CN112182737A - Parallelization high-precision flutter calculation method based on modal method - Google Patents
Parallelization high-precision flutter calculation method based on modal method Download PDFInfo
- Publication number
- CN112182737A CN112182737A CN202010872130.XA CN202010872130A CN112182737A CN 112182737 A CN112182737 A CN 112182737A CN 202010872130 A CN202010872130 A CN 202010872130A CN 112182737 A CN112182737 A CN 112182737A
- Authority
- CN
- China
- Prior art keywords
- node
- generalized
- grid
- displacement
- modal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/15—Vehicle, aircraft or watercraft design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Abstract
The invention discloses a parallelization high-precision flutter calculation method based on a modal method, and belongs to the field of aeroelasticity. Firstly, flow field meshing software carries out meshing on a flow field domain of the aircraft, fluent software outputs airfoil surface mesh node coordinates, mesh center point coordinates, surface unit global numbers and accumulated node numbers to a txt file, and structure finite element software outputs structure finite element node coordinates and modal displacement of each order. And then interpolating the modal displacement value of each order to each grid node on the surface of the aircraft, reading the modal displacement of each order and the txt file, calculating the generalized aerodynamic force of the whole wing, solving to obtain the modal generalized displacement and the generalized velocity of each order, outputting the modal displacement and the generalized velocity to the txt file, and calculating the real velocity and displacement of the center point of each grid on the surface of the aircraft. The true displacement of the structure at each time step to the termination time step is calculated. And monitoring a structural real displacement-time curve of a given position to find out the high-precision flutter speed. The invention improves the calculation speed and saves the calculation cost.
Description
Technical Field
The invention belongs to the field of aeroelasticity, and relates to a parallelization high-precision flutter calculation method based on a modal method.
Background
The flutter problem is a classic problem in aeroelasticity mechanics and is also a problem of general concern to aircraft designers: because it seriously affects flight safety, most aircraft undergo flutter checking after aerodynamic and structural design is completed.
The flutter calculation method commonly used in engineering is a frequency domain flutter calculation method based on the aerodynamic force of a surface element method, and the method has universal applicability to the classical flutter problem of the aircraft with linear structure and aerodynamic force; however, in the case of nonlinear flow phenomena such as low reynolds number flow, strong separation flow, flow containing chemical reaction and phase change, and transonic flow, which are often observed in actual engineering, the prediction of the engineering flutter calculation method is not accurate, and it is necessary to perform flutter calculation based on a high-precision Computational Fluid Dynamics (CFD) method. It is therefore desirable to develop an efficient flutter calculation method that can take into account nonlinear aerodynamic forces.
A common solution idea is a CFD/CSD (computational solid dynamics) coupling strategy, and force, displacement and other information exchange is carried out at a fluid-solid interface.
The other solution is to use an equivalent reduced-order model to identify the nonlinear aerodynamic force under any working condition, and then use the identification model to predict the aerodynamic force, so as to solve the flutter problem; this approach requires a large number of CFD data samples to train the reduced order model, at a non-trivial cost in time. Meanwhile, the normalization capability of the reduced-order model under any flight working condition needs to be tested. In general, a reduced-order model can only have a good effect on nonlinear aerodynamic force prediction under certain working conditions.
Disclosure of Invention
The invention provides a parallelization high-precision flutter calculation method based on a modal method, which aims to solve the problem of engineering flutter calculation which can consider aerodynamic nonlinearity and simultaneously give consideration to calculation precision and calculation efficiency.
The flutter calculation method comprises the following specific steps:
step one, using flow field meshing software to perform meshing on a flow field domain of the aircraft, and outputting airfoil surface mesh node coordinates, mesh center point coordinates, surface unit global numbers and accumulated node numbers to a txt file in fluent software.
Each face unit includes at least three face mesh nodes and a mesh center point.
Step two, outputting node coordinates of the finite element of the structure and modal displacement of each order in the finite element software of the structure;
and thirdly, running fluent software in a parallel mode, and interpolating the read modal displacement values of all orders to grid nodes on the surface of the aircraft by using a thin plate spline interpolation (TPS) method through a geometric relation to obtain the modal displacement of all orders corresponding to the grid nodes of the flow field.
And step four, reading modal displacement and txt files of each order respectively corresponding to the central point of each grid by a host node, and calculating the sum of generalized aerodynamic force of each order by using each branch thread node to obtain the generalized aerodynamic force of the whole wing.
