size Distribution	Asphalt Binder Content (%)	Temperature during Molding (°C)	Percentage of Void (%)	Flow Time (s)
1	Lower fluctuation threshold	5.0	160	5.5	672
2	Design gradation	5.3	160	2.7	1669

. With consideration of the unfavorable situation, the gradation level of the lower boundary of variation was selected for the supplementary TG 1 and the asphalt binder content of the upper limit of fluctuation was used for supplementary TG 2. Other conditions were consistent with the control group. According to the results of supplementary TGs 1, 2, and TGs 2, 4, and 6 (as shown in

Table 14. Comparison of impact degree of multiple factors on high-temperature performance.

Group		Drop (%)	Relative Analysis of Effect	Group		Drop (%)	Relative Analysis of Effect
Control group		-	-	OT	9	95.0	1
OT	1	39.0	7	Supplementary TG 1		70.7	3
	2	−29.5	17	Supplementary TG 2		27.2	11
	3	27.0	12	TG 1		−3.1	16
	4	20.6	13	TG 2		28.7	10
	5	−1.0	15	TG 3		33.2	8
	6	54.0	5	TG 4		51.2	6
	7	68.9	4	TG 5		17.3	14
	8	90.1	2	TG 6		30.4	9

), it was discovered that numerous components in the construction process had varying degrees of influence on high-temperature performance, ranging from strong to weak, including coarser gradation during mixing, finer gradation from segregation, temperature segregation, and increasing asphalt binder content during the mixing process. When the construction process indices are in the degree of fluctuation in various factors defined in this study, it is important to pay special attention to gradation coarsening in the asphalt mixtures’ mixing process and gradation becoming finer owing to segregation.

3.4. Tracking Rutting Resistance from the BIM Framework

3.4.1. Size-Grade Distribution and the Asphalt Binder Content

Digital image processing technology [12,13,40] and asphalt mixing plant online detection technology [6,7] can both determine the mixture gradation and asphalt dosage in the mixing process in real time [6]. Currently, the data-logging framework of the asphalt mixing plant can accurately gather data such as the weight of asphalt and aggregate, and accurately calculate the amount of asphalt in the tank. The online examination results for the asphalt binder content for each pot of asphalt mixture in a Chinese road construction project are presented in Figure 13. The online examination demonstrates that the asphalt binder content fluctuates just slightly. For example, between 13:00 and 15:00, there were only two pots of asphalt mixture whose asphalt binder content fluctuated.

Table 15. Online detection of the asphalt binder content.

Materials	10–15 mm	5–10 mm	3–5 mm	0–3 mm	Limestone Filler	Asphalt
Mass (kg)	480	579	278	566	141	106
Ratio (%)	23.5	28.3	13.6	27.7	6.9	-
Asphalt binder content (%)	5.2

depicts the current gradation of an asphalt mixed pot.

A truck can carry several pots of asphalt mix. The gradation and the asphalt proportion of transport can be computed according to the online test outcomes of the asphalt mixing plant. The division of the components of asphalt layers is carried out according to the paving distance and position of a truck of the mixture. Thus, the construction information will correspond to the components.

3.4.2. Rolling Temperature

The initial rolling temperature is measured using an infrared temperature sensor, and the temperature value is sent to the data-processing center, as shown in Figure 14. The primary task is to correspond the collected temperature values to the components of the pavement. In the course of paving, the starting and ending pile numbers of the asphalt mixture for transport have been determined, and the location information of the initial rolling temperature can be used to determine the corresponding relationship between the temperature detection values and the components. For the section between K11 + 986.2 and K12 + 44.1 of the road construction project in China,

Table 16. Initial values of rolling temperature detection from paving part starting at K11 + 986.2 and ending at K12 + 44.1.

