A real-time automated detection and positioning method for typical damage in bridge bearings

Abstract

The present invention relates to bridge detection and discloses a real-time automated detection and positioning method for typical damage of bridge bearings, comprising: jacking the main beam, arranging an ultrasonic sensor on the upper surface of the bearing, obtaining a first detection signal after sending an initial excitation signal, converting the damage signal intercepted from the first detection signal into a secondary excitation signal using a time reversal method, obtaining a corresponding inverted focus signal, and then calculating a damage characteristic index, distinguishing bearing damage based on the damage characteristic index, and when it is determined that damage has occurred, obtaining a probability value of damage occurring at a point on the upper surface of the bearing by establishing a probabilistic damage model, and imaging the bearing damage based on the probability value. The present invention establishes a bearing damage characteristic index, which does not require an initial excitation signal and has improved resistance to environmental interference; and uses a threshold method to eliminate tail wave data in the detection signal, thereby reducing data processing complexity, eliminating the need to interrupt traffic, and improving detection efficiency while ensuring detection accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of bridge detection, and in particular to a real-time automatic detection and positioning method for typical damage of a bridge support. Background Art

[0002] As a key load-transmitting component of a bridge, bridge bearings are susceptible to the external environment and reciprocating vehicle loads during their service life, making them prone to rubber aging, cracking, and fatigue damage. This in turn reduces the bearing's bearing capacity and shortens its service life. Bearing damage detection methods include direct observation, static response, and dynamic response. The direct observation method relies on manual or camera observation of the bearing's appearance, which is easy to operate. However, the bearing is located between the beam and the pier, in a small space with dim lighting, resulting in poor image quality. The static response and dynamic response methods mainly rely on the overall static and dynamic response characteristics of the bridge to reflect changes in bearing boundary conditions before and after damage, making it difficult to accurately identify and quantify local damage. In addition, the static response method often requires traffic closures, a large number of measurement points, a complex detection process, and high costs. The dynamic response method requires high sensor accuracy, and the simplified response model design deviates from reality, making detection accuracy difficult to guarantee.

[0003] Ultrasonic testing is a dynamic response method that has been applied in the field of bridge inspection. Currently, this method has the following shortcomings in bearing damage detection: (1) To reduce external interference, the inspection process usually requires traffic control, which affects the normal operation of the bridge; (2) The test results are highly dependent on the initial health detection signal, have weak resistance to environmental noise interference, and have large errors; (3) It is difficult to accurately locate complex cracks such as inclined cracks, curved cracks, or non-planar cracks, and its universality is low. Therefore, how to conduct inspections simply and efficiently and identify bearing damage conditions early is crucial to ensuring the safe operation of bridge structures. Summary of the Invention

[0004] In response to the problems existing in the prior art, the present invention provides a real-time automated detection and positioning method for typical damage of service bridge bearings, which has a simple operation process, high detection accuracy, good imaging effect, rapid data processing and can prevent traffic interruption.

[0005] To achieve the above object, the present invention adopts the following technical solutions:

[0006] A real-time automatic detection and positioning method for typical damage of bridge bearings comprises the following steps:

[0007] Step 1: Preparation for testing:

[0008] Use a jack to lift the main beam upward from the support to form a sensor arrangement space between the main beam and the support;

[0009] Step 2: Deploy the sensor network:

[0010] N (N ≥ 12, N = 12 for circular arrangement and N = 16 for square arrangement) ultrasonic transceivers are evenly arranged along the outer edge of the upper surface of the support. A coordinate system Oxy is established with the center of the upper surface of the support as the origin to obtain the coordinate position of each ultrasonic sensor relative to the center of the support.

[0011] Step 3: Obtain ultrasonic inversion focusing signal:

[0012] The initial excitation signal is sent to the support through the i-th ultrasonic sensor , the corresponding first detection signals are received respectively through the other N-1 ultrasonic sensors; i=1, 2, ..., N;

[0013] For the first detection signal received by the jth ultrasonic sensor, the threshold method is used to intercept the first damage signal, and the time reversal method is used to convert the first damage signal into a secondary excitation signal; j = 1, 2, ..., N, i ≠ j;

[0014] The jth ultrasonic sensor sends the secondary excitation signal to the support, and the threshold method is used to intercept the secondary damage signal of the secondary detection signal received by the i-th ultrasonic sensor as the inversion focus signal. ;

[0015] Step 4: Calculate the damage characteristic index:

[0016] According to the initial excitation signal obtained in step 3 and inverted focus signal , calculate the bearing damage characteristic index according to formula (4.1) :

[0017]

[0018] Where, is the initial excitation signal, is the inversion focus signal, t0 is the start time, and t1 is the end time;

[0019] The larger the DI value, the greater the probability of damage. The DI value range is 0~1;

[0020] Step 5: Determination of bearing damage:

[0021] According to the damage characteristic index The value is used to judge whether the bearing is damaged: if there are more than 15% damage characteristic indicators among all damage characteristic indicators If both are greater than 0.4, it is determined that the bearing has typical damage;

[0022] In other cases, it is determined that the bearing does not have typical damage;

[0023] Step 6: Locate typical bearing damage:

[0024] 6.1 Constructing the support probabilistic damage model:

[0025] Assume that the coordinates of any point on the upper surface of the support are , compare the sum of the distances from the point to any two ultrasonic sensors with the distance between the two ultrasonic sensors to obtain the ratio ,according to as well as The support probabilistic damage model is constructed as shown in Equations (6.1) to (6.2):

[0026]

[0027] Where, is the coordinate point The probability value of damage; N is the number of ultrasonic sensors; is the probability distribution function value, indicating the coordinate point The probability distribution of damage at the location; is the distance ratio; It is a control parameter that affects the image resolution and damage imaging range. .

