A method for identifying stress of corroded steel strand based on wave peak time delay of guided wave

Abstract

The application provides a corrosion steel strand stress identification method based on a wave peak time delay, and the steel strand stress is deduced from the calculation of the wave peak time delay index. The application can be applied to the identification of the existing stress of the corrosion steel strand, and the stress identification object is a seven-wire steel strand with a nominal diameter of 15.2 mm, and can be extended to other specifications of steel strands or general rod-shaped force components. The wave peak time delay index used in the application has the characteristics of high sensitivity and high efficiency, and the optimal guided wave excitation frequency is selected, so that the accuracy and robustness of the stress identification can be improved.

Description

Technical Field

[0001] This invention relates to the field of stress identification technology for corroded steel strands, and in particular to a method for stress identification of corroded steel strands based on guided wave peak time delay. Background Technology

[0002] Steel strands are used as tension members in prestressed concrete structures, such as cable-stayed bridges, suspension bridges, steel reinforcement bundles in offshore platforms, as well as dams and nuclear power plants. Identifying and maintaining their good condition is crucial to ensuring the safety of the structure.

[0003] During their service life, steel strands are under high-stress tension, subjected to repeated vehicle loads and environmental factors such as rain, seawater, wind, and sun exposure. Corrosion is one of the main causes of performance degradation in prestressed structures, and structural safety accidents caused by corrosion are common worldwide. Corrosion not only affects the structural stress but also causes significant economic losses and maintenance costs. Prestressed structures bear enormous loads and support important social functions. As load-bearing components, steel strands are not only affected by corrosion but also experience prestress loss. Under the influence of long-term loads and environmental factors, the mechanical properties of steel strands decline. Damage, relaxation, creep caused by steel strand corrosion, and shrinkage of concrete all affect their effective stress. Overestimating prestress loss leads to conservative and uneconomical maintenance designs, while underestimating prestress loss creates structural safety hazards or even causes disasters, impacting economic and social safety. Therefore, the detection of effective stress in steel strands is crucial. Steel strand stress detection techniques include vibration methods, magnetic flux methods, and impedance methods. While the vibration method establishes an ideal linear relationship between the characteristic quantities and stress values ​​for identifying bare steel strand stress, its accuracy is easily affected by the span, sag, and boundary conditions of the component itself. The magnetic flux method is sensitive to ambient temperature; testing in extremely cold or hot environments can have a significant impact, requiring extensive experimental calibration. When testing the stress of steel strands encased in concrete, several scholars have reached conflicting conclusions regarding the relationship between the natural frequency and stress in the vibration method, regardless of the presence or absence of cracks. The method of embedding a large, penetrating sensor within the concrete and placing it on the steel strand presents challenges. The impedance method is susceptible to temperature effects, necessitating quantitative temperature compensation experimental research.

[0004] Ultrasonic guided wave stress identification based on guided wave acoustoelasticity has unique advantages in the field of non-destructive stress testing. Currently, most ultrasonic guided wave stress identification indicators are constructed based on guided wave group velocity. However, measuring guided wave group velocity requires stripping at least 1.5–2 m of bare steel strand from the actual structure as the measurement gauge length. This is not only difficult to implement in actual testing but also seriously affects the durability of the steel strand and the structural safety. To shorten the measurement gauge length to about 20 cm (the length of this testing section is the single-twist length of the steel strand), guided wave group velocity cannot be used; instead, guided wave phase velocity must be employed. Since guided waves inevitably disperse during propagation in the steel strand, guided wave phase velocity is difficult to measure directly, necessitating the construction of other easily measurable identification indicators based on guided wave phase velocity. Furthermore, most research still focuses on the non-corroded state of the steel strand. Corrosion affects the accuracy of guided wave stress detection; therefore, it is necessary to determine the impact of corrosion on stress identification indicators. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a method for stress identification of corroded steel strands based on waveguide peak delay. The method identifies the existing stress of corroded steel strands based on the waveguide peak delay index. It features a short measurement gauge length, high sensitivity of the delay index to stress changes, high signal strength ratio, good linear relationship between the delay index and stress changes, and a logarithmic relationship between the delay index and diameter changes.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a method for stress identification of corroded steel strand based on guided wave crest time delay, comprising the following steps:

