Patch scheme optimizing calculation method

Abstract

The invention provides a patch scheme optimization calculation method, which comprises the following steps of: S1, determining a patch component, and obtaining a modal monitoring order of the patch component, an available position set of each order of patches and a union set of the available position sets of each order of patches; s2, determining modal frequency bands of all orders, the number of patches and the sizes of strain gauges; s3, according to the patch number and the union set, the number of parallel computing cores is determined, and each core independently uses a recursive algorithm to determine an available patch position combination; and S4, selecting a group of patch position combination with the maximum monitoring order and the maximum influence coefficient from the available patch position combinations as a patch scheme. According to the method, the optimal patch position scheme can be obtained, the maximum modal order is monitored by using the minimum patch number, and meanwhile, the patch position precision can be improved.

Description

Technical Field

[0001] The present invention relates to the field of engines, and in particular to the field of calculation methods for patch position selection. Background Art

[0002] Aero-engine is a highly complex and sophisticated thermal machine used to power aircraft, including turbojet / turbofan engine, turboshaft / turboprop engine, etc. The fan blades of existing engines are often made of composite materials.

[0003] During the strength test and test run of composite blades, it is necessary to paste a patch on the test object, that is, to paste a strain gauge, so as to obtain the strain change at the corresponding position in the test. On the one hand, it can ensure the safety of the test, and on the other hand, it can obtain the stress and strain at the corresponding position of the composite blade in different strength tests such as vibration fatigue, rotation and bird strike, as well as the whole machine test run, so as to obtain the dynamic strain data of the blade, provide test data for model optimization and improvement, and also provide test accumulation for the final airworthiness certification. Therefore, it is necessary to determine the pasting position of the strain gauge on the blade.

[0004] The traditional method is to manually determine several points in the available patch set as the patch position, and give the distance from the leading and trailing edges of the blade, and then use a caliper to determine the distance of the point on the actual blade, so as to perform patching. The problems with this are: it is impossible to obtain the optimal patch position, it is impossible to obtain more accurate measurement results, it is impossible to monitor the most modal orders with the least number of patches, and it is impossible to paste strain gauges at the precise position of the actual blade. Summary of the invention

[0005] An object of the present invention is to provide a patch solution optimization calculation method that can obtain an optimal patch location solution.

[0006] To achieve the above-mentioned destination patch solution optimization calculation method, the following steps are included: S1. Determine the patch component, obtain the modal monitoring order of the patch component, the available position set of each order patch and the union of the available position sets of each order patch; S2. Determine the modal frequency band of each order, the number of patches and the size of strain gauges; S3. Determine the number of parallel computing cores based on the number of patches and the union, and each core independently uses a recursive algorithm to determine the available patch position combination; S4. Select a group of patch position combinations with the most monitorable orders and the largest influence coefficient from the available patch position combinations as the patch solution.

[0007] In one or more embodiments, S1 also includes the following steps: determining the engine state for dynamic stress analysis; determining the modal monitoring order according to the Campbell diagram; performing aerodynamic analysis, thermal analysis, static analysis and modal analysis with prestress on the patch component under the engine state; obtaining the dangerous points under each mode according to the Goodman curve of the material of the patch component, and taking the set of points with an influence coefficient greater than 0.3 as the patch available position set for the mode of that order; and performing a union operation on the patch available position sets of each mode to form a union.

[0008] In one or more embodiments, the finite element model used for modal analysis is consistent with the finite element model used for static analysis.

[0009] In one or more embodiments, each modal frequency band in step S2 is obtained through static analysis and modal analysis with prestress.

[0010] In one or more embodiments, 0.9 times the minimum value and 1.1 times the maximum value of each mode are taken as the frequency band of the mode.

[0011] In one or more embodiments, in step S2, the number of patches is in a range of greater than or equal to 3 and less than or equal to 6.

[0012] In one or more embodiments, in step S2, for the composite material component, the strain gauge size is greater than or equal to 1.5 times the unit cell size.

[0013] In one or more embodiments, in step S2, for the metal patch component, the minimum size of the strain gauge is greater than or equal to 3 mm.

