Journal of Transportation Engineering, Part B: Pavements

Roller-integrated compaction monitoring technology has been widely used in highway construction and has played an important role in controlling compaction quality and construction progress. In order to detect and assess the compaction quality of the subgrade accurately in real time based on the analysis of the acceleration signal of the vibrating wheel, a total of four real-time monitoring indexes, namely, peak value of acceleration ($A_p$), compaction meter value (CMV), root-mean square value of acceleration ($A_{\mathrm{rms}}$), and crest factor value (CF), were selected to characterize the compactness of the soil. To determine which of these indexes is the most appropriate, field compaction experiments under various rolling conditions were carried out on the fifth section of the Hengyong Highway in Hunan Province, China. The data of compaction parameters, compaction quality, and integrated compaction monitoring under various working conditions were collected and subjected to statistical regression analysis. On the basis of the time-frequency analysis of vibration response signals, the CMV could more accurately characterize the compactness of low-liquid-limit silt. In addition, the mapping relations between compaction quality index and roller-integrated compaction monitoring indexes were founded to define the control criterion of roller-integrated compaction monitoring index, namely, CMV, providing scientific guidance for making sure the compaction quality of the subgrade during construction.

Evaluation of pavement deflection bowl using nondestructive testing (NDT) is an ideal method for assessing the structural capacity of individual layers of a pavement system. However, the cost of modern NDT equipment and the complexity involved in the back-calculation process makes it difficult for the road agencies to back-calculate individual layer properties and evaluate the in situ condition of existing pavements. This study developed low-cost equipment (LCE) for evaluating the pavement deflection bowl under static loading conditions and to correlate the results with the deflection bowl observed under dynamic loading on same pavement combinations. The study also analyzed the deflection bowl parameters through an understanding of the load distribution evaluated using linear elastic software. The moduli of the individual pavement layer were back-calculated using back-calculation software and correlated with the deflection bowl parameters. The study concluded that the deflection bowl under dynamic loading was only 0.65–0.75 times that of the deflection bowl under static loading. The difference of deflection measured at radial distances of 0 and 450 mm of the deflection bowl correlated well with the base-layer modulus, and the difference of deflection measured at radial distances of 300 and 600 mm correlated well with the subbase-layer modulus. Based on the deflection bowl data collected from nine test sections, it was found that emulsion-treated base layers exhibited 40%–60% higher moduli values than the conventional base layers during the monsoon season. Low-volume roads with dense surface mixes 30 mm thick exhibited significantly improved structural performance compared with that of low-volume roads with open-graded surface mixes 20 mm thick with a 6-mm seal coat.

The objective of this study was to develop a generic pavement condition assessment scale, called the unified pavement health index (UPHI), that is capable of quantifying asphalt pavement surface quality, which could be altered based on region-specific conditions. The database, consisting of 541 road sections encompassing 1,345,397 data points in the State of Andhra Pradesh, India, was used to formulate the metric. Five severity-based distress indexes were established to estimate the UPHI on a scale of 0 to 100. The metric was categorized into five groups to help inform maintenance interventions. A deep neural network was developed to automatically compute the UPHI for distresses data, which had a coefficient of determination of 96.05%. It is envisioned that the unified metric developed in this study will help practitioners in assessing the pavement functional condition based on 18 distresses and suggest maintenance strategies for network-level management in a rapid and rational manner.

This paper investigated the effect of the changed albedo of colored pavement on the urban temperature. The surface albedo of different conventional and colored pavement slabs was measured in the laboratory. Numerical simulations were conducted to determine the relationship between pavement albedo and the temperatures of pavement and its surrounding environment in the local urban. Results showed that colored porous portland cement concrete (PPCC) slabs have higher albedo values than conventional gray PPCC and black asphalt concrete, but conventional dense portland cement concrete (DPCC) has the highest albedo value owing to its smooth surface. The temperature of colored PPCC (except black PPCC) slabs at 1-cm depth was about 1°C–6°C lower than that of conventional gray PPCC due to the high albedo, and the green PPCC slab had the best cooling effect among all PPCCs. The simulated temperature profiles of the PPCC slab were found to be in good agreement with the laboratory test results. The local urban simulation showed that the colored pavement has a noticeable effect on reducing pavement surface temperature and nearby air temperature. Compared with open-graded friction course, which had an albedo value of 0.058, conventional gray DPCC with an albedo value of 0.32 had a temperature at 0.2 m below the pavement surface that was 0.83°C lower at 3:00 p.m. of a typical summer day.

