Ground penetrating radar detection and evaluation method suitable for internal defects of rammed earth ruins

Abstract

The invention discloses a ground penetrating radar detection and evaluation method suitable for internal defects of a rammed earth site, and relates to the technical field of internal defect nondestructive testing in historic building cultural relic protection. Comprising the following steps: constructing a dual-frequency dual-polarization humidity closed-loop trigger detection system, establishing a moisture form polarization mass transfer triple coupling decoupling model, and executing defect migration channel weathering state three-target inversion and self-verification; according to the invention, dual-frequency dual-polarization humidity closed-loop detection, a moisture-polarization-mass transfer decoupling model and a three-target inversion self-verification mechanism are utilized to solve the problems that defects and wet area misjudgment are caused by static dielectric hypothesis of the existing ground penetrating radar, a moisture migration channel cannot be identified and a weathering state cannot be quantified; the rammed earth ruin defect and wet area accurate distinguishing, defect parameter inversion, channel identification and weathering quantitative evaluation are realized, the cost is reduced, an accurate basis is provided for a protection scheme, and ruin safety and heritage value are guaranteed.

Description

Technical Field

[0001] This invention relates to the field of non-destructive testing technology for internal defects in the protection of ancient architectural relics, specifically a ground-penetrating radar detection and assessment method applicable to internal defects of rammed earth sites. Background Technology

[0002] As an important cultural heritage in my country, the accurate detection and assessment of internal defects such as cracks and cavities in rammed earth sites is a core prerequisite for their preservation. Ground-penetrating radar (GPR), with its advantages of being non-destructive and efficient, has become the mainstream technology for detecting internal defects in rammed earth sites. Its core principle is based on the difference in electromagnetic parameters between the rammed earth and the defect area, mainly the difference in dielectric constant. By analyzing the time delay, amplitude attenuation, and other characteristics of GPR signals, the location, size, and distribution of defects can be inverted.

[0003] Existing ground-penetrating radar (GPR) detection and assessment methods all rely on the core assumption of a transient linear relationship between dielectric constant and water content. They commonly employ models such as the modified Topp formula to establish this relationship and then interpret the signal response. In real-world engineering scenarios, rammed earth sites often face environmental disturbances such as torrential rain and sudden changes in temperature and humidity. As a porous medium, rammed earth exhibits significant mass transfer lag due to the resistance to pore connectivity during water migration. While surface water content responds rapidly to environmental changes, deeper water content remains high due to the resistance of pore connectivity paths. This results in asynchronous changes in the dielectric constant of rammed earth at different depths, creating a spatial-temporal lag.

[0004] However, existing technologies fail to recognize this dynamic hysteresis effect, still processing ground-penetrating radar signals based on static dielectric parameter assumptions. This results in the signal time delay simultaneously incorporating path differences caused by defects and dielectric constant deviations due to moisture content hysteresis. Existing wavefield separation techniques can only separate reflected waves, diffracted waves, and noise, failing to decouple time delay components from different physical causes. Consequently, this leads to significant errors in defect depth localization, frequently resulting in misjudgments where deep cavities are detected but excavated areas are merely moist. Such misjudgments directly distort defect assessment results, misleading subsequent protection plans and even causing over-repair or under-protection, seriously threatening the structural safety and heritage value of rammed earth sites. Existing technologies have yet to solve this core problem. Therefore, this paper proposes a ground-penetrating radar detection and assessment method suitable for internal defects in rammed earth sites to overcome the aforementioned issues. Summary of the Invention

[0005] The purpose of this invention is to provide a ground-penetrating radar detection and assessment method applicable to internal defects of rammed earth sites, so as to solve the problems mentioned in the background art.

[0006] To address the aforementioned technical problems, this invention provides a ground-penetrating radar detection and assessment method for internal defects in rammed earth archaeological sites, comprising the following steps:

[0007] Construct a dual-frequency, dual-polarization humidity closed-loop triggering detection system: Modify a conventional ground-penetrating radar by adding dual fixed-frequency switching units to output 100MHz and 300MHz signals, retaining the horizontal and vertical polarization switching function and synchronously linking it with frequency switching, and adding a dual-frequency phase difference discrimination unit at the receiving end; complete the detection data acquisition through three stages: establishing the reference state dual-frequency polarization response, co-triggered dual-frequency polarization for humidity disturbance, and fine scanning of high-response areas;

[0008] A triple coupling and decoupling model for water morphology polarization mass transfer was established: Based on the detection data, a water morphology discrimination criterion was established to distinguish between the free water-dominated region, the bound water-dominated region, and the mixed water region. A total time delay decoupling equation was constructed, which includes the free water migration dielectric deviation Δεwfzt, the bound water static dielectric εwbz, and the defect path difference ΔLz. The defect path difference delay ΔTdz, the free water migration dielectric deviation delay ΔTwfzt, and the bound water static dielectric delay ΔTWBz were separated.

[0009] Three-objective inversion and self-verification of weathering state of defect migration channels: Based on the inversion of defect parameters obtained by decoupling ΔTdz, moisture migration channels are identified by the time-series change rate of moisture morphology characterization values. Weathering state is evaluated by combining the proportion of bound water-dominated areas and the density of migration channels. Self-verification is completed through cross-validation and error correction.

