Nowadays, a large number of high-speed railway trains have become an inevitable trend andreal needs of Chinas railway system.Ballastless railway tracks, because of its high orbitsmoothness, long maintenance of track geometry and less maintenance work, get the favor ofmany countries.With the extension of operating time, increase in structural cracking,penetration of environmental pollution and erosion of surrounding environment, theballastless railway track may be at risk caused by unexpected accidents at any time.Theindicators requirement of project quality and safety will be higher, and the line environmentof elapsed area will be more complex, therefore the damage of the ballastless tracks must becarefully considered.To keep a high level of structural safety, durability and performance ofthe infrastructure in each country, an efficient system for early and regular structuralassessment is urgently required.Non-destructive testing (NDT) methods have a largepotential to be part of such a system.Ultrasonic testing (UT) is a non-destructive testingtechnique based on the propagation of ultrasonic waves in the object or material tested.　　The delamination in interface of ballastless tracks is one of the main defects in ballastlesstracks.Until now, there is no research concentrating on the Non-destructive testing methodin the interface of ballastless tracks.For quality assurance and the detection and localizationof possible delamination, ultrasonic echo methods are principally well suited, because thereflection at an air inclusion leads to a total reflection of ultrasonic pulses, which allows agood from a poor bonding to be clearly distinguished.　　For plate ballastless track systems, the interface of the track plate and the CA mortar layeris an interface of two different materials.The damage in the interface of the track plate andthe CA mortar layer belongs to the bonding damage of two materials.And the difference ofshear wave velocity between the track plate and the CA mortar layer is similar to thedifference of shear wave velocity between concrete and polyamide.In order to simplify thesample, the existence of rebars has been ignored.Thus, a three layer model with twoconcrete layers and one polyamide layer between them is made to simulate the realballastless track.　　In this thesis, both a low-frequency flaw detector A 1220 MONOLITH and an ultrasonictomograph A1040 MIRA have been used for data acquisition.The evaluation software usedhas been Inter-SAFT, which is a research software from Kassel University.A real Boeglsystem ballastless track at BAM provided by the German railway company Deutsche Bahn(DB) in the year 2000 has been tested first and used to check the real situation inside thetrack after 15 years.In experiments of small scale, three small samples being models of atwo and a three layer system coupled with a coupling agent transmitting shear waves(glucose) have been made to figure out the characterization of each concrete block, thecoupling material, the two layer model and the three layer model.Due to the differentvelocities in the different layers, a 2D and 3D reconstruction with velocity correction usingInter-SAFT was used to indicate the depth and the size of artificial defects.