[0028] 6.2 Imaging and positioning of typical bearing damage:

[0029] The damage probability value The values ​​are mapped to pixel values, and a damage probability cloud map of the bridge bearing is established according to the pixel values ​​of all coordinate points on the bearing. The pixel area with a damage probability value greater than or equal to 0.8 in the damage probability cloud map is regarded as the damage area, the coordinates of the damage area are obtained, and the typical damage location of the bearing is completed.

[0030] As a further improvement of the above solution, in step 1, the height of the sensor arrangement space is 5% to 20% of the support height.

[0031] As a further improvement to the above solution, in step 2, the distance between the ultrasonic sensors is 15% to 30% of the characteristic length (side length or diameter) of the support; the distance between each ultrasonic sensor and the outer contour of the support is not less than 5 mm; and the area of ​​the detection area enclosed by all ultrasonic sensors is not less than 80% of the support area.

[0032] As a further improvement of the above solution, in step three, the ultrasonic sensor uses a piezoelectric ceramic ultrasonic probe.

[0033] As a further improvement of the above scheme, in step three, the process of extracting the first damage signal from the first detection signal using the threshold method is as follows: define the starting time of the first detection signal as t0, obtain the absolute amplitude of each peak or trough in the first detection signal, and obtain the maximum absolute amplitude of the first detection signal, set the threshold Y and the cutoff time t1, so that the ratio of the absolute amplitude of the first detection signal after the cutoff time t1 to the maximum absolute amplitude of the first detection signal is less than or equal to the threshold Y, and take the first detection signal within the range of the starting time t0 to the cutoff time t1 as the first damage signal, generally taking Y=4%~8%, preferably 5%.

[0034] As a further improvement of the above scheme, in step three, the ultrasonic initial excitation signal adopts ultrasonic guided waves, and the frequency of the ultrasonic guided waves is selected according to the properties of the bearing material: for steel bearings, the frequency is 150kHz~300kHz, and for rubber bearings, the frequency is 40kHz~100kHz; the ultrasonic guided waves adopt a pulse waveform, and the wavelength and wave velocity of the guided waves are determined according to the numerical equation of the guided waves.

[0035] As a further improvement of the above scheme, for rubber bearings, the initial excitation signal of ultrasonic guided waves uses a narrow-band sinusoidal pulse signal modulated by Gaussian. The guided wave numerical equation is:

[0036]

[0037] Where: A is the voltage signal amplitude, generally 1V; f c is the signal center frequency, which is 40kHz~100kHz for rubber bearings; k is the normalized bandwidth, which is 0.1~0.5 for rubber bearings; s is the attenuation factor, which is 0.75~0.95 for rubber bearings; c is the signal delay parameter, which is 0~0.0005 for rubber bearings; t is the time; is the signal amplitude, ranging from -1 to 1; the selection of specific parameter values ​​needs to be flexibly adjusted according to the pulse width and frequency characteristics;

[0038] For steel supports, the initial excitation signal of ultrasonic guided waves uses a narrow-band sinusoidal signal Lamb wave modulated by a Hanning window. The numerical equation of the guided wave is:

[0039]

[0040] Where: f0 is the center frequency of the signal, which is 150kHz~300kHz for steel supports; n is the number of signal cycles, which is generally 4~6, with 4 being the preferred value.

[0041] As a further improvement of the above scheme, in step five, the typical damage to the bearing is bearing rubber damage or bearing steel structure damage; the bearing rubber damage includes rubber aging wear and cracks in the rubber hard layer; the bearing steel structure damage includes partial corrosion of the bearing steel structure and fatigue crack damage.

[0042] As a further improvement of the above solution, in step six, Calculate according to formula (6.3):

[0043]

[0044] Where, is the ratio of the sum of the distances from any point on the support surface to the two ultrasonic sensors on any propagation path to the distances between the two ultrasonic sensors. ; is the coordinate of the hypothetical injury center; 、 are the coordinates of the i-th and j-th ultrasonic sensors respectively.

[0045] Compared with the prior art, the advantages of the present invention are:

[0046] (a) The present invention provides a real-time automated detection and location method for typical bridge bearing damage. Based on the oblique or complex shape characteristics of typical bridge bearing damage, an ultrasonic sensor network is deployed. After an ultrasonic signal is excited at any sensor location, detection signals of bridge bearing damage are simultaneously obtained at other sensor locations. A threshold method is then used to intercept the corresponding damage signal from the initial detection signal. A time reversal method is used to convert the corresponding damage signal into a secondary excitation signal. The secondary excitation signal is then reversely excited to obtain an inverted focused signal. Based on the inverted focused signal and the initial excitation signal, characteristic indicators of bearing damage along each ultrasonic propagation path are obtained. The bearing damage status is then determined probabilistically. If a bearing is determined to have typical damage, a bearing damage probability cloud map is constructed by building a bearing probabilistic damage model, thereby visualizing the bearing damage. This method reduces the detection dependency on the initial excitation signal compared to conventional ultrasonic testing, resulting in a wider range of applications and greater practicality. Furthermore, damage imaging processing makes the detection results more intuitive, allowing operators to visually determine bearing damage based on the cloud map, reducing visual observation errors and lowering operator requirements. Detection time is shortened, and damage detection accuracy is high.

[0047] (b) The real-time automated typical damage detection and positioning method for bridge bearings provided by the present invention obtains an inverted focusing signal through a time reversal method, and then establishes a bearing damage characteristic index. The bearing damage characteristic index does not require an initial excitation signal and can also enhance the ability of damage detection to resist environmental interference. Under the premise of ensuring the normal passage of the bridge, the main beam is appropriately pushed with a jack, which reduces the impact on traffic, improves the applicability of the method, and avoids various negative effects caused by traffic interruption. The tail wave in the detection signal or the secondary detection signal is screened out through a threshold method, and the first wave and the second arrival wave are used as the basis for data analysis and probabilistic imaging, which reduces the complexity of data processing and improves the efficiency of damage detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 Schematic diagram of the process of the present invention.

