A method for testing the tensile creep of cementitious materials based on a horizontal spring-loaded device

Abstract

The present application relates to the technical field of cement-based material tensile creep test, and particularly to a cement-based material tensile creep test method based on a horizontal spring loading device, which comprises the following steps: preparing a tensile creep test piece and a stress-free shrinkage compensation test piece, making a speckle pattern on the surface of the test piece and pre-embedding a double-channel distributed optical fiber sensor with temperature and strain synchronous sensing function, establishing an initial mechanical reference model of the spring and test piece series system after multi-stage ladder pre-tension centering adjustment; collecting load attenuation, apparent total deformation, internal strain and temperature distribution and micro-crack characteristic data in real time in the long-term test stage, constructing a pressure spring stiffness time-varying evolution model and eliminating micro-crack damage deformation to extract real viscoelastic creep strain; finally, deducting free shrinkage deformation and temperature difference correction deformation to obtain pure tensile creep strain and calculate real tensile creep degree and creep coefficient. The present application has high test precision, constant and controllable stress and traceable data.

Description

Technical Field

[0001] This invention relates to the field of tensile creep testing technology for cement-based materials, specifically a method for testing the tensile creep of cement-based materials based on a horizontal spring loading device. Background Technology

[0002] The tensile creep characteristics of cement-based materials are important parameters for evaluating the long-term deformation performance and cracking risk of concrete structures. Accurate determination of tensile creep degree and creep coefficient is of great significance for the design and service life assessment of large-volume concrete, prestressed concrete and ultra-long structures.

[0003] Existing methods often employ external extensometers or resistance strain gauges to measure the apparent total deformation of the specimen. The measured deformation simultaneously includes elastic deformation, viscoelastic creep deformation, non-viscoelastic damage deformation caused by microcrack opening, and temperature deformation, making it impossible to effectively separate the true viscoelastic creep deformation from the damage deformation. When microcracks initiate and propagate during the specimen's loading process, the apparent deformation data will be significantly inflated due to crack opening, leading to distortions in the calculated creep degree and creep coefficient. Current testing methods struggle to obtain information on the strain and temperature distribution within the specimen, making it impossible to locate and verify the internal microcrack propagation region. Furthermore, the identification of microcrack characteristic data lacks multi-source sensing support.

[0004] Furthermore, regarding shrinkage and temperature correction, cement-based materials undergo autogenous shrinkage, drying shrinkage, and temperature deformation simultaneously during creep testing. These non-load-bearing deformations are coupled with creep deformation and must be subtracted to obtain the pure tensile creep strain. Existing methods typically use stress-free shrinkage specimens from the same batch for shrinkage compensation, but often neglect the potential temperature difference between tensile creep specimens and stress-free shrinkage-compensated specimens, and lack an iterative verification mechanism for the shrinkage correction coefficient and the coefficient of thermal expansion, leading to systematic biases in the correction results. Simultaneously, existing testing methods lack a systematic centering adjustment and initial mechanical benchmark modeling process in the initial loading stage. The initial state of the specimen installation eccentricity and the spring-test series system is not accurately characterized, and long-term test data lacks a traceable initial benchmark.

[0005] Therefore, there is an urgent need for a method for testing the tensile creep of cement-based materials based on a horizontal spring loading device to solve the above-mentioned technical problems. Summary of the Invention

[0006] The purpose of this invention is to provide a method for testing the tensile creep of cement-based materials based on a horizontal spring loading device, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A method for testing the tensile creep of cement-based materials based on a horizontal spring loading device, wherein the horizontal spring loading device includes a displacement sensor for directly measuring the actual compression of the spring, and includes the following steps:

[0009] S1: Prepare cement-based tensile creep specimens and stress-free shrinkage compensation specimens; create speckle patterns on the surface of the tensile creep specimens and embed a dual-channel distributed optical fiber sensor with synchronous temperature and strain sensing function inside it. The dual-channel distributed optical fiber sensor includes a tightly wrapped strain optical fiber that deforms with the matrix and a loosely wrapped temperature optical fiber that senses only temperature and is not affected by mechanical strain; after demolding, the tensile creep specimens are connected to the screw device only through the specimen connectors at both ends and installed on a horizontal spring loading device; simultaneously place the stress-free shrinkage compensation specimens in a stress-free environment.

[0010] S2: Install a non-contact image acquisition device and a through-hole force sensor 36 on a horizontal spring loading device, apply multi-stage stepped pretension for centering adjustment; synchronously acquire force sensor data, speckle image strain data and initial frequency shift data of dual-channel distributed fiber optic sensor, use displacement sensor to directly measure the actual compression of spring under each stage of stepped pretension, calculate the initial elastic modulus of tensile creep specimen and the initial stiffness of spring, and establish the initial mechanical reference model of spring-specimen series system;

[0011] S3: Keep the loading nut in a fixed position and enter the long-term creep test stage; collect load attenuation data from the force sensor, apparent total deformation data from the non-contact image acquisition device, internal strain distribution data and temperature distribution data from the dual-channel distributed fiber optic sensor, and microcrack feature data extracted from the speckle image sequence through digital image correlation calculation.

[0012] S4: Construct a time-varying evolution model of the spring stiffness and output the current actual stiffness corresponding to the current sampling time; use microcrack feature data to identify the microcrack propagation region inside the tensile creep specimen, remove the non-viscoelastic damage deformation caused by microcrack opening from the apparent total deformation data, and extract the true viscoelastic creep strain; extrapolate the load loss of the next time window based on the smoothed load time series data, and calculate the smooth compensation displacement in combination with the current actual stiffness; drive the loading nut to rotate at a preset smooth speed curve to advance the smooth compensation displacement, so that the load is maintained within the allowable fluctuation zone of the target load value at an approximately constant tensile stress.

[0013] S5: Simultaneously acquire free shrinkage deformation data of stress-free shrinkage compensation specimens, and use temperature sensing data from dual-channel distributed fiber optic sensors to calculate the temperature difference correction deformation between tensile creep specimens and stress-free shrinkage compensation specimens; subtract the free shrinkage deformation data and temperature difference correction deformation from the true viscoelastic creep strain to obtain the pure tensile creep strain; combine the equivalent constant tensile stress determined by the effective constant load segment to calculate the true tensile creep degree and creep coefficient of cement-based materials.

[0014] Furthermore, the horizontal spring loading device includes a main frame and left and right end plates symmetrically arranged at both ends of the main frame and parallel to each other;

[0015] A specimen assembly is provided between the left end plate and the right end plate, and the two ends of the specimen assembly are respectively connected to the left end plate and the right end plate through screw device B and screw device A;

[0016] The specimen assembly includes a specimen mold formed by splicing four templates and specimen connectors at both ends of the specimen mold. The specimen connectors include mounting plates with ribbed steel bars extending into the specimen mold on the side facing the mounting plates. Connectors are provided on the other side of the mounting plates, and the connectors are fixedly connected to the mating ends of screw device A or screw device B by threads or bolts. The specimen mold is only used for casting and molding. After the tensile creep specimen is demolded, it is supported only by the specimen connectors at both ends and connected to the screw device. The surface of the specimen gauge length section is exposed to allow unobstructed observation by non-contact image acquisition equipment.

[0017] A spring system is provided on the side of the right end plate facing away from the left end plate. The spring system includes a loading pad plate parallel to the right end plate and a compression spring disposed between the loading pad plate and the right end plate. The loading pad plate has a mounting groove on its surface facing the left end plate. A pressure-bearing pad is disposed in the mounting groove, and a through-hole force sensor is disposed between the pressure-bearing pad and the inner wall of the mounting groove. The two ends of the compression spring are fixed to the pressure-bearing pad and the right end plate respectively by positioning steel rings. A guide spindle is disposed inside the compression spring or a guide sleeve is disposed outside the compression spring.

[0018] The screw of the screw device A extends out after passing through the through hole of the right end plate, the loading pad plate and the center hole of the through-hole force sensor in sequence. A loading nut is provided on the extended section of the screw device A, and a thrust ball bearing is provided between the loading nut and the loading pad plate.

[0019] A displacement sensor is installed between the right end plate and the loading pad plate to directly measure the actual compression of the compression spring.

[0020] As can be seen from the above technical solution provided by the present invention, the beneficial effects of the method for testing the tensile creep of cement-based materials based on a horizontal spring loading device provided by the present invention are:

[0021] This invention directly measures the actual compression of the spring by setting a displacement sensor on a horizontal spring loading device. Based on the measured compression, the initial stiffness of the spring is calibrated and a time-varying evolution model of the spring stiffness is constructed. This accurately reflects the stiffness decay law of the spring under long-term compression, avoiding load compensation deviation caused by time-varying spring stiffness. The load loss in the next time window is predicted by extrapolating the load decay data. The displacement is calculated using the current actual stiffness and the loading nut is driven to advance the compensation in a smooth speed curve consisting of a low-speed start section, a uniform speed compensation section, and a low-speed convergence section. This not only eliminates the additional damage to the specimen caused by step stress impact, but also keeps the tensile stress at an approximately constant level within the allowable fluctuation range of the target load value, ensuring the stability of stress conditions in long-term creep tests.

[0022] This invention employs a non-contact image acquisition device combined with digital image correlation technology to acquire full-field displacement and strain data of the specimen surface. Simultaneously, a dual-channel distributed optical fiber sensor is pre-embedded inside the specimen. By demodulating the frequency shift data of the tightly packed strain fiber and the loosely packed temperature fiber, the true internal axial strain distribution and temperature distribution are separated. By comparing the spatial positions of the localized strain region on the surface and the abrupt strain region inside, the microcrack propagation region is jointly identified. The non-viscoelastic damage deformation caused by microcrack opening is removed from the apparent total deformation, and the true viscoelastic creep strain is extracted. This effectively solves the problem that existing methods cannot separate damage deformation, resulting in distorted creep test results.

[0023] This invention simultaneously prepares stress-free shrinkage compensation specimens and places them in the same environment. By synchronously triggering and collecting their free shrinkage deformation data, and combining the temperature sensing data of the dual-channel distributed fiber optic sensor, the temperature difference correction deformation between the tensile creep specimen and the stress-free shrinkage compensation specimen is calculated. The pure tensile creep strain is obtained by subtracting the free shrinkage deformation and the temperature difference correction deformation from the true viscoelastic creep strain. The shrinkage correction coefficient and the thermal expansion coefficient are iteratively updated through residual correlation test, which further improves the correction accuracy and ensures that the final output tensile creep degree and creep coefficient only reflect the viscoelastic creep characteristics under constant tensile stress.

[0024] This invention uses multi-stage stepped preload for centering adjustment during the initial loading stage, simultaneously acquiring force sensor data, speckle image strain data, and initial fiber optic frequency shift data. It calculates the initial elastic modulus of the specimen and the initial stiffness of the compression spring, and establishes an initial mechanical benchmark model for the spring-specimen series system based on deformation compatibility and static equilibrium conditions, providing a traceable initial state benchmark for long-term testing. During testing, it performs time-series alignment of multi-source data such as load attenuation characteristics, apparent deformation, internal true strain, and temperature, constructing a multi-dimensional feature time-series database. The model parameters are updated after each compensation, ensuring complete traceability and data consistency in the test results.

[0025] This invention employs a specimen installation method where the specimen is supported only by the connecting parts at both ends after demolding, fully exposing the surface of the gauge section of the specimen. This avoids additional constraints and friction from the template on the specimen and provides an unobstructed observation window for non-contact image acquisition equipment. The two ends of the compression spring are fixed by positioning steel rings and equipped with guide mandrels or guide sleeves. Combined with a through-type force sensor and a thrust ball bearing, this ensures the coaxiality of the force transmission path and the accuracy of the compensation displacement measurement. The overall device has a stable structure and is easy to operate, making it suitable for high-precision testing of long-term tensile creep of cement-based materials. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the steps of a method for testing the tensile creep of cement-based materials based on a horizontal spring loading device according to the present invention.

[0027] Figure 2 This is a schematic diagram of the horizontal spring loading device in this invention;

[0028] Figure 3 This is a schematic diagram of the mold structure for the prototype in this invention;

[0029] Figure 4 This is a schematic diagram of the connector structure of the prototype in this invention;

[0030] Figure 5 for Figure 2 Schematic diagram of the structure of region A in the middle;

[0031] Figure 6 for Figure 2 Schematic diagram of the structure of region B in the middle.