The specific process is as follows:
step 401, for each surface unit performed in a parallel manner, a host node reads modal displacement of each order corresponding to a central point of each grid and transmits the modal displacement to each branching process node;
each branch thread node obtains modal displacement of each order of all grid central points contained by the node by adopting a global numbering matching search algorithm, and the specific process is as follows:
first, all the surface units of the aircraft surface are integratedNumber vector faceGlobalId, by serial local number kserialArranging;
then, by using a function F _ ID for converting the local number into the global number in the udf function library, circularly traversing the serial local numbers of all the face units in each thread node, and returning the corresponding global number F _ ID (k)parallel,tf)；
Next, the global number F _ ID (k) in each face unit is assignedparallelTf) comparing and searching with all global numbers stored in the vector faceGlobalId, and recording the serial local number k of the position of the serial local number in the vector faceGlobalId when the same global number is foundserialAnd recording the parallel local number k of the current unit in the nodeparallelForm a set of dictionary keys and values { k }parallel:kserial}。
Finally, in the calculation of the generalized aerodynamic force, when the k-th position is calculatedparallelThe aerodynamic force of the surface unit is used for numbering k parallel partsparallelSubstitution { k }parallel:kserialReturns the corresponding serial local number kserialObtaining the ith order modal vector phi of the center point of the airfoil fluid grid(c),i (Nc×1)K of (1)serialIndividual modal value
And obtaining modal displacement of each order of the central point of each grid in the same way.
Step 402, aiming at each branching process node, the node calls each order modal value of all grid central points contained in the node, and respectively multiplies the modal values by the aerodynamic force of a surface unit where each grid central point is located to obtain each order generalized aerodynamic force of the grid unit, and sums to obtain the sum of each order generalized aerodynamic force of the node;
when the aerodynamic force of the surface unit where the central point of each grid is located is solved by using the density base, the specific process is as follows:
firstly, each surface unit of the surface of the aircraft corresponds to one body unit, and the grid center coordinate of each surface unit, the grid center coordinate of the body unit and the pressure intensity are obtained.
Then, for each individual unit, a pressure gradient from the center of the individual unit to the center of each surface unit and a coordinate vector from the center of the individual unit to the center of each surface unit are obtained.
And finally, calculating the pressure value of the center of each surface of the aircraft surface, and multiplying the pressure value by the area of the surface unit to obtain the aerodynamic force of the surface unit.
Step 403, the branch node 0 sums the aerodynamic force sums of all the branch route nodes performed in the parallel manner again to obtain the generalized aerodynamic force sum of each order.
And step 404, transmitting the sum of the generalized aerodynamic forces of each order to a host node through a branch node 0 to obtain the generalized aerodynamic force of the whole aircraft surface.
Step five, solving a generalized aeroelastic motion equation set compiled in a user-defined function in a host node by utilizing the generalized aerodynamic force of the surface of the whole aircraft to obtain the generalized displacement and the generalized speed of each order of mode corresponding to the central point of each grid, outputting the generalized displacement and the generalized speed of each order of mode at the current time step to a txt file by the host node, and transmitting the generalized displacement and the generalized speed to each branch process node;
step six, each branching thread node executes a user self-defined function: calculating the real speed and displacement of the central point of each grid on the surface of the aircraft by using the generalized displacement and the generalized speed of each order of mode at the current time step;
and seventhly, controlling all aircraft surface grid nodes on the plane unit where each grid central point is located to move by each branching process node by using the real displacement of each grid central point, and updating the speed boundary condition of the grid central point by using the real speed to finish the solving calculation of the time step.
Step eight, judging whether the solving calculation reaches the termination time step, if so, entering the step nine, otherwise, returning to the step four;
step nine, calculating the structure real displacement of each time step through the generalized displacement and txt file of the generalized speed of each time step output by the host node, the structure finite element node coordinates obtained in the step two and the modal data of each order;
step ten, monitoring the structural real displacement-time curve of the given position under different incoming flow speeds, and finding out the flutter speed meeting the calculation precision.
Compared with the prior art, the invention has the following advantages:
(1) the invention relates to a parallelization high-precision flutter calculation method based on a modal method, which is characterized in that a global numbering matching search technology is used when a host node transmits data to a node, and a generalized aeroelastic motion equation set is compiled in a user-defined function; the structural equation solving process is written into a user-defined function, so that the solid kinetic equation is avoided being solved, the flow field and the structural kinetic equation are solved simultaneously in the fluid dynamics software, and time domain flutter calculation is carried out. The fluid-solid coupling calculation speed is greatly improved, and the calculation cost is saved.
