JP4695399B2 - Simulation model generation method for filler compounding materials - Google Patents

Description

  The present invention relates to a method for generating a simulation model of a filler compounding material that can easily generate a simulation model from a filler compounding material in which a filler is dispersed and mixed in a matrix.

  In recent years, analysis by simulation has been performed to evaluate the deformation behavior, stress or strain distribution state of filler compounded material (for example, carbon reinforced rubber) in which filler (for example, carbon black) is dispersed and mixed in a matrix (for example, rubber). It has been broken. In order to perform the simulation, it is necessary to prepare a model that can be handled by a numerical analysis method such as a finite element method for the above-described filler compounding material.

  FIG. 2 shows a transmission cross-sectional image obtained by imaging the carbon reinforced rubber with a transmission electron microscope (hereinafter, abbreviated as “TEM” for abbreviated as “Transmission Electron Microscope”). A white part shows rubber | gum and the carbon particle is disperse | distributed and arrange | positioned in it. Further, FIG. 7 shows an example of a simulation model a of a conventional carbon reinforced rubber visualized. A portion surrounded by a slightly darker outline is a carbon model b obtained by modeling carbon, and an outer side thereof is a rubber model c. Each model b and c is divided by a finite number of small elements.

  The conventional simulation model a is created with reference to a transmission cross-sectional image of carbon reinforced rubber obtained by TEM or the like. However, since the outline of the carbon model b is simplified based on a circular shape, it does not accurately reflect the actual shape or the like of the carbon. Such a difference in shape may affect the simulation calculation result. It is also possible to accurately create the contour shape of the carbon model b with reference to the transmission cross-sectional image. However, since the actual carbon particles have a very complicated outline, in the conventional manual operation, it takes much time and labor to generate a model.

  The present invention has been devised in view of the above circumstances, and a method for generating a simulation model of a filler compounding material that can accurately and easily generate a simulation model from a filler compounding material in which a filler is dispersed and mixed in a matrix. The purpose is to provide.

The invention according to claim 1 of the present invention is a method for generating a simulation model of a filler compounded material in which a filler is dispersed and mixed in a matrix,
Obtaining an image containing the matrix and filler from the filler compounding material;
Extracting filler edges from the image;
After extracting the edge, setting a contour line offset by a predetermined thickness outside the edge; and
Dividing the inner region of the edge and defining as a filler model;
Dividing the outer region of the contour and defining as a matrix model ;
And dividing the region between the contour line and the edge and defining the interface model having a physical property value different from the matrix model representing the interface between the matrix and the filler .

  According to a second aspect of the present invention, in the step of acquiring the image, a step of preparing a filler compounding material having a filler filling factor smaller than a filler filling factor to be analyzed, and a transmission electron microscope from the filler compounding material A method for generating a simulation model of a filler compounding material according to claim 1, further comprising: acquiring a transmission cross-sectional image using

  According to a third aspect of the present invention, the steps of determining edge continuity after extracting the edges, and connecting the discontinuous portions when the edges are not continuous, and continuing the edges are provided. The method for generating a simulation model of a filler compounding material according to claim 1, further comprising:

  According to a fourth aspect of the present invention, after extracting the edge, determining a size of an inner area of the edge; and deleting all of the edges when the inner area is smaller than a predetermined size; The method for generating a simulation model of a filler compounding material according to claim 1, further comprising:

  In the present invention, an edge (contour) is extracted from an image of an actual filler compounding material, and a filler model is set based on the extracted edge (contour). Therefore, a simulation model including a filler model that is very close to the actual physical structure can be generated easily and efficiently.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
In the present embodiment, as a filler compounding material, a carbon reinforced rubber in which rubber as a matrix is filled with carbon as a filler will be described as an example, and a procedure for generating a two-dimensional (single plane) simulation model will be described. Most of this procedure is performed using a computer device (not shown).

  FIG. 1 shows an example of the processing procedure of the simulation model generation method of the present invention.

  In the simulation model generation method, first, a step of acquiring an image including rubber and carbon from the carbon reinforced rubber is performed (step S1). FIG. 2 shows an example of such an image 2. The image acquisition method is not particularly limited as long as the carbon C outline can be clearly obtained as an image. In the present embodiment, TEM is used for image acquisition.

  The TEM can obtain a transmission cross-sectional image of a thickness portion of about 100 nanometers from a carbon reinforced rubber in which carbon is dispersed and arranged (three-dimensionally) in rubber. In this transmission cross-sectional image, carbon contained in the thickness is projected into one planar image. Therefore, the image obtained by TEM tends to contain carbon at a rate higher than the carbon filling rate (mixing rate) of the actual carbon reinforced rubber.

