JP2013210958A - Finite element analysis system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a finite element analysis system capable of suppressing increase in an analysis time while maintaining high precision of analysis.SOLUTION: A finite element analysis system performs stress analysis of contact of a wheel body that rolls on a long rail body toward a traveling direction along it, and includes analysis means that repeats steps as follows: the rail body is subjected to virtual division into a plurality of unit blocks along a longer direction thereof; after defining an analysis section composed of back and forth unit blocks of a predetermined number while interposing the unit block at least containing the contact part of the wheel body and the rail body, finite element analysis is performed until the contact part moves in the traveling direction for a predetermined traveling distance within the analysis section; a new analysis section is defined by shifting the analysis section by one unit block in the traveling direction; and finite element analysis is further performed.

Description

  The present invention relates to a finite element analysis system using a finite element method, and in particular, a finite element analysis for contact stress analysis by a wheel body that rolls on a long rail body in a traveling direction along the rail body. About the system.

  In the finite element method, in general, the analysis accuracy can be improved as the elements of the analysis model are set finer. On the other hand, the number of elements increases and time is required for calculation processing. Therefore, in the analysis of a long object that becomes a large-scale analysis model, for example, the analysis of stress on the roadbed along the moving direction of the wheel body rolling on the roadbed, the allowable analysis accuracy and the analysis time Various ideas are required in relation.

  For example, Patent Document 1 discloses a method for analyzing stress and / or distortion generated in a conveyor belt caused by contact with a roller by using a finite element method for a long conveyor belt that is moved in contact with the roller. Yes. Here, the elements of the conveyor belt are finely set at the contact position of the roller and coarser as the distance from the contact position is increased, thereby achieving both improvement in analysis accuracy and reduction in processing time. Since the conveyor belt is supported by rollers arranged at the center interval L, it is sufficient to perform the calculation process by the finite element method only for the conveyor belt having the length L.

  Further, Patent Document 2 discloses an analysis method by a finite element method for the purpose of estimating damage states such as peeling due to rolling fatigue of rolling bearings such as ball bearings and roller bearings. Here, peeling or the like is caused by non-metallic inclusions present in the metal base of the rolling bearing. Further, when the rolling element does not pass only once on the surface of the metal base but rolls repeatedly, strain (internal damage) is accumulated inside the metal base. Therefore, the stress calculation of the matrix is performed by the finite element method around the non-metallic inclusions that cause the accumulation of strain inherent in the vicinity of the surface of the metallic substrate. In such a finite element method, the elements given to the metallic substrate are finely set at the contact positions in the vicinity of the non-metallic inclusions and within the moving range of the rolling elements, and are set coarser as they move away from these positions. Since it is a simulation of rolling fatigue based on non-metallic inclusions, even if the calculation process is performed with the rolling element's moving range limited to the periphery of the non-metallic inclusions, the analysis accuracy will not be impaired and the analysis time will also be reduced. It can be shortened.

  On the other hand, for example, Patent Document 3 discloses a method for simulating the behavior of a rolling element tire moving on a road surface provided with a water film. When the tire rolls at high speed, it is necessary to model and simulate the road surface together with the tire over a long distance. Accordingly, an analysis method is adopted in which the fluid analysis region is moved so as to follow the movement of the tire while the fluid analysis region is limited to a predetermined length range with respect to the tire traveling direction. Specifically, each time the tire model is moved by a distance corresponding to the moving speed and the step size of the analysis time, the fluid analysis area is moved by the same distance, and boundary conditions are set for each element set in the fluid analysis area. Is reset. According to this method, it is not necessary to set a fluid analysis region that is long in the moving direction of the tire model, and the analysis time can be shortened.

JP-A-2005-75617 JP 2011-64444 A JP 2004-338660 A

  By the way, in the contact stress analysis by the wheel body rolling on the long rail body in the traveling direction along the rail body, various devices are required in relation to the allowable analysis accuracy and the analysis time. The In particular, as the speed of railway vehicles increases, analysis of a longer rail body is required, and such a demand is remarkable.

  The present invention has been made in view of such circumstances, and an object of the present invention is to analyze contact stress by a wheel body that rolls on a long rail body in a traveling direction along the rail body. An object of the present invention is to provide a system that shortens analysis time while maintaining high analysis accuracy.

