A fatigue damage monitoring system and method for carbon fiber hydrogen storage cylinders based on digital twins

Abstract

This invention discloses a fatigue damage monitoring system and method for carbon fiber hydrogen storage cylinders based on digital twins, relating to the field of health monitoring technology for high-pressure hydrogen storage containers. The scheme incorporates a hydrogen-sensitive functional layer between a plastic inner liner and a carbon fiber winding layer. The lattice expansion of the hydrogen-sensitive filler in a hydrogen environment converts permeation flux into interfacial pressure changes. The interfacial pressure sensing unit is sensitive only to normal compressive stress, while the winding layer strain sensing unit is layered along the thickness direction to acquire strain gradients. The sensed signals drive the digital twin models of the inner liner and the winding layer, respectively. By assimilating and inverting the equivalent permeability coefficient of the inner liner and the stiffness degradation distribution of the winding layer, the sealing damage and load-bearing damage states are independently assessed, and the remaining life is predicted based on the evolution trend. This scheme achieves decoupled monitoring and quantitative assessment of two failure modes, improving the early warning sensitivity of decompression damage and the confidence of life prediction.

Description

Technical Field

[0001] This invention belongs to the field of safety monitoring technology for carbon fiber hydrogen storage cylinders, specifically relating to a fatigue damage monitoring system and method for carbon fiber hydrogen storage cylinders based on digital twins. Background Technology

[0002] Type IV high-pressure hydrogen storage cylinders, constructed with carbon fiber reinforced resin matrix composites, have become the mainstream technology for hydrogen storage systems due to their high specific strength, lightweight design, and excellent resistance to hydrogen embrittlement. These cylinders typically consist of a high-density polyethylene or polyamide plastic liner for hydrogen sealing, while an epoxy resin-impregnated carbon fiber winding layer bears the internal high-pressure load. Under frequent hydrogen charging and discharging cycles, the cylinders are subjected to alternating mechanical stress coupled with hydrogen permeation. Accumulated fatigue damage is the core mechanism leading to performance degradation and eventual failure. Therefore, real-time and accurate monitoring and assessment of the fatigue damage status of hydrogen storage cylinders throughout their entire lifespan is a key technical requirement for ensuring the safe operation of on-board hydrogen systems, optimizing maintenance strategies, and extending service life.

[0003] Current fatigue damage monitoring technologies for carbon fiber hydrogen storage cylinders primarily rely on deploying strain sensors on the cylinder surface or at a single interface between the inner liner and the winding layer. By collecting strain response data during hydrogen charging and discharging, and combining this data with empirical formulas or finite element models, the overall damage level of the storage cylinder is inferred. This approach treats the inner liner and carbon fiber winding layer as a mechanical unit, performing mixed collection and fusion analysis of strain data to output a general health indicator. However, the inner liner and carbon fiber winding layer of a hydrogen storage cylinder have fundamentally different functions: the inner liner is responsible for hydrogen sealing, and its failure modes mainly manifest as increased permeability due to microcrack propagation and decompression damage; the carbon fiber winding layer is responsible for bearing internal pressure loads, and its failure modes mainly manifest as matrix microcracks, interlayer delamination, and fiber breakage. The strain data obtained by existing monitoring schemes is the result of the superposition and coupling of inner liner deformation and winding layer deformation. When abnormal monitoring data occurs, it is difficult to accurately distinguish whether the root cause of the failure is a degradation of the inner liner's sealing function or a decrease in the winding layer's load-bearing capacity, making it impossible to take differentiated measures for different failure modes. Meanwhile, hydrogen permeation of the inner liner is a slow chemical and physical process. Traditional strain sensors are not sensitive to the initiation and propagation of microcracks in the inner liner material. During the hydrogen filling and discharging process, the elastic strain caused by internal pressure is dominant. The changes in permeation characteristics caused by the accumulation of microcracks are extremely weak in mechanical response. Existing technologies cannot achieve independent quantitative characterization of the degree of permeation damage to the inner liner in principle. Summary of the Invention

[0004] To address the shortcomings and problems of existing carbon fiber hydrogen storage cylinder monitoring systems, this invention provides a fatigue damage monitoring system and method for carbon fiber hydrogen storage cylinders based on digital twins. This system effectively solves the problem that existing technologies cannot distinguish between inner liner sealing failure and winding layer load-bearing failure, enabling decoupled monitoring, independent quantitative assessment, and accurate fault location for the two types of damage modes.