  For this reason, when it is desired to obtain a cross-sectional image of a carbon reinforced rubber having a specific carbon filling rate (area ratio), the carbon reinforced rubber used for imaging has a carbon reinforced with a filling rate smaller than the carbon filling rate to be analyzed. It is better to use rubber. For example, when it is desired to obtain a cross-sectional image of a carbon reinforced rubber having a carbon filling rate of 30%, the actual carbon filling rate of the carbon reinforced rubber is smaller than that, for example, 10 to 20% is desirable. That is, it is desirable to take an image from a sample having a carbon filling rate of about 33 to 66% of the carbon filling rate to be obtained as an image. As a result, an image having a substantially desired carbon filling rate (area ratio) can be obtained using TEM.

  When TEM is used, many of the cross-sectional images of carbon reinforced rubber are obtained as physical photographs. For this reason, in the present embodiment, all or a part of the photograph is read by an image scanner or the like, and necessary areas are converted into raster data (step S2). This data is stored in a storage device of a computer (not shown). If digital data can be obtained directly from the TEM, the process of step S2 is not necessary. In the present embodiment, for example, noise is previously removed from the raster data based on an algorithm such as a moving average method or a median filter (step S3).

  Next, processing is performed on the rasterized image to extract carbon edges (step S4). The edge of the raster data refers to a portion where the brightness (numerical value) in the image changes abruptly, and in this example, the contour of carbon corresponds to it.

  The process of extracting the edge of a specific object from rasterized data can be performed using various image processing applications, for example. For example, in the edge extraction process, a threshold process is performed on the raster data in FIG. 2 to obtain the binarized data 3 as shown in FIG. 3, and each object O included in the binarized data 3 (in this example, , Indicates a group of pixels having the same color value), a stage of labeling that gives the same group attributes L1, L2,..., And Ln, and a stage of extracting an edge using a differential operation or the like for each object O Can be included. FIG. 4 shows an example of image data 4 obtained by extracting the edge E of each object O from the image. The data 4 is also stored in the storage device of the computer.

  Next, in this embodiment, the image from which the edge is extracted is converted into vector data (vectorized) (step S5). Raster data is simply a collection of dots, but vector data consists of mathematically defined numerical data such as straight lines with defined start and end points and circles or arcs with defined centers and radii. . Therefore, it is suitable for determining various geometric calculations, for example, the area of the inner region of the edge E and the continuity of the edge E. Conversion processing from raster data to vector data can be performed using various image processing applications. At this time, the modified vector data can be appropriately corrected.

  Next, in this embodiment, it is determined whether or not the edge E obtained for each object O is continuous (step S6). This determination is performed using vectorized data. Since the edge E of each object O is formed by a plurality of lines, the continuity of the edge E can be easily determined by examining the start point and the end point of each line.

  If there is an object O in which the edge E is not continuous (N in step S6), the discontinuous portion of the edge E is connected by, for example, interpolation, and the edge E is continued (step S11). Thereby, each object O partitions the inner area Ai, which is a closed area surrounded by the continuous edge E, and the outer area Ao.

  In the present embodiment, for each object O, it is determined whether or not the inner area Ai of the edge E has a predetermined size (step S7). When the inner area Ai is smaller than a predetermined size (N in Step S7), a process for deleting the object including the edge E is performed (N in Step S7). Specifically, the edge E is erased, that is, the color information of the pixels constituting the edge E is rewritten to non-edge color information. Further, the size (threshold value) of the inner region is determined so as not to affect the simulation calculation result. By performing this process, noise or carbon unnecessary for the calculation that could not be removed by the noise removal process (step S3) is removed, and it is useful for reducing the subsequent simulation calculation load and improving efficiency.

  Next, in this embodiment, for each object O, the inner region closed by the edge E is divided and defined as the carbon model Mc, and the outer region of the edge E is divided and defined as the rubber model Mr. Processing is performed (steps S8 and S9). In FIG. 5, the simulation model M generated by performing such processing is visualized.

  The predetermined closed space is divided by the small element e, that is, the meshing process can be performed using various application software. In the case of a planar element, the shape of the element e is preferably a triangle or a quadrangle. The size of each element e is determined in advance in accordance with the purpose of analysis and the like, and processed so as to be within the range.

  For each element e in the inner region Ai of the edge E, physical property values such as carbon density and elastic modulus are defined as parameters for calculation. Similarly, in the element e of the outer area Ao of the edge E, physical properties such as rubber density, elastic modulus and tan δ are defined as calculation parameters. The numerical data constituting the simulation model M is stored in the storage device of the computer device (step S10), and is called up when necessary to be used for simulation calculation.