(2) According to the parallel high-precision flutter calculation method based on the modal method, the method can support multi-core parallel calculation by performing parallel processing, and the calculation speed of solution is further improved.
(3) The parallelization high-precision flutter calculation method based on the modal method is wide in application range, and can solve the flutter problem of participation of nonlinear aerodynamic force (transonic speed, low Reynolds number, large attack angle and the like) in different speed (low speed, transonic speed and supersonic speed) ranges.
Drawings
FIG. 1 is a flow chart of a parallelization high-precision flutter calculation method based on a modal method according to the present invention;
FIG. 2 is a diagram of the present invention illustrating a fluid-solid coupling computation process parallelized in one time step;
FIG. 3 is a diagram of the correspondence between vector elements in different numbering modes in the global numbering matching search technique of the present invention;
FIG. 4 is a schematic diagram of the extrapolation of the center of the body unit to the center of the face unit of the present invention;
FIG. 5 is a graph of AGARD 445.6 wing tip displacement versus time at different incoming flow velocities according to the present invention.
Detailed Description
The present invention will be described in further detail and with reference to the accompanying drawings so that those skilled in the art can understand and practice the invention.
The invention provides a fluid-structure interaction flutter calculation method for nonlinear aerodynamic force, which is suitable for various structural models such as wings, missiles and airplanes, and the method utilizes udf (user-defined function) function in fluent software, and solves structural change parameters while calculating nonlinear unsteady aerodynamic force through a time domain flutter solving algorithm based on a modal method, and in addition, programs are subjected to parallel processing, so that flutter analysis is performed, and the method has the advantages of high calculation efficiency, high precision, wide applicability and the like.
The parallel high-precision flutter calculation method of the modal method is based on fluid dynamics calculation and user-defined function functions in Ansys software, a flutter solving algorithm is developed in the user-defined function, the parallel high-precision flutter calculation method is provided with a displacement interpolation module, a body pressure to surface pressure conversion module, a parallel processing module, a global serial number matching search module and a structural dynamics equation time domain solving module, and the parallel high-precision flutter calculation method can be applied to engineering flutter calculation of nonlinear aerodynamic forces such as large attack angle, separation flow, low Reynolds number and resistance rudder.
The flutter calculation method comprises the following specific steps as shown in fig. 1:
step one, using flow field meshing software to perform meshing on a flow field domain of the aircraft, and outputting a wing surface mesh node coordinate, a mesh center point coordinate, a face unit global number and an accumulated node number to a txt file through a user-defined function in the fluent software.
Each face unit includes at least three face mesh nodes and a mesh center point.
Step two, outputting node coordinates of the finite element of the structure and modal displacement of each order in the finite element software of the structure;
and reading the node coordinates of the finite element of the structure, the modal displacement of each order, the node coordinates of the airfoil grid in the fluid grid and the coordinates of the central point of the grid in a displacement interpolation module based on an Ansys user-defined function.
And thirdly, running fluent software in a parallel mode, and interpolating the read modal displacement values of all orders to grid nodes on the surface of the aircraft by executing a thin plate spline interpolation (TPS) method in the user-defined function by utilizing a geometric relation to obtain the modal displacement of all orders corresponding to the grid nodes of the flow field.
By writing the thin plate spline interpolation method into a user-defined function, the modal displacement interpolation of the structure finite element model into the flow field grid is realized in fluent, and the method can be suitable for modal formats output by any finite element software; the interpolation modality can be applied directly to the dynamic mesh boundary of Ansys or other fluid dynamics software.
And step four, executing a user-defined function to enable host nodes to read modal displacement and txt files of each order corresponding to the central point of each grid respectively, and calculating the sum of generalized aerodynamic force of each order by using branch process nodes to obtain the generalized aerodynamic force of the whole wing.
Host node (host): the node is responsible for receiving user instructions and transmitting the user instructions to the 0 th branch node 0; meanwhile, the master node is also responsible for receiving all the information gathered by the node 0 from other nodes, processing and feeding back to Cortex. In parallel computing, the host node does not have any flow field and grid related information, and does not participate in fluid related solving computation.