[0005] The solution adopted by this invention to solve its technical problem is: a fatigue damage monitoring system for carbon fiber hydrogen storage cylinders based on digital twins, comprising a hydrogen storage cylinder body, a sensing module, a data processing module, a digital twin module, and an early warning module. The hydrogen storage cylinder body includes an inner liner and a carbon fiber winding layer arranged sequentially from the inside out. A hydrogen-sensitive functional layer is provided between the inner liner and the carbon fiber winding layer, and the hydrogen-sensitive functional layer is configured to expand in volume when in contact with hydrogen gas. The sensing module includes an interface pressure sensing unit and a winding layer strain sensing unit. The interface pressure sensing unit is located at the interface between the inner liner and the hydrogen-sensitive functional layer to detect the interface contact pressure at that interface. The winding layer strain sensing unit is located at the interface between the inner liner and the hydrogen-sensitive functional layer to detect the interface contact pressure at that interface. Within the carbon fiber winding layer, strain data along the thickness direction of the carbon fiber winding layer is detected. The data processing module is communicatively connected to the sensing module to receive sensing data transmitted by the interface pressure sensing unit and the winding layer strain sensing unit, process the sensing data, and send it to the digital twin module. The digital twin module automatically calculates the degree of permeation damage to the inner liner and the load-bearing damage state of the carbon fiber winding layer in the previous state based on the received sensing data. The digital twin module comprehensively evaluates the remaining life of the entire hydrogen storage cylinder based on the calculation results. The early warning module matches preset thresholds according to the degree of permeation damage to the inner liner and the load-bearing damage state of the carbon fiber winding layer, and executes graded early warnings.

[0006] The hydrogen-sensitive functional layer includes a polymer matrix and a hydrogen-sensitive functional filler dispersed in the polymer matrix. The hydrogen-sensitive functional filler undergoes lattice expansion or volume expansion in a hydrogen environment to drive the polymer matrix to produce a restricted volume expansion.

[0007] The hydrogen-sensitive functional filler includes palladium-based nanoparticles or palladium alloy nanoparticles, and the polymer matrix includes hydrogenated nitrile rubber or nitrile rubber.

[0008] The hydrogen-sensitive functional layer covers at least the shoulder transition area and the middle section of the body of the hydrogen storage bottle, and the thickness of the shoulder transition area and the middle section of the body is greater than the thickness of other areas.

[0009] The strain sensing unit of the winding layer includes multiple strain sensors arranged at different lay-up depths along the thickness direction of the carbon fiber winding layer. The strain sensors are connected to the data processing module to form a multi-layer strain monitoring array, which is used to collect the strain gradient in the thickness direction of the carbon fiber winding layer and identify interlayer delamination, matrix cracking and fiber breakage damage of the winding layer.

[0010] The outer surface of the inner liner has a shallow groove. The interface pressure sensing unit is a fiber optic grating pressure sensor, which is encapsulated in a flexible metal sheet and embedded in the shallow groove on the outer surface of the inner liner. It is covered and pressed from the outside by the hydrogen-sensitive functional layer.

[0011] The digital twin module incorporates a decoupled digital twin model of the inner liner and a digital twin model of the winding layer. The digital twin model of the inner liner is used to invert the degree of permeation damage and remaining life of the inner liner based on the interface contact pressure data. The digital twin sub-model of the winding layer is used to invert the stress damage state and remaining life of the winding layer based on the strain gradient signal. The digital twin module takes the minimum value between the remaining life of the inner liner and the remaining life of the winding layer as the remaining life of the entire hydrogen storage cylinder.

