WO2021179350A1

WO2021179350A1 - Method of damage detection for decks of girder bridges using an actively excited vehicle

Info

Publication number: WO2021179350A1
Application number: PCT/CN2020/080989
Authority: WO
Inventors: Tinghua YI; Jian Zhang; Chunxu QU; Hongnan LI
Original assignee: Dalian University Of Technology
Priority date: 2020-03-13
Filing date: 2020-03-25
Publication date: 2021-09-16
Also published as: CN111366317A; CN111366317B

Abstract

A method of vibration-based damage detection for a deck of a concrete girder bridge is provided. First, a test vehicle is parked on the deck to perform the swept sine excitation, and the excitation frequency bandwidth, period and amplitude during the detection process are determined according to test results. Second, according to the design drawings or on-site measurement, the span of bridge, number of girders and spacing of two adjacent girders are acquired. And then, the detection path of the test vehicle can be determined. Third, the test vehicle moves from left end to right end of the bridge according to the detection path. The acceleration of the vehicle is collected. At last, the acceleration is processed by the damage detection algorithm. The algorithm includes following five steps: the measured acceleration is divided into equal segments according to the excitation period; the auto power spectrum is computed for the segmented acceleration response; the amplitude of each auto power spectrum is sampled at equal intervals to form the vector; the similarity coefficient Q of any two vectors are calculated to form a matrix; visual results of the deck are generated by combining the matrix with location data of the test vehicle.

Description

METHOD OF DAMAGE DETECTION FOR DECKS OF GIRDER BRIDGES USING AN ACTIVELY EXCITED VEHICLE

Technical Field

The present invention belongs to the technical field of damage detection for the bridge, and relates to a method of vibration-based damage detection for the deck of the concrete girder bridge.

Background

Because of material degradation and higher traffic loads, concrete bridge decks may suffer the local damage such as delamination, crack, reinforcement corrosion and spalling. Visual inspection is most extensively used for monitoring the damages in concrete structures. The main disadvantage of visual inspection is that it detects the cracks, deterioration and damage only when it begins to affect the life of the structure or in some cases it has badly affected the internal layers of the structure while only minor cracks appear on the surface. Meanwhile, inspection at bottom of bridge deck on the stream or river is very difficult. Furthermore, non-destructive testing technology can also be used to detect the damage of the bridge deck. For example: Ground penetrating radar (GPR) , Impact echo (IE) , Ultrasonic pulse echo (UPE) , Electrical Resistivity (ER) and so on. For small and regular structures, such as the local damage detection of pressure vessels. Non-destructive testing technology is very effective. However, they are found to be difficult to implement in large civil engineering structures.

In addition to above non-destructive testing methods, the vibration-based damage detection method has already been widely studied for many years. In dynamic test, the natural frequency, mode shapes and modal damping values are measured during testing. The values of these modal parameters change whenever a structure deteriorates. Traditional vibration tests aimed at measuring the modal parameters often require on-site installation of the measurement equipment, which is not only costly, but also inconvenient. Yang et al. got an idea of extracting natural frequencies and mode shapes of bridge structures from the acceleration of a passing vehicle. This method needs only one sensor installed on the vehicle, which greatly facilitates its implementation. Inspired by Yang’s work, Zhang et al. proposed a damage detection approach by joining sinusoidal tapping force on the vehicle passing across the structure. The mode shape square could be extracted from the acceleration response of the vehicle. Then, by comparing the mode shape square before and after damage, the local damage could be located. The above studies are aimed at simply supported beams and focus on the low-order global modes of vibration. Among the small and medium span bridges, a large number of concrete girder bridges were constructed, which the reinforced concrete deck is supported by several concrete girders. The ratio of the width to the length of the bridge is large, and the bridge cannot be regarded as a simply supported beam. Therefore, the above detection methods are not applicable. According to the structural characteristics of the concrete girder bridge, the vibration modes of the concrete girder bridge can be divided into two types: the global modes of vibration and local modes of vibration in the deck slab. In summary, it is important to quickly detect the bridge deck by exciting the local vibration of the bridge deck.

Summary

The objective of the present invention is to provide a new method for deck of girder bridge using an actively excited vehicle, which can solve the problem of quickly locating the damage of the deck during the bridge detection.

