This paper proposes a novel framework to characterise the morphological pattern of Barely Visible Impact Damage using machine learning. Initially, a sequence of image processing methods is introduced to extract the damage contour, which is then described by a Fourier descriptor-based filter. The uncertainty associated with the damage contour under the same impact energy level is then investigated. A variety of geometric features of the contour are extracted to develop an AI model, which effectively groups the tested 100 samples impacted by 5 different impact energy levels with an accuracy of 96%. Predictive polynomial models are finally established to link the impact energy to the three selected features. It is found that the major axis length of the damage has the best prediction performance, with an R 2 value up to 0.97. Additionally, impact damage caused by low energy exhibits higher uncertainty than that of high energy, indicating lower predictability.

The seismic behaviors of tunnel lining passing through the fault are studied by means of large-scale shaking table tests under the load of Wolong-seismic wave. The model tunnel is installed into a model box with a length of 3.65 m, width of 1.5 m and height of 1.8 m. The sizes of model and prototype tunnel are designed by using the similar principles with a scale factor of 1:30. As the result shows, under the seismic force load, the tunnel lining has a delay of responding the seismic force when it crosses the faults geological belt. The time is 0.085seconds. Model box amplified table acceleration, the farther the distance from shaking table is, the more obvious amplification is. The seismic force load is horizontal, but the resistance applied by the faults geological belt is weak, resulting relative motion, collision, extrusion etc. so, the tunnel lining appears cracks. It is in accord with investigation conclusions of Wenchuan earthquake. Article also aims at some problems in the experiment such as: boundary effect, model box stiffness etc, and gives some suggestions.

Two full scale frames were tested on a shaking table to investigate seismic performance and fracture mechanism of fiber reinforced concrete in contrast to the plain concrete. The information about acceleration response, the maximum strain value as well as the time to reach it, the typical strain - time curves and the crack development of two test frames were presented. Test results indicate that reinforced concrete did not crack during the test; the fiber reinforced concrete could better absorb or consume energy in the process of stress redistribution after peak acceleration; maximum strain and maximum acceleration did not occur at the same time; structure came into being deformation even failure when the seismic energy in the structure gone up to certain extent, and the dynamic failure would be their main failure modes.

Sesmic dynamic response of Fiber Reinforced Concrete tunnel lining is studied in contrast to the plain concrete. Based on similar theory, model test has been carried out through the 5m×5m triaxial shaking table by inputting sesmic wave, then the damage characteristics of tunnel lining is acquired.The test results show that both the plain concrete and fiber concrete is brocken by sesmic load, but fracture form is not the same,the crack on Fiber concrete is narrow and sawtooth , the crack on plain concrete is wide and straight.Fiber concrete lining strain-time curve is sawtooth partly, it’s vibration reponse is a little lagger than that of plain concrete.It indicats that kinematic velocity of concrete granule is decreased and sesmic energe is absorbed by fiber cohesive force,then frequence amplitude can be reduced.So fiber concrete can be proved as fine anti-seismic material.