CN103971395A - Mimicry reconstruction and performance computing method of fiber filter medium structure - Google Patents

Abstract

The invention discloses a mimetic reconstruction of a fiber filter medium structure and a performance calculation method thereof, belonging to the technical field of fibrous structure mimetic reconstruction. The mimetic reconstruction process of the present invention is as follows: 1. Obtain the image of the fiber filter medium; 2. Set the optimal threshold and perform binary processing; 3. Perform centerline detection on the binary image; Four diagonal directions are compared to obtain the rotation radius di and the rotation direction df; fifth, three-dimensional rotation is performed to obtain the mimic reconstruction image of the fiber filter medium. The performance calculation method of the present invention obtains the filtration efficiency curve and the pressure loss curve of the fiber filter medium at different wind speeds by inputting a series of relevant parameters and utilizing empirical formulas. The invention is based on the internal microscopic single-layer structure image of the real fiber filter medium, and uses a simple and fast image processing method to perform mimic reconstruction on the obtained image, and the mimic reconstruction model can realistically simulate the structure of the real fiber filter medium.

Description

Mimic reconstruction of fiber filter media structure and its performance calculation method

technical field

The invention relates to the technical field of fibrous structure mimic reconstruction, and more specifically, relates to a mimetic reconstruction of a fiber filter medium structure and a performance calculation method thereof.

Background technique

Dust will cause harmful effects on the atmospheric environment, production and human health, especially the inhalable particulate matter will cause harm to the human respiratory system. Therefore, reducing suspended particulate matter in the air is of great significance to protecting the environment and safeguarding human health.

Faced with the increasingly serious problem of air pollution, the bag filter has a high efficiency of collecting particulate pollutants, so it has been widely used. As one of the core components of the bag filter, fiber filter media can meet the requirements of national standards for the filtration of fine particles, which is closely related to its filtration efficiency and pressure loss, especially the pressure loss of fiber filter media is an important indicator of filter performance. Parameters, its size is related to the energy consumption of the entire filtration system. The above-mentioned performance indicators of the filter medium are mainly determined by the internal microstructure of the fiber filter medium. Therefore, it is very important to accurately obtain the internal microstructure of filter media.

In the early research, the fiber filter medium was simplified as a single fiber or isolated fiber. On the basis of a single fiber, foreign scientists considered the interference of the surrounding fibers to establish a fiber model, and simplified the fiber filter medium to an ideal filter medium, which was generated by computer technology. Two-dimensional woven (fiber parallel and cross-arranged) filter media. However, the fiber filter medium established by the above method has a two-dimensional structure, which is quite different from the actual three-dimensional filter medium structure. The mimetic model of the two-dimensional filter medium established cannot truly reflect the performance of the filter medium and can only be used for research. simulation.

There are also researchers who use computer technology to randomly generate three-dimensional fiber filter media in space, and consider factors such as fiber diameter, length, direction, and bending. However, the model constructed only considering the factors of the fiber part is not based on the actual fiber filter media, and factors such as extrusion deformation between fibers are not considered. Therefore, there is still a large difference between the constructed model and the actual filter medium structure, and it can only be used for research and simulation, but the structural parameters of the real fiber filter medium cannot be obtained.

After searching, the technical solution for the research on the performance of fiber materials has been published, such as Chinese patent application number 201210268547.0, the application date is July 30, 2012, and the invention name is: quantitative evaluation method for fiber orientation degree of short fiber reinforced composite materials; The application relates to a method for quantitatively evaluating the fiber orientation degree of short-fiber reinforced composite materials, which is characterized by: selecting and determining the section plane and sectioning the composite material sample; obtaining micrographs through optical microscope or scanning electron microscope; Redraw the elliptical fiber section in the photomicrograph; extract the length of the major axis, the length of the minor axis and the angle between the major axis of the ellipse and the coordinate axis of each ellipse in the redrawn picture; calculate each ellipse according to the extracted ellipse section parameters The direction vector of the corresponding fiber; calculates the degree of orientation of the composite material along a specific direction parameter. This application eliminates the calculation error caused by the same elliptical cross-section corresponding to two fiber directions by selecting an appropriate section plane. It is mainly suitable for quantitative evaluation of fiber orientation degree, and is not suitable for mimetic reconstruction of fiber filter media structure.

