CN101782372B - An Intelligent Method for Damage Diagnosis of Bridge Expansion Joints Based on Longitudinal Displacement of Beam Ends - Google Patents

Abstract

The present invention relates to an intelligent diagnosis method for bridge telescopic seam injury based on girder end longitudinal displacement. The long-term monitoring data of main girder longitudinal displacement, temperature and vertical acceleration after a bridge is built are obtained by mounting a small number of sensors on the bridge; a relevance model of the main girder longitudinal displacement, the bridge temperature and the vertical acceleration in a healthy state is established step by step; the influence of the main girder temperature and the vertical acceleration on the longitudinal displacement can be eliminated on the basis of the relevance model, and the environment condition uniformization displacement which can reflect a working state of a telescopic seam is obtained. When the method is applied to the injury diagnosis of the telescopic seam, the monitoring data in an unknown state are processed just by adopting the established relevance model, and finally, the environment condition uniformization displacement of the healthy state and the unknown state is simultaneously input into a mean value control map; and if the telescopic seam generates injury, a sample point of the control map can exceed a control line, thereby realizing the intelligent recognition on the injury of the telescopic seam.

Description

An Intelligent Method for Damage Diagnosis of Bridge Expansion Joints Based on Longitudinal Displacement of Beam Ends

technical field

The invention is an intelligent method applied to damage diagnosis of expansion joints of bridge structures, and relates to the field of non-destructive testing of bridge engineering.

Background technique

When the temperature of the bridge changes, the bridge deck has longitudinal deformation of expansion and contraction. In addition, under the action of vehicle load, the bridge deck will also produce longitudinal displacement. In order to meet this deformation, it is necessary to set expansion joints between the beam end of the bridge and the abutment. In the bridge structure, expansion joints must not only meet the longitudinal deformation of the main beam caused by temperature changes and vehicle loads, but also meet the displacement caused by concrete shrinkage and creep, and reduce the secondary deformation of the superstructure caused by uneven settlement. Stress ^[1] , therefore, expansion joints are very important structural components in bridges, and whether their state is safe or not is not only related to the operation of the entire bridge structure, but also directly affects the driving state of vehicles passing through the bridge. Bridge expansion joints bear complex forces and have always been the weak link of bridges. However, due to defects in design and construction, poor management and maintenance, and vehicle overloading, damage to expansion joints in actual engineering is more serious. Therefore, it is necessary to The status of expansion joints is monitored and evaluated in order to accurately detect the occurrence of damage and repair or replace expansion joints in a timely manner.

At present, the damage detection of expansion joints mainly adopts manual periodic detection, which has the following problems: (1) manual detection is highly subjective, and cannot make quantitative judgments on the damage state of expansion joints; (2) manual Detection usually affects traffic and lacks the accumulation of historical data; (3) the real-time performance is poor, and the occurrence of expansion joint damage cannot be found in time, which may affect the safety of bridge structures and traffic; (4) due to the need for long-term and regular Assign maintenance workers to conduct on-site inspections, and the overall cost is higher. Therefore, in view of the above-mentioned shortcomings of manual detection, it is urgent to develop an intelligent method for real-time damage diagnosis of the state of bridge expansion joints. The development of bridge structure health monitoring technology provides an opportunity to achieve the above goals ^[2-3] . During the construction process, sensors can be laid on the bridge structure, and the longitudinal displacement of the beam end, the structure temperature, and the traffic load can be recorded for a long time during the bridge operation period. Through these monitoring data, bridge managers can evaluate the "health" status of expansion joints, and provide basis and guidance for maintenance, repair and management decisions of expansion joints.

references

[1] Chen Wai-Fah, Duan Lian. Bridge engineering handbook [M]. Boca Raton: CRC Press, 2000.

[2] Li Aiqun, Miao Changqing, Li Zhaoxia. Research on Structural Health Monitoring System of Runyang Yangtze River Highway Bridge [J]. Journal of Southeast University (Natural Science Edition), 2003, 33(5): 544-548.

[3] Ko J M, Ni Y Q. Technology developments in structural health monitoring of large-scale bridges [J]. Engineering Structures, 2005, 27(12): 1715-1725.