[0007] Step 1: Take m samples of corroded steel strands of the same type as those in the actual engineering structure. For each sample, randomly select n segments. The length of each segment should be selected according to the testing requirements, including whether the steel strand has a single or multiple twist. Measure the diameter d of the corroded steel strand. Calibrate the corrosion delay parameters x, y, z, and stress sensitivity coefficient k at a certain excitation frequency, and obtain the signal strength ratio c and linearity r. 2 Let d be the diameter of the j-th segment of the i-th steel strand sample. ij Corrosion delay parameter x ij y ij z ij Stress sensitivity coefficient k ij Signal strength ratio index c ij and linearity At this time, 1≤i≤m, 1≤j≤n;

[0008] Step 2: Change the frequency of the narrowband pulse signal and repeat Step 1 to obtain several corrosion delay parameters, stress sensitivity coefficient, signal strength ratio, and linear fit goodness at preset frequencies;

[0009] Step 3: By comparing the values ​​of the three indicators—stress sensitivity coefficient, signal strength ratio, and linear fit goodness—the optimal frequency is selected from several preset frequencies.

[0010] Step 4: Repeat step 1, but change the tensile stress of the corroded steel strand sample to A and the number of sections to N, where N is selected according to the actual situation; measure the relative time coordinates of the third peak in all sections of all steel strand samples on the signal acquisition device and take the average value, which is set as S.

[0011] Step 5: Select the 1st to Pth sections of the corroded steel strand to be tested, where P is selected according to the actual situation; measure the relative time coordinates of the 3rd peak of all sections displayed on the signal acquisition device and take the average value, which is Q; measure the diameter change ΔD of the corroded steel strand.

[0012] Step 6: According to SQ=K(AB)+F(ΔD), the existing prestress of the steel strand B=A-[SQF(ΔD)] / K, where F(ΔD) is a function of the influence of diameter change on the relative time coordinate change. If the third peak is not used and other numbered peaks are used for identification, the method is the same as above.

[0013] In a preferred embodiment, the diameter d ij Corrosion delay parameter x ij y ij z ij Stress sensitivity coefficient k ij Signal strength ratio index c ij linearity The steps to obtain it are as follows:

[0014] Step 11: Measure the diameter d at the j-th section of the i-th steel strand sample. ij Repeated multiple times;

[0015] Step 12: Select the excitation signal for guided wave stress identification, with a frequency of f1;

[0016] Step 13: Control the temperature of the test site to be consistent with the temperature of the measurement site, then introduce a pulse signal on one side of the j-th section of the i-th steel strand sample and receive the signal at the other end;

[0017] Step 14: Repeat this step w-1 times to obtain the arrays of stress and relative time coordinates: (1400, t1), (p2, t2), ..., (p... w ,t w ), of which 1400 <p2<p3<……<p w ;

[0018] Step 15: Fit a straight line based on the above sequence. The slope of the straight line is the stress sensitivity coefficient k. ijThe absolute value of the amplitude of the waveguide received signal is c. out The absolute value of the amplitude c of the guided wave transmitted signal in Then the signal frequency f is obtained. i Under the condition of the signal strength ratio index c of the j-th section of the i-th steel strand sample ij =c out / c in ;

[0019] Step 16: Keeping the selected stress constant, measure the diameter of the steel strand obtained by measuring the change in the degree of corrosion of the steel strand, and obtain the array of diameter versus relative time coordinates (d1,t1), (d2,t2), ..., (d... w ,t w To fit a logarithmic relationship curve x based on the changing pattern of this array. ij -y ij ln(Δd ij +z ij );

[0020] Step 17: Similarly, we obtain d. 12 ~d 1n 、……、d m1 ~d mn x 12 ~x 1n ... x m1 ~x mn y 12 ~y 1n ... y m1 ~y mn , z 12 ~z 1n ... z m1 ~z mn k 12 ~k 1n ... k m1 ~k mn , c 12 ~c 1n ... c m1 ~c mn ;

[0021] Step 18: For all d, x, y, z, k, R 2 The excitation frequency f is obtained by averaging the c value and c. i The mean value of all sections of the steel strand sample is used to minimize the influence of random errors.

[0022] In a preferred embodiment, step 3 specifically includes:

[0023] Step 31: Set the stress sensitivity coefficient threshold, and screen out the frequency range corresponding to the stress sensitivity coefficient higher than this threshold;

[0024] Step 32: Set the signal intensity ratio threshold, and screen out the frequency range corresponding to the signal intensity ratio higher than this threshold;

[0025] Step 33: Set the linear fitting goodness-of-fit threshold, and screen out the frequency range corresponding to the linear fitting goodness-of-fit higher than this threshold;

[0026] Step 34: The intersection of the above three regions, that is, f2 < f < f1, is the interval where the optimal frequency is located. Select the preset frequency falling within this interval as the best excitation frequency. If there is more than one best excitation frequency, choose any one; the sensitivity coefficient K corresponding to this frequency is the sensitivity coefficient of the steel strand sample.