[0010] Furthermore, the modification and three-stage process of the dual-frequency dual-polarization humidity closed-loop trigger detection system are as follows:

[0011] Detection system upgrade: The dual fixed frequency switching unit consists of a crystal oscillator, an analog switch, and a control circuit. The analog switch switches between 100MHz and 300MHz signals at a frequency of 0.3Hz. At the same detection point, four sets of signals are acquired simultaneously: 100MHz horizontal polarization, 100MHz vertical polarization, 300MHz horizontal polarization, and 300MHz vertical polarization. The receiving end acquires the amplitude and phase difference of the four sets of signals.

[0012] The three-stage process is as follows: the baseline establishment stage uses a 10cm×10cm grid scan with a 0.5-second pause at each point; the humidity disturbance stage uses a micro-spray device with a 2-meter spacing to perform intermittent spraying; and the high-response area triggers a fine scan.

[0013] Furthermore, the specific process for establishing the baseline dual-frequency polarization response is as follows: A scanning detection area is selected during a stable environmental period of three consecutive days without sudden temperature or humidity changes. A baseline response matrix P0(z) is constructed, containing A100H0(z), A100V0(z), Δφ1000(z), A300H0(z), A300V0(z), and Δφ3000(z). Based on P0(z), baseline parameters are calculated: Initial value of pore connectivity resistance:

[0014] R0(z)=k×Δφ1000(z) / A100H0(z),

[0015] Where k is a proportionality constant calibrated by linear fitting of indoor rammed soil samples; the initial moisture speciation value η0(z) = Δφ3000(z) / Δφ1000(z).

[0016] Furthermore, the specific process of humidity disturbance dual-frequency polarization synergistic triggering is as follows: the micro-spray device adopts an intermittent mode of spraying for 30 seconds and pausing for 120 seconds, which is repeated 3 times, with a total spray volume of 5 mL / cm³. 2 Scanning was performed at four time points after the disturbance: 1 hour after the disturbance (surface free water migration period), 6 hours after the disturbance (middle layer free water response period), 24 hours after the disturbance (deep layer free water infiltration period), and 72 hours after the disturbance (bound water stabilization period). The time-series response matrix Pt(z) was obtained, and the real-time water speciation characterization value ηt(z) = Δφ300t(z) / Δφ100t(z) was calculated.

[0017] Furthermore, the triggering conditions and parameters for fine scanning in the high-response region are as follows: fine scanning is triggered when |(ηt(z)-η0(z)) / η0(z)|>40% or |(Rt(z)-R0(z)) / R0(z)|>30% at a certain depth z; fine scanning uses a 5cm×5cm grid, and the dwell time for four sets of signal acquisition is extended to 1 second, where Rt(z) is the real-time pore connectivity resistance.

[0018] Furthermore, the specific content of the triple coupling and decoupling model of water speciation polarization mass transfer is as follows:

[0019] Water form identification criteria: η(z,t)>1.5 is the free water dominant region, η(z,t)<1.0 is the bound water dominant region, and 1.0≤η(z,t)≤1.5 is the mixed water region;

[0020] The triple coupling and decoupling equations are as follows: Total time delay ΔT(z,t) = ΔTd(z) + ΔTwf(z,t) + ΔTWB(z), where ΔTd(z) = ΔL(z) × √ε0(z) / c, c is the speed of light, ε0(z) is the initial dielectric constant of the rammed earth, ΔTwf(z,t) = z × √(ε0(z) + Δεwf(z,t)) / c, ΔTWB(z) = z × √(ε0(z) + εwb(z)) / c, Δεwf(z,t) = C1 × ηt(z) × exp(-t / τz), C1 is the free water polarization coefficient, and τz = Rt(z) × z 2 / μ, μ is the soil moisture diffusion coefficient, εwb(z)=C2×Δφ1000(z), C2 is the bound water polarization coefficient.

[0021] Furthermore, the specific criteria for the three-objective inversion are as follows:

[0022] Defect inversion: The following conditions must be met: ΔTd(z) is a fixed value and the relative error with the fine scan calculation value ΔTdcal(z) is ≤5%, and the corresponding region ηt(z) is ≤1.0; the three-dimensional coordinates, size, shape and confidence level Cd of the defect are inverted. Cd≥0.96 is a first-level confidence level, and 0.92≤Cd<0.96 is a second-level confidence level.

[0023] Water migration channel identification: The region in the free water-dominated area with a temporal change rate of ηt(z) > 20% / h is the migration channel, and the channel width W = 0.1 × ln(ηmax / ηmin), where ηmax and ηmin are the maximum and minimum values ​​of ηt(z) in this region, respectively.

[0024] Weathering status assessment: The area dominated by bound water accounts for <60% and the density of migration channels is >0.3 channels / m². 2 It is a strongly weathered area; the area dominated by bound water accounts for 60% to 85% and has 0.1 veins / m. 2 ≤Migration channel density≤0.3 channels / m 2 It is a moderately weathered area; the area dominated by bound water accounts for >85% and the density of migration channels is <0.1 channels / m². 2 It is a weakly weathered area.