Branch node 0: receiving instructions from the main node and distributing the instructions to the main node and the rest branch nodes; meanwhile, node 0 may receive data from other nodes, perform global processing (global sum, difference, etc.), and then feed back the data to the master node. In parallel computing, node 0 also undertakes the solving work of the flow field besides distributing instructions and summarizing information. The node stores the flow field and grid related information of a parallel block and is responsible for solving and calculating the flow field of the block.
Other branch nodes (nodes 1 to n): the other branch nodes receive the instruction from node 0 and also transmit their own data to node 0. In parallel computing, other branch nodes mainly undertake flow field solving work. Each branch node stores the flow field and grid related information of a corresponding parallel block and is responsible for solving and calculating the flow field of the block.
A global number matching search technology and a body unit pressure and surface unit pressure conversion technology used when the density base is used for solving and calculating aerodynamic force are used when the host node transmits data to the node; as shown in fig. 2, firstly, writing XYZ coordinates of a fluid grid node into a file 1, writing XY coordinates of a fluid grid center point and global numbers and accumulated node numbers of corresponding surface units into a file 2, reading four-order modal displacement after interpolation in the files 1 and 2 by a host node, transmitting the four-order modal displacement to each branch-thread node, reading aerodynamic force of a unit above the node by each branch-thread node, and multiplying the aerodynamic force by each order modal value of all grid center points to obtain generalized aerodynamic force of the node; and finally, summing the sum of the generalized aerodynamic forces of all the branch line node nodes again to obtain a generalized aerodynamic force global sum, and combining the given initial generalized displacement and generalized speed of the generalized aerodynamic elastic motion, or setting the result of the generalized displacement and generalized speed obtained in the previous calculation as an initial value to solve a generalized aerodynamic elastic motion equation set so as to further calculate the real speed, displacement and surface boundary speed of each grid node on the surface of the aircraft.
The specific process is as follows:
step 401, executing a user-defined function for each plane unit in a parallel mode to enable a host node to read modal displacement of each order corresponding to each grid center point and transmit the modal displacement to each branch thread node;
each branch thread node obtains modal displacement of each order of all grid central points contained by the node by adopting a global numbering matching search algorithm, and the specific process is as follows:
firstly, globally numbering vectors faceGlobalId of all surface units of the surface of the aircraft according to serial local numbering kserialArranging;
the vector faceGlobalId is Nc in length and is numbered k serially according to localserialAligned with the ith order mode vector phi of the center point of the airfoil fluid grid(c),i (Nc×1)The arrangement mode of the medium elements is consistent.
Then, use udf function library to exchangeConverting the part number into a function F _ ID of a global number, circularly traversing serial local numbers of all face units in each thread node, and returning the corresponding global number F _ ID (k)parallel,tf)；
Next, the global number F _ ID (k) in each face unit is assignedparallelTf) comparing and searching with all global numbers stored in the vector faceGlobalId, and recording the serial local number k of the position of the serial local number in the vector faceGlobalId when the same global number is foundserialAnd recording the parallel local number k of the current unit in the nodeparallelForm a set of dictionary keys and values { k }parallel:kserial}。
Finally, in the calculation of the generalized aerodynamic force, when the k-th position is calculatedparallelThe aerodynamic force of the surface unit is used for numbering k parallel partsparallelSubstitution { k }parallel:kserialReturns the corresponding serial local number kserialObtaining the ith order modal vector phi of the center point of the airfoil fluid grid(c),i (Nc×1)K of (1)serialIndividual modal value
And obtaining modal displacement of each order of the central point of each grid in the same way.
The global number matching search algorithm is formed by a face unit loop, and is executed in each node i:
the vector faceglobalsd is a global numbering vector of the surface unit of the airfoil read by the host node, and the information is contained in the grid data of the airfoil read by the interpolation module. As shown in fig. 3, the relationship between the cell numbers of the grids of each plane is shown. Since all node loops in the UDF are realized by nesting of 'surface unit loop + in-plane node loop', only the serial local number k of the surface unit is neededserialThe node position v can be positioned by summing the number n of the node in the surfaceserial(strings)Line local numbering) without searching for a match determination by the node global number. At this time, the accumulated node number N corresponding to each surface unit is usedTIt is apparent that the sequence N is formed at the number of accumulated nodes readTIn (f), the k-thserialAn element NT(kserial) Expressed as front kserial-1 surface element with a total of node numbers NT(kserial). The in-plane node partial number n represents the kthserialThe nth node within a face unit. Thus, the local number N of the in-plane node and the cumulative node number N of the plane unitT(kserial) And the ith order modal vector phi of the fluid grid node(f),i (Nf×1)Serial local numbering v of elements in (1)serialThe following relationships are formed:
vserial＝NT(kserial)+n
in fact, the method is suitable for various forms of mixed grids and has good universality.
