Improved Load Distribution for Load Rating of Low-Fill Box Structures

Abstract

Reinforced concrete box culverts are mostly used at shallow depths. Periodic evaluation of their load carrying capacities is required for load rating of the culvert by determining a rating factor (RF) or truck tonnage of a HS truck. The rating factor is defined as the capacity of the structure minus the dead load demand, and then divided by the live load demand. All the state DOTs are required to inspect and assess culvert conditions and capacities by load rating in every two years. The distribution of live loads on the top slab of a box culvert plays a major role in determining the rating factor of the culvert. The current AASHTO guidelines do not consider the effects of pavements present above the fill while determining the load distribution. The distribution of the wheel load through a pavement may be different from that suggested by the current AASHTO guidelines. In addition to the pavement effect, the fill conditions (i.e., fill thickness and fill modulus) may affect the load distribution. Currently, there is lack of a design method to address the load distribution when a pavement is present above the fill. In this research, two field tests were carried out on the culverts under rigid and flexible pavements respectively. The finite difference numerical models of the test culverts were created in the Fast Lagrangian Analysis of Continua in 3 Dimensions (FLAC3D) software and were verified against the field test results. The verified finite difference models of the culverts were used for a parametric study to analyze the effects of pavement type (i.e., flexible and rigid pavement), pavement thickness, fill depth, and culvert span on the pressure distribution. The material properties and boundary conditions used in the models for the parametric study were similar to the verified models. The parametric study demonstrated that the intensity of vertical pressure on the top slab of the culvert gradually decreased as the pavement thickness increased. The vertical pressure under a rigid pavement was lower than that under a flexible pavement at the same pavement thickness. Within the range of the fill depth covered in this study, the intensity of vertical pressure decreased gradually with an increase of the fill depth over the culvert. The effect of the traffic load on the vertical pressure on the culvert was more significant at the lower fill depth and gradually decreased with the increase of the fill depth. The distribution of the vertical pressure at the fill depth of 0.6 m was characterized by a peak pressure under the wheel load and the peak pressure shifted to the middle of the axle at the fill depth of 1.2 m and greater. There was little interaction between the wheels at the fill depth of 0.6 m and, the interaction between the wheel loads was fully developed for the fill depth of 1.2 m and greater. The calculated vertical pressure decreased by increasing the culvert span from 1.8 to 5.4 m when the top slab thickness of the culvert was the same. However, when the top slab thickness increased, the vertical pressure at the larger span was close to that at the small span. The effect of the culvert span on the vertical pressure was negligible if the thickness of the top slab was properly designed. The maximum vertical pressure obtained from the numerical analysis was compared with those by the distribution formulae in the AASHTO guidelines. The comparison showed that the current AASHTO guidelines over-estimated the pressure for low-fill culverts under a pavement. Simplified methods were developed in this study to estimate the vertical pressures under rigid and flexible pavements. In the case of the culvert under a rigid pavement, the total wheel load is distributed uniformly over the combined area with the extremities of the distributed areas by individual wheels. In the case of the culvert under a flexible pavement, the vertical pressure within the overlapped area is obtained by the superposition of the pressures due to individual wheels. The calculated vertical pressures by the simplified methods were in good agreement with the maximum vertical pressures obtained by the numerical method.