[0032] In the diagram: 1. Main frame; 11. Left end plate; 12. Right end plate; 13. Screw assembly A; 14. Screw assembly B; 2. Specimen assembly; 3. Spring system; 31. Loading pad; 32. Positioning steel ring; 33. Compression spring; 34. Loading nut; 4. Specimen mold; 41. Template; 5. Specimen connector; 51. Connector; 52. Ribbed steel bar; 53. Mounting plate. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0034] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific embodiments.

[0035] like Figure 1-6As shown, this embodiment of the invention provides a method for testing the tensile creep of cement-based materials based on a horizontal spring loading device. The horizontal spring loading device includes a displacement sensor for directly measuring the actual compression of the compression spring 33, and includes the following steps:

[0036] S1: Prepare cement-based tensile creep specimens and stress-free shrinkage compensation specimens; create speckle patterns on the surface of the gauge length section of the tensile creep specimens, and embed a dual-channel distributed optical fiber sensor with synchronous temperature and strain sensing function inside it. The dual-channel distributed optical fiber sensor includes a tightly packed strain optical fiber that deforms with the matrix and a loosely packed temperature optical fiber that senses only temperature and is not affected by mechanical strain; after demolding, the tensile creep specimens are connected to the screw device only through the specimen connectors 5 at both ends and installed on a horizontal spring loading device; simultaneously place the stress-free shrinkage compensation specimens in a stress-free environment;

[0037] S2: Install a non-contact image acquisition device and a through-hole force sensor 36 on a horizontal spring loading device, apply multi-stage stepped pretension for centering adjustment; synchronously acquire data from the force sensor 36, speckle image strain data, and initial frequency shift data from the dual-channel distributed fiber optic sensor; use a displacement sensor to directly measure the actual compression of the compression spring 33 under each stage of stepped pretension, calculate the initial elastic modulus of the tensile creep specimen and the initial stiffness of the compression spring 33, and establish the initial mechanical reference model of the spring-specimen series system;

[0038] S3: Keep the loading nut 34 in a fixed position and enter the long-term creep test stage; collect load attenuation data from the force sensor 36, apparent total deformation data from the non-contact image acquisition device, internal strain distribution data and temperature distribution data from the dual-channel distributed fiber optic sensor, and microcrack feature data extracted from the speckle image sequence through digital image correlation calculation.

[0039] S4: Construct a time-varying evolution model of the spring stiffness and output the current actual stiffness corresponding to the current sampling time; use microcrack feature data to identify the microcrack propagation region inside the tensile creep specimen, remove the non-viscoelastic damage deformation caused by microcrack opening from the apparent total deformation data, and extract the true viscoelastic creep strain; extrapolate the load loss of the next time window based on the smoothed load time series data, and calculate the smooth compensation displacement in combination with the current actual stiffness; drive the loading nut 34 to rotate at a preset smooth speed curve to advance the smooth compensation displacement, so that the load is maintained within the allowable fluctuation zone of the target load value at an approximately constant tensile stress.

[0040] S5: Simultaneously acquire free shrinkage deformation data of stress-free shrinkage compensation specimens, and use temperature sensing data from dual-channel distributed fiber optic sensors to calculate the temperature difference correction deformation between tensile creep specimens and stress-free shrinkage compensation specimens; subtract the free shrinkage deformation data and temperature difference correction deformation from the true viscoelastic creep strain to obtain the pure tensile creep strain; combine the equivalent constant tensile stress determined by the effective constant load segment to calculate the true tensile creep degree and creep coefficient of cement-based materials.

[0041] In this embodiment, the core function of step S1 is to ensure the reliability, consistency, and traceability of long-term creep test data from the source of the specimen. By simultaneously preparing tensile creep specimens and stress-free shrinkage compensation specimens, a speckle pattern that can be stably identified by digital image correlation algorithms is constructed on the surface of the tensile creep specimen. A distributed fiber optic sensor with synchronous temperature and strain sensing function is pre-embedded inside. The tensile creep specimen is assembled into a horizontal spring loading device and connected to a screw device. Simultaneously, the stress-free shrinkage compensation specimen is placed in a stress-free environment. This provides the required specimen state and sensing conditions for establishing the initial mechanical reference model of the spring-specimen series system in step S2 and for the long-term creep test and multi-source data acquisition in steps S3 to S5. The detailed steps are as follows:

[0042] Step S1-1: Based on the strength grade and mix design parameters of the target cement-based material, determine the preparation schemes for tensile creep specimens and stress-free shrinkage compensation specimens; the tensile creep specimen adopts a prism configuration, with a rectangular cross-section at the gauge length, and the cross-sectional area of ​​the gauge length is denoted as . The length of the gauge length is denoted as The two ends of the specimen are connected to the specimen connector 5 by ribbed steel bars 52 embedded at the ends to form a force-transmitting anchorage section; the stress-free shrinkage compensation specimen and the tensile creep specimen are prepared synchronously using the same mix ratio, the same batch of mixture, and the same geometric dimensions to ensure the compensability of their shrinkage characteristics and temperature deformation characteristics; the four templates 41 of the specimen mold 4 are cleaned, and a release agent is evenly applied to the inner wall of the templates 41. The four templates 41 are spliced ​​together to form the specimen mold 4 and locked with fastening bolts. The sealing of the joints is checked; the two specimen connectors 5 are installed at both ends of the specimen mold 4, so that the mounting plate 53 is attached to the end of the specimen mold 4. The ribbed steel bars 52 extend into the mold cavity along the specimen axis and are kept centered to prepare for casting; the anchorage length of the ribbed steel bars 52 is determined according to the following formula for bond bearing capacity verification: ,in, This refers to the bond strength between the ribbed steel bar 52 and the cement matrix. The diameter of the ribbed steel bar is 52. This refers to the anchorage length of the 52 ribbed steel bar. The target is a constant tensile load; if the calculation does not meet the requirements, the anchorage length is increased or end hooks, nuts, and anchor plates are used for reinforcement; the sealing condition and loading age of the specimen are specified: when testing basic creep, the remaining surfaces and both ends of the specimen are sealed except for the speckle observation surface, and the sealing method is aluminum foil or paraffin plus polyethylene film; when testing total creep, the specimen surface is kept exposed and in the same environment as the specimen without stress shrinkage compensation; the specimen is cured to the specified loading age, such as 28 days, and the loading age, curing temperature and humidity are recorded in the specimen condition file;

[0043] Step S1-2: Weigh the cementitious materials, aggregates, mixing water, and additives according to the preset mixing ratio. First, dry mix the cementitious materials and aggregates evenly, then add the mixing water and additives and continue stirring to obtain a uniform mixture. The water-cement ratio of the mixture is as follows: Control, among which, This refers to the water-to-glue ratio. For the quality of the mixing water, The total mass of the cementitious material; the mixture is poured into specimen mold 4 in layers, and each layer is fully compacted using a vibrating table or tamping method to remove internal air bubbles; during pouring, a pre-reserved conduit is embedded along the axial centerline of the gauge length section to establish a through-channel for the subsequent internal pre-embedding of distributed fiber optic sensors; after pouring, the upper surface is smoothed and covered with a moisturizing film to prevent early water loss; after the specimen reaches the demolding strength, the mold is removed, and the tensile creep specimen and the stress-free shrinkage compensation specimen are simultaneously placed in the curing environment, and the curing temperature is maintained at [temperature missing]. Temperature in degrees Celsius, relative humidity not lower than After curing to the specified age, the specimens are removed and left to stand in a constant temperature and humidity environment until they are surface dry. During the curing period, the pouring time, curing temperature and humidity curves are recorded simultaneously as part of the specimen condition archive.

[0044] Steps S1-3: After curing, visually inspect the stress-free shrinkage compensation specimen to confirm that there are no initial defects such as cracks or missing edges on its surface; clean its sides and two end faces to reserve surface conditions for making speckle patterns with the same gauge length as the tensile creep specimen in step S5; at the same time, check the geometric dimensions and mass of the stress-free shrinkage compensation specimen to confirm its consistency with the tensile creep specimen, and record its initial mass and appearance as a reference for subsequent shrinkage deformation analysis.

[0045] Steps S1-4: Clean and dry the gauge length of the tensile creep specimen after demolding. First, uniformly spray a matte white primer to form a high-contrast base. Then, spray black speckle particles using a random spraying method to form a randomly distributed speckle pattern. The physical diameter of the speckle particles must match the resolution and field of view of the non-contact image acquisition device. The number of target pixels covered by a single speckle particle in the acquired image is determined by... Calculate, where, The number of target pixels covered by a single speckle particle. The physical diameter of the speckle particles. This represents the number of effective pixels of the image sensor along the gauge length direction. The physical length of the field of view of the non-contact image acquisition device along the gauge length direction; requirements The value should be between 3 and 5; below 3 will lead to undersampling of speckle, and above 5 will reduce spatial resolution; the area coverage of speckle particles is calculated as follows: Calculate, where, For speckle coverage, The total area covered by speckled particles. The total area of ​​the measured surface within the gauge length; requirements The grayscale randomness of the surface under test is controlled between 40% and 60%. After the speckle pattern is created, a non-contact image acquisition device is used to take test shots of the speckle pattern to check the grayscale contrast and particle uniformity. Areas that do not meet the quality requirements are re-sprayed and re-inspected until the speckle pattern meets the requirements of digital image correlation calculation for grayscale gradient and randomness. The speckle pattern is created after the mold is removed. The surface of the gauge length section must not be obscured by the template to provide a reliable observation surface for the acquisition of speckle image strain data in step S2.

[0046] Step S1-5: Within the gauge length of the tensile creep specimen, a dual-channel distributed optical fiber sensor is installed along the axial centerline using the pre-reserved ducts from the casting stage. This sensor includes a tightly packed strain fiber and a loosely packed temperature fiber. The tightly packed strain fiber is sheathed and backfilled with grout matching the stiffness of the matrix, allowing it to deform in tandem with the matrix. The change in its Brillouin frequency shift satisfies a linear coupling relationship with strain and temperature. ,in, This refers to the Brillouin frequency shift variation in tightly packed strained optical fibers. The strain sensitivity coefficient of the tightly packed strained optical fiber. The change in strain The temperature sensitivity coefficient of the tightly packed strained optical fiber. This refers to the temperature change; the loose-tube temperature fiber is placed inside a PTFE loose tube or flexible conduit, and is not affected by mechanical strain, but only responds to temperature. Its frequency shift change satisfies: ,in, This represents the Brillouin frequency shift variation in loose-tube temperature-controlled optical fibers. This refers to the temperature sensitivity coefficient of loose-tube temperature optical fiber. Consistent with the above, the two equations above show that single-channel frequency shift data simultaneously contains strain and temperature effects, and the two unknowns cannot be separated by a single equation. However, the strain and temperature can be solved by combining dual-channel frequency shift data, providing a physical basis for obtaining multi-level true axial strain values ​​and initial temperature field distribution data through temperature-strain simultaneous demodulation and separation in step S2. After the fiber optic cable is laid, grout matching the stiffness of the substrate is injected into the reserved channel to backfill the strain fiber optic channel. Loose-fitting temperature fiber optic conduits do not require backfilling. A flexible protective sleeve is fitted over the fiber optic lead-out end to prevent end bending damage, and a sufficient length of fiber optic cable is reserved for connection with the demodulation equipment. The number of effective measurement points in the gauge length section is determined by the ratio of the gauge length section length to the spatial sampling interval of the distributed fiber optic sensor along the fiber length direction. After the pre-embedding is completed, the two optical fibers are tested for light transmission and a set of reference frequency shift data is collected to confirm that the sensing channel is intact.