  A finite element analysis system according to the present invention is a finite element analysis system for stress analysis by contact of a wheel body rolling on a long rail body in a traveling direction along the rail body. Is virtually divided into a plurality of unit blocks along the longitudinal direction, and an analysis section including a predetermined number of the unit blocks before and after the unit block including at least the contact portion of the wheel body and the rail body is defined. In the above, a finite element analysis is performed until the contact portion moves in the traveling direction by a predetermined traveling distance in the analysis section, and the analysis section is shifted by one unit block in the traveling direction. It further comprises analysis means for determining the section and repeating the finite element analysis.

  According to this invention, even in a finite element analysis system that repeatedly performs a predetermined calculation between given elements over a long rail body, the number of calculations can be reduced while maintaining analysis accuracy. That is, the analysis time can be shortened. Further, by making the analysis section consist of a plurality of unit blocks, the step of newly defining the analysis section can be simplified, and the analysis time can be shortened.

  In the above-described invention, an independent mesh model may be assigned to each unit block. According to this invention, the process of the step which newly defines an analysis area can be simplified, and analysis time can be shortened.

  In the above-described invention, when the finite element analysis is performed, when a new analysis section is defined, a mesh model may be additionally added to the unit block newly incorporated in the analysis. . According to this invention, in the step of newly determining the analysis section, the process of applying the mesh model can be simplified and the analysis time can be shortened.

  In the above-described invention, following the virtual division, a mesh model may be provided over the entire rail body. According to this invention, in the step of newly determining the analysis section, the process of applying the mesh model can be simplified, and the analysis time can be shortened.

  In the above-described invention, the mesh model for each unit block may be the same. According to this invention, in the step of newly determining the analysis section, the process of applying the mesh model can be simplified, and the analysis time can be shortened.

  In the above-described invention, a specific code for specifying the element node is given to the mesh model in the analysis section, and when a new analysis section is defined, the unit block newly incorporated in the analysis corresponds to the mesh model. The specific code of the unit block that is out of the analysis may be assigned to the element node. According to this invention, in the step of newly determining the analysis section by reducing the number of specific codes given to the unit block, it is possible to simplify the process of assigning the specific code to the mesh model and shorten the analysis time.

  In the above-described invention, the predetermined traveling distance may be the length of the unit block in the longitudinal direction of the rail body. According to this invention, it is possible to simplify the step of newly determining the analysis section, and to shorten the analysis time.

It is a block diagram which shows the finite element analysis system by this invention. It is a flowchart which shows the procedure of the finite element analysis in the finite element analysis system by this invention. It is a figure of the analysis model used for a finite element analysis system. It is a figure of the analysis model used for a finite element analysis system. It is a figure of the load condition of the analysis model by a finite element analysis system. It is a figure which shows the translation speed log | history of a wheel among the analysis results by a finite element analysis system. It is a figure which shows the analysis result by a finite element analysis system.

  The details of the finite element analysis system according to one embodiment of the present invention will be described with reference to FIG.

  As shown in FIG. 1, a finite element analysis system 30 is a computer-based system that simulates a wheel moving while rolling on a rail and performs stress analysis (or strain analysis) associated with the contact between the rail and the wheel. It is. The finite element analysis system 30 includes a control unit 31 and a storage unit 32, an analysis model creation unit 33 for supporting creation of an analysis model for performing finite element analysis, and an analysis interval setting for determining an analysis interval to be described later And an analysis unit 35 for performing analysis in the analysis section of the generated analysis model.

  The control unit 31 includes a CPU (not shown) that executes a predetermined program and associated ROM and RAM. Calculation processing is performed by executing each program of the analysis model creation unit 33, the analysis section setting unit 34, and the analysis unit 35 in the control unit 31. That is, the control unit 31 can function as an analysis unit of the finite element analysis system 30 by executing a predetermined program. Details of the analysis model creation unit 33, the analysis section setting unit 34, and the analysis unit 35 will be described later.

  The storage unit 32 stores the analysis model created by the analysis model creation unit 33 and stores the analysis result calculated by the analysis unit 35. That is, input data required for the operation of the control unit 31 and output data calculated by the control unit 31 can be stored as necessary.

  Next, the analysis processing procedure in the finite element analysis system 30 will be described with reference to FIGS.

  As shown in FIG. 2, in the finite element analysis system 30, the analysis model creation unit 33 is activated to create an analysis model (S1). As shown in FIG. 3, when a wheel 41 that rolls on the rail 42 is selected as the target of the analysis model 40, first, a predetermined section of the rail 42 is virtually divided into a plurality of unit blocks along the longitudinal direction. To do. Here, all the unit blocks have the same shape. The object of the analysis model 40 is stored in the storage unit 32. It should be noted that a predetermined value required for analysis, such as a specified value for determining an analysis interval, which will be described later, may be input at this time and stored together in the storage unit 32.

