A hierarchical damage detection method based on the eigenvalue of the defect state of a periodic structure

Abstract

The present invention discloses a hierarchical damage detection method based on the eigenvalue of the defect state of the periodic structure. A hierarchical damage length recognition model DI is obtained by training a damage detection algorithm. length (n, l) and a hierarchical damage position recognition model #imgabs0#. Mass blocks are periodically arranged on the beam to be detected to construct a periodic structure beam, and the frequency response signal is obtained. Whether there is a defect state frequency in the bandgap of the frequency response curve is used to judge whether there is damage to the structure. For a damaged structure, by extracting the defect state frequency #imgabs1# and comparing it with the damage recognition model, the length and position of the delamination damage can be quickly determined. The present invention can quickly evaluate the length and position of the material delamination damage and is an efficient non-destructive testing method.

Description

Technical Field

[0001] The present invention belongs to the technical field of damage detection, and particularly relates to a hierarchical damage detection method based on the eigenvalue of the defect state of a periodic structure. Background Art

[0002] Due to their low density, high strength and other characteristics, composite materials are increasingly widely used in industrial equipment. However, due to their inherent layered structure, during use, delamination damage inside the structure often occurs due to impact or internal stress fatigue, resulting in relatively large economic losses. For delamination damage, two indicators need to be concerned: the location of the delamination and the length of the delamination. The damage detection technology based on vibration frequency can estimate the damage size / location by monitoring the deviation of modal data between the intact structure and the damaged structure. However, the damage detection technology based solely on frequency shift has disadvantages such as low sensitivity and low accuracy, especially for local small-size damage.

[0003] Different from traditional structures, periodic structures such as phononic crystals have unique vibration bandgap phenomena (frequency ranges where elastic waves cannot propagate). The bandgap is very sensitive to lattice parameters. When there are defects such as unit cell mismatches in the phononic crystal, defect state frequencies can be seen in the bandgap. The length and location of delamination damage have a quasi-linear relationship with the defect state frequency, and a relationship diagram can be obtained by a damage detection algorithm. By comparing with the detected defect state frequency, the length and location of delamination damage can be quickly evaluated.

[0004] Composite beam structures are widely used in the mechanical industry, and the damage detection of beam structures has always attracted people's attention.

[0005] For example, a "thermal image data processing method for detecting delamination damage of composite materials" disclosed in a Chinese patent document, with the publication number CN111429367A, includes first obtaining a sequence of thermal images of a test piece of a composite laminate containing delamination damage; converting each frame of the thermal image into a corresponding gray matrix and extracting the maximum gray value in the gray matrix; then, according to the distribution law of the maximum temperature value of the thermal image sequence, eliminating the thermal images with unstable maximum gray value changes in the thermal image sequence to obtain an optimized thermal image sequence; and then performing thermal image reconstruction on the optimized thermal image sequence to obtain a reconstructed thermal image. After obtaining the reconstructed thermal image, it further includes judging whether there is uneven heating in the reconstructed thermal image, eliminating the influence of uneven heating according to the gray value distribution law of the reconstructed thermal image; performing image segmentation on the thermal image, and finally obtaining a damage image with high signal-to-noise ratio, which can simply and effectively process thermal image data, improve the recognition ability and detection accuracy of delamination damage of composite materials. However, this invention still cannot solve the problems other than the gray value of the thermal image, has low sensitivity, complex steps, and results in low damage detection efficiency. Summary of the Invention

[0006] The present invention mainly aims at the problems of low sensitivity and low accuracy existing in the damage detection technology under the prior art; a hierarchical damage detection method based on the eigenvalue of the defect state of a periodic structure is provided; a hierarchical damage identification model is obtained by training a damage detection algorithm, the frequency response signal of the beam structure to be detected is acquired as the signal to be identified, and whether there is damage to the structure is judged by whether there is a defect state frequency in the bandgap of the frequency response curve. For the damaged structure, the length and position of the layer are determined by comparing the damage identification model with the signal to be identified, enabling rapid evaluation and efficient non-destruction.