[0012] A fatigue damage monitoring method for carbon fiber hydrogen storage cylinders based on digital twins, applied to the system described in any one of claims 1-7, includes the following steps: Step 1: Obtain the normal contact pressure at the interface between the plastic inner liner and the hydrogen-sensitive functional layer during the hydrogen charging and discharging cycle, as well as the strain gradient of the carbon fiber winding layer along the thickness direction. Step 2: Driven by the measured data of the normal contact pressure, run the digital twin model of the inner liner, and invert the equivalent permeability coefficient of the plastic inner liner through data assimilation. Based on the evolution of the equivalent permeability coefficient, evaluate the permeation damage state and remaining life of the inner liner. Step 3: Driven by the measured strain gradient data, run the digital twin model of the winding layer, and invert the stiffness degradation distribution of each layup of the carbon fiber winding layer through data assimilation. Based on the stiffness degradation distribution, evaluate the load-bearing damage state and remaining life of the winding layer. Step 4: Determine the remaining lifespan of the hydrogen storage cylinder based on the remaining lifespan of the inner liner and the remaining lifespan of the winding layer, and trigger graded early warnings based on the damage status of the inner liner and the winding layer.

[0013] The beneficial effects of this invention: The fatigue damage monitoring system and method for carbon fiber hydrogen storage cylinders based on digital twins provided by this invention have the following beneficial effects: 1. This invention establishes a hydrogen-sensitive functional layer between the outer surface of the inner liner and the inner surface of the carbon fiber winding layer. Utilizing the lattice expansion effect of the hydrogen-sensitive filler in a hydrogen environment, the hydrogen permeation flux is converted into a change in normal contact pressure at the interface, which is independently acquired by an interface pressure sensing unit sensitive only to compressive stress. Simultaneously, strain sensing units are layered along the thickness direction of the carbon fiber winding layer to obtain strain gradient data specifically reflecting the stiffness degradation of the winding layer. Based on this structure, the bidirectional coupling interference between the inner liner tensile strain and the winding layer monitoring signal, and between the inner liner sealing signal and the load-bearing signal, is completely severed from a physical sensing perspective. This allows the inner liner permeation damage factor and the winding layer stiffness degradation field to be independently inverted, significantly improving the accuracy of damage source identification and the precision of fault location.

[0014] 2. This invention normalizes the degree of liner permeation damage into a damage factor D and quantifies the damage state of the winding layer into a stiffness degradation coefficient distribution and a layering pattern. This allows both types of damage to be output as dimensionless or standardized indicators, facilitating direct comparison with preset alarm thresholds. The system can trigger audible and visual alarms based on the liner and winding layer thresholds respectively, guiding maintenance personnel to adopt differentiated maintenance strategies such as replacing the liner or repairing the winding layer for specific failure modes. This avoids over-inspection or under-maintenance caused by unclear fault indications in traditional solutions. Simultaneously, the digital twin model uses measured data as the assimilation target for parameter inversion, eliminating the need for extensive offline destructive testing calibration and reducing the reliance on expensive experimental resources for model calibration.

[0015] 3. The digital twin model of the inner liner constructed in this invention predicts the remaining life based on the trend extrapolation of the permeability coefficient evolution sequence. When the evolution shows an accelerating trend, a nonlinear extrapolation model is automatically used to give a conservative estimate. The digital twin model of the winding layer predicts the remaining number of cycles based on the inverted damage size combined with the fatigue propagation law. The smaller value of the two is taken as the remaining life of the entire system. This dual-model collaborative evaluation mechanism covers the two dominant failure modes of sealing failure and structural failure, enabling the monitoring system to maintain stable damage tracking capabilities under harsh operating conditions such as rapid hydrogen charging and discharging and frequent start-stop, providing a reliable technical basis for safety early warning, scheduled maintenance cycle optimization, and tiered utilization decisions of hydrogen storage cylinders throughout their entire life cycle. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the system architecture of the present invention.

[0017] Figure 2 This is a schematic diagram showing the layout of the interface pressure sensing unit of the present invention.

[0018] Figure 3 This is a schematic diagram of the internal architecture and data-driven process of the digital twin module of the present invention.

[0019] Figure 4This is a curve showing the evolution of the degree of permeation damage to the inner liner of the present invention with the number of cycles.

[0020] Figure 5 This is a flowchart of the monitoring method of the present invention.