[0049] Figure 2 Schematic diagram of ultrasonic guided wave detection.

[0050] Figure 3 Schematic diagram of the structure of the rubber bearing and the spherical steel bearing in the embodiment; among them, Figure (a) is a square rubber bearing, Figure (b) is a circular rubber bearing, and Figure (c) is a spherical steel bearing; 1-upper bearing plate; 2-rubber layer; 201-stiffening plate; 202-intersecting layer; 3-lower bearing plate; 401-flat polytetrafluoroethylene plate; 402-spherical crown steel lining plate; 403-spherical polytetrafluoroethylene plate.

[0051] Figure 4 Schematic diagram of the sensor network layout of the support in the embodiment; wherein, Figure (a) is a schematic diagram of the sensor network layout of the square rubber support, Figure (b) is a schematic diagram of the sensor network layout of the circular rubber support, and Figure (c) is a schematic diagram of the sensor network layout of the spherical steel support.

[0052] Figure 5 Schematic diagram of preset damage arrangement of bearings in the embodiment; wherein, Figure (a) shows rubber aging damage of square rubber bearing, Figure (b) shows rubber aging damage of circular rubber bearing, Figure (c) shows rubber crack damage of square rubber bearing, Figure (d) shows rubber crack damage of circular rubber bearing, and Figure (e) shows crack damage of spherical crown steel bearing.

[0053] Figure 6 Schematic diagram of the finite element model of the bearing of the embodiment; wherein, Figure (a) is the finite element model of the square rubber bearing, Figure (b) is the finite element model of the circular rubber bearing, and Figure (c) is the finite element model of the spherical crown steel bearing.

[0054] Figure 7 Figure 2 is an example of detecting time domain signals; Figure (a) is a Gauss wave, Figure (b) is a Lamb wave, and Figure (c) is a time domain signal comparison example.

[0055] Figure 8 Schematic diagram of the time reversal method.

[0056] Figure 9 Figure 1 is a summary diagram of damage characteristic indicators of the embodiment; among them, Figure (a) is the rubber aging damage characteristic indicator of the square rubber bearing, and Figure (b) is the fatigue crack damage characteristic indicator of the spherical steel bearing.

[0057] Figure 10 Schematic diagram of the probabilistic damage imaging principle.

[0058] Figure 11 Figure 1 is the damage imaging and positioning result diagram of the embodiment; among them, Figure (a) is the rubber aging damage imaging of the square rubber bearing, and Figure (b) is the fatigue crack damage imaging of the spherical steel bearing.

[0059] Figure 12 Schematic diagram of the main beam push-up bearing; among them, Figure (a) is a square rubber bearing, Figure (b) is a circular rubber bearing, and Figure (c) is a spherical crown steel bearing. DETAILED DESCRIPTION

[0060] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0061] See also Figure 1 The present invention discloses a real-time automatic detection and positioning method for typical damage of bridge bearings, and the specific steps are as follows:

[0062] Step 1: Preparation for testing:

[0063] A jack is used to push the main beam upward to form a space for sensor arrangement between the main beam and the support.

[0064] The present invention requires the detection of damage to the bridge supports in service, and existing traffic cannot be interrupted. If one side is not detected and the traffic is interrupted once, it will cause huge economic losses. Figure 12 To mitigate the impact of damage detection on traffic, the present invention utilizes multiple jacks to push the main beam above the piers. Typically, one jack is positioned at each of the four corners of the pier, for a total of four jacks. This push is performed using a diagonal gradient pattern: First, one diagonal jack is pushed a certain distance, then the other diagonal jack is pushed a certain distance, repeating this cycle. This gradual increase in the height of the sensor placement space is 5% to 20% of the support height, facilitating the placement of ultrasonic sensors while also preventing excessive push on the main beam that could damage the proper expansion joint structure.

[0065] Step 2: Deploy the sensor network:

[0066] N (N ≥ 12, N = 12 for circular arrangement, N = 16 for square arrangement) ultrasonic transceivers are evenly arranged along the outer edge of the upper surface of the support. A coordinate system Oxy is established with the center of the upper surface of the support as the origin to obtain the coordinate position of each ultrasonic sensor relative to the center of the support.

[0067] See also Figure 4 The distance between ultrasonic sensors is 15% to 30% of the characteristic length of the support (side length or diameter of the support). If the spacing is too large, a detection blind spot may easily occur, and if the spacing is too small, interference may easily occur. The distance between each ultrasonic sensor and the outer contour of the support should be no less than 5 mm to reduce the influence of edge signal reflection. The area of ​​the detection area enclosed by all ultrasonic sensors should be no less than 80% of the support area, and ensure that the detection area includes most of the damaged areas of the support, especially the damaged areas that are difficult to observe directly with the naked eye.

[0068] See also Figure 3 Specifically, for the rubber bearing, the sensor is arranged on the upper surface of the rubber layer 2, and for the steel bearing, the sensor is arranged on the upper surface of the spherical cap steel liner 402.

[0069] See also Figure 2 The equipment needed for ultrasonic guided wave testing includes: an integrated transceiver ultrasonic signal processing workstation, whose functions include ultrasonic guided wave signal exciter, data acquisition instrument, data analyzer (with built-in probabilistic damage model algorithm); BNC signal connection cable; integrated piezoelectric ceramic ultrasonic probe (which can receive and excite piezoelectric signals).