[0027] In a preferred embodiment, for steel strands with any diameter, there is:

[0028] v = b(d - d0) + v0 (1)

[0029] d——The current diameter of the steel strand (unit: mm);

[0030] d0——The reference diameter of the steel strand (unit: mm);

[0031] v——The phase velocity of the guided wave signal corresponding to the diameter d of the steel strand (unit: m / s);

[0032] v0——The phase velocity corresponding to the reference diameter d0 of the steel strand (unit: m / s);

[0033] b——The sensitivity coefficient of the guided wave signal phase velocity to the diameter (unit: m·s -1 / mm);

[0034] Equation (1) is transformed into:

[0035]

[0036] The photoelastic effect is a weak effect, so the difference between ν and ν0 is very small, while the difference between d and d0 is relatively large. It can be seen that b is a small quantity compared with ν0 / (d - d0).

[0037] In a preferred embodiment, assume that the guided wave is emitted from one end of the pitch length and received at the other end of the pitch length. The propagation time of the guided wave is t, the propagation distance is l, the phase velocity of the guided wave under the reference diameter is ν0, and the corresponding propagation time is t0. At this time, l = v0t0 = vt, so there is:

[0038]

[0039]

[0040] Since b is a small quantity compared to ν0 / (d-d0), b is ignored in the denominator of equation (5), and equation (5) is transformed into:

[0041]

[0042] In the formula, t-t0 is the peak delay index;

[0043] b, t0, and l are all constants, therefore -bt0 2 / l is also a constant, assuming k cor =-bt0 2 / l;k cor The specific value is determined through numerical calculation;

[0044] Let Δd = d - d0 and Δt = t - t0, then equation (6) transforms into:

[0045] Δt=k cor Δd (7)

[0046] Considering the impact of corrosion product accumulation on the surface of corroded steel strands, the corrosion-induced peak delay index is:

[0047] Δt cor =[x-yln(Δd+z)]-ξk cor Δd (8)

[0048] ξ is the equivalent substitution coefficient between the numerical calculation of the circular bar model and the steel strand, and k cor This is a parameter sensitive to diameter changes in numerical calculation of time delay; in actual measurement, the time delay Δt caused by corrosion is subtracted from the measured peak time delay index Δt = t - t0. cor That is, to obtain the change in stress after considering the effect of corrosion compared to the reference stress state, Δσ = σ - σ0, the expression is:

[0049] Δt=Δt σ +Δt cor =k σ Δσ+[x-yln(Δd+z)]-ξk cor Δd (9)

[0050] Different guided wave excitation frequencies (k values) result in different crest delay stress sensitivities at different frequencies; the optimal excitation frequency should be k, c, and R. 2 Each value corresponds to the frequency of the larger value within its respective numerical range;

[0051] Suppose that for any corroded steel strand, it is necessary to measure its existing prestress σ; take n steel strand samples of the same type, select the optimal excitation frequency f, and measure the relative time coordinates t1, t2, t3...t of the wave peaks in the detection section of the n steel strand samples under the initial corrosion state and initial stress σ0 conditions. n Take their average By changing the corrosion state and stress state of the steel strands, the time delays under different stress levels in different initial corrosion states of the steel strands, as well as the time delays under different corrosion states in different steel strands under the same stress state, were measured. From this, the stress sensitivity coefficients of n steel strand samples were obtained as k1, k2, k3…k n The corrosion delay parameters are x1, x2, x3...x n y1, y2, y3...y n z1, z2, z3...z n Take their average value Then, by applying the optimal excitation frequency f to the object under test, the relative time coordinate t of the wave peak displayed on the signal acquisition device in the test section is obtained, thus obtaining the wave peak delay of the current corrosion state and stress state compared to the initial corrosion state and stress state. The existing prestress σ of the steel strand can be calculated using equation (9).

[0052] Compared with the prior art, the present invention has the following beneficial effects: Based on the waveguide phase velocity, the present invention constructs a waveguide peak time delay index for detecting the stress of corroded steel strands within the gauge length under single or multiple twist pitches of steel strands. This is of great significance for stress detection of existing corroded steel strands in the field of civil engineering, as well as structural safety assessment and subsequent reinforcement based on the detection results. Attached Figure Description

[0053] Figure 1 This is a schematic illustration of the sensitivity index of the peak delay to the stress in a preferred embodiment of the present invention.