  The simulation model M generated through the above processing is very close to the actual carbon contour shape and dispersion state. Therefore, the calculation result in the simulation becomes closer to the actual product, and a great improvement in simulation accuracy can be expected. In addition, the generation process of the simulation model M can be executed as a continuous automatic process by a computer for the most part (in this embodiment, for example, at least steps S3 to S10). be able to. Therefore, efficiency can be improved. In addition, for example, when meshing the outside of the edge E, for example, by changing the position without changing the shape of the edge that becomes the carbon model (for example, close contact with each other), the dispersion rate and the like can be easily achieved. Can be changed.

  In the present embodiment, the two-dimensional simulation model M is shown. However, the above processing may be repeated in the thickness direction of the carbon reinforced rubber, and a three-dimensional simulation model may be generated by supplementing between them. It becomes possible.

  By the way, how to handle the interface between carbon and rubber in an actual carbon reinforced rubber is an important problem in deformation simulation of the carbon reinforced rubber. As a result of various studies, slipping between rubber and carbon, friction, and a large energy loss associated therewith are becoming apparent at the interface. Therefore, in order to simulate a carbon reinforced rubber with high accuracy, it is desirable to incorporate such an interface as a model and give different physical property values thereto. Therefore, the present invention can further include a step of creating an interface model that models such an interface.

FIG. 6 shows a simplified example of a procedure for creating such an interface model.
In the present embodiment, the interface model is modeled as a film formed between rubber and carbon and having a small thickness. First, as shown in FIG. 6A, a contour line F offset from the edge E by a predetermined thickness t is set around the edge E of the carbon model Mc. Although this thickness t is not specifically limited, For example, about 1-10 nanometer is employable.

Next, as shown in FIG. 6B, the inside of the contour line F and the outside of the edge E are defined and meshed as the interface model Me. Further, the outside of the contour line F is defined as a rubber model Mr and meshed as in the above embodiment. For each element e constituting the interface model Me , it is desirable to define a viscoelastic characteristic having a hysteresis loss larger than, for example, the elastic modulus (complex elastic modulus) of rubber as a parameter for calculation. Thereby, a larger energy loss can be caused in the interface model Me. The inner region of the edge E is defined as a carbon model Mc and meshed as in the above embodiment. Thus, in this embodiment, it is possible to include a process of automatically setting the interface model Me around the carbon model Mc specified by the edge E.

  Although the embodiments of the present invention have been described above, the present invention can be used to generate a simulation model of a rubber composition or various elastomers using silica as a filler in addition to a carbon reinforced rubber.

It is a flowchart which shows embodiment of this invention. It is an example of the permeation | transmission cross-sectional image of a carbon reinforcement rubber. FIG. 3 is a raster data diagram obtained by binarizing FIG. 2. FIG. 4 is a raster data diagram obtained by extracting edges from FIG. 3. It is a diagram which visualizes and shows an example of a simulation model. (A) And (B) is the schematic explaining the step which sets an interface model. It is the diagram which visualizes and shows the conventional simulation model.

Explanation of symbols

Ai Inner region Ao Edge outer region E Edge M Simulation model Mc Carbon model Mr Rubber model Me Interface model

Claims (4)

A method for generating a simulation model of a filler compounded material in which a filler is dispersed and mixed in a matrix,
Obtaining an image containing the matrix and filler from the filler compounding material;
Extracting filler edges from the image;
After extracting the edge, setting a contour line offset by a predetermined thickness outside the edge; and
Dividing the inner region of the edge and defining as a filler model;
Dividing the outer region of the contour and defining as a matrix model ;
Dividing the region between the contour line and the edge and defining as an interface model having a physical property value different from the matrix model representing the interface between the matrix and the filler. A simulation model generation method for materials. The step of acquiring the image includes
Preparing a filler compounding material having a filler filling rate smaller than the filler filling rate to be analyzed;
The method of generating a simulation model for a filler compounding material according to claim 1, further comprising: obtaining a transmission cross-sectional image from the filler compounding material using a transmission electron microscope. Determining the continuity of the edges after extracting the edges;
3. The method for generating a simulation model of a filler compounding material according to claim 1, further comprising a step of connecting the discontinuous portions to make the edges continuous when the edges are not continuous. Determining the size of the inner region of the edge after extracting the edge;
The method for generating a simulation model for a filler compounding material according to any one of claims 1 to 3, further comprising the step of deleting all edges when the inner region is smaller than a predetermined size.

Images