[0070] The connection relationship is as follows: the ultrasonic signal processing workstation port is connected to the integrated piezoelectric ceramic ultrasonic probe through each BNC signal connection line. The integrated piezoelectric ceramic ultrasonic probe is arranged at the support detection point to form a sensor network, forming a signal excitation-reception path.

[0071] The excitation process is as follows: the ultrasonic signal processing workstation sends an initial excitation signal to the i-th ultrasonic sensor through the BNC signal connection line. The excitation signal detects damage to the support. Other ultrasonic sensor pairs receive the damage signal and use the piezoelectric ceramic ultrasonic probe to transmit their respective damage signals to the ultrasonic signal processing workstation via the BNC signal connection line to complete a signal detection. Then, the ultrasonic signal is converted into a secondary excitation signal after being processed by the ultrasonic signal processing workstation. The other ultrasonic sensors excite their respective secondary excitation signals one by one, and the i-th ultrasonic sensor receives the corresponding inverted focusing signal.

[0072] Piezoelectric ceramic ultrasonic probes are used as ultrasonic sensors. As piezoelectric materials, piezoelectric ceramics have a unique piezoelectric effect that can convert electrical energy into mechanical energy, achieving the purpose of stimulating and receiving ultrasonic guided waves. The piezoelectric effect can cause the piezoelectric material to undergo electrical polarization when subjected to external pressure, generating charges of opposite signs on the surface, thereby generating a pressure-conducting current. Conversely, when the piezoelectric material is affected by an external current, it will produce corresponding mechanical vibrations, forming usable ultrasonic guided waves. In bearing damage detection, when a guided wave signal needs to be excited, a specific potential signal is excited into the piezoelectric material. The inverse piezoelectric effect of the piezoelectric material converts the potential energy into mechanical energy, and the guided wave signal for detecting damage will be transmitted in the form of mechanical vibration. When the damage signal needs to be received, the transmitted mechanical vibration triggers the positive piezoelectric effect in the piezoelectric material, generating electrical polarization, and the mechanical vibration transmitted by the damage detection is converted into an electrical signal and transmitted.

[0073] Step 3: Obtain ultrasonic inversion focusing signal:

[0074] The initial excitation signal is sent to the support through the i-th ultrasonic sensor , and receive corresponding detection signals through other N-1 ultrasonic sensors respectively; i=1, 2, ..., N;

[0075] For the detection signal received by the jth ultrasonic sensor, the threshold method is used to intercept the first damage signal, and the time reversal method is used to convert the first damage signal into a secondary excitation signal; j = 1, 2, ..., N, i ≠ j;

[0076] The jth ultrasonic sensor sends the secondary excitation signal to the support, and the threshold method is used to intercept the secondary damage signal of the secondary detection signal received by the i-th ultrasonic sensor as the inversion focus signal. .

[0077] The initial ultrasonic excitation signal uses ultrasonic guided waves that propagate along a specific path in the medium. Compared to general ultrasonic testing, ultrasonic guided waves have a wider coverage range, higher sensitivity, and are more suitable for the detection of complex structures. In bearing damage detection, the frequency of ultrasonic guided waves is the most critical parameter, which determines the feasibility and accuracy of damage detection. The higher the frequency, the higher the detection resolution but the weaker the penetration ability. The lower the frequency, the stronger the guided wave penetration ability but the lower the detection accuracy. The frequency of ultrasonic guided waves is selected according to the properties of the bearing material: for steel bearings, the frequency is 150kHz~300kHz, and for rubber bearings, the frequency is 40kHz~100kHz. Ultrasonic guided waves use pulse waveforms, and the wavelength and wave velocity of the guided waves are indirectly determined by the guided wave frequency according to the guided wave numerical equation.

[0078] For rubber bearings, the initial excitation signal of ultrasonic guided waves uses a narrow-band sinusoidal pulse signal modulated by Gaussian. The guided wave numerical equation is:

[0079]

[0080] Where: A is the voltage signal amplitude, generally 1V; f c is the signal center frequency, which is 40kHz~100kHz for rubber bearings; k is the normalized bandwidth, which is 0.1~0.5 for rubber bearings; s is the attenuation factor, which is 0.75~0.95 for rubber bearings; c is the signal delay parameter, which is 0~0.0005 for rubber bearings; t is the time; is the signal amplitude, ranging from -1 to 1; the selection of specific parameter values ​​needs to be flexibly adjusted according to the pulse width and frequency characteristics.

[0081] For steel supports, the initial excitation signal of ultrasonic guided waves uses a narrow-band sinusoidal signal Lamb wave modulated by a Hanning window. The numerical equation of the guided wave is:

[0082]

[0083] Where: f0 is the center frequency of the signal, which is 150kHz~300kHz for steel supports; n is the number of signal cycles, which is generally 4~6, with 4 being the preferred value.

[0084] The process of extracting the primary damage signal from the primary detection signal using the threshold method is as follows: define the start time of the primary detection signal as t0, obtain the absolute amplitude of each peak or trough in the primary detection signal, and obtain the maximum absolute amplitude of the primary detection signal. Set the threshold value Y and the cutoff time t1 so that the ratio of the absolute amplitude of the primary detection signal after the cutoff time t1 to the maximum absolute amplitude of the entire primary detection signal is less than or equal to the threshold value Y. The primary detection signal within the range of the start time t0 to the cutoff time t1 is regarded as the primary damage signal. Generally, Y = 4% to 8%, preferably 5%. (The process of extracting the secondary damage signal from the secondary detection signal using the threshold method is the same.)