The technical solution of the present invention is as follows:

The procedure of damage detection for the bridge deck is proposed. First, the test vehicle is parked on the deck to perform the swept sine excitation, and the excitation frequency bandwidth, period and amplitude during the detection process are determined according to test results. Second, according to the design drawings or on-site measurement, the span of bridge, number of girders and spacing of two adjacent girders are acquired. And then, the detection path of the test vehicle can be determined. Third, the test vehicle moves from the left to the right end of the bridge according to the detection path. The acceleration of the vehicle is collected. At last, the acceleration is processed by the damage detection algorithm. The algorithm includes following five steps: the measured acceleration signal is divided into equal segments according to the excitation period; the auto power spectrum is computed for the segmented acceleration response; the amplitude of each auto power spectrum is sampled at equal intervals to form the vector; the similarity coefficient Q of any two vectors are calculated to form a matrix; visual results of the deck is generated by combining the matrix with location data of test vehicle.

The procedures of the damage detection for the bridge deck are as follows:

Step 1: Set excitation parameters before the detection.

Excitation parameters including frequency bandwidth, period and amplitude of excitation. The test vehicle is parked in the middle of the deck of two adjacent girders, and then the vehicle is excited by the shaker mounted on it. At the same time, the acceleration response of the vehicle is collected. The frequency f _vd of the deck with a test vehicle is obtained from the peak of the auto power spectrum of the acceleration response. The upper limit of the bandwidth be taken as f _vd plus 15%-20%, and the lower limit of the bandwidth be taken as f _vd minus 15%-20%. The distance traveled by the vehicle on the deck during a period of the excitation is considered as a detection segment. The excitation period is taken as 0.2 s-0.6 s. The amplitude of the excitation depends on the range of the excitation force provided by the type of the shaker.

Step 2: Determine the detection path.

According to the design drawings or on-site measurement, the span of bridge, number of girders and spacing of two adjacent girders are acquired. The center line of every two adjacent girders is used as the detection path, and the test vehicle can go back and forth to complete the detection of all paths.

Step 3: Test vehicle moves across the bridge.

The test vehicle travels along the centerline of two adjacent girders from the left to the right end of the bridge according to the detection path set by Step 2. The speed of the vehicle is 0.5 m/s-1.5 m/s. At the same time, the excitation is applied to the vehicle by the shaker according to the parameters set by Step 1. The acceleration response of the vehicle is synchronously collected by the accelerometer is mounted on the test vehicle.

Step 4: Get visual results of the damage detection.

The acceleration response of the vehicle contains damage information of the deck. The response is processed by the damage detection algorithm, and the visual results of the damage detection are shown. The algorithm comprises the following steps.

In the first step, the measured acceleration signal is divided into equal segments and each data segment represents a detection segment. The length of each segment is the same as the period of the excitation during the test to ensure that the excitation force is the same in each segment. For example, the signal in the time domain is divided into n segments.

In the second step, the auto power spectrum is computed for the segmented acceleration response. N number of the auto power spectrum will be obtained by n segments.

In the third step, the amplitude of each auto power spectrum is sampled at equal intervals to form a vector, and a total of n vectors are obtained.

In the fourth step, for any two vectors V _i and V _j, the value of Q is calculated as follow:

The value of Q is used to indicate the degree of similarity between the two vectors. A square matrix of order n is obtained by n vectors, The values in the square matrix represent the differences in the vibration of the bridge deck between any two detection segments.

In the fifth step, the visualization of results is performed. The square matrix obtained in the previous step is integrated with positioning data to generate the contour plot of the scanned deck.

The advantage of the invention is that the vibration of the bridge deck is excited by the actively excited vehicle, and the damage response algorithm is used to process the acceleration response of the vehicle to determine the location of the damage. The structure of the actively excited vehicle is simple, and the acceleration response is convenient to obtain.

Description of Drawings

Figure 1 is the determined detection path.

Figure 2 is the model of the actively excited vehicle.

Figure 3 is the cross section of the bridge.

Figure 4 is the acceleration response measured from the vehicle in the step 1.

Figure 5 is the auto power spectrum of the response measured from the vehicle in the step 1.

Figure 6 is the damage scenarios of delamination and section loss are set in each lane.

Figure 7 is the typical acceleration response measured from the vehicle as the vehicle travels across the bridge.

Figure 8 is the Visual results for delamination (Lane 1, Lane 2) and section loss (Lane 4, Lane 5) (a) plot of Lane 1 (b) plot of Lane 2 (c) plot of Lane 4 (d) plot of Lane 5.

Detailed Description

The present invention is further described below in combination with the technical solution and figures.