[0047] Steps S1-6: After demolding, the tensile creep specimen with speckle fabrication and fiber optic pre-embedding is transferred to the horizontal spring loading device and placed between the left end plate 11 and the right end plate 12 inside the main frame 1; the specimen is connected to the screw device B14 and the screw device A13 respectively through the connector 51 of the specimen connector 5 at both ends, wherein the screw device B14 passes through the left end plate 11 and the screw device A13 passes through the right end plate 12 and is linked to the spring system composed of the loading pad 31 and the compression spring 33; during the installation process, it is confirmed that the positioning steel rings 32 at both ends of the compression spring 33 are fixed to the pressure pad 37 and the right end plate 12 respectively, and the force sensor 36 is located between the pressure pad 37 and the inner wall of the mounting groove 35 and is in the normal zero position; the ribbed steel bar 52 is anchored. The axis is aligned with the axes of screw devices A13 and B14. A level or laser alignment instrument is used to measure the lateral deviation between the specimen axis and the loading axis at multiple measuring points. When the lateral deviation at any measuring point exceeds the preset installation alignment allowable value, the connection position of the specimen connector 5 and the screw device is adjusted and the measurement is repeated until the requirements are met. Then, the end nut of screw device B14 and the loading nut 34 are tightened to eliminate the connection gap and make the specimen in a light-touch, stress-free state. The initial position of the loading nut 34 is recorded as the zero-position reference. The specimen posture is adjusted so that the surface where the speckle pattern is located faces the installation side of the non-contact image acquisition device, keeping the imaging optical axis perpendicular to the measured surface, and confirming that the force sensor 36 and the two optical fiber leads are intact and reliable.

[0048] Steps S1-7: Place the visually inspected and qualified stress-free shrinkage compensation specimen horizontally on an independent rigid external support. A friction-reducing layer is placed between the specimen and the external support to ensure the specimen is in a state without external constraints. The friction-reducing layer is made of a low-friction coefficient material to eliminate the constraint effect of support friction on free shrinkage. Place the stress-free shrinkage compensation specimen in the same temperature and humidity environment as the tensile creep specimen, ensuring it is in a non-contact position with the horizontal spring loading device to avoid interference from device vibration and load transfer. Record the placement time, ambient temperature, and relative humidity of the stress-free shrinkage compensation specimen to establish an environmental benchmark synchronized with the tensile creep specimen. This provides placement conditions and a state benchmark for the synchronous acquisition of free shrinkage deformation data of the stress-free shrinkage compensation specimen in step S5.

[0049] In this embodiment, the core function of step S2 is to perform graded pre-tensioning and centering verification on the horizontal spring loading device based on the specimen preparation, sensing element pre-positioning and assembly completed in step S1. Simultaneously, data from the through-hole force sensor, speckle image strain data, and initial frequency shift data from the distributed fiber optic sensor are collected. Then, the initial elastic modulus of the tensile creep specimen and the initial stiffness of the compression spring 33 are calculated. These, along with the initial temperature field distribution data, are injected into the state-space equation of the series system constructed based on deformation compatibility conditions and static equilibrium conditions. This establishes an initial mechanical benchmark model characterizing the deformation compatibility and force equilibrium of the spring-specimen series system under multi-physics coupling. This provides a traceable initial state benchmark for the long-term creep testing stage in step S3 and the construction of the time-varying evolution model of the compression spring stiffness and the calculation of smooth compensation displacement in step S4. The detailed steps are as follows:

[0050] Step S2-1: Install the non-contact image acquisition device on the horizontal spring loading device and fix it on an independent bracket, making the imaging optical axis perpendicular to the surface where the speckle pattern of the tensile creep specimen is located. Adjust the working distance so that the gauge length section falls completely within the field of view. Configure the sampling clock, exposure time, and triggering mode of the image acquisition device to enable it to synchronously trigger sampling with the force sensor and the distributed fiber optic sensor. Check and zero the through-hole force sensor 36 to confirm that it is located between the pressure pad 37 and the inner wall of the mounting groove 35, the screw of the screw device A13 passes through its central hole, the force-bearing surface is flat and there is no initial off-center load. Record the full scale, zero drift, and sensitivity parameters of the force sensor 36, and set its sampling frequency to be higher than the sampling frequency of the non-contact image acquisition device to reserve interpolation space for the time alignment of multi-source data. Perform zero-position calibration on the displacement sensor and record the initial pre-compression of the compression spring 33. After installation, a joint pre-acquisition test was conducted on the non-contact image acquisition device and the force sensor 36. After confirming that the data link was normal, the loading phase was initiated.

[0051] Step S2-2: Keep the loading nut 34 at the zero reference position recorded in steps S1-6, apply graded screwing advance to the loading nut 34 according to a preset ratio, and apply multi-stage stepped preload step by step; The stepped preload is calculated based on the target constant tensile load and the first step. The product of the loading coefficients is determined, and the loading coefficients are arranged in ascending order, with the highest loading coefficient not exceeding 1, so that the preload remains within the elastic working range throughout the entire process, avoiding damage to the tensile creep specimen. After each level of preload is applied, it is stabilized and held. Strain gauges are attached to symmetrical positions on both sides of the gauge length after the specimen is rigidly attached, or dual-sided digital image correlation observation is used to extract the average strain at the symmetrical positions on both sides and calculate the centering strain deviation rate. ,in, For the first Centering strain deviation rate under stepped preload The average strain at symmetrical positions on the front side of the measured surface. The average strain at symmetrical positions on the back side of the measured surface. The arithmetic mean of the strain on both sides; under any stage of step preload If the deviation exceeds the preset allowable alignment value (e.g., 10%), axial eccentricity is determined to exist. The mating position of screw device B14 and screw device A13, or the installation posture of specimen connector 5, should be adjusted, and the screwing and re-testing should be performed. Under each stage of stepped preload... If none of the values ​​exceed the preset centering allowable value, the centering adjustment is deemed qualified, and the centering adjustment completion status is recorded.

[0052] Step S2-3: After the centering adjustment is qualified, the loading nut 34 is returned to the zero reference position and multi-level stepped pretension is applied again. Under each step of pretension, three types of initial data are collected synchronously according to a unified trigger signal. The first type is force sensor data, which continuously records the force sensor readings at a preset sampling frequency to obtain the original load time series data corresponding to the multi-level stepped pretension. The second type is speckle image strain data, which is obtained by non-contact image acquisition equipment capturing speckle patterns under each step of pretension to obtain a speckle image sequence covering the gauge length. The third type is the initial frequency shift data of the dual-channel distributed optical fiber sensor, which is obtained by distributed optical fiber demodulation equipment synchronously recording the Brillouin frequency shift distribution of the tightly packed strain fiber and the loosely packed temperature fiber along the fiber length direction, and using the reference frequency shift data collected after the pre-embedding in step S1-5 as a reference, to form the frequency shift change distribution data at each level. The acquisition time of the three types of data is recorded according to a unified sampling clock.

[0053] Step S2-4: Perform noise reduction filtering on the synchronously acquired force sensor data to eliminate random noise interference; use a moving average filter to smooth the original load time series data; the filtered load value is the arithmetic mean of the original load sample values ​​within the half-window width of the filter; divide the filtered load time series data into segments according to the loading sequence of multi-stage preload, identify the pressure stabilization interval under each stage of preload, and take the arithmetic mean of the filtered load values ​​within the pressure stabilization interval as the multi-stage load value for that stage, denoted as... ; will all levels The load values ​​are stored in association with the corresponding step preload levels to form a multi-level step load value sequence.

[0054] Step S2-5: Using the initial frequency shift data from the dual-channel distributed optical fiber sensor, demodulate simultaneously. Based on the frequency shift-strain-temperature relationship between the tightly packed strain fiber and the loosely packed temperature fiber described in Step S1-5, solve simultaneously for the temperature change and the true axial strain: , ,in, , , , , , , The meaning is consistent with the annotation of step S1-5; the true axial strain obtained by demodulating point by point along the fiber length direction under each step of pretension is correlated with the spatial position of the gauge length segment to obtain the multi-level true axial strain values ​​inside the tensile creep specimen; at the same time, the temperature change distribution along the fiber length direction during each demodulation process is superimposed with the reference temperature to obtain the initial temperature field distribution data, which is used as the temperature input item of the initial mechanical reference model.

[0055] Step S2-6: Based on the multi-stage stepped load values ​​and the cross-sectional area of ​​the tensile creep specimen, calculate the multi-stage stepped nominal stress values. The calculation formula is as follows: Step load values Cross-sectional area of ​​gauge length The ratio; linearly fitting the nominal stress values ​​of the multi-stage stepped strain with the corresponding real axial strain values ​​of the multi-stage strain yields the initial stress-strain curve; since the multi-stage stepped preload is within the elastic working range, the initial stress-strain curve is approximately a straight line passing through the origin, and its slope is the initial elastic modulus of the tensile creep specimen, denoted as . The calculation is performed according to the least squares principle; the linear correlation test is performed on the fitting results. When the goodness of fit is lower than the preset threshold, the accuracy of the extraction range of the stepped load value and the temperature-strain simultaneous demodulation is checked and the fitting is re-fitted; after the fitting is qualified, the initial elastic modulus is used as the basis for calculating the axial stiffness of the specimen and stored in the initial mechanical reference model; the nominal stress is calculated based on the net cross-sectional area, deducting the cross-sectional area of ​​the fiber optic channel; the stabilization time of each stage should be short, not exceeding 5 minutes, to limit the additional strain due to creep; the fitting adopts the average of loading and unloading data and the test age is noted;

[0056] Step S2-7: Perform digital image correlation calculations on the speckle image strain data. Using the speckle images under each step of preload as the deformation image and the speckle image at zero position as the reference image, perform sub-region matching and sub-pixel displacement solving to obtain the multi-level apparent total displacement data of the tensile creep specimen surface, which is used for cross-checking with the fiber optic strain integration results. The multi-level actual compression of the compression spring 33 corresponding to the multi-level step load values ​​is directly measured by the displacement sensor and recorded as... When displacement sensor data is unavailable, it is determined by geometric conversion: ,in, For the first The actual compression of the stepped preload compression spring 33. Consistent with the above, For the first The cumulative number of turns the nut 34 is screwed in under the stepped preload. The pitch of the screw assembly A13 is... For the first The elongation of the specimen under stepped preload is given by integrating the multiple levels of real axial strain values ​​along the gauge length or by the gauge length displacement related to the digital image; dividing the multiple levels of stepped load values ​​by the corresponding multiple levels of actual compression yields multiple initial stiffness test values: ,in, For the first Initial stiffness test values ​​corresponding to stepped loads. Consistent with the above, Consistent with the above; linear regression was performed on the force-compression data points and the linearity was tested, and the slope of the linear segment was used as the initial stiffness of the compression spring 33. Alternatively, the arithmetic mean of multiple initial stiffness test values ​​after removing outliers can be taken as the initial stiffness. The difference method of subtracting the fiber strain integral from the apparent total displacement should not be used to determine the compression of the spring, because both are observations of the deformation of the specimen itself, and their difference does not contain the compression information of the spring.

[0057] Step S2-8: Input the initial elastic modulus, initial stiffness of spring 33, and initial temperature field distribution data into the state-space equation of the series system constructed based on deformation compatibility conditions and static equilibrium conditions; the static equilibrium condition of the spring-specimen series system is that the forces on the series elements are equal, and the deformation compatibility condition is that the total deformation of the system is equal to the sum of the deformations of each element; the equivalent stiffness of the spring-specimen series system is as follows: , ,in, The equivalent stiffness of the spring-specimen series system, To determine the axial stiffness of the tensile creep specimen, The initial stiffness of compression spring 33 is given. Consistent with the above, , Consistent with the above; based on this, the geometrical parameters of the horizontal spring loading device are introduced as boundary conditions. These geometrical parameters include the gauge length, the position of the loading pad 31, the installation position of the positioning steel ring 32, and the screw extension length of the screw device A13. Based on these parameters, the degree-of-freedom constraint relationship between the axial deformation component of the specimen and the compression deformation component of the compression spring is determined. A system-level stiffness matrix is ​​assembled to obtain the system-level equilibrium equation, i.e., the product of the equivalent stiffness and the total system displacement equals the system load. Furthermore, the system-level equilibrium equation is rewritten as a discrete state-space equation: , ,in, For the first The system state vector under stepped load consists of the axial deformation component of the tensile creep specimen and the compressive deformation component of the compression spring 33. For the first The input vector under stepped load is composed of multiple stepped load values. For the first The output vector under stepped load is composed of the force sensor readings and the observed values ​​of multi-level apparent total displacement data. Here is the system state transition matrix. Input matrix to the system, The system output matrix is ​​used to identify and verify the parameters of the state transition matrix using measured data under various stepped loads, so that the solution of the state equation matches the measured response. Finally, an initial mechanical reference model is established to characterize the deformation coordination and force balance of the spring-specimen series system under multi-physics coupling. The model uses the initial elastic modulus, the initial stiffness of the compression spring 33 and the initial temperature field distribution data as the initial state reference, and saves the model parameters and initial state values.