Contents of the invention

1. The technical problem to be solved by the invention

The purpose of the present invention is to overcome existing fibrous filter medium mimic reconstruction method to exist: 1) do not take actual fiber filter medium as the basis of mimic reconstruction, the fiber model that builds and obtains is quite different from the actual filter medium structure, can only be used for research Simulation has limitations in application; 2) the mimic reconstruction process is complex, the amount of data processing is large, and the model construction time is long. A kind of mimic reconstruction of fiber filter medium structure and its performance calculation method are provided; the present invention uses real fiber Based on the microscopic single-layer structure image inside the filter medium, use simple and fast image processing methods to perform mimic reconstruction on the obtained image, and extract image information to obtain the structural parameters of the mimic reconstruction model. The mimic reconstruction model can realistically simulate real fibers The structure of the filter medium, the obtained structural parameters can represent the structural parameters of the real fiber filter medium.

2. Technical solution

In order to achieve the above object, the technical scheme provided by the invention is:

A mimetic reconstruction method of a fiber filter medium structure of the present invention, the steps of which are:

Step 1, obtaining an image of a real fiber filter medium;

Step 2. Set the optimal threshold, perform binary processing on the image obtained in step 1, and extract the binary image of the fiber filter medium and pores, the fiber pixel value is 1, and the pore pixel value is 0;

Step 3: Perform fiber filter medium centerline detection on the binary image obtained in step 2 to obtain the fiber filter medium centerline ss;

Step 4, taking the pixel point on the center line obtained in step 3 as the base point to compare the four diagonal directions, and obtain the three-dimensional rotation radius di and the three-dimensional rotation direction df of each pixel point on the center line;

Step 5. Taking the pixel point on the center line obtained in Step 3 as the base point of rotation, and using the 3D rotation radius di and 3D rotation direction df obtained in Step 4 as rotation parameters, three-dimensional rotation is performed to obtain a mimic reconstruction image of the fiber filter medium.

Furthermore, in step 1, the real fiber filter medium is scanned under a scanning electron microscope to obtain a scanning electron microscope image of the fiber filter medium.

Furthermore, the determination process of the optimal threshold described in step 2 is as follows:

a. Gray-scale stretching of the scanning electron microscope image of the fiber filter medium:

K=255/(b-a)

M=(-255*a)/(b-a)

Y=KX+M

In the above formula, a is the minimum grayscale value of the SEM image; b is the maximum grayscale value of the SEM image; X is the grayscale value of any pixel in the SEM image; Y is the grayscale stretch value corresponding to the pixel point X;

b. Determine the optimal threshold:

Assume that the ratio of the number of pixels in the target range to the entire image in the image after grayscale stretching in step a is ω0, the average gray level of pixels in the target range is μ0, and the number of pixels in the background range accounts for the entire image The proportion of ω1 is ω1, the average gray value of pixels within the background range is μ1, the segmentation threshold of the target and the background is denoted as T _m , and the inter-class variance between the target and the background is denoted as g, then:

g=ω0ω1(μ0-μ1)^2

The traversal method is used to obtain the corresponding threshold T _m when the inter-class variance g is maximized, and this threshold is the optimal threshold.

Furthermore, the method for obtaining the centerline ss of the fiber filter medium in step three is: take the pixel point with a pixel value of 1 in the binary image obtained in step two as the center and extend outward in 8 directions, and the 8 directions described above are At 45° intervals, count the number of pixels with a continuous pixel value of 1 in 8 directions. If the number of pixels counted in any two diagonal directions of the pixel is equal or the difference is 1, then the pixel is retained. Otherwise, it is discarded; the above operation is performed on all pixels with a pixel value of 1, and the centerline ss of the fiber filter medium is obtained.