[0025] Furthermore, the specific process of the self-verification mechanism is as follows:

[0026] Cross-validation: The defect region must satisfy the full-time ηt(z) ≤ 1.0 and ΔTd(z) is fixed; the migration channel must satisfy the temporal variation of ηt(z) conforming to the law of exp(-t / τz);

[0027] Error correction: If the defect confidence level Cd < 0.92, the adjacent grid data is called to correct ΔTd(z) through a linear interpolation algorithm, and the defect is re-evaluated after correction.

[0028] Compared with the prior art, the beneficial effects of the present invention are:

[0029] 1. The upgraded ground-penetrating radar is equipped with dual fixed-frequency switching units and synchronously linked polarization switching. It can simultaneously acquire the amplitude and phase difference of four sets of signals at the same detection point. Combined with the three-stage process of baseline state establishment, humidity disturbance, and fine scanning, it can accurately capture changes in water morphology and pore connectivity resistance at different depths and time points. The baseline state establishes initial parameters (R0(z), η0(z)) through a 3-day stabilization period. The humidity disturbance acquires dynamic data through four key time points (1h surface free water period, 6h middle layer free water period, 24h deep free water period, and 72h bound water stabilization period). A 5cm×5cm fine scan (1-second dwell) is triggered in high-response areas (change rate exceeding 40% / 30%) to avoid subsequent misjudgments due to insufficient data accuracy.

[0030] 2. The detection system is based on the modification of conventional ground penetrating radar. Only a dual-frequency switching unit consisting of a crystal oscillator and an analog switch needs to be added. There is no need to replace the wideband sweep frequency equipment. The cost is significantly reduced after the modification. The operation process does not require the participation of professional algorithm engineers. Conventional technicians can master it after simple training. It is suitable for on-site detection of various rammed earth sites.

[0031] 3. Existing technologies cannot separate the defect path difference delay from the dielectric deviation delay caused by moisture content hysteresis. This model solves this problem through two core design features:

[0032] First, establish water form identification criteria based on η(z,t) values ​​to clarify the water types in different regions;

[0033] Then, a triple coupling decoupling equation (ΔT(z,t)=ΔTd(z)+ΔTwf(z,t)+ΔTWB(z)) is constructed to separate the defect path difference delay ΔTd(z) which is independent of water, the dielectric deviation delay ΔTwf(z,t) which varies with the migration of free water, and the fixed bound water dielectric delay ΔTWB(z).

[0034] At a depth of 3m, η = 0.8 (dominated by bound water), and after decoupling, ΔTd = 1.3ns (defect exists); at a depth of 3.5m, η = 1.6 (dominated by free water), and after decoupling, ΔTd = 0 (only a humid area). This completely eliminates the misjudgment of "detecting a deep cavity that is actually a humid area". The accuracy of defect and humid area identification fully meets the standard, and the positioning error is controlled within the preset accuracy range.

[0035] 4. Defect inversion employs a dual standard: fixed ΔTd + relative error ≤ 5% + ηt ≤ 1.0. The inversion results include the defect's three-dimensional coordinates, size, morphology, and confidence level (Cd ≥ 0.96 for Level 1 confidence, 0.92 ≤ Cd < 0.96 for Level 2 confidence), avoiding ambiguous defect assessments. At a depth of 3m, the defect has a confidence level of Cd = 0.97 (Level 1 confidence), accurately inverting its coordinates (5.2m, 3.1m, 3m), size (0.5m × 0.4m × 0.3m), and morphology (void), providing precise parameters for targeted repair.

[0036] 5. By identifying the temporal change rate of ηt(z) within the free water-dominated zone (>20% / h), the system can also calculate the channel width (W=0.1×ln(ηmax / ηmin)) and extension depth, overcoming the limitation of existing technologies that "can only detect existing defects." For example, in the embodiment, two migration channels (0.3-0.4m wide, extending to 1.8m) were identified in the surface layer 0-0.8m, which can guide the design of targeted drainage systems and prevent water accumulation from causing new cracks or cavities.

[0037] 6. Based on the proportion of the water-dominant area and the density of migration channels, the weathering state is divided into three levels (strong weathering: proportion <60% and density >0.3 channels / m²). 2 Moderate weathering: 60% ≤ percentage ≤ 85% and 0.1 ≤ density ≤ 0.3 streaks / m 2 Weakly weathered: proportion >85% and density <0.1 stripes / m² 2 This allows for a quantitative assessment of the degree of weathering. For example, if the area identified in the embodiment is a moderately weathered zone, the protection plan offers a dual approach: repairing existing cavities and optimizing reinforcement materials to slow down weathering.

[0038] 7. Further elimination of misjudgment risk is achieved through cross-validation (the defect must satisfy the full-time ηt≤1.0 and ΔTd fixed; the migration channel must conform to the exp(-t / τz) variation law) and error correction (when Cd<0.92, linear interpolation of adjacent grids is used to correct ΔTd), as demonstrated in the example where cross-validation at a depth of 3m (full-time ηt=0.8≤1.0, ΔTd error 3%) passed, ensuring the reliability of the results.