Step 402, aiming at each branching process node, the node calls each order modal value of all grid central points contained in the node, and respectively multiplies the modal values by the aerodynamic force of a surface unit where each grid central point is located to obtain each order generalized aerodynamic force of the grid unit, and sums to obtain the sum of each order generalized aerodynamic force of the node;
when the aerodynamic force of the surface unit where the central point of each grid is located is solved by using the density base, the specific process is as follows:
firstly, each surface unit of the surface of the aircraft corresponds to one body unit, and the grid center coordinate of each surface unit, the grid center coordinate of the body unit and the pressure intensity are obtained.
Then, for each individual unit, a pressure gradient from the center of the individual unit to the center of each surface unit and a coordinate vector from the center of the individual unit to the center of each surface unit are obtained.
And finally, calculating the pressure value of the center of each surface of the aircraft surface, and multiplying the pressure value by the area of the surface unit to obtain the aerodynamic force of the surface unit.
The conversion of body pressure to surface pressure is applied to calculate the aerodynamic force of the aircraft surface.
Because the fluid dynamics solver has two solving modes: pressure basis and density basis. The pressure base solution is suitable for low-speed conditions, and the density base solution is suitable for pressure sound velocity to supersonic velocity states. The pressure value of the center of the surface grid can be directly read by a pressure base solving mode, and the density base solving can not be realized. In order to enable the method to be suitable for a density-based solver to perform flutter calculation in a high-speed state, the pressure value of the center of the grid surface unit is obtained by performing extrapolation on the pressure value of the center point of the body unit where the grid surface unit is located. The mathematical expression is as follows:
wherein the content of the first and second substances,
is the value of the pressure intensity at the center of the volume grid,
is the central pressure intensity of the surface grid,
pressure gradient at the center of the volume grid, dCCFCIs the vector from the center of the volume mesh to the center of the face mesh. As shown in fig. 4, the extrapolated relationship between the volume cell center and the face cell center. However, when a density-based solver is used, the pressure gradient at the center of the body unit is not directly available. Therefore, the ideal gas state equation is introduced:
wherein R is a general gas constant, M is a gas molar mass, and the ratio of the general gas constant to the gas molar mass
It can be obtained by a macro function in udf, where T is the thermodynamic temperature of the gas and ρ is the density. To make things of mutual relationshipGradient of ideal gas state equation
The following can be obtained:
wherein the content of the first and second substances,
is a gradient of density,
A gradient of thermodynamic temperature of the gas;
applying the equation to the center of the volume element solves for the pressure gradient at that point. At the same time, vector dCCFCThe coordinate of the center point of the surface unit and the coordinate of the center point of the body unit can be obtained through the difference, and the coordinates are obtained by means of macro functions in user-defined functions respectively. Therefore, the central pressure of the surface unit can be calculated by an extrapolation method of the central pressure value of the volume grid.
And step 403, executing user-defined function division to enable the branch node 0 to sum the aerodynamic force sums of all the branch thread nodes in the parallel mode again to obtain the generalized aerodynamic force sum of each order.
And step 404, transmitting the sum of the generalized aerodynamic forces of each order to a host node through a branch node 0 to obtain the generalized aerodynamic force of the whole aircraft surface.
Step five, solving a generalized aeroelastic motion equation set compiled in a user-defined function in a host node by utilizing the generalized aerodynamic force of the surface of the whole aircraft to obtain the generalized displacement and the generalized speed of each order of mode corresponding to the central point of each grid, outputting the generalized displacement and the generalized speed of each order of mode at the current time step to a txt file by the host node, and transmitting the generalized displacement and the generalized speed to each branch process node;
when the host node transmits data to the node, a global numbering matching search technology is used, and a generalized aeroelastic motion equation set is written in a user-defined function, so that a structural motion equation is solved simultaneously in the fluent fluid calculation process.
And solving the generalized aeroelastic motion equation set in the host to obtain the generalized displacement and the generalized speed of each order of motion, and transmitting the generalized displacement and the generalized speed to each node i. The discretization processing of the generalized aeroelastic equation set in the host node adopts a four-order Runge Kutta method with variable step length, has higher calculation precision, and can be changed into other time domain numerical integration methods.
