Prediction Model Of Punchout For Continuously Reinforced Concrete Pavement

The distress of punchout is the major distress for continuously reinforced concrete pavement(CRCP). Factors that influencing the formulation of punchout are very complex. There is not a widely used punchout prediction model so far, and the design guide of CRCP is still needed to be improved. The top-down longitudinal cracking and the erosion of the base are the two main causes for punchout. The pavement structure, properties of materials, and construction conditions are all affecting the performance of CRCP. The research is focused on the prediction model of punchout, including numerical simulation, field test and analysis of performance monitoring data from long-term pavement performance(LTPP) to illustrate the deficiency of the current design guide for CRCP and research on the prediction model of punchout basing on the fatigue properties was carrided out.Firstly, a numerical simulation program, named RP_TMP, was built, which can deal with the materials’ properties, climate information and construction conditions and is employed to predict the temperature profile within the concrete slab and the temperature profile at the moment of concrete setting.In addtion, field test was performed to investigate the influence of paving conditions on the setting temperature gradient, and the temperature monitoring data was utilized to validate the prediction result of RP_TMP. Using the temperature moment concept, the equivalent linear setting temperature difference was introduced to present the level of the built-in curling. Further, the influence of pavement construction seasons, paving activities starting time and curing methods on the setting temperature gradient was analyzed. The temperature profiles for both the setting and in-service PCC slab are nonlinear, a new curling stress methodology, which can take the nonlinearity of the setting temperature profile into account, was established. As the result of the analysis, paving activities conducted in the summer will increase the curling stress on the top of PCC as well as the probability of experiencing tensile stress, which illustrate the importance of construction season on top-down cracking. It is found from LTPP data that pavement performance corresponded with the curling stress analysis in this research, both the number of transverse cracking and length of longitudinal cracking for the summer construction are much higher than that of low temperature construction, which validates the new model for curling stress determination in this paper.Secondly,it can be concluded that the LTE along the shoulder plays an important role on the mechanical response of CRCP in accordance with the phenomenon that punchout rarely occur for the CRCPs with tie bar PCC shoulder.ISLAB2000 was employed to calculate the top tensile stress under various combinations of pavement structures, crack spacing, load transfer efficiency(LTE) at the transverse cracking and LTE along the shoulder. Artificial neural network(ANN) was trained with the results of ISLAB2000 to predict the tensile stress under wheel load. In addition, sensitivity analysis was carried out for the factors that influencing the tensile stress with perpendicular test. It was found that the tensile stress on top of PCC is more sensitive to the LTE along the shoulder than LTE at the transverse cracking. Base on the Hamburg wheel-tracking device(HWTD) test on the erosion depth of base materials performed in Texas A&M University and the analysis on the LTE and faulting depth, it was found that the load transfer at the joint decreases with the increase of faulting depth. Furthermore, the relationship between the erosion depth and LTE was built, which can be utilized to describe the LTE deterioration and the loss of effective thickness of PCC layer with the pumping of base materials along the longitudinal joint.Finally, it was realized that the hazard rate for punchout is not constant, characteristic of the number of equivalent single axle loads(ESALs) leads to punchout is believedto follow Weibull distribution. Pavement performance monitoring data from LTPP should be taken as grouped data. A methodology of punchout prediction with Weibull distribution and estimation of the parameters which consists the estimation of the equivalent single standard axles, determination of number of samples, building the likelihood equations, estimation of initial values for the scale and shape parameters with least square method and the solution of likelihood equations with Newton iteration method. Number of punchouts monitoring from LTPP was extracted to validate the new prediction model. The parameters from the Weibull distribution was also employed to illustrate the effect of setting temperature gradient and base materials with various erodibility resistance on the number of punchouts.Focusing on the mechanism of punchout formulation, the two major factors, longitudinal cracking and erosion of the base, are investigated to improve the methodologies for the determination of curling stress and wheel load stress. As a result, the prediction model that based on the Weibull distribution was introduced as well.