[0039] 8. This method, by accurately identifying defects, migration channels, and weathering conditions, can avoid over-repairing damp areas (e.g., not needing to reinforce non-defective damp areas) and under-protecting genuine defects (e.g., not overlooking deep cavities); at the same time, it optimizes the selection of reinforcement materials and the scope of reinforcement based on the evaluation results, significantly extending the structural life of rammed earth sites and maximizing the protection of their historical and structural value as cultural heritage. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of a ground-penetrating radar detection and assessment method for internal defects of rammed earth sites according to the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Please see Figure 1 The present invention provides a technical solution:

[0043] See Figure 1 As shown, an embodiment of a ground-penetrating radar detection and assessment method suitable for internal defects of rammed earth sites is presented:

[0044] I. Construction of a Dual-Frequency Dual-Polarization Humidity Closed-Loop Triggering Detection System:

[0045] 1. Detection system upgrade:

[0046] The detection system is based on a low-cost modification of conventional ground-penetrating radar. The modifications include:

[0047] A dual fixed frequency switching unit is added. This unit consists of a crystal oscillator, an analog switch, and a control circuit. The crystal oscillator generates fixed frequency signals of 100 MHz and 300 MHz respectively. The 100 MHz signal is a low-frequency signal that is sensitive to bound water, while the 300 MHz signal is a high-frequency signal that is sensitive to free water. The analog switch, driven by the control circuit, switches between the two frequency signals at a frequency of 0.3 MHz to avoid mutual interference between different frequency signals.

[0048] The horizontal and vertical polarization switching function is retained, and the polarization switching and frequency switching are synchronized. At the same detection point, four sets of signals are acquired simultaneously: 100 MHz horizontal polarization, 100 MHz vertical polarization, and 300 MHz horizontal polarization.

[0049] The receiver is equipped with a dual-frequency phase difference discrimination unit, which consists of a signal amplifier, a filter, and a phase comparator. It synchronously acquires the amplitude and phase difference of four sets of signals. The amplitudes of the four sets of signals are denoted as A100, HA100, V300, HA300, and V, respectively. The phase differences are denoted as Δφ100 and Δφ300, respectively. Δφ100 is the phase difference between the 100 MHz horizontally polarized signal and the vertically polarized signal, and Δφ300 is the phase difference between the 300 MHz horizontally polarized signal and the vertically polarized signal.

[0050] 2. Three-stage closed-loop detection process:

[0051] First-stage establishment of the reference state dual-frequency polarization response:

[0052] During a stable environmental period of three consecutive days without sudden changes in temperature and humidity, the detection area was scanned using a 10 cm x 10 cm grid. Each detection point was paused for 0.5 seconds, and four sets of signals were collected simultaneously to construct a baseline response matrix Pzeroz. Pzeroz contains six parameters: A100HzerozA100VzerozΔφ100zA300HzerozA300VzerozΔφ300z. z represents the depth range of 0 to 5 meters.

[0053] Calculate two key baseline parameters based on the baseline response matrix:

[0054] ① The initial value of pore connectivity resistance is Rzeroz. The calculation formula is Rzeroz equals k multiplied by Δφ100z divided by A100H0z, where k is a proportionality constant. The calibration method is to select a sample with the same material as the rammed earth in the detection area, measure the dual-frequency phase difference and amplitude under different pore connectivity resistance conditions in the room, and obtain the value of k through linear fitting.

[0055] ② The initial water form characterization value ηzeroz is calculated as ηzeroz equal to Δφthree hundred and zeroz divided by Δφone hundred and zeroz. The larger the η value, the higher the proportion of free water.

[0056] Second-stage humidity disturbance dual-frequency polarization coordinated triggering:

[0057] Four micro-spray devices were placed on the surface of the exploration area, with a spacing of two meters between the devices. The intermittent spraying mode was adopted, with specific parameters of spraying for 30 seconds and pausing for 120 seconds, repeated three times, and a total spray volume of 5 milliliters per square centimeter. This spray volume can simulate moderate rainfall and will not cause damage to the site.

[0058] At four key time points—one hour after disturbance (surface free water migration period), six hours after disturbance (middle layer free water response period), twenty-four hours after disturbance (deep layer free water infiltration period), and seventy-two hours after disturbance (bound water stabilization period)—the same grid area was repeatedly scanned to obtain the time-series response matrix Ptz. Ptz contains six parameters: A100HtzA100VtzΔφ100tzA300HtzA300VtzΔφ300tz. The real-time water speciation characterization value ηtz was calculated simultaneously using the formula ηtz equal to Δφ300tz divided by Δφ100tz.

[0059] Third-stage high-response region fine scanning:

[0060] Set response judgment criteria. If the absolute value of the change rate of the water morphology characterization value at a certain depth z is greater than 40%, or the absolute value of the change rate of the pore connectivity resistance is greater than 30%, then a fine scan will be automatically triggered. The change rate of the water morphology characterization value is the difference between ηtz and ηzeroz divided by ηzeroz, and the change rate of the pore connectivity resistance is the difference between Rtz and Rzeroz divided by Rzeroz.