[0007] A hierarchical damage detection method based on the eigenvalue of the defect state of a periodic structure, the hierarchical damage detection method comprising the following steps:

[0008] Step S1, periodically arranging mass blocks on the beam without layers to construct a periodic structure beam, and acquiring the corresponding frequency response signal;

[0009] Step S2, extracting the material parameters required for the damage detection algorithm from the frequency response signal obtained in Step S1; marking the two ends of the beam as D1 and D2 respectively in the simulation model;

[0010] Step S3, when exciting at the D1 end, calculating the frequency response curves of the layers at different positions under the condition that the length of each layer is l, and extracting the defect state frequencies in each order of the bandgap, denoted as:

[0011]

[0012] In the above formula, n is the order of the bandgap, x is the position where the layer is located, and l is the length of the layer;

[0013] Step S4, when exciting at the D2 end, calculating the frequency response curves of the layers at different positions under the condition of each layer length, and extracting the defect state frequencies in each order of the bandgap, denoted as:

[0014]

[0015] In the above formula, n is the order of the bandgap, x is the position where the layer is located, and l is the length of the layer;

[0016] Step S5, calculating the maximum and minimum values of the characteristic quantity DI length (n, l) under each layer length l to obtain a hierarchical damage length identification model under each layer length l, and the calculation formula is as follows:

[0017]

[0018] In the above formula, x1 and x M are the positions where the layer is located;

[0019] Step S6: Mark both ends of the beam to be detected as D1 and D2 respectively; apply an excitation signal at D1 and measure the response signal at D2, and obtain the frequency response curve of D2 from the ratio of the response signal to the excitation signal; apply an excitation signal at D2 and measure the response signal at D1, and obtain the frequency response curve of D1 from the ratio of the response signal to the excitation signal.

[0020] Step S7: If a defect state appears inside the bandgap of the frequency response curve, mark the frequencies of each order of defect state as and proceed to Step S8; if no defect state appears, then adjust the period interval of the mass block on the beam to be detected, and enter Step S6. If no defect state still appears after retesting, it is considered that the beam to be detected has no delamination damage.

[0021] Step S8: Compare with the defect state frequency signal obtained by detection through the delamination damage length identification model to identify the length range l = L1~L2 of the delamination.

[0022] Step S9: After determining the delamination length, calculate the characteristic quantity under the nth-order bandgap through the following formula to obtain the delamination damage position identification model at each delamination position x:

[0023]

[0024] Step S10: Compare with the defect state frequency signal obtained by detection through the delamination damage position identification model to identify the position of the delamination.

[0025] The delamination damage identification model is obtained by training the damage detection algorithm. The frequency response signal of the beam structure to be detected is obtained as the signal to be identified. Whether the structure has damage is judged by whether there is a defect state frequency inside the bandgap of the frequency response curve. For the damaged structure, the length and position of the delamination are determined by comparing the damage identification model with the signal to be identified.

[0026] Preferably, the weights of the mass blocks in Step S1 are equal and are fixed on the beam structure to be detected at equal intervals, and the number of arrangements is greater than or equal to 2. Having more than 2 mass blocks can ensure that more accurate frequency data at different positions on the beam can be obtained during the detection process, making the frequency curve more accurately fit the true frequency response signal on the beam.

[0027] Preferably, the marked positions D1 and D2 in Step S2 and Step S6 are the same, and the front and back standards need to be unified.

[0028] Preferably, the frequency response curve calculation described in steps S3 and S4 is obtained by the finite element method or the spectral element method, or can also be measured by multiple groups of experiments. There are various methods to obtain the frequency response curve, and a suitable method can be selected without being limited to a single means. Therefore, this method can be applied to various scenarios.

[0029] Preferably, the frequency response curves described in steps S1, S3, S4, and S6 can be displacement transmissibility curves, strain transmissibility curves, etc., or other transmissibility curves, such as acceleration transfer curves.

[0030] Preferably, for the delamination damage length identification in step S8, the DIs length (nth order) obtained from different order bandgaps are compared, and the intersection of the evaluation lengths is taken as the identification length of the delamination damage.

[0031] Preferably, for the delamination position identification in step S10, the DIs location (nth order) obtained from different order bandgaps are compared, and the intersection of the evaluation positions is taken as the identification position of the delamination damage.

[0032] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0033] The method of the present invention is simple to operate, has high detection efficiency and low detection cost. The frequency response curves are obtained by exciting both sides of the structure to be detected respectively. By comparing the defect state frequencies with the delamination damage identification model, the evaluation of the delamination length and position is realized, overcoming the problem that the existing detection methods need to detect the delamination damage of the structure point by point. It is a fast and efficient non-destructive detection method. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The drawings forming a part of the present invention are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention.