[0014] In one or more embodiments, in step S3, the formula Determine the number of cores, where crad(Ω) is the number of elements in the union set Ω, n gauge is the number of patches, and the number of cores is the maximum integer that satisfies this formula.

[0015] In one or more embodiments, in step S3, the recursive algorithm of each core includes the following steps: performing a combination operation Where crad(Ω) is the number of elements in the union set Ω, n gauge is the number of patches, n is selected in each operation gauge The nodes satisfy the following rules: each point is outside a sphere with the rest of the points as the center and the diagonal length of the strain gauge as the radius, and the mode monitored by each point has no frequency band intersection.

[0016] In one or more embodiments, in step S3, a set of computing nodes that conform to the rule is recorded as the available patch position combination, and the total number of modal orders monitored by each set of computing nodes and the sum of the influence coefficients corresponding to each order of monitored modal are calculated.

[0017] In one or more embodiments, in step S4, a master node is used to communicate with each slave node through a message passing interface, each slave node corresponds to each core, and the computing node sets formed by each slave node are collected; the master node is used to sort the computing node sets, and the sorting uses the following rules: first sort the computing node sets according to the number of detection modes, and select the computing node set with the largest number of detection modes as the patch solution; when multiple computing node sets have the maximum number of detection modes, take the group with the largest sum of influence coefficients in this part of the computing node sets as the patch solution.

[0018] In one or more embodiments, in step S4, the three-dimensional coordinates of each point in the patch solution are determined in finite element software; and the three-dimensional coordinates are marked on the patch component using a three-dimensional coordinate measuring machine.

[0019] The above patch scheme optimization calculation method can obtain the patch position that can monitor the most modal orders through calculation and algorithm optimization, and can ensure that the patch position is in the area with the largest influence coefficient of the patch component, so that the strain gauge in this area can capture more precise measurement results. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The above and other features, properties and advantages of the present invention will become more apparent through the following description in conjunction with the accompanying drawings and embodiments, in which:

[0021] Figure 1 It is a flow chart of the calculation method for optimizing the patch solution;

[0022] Figure 2A-2B It is a schematic diagram of the traditional patch location;

[0023] Figure 3 It is a schematic diagram of the positions of available patch positions corresponding to a certain mode;

[0024] Figure 4 This is a schematic diagram of accurate patch position determination based on three coordinates;

[0025] Figure 5A-5B It is the curve diagram of the measured dynamic strain results;

[0026] Figure 6 It is a flowchart of a specific embodiment of the patch solution optimization calculation method. DETAILED DESCRIPTION

[0027] The present invention is further described below in conjunction with specific embodiments and drawings. More details are elaborated in the following description to facilitate a full understanding of the present invention. However, the present invention can obviously be implemented in a variety of other ways different from the description herein. Those skilled in the art can make similar generalizations and deductions based on actual application situations without violating the connotation of the present invention. Therefore, the protection scope of the present invention should not be limited by the content of this specific embodiment.

[0028] It should be noted that these and other subsequent drawings are only examples and are not drawn to scale, and should not be used to limit the actual scope of protection required by the present invention.

[0029] When an object vibrates at a certain order of natural frequency, the displacement of each point on the object from the equilibrium position satisfies a certain pattern, which is called the mode corresponding to the natural frequency. The first-order mode is the lowest frequency at which the object resonates, and the second-order mode occurs when the excitation frequency of the external force is equal to twice the lowest resonant frequency of the object. At this time, it is a second-order vibration mode. The vibration response of an object is a superposition of vibration modes of various orders. Each mode has a specific natural frequency, damping ratio and modal vibration mode. It is understandable that the blades of an aircraft engine will produce vibrations of multiple modal orders when they are affected by external factors such as external force and temperature during operation.

[0030] Therefore, it is necessary to paste strain gauges during the composite blade strength test and test run to obtain the strain changes at the corresponding positions in the test. Strain gauges refer to components composed of sensitive grids for measuring strain. They are attached to the surface of patch components such as blades to obtain dynamic strain data under various modes of patch components, providing test data for model optimization and improvement.

