JP5269830B2 - Model generation method, apparatus thereof, and program thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a model generation apparatus capable of removing an unnecessary filler and a fine and rough interface. <P>SOLUTION: A model generation apparatus 10 binarizes an input image to generate a binarized image, and after detecting an edge from the binarized image, connects a boundary point existing on the edge by at least one spline curve to generate a smoothed image, and sets an inner area of the edge in the smoothed image in a filler area and sets the other external area in a matrix area to generate a simulation image. <P>COPYRIGHT: (C)2012,JPO&amp;INPIT

Description

  The present invention relates to a model generation technique used for analyzing a deformation state and the like of a composition.

  The raw material of the tire rubber material is a composition in which a filler such as carbon black or silica is blended in a matrix. By blending this filler, the elastic modulus and tensile strength of the vulcanized rubber are greatly improved. However, in a matrix in which a filler is blended, the hysteresis in which the stress during unloading is lower than that in loading may be more noticeable than in a matrix not blended with a filler. Since this hysteresis has a close influence on the function of the rubber material, in order to investigate the deformation behavior of the rubber material in more detail, a simulation by a finite element method using a computer is performed.

  In order to perform this simulation, it is necessary to generate a model to be used by the finite element method from a matrix in which fillers are blended.

  Conventionally, as a method for generating this model, as shown in Patent Document 1, an image including a matrix and a filler is acquired from a composition, the edge of the filler is detected from the acquired image, and the internal region of the edge is divided. And a filler model, and the outer region of the edge is divided and defined as a matrix model.

JP 2006-193560 A

  However, in the model generation method in Patent Document 1, since an edge is directly detected from an acquired image, a large amount of small unnecessary fillers can be formed or a fine and rough interface can be formed. As a result, the probability of errors occurring during analysis is high, and fine meshes and meshes that are prone to element collapse are generated during mesh generation, increasing the number of meshes, and the time required for modeling, and thus simulation is completed. There is a problem that the overall time to do this also increases.

  Therefore, in view of the above problems, the present invention provides a model generation method, an apparatus thereof, and a program thereof that can remove unnecessary fillers and fine and rough interfaces.

  The present invention includes an input step of inputting an input image of a composition containing a matrix and a filler, a binarization step of binarizing the input image to generate a binarized image, and an edge from the binarized image A detection step for generating an edge image that detects the image, a setting step for setting a plurality of boundary points on the edge in the edge image, and at least one smoothing curve represented by a polynomial expression for the plurality of boundary points. A smoothing step for generating a smoothed image formed by connecting and forming a smoothed edge, and an inner region surrounded by the smoothed edge in the smoothed image is set as a filler region and a mesh shape including a plurality of elements Divide and set the area other than the inner area as a matrix area and divide it into a mesh consisting of a plurality of elements for simulation. An area setting step of generating a Dell, a model generating method characterized by and an output step of outputting the model.

  According to the present invention, unnecessary fillers and fine and rough interfaces can be removed.

It is a block diagram of the model production | generation apparatus which shows the Example of this invention. It is a flowchart which similarly shows operation | movement of a model production | generation apparatus. It is explanatory drawing for smoothing using a spline curve. It is a figure of a binarized image. It is a figure of a blurred image. It is a figure of an edge image. FIG.

  Hereinafter, a model generation apparatus 10 according to an embodiment of the present invention will be described with reference to FIGS.

  The model generation apparatus 10 of the present embodiment is an apparatus that generates a model for simulation used in the finite element method from a matrix (for example, carbon reinforced rubber) in which a filler is blended.

  Here, the “filler” is only required to have a component that can be uniformly dispersed as fine particles or regions having a size that can be observed with a microscope or the like and an identifiable interface in the rubber composition. In addition to the suitably used fillers, the domains of each rubber component in the rubber blend can also be modeled as “fillers”.

  Fillers other than carbon black are not particularly limited, inorganic oxides such as silica and calcium carbonate, natural mineral particles such as clay and mica, carbides such as polymer gel and bamboo charcoal, synthetic resin particles such as polyethylene and phenol resin, Alternatively, natural products such as walnuts and lignin are pulverized. The rubber blend that can be modeled in the present invention is not particularly limited, but diene rubbers such as natural rubber (NR), isoprene rubber (IR), butadiene rubber (BR), styrene butadiene rubber (SBR), butyl rubber (IIR). And two or more blends of either halogenated butyl rubber (XIIR) are particularly preferred.