[0021] The diagram is labeled as follows: 10 is the inner liner, 11 is the shallow groove, 20 is the hydrogen-sensitive functional layer, and 30 is the flexible metal sheet. Detailed Implementation

[0022] The present invention will be further described below with reference to the accompanying drawings and embodiments. Example

[0023] In view of the problems raised in the background art above, this embodiment provides a fatigue damage monitoring system for carbon fiber hydrogen storage cylinders based on digital twins. Its core structure includes a hydrogen storage cylinder body, a sensing module, a data processing module, and a digital twin. The hydrogen storage cylinder body is a Type IV high-pressure composite container, comprising, from the inside out: a plastic inner liner 10, a hydrogen-sensitive functional layer 20, and a carbon fiber winding layer 40. The main function of the plastic inner liner 10 is to provide airtightness and prevent hydrogen leakage. Its material is preferably high-density polyethylene (HDPE) or polyamide (PA), and its thickness typically ranges from 8 mm to 12 mm. The main function of the carbon fiber winding layer 40 is to bear the internal high-pressure load. It is formed by winding epoxy resin-impregnated carbon fibers (such as T700 grade), and the layup angle typically includes circumferential winding and helical winding.

[0024] A hydrogen-sensitive functional layer 20 is disposed between the outer surface of the inner liner 10 and the inner surface of the carbon fiber winding layer 40. The core function of this hydrogen-sensitive functional layer 20 is to undergo restricted volume expansion upon contact with trace amounts of hydrogen permeating through the inner liner 10, thereby converting the chemical signal (hydrogen concentration) into a mechanical signal (interfacial pressure change). The coverage area of ​​the hydrogen-sensitive functional layer 20 includes at least the shoulder transition area and the middle section of the bottle body. Preferably, the thickness of the hydrogen-sensitive functional layer 20 in the shoulder transition area and the middle section of the bottle body is greater than its thickness in other areas such as the end cap; for example, the former has a thickness of 4 mm to 5 mm, and the latter has a thickness of 2 mm to 3 mm, to accommodate the stress concentration and hydrogen permeation risk levels in different areas.

[0025] Furthermore, the material composition of the hydrogen-sensitive functional layer 20 specifically includes a polymer matrix and a hydrogen-sensitive functional filler dispersed in the matrix. The polymer matrix is ​​preferably hydrogenated nitrile butadiene rubber (HNBR) or nitrile butadiene rubber (NBR) because it possesses good hydrogen compatibility, elastic recovery ability, and processability. The hydrogen-sensitive functional filler is configured to undergo lattice expansion in a hydrogen environment, thereby driving macroscopic volume expansion of the surrounding polymer matrix. In a preferred embodiment, the hydrogen-sensitive functional filler is palladium (Pd)-based nanoparticles or palladium alloy nanoparticles, with a preferred particle size range of 50 nm to 100 nm. To prevent the palladium nanoparticles from being oxidized or vulcanized and deactivated during processing and use, a dense silica (SiO2) antioxidant protective layer is coated on its surface, forming a core-shell structure of Pd@SiO2 nanoparticles.

[0026] The sensing module includes an interface pressure sensing unit and a winding layer strain sensing unit, enabling independent monitoring of the inner liner's sealing status and the winding layer's stress state. Specifically: like Figure 1 and Figure 2 As shown, the interface pressure sensing unit is disposed at the interface between the inner liner 10 and the hydrogen-sensitive functional layer 20. The interface pressure sensing unit is specifically used to detect the normal contact pressure (interface contact pressure) at this interface. In one embodiment, the interface pressure sensing unit adopts a miniature fiber optic grating (FBG) pressure sensor, which is encapsulated in a flexible metal sheet 30 and embedded in a shallow groove 11 prefabricated on the outer surface of the inner liner 10. The hydrogen-sensitive functional layer 20 covers and presses it from the outside, forming a stable mechanical contact with bidirectional clamping. The interface pressure sensing unit is only sensitive to compressive stress perpendicular to the interface and is not sensitive to tensile strain along the surface of the inner liner, thus avoiding the tensile-compression coupling interference of traditional strain gauges from a physical principle perspective.