[0061] The fine scanning uses a 5 cm by 5 cm grid, and the dwell time for four sets of signal acquisition is extended to one second, improving the accuracy of data acquisition.

[0062] II. Establishment of a triple coupling and decoupling model for water speciation polarization mass transfer:

[0063] 1. Precise identification of water form:

[0064] Criteria for identifying moisture speciation were established, and these criteria were obtained through calibration of indoor rammed soil samples.

[0065] The criterion for determining the free water-dominated region is that ηzt is greater than 1.5. In this region, the high-frequency phase difference is significantly greater than the low-frequency phase difference, indicating that the free water responds strongly to high-frequency polarization waves.

[0066] The criterion for determining the dominant region of bound water is that ηzt is less than 1.0. In this region, the low-frequency phase difference is dominant, indicating that bound water only responds to low-frequency polarization waves.

[0067] The criteria for determining a mixed water zone are that ηzt is greater than or equal to 1.0 and less than or equal to 1.5, indicating that free water and bound water coexist in this zone;

[0068] Based on the time series ηtz, a dynamic distribution map of water speciation in the detection area was drawn, and the spatial location and range of free water migration paths and bound water enrichment areas were marked on the distribution map.

[0069] 2. Construction of triple coupling decoupling equations:

[0070] Define three core parameters:

[0071] The free water migration dielectric deviation Δεwfzt exists only in the free water-dominated region and is positively correlated with ηzt. The calculation formula is Δεwfzt equal to C - multiplied by ηzt multiplied by the exponential function exp(-t) divided by τz, where C - is the free water polarization coefficient, obtained through indoor rammed soil sample calibration, and τz is the migration lag time, derived from the pore connectivity resistance Rtz. The derivation formula is τz equal to Rtz multiplied by z squared divided by μ, where μ is the rammed soil moisture diffusion coefficient.

[0072] The static dielectric constant of bound water, εwbz, is a fixed value in the bound water-dominant region. The calculation formula is εwbz equal to C2 multiplied by Δφ100z, where C2 is the polarization coefficient of bound water, which is obtained by calibration through indoor rammed soil samples.

[0073] The defect path difference ΔLz is a fixed value and is independent of the moisture form.

[0074] The total time delay decoupling formula is ΔTzt equal to ΔTdz plus ΔTwfzt plus ΔTWBz, where:

[0075] ΔTdz is the defect path difference delay, which is independent of time t. The calculation formula is ΔTdz equal to ΔLz multiplied by the square root of εzeroz and then divided by c, where c is the speed of light and εzeroz is the initial dielectric constant of the rammed earth.

[0076] ΔTwfzt is the dielectric deviation delay of free water migration, which varies with time t. The calculation formula is: ΔTwfzt equals z multiplied by the square root of εzeroz and Δεwfzt divided by c, minus z multiplied by the square root of εzeroz divided by c.

[0077] ΔTWBz is the static dielectric delay of bound water, which is a fixed value. The calculation formula is: ΔTWBz equals z multiplied by the sum of εzeroz and εwbz divided by c, minus z multiplied by the sum of εzeroz divided by c.

[0078] Decoupling process:

[0079] The first step is to extract εzeroz and Δφ10z from the reference response matrix Pzeroz, substitute them into the εwbz calculation formula to obtain εwbz, and then substitute them into the ΔTWBz calculation formula to obtain ΔTWBz.

[0080] The second step is to obtain ηtz based on the time-series response matrix Ptz, determine the free water dominant area through the water form identification criterion, substitute it into the Δεwfzt calculation formula to obtain Δεwfzt, and then substitute it into the ΔTwfzt calculation formula to obtain ΔTwfzt.

[0081] The third step is to directly extract the total delay ΔTzt from the radar signal, substitute it into the total time delay decoupling formula, solve for ΔTdz, and then substitute it into the ΔTdz calculation formula to obtain ΔLz.

[0082] III. Three-objective inversion and self-verification of weathering state in defect migration channels:

[0083] 1. Precise defect inversion:

[0084] Defect determination adopts a dual verification standard:

[0085] The first standard is that ΔTdz is a fixed value, and the absolute value of the relative error between ΔTdz and the value ΔTdcalz calculated by the second fine scan is less than or equal to five percent;

[0086] The second criterion is that the corresponding region ηzt is less than or equal to 1.0, that is, the region is dominated by bound water and there is no free water migration, thus eliminating the misjudgment of humid areas;

[0087] Defect parameter calculation: Substitute ΔLz into the electromagnetic wave propagation formula to obtain the three-dimensional coordinates xyz of the defect, its length, width, height, and shape (cavities or cracks). At the same time, calculate the defect confidence level Cd. Cd greater than or equal to 0.96 is a level 1 confidence level, and Cd greater than or equal to 0.92 and less than 0.96 is a level 2 confidence level.