[0085] The detection signals within the range t0 to t1 are the direct wave signal and subsequent secondary arrival waves. The direct wave and secondary arrival wave signals better reflect the bearing health information and the corresponding damage signature signals. However, the final coda wave signal, due to the interference between the boundary reflection signal and the damage diffraction signal, is difficult to extract the signal characteristics, resulting in poor reliability and hindering subsequent data processing. By eliminating these signals, the present invention can ensure data accuracy and detection precision while reducing data processing complexity, improving damage detection efficiency, and accelerating damage imaging.

[0086] The time reversal method is to reverse the time domain signal emitted by the sound source after the sensor receives it, without changing the amplitude (similar to mirror processing), and then transmit it back to the corresponding sensor, finally refocusing the signal to obtain an inverted focused signal. Figure 7 When there is no damage, the excitation signal and the inverted focus signal should be highly similar. When the structure is damaged, the change in the reflection characteristics of the damaged part will cause a significant difference between the inverted focus signal after the damage and the excitation signal before the damage. By comparing the difference between the inverted focus signal and the excitation signal, it can be determined whether damage has occurred.

[0087] Step 4: Calculate the damage characteristic index:

[0088] According to the initial excitation signal obtained in step 3 and inverted focus signal , calculate the bearing damage characteristic index according to formula (4.1) :

[0089]

[0090] Where, is the initial excitation signal, is the inverted focus signal, t0 is the start time, and t1 is the end time. The damage characteristic index DI is used to quantitatively describe the degree of difference between the intact signal and the damaged signal. The greater the signal difference, the larger the DI value, which means the greater the probability of damage. The value range of DI is 0~1.

[0091] Step 5: Determination of bearing damage:

[0092] According to the damage characteristic index The value is used to judge whether the bearing is damaged: if there are more than 15% damage characteristic indicators among all damage characteristic indicators If both are greater than 0.4, it is determined that the bearing has typical damage; in other cases, it is determined that the bearing does not have typical damage.

[0093] Typical bearing damage includes bearing rubber damage or bearing steel structure damage; the bearing rubber damage includes rubber aging wear and cracks in the rubber hard layer; the bearing steel structure damage includes partial corrosion of the bearing steel structure and fatigue crack damage.

[0094] The present invention mainly judges whether a bridge bearing is damaged based on damage characteristic indicators. Each excitation-receiving path has a different damage characteristic indicator DI. Considering actual operation errors and environmental disturbances, the DI threshold is set to 0.4. That is, if it is greater than 0.4, it is considered that there is a high probability of damage. Comprehensively considering the damage characteristic indicators of all excitation-receiving paths, if more than 15% of the damage characteristic indicators DI are greater than 0.4, it can be determined that the bridge bearing is damaged and in poor health.

[0095] Step 6: Locate typical bearing damage:

[0096] 6.1 Constructing the support probabilistic damage model:

[0097] Assume that the coordinates of any point on the upper surface of the support are , compare the sum of the distances from the point to the two ultrasonic sensors on any propagation path with the distance between the two ultrasonic sensors to obtain the ratio ,according to and , construct the probabilistic damage model as shown in Equations (6.1) to (6.2):

[0098]

[0099] Where, is the coordinate point The probability value of damage; N is the number of ultrasonic sensors; is the probability distribution function value, indicating the coordinate point The probability distribution of damage at each location is constructed using the principle of elliptical tomography; It is a control parameter that affects the image resolution and damage imaging range. ;

[0100] Calculate according to formula (6.3):

[0101]

[0102] Where, Any point on the upper surface of the support The ratio of the sum of the distances to the two ultrasonic sensors on any propagation path to the distance between the two ultrasonic sensors, ; is the coordinate of the hypothetical injury center; 、 are the coordinates of the i-th and j-th ultrasonic sensors respectively.

[0103] The present invention detects typical bearing damage. Damage such as bearing voids and displacement cannot be detected using ultrasonic guided waves. Typical bearing damage typically occurs on the contact surface, gradually caused by the forces acting between the bearing and the bridge structure (dynamic fatigue, wear, etc.). Damage then progresses from the bearing's exterior to the interior. However, internal bearing damage is often less severe than external damage, and its location corresponds to external damage. Therefore, detecting external bearing damage can generally characterize typical bearing damage. Furthermore, guided waves have a sufficient detection depth, allowing internal bearing damage to be detected.

[0104] Since ultrasonic guided waves are used for bearing damage detection, it is necessary to arrange multiple ultrasonic sensors in the detection area to form a detection network. Choosing a reasonable sensing path scheme (such as a one-to-one path in elliptical tomography, with only one pair of sensors) can achieve the best detection effect with a smaller number of sensors. At the same time, a large amount of signal data is generated during the detection. If some complex computational imaging methods are used, the computational pressure will be too high and the calculation speed will be slow. One of the biggest advantages of the elliptical tomography principle is that the calculation formula is simple and efficient. The elliptical tomography principle is based on weighted damage probability imaging. The weight distribution function adopts a linearly attenuated elliptical distribution. Data fusion uses the ellipses corresponding to each propagation path to superimpose to achieve tomography. The elliptical tomography principle can quickly calculate a large amount of complex data with a minimum number of sensors. The imaging resolution is high, and the detection capabilities of both single and multiple damages are strong. The data calculation is simple and efficient, suitable for on-site rapid calculation and rapid imaging. See Figure 10 Each propagation path affects an elliptical area. The closer to the excitation-sensing path, the greater the possibility of damage, and the farther away from the excitation-sensing path, the smaller the possibility of damage. A single calculation of a propagation path can only characterize the degree of damage to the bearing on this propagation path. The closer to the propagation path, the higher the degree of damage, that is, it presents an elliptical distribution. Multiply each propagation path by the corresponding damage characteristic index to form ellipses of different brightness. All ellipses of different brightness are superimposed together according to the measurement point position (ellipse focus, sensor point position), which can display the damage probability of any point on the bearing surface. The higher the degree of overlap, the greater the probability of damage. For typical damage to bridge bearings with oblique or complex shapes, the detection adaptability is relatively high.

