Abstract: A specific time reversal imaging method is proposed in this article to detect defects in concrete structures. In order to detect the damage that the scale is the same as the aggregate, a meso-scale concrete model is introduced in this article. As the concrete is a composite material composed by cement, aggregate, water and concrete admixtures, the Monte Carlo random model and the aggregate grading curve of real concrete samples are introduced for designing this finite element model. Then, the damaged model was analyzed by employing a self-adapted time reversed model to achieve the ultrasonic wave field simulation. This imaging method contains two steps:the first is the forward detection. A series of reflected echo signals with damage information are obtained in this section. These received signals can be reversed in Matlab to serve as the incident signals in the next Time Reversal process; the second step is to image the damage location via interfering the wave-fronts actuated by different transducers to illustrate the peaks of waveform amplitudes. By determining the interfering wave peak time, obtaining the original wave field of that moment, we can forming the wavelet transform energy field, and then complete the damage imaging of the concrete model with defect. In the time reversal process, we introduced the equivalent elastic parameters as the same geometric dimension of the original mesosacle concrete model to locate the damage effectively. Compared with the original wave field, the energy field modified by wavelet transform can lower the effect of the environmental noise. Finally, we discussed the equivalent elastic parameters and the damage sizes to verify the robustness of this method which is applicable in monitoring and evaluating the damage in concrete structures.