[0088] 2. Inversion of moisture migration channels and weathering state:

[0089] Water migration channel identification: In the free water-dominated region, areas with a temporal change rate of ηzt greater than 20 percent per hour are identified as water migration channels. A three-dimensional distribution map of the migration channels is drawn, and the channel width and extension depth are marked. The channel width is calculated by the formula W = zero-one multiplied by the natural logarithm lnηmax divided by ηmin, where ηmax is the maximum value of ηzt in the region and ηmin is the minimum value of ηzt in the region.

[0090] Weathering status prediction:

[0091] The criteria for identifying a strongly weathered area are that the area dominated by bound water accounts for less than 60%, and the density of migration channels is greater than 0.3 per square meter. This area has a loose pore structure and is easily permeable by water.

[0092] The criteria for determining a moderately weathered area are that the area dominated by bound water accounts for 60% or more and 85% or less, and the density of migration channels is greater than or equal to 0.11 per square meter and less than or equal to 0.3 per square meter.

[0093] The criteria for identifying a weakly weathered area are that the area dominated by bound water accounts for more than 85%, and the density of migration channels is less than one per square meter. This area has a dense pore structure, making it difficult for water to migrate.

[0094] 3. Self-verification mechanism:

[0095] Cross-validation: The defect region must simultaneously satisfy two conditions: the full-time ηzt must be less than or equal to one point zero and ΔTdz must be a fixed value. The migration channel must satisfy the condition that the temporal change of ηzt conforms to the law of exp(negativet) divided by τz.

[0096] Error correction: If the defect confidence level Cd in a certain area is less than 0.92, the data of the adjacent grid is automatically called and ΔTdz is corrected by linear interpolation algorithm. After correction, the defect judgment is re-performed.

[0097] For example:

[0098] The scenario for this example is a rammed earth city wall site from the Tang Dynasty. The exploration area is 10 meters by 6 meters. Historical explorations of this area have resulted in misjudgments that deep cavities were actually damp areas. This exploration requires accurate identification of defects and structural conditions.

[0099] I. Dual-frequency dual-polarization humidity closed-loop trigger detection system execution:

[0100] The baseline detection employed a modified dual-frequency dual-polarization ground-penetrating radar with output frequencies of 100 MHz and 300 MHz, a scanning grid of 10 cm x 10 cm, and a scanning depth of 0 to 4 m. The initial value of pore connectivity resistance, Rzeroz, was calculated as follows: for the surface layer (0 to 1 m), Rzero was 1.2 x 105 Pascals per second (Pasecond / m); for the middle layer (1 to 2.5 m), Rzero was 2.5 x 105 Pasecond / m; and for the deep layer (2.5 to 4 m), Rzero was 4.0 x 105 Pasecond / m. The initial water morphology characterization value, ηzeroz, was calculated as follows: for the surface layer, ηzero was 1.3, indicating mixed water; and for the middle and deep layers, ηzero was 0.8, indicating bound water.

[0101] Humidity disturbance was achieved using an intermittent spraying mode, with a 30-second spray followed by a 120-second pause, repeated three times, with a total spray volume of 5 ml per square centimeter. Four dynamic measurements were performed at 1 hour, 6 hours, 24 hours, and 72 hours after the disturbance. The results showed that after 1 hour, the surface ηt was 1.8, indicating a free water-dominated zone; after 6 hours, the middle layer ηt was 1.4, indicating a mixed water zone; after 24 hours, the deep layer ηt was 1.1, indicating a mixed water zone; and after 72 hours, the entire region ηt was 0.9, indicating a stable state of bound water.

[0102] High-response area fine scan trigger: at a depth of three meters, the ηtz change rate is 52 percent, which is greater than 40 percent, automatically triggering a five-centimeter by five-centimeter fine scan, and the dwell time of the four sets of signal acquisitions is extended to one second.

[0103] II. Decoupling of the triple coupling of water speciation polarization and mass transfer:

[0104] Moisture form identification results: at a depth of 3 meters, the ηt value after 72 hours is 0.8, indicating a bound water-dominated zone; at a depth of 3.5 meters, the ηt value after 72 hours is 1.6, indicating a free water-dominated zone, i.e., a humid zone.

[0105] Decoupling calculation: At a depth of 3 meters, ΔTwfzt is zero, ΔTWBz is 0.2 nanoseconds, and the total delay ΔTzt is 1.5 nanoseconds. Substituting into the decoupling formula, we get ΔTd as 1.3 nanoseconds. Further calculation yields ΔL as 0.5 meters. At a depth of 3.5 meters, ΔTwfzt is 0.4 nanoseconds, ΔTWBz is 0.2 nanoseconds, and ΔTd is zero. Therefore, this area is determined to be defect-free.

[0106] III. Three-objective inversion and self-verification of weathering state in defect migration channels:

[0107] Defect inversion: At a depth of three meters, ΔTd is a fixed value and ηzt is less than or equal to 1.0. The defect confidence level Cd is 0.97, which belongs to the first level confidence. The inversion yields the three-dimensional coordinates of the defect as 5.2 meters, 3.1 meters and 3 meters, the size as 0.5 meters by 0.4 meters by 0.3 meters, and the morphology as a cavity.