Step six, each branching thread node executes a user-defined function: calculating the real speed and displacement of each grid node on the surface of the aircraft by using the generalized displacement and the generalized speed of each order of mode at the current time step;
and seventhly, controlling the movement of all the aircraft surface grid nodes by using the real displacement of each grid node on the surface of the aircraft by each branch thread node, and updating the speed boundary conditions of the grid nodes by using the real speed to finish the solving calculation of the time step.
The effect of the velocity boundary condition is to calculate the aerodynamic force.
Step eight, judging whether the solving calculation reaches the termination time step, if so, entering the step nine, otherwise, returning to the step four;
step nine, calculating the structure real displacement of each time step through the generalized displacement and txt file of the generalized speed of each time step output by the host node, the structure finite element node coordinates obtained in the step two and the modal data of each order;
step ten, monitoring the structural real displacement-time curve of the given position under different incoming flow speeds, and finding out the flutter speed meeting the calculation precision.
As shown in FIG. 5, the abscissa represents time and the ordinate represents the trailing edge tip displacement. The displacement amplitude from left to right is respectively in a convergence state, a constant amplitude state and a divergence state along with time. From this it can be determined that the flutter velocity of the wing is about 175.3 m/s.
TABLE 1 comparison of calculated flutter speed with experimental results
The method for converting the body pressure to the surface pressure based on the Ansys user-defined function and the extrapolation method based on the body grid center data realize the function of obtaining the surface pressure data from the body pressure data, and can be suitable for flutter calculation of high-speed flowing density-based aerodynamic force solution. A program parallelization method and a global number matching search method based on an Ansys user-defined function. The global number matching search method realizes data extraction and transmission between each node and a host node during parallelization calculation by establishing the relation between the global number and the local number of the flow field data. The program parallelization method established on the basis improves the operation efficiency and greatly saves the operation time by parallelizing the programs of flutter data distribution, node division solution and collection. And solving a structural dynamic equation of a variable-step length Runge Kutta algorithm based on the Ansys user-defined function. The method realizes the adjustment of the solving time step according to the integral truncation error in the solving process, thereby ensuring the time domain flutter calculation precision.
Claims (4)
1. A parallelization high-precision flutter calculation method based on a modal method is characterized by comprising the following specific steps:
step one, using flow field meshing software to perform meshing on a flow field domain of the aircraft, and outputting a wing surface mesh node coordinate, a mesh center point coordinate, a face unit global number and an accumulated node number to a txt file in fluent software;
each face unit comprises at least three face mesh nodes and a mesh central point;
step two, outputting node coordinates of the finite element of the structure and modal displacement of each order in the finite element software of the structure;
running fluent software in a parallel mode, and interpolating the read modal displacement values of each order to grid nodes on the surface of the aircraft by using a thin plate spline interpolation method through a geometric relation to obtain the modal displacement of each order corresponding to each flow field grid node;
step four, reading modal displacement and txt files of each order respectively corresponding to the central point of each grid by a host node, and calculating the sum of generalized aerodynamic force of each order by using each branch thread node to obtain the generalized aerodynamic force of the whole wing;
step five, solving a generalized aeroelastic motion equation set compiled in a user-defined function in a host node by utilizing the generalized aerodynamic force of the surface of the whole aircraft to obtain the generalized displacement and the generalized speed of each order of mode corresponding to the central point of each grid, outputting the generalized displacement and the generalized speed of each order of mode at the current time step to a txt file by the host node, and transmitting the generalized displacement and the generalized speed to each branch process node;
step six, each branching thread node executes a user self-defined function: calculating the real speed and displacement of the central point of each grid on the surface of the aircraft by using the generalized displacement and the generalized speed of each order of mode at the current time step;
step seven, each branching process node controls all aircraft surface grid nodes on the plane unit where each grid central point is located to move by using the real displacement of each grid central point, and the speed boundary condition of the grid central point is updated by using the real speed, so that the solving calculation of the time step is completed;
step eight, judging whether the solving calculation reaches the termination time step, if so, entering the step nine, otherwise, returning to the step four;
step nine, calculating the structure real displacement of each time step through the generalized displacement and txt file of the generalized speed of each time step output by the host node, the structure finite element node coordinates obtained in the step two and the modal data of each order;
step ten, monitoring the structural real displacement-time curve of the given position under different incoming flow speeds, and finding out the flutter speed meeting the calculation precision.