Contents of the invention

Technical problem: The purpose of the present invention is to provide a method for diagnosing expansion joint damage based on abnormal monitoring of beam end longitudinal displacement, so as to solve the shortcomings of existing detection technologies.

Technical solution: The basic idea of the present invention is: since the purpose of setting the expansion joints at the end of the main girder of the bridge is to meet the needs of the longitudinal displacement of the main girder, the change law of the longitudinal displacement of the beam end implies the expansion joints. State information, when the expansion joint is damaged, the longitudinal displacement of the beam end will change abnormally. Based on this displacement change, the damage of the bridge expansion joint can be identified.

However, as described in the background technology, under the operating conditions of the bridge, the effects of temperature and vehicle load will cause the longitudinal displacement of the beam end to fluctuate in a wide range (the principle of the temperature and vehicle load causing the beam end displacement can be found in the specification sheet Figure 1 and Figure 2), this fluctuation will submerge or cover up the displacement change of the structure caused by the expansion joint damage. Therefore, the method of the present invention is to establish a mathematical model of the longitudinal displacement of the beam end under the normal state of the expansion joint and the environmental conditions such as temperature and vehicle load, and eliminate the influence of the environmental condition on the longitudinal displacement on this basis, so as to obtain a model that can truly reflect the expansion joint. The longitudinal displacement of the healthy state, and the method of mean value control chart is used to identify the abnormal change of the longitudinal displacement caused by the damage of the expansion joint, thus establishing an intelligent method for real-time online damage diagnosis of the expansion joint state.

The intelligent method of bridge expansion joint damage diagnosis based on beam end longitudinal displacement proposed by the present invention is:

1) The longitudinal displacement of the girder end of the main beam and the setting of the environmental condition sensor of the bridge

During bridge construction, a longitudinal displacement sensor is installed at the girder end of the main girder. At the same time, a temperature sensor and an acceleration sensor are installed at the mid-span position of the main girder to monitor the temperature of the main girder and the vertical acceleration of the main girder caused by the vehicle load. ;

2) Processing of monitoring data

With 10-min as the calculation interval, the raw data acquired by the sensor are processed, and the representative values of longitudinal displacement, temperature and vehicle load at the beam end are calculated;

3) Mathematical correlation model of longitudinal displacement and environmental conditions in intact state

a) Select the monitoring data n days after the bridge construction is completed to establish the relevant model, the longitudinal displacement D, the temperature T and the representative value R of the vehicle load,

b) The linear regression method is used to establish the relationship between the temperature T and the longitudinal displacement D of the beam end, and the regression model parameters are calculated by the least square method,

c) Before establishing the correlation model between the vehicle load representative value R and the displacement, the influence of temperature on the longitudinal displacement is first eliminated, the reference temperature is selected as T _r , and the original test value D of the displacement is "normalized" to the reference temperature T _r , to obtain The longitudinal displacement value D ₁ of the beam end to eliminate the influence of temperature,

d) Use linear regression to establish a correlation model between D ₁ and the representative value R of vehicle load, and then similar to step c), select the reference value of vehicle load R _r , and "normalize" D ₁ obtained in step c) To the reference value R _r of the vehicle load, the displacement value D ₂ to eliminate the influence of the vehicle load is obtained;

4) Determination of the significance level of the control chart

Take the daily average value of the displacement value D ₂ calculated in step 3), record it as D ₂ , input it into the mean value control chart, and adjust the significance level of the control chart so that the above n sample points all fall on the top of the control chart, within the Lower Line of Control;

5) Intelligent diagnosis of expansion joint damage

For the m-day monitoring data in the unknown state, the correlation model under the intact state of the expansion joint is used to eliminate the influence of temperature and vehicle load, and m daily average displacement values are obtained on this basis, which is recorded as D ₃ . Keep the significance determined in step 4) unchanged, and input D ₂ and D ₃ into the mean value control chart at the same time. At this time, if all n+m samples are still within the upper and lower control lines, it means that the state of the expansion joint is normal , if any sample falls outside the control line, it indicates that the state of the expansion joint is abnormal, and an early warning of damage to the expansion joint can be given.