[0027] like Figure 1 and Figure 3 As shown, the strain sensing unit of the winding layer is arranged inside the carbon fiber winding layer 40. In order to realize the perception of delamination damage of the winding layer, the strain sensing unit of the winding layer preferably includes multiple strain sensors arranged at different ply depths along the thickness direction of the carbon fiber winding layer. For example, the first group of sensors can be arranged between the 2nd to 4th plies near the inner layer, the second group of sensors can be arranged between the 7th to 9th plies in the middle layer, and the third group of sensors can be arranged between the 12th to 14th plies in the outer layer. Each group of sensors is evenly distributed along the circumference to obtain the strain gradient data of the carbon fiber winding layer in the thickness direction, specifically reflecting the evolution process of load-bearing damage such as interlayer delamination and fiber breakage of the winding layer.

[0028] The data processing module is communicatively connected to the sensing module. It receives sensing data transmitted from the interface pressure sensing unit and the winding layer strain sensing unit connected to the sensing module, processes the sensing data, and then sends it to the digital twin module. The digital twin module automatically calculates the degree of permeation damage to the inner liner and the stress damage state of the carbon fiber winding layer in the previous state based on the received sensor data, and sends the calculation results to the early warning module. The digital twin module comprehensively evaluates the remaining life of the hydrogen storage cylinder based on the calculation results. The early warning module compares the calculation results with the preset alarm threshold. When the calculation results reach the early warning level, the early warning module triggers an audible and visual alarm.

[0029] The digital twin module internally constructs and runs two decoupled digital twin sub-models: the inner liner digital twin model and the winding layer digital twin model. The task of the inner liner digital twin model is to synchronize the hydrogen permeation state of the virtual inner liner with the physical entity, driven by measured data of interface contact pressure, and then calculate the degree of permeation damage and remaining lifespan from the model's internal parameters. Specifically: The digital twin model of the inner liner is constructed based on the theory of hydrogen diffusion and permeation in polymers. Geometrically, the model is discretized into one-dimensional or two-dimensional diffusion units according to the actual wall thickness of the inner liner. Each unit is assigned initial material parameters, including the hydrogen solubility coefficient, diffusion coefficient, and permeation coefficient defined by their product. The model boundary conditions are set as follows: the inner wall of the inner liner is subjected to a periodic hydrogen pressure load consistent with actual working conditions; the outer wall of the inner liner is in contact with the hydrogen-sensitive functional layer, which is equivalent to an elastic constraint boundary with a hydrogen-induced expansion effect. A quantitative relationship is established between the expansion response of the hydrogen-sensitive functional layer and the hydrogen flux permeating through the outer wall of the inner liner through a pre-calibrated hydrogen-induced expansion coefficient curve. This allows the hydrogen escape flux at the outer wall of the inner liner to be converted into an equivalent interfacial pressure acting on the outer wall of the inner liner.

[0030] After each hydrogen charge / discharge cycle, the data processing module acquires the residual interface pressure value measured by the interface pressure sensing unit during the pressure relief phase, and compares it with the residual pressure baseline value of the factory reference cycle to obtain the hydrogen-induced pressure increment for the current cycle. This increment serves as the data assimilation target for the inner tank sub-model in this cycle.

[0031] The digital twin model of the inner liner uses a preset permeability coefficient as the variable to be calibrated. It recreates the hydrogen pressure history of this cycle in a virtual environment, calculates the hydrogen escape flux at the virtual inner liner's outer wall and the corresponding virtual interface pressure increment, and adjusts the equivalent permeability coefficient of the inner liner material through an iterative optimization algorithm until the virtual interface pressure increment matches the measured value within the allowable error range. At this point, the permeability coefficient of the inner liner digital twin model has successfully "imitated" the current state of the physical entity, meaning the model's internal parameters are synchronized with the actual permeability capacity of the physical inner liner.

[0032] The equivalent permeability coefficient of the synchronized inner liner sub-model reflects the macroscopic permeability of the current inner liner material in the presence of microcracks. The degree of permeability damage to the inner liner is quantified by a normalized damage factor D. D = (P isointense - P initial) / (P imminent infiltration - P initial) Wherein, Pinitial is the initial equivalent permeability coefficient obtained by inversion under factory conditions, Pisopermeable is the equivalent permeability coefficient obtained by inversion in the current cycle, and Pcriticalpermeability is the critical permeability coefficient threshold for the failure of the inner liner's sealing function (determined by material testing). The damage factor increases from 0 to 1, intuitively characterizing the degradation process of the inner liner from intact to near-sealing failure.