2. The parallelization high-precision flutter computing method based on the modal method according to claim 1, wherein the step four comprises the following specific processes:
step 401, for each surface unit performed in a parallel manner, a host node reads modal displacement of each order corresponding to a central point of each grid and transmits the modal displacement to each branching process node;
step 402, aiming at each branching process node, the node calls each order modal value of all grid central points contained in the node, and respectively multiplies the modal values by the aerodynamic force of a surface unit where each grid central point is located to obtain each order generalized aerodynamic force of the grid unit, and sums to obtain the sum of each order generalized aerodynamic force of the node;
step 403, the branch node 0 sums the aerodynamic force sums of all the branch process nodes performed in the parallel mode again to obtain the generalized aerodynamic force sum of each order;
and step 404, transmitting the sum of the generalized aerodynamic forces of each order to a host node through a branch node 0 to obtain the generalized aerodynamic force of the whole aircraft surface.
3. The method according to claim 2, wherein in step 401, each sub-thread node obtains modal displacement of each order of all grid center points included in itself by using a global numbering matching search algorithm, and the specific process is as follows:
firstly, globally numbering vectors faceGlobalId of all surface units of the surface of the aircraft according to serial local numbering kserialArranging;
then, by using a function F _ ID for converting the local number into the global number in the udf function library, circularly traversing the serial local numbers of all the face units in each thread node, and returning the corresponding global number F _ ID (k)parallel,tf)；
Next, the global number F _ ID (k) in each face unit is assignedparallelTf) comparing and searching with all global numbers stored in the vector faceGlobalId, and recording the serial local number k of the position of the serial local number in the vector faceGlobalId when the same global number is foundserialAnd recording the parallel local number k of the current unit in the nodeparallelForm a set of dictionary keys and values { k }parallel:kserial}；
Finally, in the calculation of the generalized aerodynamic force, when the k-th position is calculatedparallelThe aerodynamic force of the surface unit is used for numbering k parallel partsparallelSubstitution { k }parallel:kserialReturns the corresponding serial local number kserialObtaining the ith order modal vector phi of the center point of the airfoil fluid grid(c),i (Nc×1)K of (1)serialIndividual modal value
And obtaining modal displacement of each order of the central point of each grid in the same way.
4. The method according to claim 2, wherein in the step 402, when the aerodynamic force of the surface unit with the central point of each grid is solved by using the density basis, the specific process is as follows:
firstly, each surface unit on the surface of the aircraft corresponds to a body unit, and the grid center coordinate of each surface unit and the grid center coordinate and pressure of the body unit are obtained;
then, aiming at each individual unit, respectively solving the pressure gradient from the center of the individual unit to the center of each surface unit and the coordinate vector from the center of the individual unit to the center of each surface;
and finally, calculating the pressure value of the center of each surface unit on the surface of the aircraft, and multiplying the pressure value by the area of the surface unit to obtain the aerodynamic force of the surface unit.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2020107955752 | 2020-08-10 | ||
| CN202010795575 | 2020-08-10 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112182737A true CN112182737A (en) | 2021-01-05 |
| CN112182737B CN112182737B (en) | 2022-05-03 |
Family
ID=73924453
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010872130.XA Active CN112182737B (en) | 2020-08-10 | 2020-08-26 | Parallelization high-precision flutter calculation method based on modal method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112182737B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114117950A (en) * | 2021-10-28 | 2022-03-01 | 中国运载火箭技术研究院 | Flutter judgment method for shuttle aircraft based on acting principle |
| CN115422654A (en) * | 2022-08-21 | 2022-12-02 | 西北工业大学 | CFD/CSD technology-based efficient high-precision flutter time domain analysis method for cross/supersonic aircraft |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101599104A (en) * | 2009-07-16 | 2009-12-09 | 北京航空航天大学 | A kind of analogy method of blade flutter boundary of aviation turbine engine |