[0054] Figure 2 This is a schematic illustration of the relationship between peak delay and diameter variation in a preferred embodiment of the present invention.

[0055] Figure 3 This is a schematic illustration of the signal strength ratio index in a preferred embodiment of the present invention.

[0056] Figure 4 This is a schematic illustration of the goodness-of-fit index between peak delay and stress in a preferred embodiment of the present invention.

[0057] Figure 5 The time-domain diagram (a) and frequency-domain diagram (b) of the five-peak wave at 100kHz are shown in the preferred embodiment of the present invention.

[0058] Figure 6 This is a schematic illustration of the wave peak identification of the waveguide received signal and the transmitted signal, and the concepts of the first, second, third... wave peaks, in a preferred embodiment of the present invention.

[0059] Figure 7 This is a schematic illustration of the preferred excitation frequency of the guided wave in a preferred embodiment of the present invention. Detailed Implementation

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

[0061] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0062] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations according to this application; as used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise; furthermore, it should be understood that when the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0063] refer to Figures 1 to 7 According to the guided wave acoustoelasticity theory, for steel strands under arbitrary corrosion and stress states, we have:

[0064] v=b(d-d0)+v0 (1)

[0065] d—Current diameter of the steel strand (unit: mm);

[0066] d0—Reference diameter of steel strand (unit: mm);

[0067] v — the phase velocity of the guided wave signal corresponding to the diameter d of the steel strand (unit: m / s);

[0068] v0 — Phase velocity corresponding to the reference diameter d0 of the steel strand (unit: m / s);

[0069] b — Sensitivity coefficient of the phase velocity of the guided wave signal to the diameter (unit: m·s) -1 / mm);

[0070] Equation (1) is transformed into:

[0071]

[0072] The acoustic elastic effect is a weak effect, so the difference between ν and ν0 is very small, while the difference between d and d0 is relatively large. Therefore, b is a small quantity compared to ν0 / (d-d0).

[0073] In a preferred embodiment, it is assumed that the guided wave is emitted from one end of the twist length and received from the other end, the wave propagation time is t, the propagation distance is l, the wave phase velocity at the reference diameter is ν0, and the corresponding propagation time is t0. At this time, l = v0t0 = vt, therefore:

[0074]

[0075] Since b is a small quantity compared to ν0 / (d-d0), b can be ignored in the denominator of equation (5), and equation (5) can be transformed into:

[0076]

[0077] In the formula, t-t0 is the peak delay index;

[0078] b, t0, and l are all constants, therefore -bt0 2 / l is also a constant, assuming k cor =-bt0 2 / l;k cor The specific value can be determined through numerical calculation;

[0079] Let Δd = d - d0 and Δt = t - t0, then equation (6) transforms into:

[0080] Δt=k cor Δd (7)

[0081] Considering the impact of corrosion product accumulation on the surface of corroded steel strands, the corrosion-induced peak delay index is:

[0082] Δt cor =[x-yln(Δd+z)]-ξk cor Δd (8)

[0083] The stress sensitivity coefficient k corresponding to the peak delay is determined by referring to the technical implementation method in the inventor's published invention patent CN116678530A. σ The time delay Δt caused by stress change σ ξ is the equivalent substitution coefficient between the numerical calculation circular bar model and the steel strand, k cor This is a parameter sensitive to diameter changes in numerical calculation of time delay; in actual measurement, the time delay Δt caused by corrosion is subtracted from the measured peak time delay index Δt = t - t0. cor This yields the change in stress, considering corrosion, compared to the reference stress state, Δσ = σ - σ0, expressed as:

[0084] Δt=Δt σ +Δt cor =k σ Δσ+[x-yln(Δd+z)]-ξk cor Δd (9)

[0085] The experiment calibrated the value of k corresponding to a certain type of steel strand, and determined the peak delay Δt caused by corrosion changes. cor In actual measurement, by measuring the value of the peak delay index Δt = t - t0, the change in the current stress state compared to the reference stress state under the next corrosion state can be obtained as Δσ = σ - σ0.