Subgrade strength is characterized by many indices such as the California bearing ratio, resilient modulus, and dynamic modulus. The California bearing ratio (CBR) of the subgrade soil is an important input parameter in the design of flexible pavements and the CBR of soil is largely influenced by its basic and mechanical properties such as gradation, maximum dry density (MDD), optimum moisture content (OMC), liquid limit (LL), plastic limit (PL), and plasticity index (PI). In many cases, the existing statistical relations between CBR and basic soil properties do not show good accuracy as they are developed based on limited data points of specific soil types. It is imperative to use more sophisticated and advanced algorithms to predict the CBR of the soil. In recent years, artificial intelligence (AI)/machine learning techniques have gained popularity in geotechnical engineering in solving problems when deterministic solutions are unavailable or are very expensive in terms of computational cost. The present study uses a large amount of input data consisting of 679 data points from published literature and attempts to predict the CBR of soil using a K nearest neighbor (KNN) regressor, random forest regressor (RFR), support vector regressor (SVR), multilayer perceptron (MLP), and multilinear regression (MLR). The RFR is found to result in the maximum accuracy of predicted CBR followed by KNN, MLP, SVR, and MLR methods. Additionally, sensitivity analysis is performed and the results show that MDD followed by the percentage of gravels are the most influencing parameters affecting CBR compared to the percentage of fines and Atterberg limits (LL and PL) of the soil. Furthermore, the conventional curve-fitting approach is used to find the best-fit curve for the variation of CBR with MDD and the percentage of gravel (the two most sensitive parameters). The coefficient of determination of CBR based on curve fitting and the RFR model trained on the stated two features are compared. The RFR model is found to give more accurate predictions than the conventional regression analysis examined in this study.

Most highway agencies use the default values of the lane distribution factor (LDF) as suggested by their pavement design guidelines. However, the use of the default value of the LDF results in over- or underestimation of the number of standard axle repetitions, especially for multilane roads, which have a larger variation in the movement of vehicles across the entire carriageway compared with one- and two-lane roads. Therefore, this study focused on the accurate measurement of the LDF for multilane roads. Currently, the LDF is estimated as the ratio of the number of heavy vehicles in a design lane to the total number of heavy vehicles moving on the carriageway. However, in the case of multilane roads, the same vehicle can occupy more than one lane, which ultimately results in inaccurate measurement of the LDF. To overcome this issue, a strip-based approach is introduced for a microlevel analysis of the LDF of multilane roads. The vehicle damage factor (VDF) and LDF were obtained by extracting the traffic data collected through the video recording method. The results of the LDF indicate that the channelization of trucks occurred in the intermediate lane instead of the outer lane, which commonly is assumed for the estimation of the number of standard axle repetitions for flexible pavement design. A significant variation was observed in the number of standard axle repetitions estimated using actual LDF and the default LDF values suggested by the Indian Roads Congress. The study finding suggests that the estimation of the actual LDF from field study is essential for the estimation of the number of standard axle repetitions on multilane roads for the design of flexible pavement.

The main objective of this study is to investigate the effect of moisture damage on the fatigue and cracking behaviors of warm mix asphalt (WMA) containing reclaimed asphalt pavement (RAP). To achieve this aim, different fatigue and cracking tests such as indirect tensile fatigue failure, four point beam fatigue, and semicircular bending tests have been conducted, and several parameters have been obtained. These parameters include tensile strength ratio (TSR) and fracture energy ratio (FER), indirect tensile fatigue ratio (IDTR), peak load ratio (PLR), J-integral ratio, and flexibility index (FI), and ${\mathrm{N}}_{\mathrm{f}}$ and flexural stiffness ratios. The results showed that moisture-damaged RAP-WMA demonstrated slightly weaker fatigue and cracking resistance than other mixtures, which is not significant considering the environmental advantages of the mixture. Hence, RAP-WMA would be a proper and beneficial alternative as pavement material where there are significant problems in supplying natural aggregates, and issues like moisture damage and being environment friendly are big concerns. Also, it has been concluded that the fatigue test condition (stress and strain control modes) had a significant effect on the moisture damage investigation.

To improve the efficiency and reliability of intelligent compaction technology, there is a need to evaluate the influencing factors during the compaction process. However, available research is limited, especially in terms of studies based on the dynamic theory. In this paper, the intelligent compaction process is divided into two stages: the compaction process stage, and the compaction completion stage. By establishing dynamic models for different compaction stages, the existing intelligent compaction dynamic model is modified to achieve refined modeling of the intelligent compaction process. The influencing weight of compacting parameters on intelligent compaction quality was analyzed by designing controlled tests. From the results of this study, it was found that roller weight affects the static compaction depth, whereas vibration force amplitude and frequency dominantly affect the amplitude of displacement period changes during compaction but are not predominant influences of the static compaction depth. The influence proportions of roller weight, vibration force amplitude, and vibration frequency on the final compaction quality were 82.62%, 4.92%, and 12.46%, respectively. In addition, a larger-tonnage vibratory roller and a larger vibration model with low frequency and high vibration force amplitude are suggested for compaction roadworks if rapid subgrade compaction is required. The research in this paper provides theoretical support for intelligent compaction technology so that intelligent compaction will become more applicable in practical engineering.

Analysis of overloaded axles has received scant attention during the pavement design process. Such overloaded axles are well known to cause extensive damage to the pavement, though the number of repetitions of such axles is considerably small. This investigation proposes using extreme value distributions to describe the overloaded axles from the axle load data collected from the two National Highways in India. A detailed statistical analysis is carried out under different cases. In the first case, a mixed-normal distribution function is used to describe the entire axle load spectra without explicitly considering the overloading. Different threshold values are chosen for the second and the third cases to consider an appropriate extreme value distribution for analyzing the overloaded data. The right tail region of the axle load data related to overloading is described using Pareto, Burr, and Gamma distributions, while the rest of the data is analyzed using a mixed-normal distribution. The design traffic is estimated for all three cases by calculating the load spectra factor, which uses the moment statistics of the axle load distribution. A significant increase (20%–50%) in the design traffic is observed for the second and third cases.