  These compositions usually contain multiple rubber components and / or additives, such as natural rubber / butadiene rubber / carbon / polymer gel compositions, depending on the size of each component, the dispersion conditions, and the analytical purpose. Thus, the component to be modeled as “filler” can be selected as appropriate.

  For example, in the case of the above composition, when the carbon is sufficiently smaller than the polymer gel and the carbon is dispersed only in the rubber region, the polymer gel is selected as a filler, and the remaining rubber component and carbon are used as a matrix. Model. Conversely, when the polymer gel is smaller, carbon may be selected as the filler, and the rubber component and the polymer gel may be modeled as a matrix. Furthermore, when it is desired to analyze the dispersion state of rubber components, one of natural rubber or butadiene rubber may be selected as a filler instead of carbon or polymer gel, and the other may be modeled as a matrix together with carbon and polymer gel.

  Here, the “finite element method” is a numerical analysis method in which a surface or space closed by a finite boundary is divided into finite partial areas, and the object generated by the mesh division is a small and simple element. This is a technique for obtaining an approximate value of the overall behavior by analyzing each element by dividing it into each element.

(1) Configuration of Model Generation Device 10 The configuration of the model generation device 10 will be described with reference to FIG. FIG. 1 is a block diagram of the model generation apparatus 10.

  As shown in FIG. 1, the model generation apparatus 10 includes an input unit 12, a binarization unit 14, a blurring unit 16, an edge detection unit 18, a point setting unit 20, a smoothing unit 22, a region setting unit 24, and an output unit 26. have.

  Note that the model generation device 10 can also be realized by using, for example, a general-purpose computer as basic hardware. That is, the input unit 12, the blurring unit 16, the binarization unit 14, the edge detection unit 18, the point setting unit 20, the smoothing unit 22, the region setting unit 24, and the output unit 26 are included in a processor mounted on the computer. This can be realized by executing a program. At this time, the model generation apparatus 10 may be realized by installing the above program in a computer in advance, or may be stored in a storage medium such as a CD-ROM or distributed through the network. Thus, this program may be realized by appropriately installing it in a computer.

  Hereinafter, functions of the respective units 12 to 26 will be described in order.

(2) Input unit 12
First, the input unit 12 acquires an input image of a composition including a matrix and a filler. For example, a transmission cross-sectional image obtained by photographing a carbon reinforced rubber as a composition with a transmission electron microscope (hereinafter referred to as “TEM”) is input as an input image. The TEM takes a transmission cross-sectional image of a thickness portion of about 100 nm from a carbon reinforced rubber in which carbon is dispersedly arranged (three-dimensionally arranged) in a matrix. In general, a transmission cross-sectional image is taken at 100 times or less the size of the target filler, but at this size, it is often not uniformly dispersed, and the filler area ratio in the image depends on the shooting location. fluctuate. For this reason, it is necessary to select an image having an area ratio comparable to the actual filler volume ratio. For example, in the case of carbon reinforced rubber having a volume ratio of 40%, it is preferable to model by selecting an image having an area ratio of 36 to 44%.

  However, this input image acquisition method may be any method as long as the carbon outline is clearly obtained as an image.

  Next, the input image acquired by the input unit 12 may be directly taken in as digital data from the TEM. However, after taking a picture, all or part of the input image is read by raster scanning with an image scanner or the like, and a necessary area is read. A method of generating a digital input image by digitizing may be used.

(3) Binarization unit 14
The binarization unit 14 binarizes the digital input image based on a predetermined reference value, and generates a binarized image (see FIG. 4).

(4) Blur part 16
The blurring unit 16 performs a blurring process on the binarized image to generate a fool image (see FIG. 5). By performing this blurring process, small and unnecessary fillers and fine and rough portions of the interface can be removed.

  As a method of blurring a binarized image, there are a method of processing by a Gaussian filter, or a method of blurring a binarized image by performing Fourier transform to remove high frequency components and then performing inverse Fourier transform. When processing with a Gaussian filter, the radius blur width of the filter is set to about the size of the filler determined to be unnecessary in advance, and when removing high frequency components by Fourier transform, it corresponds to the size of the filler determined to be unnecessary in advance. It is preferable to remove more than the frequency.

