Evaluation Of Interlayer Bonding Condition And Transverse Crack Properties In Continuously Reinforced Concrete Pavement

With the fast development of transportation industry in China,the need for adequate bearing capacity and long-term performance of the pavement structure is increasing.Continuously reinforced concrete pavement(CRCP)is a type of high performance pavement structure which has a longer service life and smoothness,which has resulted in great expansion space in China.Punchout is one of the major distresses in CRCP and is caused by poor support from the support layers and insufficient load transfer efficiency(LTE)between transverse cracks.Therefore,the evaluation of in-situ CRCP interface bonding condition between the slab-base and the prediction of transverse crack width in CRCP is essential for performance prediction.Through development of analytical and finite element models,constructing continuously reinforced concrete(CRC)beam test section,and conducting in-situ FWD tests,this study evaluates the in-situ CRCP-baser bonding condition and theoretically predicts transverse crack spacing and crack width.The limitation of traditional concrete pavement backcalculation method in CRCP was analyzed.Since the traditional backcalculation assumption of an infinite slab can lead to significant errors in the backcalculated results,a new backcalculation procedure for finite-sized CRCP panel sizes given either center-and edge-loaded case was established.Given simplified friction model and CRCP deflection data,a new procedure for calculating bonding condition between the slab and base was proposed.Furthermore,the backcalculation procedure and evaluation method of interface bonding condition were applied to field CRCP test sections.Lastly,the analytical equations for predicting transverse crack spacing and crack width in CRCP given linear(bond)relationship between steel reinforcement and concrete and bilinear interface relationship between slab and base was proposed.The analytical model was verified with full-scale CRC beam test sections.The main conclusions from the thesis research are as follows:(1)When applying traditional backcalculation method,the error in backcalculated k and E values and heff decreases with the increase of LTE and non-dimensional slab size(L/l).For a 10 in.(25.4cm)CRCP,the traditional backcalculation method produces significant errors given crack spacing being less than 8ft(2.4m)with LTE greater than 80%,and crack spacing being less than 10ft(3.7m)with LTE less than 80%.(2)An improved backcalculation procedure for center-loaded and edge-loaded CRCP was proposed for finite-sized slabs with various slab aspect ratios and L/l.The unique relation of l and AREA as well as ?int*and a/l were developed for multiple CRCP crack spacing.An average LTE of 90%was recommended for CRCP backcalculation with the standard AASHTO sensor configuration of 0,12 in.(30.5cm),24 in.(61.0cm),36 in.(91.4cm)for CRCP crack spacing greater than or equal to 6ft.For crack spacing of 4ft and 5ft,a backcalculation procedure with three sensors(0,12 in,24 in.)was proposed.The proposed backcalculation method for CRCP was applied to a test section on the Illinois Tollway in the United States.(3)By analyzing temperature differential,longitudinal plate offsets and lateral plate offsets,suggestions were made for in-situ FWD tests.When crack spacing is higher than 10ft(3.0m),it is recommended to avoid conducting FWD test given temperature differential being higher than 24?(13.32?)and 16?(8.88?)for center-loaded CRCP and for edge-loaded CRCP,respectively.The theoretical analysis demonstrated the importance of dropping the plate at the center of the CRCP panel given smaller crack spacing(? 8ft(2.4m))in orderto achieve accurate backcalculation results compared to larger crack spacing(12ft(3.7m),15ft(4.6m))for which longitudinal plate offsets by a few feet does not have significant effect on the backcalculated properties.For edge-loaded condition,it is important to drop the plate within 2in.(5.1cm),no more than 4in.(10.2cm)transverse offset from the center of pavement edge during FWD test in order to achieve the accuracy of the backcalculated CRCP properties given any crack spacing.(4)For edge-loaded CRCP,the over-width of the base can be described with an equivalent base thickness.The equivalent base thickness increases as the over-width of the base and transverse crack spacing in CRCP increase.When crack spacing and over-width are 12ft(3.7m)and 3ft(0.9m),respectively,the equivalent base thickness reaches 7.21in.(18.3cm),1.21in.(3.1cm)more than designed base thickness.As a result,higher backcalculated heff can be obtained given the existence of over-width base.(5)The higher the backcalculated heff,the higher the friction factor ?*between slab and base.When friction factor reaches 1.0,backcalculated heff is around 3in.(7.6cm)higher than the designed slab thickness,it is suggested the interlayer bond condition is excellent.When friction factor is higher than 0.6 or less than 0.2,it is suggested there is a good or poor interlayer bond condition,respectively.Cement treated base(CTB)can provide good support to CRCP slab given the bonding condition between slab and base is stable in the long term(more than 20 years).The combination of warm mixed asphalt(WMA)and CTB also has good performance,even though the dynamic modulus of asphalt changes with the temperature causing decreased bonding condition in summer.(6)The prediction of crack spacing in CRC test beams agreed reasonably well with the field measurements,but overpredicted the crack width.Even so,the variation for prediction results of different reinforcement ratio and concrete material in CRC beams was the same with visual survey results.The test beam with higher reinforcement ratio had both lower transverse crack spacing and crack width,and the test beam with fine lightweight aggregate concrete had higher transverse crack spacing and lower transverse crack width.(7)Based on the sensitivity analysis of the analytical model,the coefficient of bond stiffness between reinforcement and concrete and reinforcement modulus have great influence on prediction results of crack spacing and crack width.However,the impact of coefficient of friction between slab and base and modulus of concrete is much smaller.While the linear expansion coefficient and shrinkage has greater effect on crack spacing than crack width.The sensitivity analysis of different input parameters illustrates the importance of determining accurate model inputs to have more confidence in the predicted results.