Superload (SL) vehicles have unique axle configurations and high axle weights, indicating the potential for a low number of SL applications to cause significant fatigue damage to a jointed plain concrete pavement (JPCP). The current JPCP fatigue model in the AASHTO Pavement Mechanistic-Empirical (ME) Design Guide is unable to account for damage caused by SLs because the stress prediction models within it do not consider these unique axle configurations. As a result, it is not possible to account for the potential fatigue damage accumulation caused by SL applications or identify critical conditions that contribute to damage accumulation using Pavement ME. To address this, a critical, SL stress-prediction model was developed that can be used along with the current fatigue damage model. Typical SL configurations in Pennsylvania were identified based on available permit data, and a database of critical tensile stresses generated by these SLs for various JPCP structures was developed using finite element analysis. This database was used to train a series of artificial neural networks (ANNs), which predict critical tensile stress as a function of SL axle configuration, temperature gradient, and pavement structure. The method for determining fatigue damage accumulation using the ANNs to predict stresses is then presented.

An increasing number of field investigations have revealed that the transverse cracks in semirigid asphalt pavement are more likely to initiate in the top of the asphalt layer rather than in the bottom of the asphalt layer, which conflicts with the widely accepted view that bottom-initiated reflective cracking is the dominant distress. To analyze the initiation cause of this kind of surface crack, a numerical simulation model coupled with multiphysics fields was built. It was based on the viscoelastic-plastic damage model, using the finite-element software Abaqus to calculate the damage distribution coupled with the temperature field as well as the loading field. The simulation results show that asphalt layers have two main damage regions in the semirigid base crack section, located in the top asphalt layer and the bottom asphalt layer. The increase of asphalt layer thickness and the decrease of heavy loads are the most possible causes of surface-initiated cracks replacing reflective cracks in semirigid pavements in recent years in China.

The bonding performance between sulfur-based polymers and silicate cement mortars was evaluated in this study. A sulfur-based polymer with high strength and rapid setting properties was developed using fly ash, sand, sulfur, and rubber, without water. The correlation of sulfur and rubber contents with flexural bonding strength, as well as the correlation of the polymerization process and pore distribution with flexural bonding strength, were examined. Results revealed that a certain threshold value exists for the effect of porosity on the flexural bonding strength, and the sulfur-based polymer shows improved adhesion performance when the pore-size distribution is roughly equivalent. The incorporation of rubber particles can significantly improve the matrix compactness and increase the flexural bonding strength of sulfur-based polymers. The flexural bonding of sulfur-based polymers tended to initially increase and then decrease at the sulfur contents of 20%, 25%, 30%, 35%, and 40%, and rubber contents of 1%, 2%, 3%, 4%, and 5%. The optimal flexural bonding strength of 6.29 MPa was achieved at 30% sulfur and 2% rubber, which was close to the original flexural strength of silicate cement mortar, and the compressive strength reached 36.7 MPa. Hence, the proposed sulfur-based polymer meets some basic mechanical property requirements of road repair materials.

Abstract

Superpave balanced mix design (BMD) approaches have been adopted using the Illinois Flexibility Index test (I-FIT) for cracking and Hamburg wheel tracking test (HWTT) for rutting. The objective of this study was to evaluate the impact of the asphalt concrete (AC) properties on I-FIT’s flexibility index and HWTT’s rut depth. The study was intended to determine the most important parameters the influence the prediction of flexibility index and rut depth. An extensive database of I-FIT and HWTT results was collected from the Illinois Department of Transportation. A total of 18,594 I-FIT data sets were collected from 2061 mix designs. For HWTT, 8,263 data sets were collected from 3,782 mix designs. Data exploration analysis was conducted to evaluate the impact of the AC properties on the I-FIT and the HWTT results. Finally, feature ranking analysis was performed to determine the properties that significantly influence the flexibility index and rut depth. The result indicates that most of the AC properties identified in the database had an impact on flexibility index and rut depth. To rank the influential parameters, a random forest regression model was developed to execute recursive feature elimination analysis. The parameters $G_{mb}$, recycled content, air voids, and asphalt binder replacement were the most impactful on flexibility index and rut depth results.

Practical Applications

In this study, a large database of Illinois Flexibility Index and Hamburg wheel tracking data was collected. Then an evaluation of the impact of each AC property on the flexibility index and rut depth was conducted. Finally, a feature ranking analysis was conducted to identify the properties that impacted the flexibility index and rut depth more significantly. Contractors and AC designers may be challenged to adjust an AC mix design to control potential AC cracking or rutting. The results of this research identified the critical properties to predict flexibility index and rut depth. The outcome would allow the AC mix designer to achieve the required thresholds for cracking and rutting potential effectively by controlling the mix design parameters.