[0035] Figure 1 Comparison of the frequency response curves obtained from the simulation and experiment of the beam without delamination of the present invention;

[0036] Figure 2 Periodic structure beam model of the concentrated mass addition method of the present invention;

[0037] Figure 3 Frequency response curves obtained from the experiment of exciting both sides of the delaminated beam of the present invention;

[0038] Figure 4 Delamination length evaluated based on the delamination damage length identification model of the present invention;

[0039] Figure 5This is the delamination position evaluated based on the delamination location identification model of the present invention;

[0040] Figure 6 This is the flowchart of the detection method of the present invention. Detailed implementation manners

[0041] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0042] The following further details the present invention in conjunction with the accompanying drawings. The process of the detection method of the present invention can adopt the following steps:

[0043] S1. Periodically arrange mass blocks on the carbon fiber composite beam without delamination to construct a periodic structure beam, and obtain the corresponding frequency response signal;

[0044] S2. As Figure 1 shown, the simulation uses the spectral element method to calculate. Extract the material parameters required for the damage detection algorithm from the frequency response signal obtained in step S1. Mark the two ends of the beam as D1 and D2 in the simulation model. Either end can be marked as D1 and the other end as D2;

[0045] S3. When exciting at the D1 end, calculate the frequency response curves of the delamination at different positions under different delamination lengths l, and extract the defect state frequencies within each order of the bandgap, denoted as

[0046] S4. When exciting at the D2 end, calculate the frequency response curves of the delamination at different positions under different delamination lengths, and extract the defect state frequencies within each order of the bandgap, denoted as

[0047] S5. Calculate the maximum and minimum values of the characteristic quantity DI length (n, l) under the second and third order bandgaps for each delamination length l, and obtain the delamination damage length identification model for each delamination length l; for the delamination damage length identification model described in step S5, coefficients and other frequency and other parameters can be added as part of the formula;

[0048] S6. As Figure 2As shown, mark the two ends of the carbon fiber composite beam to be detected as D1 and D2 respectively (the delaminated length of the test piece is 120.0 mm, located at the middle position of the 5th interval). The D1 and D2 ends are the same as those in D1 and D2 in step S2. Apply an excitation signal at D1 and measure the response signal at D2. Obtain the frequency response curve of D2 from the ratio of the response signal to the excitation signal; apply an excitation signal at D2 and measure the response signal at D1. Obtain the frequency response curve of D1 from the ratio of the response signal to the excitation signal;

[0049] S7, as Figure 3 shown, a defect state appears inside the band gap of the frequency response curve. When excited at the D1 end, the defect state frequency in the 2nd order band gap is and in the 3rd order band gap is When excited at the D2 end, the defect state frequency in the 2nd order band gap is and in the 3rd order band gap is Proceed to step S8;

[0050] S8, as Figure 4 shown, evaluate the delaminated length by the delaminated length evaluation parameter DI length (2nd order) = 3456 Hz to evaluate that the delaminated length is in the range of 90.1 - 125.8 mm; by DI length (3rd order) = 7065 Hz to evaluate that the delaminated length is in the range of 118.9 - 139.1 mm. Thus, determine that the evaluation range of the delaminated length is 118.9 - 125.8 mm, with a relatively small error;

[0051] S9, after determining the delaminated length, calculate the characteristic quantities under the 2nd and 3rd order band gaps to obtain the delamination damage position identification model at each delamination position x;

[0052] S10, as Figure 5 shown, evaluate that the delamination is in the 4th or 5th interval of the beam by the delamination position evaluation parameter DI location (2nd order) = 34 Hz; evaluate that the delamination is in the 5th interval of the beam by DI location (3rd order) = 93 Hz. Thus, determine that the delamination is in the 5th interval of the beam, which is consistent with the actual position.

[0053] Among them, steel balls are used as concentrated mass blocks and are pasted on the beam to be detected to achieve point contact.

[0054] The frequency response curve calculation link in step S3 can be obtained by simulation methods such as the spectral element method and finite element method, or can be measured through multiple groups of experiments.

[0055] At the same time, 5 steel balls are evenly arranged on the beam to be detected, forming 6 cycles. The mass ratio of the total mass of the steel balls to the mass of the beam to be detected is 0.6:1.

[0056] In the excitation steps of S1 and S6, white noise random excitation vibration signals are adopted; an optical fiber grating sensing system is used to collect the displacement signals on the response side of the beam to be detected; and the frequency response curve is obtained by calculating the ratio between the input signal and the response signal.

[0057] In step S9, only the delamination damage location identification model at a delamination length of 120 mm is calculated. When the evaluated delamination length range is large, it is necessary to calculate the delamination damage location identification models at different delamination lengths.