| CN106508028B (en) * | 2010-09-30 | 2014-07-02 | 上海机电工程研究所 | A kind of determination complex appearance aircraft supersonic speed, the hypersonic method for having angle of attack tremor secure border |
| WO2017084106A1 (en) * | 2015-11-20 | 2017-05-26 | 田川 | System and method for numerical simulation of aircraft flow field |
| CN109933876A (en) * | 2019-03-03 | 2019-06-25 | 西北工业大学 | A kind of unsteady aerodynamic force order reducing method based on broad sense aerodynamic force |
-
2020
- 2020-08-26 CN CN202010872130.XA patent/CN112182737B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101599104A (en) * | 2009-07-16 | 2009-12-09 | 北京航空航天大学 | A kind of analogy method of blade flutter boundary of aviation turbine engine |
| CN106508028B (en) * | 2010-09-30 | 2014-07-02 | 上海机电工程研究所 | A kind of determination complex appearance aircraft supersonic speed, the hypersonic method for having angle of attack tremor secure border |
| WO2017084106A1 (en) * | 2015-11-20 | 2017-05-26 | 田川 | System and method for numerical simulation of aircraft flow field |
| CN109933876A (en) * | 2019-03-03 | 2019-06-25 | 西北工业大学 | A kind of unsteady aerodynamic force order reducing method based on broad sense aerodynamic force |
Non-Patent Citations (1)
| Title |
|---|
| 互峰强，陈艾荣: "全模态颤振分析的实用方法", 《公路交通科技》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114117950A (en) * | 2021-10-28 | 2022-03-01 | 中国运载火箭技术研究院 | Flutter judgment method for shuttle aircraft based on acting principle |
| CN114117950B (en) * | 2021-10-28 | 2022-09-06 | 中国运载火箭技术研究院 | Flutter judgment method for shuttle aircraft based on acting principle |
| CN115422654A (en) * | 2022-08-21 | 2022-12-02 | 西北工业大学 | CFD/CSD technology-based efficient high-precision flutter time domain analysis method for cross/supersonic aircraft |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112182737B (en) | 2022-05-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Biancolini et al. | Sails trim optimisation using CFD and RBF mesh morphing | |
| CN112182737B (en) | Parallelization high-precision flutter calculation method based on modal method | |
| CN112414668B (en) | Wind tunnel test data static bomb correction method, device, equipment and medium | |
| Zhang et al. | Efficient aerodynamic shape optimization of the hypersonic lifting body based on free form deformation technique | |
| Oruc et al. | Towards Real-Time Fully Coupled Flight Dynamics and CFD Simulations of the Helicopter-Ship Dynamic Interface | |
| CN111563297A (en) | Supersonic aircraft thermal environment calculation method based on BP network | |
| Neumann et al. | Coupling strategies for large industrial models | |
| CN115407680A (en) | Parallelized high-precision time domain pneumatic elastic control coupling simulation method | |
| Chen et al. | Adding-point strategy for reduced-order hypersonic aerothermodynamics modeling based on fuzzy clustering | |
| Duong et al. | Fluid–structure coupling in time domain for dynamic stall using purely lagrangian vortex method | |
| Byun et al. | Wing-body aeroelasticity using finite-difference fluid/finite element structural equations on parallel computers | |
| Woodcock | A comparison of methods used in lifting surface theory | |
| Cusick et al. | Creation of Design and Analysis Tools for Large Design Space Reusable Launch Vehicle Shape Optimization | |
| Byun et al. | Wing-body aeroelasticity on parallel computers | |
| Tian et al. | Efficient Parallel Dynamic Overset Unstructured Grid for Flow Simulation with Moving Bodies | |
| Phillips et al. | Three-Dimensional Uncoupled Fluid Structure Analysis of Small Unmanned Aerial Vehicle Wing | |
| Peng et al. | Rapid Aerodynamic Shape Optimization With Payload Size Constraints for Hypersonic Vehicle | |
| Zhang et al. | Numerical optimization algorithm for unsteady flows of rotor based on web service. | |
| Love et al. | Enhanced maneuver airloads simulation for the Automated Structural Optimization System-ATROS | |
| Scherding et al. | Data-driven reduction of lookup libraries for the simulation of non-equilibrium effects in hypersonic flows | |
| Smith et al. | Aerodynamic Database Requirements for the Detailed Design of Tactical Aircraft: Implications for the Expanded Application of CFD | |
| Morton et al. | Computational aircraft and armament stability and control techniques applied to the F-16 | |
| Melville et al. | Aeroelastic tailoring for gust-energy extraction | |
| Cunningham et al. | Multi-Fidelity Probabilistic Aerodynamic Database Generation with the ProForMA Tool | |
| Cheng et al. | A pixelation method for loads transfer on fluid structure interface |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |