A Study Of Ground Penetration Radar Methods And Techniques In Concrete Detection

As the rapid development of economy in our contrary, engineering construction has also been an unprecedented development. At construction, concrete is a common and important material. So the quality of the concrete structure is directly related to the quality of the projects and construction safety. Concrete structures usually exist common quality problems: concrete thickness does not satisfy the requirements; structural damage and cracks, interlaminar cavity and internal support frame(reinforced) uneven distribution and so on. In order to guarantee project quality and safety of concrete structure, the concrete structure required for testing and monitoring. traditional detection and monitor methods, such as cutting ring method, core drilling method, etc. are need to sample to obtain the internal defects or in the room to get relative parameters such as: the thickness of concrete layer, degree of compaction, water content, compressive strength, etc. The disadvantage of these conventional methods are distance sampling, representativeness bad , detection is destructive , high costs and long period, etc., so a better detection methods is needed. Ground-penetrating radar(GPR) technology is a nondestructive detection technique, its advantages are detection rapidly and continuous, easy to operate, high precision, etc., so GPR technology is widely applied, and play an important role in concrete quality inspection.The content of this paper focuses on nondestructive detection of concrete structure using GPR. Research can be divided into the following parts: using high-order finite difference time domain method to simulate numerical model of GPR; using GPR to detect random medium of concrete body; detecting rough interface of concrete body and analyze its influence to the targets; mesh reinforcement and defect's detection in reinforced concrete; concrete parameter's detection and quality evaluation using GPR. The main results of study are as follows:⑴Using high-order finite difference time domain(FDTD) method to simulate concrete body's ground-penetrating radar detection. Using Taylor series expansion formula to get high-order difference formula and put it to Maxwell's equations of the two curl equations to replace the differential form in order to get high-order finite difference equation. This method solved the problem of large numerical dispersion caused by the second order accuracy of differential approximation. In order to solve the electromagnetic reflection in truncation surface, using UPML absorbing boundary. We use Taylor series expansion method to make UPML absorbing boundary condition can be used for high-order FDTD method. In this paper, we give in detail to the process.⑵Media non-uniformity affects signal quality of GPR, and affects the radar data processing and interpretation. Through higher-order FDTD methods to simulate what happened to GPR wave when concrete body's dielectric constant and conductivity changed with non-uniform random change, and analyze its affects to target detection. Through the study we found: when the dielectric permittivity of inhomogeneous medium changed larger, the radar wave propagating in concrete body will occur stronger scattering and diffraction, which affect the single noise ratio(SNR) of target's reflection signal. Medium changed inhomogeneous also affect the propagation path and the propagation velocity of electromagnetic waves, result in the evaluation error of travel time to the reflection signal. When conductivity of inhomogeneous medium happened change, form snapshot of radar wave we can not find electromagnetic wave scattering and diffraction, but its absorptive capacity get stronger, and also result in reflected signals get weaker, which reduces the signal noise ratio too.⑶When we use GPR to detect concrete body, targets between surfaces or near the surfaces are often our aim, but the surface in the concrete mostly fluctuant and rough, which causes radar wave diffraction and scattering. this phenomenon not only serious affects us to determine the surface's location, but also affect their bottom targets detection. Using a number of uniformly distributed random numbers to structure a random number sequences which content the Gaussian distribution. And using filter method to make the random number sequences accord with realistic random rough surface, finally using this number sequences to structure numerical model of concrete body which surface is rough and its roughness can be changed in accordance with practical requirements. By numerical simulation we can find that with roughness of the rough surface increase, the electromagnetic wave in the rough surface happened diffraction and scattering stronger, which affect the bottom targets detection. Especially for circular targets near the rough surface, its influence is greater than layered targets. When the rough interface changed to a certain roughness, the interface's roughness will become to a main influencing factor, the other objectives can not be detected and position.⑷To reinforced concrete structure, the electromagnetic wave was shielded by mesh reinforcement and the strong multiple waves existed, which increase the detection difficulty of GPR to defect bodies under the mesh reinforcement. By GPR numerical simulation about concrete models of different density mesh reinforcement, and analyses and compare the reflect signals of mesh reinforcement and targets about different models, we can find that if the upper mesh reinforcement's density get smaller, defect bodies can easy to detect. When the mesh reinforcement density is the same, defect bodies'reflective ability will determine detection ability of GPR. In order to achieve the detection of reinforces and defects in reinforced concrete, we use F-K migration technology to deal with radar data. The results showed that: this method can very well make diffraction wave's energy located to the vertex, vertex energy is much higher than the energy of multiple wave and background signal, which can effective extrude the defects.⑸To set up an asphalt dam model and numerical simulation, then analyze these radar data. Reference to the numerical simulation results, extract the effective signal of concrete body such as amplitude, main frequency, etc. from a radar data which was detected form a actual dam (the dam of Inner Mongolia Chifeng), and combined to the porosity and density data which got form samples, we can acquire the relationship between the magnetic parameters of radar wave such as amplitude,main frequency,transient spectroscopy and the asphalt concrete parameters of the porosity and the density. From the fitting results, we can find the correlation of radar signal about amplitude and main frequency with concrete density and porosity are good, which can effectively reflect asphalt concrete's weak change about porosity and density. Through verification about the fitting results we can find that these fitting formulas can effective detect concrete structures'parameters by GPR.⑹Under the concrete material composition and properties, we establish some different water saturation of porosity asphalt concrete numerical models, through numerical simulation study of GPR and analyze the correlation between water saturation and the response of radar waves, we can find that along with the increase about water content the porosity will cause more scattering and diffraction and the absorption capacity to high-frequency radar wave also get stronger. At the same time, the detected radar signal cause more disturbances in whole region. However, the amplitude (energy) of reflected wave significantly reduces too.⑺By analyses the strength of concrete pavement, we can find that the strength of concrete have close connection between radar wave amplitude, frequency and energy. We can get linear fitting formula about radar wave's amplitude, main frequency, energy between the strength of concrete pavement by linear regression calculation method. Using this relationship formula, we calculated the radar data about yibo highway, in order to verify the results of our calculation, combined with the data of samples. By analyzing, we got the following conclusion that using this relationship formula to calculate the concrete pavement's strength by radar data, the largest relative error is only 8%, by this method we can get accuracy results about the strength of concrete pavement.