[0058] In this embodiment, the core function of step S3 is to keep the loading nut 34 fixed after the centering adjustment and initial mechanical benchmark model are established in step S2, and to continuously acquire the load attenuation data of the force sensor, the apparent total deformation data of the non-contact image acquisition device, and the internal strain distribution and temperature distribution data of the dual-channel distributed optical fiber sensor through three synchronous sensing channels. Furthermore, time-series alignment and feature extraction are performed on the multi-source data to construct a multi-dimensional feature time-series database. The detailed steps are as follows:

[0059] Step S3-1: Keep the loading nut 34 in a fixed position, record the position of the loading nut 34 and the force sensor reading at the moment of completion of loading in step S2, and take the initial value of the system state in the initial mechanical reference model in step S2 as the initial state of the long-term creep test stage, and enter the long-term creep test stage; continuously sample the force sensor 36 according to the preset sampling frequency to obtain the original load time series data containing the load decay trend; since the loading nut 34 is fixed in position and the tensile creep specimen continues to creep elongate under constant nominal load, the deformation coordination relationship of the spring-specimen series system causes the compression amount of the compression spring 33 to shrink accordingly, and the force sensor reading shows a continuous decay trend. The original load time series data completely records this load decay process;

[0060] Step S3-2: Denoise the original load time series data by filtering to eliminate random noise interference, and use the moving average filtering method from step S2-4 to obtain smooth load time series data; perform differential operations on the smooth load time series data, and use the central difference scheme to calculate the instantaneous load decay rate at each sampling time; fit the load decay curve based on the instantaneous load decay rate, and extract the load decay amplitude characteristic parameter and load decay time constant characteristic parameter from the load decay curve to characterize the long-term load loss law; arrange the load decay characteristic parameters according to the sampling time to form the load decay characteristic parameter time series data, and record the goodness of fit and residual sequence; continuously monitor the residual sequence during the fitting process, and switch to piecewise fitting when the residual shows systematic deviation;

[0061] Step S3-3: Acquire a continuous speckle image sequence captured by a non-contact image acquisition device at preset time intervals. Each image is marked with the acquisition time indicated by the preset sampling clock. Perform digital image correlation matching calculation on the continuous speckle image sequence. Using the speckle image acquired at the beginning of the long-term creep test as the reference image, perform sub-region matching and sub-pixel displacement calculation on the images at other times to obtain the full-field displacement data of the tensile creep specimen surface. Perform spatial gradient calculation on the full-field displacement data to obtain the full-field strain data of the tensile creep specimen surface. Integrate the full-field displacement data along the axial direction of the tensile creep specimen to obtain the apparent total deformation data. The apparent total deformation data includes the elastic deformation, viscoelastic creep deformation, non-viscoelastic damage deformation caused by microcrack opening, and system deformation caused by the change in spring compression of the tensile creep specimen. At the suspected surface damage area, take virtual extensometer reference lines on both sides of the crack as the boundary, calculate the microcrack opening displacement by the difference in axial displacement on both sides, and record it as COD. Combine it with the crack location to form microcrack feature data.

[0062] Step S3-4: Identify strain localization concentration regions from the full-field strain data; scan the full-field strain data at each sampling time point along the axial direction, and record the set of continuous measurement points that exceed the strain localization criterion threshold as strain localization concentration regions. This threshold is preset based on the material properties of the tensile creep specimen and the noise level of the full-field strain data; the determination of strain localization concentration regions also requires that the length of the region along the axial direction is not less than the preset minimum region length to exclude isolated false measurement points caused by speckle noise; mark the strain localization concentration regions as suspected surface damage areas, and record their axial position range, region length, and peak strain value within the region to form suspected surface damage area location data;

[0063] Step S3-5: Perform temperature-strain demodulation on the frequency shift data acquired in real time by the dual-channel distributed fiber optic sensor. Using the demodulation relationship from step S2-5, separate the temperature effect component and strain effect component in the real-time frequency shift data to obtain the true axial strain distribution data and temperature distribution data inside the tensile creep specimen. Perform curvature calculation along the fiber length direction on the true axial strain distribution data to obtain strain curvature distribution data. Identify the region where the strain curvature distribution data exceeds the preset curvature threshold as the internal strain abrupt change region. The preset curvature threshold is preset based on the fiber measuring point spacing and strain noise level. The spatial resolution of the distributed fiber is on the order of 0.2 meters to 1 meter. The internal strain abrupt change region reflects the averaged indication of strain localization and is only used for qualitative location of microcrack regions. The microcrack opening displacement is given by the digital image correlation calculation in step S3-3.

[0064] Step S3-6: Compare the spatial positions of the internal strain abrupt change zone and the suspected surface damage zone to obtain the comparison results; during the comparison, unify the axial center coordinates of the internal strain abrupt change zone and the axial center coordinates of the suspected surface damage zone to the same axial coordinate reference, calculate the spatial position deviation, and determine that the two are spatially consistent when the deviation does not exceed the preset spatial consistency tolerance; the spatial position consistency tolerance is 1 to 2 times the fiber optic measuring point spacing; based on the comparison results, the suspected surface damage zone and the internal strain abrupt change zone with consistent spatial positions are identified as microcrack propagation areas; for suspected surface damage zones or internal strain abrupt change zones that only appear in a single sensing channel, they are determined to be caused by noise or non-microcrack factors and are not included in the microcrack propagation area for the time being, but are only observed and recorded;

[0065] Step S3-7: The microcrack opening displacement (COD) is given by the digital image correlation calculation in step S3-3; the microcrack characteristic data includes the microcrack opening displacement, the location of the internal strain abrupt change zone, and the strain curvature distribution data, which are arranged into a time series of microcrack characteristic data according to the sampling time; the damage influence in the microcrack propagation area is subtracted from the true axial strain distribution data of the tensile creep specimen over the entire length range, and the damage influence at other locations is set to zero, to obtain the internal viscoelastic strain distribution data without damage deformation;

[0066] Steps S3-8: Perform multi-source data time-series alignment processing on the load attenuation characteristic parameters, apparent total deformation data, and internal viscoelastic strain distribution data; take the start time of the long-term creep test phase as the alignment zero point, set the alignment reference time step, and take the smaller of the sampling time interval of the force sensor and the acquisition time interval of the non-contact image acquisition device or their common value, so that the three types of data can be aligned to a unified timestamp; interpolate and resample the time-series data of the load attenuation characteristic parameters and the apparent total deformation data according to their original sampling time on the aligned timestamp, interpolate the internal viscoelastic strain distribution data along the axial direction according to the position of the fiber optic measuring point to a unified spatial grid, and synchronously include the temperature distribution data in the alignment processing; construct a multi-dimensional characteristic time-series database containing load evolution, apparent deformation, internal true strain and temperature, to provide time-consistent multi-source data support for step S4.

[0067] In this embodiment, the core function of step S4 is to solve the problem of non-constant stress caused by the superposition of two major factors in the long-term creep test: the time-varying decay of the compression spring stiffness and the damage evolution of the tensile creep specimen. First, a time-varying evolution model of the compression spring stiffness is constructed, and the current actual stiffness is output. Then, microcrack feature data is used to identify the microcrack propagation region, and the non-viscoelastic damage deformation caused by microcrack opening is removed from the apparent total deformation data to extract the true viscoelastic creep strain. The load loss in the next time window is predicted by extrapolation based on smoothed load time series data, and the smoothed compensation displacement is calculated in combination with the current actual stiffness. Finally, the loading nut 34 is driven to rotate at a preset smoothed speed curve to maintain an approximately constant tensile stress within the allowable fluctuation zone. The detailed steps are as follows:

[0068] Step S4-1: Using the initial stiffness of the compression spring in the initial mechanical benchmark model in Step S2-8 as the benchmark stiffness, and extracting three types of input features of the compression spring 33 under long-term compression from the multi-dimensional feature time series database in Step S3-8; the first type is the load variation feature, which is taken from the time series data of the load attenuation feature parameters and characterized by smoothed load time series data and its instantaneous load attenuation rate; the second type is the cumulative compression deformation feature, which is characterized by the actual compression deformation of the compression spring directly measured by the displacement sensor; the third type is the temperature drift feature, which is taken from the temperature sensing data of the dual-channel distributed optical fiber sensor and characterized by the difference between the temperature distribution data and the initial temperature field distribution data.

[0069] Step S4-2: The actual compression deformation of the compression spring 33 is directly measured by the displacement sensor and recorded as... When displacement sensor data is missing, it is determined by geometric conversion: ,in, For compression spring 33 in the first The actual compression deformation at each sampling time. Consistent with the previous text, The elongation of the tensile creep specimen is obtained by integrating the internal strain distribution data along the gauge length. The compression of the compression spring must not be determined here by the indirect difference method of subtracting the actual deformation of the specimen from the apparent total deformation, because both are observations of the specimen's own deformation, and their difference does not include information on the compression of the compression spring.

[0070] Step S4-3: Based on the stiffness mapping relationship between load attenuation data and actual compression deformation, determine the apparent stiffness change sequence of compression spring 33 at different sampling times: ,in, For compression spring 33 in the first Apparent stiffness at each sampling time point To smooth the load time series data in the first... The value at each sampling time, Consistent with the above; the apparent stiffness change sequence contains both the true information of time-varying stiffness decay and spurious changes caused by temperature drift, requiring temperature drift correction: ,in, For the first Temperature correction stiffness of spring 33 at each sampling time. To account for the stiffness test deviation caused by temperature drift, the spring steel's elastic modulus temperature coefficient is pre-calibrated. Consistent with the above; the temperature coefficient of the elastic modulus of steel is approximately For every degree Celsius, a fluctuation of 1 to 2 degrees Celsius in room temperature causes a stiffness deviation of approximately 0.02% to 0.04%, which is very small. Temperature correction is only performed when temperature fluctuations are significant to avoid introducing additional noise.

[0071] Step S4-4: Based on the variation trend of the temperature-corrected stiffness change sequence with the holding time, cumulative compression deformation, and load level, the stiffness decay process of the compression spring 33 is divided into an initial stable decay stage, a linear slow decay stage, and a local accelerated decay stage. The holding time is piecewise linearly fitted with the temperature-corrected stiffness change sequence to determine the boundary time between the two stages. The stiffness decay rate parameters of the compression spring 33 within the three stages are identified according to the linear fitting slope of the temperature-corrected stiffness change sequence within each stage, and the stiffness decay rate parameter of the local accelerated decay stage is greater than that of the linear slow decay stage. The cumulative duration of each stage at each sampling time is calculated using the stage boundary time as the segmentation point.

[0072] Steps S4-5: Using the initial stiffness of the compression spring, the stiffness decay rate parameters at each stage, the current cumulative compression deformation, and the current temperature as model inputs, construct a time-varying evolution model of the compression spring stiffness, characterizing the dynamic change of the stiffness of compression spring 33 with holding time and cumulative compression deformation. The current cumulative compression deformation is directly given by the measured value from the displacement sensor, and the current temperature is taken as the average value of the temperature distribution data within the gauge length range. The stiffness decay factor is characterized by the ratio of the sum of the products of the stiffness decay rate parameters at the three stages and the cumulative duration to the initial stiffness of the compression spring. The current cumulative compression deformation is used to determine the current stage of stiffness decay and trigger the stage boundary. The current temperature is used to continue temperature drift correction, ensuring that the model output corresponds to the current actual stiffness at the current sampling time, denoted as . This is used for subsequent calculations of spring compression deformation components and smoothing compensation displacement.

[0073] Step S4-6: Based on the current actual stiffness output by the time-varying evolution model of spring stiffness, perform stiffness mapping on the load attenuation data to obtain the spring compression deformation component corresponding to spring 33 under the current load. Its value is the ratio of the current load to the current actual stiffness. Subtract the spring compression deformation component from the apparent total deformation data to obtain the axial deformation component of the specimen that reflects the deformation state of the tensile creep specimen itself. It includes the elastic deformation, viscoelastic creep deformation and damage deformation caused by microcrack opening of the specimen.