[0086] For a given steel strand, different guided wave excitation frequencies k σ Different values, i.e., different frequencies, result in different sensitivities to stress due to peak delay. Therefore, in practical testing, a larger k value is required. σ The value corresponds to the guided wave excitation frequency, but frequency selection also needs to consider other factors. Therefore, this invention proposes a method for selecting the optimal guided wave excitation frequency range for steel strands by comprehensively considering three indicators: the sensitivity of crest delay to stress changes, the signal strength ratio of the received and transmitted guided wave signals, and the goodness of fit between crest delay and stress. The sensitivity of crest delay refers to the sensitivity coefficient k. σ The value of k varies even for the same steel strand; different waveguide excitation frequencies correspond to different k values. σ Value, obviously k σ A higher value is better. Sensitivity metrics for peak delay are shown in the attached table. Figure 1 As shown by the linear fitting line w between the peak delay and stress change, the greater the slope of line w, the higher the sensitivity of the peak delay to stress. The signal strength ratio index c is attached. Figure 3 absolute value of the amplitude c of the middle-guided wave received signal out The absolute value of the excitation signal amplitude c in The ratio of the wave crest delay to the stress varies with the guided wave excitation frequency, and the value of c also varies. The goodness of fit between the wave crest delay and the stress is attached. Figure 4 The coefficient of determination R of the fitted straight line of time delay and stress change shown is... 2 , Figure 4 Which peak in the middle corresponds to Figure 3 Peak number in R 2 The larger the value, the better the linearity of the fit between the peak delay and stress change. Different guided wave excitation frequencies correspond to different R values. 2 Therefore, the optimal excitation frequencies should be k, c, and R. 2 Each value corresponds to the larger value within its respective numerical range.

[0087] Furthermore, assuming that for any corroded steel strand, it is now necessary to measure its existing prestress σ (its initial prestress is known to be σ0), n steel strand samples of the same type can be taken. The optimal excitation frequency f is selected, and the relative time coordinates t1, t2, t3…t of the wave peaks displayed on the signal acquisition device are measured under the initial corrosion state and initial stress σ0 conditions, along the length of the detection section of each of the n steel strand samples (including but not limited to the single-twist or multi-twist pitch of the steel strand). n Take their average By changing the corrosion state and stress state of the steel strands, the time delays under different stress levels in different initial corrosion states of the steel strands, as well as the time delays under different corrosion states in different steel strands under the same stress state, were measured. From this, the stress sensitivity coefficients of n steel strand samples were obtained as k1, k2, k3…k n The corrosion delay parameters are x1, x2, x3...x n y1, y2, y3...y n z1, z2, z3...z n Take their average value Then, by applying the optimal excitation frequency f to the object under test, the relative time coordinate t of the peak displayed on the signal acquisition device under the length of the test section (including but not limited to the single-twist or multi-twist pitch of the steel strand) is obtained, thus obtaining the peak time delay of the current stress state compared to the initial stress state. The existing prestress σ of the steel strand can be calculated using equation (9).

[0088] Example:

[0089] Assume there is a 7-wire prestressed corrosion-resistant steel strand with a nominal diameter of 15.2 mm in the project. The existing prestress is unknown, but the initial stress is known to be A MPa. Let the existing prestress be B MPa. The steps for measuring B are as follows:

[0090] Step 1: Take m corroded steel strand samples of the same type as those in the actual engineering structure. Randomly select n sections from each sample, assuming the length of each section is equal to the single-twist pitch of the steel strand. Calibrate the corrosion delay parameters x, y, z, and stress sensitivity coefficient k at a certain excitation frequency, and obtain the signal strength ratio c and linearity r. 2 Let d be the diameter of the j-th segment of the i-th steel strand sample. ij Corrosion delay parameter x ij y ij z ij Stress sensitivity coefficient k ij Signal strength ratio index c ij and linearity At this point, 1≤i≤m, 1≤j≤n; the diameter d of the first section of the first steel strand sample. 11 Corrosion delay parameter x11 , y 11 , z 11 , stress sensitivity coefficient k 11 , signal strength ratio index c 11 , linearity The first 1 in the subscript represents the 1st steel strand, the second 1 represents the 1st section, and the method for obtaining the indexes of other sections of other steel strand samples is the same. d 11 , x 11 , y 11 , z 11 , k 11 , c 11 and The acquisition steps are as follows:

[0091] (1) Measure the diameter d at the 1st section of the 1st steel strand sample 11 = 14.5 mm, repeat multiple times

[0092] (2) Select the excitation signal for guided wave stress identification. Assume that the narrowband pulse signal shown in Figure 5 is used first, and its frequency is f1.

