Time-lapse real-time mixing test method based on multi-task local restart loading

Abstract

The invention discloses a time-lapse real-time hybrid test method based on multi-task local restart loading, and aims to solve the problems that the test efficiency is low and remarkable uncertainty and safety risk are brought due to the fact that the loading of each integral time step of a current test substructure needs to be recovered to an initial state and each loading needs a seismic period. According to the method, the mechanical property of a test substructure is dynamically evaluated to determine the local restart loading step number, and a loading target at the initial calculation moment is sequentially subjected to one-time delay compensation and multi-task loading of a servo loading system to form a preparatory loading target; completing a local multi-round restart loading process on the single test substructure one by one according to the loading command of each partition for the pre-loading target along with the increase of the integral time step; according to the invention, the requirements on test technical operation and test cost are effectively reduced, and meanwhile, the loading process of a plurality of test substructures can be completed on the premise of high precision and efficiency.

Description

Technical Field

[0001] The present invention specifically relates to a time-lapse real-time hybrid test method based on multi-task local restart loading, and belongs to the technical field of structural hybrid testing. Background Art

[0002] Hybrid testing methods organically combine numerical simulation with experimental methods. Numerical simulation is used for components that can be accurately simulated numerically, while physical testing is used for components that are difficult to simulate numerically. This approach not only effectively reduces testing costs but also achieves high test accuracy. For structures with velocity-dependent components, a real-time hybrid testing method is used. This method uses a high-speed communication channel between the numerical simulation and physical testing components to exchange data within each integration step. When studying large, complex structures, real-time hybrid testing often uses simplified numerical models and fast integration algorithms. This approach leads to large model errors and inevitably sacrifices test accuracy. When complex structures contain multiple nonlinear components, real-time hybrid testing often requires coordinated loading from multiple loading devices. This loading method places high demands on laboratory hardware and software, making such tests difficult to conduct.

[0003] In recent years, a real-time hybrid test method based on single-specimen restart and multi-task loading has been developed. This method can effectively solve the computational time lag problem of real-time hybrid testing. In addition, the dynamic performance test of multiple nonlinear components is effectively realized through multi-task loading technology. However, it still has limitations. The main reason is that when this type of method is used for large and / or complex structures, the complexity of the numerical model will lead to repeated reassembly of the test substructure. Long-term loading may cause unpredictable changes such as slip / relaxation of the connection interface between the actuator and the specimen, degradation of mechanical properties, and drift of system dynamic parameters. The above unpredictable changes do not meet the requirements of restarting, which may cause problems in the test or inconsistent results, making the test results unable to be put into subsequent use.

[0004] At present, the computational delay in real-time hybrid test methods is difficult to overcome. The main reasons are as follows:

[0005] First, the complexity of the numerical substructure. Real-time hybrid testing methods often involve complex numerical substructures with numerous degrees of freedom and nonlinear characteristics. As the complexity of the numerical substructure increases, the computational time required to solve the equations of motion also increases significantly. This computational time can lead to integration lags, where the computation time for certain time steps exceeds the integration step size, preventing the finite element software from completing the solution within the specified timeframe, resulting in a brief pause in the entire testing system.

[0006] Second, the choice of integration algorithm significantly influences computational lag. Implicit integration algorithms require iterative calculations to solve the equations of motion, which typically results in longer computation times. In contrast, explicit integration algorithms solve the equations directly without iteration, resulting in shorter computation times. However, in practical applications, implicit integration algorithms may be chosen due to their greater numerical stability, further increasing the potential for computational lag.

[0007] Third: The real-time hybrid test method requires that the simulation results of the numerical substructure and the measurement results of the experimental substructure be synchronously coupled at the substructure interface. This real-time requirement makes computational delay a difficult problem to overcome. This is because any computational delay may cause the information exchange between the substructures to be asynchronous, thus affecting the accuracy and stability of the test. In addition, hardware system limitations also make computational delay difficult to overcome. In order to reduce the impact of computational delay, some measures can be taken, such as simplifying the numerical model, optimizing the integration algorithm, and improving the performance of the hardware system. However, these measures often require a trade-off between experimental accuracy and stability and still fail to completely solve the problem of computational delay.

[0008] The root cause of computational lag lies in the complexity of numerical substructure modeling. Its refined finite element analysis requires a large amount of computing resources, resulting in a single-step calculation time exceeding the millisecond time threshold required by the real-time hybrid test method. This problem has not been effectively solved so far.