[0031] Traditional patch position determination is as follows Figure 2A-2B As shown, taking the blade as an example, first determine the blade height distance U and V from the bottom of the tenon, then determine the distance W and X from the leading edge L and the trailing edge T, determine the patch position point, and then determine the patch direction through the angle θ. In this method, the dimensions U, V, W and X are difficult to measure accurately, resulting in the patch position being inconsistent with the design state, making the test measurement results inaccurate. Therefore, when determining the patch position by the traditional method, several positions are manually determined from the optional nodes, which cannot achieve the maximum number of monitored modes and cannot guarantee the maximum influence coefficient of each monitored mode.

[0032] The patch scheme optimization calculation method disclosed in the present invention can quickly optimize the patch scheme through calculation, and can also monitor with the least number of patches, while improving the patch position accuracy on the actual object.

[0033] Reference Figure 1 and Figure 6It is understood that the method includes the following steps: S1. Determine the patch component, obtain the modal monitoring order of the patch component, the available position set of each order patch and the union of the available position set of each order patch, and the patch component selects the component area such as blades and casings that need to be tested; S2. Determine the modal frequency band of each order, the number of patches and the size of strain gauges; S3. Determine the number of parallel computing cores based on the number of patches and the union, and each core independently uses a recursive algorithm to determine the available patch position combination; S4. Select a group of patch position combinations with the most monitorable orders and the largest influence coefficients from the available patch position combinations as the patch solution.

[0034] Specifically, in step S1, obtaining the modal monitoring order of the patch component, the available position sets of patches of each order, and the union of the available position sets of patches of each order is completed using the following sub-steps.

[0035] (1) First, determine the engine state for dynamic stress analysis. Usually, the static state and the highest speed state during engine test or in the flight envelope are taken as the starting and ending states of the analysis, and several typical state points are selected between the two as intermediate states. A typical state selection is: static state, ground slow car, economic cruise, and three red lines.

[0036] (2) Determine the modal monitoring order. Usually, according to the Campbell diagram, all modal frequency orders that intersect with the known excitation source in the range from static state to the given maximum state of assessment need to be monitored. The Campbell diagram is the vibration amplitude of the monitoring point as a function of speed and frequency, so as to show the change characteristics of all components of the rotor vibration in the entire speed range, with the horizontal axis representing the speed and the vertical axis representing the frequency.

[0037] It should be noted that in engine testing, generally higher-order intersection points correspond to lower speeds and less energy and are no longer monitored, but at least the first 6 knots of frequencies should be considered.

[0038] (3) For the engine analysis state point determined in step (1), aerodynamic analysis, thermal analysis, static analysis and prestressed modal analysis are performed on the patch component under the engine state.

[0039] Aerodynamic CFD analysis and thermal analysis are carried out to obtain the aerodynamic loads and temperature loads on the blades.

[0040] For the analysis state point determined in step (1), static analysis and modal analysis with prestress are carried out. Static analysis obtains the stress and strain distribution of the model, and modal analysis obtains the frequency of the blade in static state and the dynamic frequency considering the rotational stiffening effect.

[0041] When performing modal analysis with prestress, the centrifugal force, aerodynamic load and temperature load on the blade must be considered. This way, the changes in blade frequency caused by changes in speed and temperature can be considered. The finite element model for modal analysis is consistent with the finite element model for static analysis.

[0042] The mesh size of the finite element analysis should be determined according to the size of the strain gauge. When the accuracy of the constructed finite element model is a second-order unit, the strain gauge should cover 2 mesh units; when the accuracy of the constructed finite element model is a first-order unit, the strain gauge should cover 3 mesh units.

[0043] (4) According to the Goodman curve of the material of the patch component, the dangerous points under each mode are obtained, and the set of points with an influence coefficient greater than 0.3 is used as the patch available position set of the mode.

[0044] In the Goodman curve, the horizontal axis is static stress and the vertical axis is fatigue stress. The calculation result of step (3) is used to obtain the dangerous point under each mode. For each order of modal stress, all nodes are traversed to determine the node set whose modal strain is greater than 0.3 times the dangerous point stress, which is the available patch set of the mode. The available patch sets of each order are combined to form the union Ω of the patch position set.

