The asphalt mixture, AC-13C, is aged in an oven for five different periods, and then, stress relaxation tests are conducted at -5℃ on the aged samples to study the effect of aging on the relaxation of the asphalt mixtures and to build the relaxation modulus of the asphalt mixture coupled with the aging factor. The relaxation modulus curves of the asphalt mixtures with different aged degrees show that there exists equivalence between the ageing degree and relaxation time. The shift factor formula between the aging degree and the relaxation time is obtained based on free volume theory of the material viscosity through hypothesizing that there exists a linear relationship between the fractional free volume and the aging degree. The relaxation modulus master curve is nonlinear, fitted with Burgers model and general Maxwell models with different element numbers. The general Maxwell model was superior to the Burgers model in expressing the obtained relaxation modulus master curve; the more the elements are, the higher the fitting precision is. The general Maxwell model with six elements is recommended and the relaxation modulus of asphalt mixture coupled with aging factor is built according to the fitting precision and elements.

In this study, a hard-grade paving asphalt with a penetration of 10/20 (0.1 mm) and a limestone aggregate were used to produce a hard-grade asphalt mixture AC-20 (HaGAM). Basic performance tests, such as a wheel rutting test at 60℃, a three-point beam bending test at -10℃, and an indirect tensile strength test during freezing-thawing cycles, were conducted to evaluate the conventional properties of HaGAM. The viscoelastic property of HaGAM was investigated by conducting a uniaxial compression test, through which the resilient moduli and complex moduli were determined under variable conditions. The master curve of the complex moduli of HaGAM was constructed using a sigmoid model. The road performance of HaGAM is comparable to that of a styrene-butadiene-styrene (SBS)-modified asphalt mixture based on conventional properties. However, HaGAM shows better performance than the SBS-modified mixture in terms of resilient and complex moduli. Therefore, hard-grade asphalt could be used to resist permanent deformation and ultimately improve the long-term performance of asphalt pavements.

The processing method of asphalt mixture digital images was investigated on the basis of the gray-level distribution of asphalt concrete images. Some bond areas were found between different aggregates. Watershed segmentation method was used to separate aggregates, the results of which were verified. Aggregate orientations in binary images were analyzed statistically. The correlation between aggregate orientation and aspect ratio was determined. Results showed that the watershed segmentation method can separate aggregates, and that aggregate orientations are affected by the aspect ratio. A more oblate aggregate corresponds to a more concentrated placed orientation.

A double-layer structure model of pavements that considered interlayer contact status was established to manage the dowel-bar position deviation problem in rigid pavements. The deviation effect of three-dimensional positions, such as horizontal angle, vertical angle, and embedded depth, on joint load-transfer capacity was analyzed. A load-transfer capacity prediction model that considered dowel bar position deviation was established via ternary nonlinear regression. Load correction factor and its range were also proposed. This prediction model can effectively reflect the joint load-transfer capacity during dowel position deviation after verification via falling weight deflectometer testing. The horizontal angle of the dowel bar minimally affected joint load-transfer coefficient. By contrast, the joint load-transfer coefficient decreased almost linearly as the vertical angle increased. The coefficient reduced by approximately 12% when the vertical angle was 15°. Meanwhile, the load-transfer coefficient was maximized when a dowel bar was embedded in the middle of a surface. The coefficient would decline either upward or downward. The coefficient particularly decreased by 10% when the position was 2 cm downward.

The method that uses only a compaction degree indicator in the core specimens of asphalt pavement for evaluating construction quality cannot effectively reflect the volumetric property and inner structure of asphalt mixtures. Virtually reconstructed specimens are generated by cone-beam scanning field cores with a high-accuracy industrial X-ray computed tomography, and the vertical slice images and digital processing of the aggregate are adopted. To evaluate the vertical structural segregation of the field core specimens, indicators such as the non-uniformity coefficient of coarse aggregate and the vertical variation coefficient of specimens are presented. Results indicate that (1) in various construction control situations, the inner structures of asphalt pavement with the same graded mixture have obvious differences; (2) coarse aggregate particles in the specimens show the regularity of homogeneity in the middle and segregation at both bottoms; and (3) pavement specimens have a gentle homogeneity curve and a small vertical variation coefficient in a road section with superior construction control.

