Research On Skid Resistance Performance Evolution Of Epoxy Asphalt Pavement For Steel Bridge Deck During Full Lifecycle

Good skid resistance performance of pavement is not only an effective guarantee for driving safety,but also an important condition to ensure the accessibility of urban transportation networks.The long-span steel bridge has put forward more stringent requirements for the skid resistance of bridge deck pavement due to its special use conditions and material properties.However,many problems have existed in the current researches,including the uncertainty of the evaluation index in skid resistance,the incompleteness of evolutionary behavior,the research of friction mechanism and the establishment of long-term antiskid evolution model,which need further research.Long term monitoring on long-span steel bridge and accelerated wear test in laboratory were used as research methods in this paper.Then,the skid resistance evolution of epoxy asphalt mixture for steel bridge deck pavement during the whole life cycle was analyzed.And a prediction method of skid resistance based on neural network time series model was proposed.Firstly,starting from the performance of steel bridge deck skid resistance attenuation,the anti-skid demand of the steel bridge pavement was analyzed according to the environment and material characteristics of the long-span steel bridge.According to the two layers of meaning in skid resistance,the influence of various factors on the skid resistance was analyzed respectively.Based on the multi-scale texture of the pavement surface,the evolution behavior mechanism of the anti-skid performance of the steel bridge deck epoxy asphalt pavement was studied.Secondly,a steel bridge in China was selected to track and monitor the changes of the road surface friction coefficient during the service life of epoxy asphalt concrete pavement structure.The temporal and spatial evolution characteristics of skid resistance of epoxy asphalt concrete were analyzed.The reliability of indoor accelerated grinding test was confirmed on the basic of field measured data.Thirdly,the typical pavement structure specimens of the formed steel bridge and the load simulation system(MMLS3)were used for the indoor accelerated wear test.And the structural depth(MTD)and friction coefficient(BPN)were chosen as the anti-sliding performance evaluation indexes.The change laws of surface morphology,friction coefficient and structural depth of pavement surface during wear would be revealed in the process of those tests.Finally,according to the evolution mechanism of the anti-skid performance of the epoxy asphalt pavement of the steel bridge,the nonlinear autoregressive neural network model was applied to the prediction process of the anti-skid performance of the epoxy asphalt pavement of the steel bridge to realize the multi-step prediction of the friction coefficient.The Markov error state transition matrix was introduced into the error correction model to eliminate the influence of historical prediction residual sequence.The results showed that the long-term skid resistance time series prediction model based on the modified Markov error state transfer matrix could have better prediction accuracy and error stability.