[0045] For example, assuming that the monitored modes are 1 to 6 orders, the modal stress at the dangerous point is σ cric_i , i = 1, 2, 3, 4, 5, 6, then the dangerous point set Ω corresponding to each order mode i ={n j}, where σ j For node n j The corresponding modal stress is, is the influence coefficient, which is the ratio of the modal stress to the modal stress at the dangerous point. It characterizes the sensitivity of the material and can be adjusted according to the test requirements. It is generally taken as greater than 0.3.

[0046] The patch position sets of each order are combined to form a union. If the monitoring mode is 1 to 6 orders, the calculation method of the union is Ω = Ω 1 ∪Ω 2 ∪Ω 3 ∪Ω 4 ∪Ω 5 ∪Ω 6 .

[0047] In step S2, according to step S1 (3), the frequency band is obtained through static analysis and modal analysis with prestress. Preferably, 0.9 times the minimum value and 1.1 times the maximum value of each mode are taken as the frequency band of the mode, that is, In this way, the effects of rotational softening and stress stiffening on the frequency can be considered.

[0048] Determine the number of patches n gauge It is at least greater than 3, and there is no upper limit. Generally, it is within 6, and it can be adjusted according to the calculation results of the actual patch positions.

[0049] When determining the strain gauge size, for metal components, such as the laminated blade body and the metal reinforcement edge positions, small-sized strain gauges are generally 3 mm; for woven blades, it should be determined according to the unit cell size. The strain gauge size cannot be less than 1.5 times the unit cell size. The unit cell refers to the size of the unit formed by multiple fiber bundles in the composite material.

[0050] In step S3, according to the set of available patch positions formed in step S1 and the number of patches determined in step S2, determine the number of parallel computing cores. Specifically, use the formula to determine the number of cores, where crad(Ω) is the number of elements in the union set Ω, and n gauge is the number of patches, and the number of cores is the largest integer that satisfies this formula.

[0051] Each core uses a recursive algorithm to determine the available combinations of patch positions, including the following sub-steps: perform combination operations where crad(Ω) is the number of elements in the union set Ω, and n gauge is the number of patches.

[0052] The n gauge nodes selected in each operation satisfy the following rules: each point is outside the sphere with the other points as the center and the diagonal length of the strain gauge as the radius, and the modes monitored by each node have no frequency band crossing.

[0053] The above rules assume that each node is the center of the sphere, and the influence range is a spherical region with the node as the center and the diagonal length of the strain gauge as the radius. Each point being outside the sphere with the other points as the center and the diagonal length of the strain gauge as the radius means that there is no overlapping area in the detection areas of each node (patch position). Taking each node as the patch position, the modes monitored by the strain gauge have no frequency band crossing.

[0054] Record the set of computing nodes Ψ k , k = 1, 2, 3... n gauge as the available combinations of patch positions, calculate the total number of modal orders that each set of computing nodes can monitor, and the sum of the influence coefficients corresponding to each monitored mode.

[0055] Since each core performs parallel and independent operations and there is no data exchange between cores, in step S4, the master node uses MPI (Message Passing Interface) to communicate with each slave node and collect the data formed by each slave node, which contains n gaugeThe node set Ψ k . And use the main node for sorting, and select the computing node set with the largest number of detection modes as the patch solution; when multiple computing node sets have the largest number of detection modes, that is, when the number of monitoring modes is the same, take the set with the largest sum of influence coefficients in this part of the computing node set as the patch solution.

[0056] The above determination basis means that the patch solution can monitor the most strain gauge orders and is located in the most sensitive area on the patch component.

[0057] Finally, determine the coordinates of the patch position in the xyz directions.

[0058] For example, first determine the precise three-dimensional coordinates of each point in the patch solution in the finite element software, and then k The midpoint forms the patch direction G;

[0059] The actual position of the reference point of each point is determined by a three-dimensional coordinate measuring machine, and marked to obtain the patch position P, and the strain gauge is pasted at the corresponding position.