The effects of sample size, aggregate lithology, and compaction pattern on image scanning quality were determined using 225 kV industrial computer tomography equipment. Moreover, the scanning parameters of common Marshall asphalt mixture specimens were optimized. Results show that: (1) A large sample worsens image quality; (2) Granite mixtures are more suitable for subsequent image processing analysis than other counterparts are; and (3) Compaction patterns can affect imaging quality significantly. Finally, the ideal combination of scanning parameters was determined for Marshall asphalt mixture specimens. This combination can improve scanning accuracy and efficiency.

The sand replacement rate and fineness modulus of rubber particles are considered as variables in this study. The effects of the fine aggregate content and fineness of the rubber particles on the impermeability and frost salt resistance of road cement concrete are investigated through a two-factor and four-level orthogonal test. This study shows that the rubber particles used in the concrete effectively improve the resistance of the concrete to chloride ion permeability and frost salt resistance. The replacement rate and fineness modulus of the rubber particles significantly influence the resistance to chloride ion permeability of the concrete. The replace rate of rubber particles has a particularly significant influence and fineness modulus has a significant influence on frost salt resistance of concrete. When adding the fineness modulus whose fineness modulus is 1.38 and the 40% dosage of the fine aggregate (volume ratio) into concrete, the electric flux for the concrete to chloride ion penetration resistance is 527 C, and the salt resistant level is F500.

To investigate the laws of creep and shrinkage in high-strength concrete for long-span pre-stressed continuous box girder bridge and predict long-term deflection, strain sensors were embedded in the midspan of a box girder bridge. The shrinkage strain of concrete was then directly measured, and the creep strain was obtained using the incremental algorithm of measured data. The measured data show that if the creep and shrinkage model in the existing specification is used, then the shrinkage effect is underestimated as a whole and the creep effect is overrated for the C55 high-strength concrete of the bridge. The long-term deflection is predicted using the modified prediction models of creep and shrinkage that are based on short-term measured data when the bridge is under dead load. The deflection prediction values of two modified prediction models were compared with the measured deflection. Results of the comparison indicate that the integrality of the short-term measured strain data in concrete, measurement accuracy, and modified prediction model that is based on the measured data influence the prediction accuracy of the modified prediction models.

Based on the thermodynamic equilibrium principle and elastic mechanics theory, the mechanical calculation of borehole wall stress under ice-formation pressure effect is carried out for the water saturation condition of cement mortar closed pore. The pore configuration cement mortar can be approximated as spherical or cylindrical. A micro-mechanics calculation model is established. The borehole wall inner and outer surface tensile stress is deduced in different pores under the ice-formation pressure effect. For the cement mortar component, the borehole wall stress of spherical and cylindrical pores is calculated at -20℃. Results show that the spherical pore is more constant than the cylindrical pore under the same ice-formation pressure effect. The borehole wall tensile stress of the spherical pore is much smaller than that of the cylindrical one, which is advantageous in frost resistant. Finally, according to the theoretic calculation method, the pore configuration, radius, borehole thickness and freezing temperature are the main factors in borehole wall tensile stress, as shown by the parameter analysis. Trying to useful practice in freezing and thawing destruction mechanisms of concrete in cement mortar, it can contribute to concrete durability design theories in bridge engineering.

The critical moisture content of concrete with different fibers and the quality loss of different moisture contents are analyzed by simulating the real heating curve of the building material in a fire, and a fire test on the concrete slab is conducted. Results show that (1) moisture content greatly influences the surface burst degree, number of cracks, time of water seepage, and quality loss of the specimen after a fire; (2) the critical moisture content of fireproof fiber concrete burst is higher than that of three other kinds of concrete (normal concrete, polyethylene fiber concrete, and polypropylene fiber concrete), and has the highest crack resistance; and (3) the observation of fire spalling and the burn crisp degree of the concrete slab with the same moisture content verified that the fireproof fiber has a better anti-burst effect. Testing the post-fire side face of concrete slabs and the regional rebound value show that rebound value losses are not the same because of different temperature distributions on the side faces of the concrete slabs. The loss of mechanical property of the concrete is related to the temperature field.