  Next, the analysis section setting unit 34 sets an analysis section for the rail 42 (S2). The analysis section is determined based on the position of the contact portion P that contacts the rail 42 of the wheel 41. That is, as shown in FIG. 3A, when the contact portion P is on the unit block 42b, a unit block consisting of the unit block 42b and a predetermined number of unit blocks before and after the block is used as an analysis section. Determine. At this time, it is preferable that the predetermined number is determined so as to include at least a unit block having a relatively large influence on the analysis result in the analysis section so that the analysis accuracy can be maintained high. Here, a total of three continuous unit blocks 42 a, 42 b, and 42 c, one in front of the traveling direction of the wheel 41 (X direction) and one in the rear, are determined as the analysis section of the rail 42.

  Subsequently, the analysis model creation unit 33 gives a mesh model to the wheel 41 and the rail 42 (S3). In addition, although the mesh model is given to the wheel 41 according to the conditions inputted in advance, this mesh model is omitted in the drawing. The rail 42 is provided with a mesh model that is independent for each unit block with respect to the four unit blocks of the unit blocks 42a to 42c defined as the analysis section and the unit block 42d that is a dummy block. That is, the mesh model is given so that one element constituting the mesh model is accommodated in the unit block without straddling the plurality of unit blocks. Further, the mesh models given to the unit blocks 42a to 42d are the same.

  Subsequently, the analysis unit 35 performs calculation processing by the finite element method until the contact portion P is moved by a predetermined distance (S4). Specifically, the calculation process is performed from when the position of the contact portion P is at the rear end of the unit block 42b until it reaches the front end of the unit block 42b by rolling of the wheel 41.

  Further, as shown in FIG. 3B, after the calculation process until the contact portion P reaches the front end of the unit block 42b, an end determination is performed to determine whether the end condition is satisfied (S5). If the end determination is negative (S5; N), the process returns to the analysis section setting (S2). Note that the end condition can be set in various cases, for example, when the contact portion P moves to a position determined as the end point of the analysis, or when the value calculated by the analysis reaches a predetermined value. is there.

  When the end determination is negative, the analysis section setting unit 34 performs the analysis section setting (S2) again. At this time, the contact portion P reaches the rear end of the unit block 42c and is above the unit block 42c. That is, the analysis section of the rail 42 is defined as a unit block 42c and one unit block 42b, 42c, and 42d one before and after the other. That is, the rear end and the front end of the analysis section are newly set by shifting each unit block in the traveling direction (X direction). At this time, the analysis result for the unit block 42 a outside the analysis section is stored in the storage unit 32.

  Subsequently, the analysis model creation unit 33 performs the addition of the mesh model (S3) again. The newly determined analysis section includes unit blocks 42b to 42c, which have already been given a mesh model in the previous calculation process. That is, the same mesh model as that of the unit block 42a is additionally given only to the unit block 42e newly incorporated in the analysis as a dummy block. Note that the mesh model of the wheel 41 is the same as that already given.

  Subsequently, similarly to the above, calculation processing is performed until the contact portion P reaches the front end of the unit block 42c in the analysis section newly determined by the analysis unit 35, and thereafter the calculation processing is continued in the same manner.

  On the other hand, if the end condition is satisfied in the end determination (S5), the end determination is affirmed (S5; Y), and the analysis process in the finite element analysis system 30 ends.

  As described above, the analysis is not always performed for all elements extending over the entire length of the rail 42, but the analysis section is limited to a partial range of the rail 42 based on the position of the contact portion P, and the analysis section is defined as the contact portion P. The calculation scale of the analysis can be reduced by moving according to the position. That is, the analysis time can be shortened while maintaining high analysis accuracy by maintaining the fineness of the mesh model.

  The process of moving the analysis section in the analysis section setting (S2) may be performed every time the contact portion P moves by a predetermined distance, and is performed except when the contact portion P reaches the end of the unit block. You can also. At this time, if the predetermined distance is the same as the length of the unit block, the positional relationship between the analysis section and the wheel 41 can always be within a certain range, and the analysis accuracy can be stabilized.