[0093] (3) Control the temperature of the test field to be the same as that of the measurement site. Tension the stress of the 1st steel strand sample to 1400 MPa (elastic limit of the steel strand), then introduce the pulse signal in (1) on one side of the 1st section of the 1st steel strand sample, and receive the signal at the other end. Assume that the relative time coordinate of the 3rd peak of the guided wave received signal shown in Figure 6 displayed on the signal acquisition device is t1, and the peaks of other serial numbers can also be used.

[0094] (4) Tension the stress of the 1st steel strand sample in the test field to p2 MPa (0 < p2 < 1400). Assume that the relative time coordinate of the 3rd peak of the signal at this time displayed on the signal acquisition device is t2. This step is repeated w - 1 times to obtain the array of stress and relative time coordinates (1400, t1), (p2, t2,), ……, (p w , t w ), where 1400 < p2 < p3 < …… < p w .

[0095] (5) According to the above sequence, a straight line can be fitted, and the slope of the straight line is k 11 . The determination coefficient of the fitted straight line is Assume that the guided wave signal obtained by experiment at a certain excitation frequency in the 1st section of the 1st steel strand sample is as shown in Figure 2 , the absolute value of the amplitude of the guided wave received signal is c out , and the absolute value of the amplitude of the guided wave transmitted signal c inThen, the signal strength ratio index c of the first section of the first steel strand sample under the signal frequency f1 can be obtained. 11 =c out / c in .

[0096] (6) Keeping the selected stress constant, measure the diameter of the steel strand by measuring the change in the degree of corrosion, and obtain the array of diameter versus relative time coordinates: (15.2, t1), (15.15, t2), ..., (14.9, t...). w To fit a logarithmic relationship curve x based on the changing pattern of this array. 11 -y 11 ln(Δd 11 +z 11 );

[0097] (7) Similarly, we can obtain d 12 ~d 1n 、……、d m1 ~d mn x 12 ~x 1n ... x m1 ~x mn y 12 ~y 1n ... y m1 ~y mn , z 12 ~z 1n ... z m1 ~z mn k 12 ~k 1n ... k m1 ~k mn , c 12 ~c 1n ... c m1 ~c mn ;

[0098] (8) For all d, x, y, z, k, R 2 By averaging the values ​​of c and d, the diameter d and corrosion parameters x of the third peak delay of the steel strand sample of this type are obtained when the excitation frequency is f1. f1 y f1 z f1 Stress sensitivity coefficient K f1 Signal strength ratio c f1 and linear fit goodness The average value of all sections of all steel strand samples was taken in order to minimize the influence of random errors.

[0099] Step 2: Change the frequency of the narrowband pulse signal, and repeat Step 1 to obtain the corrosion time delay parameters, stress sensitivity coefficients, signal intensity ratios, and linear fitting goodness-of-fit at several preset frequencies.

[0100] Step 3: By comparing the numerical values of the three indicators of stress sensitivity coefficient, signal intensity ratio, and linear fitting goodness-of-fit, select the optimal frequency from several preset frequencies:

[0101] (1) Set the stress sensitivity coefficient threshold, and screen out the frequency range corresponding to the stress sensitivity coefficient higher than this threshold. Assume the screening result f is as shown in Attachment Figure 7 as f < f1.

[0102] (2) Set the signal intensity ratio threshold, and screen out the frequency range corresponding to the signal intensity ratio higher than this threshold. Assume the screening result f is as shown in Attachment Figure 7 as f2 < f < f3 or f < f4.

[0103] (3) Set the linear fitting goodness-of-fit threshold, and screen out the frequency range corresponding to the linear fitting goodness-of-fit higher than this threshold. Assume the screening result f is as shown in Attachment Figure 7 as f5 < f < f6.

[0104] (4) The intersection of the above three regions, that is, f2 < f < f1, is the interval where the optimal frequency is located. The preset frequency falling within this interval can be taken as the optimal excitation frequency. If there is more than one optimal excitation frequency, any one can be selected. The sensitivity coefficient K corresponding to this frequency is the sensitivity coefficient of the steel strand sample.

[0105] Step 4: Repeat Step 1, but change the tensile stress of the steel strand sample to A, and change the number of sections to N, where N can be selected according to the actual situation. Measure the relative time coordinates of the third wave peak displayed on the signal acquisition device for all sections of all steel strand samples and take the average value, denoted as S.

[0106] Step 5: Take the 1st to Pth sections of the steel strand to be measured for corrosion, where P is selected according to the actual situation; measure the relative time coordinates of the third wave peak for all sections and take the average value, which is Q, and measure the change in diameter ΔD of the steel strand for corrosion.