[0009] In summary, because the current loading of the test substructure must be restored to its initial state at each integral time step, and each loading cycle requires a full earthquake cycle, this leads to low test efficiency and introduces significant uncertainty and safety risks. The relatively high test time cost of the real-time hybrid testing method based on single-specimen restart and multi-task loading cannot be improved. These issues have seriously restricted the development and application of the real-time hybrid testing method based on single-specimen restart and multi-task loading. Summary of the Invention

[0010] Based on the above-mentioned difficult problems, the present invention provides a time-lapse real-time hybrid test method based on multi-task local restart loading.

[0011] A time-lapse real-time hybrid test method based on multi-task local restart loading is proposed. The time-lapse real-time hybrid test method determines the number of local restart loading steps according to the mechanical properties of the test substructure. The loading target at the initial moment is calculated and subjected to a time lag compensation and multi-task loading of the servo loading system to form a preliminary loading target. The preliminary loading target is completed one by one with the increase of the integral time step and according to the loading commands of each partition to complete the local multi-round restart loading process of each test substructure.

[0012] As a preferred solution: the time-lapse real-time mixing test method includes the following steps, specifically:

[0013] Step 1: The prototype structure includes a numerical substructure and a test substructure. The number of restart steps for the mechanical properties of the test substructure is determined based on the preliminary test.

[0014] Step 2: Establish numerical models of the numerical substructure of the prototype structure and the numerical model of the test substructure respectively, and obtain the loading target at the initial moment through the numerical simulation system;

[0015] Step 3: First, perform a time lag compensation on the initial loading target. Then, use the servo loading system to perform multi-task loading on a single test substructure to obtain the initial loading target. Finally, reset the actuator to the initial time and wait for the next loading target.

[0016] Step 4: Use the step-by-step integration algorithm to calculate the motion equation of the multi-degree-of-freedom system, solve the loading command in step i, and then feed the new target back to the multi-tasking part through the predetermined communication mechanism to restart the loading module. The motion equation is:

[0017] M[a i ]+C[v i ]+K[d i ]+[F Em,i ]=-Ma i Formula 1

[0018] In the above formula, M is the mass matrix of the numerical substructure; C is the damping matrix of the numerical substructure; K is the stiffness matrix of the numerical substructure; [a i ] is the acceleration matrix of the numerical substructure; [v i ] is the velocity matrix of the numerical substructure; [d i ] is the displacement matrix of the numerical substructure; [F Em ,i] is the restoring force matrix of the test substructure; a i is the earthquake acceleration excitation; the subscript i represents the integration time step;

[0019] Step 5: The multi-task part restarts the loading module to store the new target, integrates the new target with the initial displacement to form a global loading command, and generates a reset-wait command for the test substructure through the global loading command;

[0020] Step 6: The global load command obtained in step 5 is converted into multiple subtask commands. The conversion process is as follows:

[0021] After determining the partial loading command for each subtask, multiple subtask commands are subjected to secondary time lag compensation to form multiple compensated partial loading commands. The test system sequentially executes the multiple compensated partial loading commands to complete the process of applying displacement loads to the test substructure. After the loading command is executed, the test substructure executes the reset-wait command, and the acquisition system sends back the measured reaction force and displacement.

[0022] Step 7: Restart the module to check whether the measured response meets the required accuracy. If the measured response meets the accuracy requirement, proceed to the next step to execute the reset and wait command. If the measured response does not meet the accuracy requirement, repeat the loading until the accuracy is met.

[0023] Repeat steps 4 to 7 until all integration steps are loaded and the test ends.

[0024] As a preferred solution, the process of forming a preliminary loading target by sequentially subjecting the loading target at the initial calculation moment to time lag compensation and multi-task loading by the servo loading system includes a process of obtaining the total completion time in the time-lapse real-time hybrid test. The process of obtaining the total completion time in the time-lapse real-time hybrid test is as follows:

[0025] The duration of the seismic wave is T, the time interval is Δt, and the total reset-wait time of each integration step is a constant λ total , the partial restart loading step is k, the number of test substructures (3) is n, and the equation corresponding to the total completion time is Formula 2:

[0026]

[0027] As a preferred solution: the loading time corresponding to each test substructure is kΔt, that is, the loading process experienced at the i-th step is from Δt to (1+k)Δt, where 0≤(1+k)Δt≤T; when the restart loading command of each layer from the first step to the i-th step is processed into a data segment of length kΔt according to the i-th step, (i+1)-th step, and (i+2)-th step (i≥1), each data segment corresponds to a partition, and there are a total of (i-j+1) data partitions, that is, each data partition is a local restart loading command, where k=(i-j+1).

[0028] As a preferred solution: the time-lapse real-time hybrid test method is a hybrid test loading process in which a set of loading equipment is used to perform multi-task local restart loading on a test substructure, and the dynamic performance of all test substructures is verified with limited test facilities. In the i-th integration step, a set of actuators is first used to execute partial restart loading commands of task 1, task 2, ..., task n on a single test substructure, and then n measured force and displacement data are collected as measured data, and the measured data are fed back to the numerical system. At the same time, the test substructure executes a reset-wait command.