[0033] The digital twin model of the inner liner records the equivalent permeability coefficient after each calibration cycle, forming an evolution sequence of the permeability coefficient with the number of cycles. Based on the trend extrapolation of this sequence, the digital twin model of the inner liner predicts the remaining number of cycles required for the permeability coefficient to reach the critical threshold P before leakage occurs, i.e., the remaining service life of the inner liner. If the evolution sequence shows an accelerating growth trend, the model automatically adopts a nonlinear extrapolation model to give a more conservative life estimate.

[0034] The task of the digital twin model of the winding layer is to use measured strain gradient data at each layup depth as a driving force to synchronize the stiffness degradation state of the virtual winding layer with the physical entity, thereby inverting the internal microcrack distribution and assessing the remaining lifetime. Specifically: The winding layer sub-model is constructed based on the progressive damage theory of composite laminates. Following the actual winding process, the carbon fiber winding layer is divided into multiple single-layer plate elements along its thickness direction. Each single layer is assigned an initial stiffness matrix and strength parameters, including longitudinal tensile / compressive strength, transverse tensile / compressive strength, and in-plane shear strength. Interface elements are set between the layers, assigned initial interlaminar fracture toughness parameters to simulate delamination behavior. The model boundary conditions are: the inner surface bears a uniformly distributed internal pressure load transmitted from the inner liner, and the outer surface is a free boundary. The constitutive relation of the model adopts a stiffness degradation scheme based on continuum damage mechanics; when the element stress state meets a specific failure criterion, the stiffness in the corresponding direction is reduced by a preset degradation coefficient.

[0035] During the pressure holding phase of each hydrogen charge / discharge cycle, the system collects strain values ​​output from the strain sensing units of the inner, middle, and outer layers, forming strain gradient distribution data along the thickness direction. The wound layer sub-model uses this measured strain gradient as the data assimilation target. The model executes a finite element model correction program: using the damage variable field to be inverted (i.e., the stiffness degradation coefficient of each ply element and the connection stiffness of the interlayer interface elements) as optimization variables, it calls the virtual model to calculate the virtual strain values ​​of each layer under the same internal pressure load, constructing a residual objective function between the virtual strain and the measured strain. Through iterative optimization, the virtual strain gradient and the measured strain gradient achieve the best fit. The converged stiffness degradation coefficient distribution represents the damage state of each region inside the wound layer under the current state.

[0036] The in-plane stiffness degradation coefficient of the ply unit obtained during the above calibration process reflects the cumulative degree of matrix cracks and fiber fractures in that region. Units with stiffness degradation coefficients exceeding a preset threshold are marked as damaged areas: areas with significant transverse stiffness degradation but slight longitudinal stiffness degradation are identified as matrix microcrack areas; areas with significant longitudinal stiffness degradation are identified as fiber fracture areas. Locations where the stiffness of interlayer interface units decreases significantly are identified as delamination damage areas. Summarizing the spatial coordinates of each damaged unit yields a three-dimensional distribution map of microcracks and delamination within the carbon fiber winding layer under the current state, including quantitative information such as damage location, damage type, and damage area.

[0037] The winding layer sub-model is based on the maximum delamination size or cumulative damage area currently derived. Combined with the fatigue damage propagation law of carbon fiber composites, it calculates the remaining number of hydrogen charge / discharge cycles required for the damage to propagate from the current size to the critical failure size (such as delamination penetrating the wall thickness, fiber fracture area exceeding the limit, or stiffness degradation reaching the preset limit). This is the remaining service life of the winding layer.

[0038] The digital twin module sends the estimated permeation damage level of the inner liner and the load-bearing damage level of the carbon fiber winding layer in the previous state to the early warning module. The digital twin module takes the smaller of the estimated remaining lifespan of the inner liner and the remaining lifespan of the winding layer as the remaining lifespan of the entire hydrogen storage cylinder.