[0074] Step S4-7: Calculate the strain gradient along the axial direction for the speckle image strain data, and mark the regions in the speckle image strain data whose strain gradient exceeds a preset surface gradient threshold as suspected surface damage areas; use the microcrack opening displacement, the location of the internal strain abrupt change zone, and the strain curvature distribution data in the microcrack feature data to verify the suspected surface damage areas. The verification includes the consistency of the spatial position between the suspected surface damage areas and the internal strain abrupt change zone and the rationality of the value of the microcrack opening displacement, eliminating false markings caused by speckle noise and surface deposits; the microcrack opening displacement is derived from digital image related calculations;

[0075] Step S4-8: Based on the verification results of step S4-7, suspected surface damage areas and internal strain abrupt change areas with consistent spatial location and continuous increase in microcrack opening displacement within a continuous time window are identified as microcrack propagation areas; the time persistence criterion is that the microcrack opening displacement continuously increases within at least two consecutive time windows; areas that do not meet the time persistence criterion are judged as intermittent local damage or noise interference, and are not included in the microcrack propagation area, only the observation record is retained;

[0076] Step S4-9: Map the microcrack opening displacement within the microcrack propagation area to the axial deformation component of the specimen according to the measurement point position to obtain the damage opening deformation component; subtract the damage opening deformation component from the axial deformation component of the specimen to obtain the net axial deformation data without the influence of microcrack opening.

[0077] Step S4-10: Average the net axial deformation data within the gauge length of the tensile creep specimen to obtain the true viscoelastic creep strain; arrange the true viscoelastic creep strain according to the sampling time to form viscoelastic creep strain time series data, which will serve as the verification sequence for the load loss prediction in the next time window.

[0078] Step S4-11: Divide the smoothed load time series data into historical time windows and current time windows; extract the load decay rate from the current time window. Load decay acceleration Creep strain rate and temperature change rate are used as rolling prediction features; based on the rolling prediction features, a second-order extrapolation is performed on the load attenuation trend in the next time window to obtain the predicted load loss: ,in, This represents the predicted load loss caused by the continuous decay of the load within the next time window. This represents the load decay rate within the current time window. The load decay acceleration within the current time window. The time window length is used; simultaneously, the predicted creep deformation increment for the next time window is extrapolated using viscoelastic creep strain time series data. The deformation compatibility relationship of the spring-specimen series system is converted into the load loss check value, and the conversion formula is: ,in, For the current actual stiffness, To predict creep deformation increment, Consistent with the above; when the deviation between the load time extrapolation value and the strain check conversion value exceeds the preset ratio, the load time extrapolation value shall be used as the standard and the deviation shall be recorded;

[0079] Step S4-12: Take the load currently sampled by force sensor 36 as the current load value, denoted as... Take the value of the smoothed load time series data at that moment; and take the load corresponding to the initial constant tensile stress as the target load value, denoted as . The load compensation gap is obtained by comparing the deviations between the target load value, the current load value, and the predicted load loss. ,in, This refers to the load compensation gap, which is the total load that needs to be compensated by screwing in the loading nut 34 within the next time window. The target load value is consistent with the target constant tensile load in step S1-1. This is the current load value. Consistent with the above;

[0080] Step S4-13: Determine the initial compensation displacement based on the stiffness mapping relationship between the load compensation gap and the current actual stiffness: ,in, This is the initial compensation displacement. Consistent with the above, Consistent with the above; by combining the pitch of the screw device that mates with the loading nut 34, the rotational stroke of the loading nut 34, and the allowable stress fluctuation range of the tensile creep specimen, the initial compensation displacement is limited and verified to obtain the smooth compensation displacement: ,in, To smooth out the compensation displacement, that is, the compensation displacement that is finally executed by the loading nut 34 after the amplitude limiting verification, Consistent with the above, To limit the number of turns that the load nut 34 can be screwed in, Consistent with the above, To allow for stress fluctuation amplitude, values ​​are taken from ±2% to ±5% of the target stress. Consistent with the above, Consistent with the above;

[0081] Step S4-14: Determine the target rotation stroke of the loading nut 34 based on the smooth compensation displacement and the pitch of the screw device; the target rotation number is the ratio of the smooth compensation displacement to the pitch, and the axial displacement of the target rotation stroke is equal to the smooth compensation displacement; decompose the target rotation stroke into a low-speed start-up segment, a constant-speed compensation segment, and a low-speed convergence segment, and the sum of the three segments equals the target rotation stroke; generate a smooth speed curve matching the target rotation stroke based on the magnitude of the smooth compensation displacement and the predicted load loss; the loading nut 34 rotates at a speed lower than the impact sensitivity threshold in the low-speed start-up segment, at a constant speed matching the load loss rate in the constant-speed compensation segment, and at a gradually decreasing speed to the target rotation stroke in the low-speed convergence segment; the impact sensitivity threshold is calibrated by the maximum allowable loading rate at which the tensile creep specimen does not produce additional damage under step load;

[0082] Step S4-15: During the low-speed start-up phase, monitor the real-time load change rate of the force sensor 36 and the real-time strain change rate of the dual-channel distributed fiber optic sensor. The real-time load change rate is obtained by the first-order difference of the smoothed load time series data with respect to time, and the real-time strain change rate is obtained by the first-order difference of the internal viscoelastic strain distribution data with respect to time at the mean of the gauge length. When the real-time load change rate or the real-time strain change rate exceeds the preset safety threshold, pause the screwing of the loading nut 34 and maintain its current position to avoid damage to the impact-sensitive specimen during the start-up transient. At the same time, continue to record the real-time data of the force sensor 36 and the dual-channel distributed fiber optic sensor without interruption.

[0083] Step S4-16: When the real-time load change rate and real-time strain change rate recover to within the preset safety threshold, continue to execute the uniform speed compensation segment according to the smooth speed curve, so that the compression of the compression spring 33 gradually recovers as the loading nut 34 rotates, offsetting the load loss caused by the predicted load loss; in the uniform speed compensation segment, the loading nut 34 rotates at a constant speed, the compression recovery rate of the compression spring matches the load loss rate, and the force sensor reading gradually rises and approaches the target load value; after the uniform speed compensation segment ends, enter the low-speed convergence segment, and the loading nut 34 rotates at a gradually decreasing speed to the target rotation stroke, completing the entire execution of this smooth compensation displacement;

[0084] Step S4-17: After the loaded nut 34 reaches the target rotation stroke, maintain a preset stabilization time; during the stabilization time, synchronously collect the compensated load data from the force sensor 36, the compensated apparent total deformation data from the non-contact image acquisition device, and the compensated internal strain distribution data from the dual-channel distributed fiber optic sensor to confirm that the compensated load data is within the allowable fluctuation band of the target load value; when it exceeds the allowable fluctuation band, calculate the fine-tuning displacement according to the ratio of the remaining deviation to the current actual stiffness and perform fine-tuning rotation; update the compensated load data, compensated apparent total deformation data, and compensated internal strain distribution data to a multi-source system for prediction in the next time window. The time-series data is used to correct the time-varying evolution model of the spring stiffness. The correction method is to re-execute steps S4-3 to S4-5 for apparent stiffness calculation, temperature drift correction and stiffness decay rate parameter identification, and update the stiffness decay rate parameter and stage boundary time so that the time-varying evolution model of the spring stiffness always reflects the latest stiffness state of the spring 33, and keeps the tensile stress on the tensile creep specimen at an approximately constant level within the allowable fluctuation range of the target load value. After each compensation is completed, the process returns to the long-term real-time acquisition process in step S3-1, and the load loss prediction and smoothing compensation for the next time window are executed in a rolling manner until the long-term creep test stage ends.

[0085] In this embodiment, the core function of step S5 is to further purify the true viscoelastic creep strain extracted in step S4, subtracting the free shrinkage deformation data of the stress-free shrinkage compensation specimen and the temperature difference correction deformation, to obtain the pure tensile creep strain that only reflects the viscoelastic creep characteristics under constant tensile stress. Then, combined with the equivalent constant tensile stress, the true tensile creep degree and creep coefficient of the cement-based material are calculated, and the calculation results are statistically analyzed in segments and damage is verified. The detailed steps are as follows:

[0086] Step S5-1: Place the stress-free shrinkage compensation specimen in the same environmental conditions as the tensile creep specimen, including the same ambient temperature, relative humidity, and ventilation. Isolate the stress-free shrinkage compensation specimen from the external support using a friction-reducing layer, the same layer used in step S1-7. Create a speckle pattern on the surface of the stress-free shrinkage compensation specimen with the same gauge length as the tensile creep specimen. The method and quality requirements for creating the speckle pattern are the same as in step S1-4, i.e., the number of target pixels covered by a single speckle particle is between 3 and 5, and the speckle coverage is controlled between 40% and 60%, forming an unconstrained shrinkage observation state. The stress-free shrinkage compensation specimen and the tensile creep specimen can be photographed alternately by the same non-contact image acquisition device at different times, with the time error noted. Alternatively, they can be photographed separately by two synchronously triggered non-contact image acquisition devices to ensure strict consistency of the sampling clock. Confirm that the surface of the speckle pattern on the stress-free shrinkage compensation specimen faces the non-contact image acquisition device.

[0087] Step S5-2: Using the same sampling clock as the tensile creep specimen image acquisition, synchronously trigger the acquisition of the stress-free shrinkage compensation specimen using a non-contact image acquisition device to obtain a shrinkage speckle image sequence of the stress-free shrinkage compensation specimen; synchronous trigger acquisition ensures that each image in the shrinkage speckle image sequence has an acquisition time label consistent with the continuous speckle image sequence of the tensile creep specimen; during the acquisition, the unconstrained state of the stress-free shrinkage compensation specimen remains unchanged to ensure that the shrinkage speckle image sequence completely records the entire process of free shrinkage deformation;

[0088] Step S5-3: Perform digital image correlation calculation on the shrinkage speckle image sequence. Using the shrinkage speckle image acquired at the time of establishing the unconstrained shrinkage observation state as the reference image, perform sub-region matching and sub-pixel displacement calculation on the images at other times to obtain the shrinkage displacement field within the gauge length of the stress-free shrinkage compensation specimen. Normalize the shrinkage displacement field along the gauge length of the stress-free shrinkage compensation specimen to obtain the free shrinkage deformation data expressed in strain form, denoted as... Free shrinkage deformation data includes drying shrinkage, self-shrinkage, and temperature deformation under stress-free conditions, and is the first correction amount in subsequent subtraction processing.

[0089] Step S5-4: Using temperature sensing data from a dual-channel distributed fiber optic sensor, and taking the initial temperature state as the temperature reference, obtain the average temperature change of the tensile creep specimen and the stress-free shrinkage compensation specimen during the holding load process; the temperature difference correction deformation is calculated as follows: ,in, To correct for deformation due to temperature difference, The coefficient of thermal expansion of cement-based materials. This represents the average temperature change of the tensile creep specimen. This is the average temperature change of the stress-free shrinkage compensation specimen. When two specimens are in the same constant temperature environment and the temperature difference is negligible, the temperature difference correction deformation is zero. This temperature difference correction deformation is only used to correct the temperature difference between the two specimens and should not be deducted again from the temperature deformation component already included in the free shrinkage deformation data to avoid double deduction of the same temperature deformation.

[0090] Step S5-5: Unify the actual viscoelastic creep strain, free shrinkage deformation data and temperature difference corrected deformation amount under the same sampling clock for time alignment. The alignment reference is consistent with the timestamp after alignment in step S3-8. After alignment, the aligned actual viscoelastic creep strain, aligned free shrinkage deformation data and aligned temperature difference corrected deformation amount are obtained respectively.

[0091] Step S5-6: Calculate the pure tensile creep strain according to the revised formula: ,in, This represents the strain of pure tensile creep, i.e., the strain value that reflects the viscoelastic creep characteristics under constant tensile stress after deducting free shrinkage deformation and temperature-corrected deformation. To represent the true viscoelastic creep strain after alignment. This is the shrinkage correction factor, initially set to 1. The data is the aligned free shrinkage deformation data. Consistent with the above; pure tensile creep strain is arranged according to the aligned timestamps to form pure tensile creep strain time series data;

[0092] Step S5-7: Perform residual correlation tests on the pure tensile creep strain, free shrinkage deformation data, and temperature-corrected deformation. The test is performed in two paths: the first path tests the residual correlation between pure tensile creep strain and free shrinkage deformation data, and the second path tests the residual correlation between pure tensile creep strain and temperature-corrected deformation. A significance threshold is preset, such as 0.3. When the absolute value of the residual correlation coefficient of the first path is not less than the significance threshold, the shrinkage correction coefficient is updated using least squares regression. ,in, , , , The meaning is consistent with the annotation of step S5-6; after updating, re-execute step S5-6; when the absolute value of the correlation coefficient of the second residual is not less than the significance threshold, update the thermal expansion coefficient by regression. Then repeat steps S5-4 to S5-6; when the absolute values ​​of the correlation coefficients of the two residuals are both less than the significance threshold, or the number of iterations reaches the preset upper limit such as 3 times, output the pure tensile creep strain.