[0029] As a preferred solution: the time-lapse real-time hybrid test method also includes a reset and synchronization waiting accuracy process. The reset and synchronization waiting accuracy process is to compare the numerical simulation system with the measured data of all previous steps of the test substructure after each step of loading is completed, thereby completing the process of verifying the loading and reset of the test substructure and the waiting accuracy.

[0030] As a preferred solution: after the loading target at the initial moment of calculation is subjected to a time lag compensation and multi-task loading of the servo loading system to form a preliminary loading target, when i=i+1 in the preliminary loading target, the numerical simulation system is used for step-by-step integration calculation to generate partial loading, reset and wait commands, and then after a second time lag compensation, the multi-task partial loading command is executed until the structural response meets the accuracy requirements and the reset and wait commands are executed.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1. The present invention is a multi-task local restart loading real-time hybrid test method that avoids computational lags, and uses finite element software to carry out refined modeling, thereby improving test accuracy. The present invention can determine the number of steps of local restart loading based on the mechanical properties of the test substructure, and achieve the process of performing local multiple-round restart loading on a single test substructure one by one according to the loading commands of each partition as the integral time step increases, while ensuring the matching of the current step physical quantity of the test substructure, and efficiently and accurately revealing the dynamic performance of multiple test substructures in a time-increasing manner. The present invention significantly shortens the test loading time. According to the test efficiency data, the present invention can shorten the loading time by at least 50%, and improve the influence of computational lag.

[0033] Second, this invention improves the conversion module for partial restart loading commands based on a real-time hybrid test method based on multi-task partial restart loading technology. While ensuring that the current mechanical state variables of the current test substructure match, this invention performs local loading of the test substructure in a time increment based on partition commands, significantly reducing the number of restart loading steps for the test substructure and improving test efficiency to a quantitative degree.

[0034] 3. The present invention can complete multiple rounds of partial restart loading from step j to step i on a single test substructure, thereby testing the dynamic performance of multiple test substructures, effectively reducing the demand for test technical operations and test costs, and at the same time being able to complete the loading process of multiple test substructures with high accuracy and efficiency.

[0035] 4. The real-time hybrid test method based on the multi-task local restart loading technology proposed in the present invention also includes a restart accuracy verification operation. The restart accuracy verification operation is to verify the mechanical quantities of all previous integral time steps of the test substructure every time the loading of an integral time step is completed, so as to stably, efficiently and accurately realize the partial restart loading, resetting and waiting for subsequent loading commands of the test substructure. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is a flow chart of the real-time hybrid test method based on multi-task local restart loading technology;

[0037] Figure 2 This is a flow chart showing the principle of a real-time hybrid test method based on multi-task local restart loading technology. The figure takes a seven-story frame vibration / seismic structure with six isolation bearings as an example.

[0038] Figure 3 This is a flow chart showing the principle of a real-time hybrid test method based on multi-task local restart loading technology. The figure uses a four-span bridge vibration / seismic structure equipped with three viscous dampers as an example.

[0039] Figure 4 This is a flow chart showing the principle of a real-time hybrid test method based on multi-task local restart loading technology. The figure uses a seven-story frame vibration / seismic structure equipped with seven viscous dampers as an example.

[0040] In the figure, 1-prototype structure; 2-numerical substructure; 3-experimental substructure. DETAILED DESCRIPTION

[0041] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0042] Specific implementation method 1: Combination Figures 1 to 4 The present embodiment is described. In the present embodiment, the time-lapse real-time hybrid test method based on multi-task local restart loading is specifically as follows: the number of steps of local restart loading is determined according to the mechanical properties of the test substructure 3, the loading target at the initial moment of calculation is subjected to a time lag compensation and multi-task loading of the servo loading system in sequence to form a preliminary loading target, and the preliminary loading target is completed one by one with the increase of the integral time step and according to the loading commands of each partition to complete the local multi-round restart loading process of a single test substructure.

[0043] In the specific implementation process of the present invention, when i=1, the number of partial restart loading steps is determined, the loading target at the initial moment of calculation is calculated, and the calculated loading target is the preliminary loading target. The preliminary loading target is subjected to time lag compensation, and the time lag compensation is the first time lag compensation. The preliminary loading target after the time lag compensation is subjected to a multi-task loading process through the servo loading system. The number of partial restart steps for the mechanical properties of the test substructure 3 is k. When loading is performed until the i=i+1th time, when i=i+1=k, the corresponding structural response is obtained; when i=i+1≠k, the numerical simulation system is used to perform step-by-step loading. After performing integral calculation and generating partial loading and reset-wait commands, a second time delay compensation is performed. After executing the multi-task partial loading command, an accuracy judgment is performed. When the accuracy value meets the specific corresponding requirements, the corresponding structural response is obtained in the state of i=i+1=k after executing the reset-wait command, and the loading process is completed. When the accuracy value does not meet the specific corresponding requirements, the multi-task partial loading command is repeatedly executed and an accuracy judgment is performed again. When the accuracy value meets the specific corresponding requirements, the corresponding structural response is obtained in the state of i=i+1=k after executing the reset-wait command, and the loading process is completed.