[0108] Migration channel inversion: There are two water migration channels in the surface layer from 0 to 0.8 meters, with a channel width of 0.3 to 0.4 meters and a depth extending to the middle layer of 1.8 meters.

[0109] Weathering status prediction: The area dominated by bound water accounts for 78%, and the migration channel density is 0.2 channels per square meter, indicating that the area is a medium weathering zone.

[0110] Self-verification: At a depth of 3 meters, the full-time sequence ηzt is less than or equal to 1.0, and the ΔTd error is 3%, thus the verification is passed; at a depth of 3.5 meters, the ηt time sequence change conforms to the exp negative t divided by 48 rule, thus it is determined that there are no defects in the humid area.

[0111] Summarize:

[0112] This method achieves accurate identification of moisture morphology by using the phase difference ratio of dual fixed-frequency dual-polarization signals. Combined with a triple coupling and decoupling model, it separates the defect path difference delay, free water migration dielectric deviation delay, and bound water static dielectric delay from a physical perspective. The accuracy of defect and humid area identification fully meets the standard, and the positioning error is controlled within the preset accuracy range. This completely eliminates the problem of misjudging deep cavities as humid areas in existing technologies, and provides a precise basis for the formulation of subsequent protection schemes.

[0113] This method uses two quantitative indicators—the proportion of water-dominated areas and the density of migration channels—to achieve a graded assessment of the weathering degree of rammed earth sites, clearly distinguishing between strongly weathered, moderately weathered, and weakly weathered areas, thus providing a dual basis for protection plans to repair existing defects and prevent further weathering.

[0114] This method accurately identifies the three-dimensional distribution, width, and depth of water migration channels by measuring the temporal change rate of water speciation values ​​in the free water-dominated zone. This can guide protection personnel to design targeted drainage systems in advance, preventing water accumulation from causing new defects and overcoming the limitation of existing technologies that can only detect existing defects.

[0115] The detection system is based on a conventional ground-penetrating radar. The dual fixed-frequency switching unit only requires the addition of a simple frequency switching circuit, resulting in a significantly lower equipment cost compared to broadband sweep frequency equipment. The operation process does not require the involvement of specialized algorithm engineers; ordinary technicians can master it after simple training. It is suitable for various on-site detection scenarios of rammed earth archaeological sites.

[0116] This method avoids over-repairing of damp areas and insufficient protection of actual defects. At the same time, based on the assessment results of weathering status and the distribution of migration channels, it can optimize the selection of reinforcement materials and the design of reinforcement range, significantly extend the structural life of rammed earth sites, and maximize the protection of the structural safety and heritage value of rammed earth sites.

Claims

1. A method for ground penetrating radar detection and evaluation of internal defects in rammed earth sites, comprising: The method comprises the following steps: A dual-frequency dual-polarization humidity closed-loop triggering detection system is constructed: a conventional ground penetrating radar is modified, a dual-fixed-frequency switching unit is added to output 100 MHz and 300 MHz signals, the horizontal and vertical polarization switching functions are retained and are synchronized and linked with the frequency switching, and a dual-frequency phase difference discrimination unit is added to the receiving end; detection data acquisition is completed in three stages through three stages of establishing a reference state dual-frequency polarization response, humidity disturbance dual-frequency polarization collaborative triggering and fine scanning of a high-response area; A water morphology polarization mass transfer triple-coupling decoupling model is established: a water morphology discrimination criterion is established based on the detection data to distinguish a free water dominant area, a bound water dominant area and a mixed water area, a total time delay decoupling equation containing a free water migration dielectric deviation Δεwfzt, a bound water static dielectric εwbz and a defect path difference ΔLz is constructed, and a defect path difference delay ΔTdz, a free water migration dielectric deviation delay ΔTwfzt and a bound water static dielectric delay ΔTWBz are separated; Three target inversions of defect migration channel weathering state and self-verification are performed: defect parameters are inverted based on the decoupling, water migration channels are identified through the time sequence change rate of the water morphology characterization value, the weathering state is evaluated in combination with the proportion of the bound water dominant area and the migration channel density, and self-verification is completed through cross-verification and error correction.

2. The method for detecting and evaluating internal defects of rammed earth sites using ground penetrating radar according to claim 1, wherein: The modification of the dual-frequency dual-polarization humidity closed-loop triggering detection system and the three-stage process are as follows: Modification of the detection system: the dual-fixed-frequency switching unit is composed of a crystal oscillator, an analog switch and a control circuit, and the analog switch switches to output 100 MHz and 300 MHz signals at a frequency of 0.3 Hz; 100 MHz horizontal polarization, 100 MHz vertical polarization, 300 MHz horizontal polarization and 300 MHz vertical polarization signals are synchronously collected at the same detection point, and the amplitude and phase difference of the four groups of signals are collected at the receiving end; Three-stage process: a 10 cm*10 cm grid scanning is adopted in the reference state establishment stage, and each point stays for 0.5 seconds; intermittent spraying is performed through a micro-spraying device with a spacing of 2 meters in the humidity disturbance stage; fine scanning is performed in the high-response area triggering stage.