[0093] Steps S5-8: The load data collected by force sensor 36 under the action of smooth compensation displacement and confirmed to be stable are used as load history data; the load history data is subjected to a stationarity test to obtain an effective constant load segment within the preset allowable fluctuation range; the time-weighted average value of the load within the effective constant load segment is taken. ,according to The equivalent constant tensile stress is determined by the cross-sectional area of ​​the gauge length: ,in, For equivalent constant tensile stress, The time-weighted average of the loads within the effective constant load segment. Consistent with the above; when the load sawtooth fluctuation caused by compensation cannot be ignored, the creep strain should be converted according to the superposition principle or time-weighted equivalent stress, and instantaneous load should not be directly treated as constant stress.

[0094] Step S5-9: Align the pure tensile creep strain and the equivalent constant tensile stress according to the sampling time to obtain synchronous time-series data; convert the pure tensile creep strain to the unit equivalent constant tensile stress to obtain the true tensile creep: ,in, The duration of the load is The true tensile creep at time, i.e., the pure tensile creep strain corresponding to a unit equivalent constant tensile stress. This is time-series data of pure tensile creep strain. Consistent with the above; the equivalent constant tensile stress is converted into initial elastic strain using the initial elastic modulus, and the ratio between the pure tensile creep strain and the initial elastic strain is determined as the creep coefficient: , ,in, For the initial elastic strain, Consistent with the above, Consistent with the above, The duration of the load is creep coefficient over time Consistent with the above;

[0095] Step S5-10: Perform segmented statistics on the true tensile creep and creep coefficient based on the holding time to obtain the true tensile creep subsequence and creep coefficient subsequence corresponding to the early creep stage, the stable creep stage, and the late creep stage respectively; the holding time between the two stages is preset according to the growth rate change characteristics of the creep coefficient time series data; the mean and standard deviation of each stage subsequence are output in segmented statistics to characterize the stage evolution characteristics of the creep performance of cement-based materials.

[0096] Step S5-11: Use microcrack characteristic data to perform damage verification on the true tensile creep degree subsequence and creep coefficient subsequence; when either the true tensile creep degree subsequence or the creep coefficient subsequence has an abrupt change segment corresponding to the microcrack propagation area, remove the pure tensile creep strain corresponding to the abrupt change segment, and re-execute steps S5-9 and S5-10 using the removed pure tensile creep strain to recalculate the true tensile creep degree and creep coefficient. The removed interval is marked in the output results, and the connection between the removed segment and the retained segment is connected by interpolation and noted; when neither subsequence has an abrupt change segment corresponding to the microcrack propagation area, no removal is required, and the output is directly output; the output content includes the true tensile creep degree time series data, the creep coefficient time series data, the statistical results of the subsequences of the three creep stages, as well as the loading age, sealing status, temperature and humidity records, specimen size and mix ratio, load fluctuation range, and post-test crack inspection results, which serve as the final basis for evaluating the creep performance of cement-based materials and structural design calculations.

[0097] In this embodiment, the horizontal spring loading device is a dedicated load-bearing and loading device for the tensile creep testing method of the present invention. Its core function is to install the tensile creep specimen horizontally in a closed force transmission chain, and apply and maintain an approximately constant axial tensile load within the allowable fluctuation band through an elastic load-bearing system composed of a screw device and a compression spring 33. At the same time, it provides load monitoring by a force sensor, direct measurement of the compression amount of the compression spring by a displacement sensor, a speckle observation window of a non-contact image acquisition device, and a smooth compensation screw-in interface for the loading nut 34 during the load-bearing process. The device is horizontally laid out, with the specimen assembly 2 in the middle and the spring system 3 located at the right end. The screw device passes through the end plate and is connected in series with the specimen assembly 2 and the spring system 3, forming a complete force transmission path from the right end plate 12 through the compression spring 33, the loading pad 31, the force sensor 36, the screw device A13, the specimen assembly 2 to the screw device B14 and terminating at the left end plate 11.

[0098] In this embodiment, the main frame 1 is the basic bearing structure of the horizontal spring loading device. It is welded from shaped steel or thick steel plates and has sufficient bending stiffness and overall stability. It is used to bear all the reaction force applied by the spring system 3 during long-term load holding and to enclose the reaction force inside the frame. The main frame 1 is arranged horizontally along the axial direction of the specimen, and its two ends are symmetrically provided with a left end plate 11 and a right end plate 12 that are parallel to each other. The functions of the left end plate 11 and the right end plate 12 include three aspects: First, they serve as the two ends of the closed force transmission chain; Second, their parallelism directly determines the straightness of the loading axis. Therefore, when they are installed, the plate surfaces must be parallel to each other and perpendicular to the longitudinal axis of the main frame 1 to provide a geometric reference for the centering of the specimen; Third, the right end plate 12 also serves as the fixed bearing end of the spring system 3. The compression spring 33 is fixed to the inside of the right end plate 12 by the positioning steel ring 32, so that the spring reaction force is directly transmitted to the right end plate 12 and balanced by the main frame 1. The right end plate 12 is provided with a through hole for the screw of the screw device A13 to pass through.

[0099] In this embodiment, specimen assembly 2 is the forming, anchoring, and force transmission unit of the tensile creep specimen, consisting of specimen mold 4 and specimen connector 5. Specimen mold 4 is formed by splicing four templates 41 and is used only for casting. Its cavity size determines the cross-sectional area of ​​the gauge section. Specimen connector 5 includes mounting plate 53, ribbed steel bars 52, and connectors 51. The mounting plate 53 has ribbed steel bars 52 extending into the cavity on the side facing specimen mold 4, which are wrapped by cement-based material during casting to form anchoring ends. The other side of mounting plate 53 has... A connector 51 is provided, and the connector 51 is fixedly connected to the mating end of the screw device A13 or the screw device B14 by threads or bolts; the anchorage length of the ribbed steel bar 52 is determined according to the verification formula in step S1-1; after the specimen is demolded and reaches the specified age, all templates are removed, and the tensile creep specimen is supported only by the specimen connectors 5 at both ends and connected to the screw device to bear the load, and the surface of the gauge length section is completely exposed, so that the non-contact image acquisition equipment can observe without obstruction, while avoiding additional constraints and friction caused by the contact between the template and the specimen.

[0100] In this embodiment, the spring system 3 is the core elastic load-bearing component of the horizontal spring loading device. It is located on the side of the right end plate 12 facing away from the left end plate 11, i.e., the outer side. It includes a loading pad 31, a compression spring 33, a positioning steel ring 32, a mounting groove 35, a pressure-bearing pad 37, and a through-hole force sensor 36. The loading pad 31 is parallel to the right end plate 12, and the compression spring 33 is sandwiched between the loading pad 31 and the right end plate 12. The loading pad 31 has a mounting groove 35 on its surface facing the left end plate 11. A pressure-bearing pad 37 is installed in the groove, and a through-hole force sensor 36 is installed between the pressure-bearing pad 37 and the inner wall of the mounting groove 35. The screw of the screw device A13 passes through the central hole of the force sensor 36. The pressure-bearing pad 37 is of high strength. The pressure equalization component evenly distributes the concentrated compressive force at the end of the compression spring 33 to the force sensor 36. The force sensor 36 is located in the middle of the force transmission path of the compression spring, directly sensing the axial pressure generated by the compression of the compression spring 33. The two ends of the compression spring 33 are fixed to the pressure-bearing pad 37 and the right end plate 12 respectively by positioning steel rings 32. A guide mandrel is set inside the compression spring 33 or a guide sleeve is set outside to constrain the lateral deformation of the compression spring 33, prevent instability and end slippage under long-term compression, and ensure that the axis of the compression spring 33 always coincides with the loading axis. The working pressure of the compression spring 33 does not exceed 80% of its allowable ultimate load, and the compression deformation of the compression spring 33 during operation meets the requirements of the elastic working range. The initial pre-compression of the compression spring 33 is recorded when it is mounted. The relationship between the elastic force and the amount of compression deformation of spring 33 is as follows: ,in, For the initial stiffness of the compression spring, This represents the actual compression of the compression spring. This relationship indicates that when the tensile creep specimen undergoes creep elongation under constant load, the total length of the spring-specimen series system remains unchanged, and the compression deformation of the compression spring 33 decreases accordingly. The elastic force of the compression spring 33, i.e., the tensile load borne by the specimen, is correspondingly reduced. Therefore, the compression deformation margin of the spring system 3 determines the upper limit of the load-bearing capacity of the device under conditions without human intervention.

[0101] In this embodiment, the displacement sensor 38 is disposed between the right end plate 12 and the loading pad 31 or at the extended end of the screw device A13, and is used to directly measure the actual compression of the compression spring 33. The displacement sensor 38 uses an LVDT, a grating ruler, or a digital dial indicator. Its range covers the sum of the maximum working compression of the compression spring 33 and the compensation stroke. Its resolution is not less than 0.01 mm, and it samples synchronously with the force sensor 36. The measured values ​​are the basis for the initial stiffness calibration of the compression spring in step S2-7, the calculation of the actual compression amount and apparent stiffness of the compression spring in steps S4-2 and S4-3, and the limit verification of the smooth compensation displacement in step S4-13.

[0102] In this embodiment, screw device A13 and screw device B14 are two sets of threaded force transmission components connecting the spring system 3 side and the fixed end side, respectively; the screw of screw device B14 passes through the left end plate 11, and its end is connected to the connector 51 of the specimen assembly 2 near the left end plate 11, and is fixed on the outside with a lock nut to provide fixed end constraint for the specimen; the screw of screw device A13 passes through the through hole of the right end plate 12, the loading pad 31 and the center hole of the through-hole force sensor 36 in sequence and then extends out, and its end located inside the main frame 1 is connected to the specimen assembly 2 near the right end plate 11. 2. Connector 51 is connected at one end, and a loading nut 34 is provided on the extended section; a thrust ball bearing is provided between the loading nut 34 and the loading pad 31 to decouple the screwing friction torque from the axial force and reduce the measurement error of the compensation displacement; screwing in the loading nut 34 causes the loading pad 31 to move towards the compression spring 33, increasing the compression deformation of the compression spring 33 and increasing the elastic force of the compression spring 33. The increased tension is transmitted to the specimen through the screw device A13 to achieve load compensation; screwing out in the opposite direction unloads the load; the conversion relationship between the number of screwing in the loading nut 34 and the increase in the compression of the compression spring 33 is as follows: ,in, This is the increment of the compression of spring 33. For the number of rotations, The pitch of the screw assembly A13 is given.

[0103] In this embodiment, the complete load transfer path of the horizontal spring loading device is as follows: the elastic pressure of the compression spring 33 acts on the loading pad 31 through the bearing pad 37 and the through-hole force sensor 36. The loading pad 31 transmits the reaction force to the screw of the screw device A13 through the thrust ball bearing and the loading nut 34, and then transmits it to the tensile creep specimen through the mounting plate 53 and the ribbed steel bar 52 of the right end specimen connector 5. The other end of the tensile creep specimen transmits the reaction force to the left end plate 11 through the left end specimen connector 5 and the screw device B14. Finally, the main frame 1 closes and balances the reaction forces at both ends. In this force transmission path, the tensile creep specimen... The creep specimen and the compression spring 33 form a series force system that satisfies the static equilibrium condition, i.e., the axial force of the specimen, the axial force of the compression spring, and the force sensor reading are equal; at the same time, it satisfies the deformation compatibility condition, i.e., the total deformation of the system is equal to the sum of the deformation of the specimen and the compression deformation of the compression spring 33; the equivalent stiffness of the spring-specimen series system is calculated according to the formula in step S2-8. The geometric dimensional parameters of the device include the gauge length, the position of the loading pad 31, the installation position of the positioning steel ring 32, and the screw extension length of the screw device A13, all of which are used as boundary condition inputs when establishing the initial mechanical reference model in step S2-8.