[0039] The early warning module compares the calculated results with preset alarm thresholds, including the inner liner threshold and the winding layer threshold. The audible and visual alarms include the inner liner audible and visual alarm and the winding layer audible and visual alarm. After receiving the calculated results, the early warning module compares the degree of permeation damage of the inner liner with the inner liner threshold and the load-bearing damage state of the carbon fiber winding layer with the winding layer threshold. When the degree of permeation damage of the inner liner reaches the early warning level, the early warning module triggers the inner liner audible and visual alarm; when the load-bearing damage state of the carbon fiber winding layer reaches the early warning level, the early warning module triggers the winding layer audible and visual alarm. Example

[0040] This embodiment provides a fatigue damage monitoring method for carbon fiber hydrogen storage cylinders based on digital twins. This method can be implemented based on the monitoring system of Embodiment 1, and includes the following steps: Step S100: Obtain the normal contact pressure at the interface between the plastic inner liner and the hydrogen-sensitive functional layer during the hydrogen charging and discharging cycle, as well as the strain gradient of the carbon fiber winding layer along the thickness direction.

[0041] In this step, the normal contact pressure is preferably acquired during the pressure relief phase after each hydrogen charge / discharge cycle. At this time, the residual interface pressure value output by the interface pressure sensing unit can effectively reflect the cumulative effect of permeated hydrogen. The strain gradient is preferably acquired during the pressure holding phase of the hydrogen charge / discharge cycle. At this time, the internal pressure is stable, and the winding layer is in a typical load-bearing state, which is conducive to obtaining stable and reliable strain distribution data.

[0042] Step S200: Driven by the measured data of normal contact pressure, run the digital twin model of the inner liner, assimilate and invert the equivalent permeability coefficient of the plastic inner liner through data, and evaluate the permeation damage state and remaining life of the inner liner based on the evolution of the equivalent permeability coefficient.

[0043] The inversion process of the digital twin model of the inner liner is consistent with that described in Example 1. It should be noted that as the number of hydrogen charging and discharging cycles increases, the inversion results of the equivalent permeability coefficient will gradually form a time series. This time series can be used not only for assessing the current damage state but also for training a prediction model to predict permeation behavior in subsequent cycles.

[0044] Step S300: Driven by the measured strain gradient data, run the digital twin model of the winding layer, and invert the stiffness degradation distribution of each layer of the carbon fiber winding layer through data assimilation. Based on the stiffness degradation distribution, evaluate the load-bearing damage state and remaining life of the winding layer.

[0045] The inversion process of the digital twin model of the winding layer is consistent with that described in Example 1. In a preferred embodiment, the digital twin model of the winding layer can also utilize the strain gradient change rate between adjacent cycles as an auxiliary assimilation target to further improve the detection sensitivity of early micro-damage.

[0046] Step S400: Determine the remaining life of the hydrogen storage cylinder based on the remaining life of the inner liner and the remaining life of the winding layer, and trigger an alarm when the damage status or remaining life reaches a preset threshold.

[0047] The principle for determining the remaining lifespan of the entire machine is the same as that described in Example 1, that is, taking the smaller value between the remaining lifespan of the inner liner and the remaining lifespan of the winding layer.

[0048] It should be understood that the specific embodiments described above are merely illustrative or explanatory of the principles of the present invention and do not constitute a limitation thereof. Therefore, any modifications, equivalent substitutions, improvements, etc., made without departing from the spirit and scope of the present invention should be included within the protection scope of the present invention.

Claims

1. A fatigue damage monitoring system for carbon fiber hydrogen storage cylinders based on digital twins, comprising a hydrogen storage cylinder body, a sensing module, a data processing module, a digital twin module, and an early warning module, characterized in that, The hydrogen storage cylinder body includes an inner liner and a carbon fiber winding layer arranged sequentially from the inside out. A hydrogen-sensitive functional layer is provided between the inner liner and the carbon fiber winding layer. The hydrogen-sensitive functional layer is configured to expand in volume when in contact with hydrogen gas. The sensing module includes an interface pressure sensing unit and a winding layer strain sensing unit. The interface pressure sensing unit is located at the interface between the inner liner and the hydrogen-sensitive functional layer to detect the interface contact pressure at that interface. The winding layer strain sensing unit is located within the carbon fiber winding layer to detect the strain data of the carbon fiber winding layer along the thickness direction. The processing module is communicatively connected to the sensing module to receive sensing data transmitted from the interface pressure sensing unit and the winding layer strain sensing unit, process the sensing data, and send it to the digital twin module. The digital twin module automatically calculates the degree of permeation damage to the inner liner and the load-bearing damage state of the carbon fiber winding layer in the previous state based on the received sensing data. The digital twin module comprehensively evaluates the remaining life of the entire hydrogen storage cylinder based on the calculation results. The early warning module matches preset thresholds according to the degree of permeation damage to the inner liner and the load-bearing damage state of the carbon fiber winding layer, and executes graded early warnings.