[0105] 6.2 Imaging and positioning of typical bearing damage:

[0106] The damage probability value The values ​​are mapped to pixel values, and a damage probability cloud map of the bridge bearing is established according to the pixel values ​​of all coordinate points on the bearing. The pixel area with a damage probability value greater than or equal to 0.8 in the damage probability cloud map is taken as the damage area, and the coordinates of the damage area are obtained. The position with the largest damage probability value in the damage probability cloud map is the center of the damage area, and the typical damage positioning of the bearing is completed.

[0107] The method provided by this invention involves real-time automated detection and location of typical bridge bearing damage. While ensuring normal bridge traffic, the main beam is pushed forward. A sensor network is deployed on the bearing to be inspected, stimulating ultrasonic guided waves to obtain corresponding time-series sensor data. A time reversal method is used to reduce the guided wave detection's reliance on the initial non-destructive signal. Damage characteristic indicators under the corresponding operating conditions are calculated and combined with a probabilistic damage imaging algorithm for visual damage analysis. This method allows for rapid and accurate detection and location of bridge bearing damage. This method is rational, has broad applicability, and can be used for damage detection and location in serving bridge bearings.

[0108] In response to the shortcomings of current ultrasonic damage detection, in the field of bridge bearing damage detection, the time reversal method is combined to enhance the signal quality of this ultrasonic damage detection and improve the ability of damage detection to resist environmental interference. At the same time, no lossless reference signal is required, thereby improving the robustness of this ultrasonic damage detection system. Combined with the principle of probabilistic damage imaging, a complete set of probabilistic damage imaging models is constructed to batch and quickly process the obtained ultrasonic signal data, and intuitively interpret damage and image damage positioning based on damage factors, making the process of this ultrasonic damage detection system simple and reducing the requirements for operators. At the same time, the main beam of the bridge is pushed up and damage detection is carried out under normal driving conditions, which broadens the applicable scenarios of this detection method and improves the applicability of the method.

[0109] This method was later developed into a systematic process approach, overcoming the challenges of high personnel requirements, complex data processing and interpretation, high environmental interference, and the need for non-destructive signals. This provides a better solution for bearing damage detection. By incorporating jacking support, this integrated system can be better applied to bridge bearing damage detection during normal service, a new use case. The coordination and process-based nature of these various methods make the overall detection system more effective than using each component individually or using only one. This makes the system more efficient, simplifies the process, reduces time costs, enhances operability, and increases the scope of application, enabling users to obtain high-quality results faster when detecting bridge bearing damage.

[0110] Example 1;

[0111] This embodiment provides a real-time automated detection and location method for typical damage in bridge rubber bearings. Finite element software is used to establish a rubber bearing damage model, generate ultrasonic detection signals, and compare the ultrasonic detection imaging results with the damage in the original damage model to verify the accuracy of the method. Numerical simulations do not involve the process of jacking the main beam, but rather require the establishment of a rubber bearing damage model.

[0112] The detection and location of rubber bearing damage includes the following steps:

[0113] Step 1: Establish a finite element numerical model:

[0114] Typical rubber bearings are of two types: square and round. In this embodiment, the square rubber bearing is selected , its specific structure can be seen in Figure 3 (a), circular rubber bearing selection , its specific structure can be seen in Figure 3 (b); The unit is millimeters.

[0115] The finite element software used in this invention is COMSOL Multiphysics commercial finite element software. The square rubber bearing and the circular rubber bearing are modeled in the software according to their geometric structure and material properties. The boundary conditions are simplified. The bottom of the bearing adopts fixed constraints, and the other boundaries of the bearing adopt free boundary conditions. The extremely fine grid controlled by the physical field is adopted. The time step is 10 -7 The model transient calculation is performed in 0.0025 seconds with a total calculation time of 0.0025 seconds.

[0116] By changing the elastic modulus Es of the preset damage area, the rubber aging damage condition is simulated. Figure 5 (a) and Figure 5 (b) By setting cracks in the preset damage area, the crack damage condition of the rubber hard layer is simulated. Figure 5 (c) and Figure 5 (d); Specifically: the rubber aging zone of the rubber bearing is preset as a small circular area with a radius of 15 mm, which is located at the lower left of the center of the bearing, and the horizontal and vertical distances between the center of the circle and the center of the bearing are both 35 mm; the rubber elastic modulus set in the rubber aging zone is reduced to 0.5Es; the rubber hard layer cracks of the rubber bearing are preset as a series of small strip cracks, the crack width does not exceed 1 mm, the maximum characteristic length is 50 mm, and the position is located at the lower left of the center of the bearing, and the horizontal and vertical distances between the center and the center of the bearing are both 35 mm.

[0117] The final finite element model of the square rubber bearing is shown in Figure 6 (a), finite element model of circular rubber bearing Figure 5 (b).

[0118] Step 2: Deploy the sensor network:

[0119] In order to enhance the damage identification capability and improve the energy efficiency of ultrasonic guided wave excitation sensing path, such as Figure 4 The sensor network is arranged in a mesh layout, with a total of 16 sensors in the square rubber bearing and 12 sensors in the circular rubber bearing. A one-transmit-multiple-receive signal mode is adopted to stimulate the signal in turn so that the signal detection path covers the entire support range.