[0104] In this embodiment, the coordination relationship between the horizontal spring loading device and each step of the testing method is as follows: In step S1, the specimen mold 4 is only used for casting and molding. After demolding, the tensile creep specimen is connected to the screw device B14 and screw device A13 by the specimen connectors 5 at both ends, and the speckle observation surface is not obstructed by the template; In step S2, the force sensor 36 is installed between the pressure pad 37 and the inner wall of the mounting groove 35, the displacement sensor 38 is installed between the right end plate 12 and the loading pad 31, and the non-contact image acquisition device is installed on the outside of the device with an independent bracket and aligned with the surface where the speckle pattern is located. The centering adjustment is completed by applying a multi-stage stepped preload by screwing in the loading nut 34, and the force sensor data, speckle image strain data and initial frequency shift data are collected simultaneously; In step S3, the loading... The nut 34 remains fixed in position, the force sensor 36 continuously records load attenuation data, the non-contact image acquisition device continuously captures speckle images, and the dual-channel distributed fiber optic sensor continuously outputs internal strain distribution data and temperature distribution data. In step S4, the smooth compensation displacement is calculated based on the current actual stiffness output by the time-varying evolution model of the spring stiffness. The loading nut 34 is driven to rotate at a preset smooth speed curve, so that the compression deformation of the spring 33 recovers as it rotates, offsetting the load loss caused by the predicted load loss and maintaining an approximately constant tensile stress within the allowable fluctuation zone. In step S5, the compensated constant load data is taken from the stable and confirmed reading of the force sensor 36, and is used to calculate the equivalent constant tensile stress and finally obtain the true tensile creep degree and creep coefficient.

[0105] In this embodiment, the main frame 1, left end plate 11, right end plate 12, and loading pad 31 are made of structural steel with sufficient rigidity. The bearing surfaces are machined to ensure flatness, and the flatness deviation of the pressure plate bearing surface is controlled to be no more than 0.1 mm per 100 mm. The screw device A13 and screw device B14 are made of high-strength screws, and the tensile stress under the test load is no more than 30% of the material yield point. The compression spring 33 is made of spring steel with long fatigue life and small permanent deformation, and the working pressure does not exceed 10% of its allowable ultimate load. 80; The mating surfaces of the positioning steel ring 32, the pressure-bearing gasket 37, and the mounting groove 35 are precision machined and kept coaxial to ensure that the force axis of the compression spring 33, the force sensor 36, and the screw device are consistent; After the displacement sensor 38 is installed, it is calibrated together with the force sensor 36. The calibration range covers the entire long-term load and compensation stroke, and the synchronization of the sampling clocks of the two is verified; After the overall assembly of the device is completed, no-load centering verification and step-by-step loading calibration are performed. Only after confirming that there is no looseness, no off-center load, and no additional friction in each link of the force transmission chain can it be put into use.

[0106] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for testing the tensile creep of cement-based materials based on a horizontal spring loading device, characterized in that: The horizontal spring loading device includes a displacement sensor for directly measuring the actual compression of the compression spring (33), comprising the following steps: S1: Prepare cement-based tensile creep specimens and stress-free shrinkage compensation specimens; create speckle patterns on the surface of the tensile creep specimens and embed a dual-channel distributed optical fiber sensor with synchronous temperature and strain sensing function inside it. The dual-channel distributed optical fiber sensor includes a tightly wrapped strain optical fiber that deforms with the matrix and a loosely wrapped temperature optical fiber that senses only temperature and is not affected by mechanical strain; after demolding, the tensile creep specimens are connected to the screw device only through the specimen connectors (5) at both ends and installed on the horizontal spring loading device; simultaneously place the stress-free shrinkage compensation specimens in a stress-free environment. S2: Install a non-contact image acquisition device and a through-hole force sensor (36) on a horizontal spring loading device, apply multi-level stepped pretension for centering adjustment; synchronously acquire data from the force sensor (36), speckle image strain data and initial frequency shift data from the dual-channel distributed fiber optic sensor, use the displacement sensor to directly measure the actual compression of the compression spring (33) under each level of stepped pretension, calculate the initial elastic modulus of the tensile creep specimen and the initial stiffness of the compression spring (33), and establish the initial mechanical reference model of the spring-specimen series system; S3: Keep the loading nut (34) in a fixed position and enter the long-term creep test stage; collect the load attenuation data of the force sensor (36), the apparent total deformation data of the non-contact image acquisition device, the internal strain distribution data and temperature distribution data of the dual-channel distributed optical fiber sensor, and the microcrack feature data extracted from the speckle image sequence by digital image correlation calculation in real time. S4: Construct a time-varying evolution model of spring stiffness and output the current actual stiffness corresponding to the current sampling time; use microcrack feature data to identify the microcrack propagation area inside the tensile creep specimen, remove the non-viscoelastic damage deformation caused by microcrack opening from the apparent total deformation data, and extract the real viscoelastic creep strain; extrapolate the load loss of the next time window based on the smooth load time series data, and calculate the smooth compensation displacement in combination with the current actual stiffness; drive the loading nut (34) to rotate the smooth compensation displacement at the preset smooth speed curve, so that the load is maintained in the approximately constant tensile stress within the allowable fluctuation zone of the target load value; S5: Simultaneously acquire free shrinkage deformation data of stress-free shrinkage compensation specimens, and use temperature sensing data from dual-channel distributed fiber optic sensors to calculate the temperature difference correction deformation between tensile creep specimens and stress-free shrinkage compensation specimens; subtract the free shrinkage deformation data and temperature difference correction deformation from the true viscoelastic creep strain to obtain the pure tensile creep strain; combine the equivalent constant tensile stress determined by the effective constant load segment to calculate the true tensile creep degree and creep coefficient of cement-based materials.

2. The method for testing the tensile creep of cement-based materials based on a horizontal spring loading device according to claim 1, characterized in that: The horizontal spring loading device includes a main frame (1) and a left end plate (11) and a right end plate (12) symmetrically arranged at both ends of the main frame (1) and parallel to each other. A specimen assembly (2) is provided between the left end plate (11) and the right end plate (12). The two ends of the specimen assembly (2) are respectively connected to the left end plate (11) and the right end plate (12) through screw device B (14) and screw device A (13). The specimen assembly (2) includes a specimen mold (4) formed by splicing four templates (41) and specimen connectors (5) set at both ends of the specimen mold (4); the specimen connectors (5) include mounting plates (53), the mounting plates (53) facing the specimen mold (4) are provided with ribbed steel bars (52) extending into the specimen mold (4), and the other side of the mounting plates (53) is provided with connectors (51), the connectors (51) are fixedly connected to the mating end of the screw device A (13) or the screw device B (14) by threads or bolts; the specimen mold (4) is only used for casting and molding, after the tensile creep specimen is demolded, it is only supported by the specimen connectors (5) at both ends and connected to the screw device, and the surface of the specimen gauge section is exposed so that the non-contact image acquisition equipment can observe it without obstruction; A spring system (3) is provided on the side of the right end plate (12) facing away from the left end plate (11). The spring system (3) includes a loading pad (31) parallel to the right end plate (12) and a compression spring (33) disposed between the loading pad (31) and the right end plate (12). An installation groove (35) is provided on the surface of the loading pad (31) facing the left end plate (11). A pressure-bearing pad (37) is provided in the installation groove (35). A through-hole force sensor (36) is provided between the pressure-bearing pad (37) and the inner wall of the installation groove (35). The two ends of the compression spring (33) are fixed to the pressure-bearing pad (37) and the right end plate (12) respectively by positioning steel rings (32). A guide spindle is provided inside the compression spring (33) or a guide sleeve is provided outside. The screw of the screw device A (13) extends out after passing through the through hole of the right end plate (12), the loading pad plate (31) and the center hole of the through-hole force sensor (36) in sequence. A loading nut (34) is provided on the extended section of the screw device A (13), and a thrust ball bearing is provided between the loading nut (34) and the loading pad plate (31). A displacement sensor is provided between the right end plate (12) and the loading pad plate (31) to directly measure the actual compression of the compression spring (33).

3. The method for testing the tensile creep of cement-based materials based on a horizontal spring loading device according to claim 2, characterized in that: In S2, the initial elastic modulus, the initial stiffness of the compression spring (33), and the initial mechanical reference model are calculated, specifically including: The force sensor (36) data acquired synchronously is denoised and filtered to extract the multi-level stepped load values ​​corresponding to the multi-level stepped preload; the initial frequency shift data of the dual-channel distributed optical fiber sensor is used to demodulate and obtain the multi-level true axial strain values ​​and initial temperature field distribution data. The demodulation formula is: ； ； ， ； in, This represents the frequency shift change in a tightly packed strained optical fiber. This represents the frequency shift change of the loose-tube temperature-controlled optical fiber. , The strain sensitivity coefficient and temperature sensitivity coefficient of tightly packed strained optical fiber. The change in strain This refers to the temperature sensitivity coefficient of loose-tube temperature optical fiber. This refers to the change in temperature. The multi-step nominal stress value is calculated based on the multi-step load value and the cross-sectional area of ​​the tensile creep specimen. The multi-step nominal stress value is linearly fitted with the corresponding multi-step true axial strain value to obtain the initial stress-strain relationship curve. The slope of the curve is used as the initial elastic modulus of the tensile creep specimen. The actual compression of the compression spring (33) under each step pretension is directly measured using a displacement sensor. The multi-step load value is divided by the corresponding multi-step actual compression value to obtain multiple initial stiffness test values. Linear regression is performed on the force-compression data points and the linearity is checked. The slope of the linear segment or the arithmetic mean of multiple initial stiffness test values ​​after removing outliers is taken as the initial stiffness of the compression spring (33). The initial elastic modulus, initial stiffness, and initial temperature field distribution data are input into the state space equation of the series system constructed based on deformation compatibility conditions and static equilibrium conditions. The geometric dimensional parameters of the horizontal spring loading device are introduced as boundary conditions. The system-level stiffness matrix is ​​assembled and the state equation is solved to establish an initial mechanical benchmark model characterizing the deformation compatibility and force equilibrium of the spring-specimen series system.

4. The method for testing the tensile creep of cement-based materials based on a horizontal spring loading device according to claim 1, characterized in that: Real-time acquisition of load attenuation data from the through-hole force sensor (36), apparent total deformation data from the non-contact image acquisition device, and internal strain distribution and microcrack characteristic data from the distributed fiber optic sensor; Specifically, it includes: The force sensor (36) is continuously sampled according to the preset sampling frequency to obtain the original load time series data; the original load time series data is denoised and filtered to obtain smooth load time series data; the smooth load time series data is differentially processed to calculate the instantaneous load decay rate at each sampling time, the load decay curve is fitted, and the load decay characteristic parameters used to characterize the long-term load loss law are extracted. A sequence of continuous speckle images captured by a non-contact image acquisition device is acquired at preset time intervals. Digital image correlation matching calculation is performed on the continuous speckle image sequence to obtain the full-field displacement data of the tensile creep specimen surface. Spatial gradient calculation is performed on the full-field displacement data to obtain the full-field strain data. Localized strain concentration areas are identified from the full-field strain data and marked as suspected surface damage areas. The displacement difference between the two sides of the suspected surface damage area with the crack as the boundary is extracted as the microcrack opening displacement. The full-field displacement data is integrated along the axial direction of the tensile creep specimen to obtain the apparent total deformation data. Temperature-strain simultaneous demodulation was performed on the frequency shift data acquired in real time by the dual-channel distributed fiber optic sensor to separate the true axial strain distribution data and temperature distribution data inside the tensile creep specimen. Curvature calculation was performed along the fiber length direction on the true axial strain distribution data to obtain strain curvature distribution data. Regions where the strain curvature distribution data exceeded a preset curvature threshold were identified as internal strain abrupt change regions. The spatial positions of the internal strain abrupt change regions and suspected surface damage regions were compared, and the suspected surface damage regions and internal strain abrupt change regions with consistent spatial positions were identified as microcrack propagation regions. The damage influence within the microcrack propagation regions was subtracted from the true axial strain distribution data over the entire length of the tensile creep specimen to obtain internal viscoelastic strain distribution data without damage deformation. By performing multi-source data time-series alignment processing on load attenuation characteristic parameters, apparent total deformation data, internal viscoelastic strain distribution data, temperature distribution data, and microcrack characteristic data, a multi-dimensional characteristic time-series database containing load evolution, apparent deformation, internal true strain, and temperature is constructed.