A new method to apply traffic tunnel with large deformation of soft rock is produced based on the research and development of a new yielding anchor, and on the successful application of large deformation tunnel to a coal mine system. Moreover, this new method should adopt the basic principle of "timely support, strong support, and yieldable support" to maintain a certain supporting resistance to preserve the stability of surrounding rocks and somehow allow displacement for the support system. Therefore, the support system functions include supporting, yielding, and unloading. A yielding support system can then be designed according to the basic principle, comprising parts such as yielding anchor, sprayed concrete, pressure controller, and arch frame. Combinations of the yielding system components and key technical problems for future resolution are proposed by discussing the basic yielding anchor mechanism, and analyzing the components of the system and these problems. The main technical problems include deformation coordination and yielding support system stability; determining the relationship and amount of free surrounding rock deformation; surrounding rock and support joint deformations, and constant resistance yielding; setting the initial yielding load scientifically; and maintenance of the constant resistance characteristic of the components. Meanwhile, field test research on the yielding support system should be combined with a suitable project background to further verify the suitability and reliability of the system.

This research studies ways to coordinate timetables in a regional bus transit with multiple transport modes to satisfy numerous realistic constraints, such as maximum and minimum departure intervals, by considering transfer among buses, subways, and passenger special lines with intersecting routes. The primary objective is to minimize the total waiting time for non-transferring and transferring passengers in all stops, while the secondary objective is to maximize the number of berths for all vehicles arriving at stations during a certain period. The constraint method converts the problem into a single objective programming problem. Based on the characteristic of the model, this study proposes an improved bacterial foraging optimization (BFO) design, which defines solution coding, redesigns the heuristic procedure to initialize chromosomes randomly, and uses the "ladder" concept to enhance bacterial foraging operation, to resolve the problem. Finally, a numerical example is provided to reveal the differences in schedule coordination between single and multiple transport modes as well as to analyze the influence of the best schemes of station capacity. Furthermore, the improved BFO is compared with other intelligent algorithms to verify the model as well as the accuracy and effectiveness of the algorithm.

A utility function of driving behavior for non-motor and motor vehicles is built based on dynamic reduplicate game theory to study the decision-making behavior of a driver for non-motor and motor vehicles in a mixed-traffic environment. This utility is built according to a driver's personality that affects decision-making behavior, time refinement, and time difference to a conflict point between non-motor vehicles and motor vehicles in each period. Analyzing the utility function of driving behavior for non-motor and motor vehicles, shows that a Nash equilibrium exists in the game process, and the optimal decision-making for non-motor and motor vehicles in a dynamic game is obtained. The illustration shows that during decision-making (1) impulsive drivers choose the acceleration strategy; (2) mild drivers prefer the uniform or deceleration strategy; and (3) cautious drivers choose the deceleration strategy.

For the optimization of city logistics distribution under time restriction caused by traffic organization, a bi-level programming model considering the distribution cost, environment cost, and distribution time of enterprises is developed. The primary function of this upper-level model is to minimize the general cost of distribution enterprises and determine the scheme to optimize city logistics distribution. The lower-level model is the urban traffic equilibrium distribution model for analyzing the distribution route determined by choice behavior for the optimization scheme of city logistics distribution given by the upper-level model. Results indicate that the distribution time, distribution cost, and environment cost of distribution enterprises increased significantly, and that the income of distribution enterprises was reduced because of the traffic organization time window in the process of city logistics distribution. When a distribution task was required to be completed within the formulary time window, the distribution time increased from 5.91% to 7.32%, the distribution cost increased from 7.6% to 13.5%, and the environment cost increased slightly. During the traffic organization period, the distribution cost in the second time-restricted scene increased to 26.85% because of the two-fold impacts of the average freight charge and the traffic demand. When the change in network traffic demand was equal to the basic scene, the environment cost and the distribution time in the first time-restricted scene increased slightly. However, when comparing the restriction time under traffic organization and the total distribution time, the incremental quantities of the former distribution time and distribution cost were always more than the latter.

This study presents a multi-objective optimal design of automobile suspension systems to improve vehicle ride comfort and reduce tire-induced dynamic excitations on road surface simultaneously. In the optimal model, spring stiffness and the damper coefficient are considered design variables, whereas the maximum deflection of the suspension system is regarded as a constraint. Meanwhile, the root-mean-square values of the vertical acceleration of the vehicle body and the dynamic loadings of the front and rear tires are treated as objective functions. Multi-objective optimization is implemented using the gray particle swarm algorithm. Globally optimal solutions are obtained by introducing the variance of relevant sequence numbers into gray relevant theory. A half-car model is used to illustrate the proposed optimal model and solution method. Results show that the minimum acceleration of the vehicle body and the minimum dynamic loads exerted by tires on road surfaces can be achieved through the proposed multi-objective optimal design.