Beneficial effects: Aiming at the engineering reality that bridge expansion joints are extremely prone to damage under operating conditions, the present invention proposes an intelligent method for damage diagnosis based on longitudinal displacement of beam ends by comprehensively using on-site testing, linear regression, and mean value control charts, which has the following benefits Effect:

(1) The number of sensors to be installed in the present invention is relatively small, and only displacement sensors, temperature sensors and acceleration sensors are needed. At the same time, the method adopted by the present invention is simple and easy to implement, and can be programmed and implemented in a computer more conveniently, which is convenient for the application in actual engineering.

(2) The present invention fully considers the influence of environmental factors on the longitudinal displacement of the beam end, and eliminates the influence of structural temperature and vehicle load on the displacement test value step by step, and the displacement value after eliminating the influence of environmental factors can accurately reflect the operating state Health status of lower bridge expansion joints.

(3) The method of introducing the mean value control chart in the present invention performs multi-sample hypothesis testing on abnormal changes in the longitudinal displacement of the beam end, which can reduce the possibility of misjudgment.

(4) The present invention can monitor the state of expansion joints on-line, without manual intervention in the implementation process of the method, which reduces the expenditure of human labor, can realize unattended intelligent monitoring of bridge expansion joints, and has broad engineering application prospects.

Description of drawings

Figure 1 is the longitudinal displacement and deformation diagram of a simply supported beam when the temperature rises. In the figure, T↑ represents the temperature rise, and _dt represents the longitudinal displacement of the beam end caused by the temperature;

Figure 2 is the longitudinal displacement and deformation diagram of simply supported beam under the action of vehicle load, in which F represents the vehicle load, and _df represents the longitudinal displacement of the beam end caused by the vehicle load;

Fig. 3 The correlation scatter diagram of longitudinal displacement and temperature at the north end of the main girder;

Fig. 4 The correlation scatter diagram of longitudinal displacement and temperature at the south end of the main girder;

Figure 5 is a scatter diagram of the correlation between the longitudinal displacement and the root mean square acceleration of the north end of the main girder, and the acceleration RMS in the figure means the root mean square acceleration;

Figure 6 is a scatter diagram of the correlation between the longitudinal displacement and root mean square acceleration at the south end of the main girder. The acceleration RMS in the figure means the root mean square acceleration;

Figure 7 is the daily average measured value of the longitudinal displacement at the north end of the main girder and the normalized value after eliminating the environmental impact;

Figure 8 is the daily average measured value of the longitudinal displacement at the south end of the main girder and the normalized value after eliminating the environmental impact;

Figure 9 is the average displacement control diagram of expansion joints under normal conditions, UCL in the diagram represents the upper control line, LCL represents the lower control line, and CL is the center line of the control diagram;

Figure 10 is the average displacement control diagram of expansion joints under damaged state. UCL in the figure represents the upper control line, LCL represents the lower control line, and CL is the center line of the control diagram.

Detailed ways

Specific embodiments of the present invention are further described below:

(1) During the setting process of the longitudinal displacement of the girder end and the environmental condition sensors of the bridge, the number, position and parameter setting of the sensors can be determined according to the specific conditions such as the type of bridge, the span diameter, the bridge deck width and the environment of the bridge site. , usually a longitudinal displacement sensor is set at each end of the main girder, and a temperature sensor and an acceleration sensor are set in the middle of the main girder span, which can meet the needs of the present invention.

(2) The original monitoring data is processed as follows: the longitudinal displacement of the beam end and the structural temperature data are calculated in a 10-min interval, and the average value is used as the representative value of the longitudinal displacement of the beam end and the structural temperature in this time period , and calculate the root mean square (Root Mean Square, abbreviated as RMS) of the vertical acceleration of the main beam in a 10-min interval, and use it as the representative value of the strength of the vehicle load in this period of time.

(3) Select the monitoring data of n days after the completion of the bridge construction to establish the relevant model, because the expansion joints during this period can be considered to be in a good state, and D, T and R represent the displacement, temperature and vehicle load respectively value, the total number of samples is 144×n.