2. The fatigue damage monitoring system for carbon fiber hydrogen storage cylinders based on digital twins according to claim 1, characterized in that, The hydrogen-sensitive functional layer includes a polymer matrix and a hydrogen-sensitive functional filler dispersed in the polymer matrix. The hydrogen-sensitive functional filler undergoes lattice expansion or volume expansion in a hydrogen environment to drive the polymer matrix to produce a restricted volume expansion.

3. The fatigue damage monitoring system for carbon fiber hydrogen storage cylinders based on digital twins according to claim 2, characterized in that, The hydrogen-sensitive functional filler includes palladium-based nanoparticles or palladium alloy nanoparticles, and the polymer matrix includes hydrogenated nitrile rubber or nitrile rubber.

4. The fatigue damage monitoring system for carbon fiber hydrogen storage cylinders based on digital twins according to claim 1, characterized in that, The hydrogen-sensitive functional layer covers at least the shoulder transition area and the middle section of the body of the hydrogen storage bottle, and the thickness of the shoulder transition area and the middle section of the body is greater than the thickness of other areas.

5. The fatigue damage monitoring system for carbon fiber hydrogen storage cylinders based on digital twins according to claim 1, characterized in that, The strain sensing unit of the winding layer includes multiple strain sensors arranged at different lay-up depths along the thickness direction of the carbon fiber winding layer. The strain sensors are connected to the data processing module to form a multi-layer strain monitoring array, which is used to collect the strain gradient in the thickness direction of the carbon fiber winding layer and identify interlayer delamination, matrix cracking and fiber breakage damage of the winding layer.

6. The fatigue damage monitoring system for carbon fiber hydrogen storage cylinders based on digital twins according to claim 1, characterized in that, The outer surface of the inner liner has a shallow groove. The interface pressure sensing unit is a fiber optic grating pressure sensor, which is encapsulated in a flexible metal sheet and embedded in the shallow groove on the outer surface of the inner liner. It is covered and pressed from the outside by the hydrogen-sensitive functional layer.

7. The fatigue damage monitoring system for carbon fiber hydrogen storage cylinders based on digital twins according to claim 1, characterized in that, The digital twin module incorporates a decoupled digital twin model of the inner liner and a digital twin model of the winding layer. The digital twin model of the inner liner is used to invert the degree of permeation damage and remaining life of the inner liner based on the interface contact pressure data. The digital twin sub-model of the winding layer is used to invert the stress damage state and remaining life of the winding layer based on the strain gradient signal. The digital twin module takes the minimum value between the remaining life of the inner liner and the remaining life of the winding layer as the remaining life of the entire hydrogen storage cylinder.

8. A method for monitoring fatigue damage of carbon fiber hydrogen storage cylinders based on digital twins, applied to the system described in any one of claims 1-7, characterized in that, Includes the following steps: Step 1: Obtain the normal contact pressure at the interface between the plastic inner liner and the hydrogen-sensitive functional layer during the hydrogen charging and discharging cycle, as well as the strain gradient of the carbon fiber winding layer along the thickness direction. Step 2: Driven by the measured data of the normal contact pressure, run the digital twin model of the inner liner, and invert the equivalent permeability coefficient of the plastic inner liner through data assimilation. Based on the evolution of the equivalent permeability coefficient, evaluate the permeation damage state and remaining life of the inner liner. Step 3: Driven by the measured strain gradient data, run the digital twin model of the winding layer, and invert the stiffness degradation distribution of each layup of the carbon fiber winding layer through data assimilation. Based on the stiffness degradation distribution, evaluate the load-bearing damage state and remaining life of the winding layer. Step 4: Determine the remaining lifespan of the hydrogen storage cylinder based on the remaining lifespan of the inner liner and the remaining lifespan of the winding layer, and trigger graded early warnings based on the damage status of the inner liner and the winding layer.

Images