[0107] Step 6: From Equation (9), using the formula S - Q = K(A - B) + F(ΔD), the existing prestress B of the corroded steel strand can be obtained as B = A - [S - Q - F(ΔD)] / K. If other numbered wave peaks are used instead of the third wave peak for identification, the method is the same.

Claims

1. A method for stress identification of corroded steel strands based on guided wave crest time delay, characterized in that, It includes the following steps: Step 1: Take m samples of corroded steel strands of the same type as those in the actual engineering structure. For each sample, randomly select n sections. The length of each section should be selected according to the testing requirements, including whether the steel strand has a single or multiple twist. Measure the diameter d of the corroded steel strand and calibrate the corrosion delay parameters x, y, z, and stress sensitivity coefficient k at a certain excitation frequency to determine the peak delay. ij And obtain the signal strength ratio index c and the linear fit goodness; let d be the diameter of the j-th segment of the i-th steel strand sample. ij Corrosion delay parameter x ij y ij z ij Stress sensitivity coefficient k ij Signal strength ratio index c ij and linear fit goodness At this point, 1≤i≤m, 1≤j≤n; Step 2: Change the frequency of the narrowband pulse signal, and repeat Step 1 to obtain the corrosion time delay parameters, stress sensitivity coefficients, signal strength ratios, and linear fitting goodness at several preset frequencies; Step 3: By comparing the numerical values of the three indicators of stress sensitivity coefficient, signal strength ratio, and linear fitting goodness, select the best frequency from several preset frequencies; Step 4: Repeat Step 1, but change the tensile stress of the corroded strand sample to A, and change the number of sections to N, where N is selected according to the actual situation; measure and take the average value of the relative time coordinates of the third wave peak displayed on the signal acquisition device in all sections of all strand samples, and set the average value as S; Step 5: Take the 1st to the Pth sections of the corroded strand to be measured, where P is selected according to the actual situation; measure and take the average value of the relative time coordinates of the third wave peak displayed on the signal acquisition device in all sections, and its average value is Q, and measure the change in the diameter of the corroded strand ΔD; Step 6: According to S - Q = K(A - B) + F(ΔD), obtain the existing prestress B of the strand = A - [S - Q - F(ΔD)] / K, where F(ΔD) is a function of the influence of diameter change on the change of relative time coordinates. If other serial number wave peaks are used instead of the third wave peak for identification, the method is the same; The guided wave is emitted from one end of the twist length and received at the other end. The wave propagation time is t, the propagation distance is l, and the wave phase velocity at the reference diameter is V0, corresponding to a propagation time of t0. In actual measurement, the wave crest delay index Δt = t - t0 is measured, and the delay Δt caused by corrosion is subtracted from the value. cor That is, to obtain the change in stress after considering the effect of corrosion compared to the reference stress state, Δσ = σ - σ0, the expression is: (9) ξ is the equivalent substitution coefficient between the numerical calculation of the circular bar model and the steel strand, and k cor This is a parameter sensitive to changes in diameter during numerical calculation delay; This represents the difference between the current diameter d of the steel strand and the reference diameter d0 of the steel strand.

2. The method for stress identification of corroded steel strand based on guided wave crest time delay according to claim 1, characterized in that, diameter d ij Corrosion delay parameter x ij y ij z ij Stress sensitivity coefficient k ij Signal strength ratio index c ij Goodness of linear fit The steps to obtain it are as follows: Step 11: Measure the diameter d at the j-th section of the i-th steel strand sample. ij Repeated multiple times; Step 12: Select the excitation signal for guided wave stress identification, and its frequency is f1; Step 13: Control the temperature of the test site to be the same as that of the measurement site, then introduce a pulse signal on one side of the jth section of the ith strand sample, and receive the signal at the other end; Step 14: Repeat this step w-1 times to obtain the arrays of stress and relative time coordinates: (1400, t1), (p2, t2), ..., (p... w , t w ), of which 1400 <p 2< p 3< ...< p w ; Step 15: Fit a straight line based on the above array. The slope of the line is the stress sensitivity coefficient k. ij The absolute value of the amplitude of the waveguide received signal is c. out The absolute value of the amplitude c of the guided wave transmitted signal in Then the signal frequency f is obtained. i Under the condition of the signal strength ratio index c of the j-th section of the i-th steel strand sample ij =c out / c in ; Step 16: Keeping the selected stress constant, measure the diameter of the steel strand obtained by measuring the change in the degree of corrosion of the steel strand, and obtain the array of diameter versus relative time coordinates (d1,t1), (d2,t2), ..., (d... w , t w To fit a logarithmic relationship curve x based on the changing pattern of this array. ij -y ij ln(Δd ij +z ij ); Step 17: Similarly, we obtain d. 12 ~d 1n 、……、d m1 ~d mn x 12 ~x 1n ... x m1 ~x mn y 12 ~y 1n ... y m1 ~y mn , z 12 ~z 1n ... z m1 ~z mn k 12 ~k 1n ... k m1 ~ k mn , ~ ... ~ c 12 ~c 1n ... c m1 ~c mn ; Step 18: For all d, x, y, z, k, R 2 The excitation frequency f is obtained by averaging the c value and c. i The mean value of all sections of the steel strand sample is used to minimize the influence of random errors.