(4) The linear regression method is used to establish the relationship between the temperature T and the longitudinal displacement D of the beam end, and the model expression is:

D＝β ₀ +β ₁ T (1)

In the formula, β ₀ and β ₁ are the regression coefficients, which can be obtained by the method of least squares:

${\mathrm{\β}}_1=\frac{S_{\mathrm{DT}}}{S_{\mathrm{TT}}},$ β ₀ ＝D-β ₁ T (2)

In the formula, S _DT is the covariance of displacement and temperature; S _TT is the variance of temperature; D and T are the mean values of displacement and temperature, respectively.

(5) Select the reference temperature as T _r , and substitute it into formula (1) to obtain the reference value of displacement D _r , and substitute the representative value of temperature T into formula (1) to obtain the calculated value of displacement D _t , then Calculate the longitudinal displacement value D ₁ of the beam end to eliminate the influence of temperature:

D ₁ =D-(D _t -D _r ) (3)

(6) Establish the linear regression model of the displacement value D ₁ and the vehicle load representative value R that have eliminated the influence of temperature:

D ₁ =β ₂ +β ₃ R (4)

In the formula, β ₂ and β ₃ are regression coefficients, which can be calculated similarly to formula (2). Select the reference value of the vehicle load R _r and substitute it into formula (4) to obtain the reference value D _rr of the displacement, and substitute the representative value R of the vehicle load into formula (4) to obtain the calculated value D _tt of the displacement, which can be calculated and eliminated Displacement value D ₂ affected by vehicle load:

D ₂ =D ₁ -(D _tt -D _rr ) (5)

(7) Take the daily average of 144×n displacement values D ₂ that have eliminated the influence of temperature and vehicle load, and obtain n displacement samples, which are recorded as D ₂ , input D ₂ into the mean control chart, and adjust the significance of the control chart level, so that the above n sample points all fall within the upper and lower control lines (UCL, LCL).

(8) For the m-day monitoring data in the unknown state, first process it into 10-min as the representative value of the calculation interval, and then use the relevant model under the intact state of the expansion joint to eliminate the influence of temperature and vehicle load, and calculate 144×m A displacement value D ₃ that eliminates the influence of temperature and vehicle load, and then take the daily average value to obtain m displacement samples, which are recorded as D ₃ . Keep the significance level of the control chart consistent with that in the intact state, and input D ₂ and D ₃ into the mean control chart at the same time. At this time, if all n+m samples are still within the upper and lower control lines, it means that the expansion joint The state is normal. If any sample falls outside the control line, it indicates that the state of the expansion joint is abnormal, and an early warning of damage to the expansion joint can be given.

Take the South Branch Suspension Bridge of Runyang Bridge as an example below to illustrate the specific implementation process of the present invention:

The monitoring data of 100 days from January to June 2006 were selected to establish the correlation between the longitudinal displacement of the beam end and the temperature and the vehicle load. The sensors used were the displacement sensors at both ends of the main beam, the temperature sensor in the middle of the main beam span and the vertical acceleration The sensors calculated representative values of longitudinal displacement, temperature and vehicle load at 10-min time intervals, totaling 144×100=14400 samples.

Figure 3 and Figure 4 show the correlation scatter diagrams of displacement and temperature at the north end and south end of the main beam respectively. From these two figures, it can be seen that there is a strong linear correlation between the longitudinal displacement and temperature at the beam end , and show the characteristics of "large displacement at high temperature and small displacement at low temperature". At the same time, it is found from Fig. 3 that the variation range of the displacement at the north end of the main beam is [-23.0cm, 28.7cm], and from Fig. 4 that the variation range at the south end is [-20.6cm, 33.0cm]. From this, the variation of the longitudinal displacement at the beam end can be obtained The amplitude is 51.7cm and 53.6cm.

Table 1 shows the correlation model between displacement D and temperature T established by linear regression analysis.