Pile Number	Temperature (°C)	Pile Number	Temperature (°C)	Pile Number	Temperature (°C)	Pile Number	Temperature (°C)
K11 + 994.29	153.00	K12 + 010.20	179.00	K12 + 023.58	153.00	K12 + 034.75	145.00
K11 + 994.41	153.00	K12 + 010.22	172.00	K12 + 023.58	153.00	K12 + 035.07	160.00
K11 + 997.38	161.00	K12 + 011.30	172.00	K12 + 024.11	152.00	K12 + 036.82	148.00
K12 + 001.01	153.00	K12 + 011.80	175.00	K12 + 024.54	152.00	K12 + 036.99	145.00
K12 + 001.86	152.00	K12 + 012.60	175.00	K12 + 025.05	157.00	K12 + 037.17	145.00
K12 + 002.75	167.00	K12 + 013.73	145.00	K12 + 026.25	157.00	K12 + 038.52	142.00
K12 + 002.82	168.00	K12 + 013.93	145.00	K12 + 026.65	147.00	K12 + 038.90	142.00
K12 + 003.22	159.00	K12 + 014.07	151.00	K12 + 027.92	147.00	K12 + 039.25	147.00
K12 + 003.54	158.00	K12 + 015.72	151.00	K12 + 028.39	168.00	K12 + 039.53	147.00
K12 + 004.15	158.00	K12 + 016.06	159.00	K12 + 028.76	168.00	K12 + 040.88	142.00
K12 + 004.30	171.00	K12 + 016.17	159.00	K12 + 029.23	155.00	K12 + 040.99	151.00
K12 + 005.21	155.00	K12 + 017.28	145.00	K12 + 030.53	155.00	K12 + 041.23	152.00
K12 + 005.38	158.00	K12 + 018.19	145.00	K12 + 030.87	153.00	K12 + 041.33	152.00
K12 + 006.19	171.00	K12 + 018.61	147.00	K12 + 031.60	153.00	K12 + 042.16	170.00
K12 + 006.87	158.00	K12 + 019.89	149.00	K12 + 031.82	142.00	K12 + 042.96	170.00
K12 + 007.52	176.00	K12 + 020.31	152.00	K12 + 032.67	142.00	K12 + 043.41	135.00
K12 + 008.53	179.00	K12 + 021.03	152.00	K12 + 032.98	144.00	-	-
K12 + 008.75	160.00	K12 + 021.54	153.00	K12 + 033.82	144.00	-	-
K12 + 009.66	166.00	K12 + 022.43	153.00	K12 + 034.40	145.00	-	-

provides the temperature readings obtained via infrared thermography.

3.4.3. Tracking of Construction Parameters

The obtained or forecasted construction data are promptly uploaded to the BIM platform. These data are associated with the BIM components of the highway. Then, the construction personnel can sign into the platform network to visualize the construction information and observe the values of construction parameters at all places and times. Figure 15 presents the gradation, asphalt binder content, and the initial rolling temperature associated with a BIM component. Moreover, the construction information can also be analyzed centrally. Every participant involved in the asphalt pavement’s construction can access the BIM system to check the construction details. The login personnel can directly locate the asphalt pavement from the construction information. When the values of the construction indicators exceed the warning range, construction should be suspended and remedial measures should be taken. This can reduce the rework and the waste of asphalt mixtures.

4. Conclusions

In this study, the major construction management indices for rutting resistance were determined to reduce the rework of asphalt layers, and a method for real-time monitoring of these indices using BIM technology was proposed. Rutting resistance was determined by measuring the flow time. Furthermore, the influence of the non-segregation zone on the segregation zone was considered for the asphalt mixtures’ performance at high temperatures. This research’s BIM platform was the basis for the real-time tracking of the construction control indices.

(1)

With the factor levels set, the order of impact that the three factors exert on the asphalt mixtures’ performance at high temperatures is size distribution, asphalt binder content, and temperature for molding. Consequently, it is crucial to manage the variation in gradation for the asphalt mixing plant.

(2)

A slight increase in coarseness in gradation due to segregation can positively impact the asphalt mixtures’ performance at high temperatures. However, when the gradation variation surpasses a certain degree, the flow time decreases rapidly. In contrast, the asphalt mixtures’ performance at high temperatures is adversely affected when the gradation becomes finer due to segregation.

(3)

The asphalt mixtures’ performance at high temperatures declines with an increasing degree of temperature segregation. In addition, there’s a decrease of about 17.3~30.4% in flow time due to the temperature segregation.

(4)

Rutting resistance is notably affected by the fluctuation in gradation, which includes the process of gradation coarsening during mixing and the occurrence of finer gradation due to segregation.

(5)

The BIM system was put forward to follow the instantaneous detection information from the asphalt layer’s construction. This platform allows construction workers to rapidly appraise the quality of the construction, thereby avoiding the asphalt mixture waste.