(5) Edge detector 18
The edge detection unit 18 detects an edge shown in the blurred image and generates an edge image (see FIG. 6). This edge becomes an interface surrounding the filler region. As an edge detection method, for example, a portion where the luminance of each pixel of the binarized blurred image changes more than the luminance of the surrounding pixels and a threshold is detected as an edge.

(6) Point setting unit 20
The point setting unit 20 sets a plurality of boundary points on the edge in the edge image.

  The setting of this boundary point is important. If the interval between the boundary points is wide, the shape of the filler region becomes too smooth. Conversely, if the interval between the boundary points is narrow, the shape of the filler region becomes too fine. Therefore, as a method of setting the boundary point, for example, the operator may set the boundary point artificially with a mouse or the like, or the boundary point may be generated at every predetermined distance on the edge. As a rule to set, for example, when an input image of carbon reinforced rubber having a volume ratio of 40% is used for modeling, the boundary is set so that the area ratio after the point setting is within a range of about 36 to 44%. It is preferable to set points.

(7) Smoothing unit 22
The smoothing unit 22 generates a smoothed image in which a smoothed edge is formed by connecting a plurality of boundary points on the edge of the edge image with at least one smoothing curve represented by a polynomial.

  The smoothing curve is not particularly limited as long as it is differentiable, but any one of an nth-order spline curve, a B-spline curve, a Bezier curve, or a curve generated by polynomial interpolation so as to pass through a specified boundary point is used. It is preferable. When this n-order spline curve is used, as shown in FIG. 3, a smoothing edge is generated by connecting a plurality of boundary points existing in a certain section, and an n-order spline curve is generated for each section. Generate. At this time, n represents an arbitrary numerical value among natural numbers of n = 1, 2, 3,. In addition, when a curve generated by polynomial interpolation is used, it is preferable to interpolate using Newton interpolation, Lagrange interpolation, or the like.

(8) Area setting unit 24
The region setting unit 24 sets the inner region of the smoothed edge in the smoothed image as a filler region and divides it into a mesh shape composed of a plurality of elements used in the finite element method. Further, the outer region is set as a matrix region and divided into a mesh shape composed of a plurality of elements that are also used in the finite element method to generate a model (see FIG. 7). At this time, other information is added for each region in order to perform a simulation by the finite element method. As information to be added, for example, information on elastic modulus and viscosity is added to the matrix region.

  The filler area may be set by an operator specifying an area on the smoothed image with a mouse or the like, or the area setting unit 24 determines a closed area, and the determined area is set as a filler. It may be set as an area. Then, the area setting unit 24 sets a matrix area other than the filler area. Conversely, the matrix area may be set first, and the other area may be set as the filler area.

  The shape of each element divided into meshes is preferably, for example, a triangle or a quadrangle. The size of each element is determined in advance in accordance with the purpose of simulation analysis and the like, and is processed so as to be within the range.

(9) Output unit 26
The output unit 26 outputs the model in which the filler region and the matrix region are set by the region setting unit 24 to a simulation apparatus that performs a finite element method analysis.

(10) Operation State of Model Generation Device 10 Next, the operation state of the model generation device 10 will be described with reference to FIG. FIG. 2 is a flowchart of the model generation apparatus 10.

  In step 1, the input unit 12 acquires a transmission cross-sectional image from the TEM as an input image, and then digitizes it to generate a digital input image.

  In step 2, the binarization unit 14 binarizes the digital input image to generate a binarized image.

  In step 3, the blurring unit 16 blurs the binarized image to generate a blur image.

  In step 4, the edge detection unit 18 detects an edge on the blurred image and generates an edge image.

  In step 5, the point setting unit 20 sets a plurality of boundary points on the edge in the edge image.

  In step 6, the smoothing unit 22 generates a smoothed image by connecting and smoothing a plurality of boundary points in the edge image with at least one smoothing curve represented by a polynomial.

  In step 7, the region setting unit 24 sets a filler region and a matrix region for the smoothed image, and divides each into a mesh shape composed of a plurality of elements used in the finite element method to generate a model.

  In step 8, the output unit 26 outputs the model.

(11) Effect According to the present embodiment, since the blurred image is generated after binarization, it is possible to remove small and unnecessary fillers and finely and coarsely, compared to the case where the processing is performed by detecting the edge as it is conventionally. The interface can be removed. Therefore, the model does not become complicated and mesh crushing hardly occurs in the analysis by the finite element method.