Table 1 Linear regression model of temperature-displacement

Location Regression function (displacement: D(cm) temperature: T(℃)) North end of main beam D＝-17.6681+1.0032T South end of main beam D＝-15.5170+1.0097T

Select the reference temperature T _r as 20°C, and substitute the reference temperature value into the linear model in Table 1 to obtain the reference displacement D _r of the beam end. At the same time, substitute the temperature representative value T into the model in Table 1 to obtain the calculated value of displacement D _t , and obtain The longitudinal displacement value D ₁ of the beam end that eliminates the influence of temperature.

Figure 5 and Figure 6 show the correlation scatter diagrams of the longitudinal displacement D ₁ at the north end and the south end of the main girder and the representative value R (acceleration RMS) of the vehicle load. It can be seen from the figure that the distribution of data points is scattered, but It can still be seen that there is an obvious correlation between R and _D1 , showing the characteristics of "large load with small displacement, small load with large displacement".

Similarly, Table 2 shows the correlation model of the displacement _D1 and the representative value R of the vehicle load established by regression analysis.

Table 2 Linear regression model of vehicle load-displacement

Location Regression function (displacement: D ₁ (cm) acceleration RMS: R (cm/s ² )) North Section D ₁ =2.8834-0.5391R south end D ₁ =5.6170-0.6646R

Select the representative value R _r of the reference vehicle load as 1cm/s ² , and substitute the representative value of the vehicle load into the linear model in Table 2 to obtain the reference displacement D _rr of the beam end. At the same time, substitute the representative value R of the vehicle load into the model in Table 2 to obtain the displacement The calculated value D _tt of D tt is obtained to obtain the longitudinal displacement value D ₂ of the beam end that eliminates the influence of temperature and vehicle load.

Take the daily average value of D ₂ to obtain 100 samples, which are recorded as D ₂ .

Take another 48 days of monitoring data as the data when the state of the expansion joint is unknown, and use the method of the present invention to obtain 48 daily average displacement samples, which are recorded as D ₃ .

Fig. 7 and Fig. 8 show the daily average longitudinal displacement measured value and the normalized value after eliminating the environmental influence, respectively, at the north end and south end of the main girder. The solid line in the figure represents the daily average measured value of the displacement, the first 100 samples of the dotted line represent D ₂ , and the last 48 samples represent D ₃ . From these two figures, it can be seen that the normalized value curves of the displacement at the north and south ends are all the same. very stable, and the amplitude is very small, which shows that the method of the present invention effectively removes the influence of environmental factors on the longitudinal displacement of the beam end,

Input D ₂ and D ₃ into the mean value control chart at the same time, and Figure 9 shows the mean value control chart of the expansion joint under normal conditions. The first 100 data in the figure represent the intact state of the expansion joint, that is, D ₂ ; the last 48 data represent the unknown state of the expansion joint, that is, D ₃ . By adjusting the significance level of the control chart, the first 100 samples are just within the upper and lower control lines. At the same time, it can be found that the last 48 samples are also within the control line, which shows that the expansion joints are in a healthy state at this time.

In order to check the effect of the present invention on the damage assessment of expansion joints, a certain change is applied to the above-mentioned 10-min average displacement representative value of another 48 days, so as to simulate the impact of expansion joint damage on displacement:

D _m =D-εΔD (6)

In the formula, D is the measured value of 10-min displacement in the last 48 days; D _m is the simulated value of 10-min displacement in the last 48 days of expansion joint damage; ε represents the damage level, which is taken as 1.0% here; The annual variation range of the longitudinal displacement of the beam end is 53.6cm and 51.7cm introduced above.

Then D _m is calculated according to the method of the present invention to obtain 48 daily average displacement samples, which are recorded as D ₃ . Then input D ₂ and D ₃ into the mean value control chart at the same time, keeping the significance level of the control chart unchanged. Figure 10 shows the mean displacement control chart of the expansion joint under simulated damage state. When the damage causes a 1.0% change in displacement, the last 48 samples obviously approach the lower control line, and some samples have exceeded the control range. At this time, it can be judged that the expansion joint is damaged.

The above calculation examples show that the method proposed by the present invention can effectively eliminate the influence of environmental conditions on the longitudinal displacement of the beam end, extract the normalized value of the longitudinal displacement that can reflect the state of the expansion joint, and can be applied to the long-term online monitoring of the expansion joint of the bridge and injury diagnosis.