  Here, among the nodes of the elements determined by the mesh model, each node of the unit block added to the analysis section is given a node number for specifying these. Further, at the nodes located at the front end and the rear end of the unit block, the adjacent unit blocks share the node numbers. That is, referring again to FIG. 3A, the node numbers 1 to 20 are assigned to the nodes of the unit blocks 42a to 42c in the analysis section, but the node number 7 is assigned to the unit block 42a and the unit block 42b. 8 and 8 and the unit block 42b and the unit block 42c share the node numbers 13 and 14.

  In addition, a new node number is added to the analysis so that the same node number will not be duplicated in the unit block of the analysis section, because a new node number will not be added when moving the analysis section. Used as the node number of the unit block. The unit block incorporated in the analysis includes a dummy block described later. Further, in order to efficiently assign node numbers to each node, it is preferable to determine combinations of node numbers within a unit block.

  Specifically, in FIG. 3A, the unit block 42d not in the analysis section shares the node numbers 19 and 20 with the unit block 42c in the analysis section as a dummy block, and in addition, node numbers 19 to 24 are added. Is granted. That is, the unit block 42d is not in the analysis section, but some of the node numbers 19 and 20 are incorporated into the analysis as dummy blocks shared with the unit block 42c. In FIG. 3B, the unit block 42d is newly added to the analysis section, and at the same time, the same mesh model corresponding to the unit block 42a is added to the unit block 42e newly incorporated in the analysis as a dummy block. Is granted. That is, node numbers 1 to 6 are set, but node numbers 1 and 2 are shared with the unit block 42d. Thereby, the number of node numbers can be suppressed. Subsequently, in FIG. 3C, a unit block 42e is newly added to the analysis section, and the same processing is performed.

  Next, FIG. 4 thru | or FIG. 4 thru | or about the example which compared the calculation processing result obtained by analyzing with the above-mentioned finite element analysis system 30 with the calculation processing result by the analysis by the conventional type which always calculates about the element of the full length of a rail. This will be described with reference to FIG. The analysis by the finite element analysis system 30 is displayed as “moving type”.

  As shown in FIG. 4, the analysis includes a wheel 51 provided with a mesh model as a moving type, and a rail 52 provided with a mesh model applied to a predetermined unit block after being divided into unit blocks over the entire length of the rail 52. In the model 50, the rail 52 is divided into a plurality of unit blocks 52a to 52f having the same shape. The display of other unit blocks was omitted. In this mobile analysis, a total of five unit blocks including the unit block 52b with which the wheel 51 contacts, the rear unit block 52a, and the front unit blocks 52c to 52e are analysis sections. The unit block 52f displayed in the forefront is set to be incorporated into the analysis as a dummy block that shares some node numbers with the unit block 52e added to the analysis section. The length of each unit block in the traveling direction can be set arbitrarily, but here it is 0.125 m. In addition, the element 54 which protrudes to a side surface is set to the wheel 51, and it was used as a marker for clearly showing the rolling of the wheel 51 by setting the rigidity and the density so small as not to affect the analysis. .

  In the mobile analysis, there is no need to set a unit block that is not included in the analysis. That is, the total length of the rail is the total length of the unit block added to the analysis section from the start to the end of the analysis and the last unit block set as a dummy block.

  As shown in FIG. 5, several analysis conditions are given to the analysis model 50. Here, in order to give torque to the axle 53, a concentrated load of 0.75 kN is equally applied to the rolling direction (circumferential direction) of the outer periphery of the axle 53, and 50.0 kN corresponding to the weight of the vehicle body is applied downward at the center. A load was applied. In addition, a constraint condition for completely fixing the bottom surface of the rail 52 was given, and a constraint condition for fixing the end surface 53a of the axle 53 in the Y-axis direction was given so that the axle 53 did not move in the Y-axis direction.

  In the conventional analysis, the total length of the rail was set to two types of 5 m and 10 m, and the same mesh model as the above-described analysis model 50 was given over the entire length of the rail. Furthermore, in the conventional analysis, the entire length of the rail was always analyzed without defining the analysis section. The other analysis conditions are the same as the mobile analysis conditions.

  FIG. 6 shows a history of the moving speed in the traveling direction of the central axis of the wheel when the rolling of the wheel is simulated under the above-described analysis conditions, that is, a history of the translational speed of the wheel. Since the same torque is always applied to the axle, the translational speed of the wheel increases linearly. At such a translational speed, the analysis result obtained when the total length of the conventional rail is 5 m (conventional equation (5 m)) and the analysis result obtained when the total length is 10 m (conventional equation (10 m)) are compared with the analysis result obtained by the moving method. It can be seen that a match as shown in FIG.