3. The method for detecting and evaluating internal defects of rammed earth sites using ground penetrating radar as claimed in claim 1, wherein: The specific process of establishing the reference state dual-frequency polarization response is as follows: a scanning and detection area is selected in an environment stable period without temperature and humidity mutation for three consecutive days, a reference response matrix P0(z) is constructed, P0(z) contains A100H0(z), A100V0(z), Δφ1000(z), A300H0(z), A300V0(z) and Δφ3000(z); reference parameters are calculated based on P0(z): a pore connectivity resistance initial value: R0(z)=k×Δφ1000(z) / A100H0(z), wherein k is a proportional constant calibrated through linear fitting of an indoor rammed soil sample; an initial water morphology characterization value η0(z) is Δφ3000(z) / Δφ1000(z).

4. The method for detecting and evaluating internal defects of rammed earth sites using ground penetrating radar as claimed in claim 2, wherein: The specific process of the humidity disturbance dual-frequency polarization cooperative triggering is that the micro-spraying device adopts an intermittent mode of spraying for 30 seconds and pausing for 120 seconds, and the total spraying amount is 5 mL / cm 2 ; at 1 hour after disturbance, that is, the surface free water migration period, 6 hours, that is, the middle layer free water response period, 24 hours, that is, the deep layer free water penetration period, and 72 hours, that is, the bound water stable period, four time points are scanned to obtain the time response matrix Pt(z), and the real-time water form characteristic value ηt(z) = Δφ300t(z) / Δφ100t(z) is calculated.

5. The method for detecting and evaluating internal defects of rammed earth sites using ground penetrating radar as claimed in claim 2, wherein: The trigger condition and parameters of the fine scanning of the high response area are: when |(ηt(z)-η0(z)) / η0(z)|>40% or |(Rt(z)-R0(z)) / R0(z)|>30% at a certain depth z, the fine scanning is triggered; the fine scanning adopts a 5cm*5cm grid, and the four groups of signal acquisition stay time is extended to 1 second, wherein Rt(z) is the real-time pore connectivity resistance.

6. The method for detecting and evaluating internal defects of rammed earth sites using ground penetrating radar as claimed in claim 1, wherein: The specific content of the water morphology polarization mass transfer triple coupling decoupling model is: The water morphology discrimination criterion is: η(z, t)>1.5 is the free water dominant area, η(z, t)<1.0 is the bound water dominant area, and 1.0≤η(z, t)≤1.5 is the mixed water area; Triple coupling decoupling equation: total time delay ΔT(z, t) = ΔTd(z) + ΔTwf(z, t) + ΔTWB(z), wherein ΔTd(z) = ΔL(z) x Vε0(z) / c, c is the speed of light, ε0(z) is the initial dielectric constant of rammed earth, ΔTwf(z, t) = z x V(ε0(z) + Δεwf(z, t)) / c - z x Vε0(z) / c, ΔTWB(z) = z x V(ε0(z) + εwb(z)) / c - z x Vε0(z) / c; Δεwf(z, t) = C1 x ηt(z) x exp(-t / τz), C1 is the free water polarization coefficient, τz = Rt(z) x z 2 / μ, μ is the water diffusion coefficient of rammed earth, εwb(z) = C2 x Δφ1000(z), C2 is the bound water polarization coefficient.

7. The method for detecting and evaluating internal defects of rammed earth sites using ground penetrating radar as claimed in claim 1, wherein: The specific standard of three target inversion is: Defect inversion: ΔTd(z) meets the fixed value and the relative error of the calculated value ΔTdcal(z) is ≤5%, and the corresponding area ηt(z)≤1.0; the inversion defect three-dimensional coordinates, size, shape and confidence Cd, Cd≥0.96 is the first level of confidence, 0.92≤Cd<0.96 is the second level of confidence; Water migration channel identification: the area with a time series change rate of ηt(z)>20% / h in the free water dominant area is the migration channel, the channel width W=0.1*ln(ηmax / ηmin), and ηmax and ηmin are the maximum and minimum values of ηt(z) in the area, respectively; Weathering state evaluation: < 60% of the proportion of bound water dominated area and > 0.3 of the migration channel density per meter 2 For strong weathering area; 60% ≤ the proportion of bound water dominated area ≤ 85% and 0.1 ≤ the migration channel density per meter 2 ≤ the migration channel density ≤ 0.3 per meter 2 For moderate weathering area; > 85% of the proportion of bound water dominated area and < 0.1 of the migration channel density per meter 2 For weak weathering area.

8. The method for detecting and evaluating internal defects of rammed earth sites using ground penetrating radar as claimed in claim 7, wherein: The specific process of the self-verification mechanism is: Cross-validation: the defect area needs to meet ηt(z)≤1.0 and ΔTd(z) fixed in the whole time series; the migration channel needs to meet the time series change of ηt(z) conforms to the rule of exp(-t / τz); Error correction: if the defect confidence Cd<0.92, the adjacent grid data is called to correct ΔTd(z) by linear interpolation algorithm, and the defect is re-determined after correction.

Images