[0120] This embodiment does not consider signal transmission between sensors on the outermost edges of the square rubber support (e.g., 1#-5#, 1#-2#, etc.). The primary reason is that other sensors should be avoided along any path in the sensor network to prevent network layout clutter and interference. Other minor details include: excluding signal transmission between edges reduces boundary effects, lowers the signal-to-noise ratio, and enhances system robustness. Furthermore, the reflection angle of a square edge is greater than that of a circle, which is affected by the angle. A circular edge reflects in an arc shape, resulting in a longer reflection surface and no significant angular variation. Therefore, the boundary effect of the square edge is greater than that of a circle.

[0121] Step 3: Get the inverted focus signal:

[0122] For ultrasonic damage identification of rubber bearings, due to the high elasticity and strong sound absorption ability of rubber materials, relatively low-frequency ultrasonic guided waves should be adopted. The excitation signal used in this finite element simulation is a Gaussian modulated narrow-band five-peak sinusoidal pulse signal Gauss wave, such as Figure 7 As shown in (a), its parameters are voltage signal amplitude A = 1V, attenuation factor s = 0.8, and the center frequency of the signal is f c =60 kHz, normalized bandwidth k =0.15, signal delay c=0.0001, and its waveguide numerical equation is as follows:

[0123]

[0124] Where: A is the voltage signal amplitude, which is 1V; f c is the signal center frequency, which is 60 kHz in this embodiment; k is the normalized bandwidth, which is 0.15 in this embodiment; s is the attenuation factor, which is 0.8 in this embodiment; c is the signal delay parameter, which is 0.0001 in this embodiment.

[0125] Step 4: Calculate damage characteristic index and Step 5: Determine support damage:

[0126] The damage characteristic index DI on each ultrasonic propagation path in each working condition is calculated respectively, and DI is used as the damage characteristic index for research. The larger the DI, the greater the possibility of damage. Figure 9 (a) Using the aging damage of a square rubber bearing as an example, the damage characteristic indicators of all propagation paths are shown. Among the 80 excitation-sensing paths, 45 paths have damage characteristic indicator values ​​greater than the DI threshold of 0.4, accounting for 56.25%. This data is greater than the judgment threshold of 15%, so it can be determined that the bearing in this embodiment is damaged and in poor health.

[0127] Step 6: Locate typical bearing damage:

[0128] Figure 11 (a) Schematic diagram of the rubber aging damage imaging of the square rubber bearing in Example 1. Darker red indicates a greater probability of damage. The actual damage area and the damage imaging diagram closely overlap, demonstrating the effectiveness of the constructed damage imaging algorithm. After imaging, the operator maintains or replaces the bearing based on the damage location identified in the imaging results, then removes the jack to restore normal function.

[0129] Example 2;

[0130] This embodiment provides a real-time automatic detection and positioning method for typical damage of spherical steel bearings of bridges. The size of the steel ball cap lining is mm, its specific structure is shown in Figure 3 (c).

[0131] The detection and location of spherical steel bearing damage includes the following steps:

[0132] Step 1: Establish a finite element numerical model:

[0133] The finite element software used in this paper is COMSOL Multiphysics commercial finite element software. The model is built in the software according to the geometric structure and material properties of the spherical steel support. The boundary conditions are simplified. The bottom of the support adopts fixed constraints, and the other boundaries of the support adopt free boundary conditions. The extremely fine grid controlled by the physical field is adopted. The time step is 10 -7 The model transient calculation is performed in 0.0025 seconds with a total calculation time of 0.0025 seconds.

[0134] By presetting cracks in the pre-set damage area, the fatigue crack damage condition of the steel ball cap liner is simulated in Example 2. Figure 5 (e); Specifically: the fatigue crack of the steel structure of the steel ball crown liner is a small strip crack with a crack width not exceeding 1 mm and a maximum characteristic length of 50 mm. The crack is located at the lower left of the center point of the support, and the horizontal and vertical distances between the center and the center point of the support are both 35 mm.

[0135] Finally, the finite element model of the spherical steel support is established. Figure 6 (c).

[0136] Step 2: Deploy the sensor network:

[0137] In order to enhance the damage identification capability and improve the energy efficiency of ultrasonic guided wave excitation sensing path, such as Figure 4 This embodiment adopts a mesh-shaped sensor network arrangement, with a total of 12 sensors on the steel ball crown liner; a one-transmit-multiple-receive signal mode is adopted, and the signal is stimulated in turn, so that the signal detection path covers the entire detection range.

[0138] Step 3: Get the inverted focus signal:

[0139] For ultrasonic damage identification of spherical steel bearings, since the grain size of steel materials is relatively fine and the material has good acoustic transparency, a relatively high-frequency ultrasonic guided wave should be adopted. The excitation signal used in this finite element simulation is a narrow-band four-peak sinusoidal signal Lamb wave modulated by a Hanning window, such as Figure 7 As shown in (b), its parameters include the center frequency of the signal f0 = 240 kHz and the number of signal cycles n = 4. The specific formula is as follows:

[0140]

[0141] Wherein: f0 is the center frequency of the signal, which is 240 kHz in this embodiment; n is the number of signal cycles, which is 4 in this embodiment.

[0142] Step 4: Calculate damage characteristic index and Step 5: Determine support damage:

[0143] The damage index DI on each ultrasonic propagation path in each working condition is calculated respectively, and DI is used as the damage characteristic index for research. The larger the DI, the greater the possibility of damage. Figure 9 (b) Taking the fatigue crack damage of the steel ball crown liner in Example 2 as an example, the damage characteristic indicators of all propagation paths are displayed. Among the 66 excitation-sensing paths, 12 paths have damage characteristic indicator values ​​greater than the DI threshold of 0.4, accounting for 18.18%. This data is greater than the judgment threshold of 15%, so it can be determined that the bearing in this example is damaged and in poor health.