3. The method for stress identification of corroded steel strand based on guided wave crest time delay according to claim 1, characterized in that, The specific content of Step 3 includes: Step 31: Set the stress sensitivity coefficient threshold, and screen out the frequency range corresponding to the stress sensitivity coefficient higher than this threshold; Step 32: Set the signal strength ratio threshold, and screen out the frequency range corresponding to the signal strength ratio higher than this threshold; Step 33: Set the linear fitting goodness threshold, and screen out the frequency range corresponding to the linear fitting goodness higher than this threshold; Step 34: The intersection of the above three regions, that is, f2 < f < f1, is the interval where the optimal frequency is located. Select the preset frequency falling within this interval as the best excitation frequency. If there is more than one best excitation frequency, choose any one; the sensitivity coefficient K corresponding to this frequency is the sensitivity coefficient of the strand sample.

4. The method for stress identification of corroded steel strand based on guided wave crest time delay according to claim 1, characterized in that, For strands with any diameter, there is: (1) d—the current diameter of the strand, with the unit of mm; d0—the reference diameter of the strand, with the unit of mm; v—the phase velocity of the guided wave signal corresponding to the strand diameter d, with the unit of m / s; v0—the phase velocity corresponding to the reference diameter d0 of the strand, with the unit of m / s; b—the sensitivity coefficient of the guided wave signal phase velocity to the diameter, in m·s. -1 / mm; Equation (1) is transformed into: (2) The acoustoelastic effect is a weak effect, so the difference between v and v0 is very small, while the difference between d and d0 is relatively large. It can be seen that b is a small quantity compared with v0 / (d - d0).

5. The method for stress identification of corroded steel strand based on guided wave crest time delay according to claim 1, characterized in that, The propagation time of the guided wave is t, the propagation distance is l, the phase velocity of the guided wave under the reference diameter is V0, and the corresponding propagation time is t0. At this time, l = v0t0 = vt, so there is: (3) (4) (5) Since b is a small quantity compared with V0 / (d - d0), b is ignored in the denominator of Equation (5), and Equation (5) is transformed into: (6) In the formula, t - t0 is the wave peak time delay index; b, t0, and l are all constants, therefore -bt0 2 / l is also a constant, assuming k cor =-bt0 2 / l;k cor The specific value is determined through numerical calculation; Let Δd = d - d0 and Δt = t - t0, then equation (6) transforms into: (7) Considering the impact of corrosion product accumulation on the surface of corroded steel strands, the corrosion-induced peak delay index is: (8) Different guided wave excitation frequencies (k values) result in different crest delay stress sensitivities at different frequencies; the optimal excitation frequency should be k, c, and R. 2 Each value corresponds to the frequency of the larger value within its respective numerical range; Suppose that for any corroded steel strand, it is necessary to measure its existing prestress σ; take n steel strand samples of the same type, select the optimal excitation frequency f, and measure the relative time coordinates t1, t2, t3...t of the wave peaks in the detection section of the n steel strand samples under the initial corrosion state and initial stress σ0 conditions. n Take their average ; By changing the corrosion state and stress state of the steel strands, the time delays under different stress levels in different initial corrosion states of the steel strands, as well as the time delays under different corrosion states in different steel strands under the same stress state, were measured. From this, the stress sensitivity coefficients of n steel strand samples were obtained as k1, k2, k3…k n The corrosion delay parameters are x1, x2, x3...x n y1, y2, y3...y n z1, z2, z3...z n Take their average Then, using the optimal excitation frequency f on the object under test, the relative time coordinate t of the wave peak displayed on the signal acquisition device in the test section is obtained, and the wave peak delay t_ ... The existing prestress σ of the steel strand can be calculated by equation (9).

Images