[0044] During the test process of the present invention, local loading is applied to test substructure 3 in time increments according to partitioning commands, while ensuring that the current mechanical state variables of the test substructure 3 match. This invention significantly reduces the number of restart loading steps required for test substructure 3, significantly improving test efficiency and providing a quantitative comparison. Specific data is shown in Table 1 below, where "Convention" represents the test time required for the conventional loading technique, while "Novel" represents the test time required for the proposed loading technique.

[0045] Table 1 Comparison of loading time under different working conditions

[0046]

[0047] As can be seen from the above table, the loading time of the present invention is quantitatively and significantly reduced, which is also conducive to significantly reducing the number of restart loading steps of the test substructure 3 and significantly improving the test efficiency.

[0048] The present invention can implement a partial loading process that is different from the full-load principle, and is a real-time hybrid test process based on multi-task partial restart loading technology. The present invention includes a multi-task command conversion module for partial restart loading. Specifically, in step i, a single test substructure 3 can be subjected to partial restart loading from step j to step i in multiple rounds, thereby testing the dynamic performance of multiple test substructures 3. Wherein, (i-j+1)=k. This technology effectively reduces the requirements for test technical operations and test time costs, while achieving a high degree of accuracy and efficiency in the loading process of multiple test substructures 3.

[0049] The present invention also includes a restart accuracy verification module. Its verification principle is to verify the mechanical quantities of all previous integration time steps of the test substructure 3 after each integration time step. This allows for stable and efficient partial restart loading, resetting, and waiting for subsequent loading commands for the test substructure 3.

[0050] The present invention not only solves the computational time lag problem of real-time hybrid testing, but also can form a refined modeling process, specifically by carrying out refined modeling through the finite element method, thereby facilitating improved test accuracy.

[0051] Specific embodiment 2: This embodiment is a further limitation of specific embodiment 1. The time-lapse real-time mixing test method includes the following steps, specifically:

[0052] Step 1: The prototype structure 1 includes a numerical substructure 2 and a test substructure 3. The number of restart steps for the mechanical properties of the test substructure 3 is determined based on the preliminary test.

[0053] Step 2: Establish numerical models of the numerical substructure 2 of the prototype structure 1 and the numerical model of the test substructure 3 respectively, and obtain the loading target at the initial moment through the numerical simulation system;

[0054] Step 3: First, perform a time lag compensation on the loading target at the initial moment. Then, perform multi-task loading on the single test substructure 3 through the servo loading system to obtain the loading target at the initial moment. Finally, reset the actuator to the initial moment and wait for the loading target at the next moment.

[0055] Step 4: Use the step-by-step integration algorithm to calculate the motion equation of the multi-degree-of-freedom system, solve the loading command in step i, and then feed back the new target to the multi-task part restart loading module through the predetermined communication conditions. The predetermined communication conditions are also predetermined communication mechanisms, which are communication conditions that are predetermined and adapted according to specific requirements. When based on the shared memory architecture of the OpenFresco middleware platform, the corresponding predetermined communication conditions are to build a high-speed data interaction channel between the OpenSEES finite element analysis software and the dSpace controller. After the predetermined communication conditions are determined, the new target is fed back to the multi-task part restart loading module through the predetermined communication conditions. The corresponding calculated motion equation is:

[0056] M[a i ]+C[v i ]+K[d i ]+[F Em,i ]=-Ma i Formula 1

[0057] In the above formula, M is the mass matrix of numerical substructure 2; C is the damping matrix of numerical substructure 2; K is the stiffness matrix of numerical substructure 2; [a i ] is the acceleration matrix of numerical substructure 2; [v i ] is the velocity matrix of numerical substructure 2; [d i ] is the displacement matrix of numerical substructure 2; [F Em ,i] is the restoring force matrix of the test substructure 3; a i is the earthquake acceleration excitation; the subscript i represents the integration time step;

[0058] Step 5: The multi-task part restarts the loading module to store the new target, integrates the new target with the initial displacement to form a global loading command, and generates a reset-wait command for the test substructure 3 through the global loading command;

[0059] Step 6: The global load command obtained in step 5 is converted into multiple subtask commands. The conversion process is as follows:

[0060] After determining the partial loading command for each subtask, multiple subtask commands are subjected to secondary time lag compensation to form multiple compensated partial loading commands. The test system sequentially executes the multiple compensated partial loading commands to complete the process of applying the displacement load to the test substructure 3. After the loading command is executed, the test substructure 3 executes the reset-wait command, and the acquisition system sends back the measured reaction force and displacement.

[0061] Step 7: Restart the module to check whether the measured response meets the required accuracy. If the measured response meets the accuracy requirement, proceed to the next step to execute the reset and wait command. If the measured response does not meet the accuracy requirement, repeat the loading until the accuracy is met.

[0062] Repeat steps 4 to 7 until all integration steps are loaded and the test is completed.

[0063] Specific embodiment three: This embodiment is a further limitation of specific embodiment one or two. The process of forming a preliminary loading target after the loading target at the initial calculation time is subjected to a time delay compensation and multi-task loading of the servo loading system includes a process of obtaining the total completion time in the time-lapse real-time hybrid test. The process of obtaining the total completion time in the time-lapse real-time hybrid test is as follows: the duration of the seismic wave is T, the time interval is Δt, the reset-waiting time of each integration step is a constant λ, the partial restart loading step is k, the number of test substructures 3 is n, and the calculation equation of the total completion time is Formula 2, specifically:

[0064]

[0065] Specific embodiment four: This embodiment is a further limitation of specific embodiments one, two or three. In this embodiment, the loading time corresponding to each test substructure 3 is kΔt, that is, the loading process experienced at the i-th step is from Δt to (1+k)Δt, where 0≤(1+k)Δt≤T; when the restart loading command of each layer from the first step to the i-th step is processed into a data segment with a length of kΔt in the manner of the i-th step, the (i+1)th step, and the (i+2)th step (i≥1), each data segment corresponds to a partition, and there are a total of (i-j+1) data partitions, that is, each data partition is a local restart loading command, where k=(i-j+1).

[0066] Specific embodiment five: This embodiment is a further limitation of specific embodiments one, two, three or four. The time-lapse real-time hybrid test method described in this embodiment is a hybrid test loading process in which a set of loading equipment is used to perform multi-task local restart loading on a test substructure 3, and the dynamic performance of all test substructures is verified with limited test facilities. In the i-th integration step, a set of actuators is first used to execute partial restart loading commands of task 1, task 2, ..., task n on a single test substructure, and then n measured force and displacement data are collected as measured data, and the measured data are fed back to the numerical system. At the same time, the test substructure 3 executes a reset-wait command.

[0067] Specific embodiment six: This embodiment is a further limitation of specific embodiments one, two, three, four or five. In this embodiment, the time-lapse real-time hybrid test method also includes a reset and synchronization waiting accuracy process. The reset and synchronization waiting accuracy process is to compare the numerical simulation system with the measured data of all previous steps of the test substructure 3 after each step of loading is completed, thereby completing the process of verifying the loading and reset taking into account the waiting accuracy of the test substructure 3.

[0068] Specific embodiment seven: This embodiment is a further limitation of specific embodiments one, two, three, four, five or six. After the loading target at the initial moment of calculation is subjected to one time lag compensation and multi-task loading of the servo loading system to form a preliminary loading target, when i=i+1 in the preliminary loading target, the numerical simulation system is used for step-by-step integration calculation to generate partial loading, reset and waiting commands, and then after secondary time lag compensation, the multi-task partial loading command is executed until the structural response meets the accuracy requirements and the reset and waiting command is executed.

[0069] During the specific loading process, the present invention can locally load the test substructure in a time increment manner according to the partition command, under the premise of ensuring the matching of the current mechanical state variables of the current test substructure, greatly reducing the number of test substructure restart loading steps, and improving the test efficiency in a quantitative sense. Specifically:

[0070] Combine Figure 2 As shown, taking a high-rise vibration reduction / vibration structure equipped with multiple isolation supports as an example, the basic principles and usage steps of the method of the present invention are explained. In this embodiment, when the prototype structure 1 is a high-rise vibration reduction structure, the high-rise structure is taken as the numerical substructure 2, and finite element software is used for detailed modeling; one of the isolation supports is taken as the test substructure 3, and a servo loading system is used to load it in sequence, and a real-time hybrid test based on multi-task local restart loading technology is carried out. The specific test process is as follows:

[0071] Step 1: Determine the number of partial restart steps k that can ensure the mechanical performance of the isolation bearing based on preliminary tests;

[0072] Step 2: Establish a finite element model of the seven-story frame structure. Assuming that the initial restoring force of the isolation bearings is 0, the finite element model calculates the loading targets of the six isolation bearings under external excitation and initial conditions.

[0073] Step 3: Compensate for time lag in the initial loading target. The servo loading system then applies multiple rounds of loading to each isolation bearing to achieve the initial loading target for all isolation bearings. Finally, the actuator resets to the initial time and waits for the next loading target.

[0074] Step 4: Use the step-by-step integration algorithm to calculate the motion equation of the multi-degree-of-freedom system, that is, Formula 1, solve the loading command in step i, and then feed the new target back to the multi-tasking part through certain communication conditions to restart the loading module;

[0075] M[a i ]+C[v i ]+K[d i ]+[F Em,i ]=-Ma i Formula 1

[0076] In the above formula, M, C, and K are the mass matrix, damping matrix, and stiffness matrix of the seven-story frame structure respectively; [a i ]、[v i ]、[d i ] are the acceleration matrix, velocity matrix and displacement matrix of the seven-layer frame structure; [F Em ,i] is the restoring force matrix of the isolation support; a i is the earthquake acceleration excitation; the subscript i represents the integration time step.

[0077] Step 5: Restart the multi-task part and load the module to store the new target, integrate the new target with the initial displacement to form a global load command [D i ] and generates a reset-wait command for the isolation bearing.

[0078] Step 6: Load the global command [D E,i ] is converted into multiple subtask commands, namely d E1,i d E2,i ,...,d E7,i , determine the partial loading command for each subtask, i.e. d E1,i_p d E2,i_p ,...,d E7,i_p Then, they are time-delay compensated to generate multiple compensated partial load commands, namely d Ec1,i_p d Ec2,i_p ,...,d Ec7,i_p Subsequently, the test system executes the loading commands in sequence to apply displacement loads to the isolation bearings. After the loading commands are executed, the isolation bearings execute the reset-wait command, and the acquisition system measures the reaction force [F Em,i ] and displacement [D Em,i ]Send back.

[0079] Step 7: Restart the module and check whether the measured response meets the required accuracy. If it does, proceed to the next step; if not, repeat the loading until the accuracy is achieved.

[0080] Repeat steps 4 to 7 until all integration steps are loaded and the test ends.

[0081] The present invention can also be applied to the time-lapse real-time hybrid test process of multi-task local restart loading of other large and complex vibration-damping structures, specifically:

[0082] Combine Figure 3 As shown, the basic principles and application steps of the present method are described using prototype structure 1 as a bridge vibration reduction / seismic structure. In this embodiment, when prototype structure 1 is a four-span bridge vibration reduction / seismic structure, the numerical substructure 2 is specifically a bridge structure, which is refined and modeled using finite element software; the test substructure 3 is a viscous damper, which is loaded sequentially using a servo loading system to conduct a real-time hybrid test based on multi-task local restart loading technology. The specific test process is as follows:

[0083] Step 1: Determine the number of partial restart steps k that can ensure the mechanical performance of the viscous damper based on preliminary tests;

[0084] Step 2: Establish a finite element model of the bridge structure. Assuming that the restoring forces of the viscous dampers are all zero at the initial moment, the finite element model calculates the loading targets of the six viscous dampers under external excitation and initial conditions.

[0085] Step 3: Compensate for time lag in the initial load target. The servo loading system then applies multiple rounds of loading to each viscous damper to achieve the initial load target for all viscous dampers. Finally, the actuator resets to the initial time and waits for the next load target.

[0086] Step 4: Use the step-by-step integration algorithm to calculate the motion equation of the multi-degree-of-freedom system, that is, Formula 1, solve the loading command in step i, and then feed the new target back to the multi-tasking part through certain communication conditions to restart the loading module;

[0087] M[a i ]+C[v i ]+K[d i ]+[F Em,i ]=-Ma i Formula 1

[0088] In the above formula, M, C, and K are the mass matrix, damping matrix, and stiffness matrix of the bridge structure respectively; [a i ]、[v i ]、[d i ] are the acceleration matrix, velocity matrix and displacement matrix of the bridge structure; [F Em ,i] is the restoring force matrix of the viscous damper; a i is the earthquake acceleration excitation; the subscript i represents the integration time step;

[0089] Step 5: Restart the multi-task part and load the module to store the new target, integrate the new target with the initial displacement to form a global load command [D i ] and generates a reset-wait command for the viscous damper.

[0090] Step 6: Load the global command [D E,i ] is converted into multiple subtask commands, namely d E1,i d E2,i ,...,d E7,i , determine the partial loading command for each subtask, i.e. d E1,i_p d E2,i_p ,...,d E7,i_p Then, they are time-delay compensated to generate multiple compensated partial load commands, namely d Ec1,i_p d Ec2,i_p ,...,d Ec7,i_p Subsequently, the test system executes the loading command in sequence to apply displacement load to the viscous damper. After the loading command is executed, the viscous damper executes the reset-wait command, and the acquisition system measures the reaction force [F Em,i ] and displacement [D Em,i ]Send back;

[0091] Step 7: Restart the module and check whether the measured response meets the required accuracy. If it does, proceed to the next step; if not, repeat the loading until the accuracy is achieved.

[0092] Repeat steps 4 to 7 until all integration steps are loaded and the test ends.

[0093] When the present invention is applied to other large and complex vibration reduction structures, Figure 4 As shown in the figure, when the prototype structure 1 is a frame vibration / shock structure equipped with three viscous dampers, the corresponding numerical substructure 2 is a frame structure, which is finely modeled using finite element software; the test substructure 3 is a viscous damper on the first floor of the structure, which is loaded sequentially using a servo loading system to carry out a force correction update iterative hybrid test based on multi-task loading. The specific test process is as follows:

[0094] Step 1: Determine the number of partial restart steps k that can ensure the mechanical performance of the viscous damper based on preliminary tests;

[0095] Step 2: Establish a finite element model of the three-layer frame structure. Assuming that the restoring forces of the viscous dampers are all zero at the initial moment, the finite element model calculates the loading targets of the three viscous dampers under external excitation and initial conditions.

[0096] Step 3: Compensate for time lag in the initial load target. The servo loading system then applies multiple rounds of loading to each viscous damper to achieve the initial load target for all viscous dampers. Finally, the actuator resets to the initial time and waits for the next load target.

[0097] Step 4: Use the step-by-step integration algorithm to calculate the motion equation of the multi-degree-of-freedom system, that is, Formula 1, solve the loading command in step i, and then feed the new target back to the multi-tasking part through certain communication conditions to restart the loading module;

[0098] M[a i ]+C[v i ]+K[d i ]+[F Em,i ]=-Ma i Formula 1

[0099] In the above formula, M, C, and K are the mass matrix, damping matrix, and stiffness matrix of the three-layer frame structure respectively; [a i ]、[v i ]、[d i ] are the acceleration matrix, velocity matrix and displacement matrix of the three-layer frame structure; [F Em ,i] is the restoring force matrix of the viscous damper; a i is the earthquake acceleration excitation; the subscript i represents the integration time step.

[0100] Step 5: Restart the multi-task part and load the module to store the new target, integrate the new target with the initial displacement to form a global load command [D i ] and generates a reset-wait command for the viscous damper.

[0101] Step 6: Load the global command [D E,i ] is converted into multiple subtask commands, namely d E1,i d E2,i ,...,d E7,i , determine the partial loading command for each subtask, i.e. d E1,i_p d E2,i_p ,...,d E7,i_p Then, they are time-delay compensated to generate multiple compensated partial load commands, namely d Ec1,i_p d Ec2,i_p ,...,d Ec7,i_p Subsequently, the test system executes the loading command in sequence to apply displacement load to the viscous damper. After the loading command is executed, the viscous damper executes the reset-wait command, and the acquisition system measures the reaction force [F Em,i ] and displacement [D Em,i ]Send back.

[0102] Step 7: Restart the module and check whether the measured response meets the required accuracy. If it does, proceed to the next step; if not, repeat the loading until the accuracy is achieved.

[0103] Repeat steps 4 to 7 until all integration steps are loaded and the test ends.

[0104] The present invention can be adapted for use in high-rise vibration / shock reduction structures and large and complex vibration reduction structures, and can achieve a precise loading process that avoids calculation time delays.

Claims

1. A time-lapse real-time hybrid test method based on multi-task partial restart loading, characterized by: The time-lapse real-time hybrid test method determines the number of local restart loading steps based on the mechanical properties of the test substructure. The loading target at the initial moment of calculation is subjected to a time delay compensation and multi-task loading of the servo loading system to form a preliminary loading target. The preliminary loading target is then increased with the increase of the integral time step and according to the loading commands of each partition to complete the local multi-round restart loading process of each test substructure one by one.

2. The time-lapse real-time hybrid test method based on multi-task partial restart loading according to claim 1 is characterized in that: The time-lapse real-time mixing test method includes the following steps, specifically: Step 1: The prototype structure (1) includes a numerical substructure (2) and a test substructure (3). The number of restart steps for the mechanical properties of the test substructure (3) is determined based on the preliminary test. Step 2: Establish the numerical model of the numerical substructure (2) and the numerical model of the test substructure (3) of the prototype structure (1) respectively, and obtain the loading target at the initial moment through the numerical simulation system; Step 3: First, perform a time lag compensation on the loading target at the initial moment, then obtain the loading target at the initial moment after multi-tasking loading on a single test substructure (3) through the servo loading system, and finally reset the actuator to the initial moment and wait for the loading target at the next moment; Step 4: Use the step-by-step integration algorithm to calculate the motion equation of the multi-degree-of-freedom system, solve the loading command in step i, and then feed the new target back to the multi-tasking part through the pre-established real-time communication mechanism to restart the loading module. The motion equation is: M[a i ]+C[v i ]+K[d i ]+[F Em,i ]=-Ma i Formula 1 In the above formula, M is the mass matrix of the numerical substructure (2); C is the damping matrix of the numerical substructure (2); K is the stiffness matrix of the numerical substructure (2); [a i ] is the acceleration matrix of the numerical substructure (2); [v i ] is the velocity matrix of the numerical substructure (2); [d i ] is the displacement matrix of the numerical substructure (2); [F Em ,i] is the restoring force matrix of the test substructure (3); a i is the earthquake acceleration excitation; the subscript i represents the integration time step; Step 5: The multi-task part restarts the loading module to store the new target, integrates the new target with the initial displacement to form a global loading command, and generates a reset-wait command for the test substructure (3) through the global loading command; Step 6: The global load command obtained in step 5 is converted into multiple subtask commands. The conversion process is as follows: After determining the partial loading command of each subtask, a plurality of subtask commands are subjected to secondary time lag compensation to form a plurality of compensated partial loading commands, and the test system is used to sequentially execute the plurality of compensated partial loading commands to complete the process of applying displacement load to the test substructure (3). After the loading command is executed, the test substructure (3) executes a reset-wait command, and the acquisition system sends back the measured reaction force and displacement; Step 7: Restart the module to check whether the measured response meets the required accuracy. If the measured response meets the accuracy requirement, proceed to the next step to execute the reset and wait command. If the measured response does not meet the accuracy requirement, repeat the loading until the accuracy is met. Repeat steps 4 to 7 until all integration steps are loaded and the test ends.

3. The time-lapse real-time hybrid test method based on multi-task partial restart loading according to claim 1 or 2, characterized in that: The process of forming a preliminary loading target by sequentially subjecting the loading target at the initial calculation time to time delay compensation and multi-task loading by the servo loading system includes a process of obtaining the total completion time in the time-lapse real-time hybrid test. The process of obtaining the total completion time in the time-lapse real-time hybrid test is as follows: The duration of the seismic wave is T, the time interval is Δt, and the total reset-wait time of each integration step is a constant λ total , the partial restart loading step is k, the number of test substructures (3) is n, and the equation corresponding to the total completion time is Formula 2:

4. The time-lapse real-time hybrid test method based on multi-task partial restart loading according to claim 3 is characterized in that: The loading time corresponding to each experimental substructure (3) is kΔt, that is, the loading process experienced at the i-th step is from Δt to (1+k)Δt, where 0≤(1+k)Δt≤T; when the restart loading command of each layer from the first step to the i-th step is processed into a data segment of length kΔt in the manner of the i-th step, the (i+1)-th step, and the (i+2)-th step (i≥1), each data segment corresponds to a partition, and there are a total of (i-j+1) data partitions, that is, each data partition is a local restart loading command, where k=(i-j+1).

5. The time-lapse real-time hybrid test method based on multi-task partial restart loading according to claim 4 is characterized in that: The time-lapse real-time hybrid test method is a hybrid test loading process in which a set of loading equipment is used to perform multi-task local restart loading on a test substructure (3), and the dynamic performance of all test substructures is verified with limited test facilities. In the i-th integration step, a set of actuators is first used to execute partial restart loading commands of task 1, task 2, ..., task n on a single test substructure, and then n measured force and displacement data are collected as measured data, and the measured data are fed back to the numerical system. At the same time, the test substructure (3) executes a reset-wait command.

6. The time-lapse real-time hybrid test method based on multi-task partial restart loading according to claim 5 is characterized in that: The time-lapse real-time hybrid test method also includes a reset and synchronization waiting accuracy process. The reset and synchronization waiting accuracy process is to compare the numerical simulation system with the measured data of all previous steps of the test substructure (3) after each step of loading is completed, thereby completing the process of verifying the loading and reset of the test substructure (3) and the waiting accuracy.

7. The time-lapse real-time hybrid test method based on multi-task partial restart loading according to claim 1 is characterized in that: After the loading target at the initial moment of calculation is subjected to a time delay compensation and multi-task loading of the servo loading system to form a preliminary loading target, when i=i+1 in the preliminary loading target, the numerical simulation system is gradually integrated and calculated to generate partial loading, reset and wait commands. After a second time delay compensation, the multi-task partial loading command is executed until the structural response meets the accuracy requirements and the reset and wait commands are executed.

Images