  Further, since smoothing is performed using a spline function, generation of a fine and rough interface can be prevented, setting of the filler region and the matrix region becomes complicated, and mesh crushing hardly occurs in the analysis by the finite element method.

  Furthermore, since the small unnecessary filler and the fine rough interface are removed, the time until the model is generated can be shortened.

(12) Modification Examples The present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the gist thereof.

  In the above-described embodiment, blurring is performed after binarization, but binarization may be performed after blurring instead.

  In the above embodiment, the spline function is used as the smoothing process, but smoothing may be performed using other functions.

  In the above embodiment, both the blurring process and the smoothing process are performed, but only the smoothing process may be performed.

(13) Experimental Results Experimental results in the above examples, modified examples, and comparative examples 1 and 2 are described.

  In this embodiment, as described above, both the blurring process and the smoothing process are performed.

  Further, as described above, the modification example is a case where only the smoothing process is performed without performing the blurring process.

  Comparative Example 1 is a case where only the blurring process is performed on the binarized image and the smoothing process is not performed.

  Moreover, the comparative example 2 is a case where neither a blurring process nor a smoothing process is performed.

  When the model generated from the model generation apparatus of the above-described embodiment, modified example, and comparative examples 1 and 2 is analyzed using the finite element method, the analysis error rate is almost zero in the example and modified example. In Comparative Examples 1 and 2, an error occurred. Here, the “analysis error rate” is a probability that when a model is analyzed using the finite element method, mesh collapse occurs and analysis becomes impossible.

DESCRIPTION OF SYMBOLS 10 Model generator 12 Input part 14 Binarization part 16 Blur part 18 Edge detection part 20 Point setting part 22 Smoothing part 24 Area setting part 26 Output part

Claims (5)

An input step for inputting an input image of a composition including a matrix and a filler;
A binarization step of binarizing the input image to generate a binarized image;
A detection step of generating an edge image in which an edge is detected from the binarized image;
A setting step of setting a plurality of boundary points on the edge in the edge image;
A smoothing step of generating a smoothed image in which the plurality of boundary points are connected to each other by at least one smoothing curve represented by a polynomial to form a smoothed edge;
An inner region surrounded by the smoothed edge in the smoothed image is set as a filler region and divided into a mesh shape composed of a plurality of elements, and other than the inner region is set as a matrix region and composed of a plurality of elements. An area setting step for generating a model for simulation by dividing into meshes,
An output step of outputting the model;
A model generation method characterized by comprising: The smoothing curve is either an nth-order spline curve, a B-spline curve, a Bezier curve, or a curve generated by polynomial interpolation so as to pass through designated boundary points.
The model generation method according to claim 1. A blur step for blurring the binarized image;
In the detection step, an edge image is generated from the blurred binary image.
The model generation method according to claim 1. An input unit for inputting an input image of a composition containing a matrix and a filler;
A binarization unit that binarizes the input image to generate a binarized image;
A detection unit that generates an edge image in which an edge is detected from the binarized image;
A setting unit for setting a plurality of boundary points on the edge in the edge image;
A smoothing unit that generates a smoothed image by connecting the plurality of boundary points with at least one smoothing curve represented by a polynomial to form a smoothed edge;
An inner region surrounded by the smoothed edge in the smoothed image is set as a filler region and divided into a mesh shape composed of a plurality of elements, and other than the inner region is set as a matrix region and composed of a plurality of elements. An area setting unit that generates a model for simulation by dividing into meshes,
An output unit for outputting the model;
A model generation apparatus characterized by comprising: On the computer,
An input function for inputting an input image of a composition including a matrix and a filler;
A binarization function for binarizing the input image to generate a binarized image;
A detection function for generating an edge image in which an edge is detected from the binarized image;
A setting function for setting a plurality of boundary points on the edge in the edge image;
A smoothing function for generating a smoothed image in which the plurality of boundary points are connected to each other by at least one smoothing curve represented by a polynomial to form a smoothed edge;
An inner region surrounded by the smoothed edge in the smoothed image is set as a filler region and divided into a mesh shape composed of a plurality of elements, and other than the inner region is set as a matrix region and composed of a plurality of elements. A region setting function that generates a model for simulation by dividing into meshes,
An output function for outputting the model;
Model generation program for realizing

Images