Claims (1)

1. bridge telescopic seam injury intelligent diagnosing method based on girder end longitudinal displacement is characterized in that this damage intelligent diagnosing method is:

1) setting of the length travel of girder beam-ends and bridge environmental conditions ensor:

When bridge construction is built, the length travel sensor is set in girder beam-ends position, simultaneously, at the span centre position mounting temperature sensor and the acceleration transducer of girder, in order to the temperature of monitoring girder with because the vertical acceleration of girder that vehicular load causes;

2) processing of Monitoring Data:

With 10min is computation interval, and the raw data that sensor obtains is handled, and calculates the typical value of girder end longitudinal displacement, temperature and vehicular load; The typical value of described vehicular load is to be the interval root mean square that calculates the vertical acceleration of girder with 10min, with this intensity typical value as the vehicular load in this time period;

3) the mathematics correlation model of length travel and environmental baseline under the serviceable condition:

A) choose bridge construction and build up n days the Monitoring Data in back and set up correlation model, length travel D, temperature T and vehicular load typical value R,

B) method of employing linear regression is set up the relation between temperature T and the girder end longitudinal displacement D, and the regression model parameter is calculated by least square method,

C) before the correlation model of setting up vehicular load typical value R and displacement, eliminate the influence of temperature earlier to length travel, choosing reference temperature is T _r, with the original test value D of displacement " normalization " to reference temperature T _r, the girder end longitudinal displacement value D of the temperature effect that is eliminated ₁,

Concrete grammar is: the method for employing linear regression is set up the relation between temperature T and the girder end longitudinal displacement D, and the model tormulation formula is:

D＝β ₀+β ₁T (1)

In the formula, β ₀And β ₁Be regression coefficient, can obtain by the method for least square:

In the formula, S _DTCovariance for displacement and temperature; D _TTVariance for temperature;

With

Be respectively the average of displacement and temperature, choosing reference temperature is T _r, with its substitution formula (1), the reference value that obtains displacement is D _r,, obtain calculation of displacement value D simultaneously with also substitution formula of temperature typical value T (1) _tSo, can calculate the girder end longitudinal displacement value D that eliminates temperature effect ₁:

D ₁＝D-(D _t-D _r) (3)

D) adopt linear regression to set up D ₁With the correlation models of vehicular load typical value R, similar with step c) then, the reference value of choosing vehicular load is R _r, with the D that obtains in the step c) ₁" normalization " is to the reference value R of vehicular load _r, the shift value D of the vehicular load that is eliminated influence ₂

Concrete grammar is: set up the shift value D that has eliminated temperature effect ₁Linear regression model (LRM) with vehicular load typical value R:

D ₁＝β ₂+β ₃R (4)

In the formula, β ₂And β ₃Be regression coefficient, can similar formula (2) calculate that the reference value of choosing vehicular load is R _r,, obtain the reference value D of displacement with its substitution formula (4) _Rr,, obtain calculation of displacement value D with also substitution formula of vehicular load typical value R (4) _Tt, can calculate the shift value D that has eliminated the vehicular load influence ₂:

D ₂＝D ₁-(D _tt-D _rr) (5)

4) determining of control chart level of significance:

The shift value D that step 3) is calculated ₂Get daily mean, be designated as , with its input mean chart, adjust the level of significance of control chart, make a said n sample point all drop within the upper and lower control line of control chart;

5) intelligent diagnostics of telescopic seam injury:

To m days monitoring data of unknown state, the correlation model under the serviceable condition of employing expansion joint is eliminated the influence of temperature and vehicular load, obtains m per day shift value on this basis, is designated as

, the conspicuousness that keeps step 4) to determine is constant, will With

Import mean chart simultaneously, at this moment, if all n+m sample still all is positioned at upper and lower control line, illustrate that then the expansion joint state is for normal, if there is sample to drop on beyond the control line, the expansion joint abnormal state then is described, can make the early warning that damage takes place at the expansion joint.

Images