This embodiment uses numerical simulations to verify the validity of the method proposed. The bridge has a roadway width of 13.30 m and a bridge deck thickness of 0.20m, as shown in Figure 3. The span of the bridge is 20m. The number of crossbeams is three, and the crossbeams are located at the end and middle of the bridge. The test vehicle is simplified to the spring mass damping system. The mass of the car body is 400 kg, the mass of the wheel is 100 kg, spring stiffness is 1e6 kN/m and damping ratio is 0.02. Damage scenarios of the delamination and section loss are set in each lane, as shown in Figure 7. The procedures are described as follows:

(1) The excitation parameters are set before the detection. The frequency f _vd of the deck with a test vehicle is obtained by swept sine excitation in the step 1 of damage detection. The parameters of excitation are as follows: the total time of excitation is 5s, the frequency bandwidth is 50 Hz-150 Hz, and the amplitude is 0.1 kN.The acceleration response of the vehicle is measured, as shown in Figure 4. Auto power spectrum of the response is given in Figure 5. The frequency corresponding to the highest peak in the spectrum is 99.71 Hz, which is the f _vd. The periodic swept sine excitation is applied to the vehicle to continuously excite the bridge deck during the vehicle traveling across the bridge. The parameters of excitation are as follows: the period of excitation is 0.4 s, the frequency bandwidth is 80 Hz -120 Hz, and the amplitude is 0.1 kN. The speed of the test vehicle is 1 m/s.

(2) The vehicle travels along the centerline of two adjacent girders during the test, which is the direction pointed by the arrow in Figure 1. The detection of all lanes is performed in the order marked in Figure 1.

(3) The test vehicle moves from the left to the right end of the bridge according to the detection path set by Step 2. The acceleration response of the vehicle is shown in Figure 7.

(4) The acceleration response of the vehicle collected in each lane is processed by the damage detection algorithm. Visual results for damage scenarios of delamination and section loss are shown in Figure 8.

For various degrees and forms of the damage, Visual results for damage scenarios appear abnormal at the corresponding position of the damage. As the degree of damage decreases, the degree of abnormality also decreases. Therefore, the proposed invention method can determine the location of the damage.

Claims

A method of damage detection for decks of girder bridges using an actively excited vehicle, wherein, comprising the following steps:

step 1: set excitation parameters before the detection:

excitation parameters including frequency bandwidth, period and amplitude of excitation; a test vehicle is parked in the middle of the deck of two adjacent girders, and then the vehicle is excited by the shaker mounted on it; at the same time, the acceleration response of the vehicle is collected; the frequency f _vd of the deck with a test vehicle is obtained from the peak of the auto power spectrum of the acceleration response; the upper limit of the bandwidth be taken as f _vd plus 15%-20%, and the lower limit of the bandwidth be taken as f _vd minus 15%-20%; the distance traveled by the vehicle on the deck during a period of the excitation is considered as a detection segment; the excitation period is taken as 0.2s-0.6s; the amplitude of the excitation depends on the range of the excitation force provided by the type of the exciter;

step 2: determine the detection path:

according to the design drawings or on-site measurement, the span of bridge, number of girders and spacing of two adjacent girders are acquired; the center line of every two adjacent girders is used as the detection path, and the test vehicle can go back and forth to complete the detection of all paths;

step 3: test vehicle moves across the bridge:

the test vehicle travels along the centerline of two adjacent girders from the left to the right end of the bridge according to the detection path set by step 2; the speed of the vehicle is 0.5m/s-1.5m/s; at the same time, the excitation is applied to the vehicle by the shaker according to the parameters set by step 1; the acceleration response of the vehicle is synchronously collected by the accelerometer is mounted on the test vehicle;

step 4: get visual results of the damage detection:

the acceleration response of the vehicle contains damage information of the deck; the response is processed by the damage detection algorithm, and the visual results of the damage detection are shown; the algorithm comprises the following steps:

first step, the measured acceleration signal is divided into equal segments and each data segment represents a detection segment; the length of each segment is the same as the period of the excitation during the test to ensure that the excitation force is the same in each segment; the signal in the time domain is divided into n segments;

second step, the auto power spectrum is computed for the segmented acceleration response; N number of the auto power spectrum will be obtained by n segments;

third step, the amplitude of each auto power spectrum is sampled at equal intervals to form a vector, and a total of n vectors are obtained;

fourth step, for any two vectors V _i and V _j, the value of Q is calculated as follow:

the value of Q is used to indicate the degree of similarity between the two vectors; a square matrix of order n is obtained by n vectors, the values in the square matrix represent the differences in the vibration of the bridge deck between any two detection segments;

fifth step, the visualization of results is performed; the square matrix obtained in the previous step is integrated with positioning data to generate the contour plot of the scanned deck.