5. The method for testing the tensile creep of cement-based materials based on a horizontal spring loading device according to claim 1, characterized in that: The construction of the time-varying evolution model of compression spring stiffness specifically includes: The initial stiffness of the compression spring in the initial mechanical reference model is used as the reference stiffness, and the load change characteristics, compression deformation accumulation characteristics and temperature drift characteristics of the compression spring (33) under long-term compression are extracted from the load attenuation data, apparent total deformation data, internal strain distribution data and temperature sensing data. The value directly measured by the displacement sensor is taken as the actual compression deformation of the compression spring (33), or determined according to the geometric conversion relationship when the displacement sensor data is missing: ,in, This represents the actual compression deformation of the compression spring (33). This is the initial pre-compression amount during mounting. Elongation of the tensile creep specimen; Based on the stiffness mapping relationship between load attenuation data and actual compression deformation, the apparent stiffness change sequence of compression spring (33) at different sampling times is determined; Temperature drift correction is performed on the apparent stiffness change sequence using temperature sensing data to eliminate stiffness test deviations caused by temperature changes, resulting in a temperature-corrected stiffness change sequence. Based on the variation trend of temperature-corrected stiffness change sequence with holding time, cumulative amount of compression deformation and load level, the stiffness decay process of compression spring (33) is divided into initial stable decay stage, linear slow decay stage and local accelerated decay stage, and stiffness decay rate parameters of each stage are identified respectively. The initial stiffness of the compression spring, the stiffness decay rate parameters at each stage, the current cumulative amount of compression deformation, and the current temperature are used as model inputs to construct a time-varying evolution model of the compression spring (33) stiffness that characterizes the dynamic change of the stiffness of the compression spring (33) with the holding time and the cumulative amount of compression deformation. The output of the time-varying evolution model of the compression spring stiffness is the current actual stiffness corresponding to the current sampling time.

6. The method for testing the tensile creep of cement-based materials based on a horizontal spring loading device according to claim 5, characterized in that: Identifying microcrack propagation regions and extracting true viscoelastic creep strain using microcrack feature data specifically includes: Based on the current actual stiffness output by the time-varying evolution model of spring stiffness, the load attenuation data is stiffened and the spring compression deformation component corresponding to the spring (33) under the current load is obtained. Subtracting the spring compression deformation component from the apparent total deformation data yields the axial deformation component of the specimen, which reflects the deformation state of the tensile creep specimen itself. The strain gradient along the axial direction is calculated from the strain data of the speckle image. Regions with strain gradients exceeding a preset surface gradient threshold are marked as suspected surface damage areas. The suspected surface damage areas are then verified using microcrack opening displacement, internal strain abrupt change location, and strain curvature distribution data from the microcrack feature data. Suspected surface damage areas and internal strain abrupt change areas with consistent spatial location and continuously increasing microcrack opening displacement within a continuous time window are identified as microcrack propagation regions. The microcrack opening displacement within the microcrack propagation region is mapped to the axial deformation component of the specimen according to the measurement point location to obtain the damage opening deformation component. Subtract the damage opening deformation component from the axial deformation component of the specimen to obtain the net axial deformation data that does not contain the influence of microcrack opening. The net axial deformation data were averaged within the gauge length of the tensile creep specimen to obtain the true viscoelastic creep strain, and the true viscoelastic creep strain was arranged according to the sampling time to form viscoelastic creep strain time series data.

7. The method for testing the tensile creep of cement-based materials based on a horizontal spring loading device according to claim 6, characterized in that: The load loss in the next time window is predicted based on the actual viscoelastic creep strain, and the smooth compensation displacement is calculated by combining the current actual stiffness output from the time-varying evolution model of the compression spring stiffness; specifically including: The smoothed load time series data is divided into historical time windows and current time windows, and the load decay rate, load decay acceleration, creep strain rate and temperature change rate are extracted from the current time window as rolling prediction features. Based on the rolling prediction characteristics, the load attenuation trend in the next time window is extrapolated to obtain the predicted load loss. At the same time, the creep deformation increment in the next time window is extrapolated using viscoelastic creep strain time series data, and converted into a load loss check value through the deformation coordination relationship of the spring-specimen series system, and the predicted load loss is checked and corrected. The load currently sampled by the force sensor (36) is taken as the current load value, and the load corresponding to the initial constant tensile stress is taken as the target load value. The load compensation gap is obtained by comparing the deviations between the target load value, the current load value, and the predicted load loss. ,in, For load compensation gap, For the target load value, This is the current load value. To predict load loss; The initial compensation displacement is determined based on the stiffness mapping relationship between the load compensation gap and the current actual stiffness. The initial compensation displacement is then limited and verified by considering the pitch of the screw device that mates with the loading nut (34), the rotational stroke of the loading nut (34), and the allowable stress fluctuation range of the tensile creep specimen. The limit verification formula is as follows: ,in, To smoothly compensate for displacement, This is the initial compensation displacement. To limit the number of turns the load nut (34) can be screwed in, The pitch of the screw assembly. To allow for stress fluctuation amplitude, To determine the cross-sectional area of ​​the tensile creep specimen, This represents the current actual stiffness.

8. The method for testing the tensile creep of cement-based materials based on a horizontal spring loading device according to claim 7, characterized in that: The drive loading nut (34) is rotated at a preset smooth speed curve to smoothly compensate for the displacement, so as to eliminate the step stress impact and maintain the constant tensile stress on the tensile creep specimen. Specifically, it includes: The target rotation stroke of the loading nut (34) is determined based on the smooth compensation displacement and the pitch of the screw device, and the target rotation stroke is decomposed into a low-speed start-up segment, a uniform speed compensation segment and a low-speed convergence segment. Based on the magnitude of the smooth compensation displacement and the load loss in the next time window, a smooth speed curve matching the target rotation path is generated, so that the loading nut (34) rotates at a speed lower than the impact sensitivity threshold in the low-speed start-up section, rotates at a constant speed matching the load loss rate in the uniform speed compensation section, and rotates at a gradually decreasing speed to the target rotation path in the low-speed convergence section. During the low-speed start-up phase, the real-time load change rate of the force sensor (36) and the real-time strain change rate of the dual-channel distributed optical fiber sensor are monitored. When the real-time load change rate or the real-time strain change rate exceeds the preset safety threshold, the screwing of the loading nut (34) is paused and the current position is maintained. When the real-time load change rate and the real-time strain change rate recover to within the preset safety threshold, the uniform speed compensation segment continues to be executed according to the smooth speed curve, so that the compression amount of the compression spring (33) gradually recovers as the loading nut (34) rotates, thus offsetting the load loss caused by the predicted load loss. After the loaded nut (34) reaches the target rotation stroke, maintain the preset stabilization time, and synchronously collect the compensated load data, the compensated apparent total deformation data and the compensated internal strain distribution data during the stabilization time; compare the compensated load data with the target load value, and when it exceeds the allowable fluctuation zone, calculate the fine-tuning displacement according to the remaining deviation and execute the fine-tuning rotation. The compensated load data, compensated apparent total deformation data, and compensated internal strain distribution data are updated into multi-source time series data for prediction of the next time window. The updated multi-source time series data are then used to correct the time-varying evolution model of the compression spring stiffness, so that the tensile stress on the tensile creep specimen is maintained at an approximately constant level within the allowable fluctuation range of the target load value.

9. The method for testing the tensile creep of cement-based materials based on a horizontal spring loading device according to claim 1, characterized in that: Simultaneously collect free shrinkage deformation data of stress-free shrinkage compensation specimens, and use temperature sensing data from distributed fiber optic sensors to calculate temperature compensation deformation. Subtracting the free shrinkage deformation data and temperature-compensated deformation from the actual viscoelastic creep strain yields the pure tensile creep strain; specifically including: The stress-free shrinkage compensation specimen was placed in the same environmental conditions as the tensile creep specimen, and the stress-free shrinkage compensation specimen was isolated from the external support by a friction-reducing layer, and was in a state of no external constraint; a speckle pattern with the same gauge length as the tensile creep specimen was made on the surface of the stress-free shrinkage compensation specimen. Using the same sampling clock as the image acquisition of the tensile creep specimen, the stress-free shrinkage compensation specimen was synchronously triggered and acquired through a non-contact image acquisition device to obtain a shrinkage speckle image sequence; digital image correlation calculations were performed on the shrinkage speckle image sequence and normalized by gauge length to obtain free shrinkage deformation data expressed in strain form; Using temperature sensing data from a dual-channel distributed fiber optic sensor, and taking the initial temperature state as the temperature reference, the average temperature change of the tensile creep specimen and the average temperature change of the stress-free shrinkage compensated specimen were obtained respectively. The temperature difference corrected deformation was then calculated according to the temperature difference correction formula. ,in, To correct for deformation due to temperature difference, The coefficient of thermal expansion of cement-based materials. This represents the average temperature change of the tensile creep specimen. The average temperature change of the stress-free shrinkage compensation specimen; when the two specimens are in the same constant temperature environment and the temperature difference is negligible, the temperature difference correction deformation is zero, and the temperature deformation component already included in the free shrinkage deformation data is no longer deducted. The actual viscoelastic creep strain, free shrinkage deformation data, and temperature-corrected deformation data are unified under the same sampling clock for timing alignment, and the pure tensile creep strain is calculated according to the correction formula: ,in, This represents pure tensile creep strain. To represent the true viscoelastic creep strain after alignment. The data is the aligned free shrinkage deformation data. This is the shrinkage correction factor calibrated by the residual correlation test, initially set to 1. To correct for deformation due to temperature difference; A residual correlation test was performed on the pure tensile creep strain, free shrinkage deformation data, and temperature-corrected deformation: when a residual component exists in the pure tensile creep strain that shows a trend consistent with the free shrinkage deformation data, the shrinkage correction coefficient was updated using regression. And recalculate; when there is a residual component consistent with the trend of the deformation amount corrected by temperature difference, update the coefficient of thermal expansion by regression. And recalculate; when the absolute value of the correlation coefficient between the two residuals is less than the preset significance threshold or the number of iterations reaches the preset upper limit, output the pure tensile creep strain.

10. The method for testing the tensile creep of cement-based materials based on a horizontal spring loading device according to claim 9, characterized in that: Based on the compensated constant load data, the true tensile creep and creep coefficient of cement-based materials were calculated; specifically including: The load data collected by the force sensor (36) under the action of smooth compensation displacement and confirmed to be stable are used as load history data; the stability of the load history data is checked to obtain an effective constant load segment within the preset allowable fluctuation range. Obtain the cross-sectional area of ​​the tensile creep specimen, and determine the equivalent constant tensile stress based on the time-weighted average of the load within the effective constant load segment and the cross-sectional area: ,in, For equivalent constant tensile stress, It is the time-weighted average of the loads within the effective constant load segment; The pure tensile creep strain and the equivalent constant tensile stress are aligned according to the sampling time, and the pure tensile creep strain is converted to the unit equivalent constant tensile stress to obtain the true tensile creep: ,in, To accurately represent the stretch creep. This is time-series data of pure tensile creep strain; The equivalent constant tensile stress is converted into initial elastic strain using the initial elastic modulus, and the ratio between the pure tensile creep strain and the initial elastic strain is determined as the creep coefficient: , ,in, For the initial elastic strain, The initial elastic modulus, The creep coefficient; Based on the holding time, the true tensile creep and creep coefficient are statistically analyzed in segments to obtain the true tensile creep and creep coefficient subsequences corresponding to the early creep stage, the stable creep stage, and the late creep stage. Damage verification was performed on the true tensile creep subsequence and creep coefficient subsequence using microcrack characteristic data. When any subsequence had an abrupt change segment corresponding to the microcrack propagation region, the pure tensile creep strain corresponding to the abrupt change segment was removed and recalculated, and the removed interval was marked in the output results. The recalculated true tensile creep and creep coefficient were output as the true tensile creep and creep coefficient of cement-based materials. The output also included loading age, sealing condition, temperature and humidity records, specimen size and mix proportion, load fluctuation range, and crack inspection results after the test.

Images