[0058] Meanwhile, the material of the beam to be detected is not limited, and it can be a composite material or a metal material, etc.

[0059] The above embodiments only represent one implementation mode of the present invention, and the description thereof is relatively specific and detailed, but it should not be construed as a limitation to the scope of the patent of the present invention. For those of ordinary skill in the art, without departing from the concept of the present invention, several deformations and improvements can still be made, and these all belong to the protection scope of the present invention.

Claims

1. A hierarchical damage detection method based on the eigenvalue of the defect state of a periodic structure, characterized in that The described hierarchical damage detection method includes the following steps: Step S1, periodically arrange mass blocks on the beam without delamination to construct a periodic structure beam, and obtain the corresponding frequency response signal; Step S2, extract the material parameters required for the damage detection algorithm from the frequency response signal obtained in Step S1; mark the two ends of the beam as D1 and D2 in the simulation model; Step S3, when exciting at the D1 end, calculate the frequency response curves of the delamination at different positions for each delamination length l, and extract the defect state frequencies within each order of the bandgap, denoted as: In the above formula, n is the order of the bandgap, x is the position where the delamination is located, and l is the length of the delamination; Step S4, when exciting at the D2 end, calculate the frequency response curves of the delamination at different positions for each delamination length, and extract the defect state frequencies within each order of the bandgap, denoted as: In the above formula, n is the order of the bandgap, x is the position where the delamination is located, and l is the length of the delamination; Step S5, calculate the maximum and minimum values of the feature quantity DI(n, l) under the nth-order bandgap at each layer length l, and obtain the layer damage length identification model at each layer length l. The calculation formula is as follows: length (n, l), and obtain the layer damage length identification model at each layer length l. The calculation formula is as follows: In the above formula, x1 and x M are the positions where the layering is located; Step S6, mark the two ends of the beam to be detected as D1 and D2 respectively; apply an excitation signal at D1 and measure the response signal at D2, and obtain the frequency response curve of D2 from the ratio of the response signal to the excitation signal; apply an excitation signal at D2 and measure the response signal at D1, and obtain the frequency response curve of D1 from the ratio of the response signal to the excitation signal; Step S7. If there are defect states inside the bandgap of the frequency response curve, mark the frequencies of each order of defect states as and proceed to Step S8. If there are no defect states, adjust the period interval of the mass block on the beam to be detected, and enter Step S6. If there are still no defect states after re-testing, it is considered that the beam to be detected has no delamination damage; Step S8, identify the delamination length range l = L1~L2 by comparing the delamination length identification model with the defect state frequency signal obtained by detection; where L1 and L2 are the minimum delamination length and the maximum delamination length respectively; ​ Step S9, after determining the lamination length, calculate the characteristic quantity under the nth-order bandgap through the following formula Obtain the lamination damage position recognition model at each lamination position x: Step S10, identify the delamination location by comparing the delamination damage location recognition model with the detected defective state frequency signal to identify the location of delamination.

2. The hierarchical damage detection method based on the eigenvalue of the defect state of the periodic structure according to claim 1, wherein, The mass blocks described in Step S1 have equal weights and are fixed at equal intervals on the beam structure to be detected, and the number of arrangements is greater than or equal to 2.

3. The hierarchical damage detection method based on the eigenvalue of the defect state of the periodic structure according to claim 1, wherein The marked positions of D1 and D2 described in Step S2 and Step S6 are the same.

4. The hierarchical damage detection method based on the eigenvalue of the defect state of the periodic structure according to claim 1, wherein The calculation of the frequency response curves described in Step S3 and Step S4 is obtained by the finite element method or the spectral element method.

5. The hierarchical damage detection method based on the eigenvalue of the defect state of the periodic structure according to claim 1, characterized in that, The frequency response curves described in Step S1, Step S3, Step S4, and Step S6 are displacement transfer rate curves or strain transfer rate curves.

6. The hierarchical damage detection method based on the eigenvalue of the defect state of the periodic structure according to claim 1, wherein The hierarchical damage length identification described in step S8 compares the DIs length (n-th order) obtained from different order bandgaps, and takes the intersection of the evaluated lengths as the identified length of the delamination damage.

7. The hierarchical damage detection method based on the eigenvalue of the defect state of the periodic structure according to claim 1, wherein The hierarchical position recognition described in step S10 compares the DIs location (n-th order) obtained from different order bandgaps, and takes the intersection of the evaluation positions as the recognition position of the hierarchical damage.

Images