[0144] Step 6: Locate typical bearing damage:

[0145] Figure 11 (b) is a schematic diagram of fatigue crack damage imaging of the steel ball crown liner of the spherical steel support in Example 2. The darker the red color, the greater the probability of damage. It can be seen that the set damage area is highly overlapped with the damage area in the damage imaging schematic diagram, indicating that the constructed damage imaging algorithm is relatively effective.

[0146] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions based on the principles of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications made without departing from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A real-time automated detection and location method for typical damage in bridge bearings, characterized by: The steps include: Step 1: Preparation for testing: Push the main beam away from the support to form a sensor arrangement space between the main beam and the support; Step 2: Deploy the sensor network: N ultrasonic transceivers are evenly arranged along the outer edge of the upper surface of the support. A coordinate system Oxy is established with the center of the upper surface of the support as the origin to obtain the coordinate position of each ultrasonic sensor relative to the center of the support; N ≥ 12; Step 3: Obtain ultrasonic inversion focusing signal: The initial excitation signal is sent to the support through the i-th ultrasonic sensor , receive the first detection signal through the other N-1 ultrasonic sensors respectively; i=1, 2, ..., N; For the first detection signal received by the jth ultrasonic sensor, the threshold method is used to intercept the first damage signal, and the time reversal method is used to convert the first damage signal into a secondary excitation signal; j = 1, 2, ..., N, i ≠ j; The jth ultrasonic sensor sends the secondary excitation signal to the support, and the threshold method is used to intercept the secondary damage signal of the secondary detection signal received by the i-th ultrasonic sensor as the inversion focus signal. ; Step 4: Calculate the damage characteristic index: According to the initial excitation signal obtained in step 3 and inverted focus signal , calculate the bearing damage characteristic index according to formula (4.1) : (4.1)； Where, is the initial excitation signal, is the inversion focus signal, t0 is the start time, and t1 is the end time; Step 5: Determination of bearing damage: If there are more than 15% of all damage characteristic indicators If both are greater than 0.4, it is determined that the bearing has typical damage; In other cases, it is determined that the bearing does not have typical damage; Step 6: Locate typical bearing damage: 6.1 Constructing a probabilistic bearing damage model: Assume that the coordinates of any point on the upper surface of the support are , compare the sum of the distances from the point to any two ultrasonic sensors with the distance between the two ultrasonic sensors to obtain the ratio ,according to and , construct the support probabilistic damage model as shown in Equations (6.1) to (6.2): (6.1)； (6.2)； Where, is the coordinate point The probability value of damage; N is the number of ultrasonic sensors; is the probability distribution function value; is the control parameter; 6.2 Imaging and positioning of typical bearing damage: The damage probability value Mapped into pixel values, a bridge bearing damage probability cloud map is established based on the pixel values ​​of all coordinate points on the bearing. The pixel area with a damage probability value greater than or equal to 0.8 in the damage probability cloud map is regarded as the damage area. The coordinates of the damage area are obtained to complete the positioning of typical bearing damage.

2. The method according to claim 1, wherein: In step 1, the height of the sensor arrangement space is 5% to 20% of the support height.

3. The method according to claim 1, wherein: In step 2, the distance between ultrasonic sensors is 15% to 30% of the characteristic length of the support; the distance between each ultrasonic sensor and the outer contour of the support is not less than 5 mm; the area of ​​the detection area enclosed by all ultrasonic sensors is not less than 80% of the support area; the characteristic length refers to the side length or diameter of the support.

4. The method according to claim 1, wherein: In step three, the ultrasonic sensor uses a piezoelectric ceramic ultrasonic probe.

5. The method according to claim 1, wherein: In step three, the process of extracting the first damage signal from the first detection signal using the threshold method is as follows: define the starting time of the first detection signal as t0, obtain the absolute amplitude of each peak or trough in the first detection signal, and obtain the maximum absolute amplitude of the first detection signal, set the threshold Y and the cutoff time t1, so that the ratio of the absolute amplitude of the first detection signal after the cutoff time t1 to the maximum absolute amplitude of the first detection signal is less than or equal to the threshold Y, and take the first detection signal within the range of the starting time t0 to the cutoff time t1 as the first damage signal.

6. The method according to claim 1, wherein: In step three, the ultrasonic initial excitation signal uses ultrasonic guided waves, and the frequency of the ultrasonic guided waves is selected according to the properties of the bearing material: for steel bearings, the frequency is 150kHz~300kHz, and for rubber bearings, the frequency is 40kHz~100kHz; the ultrasonic guided waves use a pulse waveform, and the wavelength and wave velocity of the guided waves are determined according to the numerical equation of the guided waves.

7. The method according to claim 6, wherein: For rubber bearings, the initial excitation signal of ultrasonic guided waves uses a narrow-band sinusoidal pulse signal modulated by Gaussian. The guided wave numerical equation is: (3.1)； Where: A is the voltage signal amplitude; f c is the signal center frequency; k is the normalized bandwidth; s is the attenuation factor; c is the signal delay parameter; t is the time; is the signal amplitude; For steel supports, the initial excitation signal of ultrasonic guided waves uses a narrow-band sinusoidal signal Lamb wave modulated by a Hanning window. The numerical equation of the guided wave is: (3.2)； Where: f0 is the center frequency of the signal; n is the number of signal cycles; is the signal amplitude.

8. The method according to claim 1, wherein: In step five, typical bearing damage includes bearing rubber damage or bearing steel structure damage; the bearing rubber damage includes rubber aging wear and cracks in the rubber hard layer; the bearing steel structure damage includes partial corrosion of the bearing steel structure and fatigue crack damage.

9. The method according to claim 1, wherein: In step six, Calculate according to formula (6.3): (6.3)； Where, are the coordinates of the i-th and j-th ultrasonic sensors respectively.

Images