  As shown in FIG. 7, when the mobile analysis is compared with the conventional analysis, the number of elements and contacts used for the analysis can be greatly reduced. The number of mobile elements and the number of contacts are calculated including dummy blocks. Moreover, in the analysis simulating until the translation speed of the wheel reaches 10 km / h, the calculation time in the moving type requires 54023 seconds, 188880 seconds in the conventional type (5 m) and in the conventional type (10 m) Compared to 583448 seconds, the calculation time ratio is very small, 0.29 times and 0.09 times, respectively. In other words, the mobile type can greatly reduce the analysis time. Moreover, if the total length of the rail is increased, the calculation time ratio is further reduced.

  On the other hand, the maximum stress in the Z direction applied to the rail when the wheel translation speed was 10 km / h was −768.253 MPa for the moving type and −7680.2252 MPa for the conventional type (5 m and 10 m). In other words, high analysis accuracy was maintained.

  According to the present embodiment, the analysis time can be shortened by limiting the analysis section to a predetermined range and moving the rail 42 (52) based on the position of the contact portion P as a reference. Further, when the analysis section is moved, since the unit P is shifted by one unit block for every movement of the contact portion P by a predetermined distance, the number of movements of the analysis section can be reduced. Furthermore, an independent mesh model is given to the unit block, for example, complicated processing such as processing for an element straddling a plurality of unit blocks can be avoided, and processing of a step for newly determining an analysis section can be simplified. In addition, by making the mesh model for each unit block the same, processing such as changing the mesh model in the step of newly determining the analysis section is unnecessary, and the processing for adding the mesh model can be further simplified. Therefore, the mesh model of the rail 42 (52) can be made fine to keep the analysis accuracy high, while the analysis time can be shortened.

  In moving the analysis section, the analysis section may be shifted in the traveling direction by one unit block according to the position of the contact portion P. That is, a pattern other than the above can be used for adding a unit block to the analysis section and removing it from the analysis section. For example, after one unit block is newly added to the front and analyzed, one of the unit blocks is removed from the rear, the analysis is performed, and this can be repeated. It is also possible to change the number of unit blocks added to the analysis section according to the position of the contact portion P according to a predetermined pattern.

  Further, when the rail is virtually divided into a plurality of unit blocks, a mesh model may be created in advance for the entire length of the rail. A unit block given a mesh model is selected as an analysis section. In comparison with the above-described embodiment, it is preferable to select a method that can simplify the process of applying the mesh model and reduce the analysis time.

  So far, representative embodiments and modifications based thereon have been described. However, the present invention is not necessarily limited thereto, and those skilled in the art will understand the gist of the present invention or the scope of the appended claims. Various alternative embodiments and modifications may be found without departing from the invention.

30 Finite Element Analysis System 34 Analysis Section Setting Unit 35 Analysis Unit 41, 51 Wheel 42, 52 Rail P Contact Unit

Claims (7)

A finite element analysis system for stress analysis by contact of a wheel body rolling on a long rail body in a traveling direction along the rail body,
The rail body is virtually divided into a plurality of unit blocks along the longitudinal direction thereof, and an analysis section comprising a predetermined number of the unit blocks before and after the unit block including at least the wheel body and the contact portion of the rail body. After defining
A finite element analysis is performed until the contact portion moves in the traveling direction by a predetermined traveling distance in the analysis section, and a new analysis section is defined by shifting the analysis section by one unit block in the traveling direction. And a finite element analysis system further comprising analysis means for repeating the finite element analysis.   The finite element analysis system according to claim 1, wherein an independent mesh model is assigned to each unit block.   3. The finite element according to claim 2, wherein when performing the finite element analysis, when a new analysis section is defined, a mesh model is additionally provided to the unit block newly incorporated in the analysis. Element analysis system.   3. The finite element analysis system according to claim 2, wherein a mesh model is applied over the entire rail body following the virtual division.   The finite element analysis system according to claim 2, wherein the mesh model for each unit block is the same.   The mesh model of the analysis section is given a specific code for specifying the element node, and when a new analysis section is defined, the analysis is performed on the corresponding element node of the unit block newly incorporated in the analysis. 6. The finite element analysis system according to claim 5, wherein the specific code of the deviated unit block is assigned.   The finite element analysis system according to claim 1, wherein the predetermined traveling distance is a length of the unit block in a longitudinal direction of the rail body.