[0060] The use of a three-coordinate measuring machine can accurately determine the patch position and improve the patch accuracy.

[0061] The above parallel computing method can propose the optimal patch solution and position, monitor with the least number of patches, and improve the patch position accuracy on the actual object. Figure 5A-5B The measured dynamic strain results are shown, and it can be seen that it has better monitoring capabilities.

[0062] Although the present invention is disclosed as above with preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art may make possible changes and modifications without departing from the spirit and scope of the present invention. Therefore, any modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention shall fall within the protection scope defined by the claims of the present invention.

Claims

1. The patch solution optimization calculation method includes the following steps: S1. Determine a patch component, obtain the modal monitoring order of the patch component, the available position set of each order patch, and the union of the available position set of each order patch; S2. Determine the frequency band of each modal order, number of patches and size of strain gauges; S3. Determine the number of parallel computing cores according to the number of tiles and the union, and each core independently uses a recursive algorithm to determine the available tile position combination; S4. Selecting a group of patch position combinations with the largest number of monitorable orders and the largest influence coefficients from the available patch position combinations as a patch solution.

2. The method according to claim 1, It is characterized in that Step S1 also includes the following steps: Determine engine conditions for dynamic stress analysis; Determining the modal monitoring order according to the Campbell diagram; Under the engine state, aerodynamic analysis, thermal analysis, static analysis and modal analysis with prestress are performed on the patch component; According to the Goodman curve of the material of the patch component, the dangerous points under each mode are obtained, and the set of points with an influence coefficient greater than 0.3 is used as the patch available position set of the mode; The patch available position sets of each order mode are combined to form a union.

3. The method according to claim 2, It is characterized in that The finite element model for modal analysis is the same as the finite element model for static analysis.

4. The method according to claim 2, It is characterized in that Through static analysis and modal analysis with prestress, the modal frequency bands of various orders in step S2 are obtained.

5. The method according to claim 4, It is characterized in that 0.9 times the minimum value and 1.1 times the maximum value of each mode are taken as the frequency band of the mode.

6. The method according to claim 1, It is characterized in that In step S2, the number of patches is in a range of greater than or equal to 3 and less than or equal to 6.

7. The method according to claim 1, It is characterized in that In step S2, for a composite material component, the strain gauge size is greater than or equal to 1.5 times the unit cell size.

8. The method according to claim 1, It is characterized in that In step S2, for the metal patch component, the minimum size of the strain gauge is greater than or equal to 3 mm.

9. The method according to claim 1, It is characterized in that In step S3, use the formula Determine the number of cores, where crad(Ω) is the number of elements in the union set Ω, n gauge is the number of patches, and the number of cores is the maximum integer that satisfies this formula.

10. The method according to claim 1, It is characterized in that In step S3, the recursive algorithm of each core includes the following steps: Perform combinatorial operations Where crad(Ω) is the number of elements in the union set Ω, n gauge is the number of patches, n is selected in each operation gauge The nodes satisfy the following rules: Each point is outside a sphere with the rest of the points as the center and the diagonal length of the strain gauge as the radius, and the mode monitored at each point has no frequency band intersection.

11. The method according to claim 10, It is characterized in that In step S3, the computing node sets that meet the rules are recorded as the available patch position combinations, and the total number of modal orders monitored by each computing node set and the sum of the influence coefficients corresponding to each order of monitored modal are calculated.

12. The method according to claim 10, It is characterized in that In step S4, the master node is used to communicate with each slave node through a message passing interface, each slave node corresponds to each core, and the computing node sets formed by each slave node are collected; Use the master node to sort the set of computing nodes, and the sorting uses the following rules: First, the computing node set is sorted according to the number of detection modes, and the computing node set with the largest number of detection modes is selected as the patch solution; When multiple computing node sets all have the maximum number of detection modes, the set with the largest sum of influence coefficients in the computing node sets is taken as the patch solution.

13. The method of claim 1, It is characterized in that In step S4, Determine the three-dimensional coordinates of each point in the patch scheme in finite element software; The three-dimensional coordinates are marked on the patch component using a three-dimensional coordinate measuring machine.

Images