Furthermore, the comparison of the four diagonal directions mentioned in step four is specifically: for the pixels retained in step three, compare the number of pixels with a continuous pixel value of 1 in the four diagonal directions, and compare the number of pixels with the least number of pixels The angular direction is taken as the three-dimensional rotation direction df, and half of the number of pixels in the diagonal direction is taken as the three-dimensional rotation radius di.

The performance calculation method of a kind of fiber filter medium structure of the present invention, its steps are:

Step A, obtaining the image of the real fiber filter medium;

Step B, setting the optimal threshold, performing binary processing on the image obtained in step A, extracting the binary image of the fiber filter medium and pores, the fiber pixel value is 1, and the pore pixel value is 0;

Step C, performing fiber filter medium centerline detection on the binary image obtained in step B, to obtain the fiber filter medium centerline ss;

Step D, taking the pixel points on the center line obtained in step C as the base point to compare the four diagonal directions, and obtain the three-dimensional rotation radius di and the three-dimensional rotation direction df of each pixel point on the center line;

Step E, inverting the binary image obtained in step B, the fiber pixel value is 0, the pore pixel value is 1, and the number of pixel points with a statistical pixel value of 1 is denoted as a; the statistical step A obtains The number of pixels in the real fiber filter medium image, denoted as b, obtains the fiber filter medium image porosity a/b; counts the number of pixel point regions whose pixel value is 1, and obtains the fiber pore histogram;

Step F, performing statistics on the three-dimensional radius of rotation di obtained in step D, and obtaining a statistical diagram of fiber diameter and an average fiber diameter _df ;

Step G, substituting the average fiber diameter d _f obtained in step F and the absolute temperature T of the filtered fluid, the thickness of the filter material t, the head-on wind speed V and the particle diameter d _p , apply the empirical formula to obtain the filtration efficiency curve and pressure loss curve.

Further, the determination process of the optimal threshold described in step B is as follows:

a. Gray-scale stretching of the scanning electron microscope image of the fiber filter medium:

K=255/(b-a)

M=(-255*a)/(b-a)

Y=KX+M

In the above formula, a is the minimum grayscale value of the SEM image; b is the maximum grayscale value of the SEM image; X is the grayscale value of any pixel in the SEM image; Y is the grayscale stretch value corresponding to the pixel point X;

b. Determine the optimal threshold:

Assume that the ratio of the number of pixels in the target range to the entire image in the image after grayscale stretching in step a is ω0, the average gray level of pixels in the target range is μ0, and the number of pixels in the background range accounts for the entire image The proportion of ω1 is ω1, the average gray value of pixels within the background range is μ1, the segmentation threshold of the target and the background is denoted as T _m , and the inter-class variance between the target and the background is denoted as g, then:

g=ω0ω1(μ0-μ1)^2

The traversal method is used to obtain the corresponding threshold T _m when the inter-class variance g is maximized, and this threshold is the optimal threshold.

Further, the empirical formula described in step G is:

Empirical formula for pressure loss:

$\mathrm{<span\; class="notranslate">\; \&Delta;p</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>\frac{\mathrm{<span\; class="notranslate">\; \&mu;tV</span>}}{{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; f</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}$

in: $f\left(\mathrm{\&alpha;}\right)=\frac{16\mathrm{\&alpha;}}{-0.5\ln \mathrm{\&alpha;}-0.5\frac{1-{\mathrm{\&alpha;}}^2}{1+{\mathrm{\&alpha;}}^2}};$ $f\left(\mathrm{\&alpha;}\right)=\frac{16\mathrm{\&alpha;}}{\mathrm{Ku}},$ Ku=-0.5lnα-0.75+α-0.25α ² ; or f(α)=64α ^3/2 (1+56α ³ ), μ is the air viscosity coefficient, which is a function of the absolute temperature T of the filtered fluid, and α is the solid volume Fraction;

Filtration efficiency empirical formula:

$\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; (</span>\frac{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 4</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}{\mathrm{<span\; class="notranslate">\; E.</span>}}_{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; t</span>}}{\mathrm{<span\; class="notranslate">\; \&pi;</span>}{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span>$

Among them, E _Σ ＝1-(1-E _R )(1-E _I )(1-E _D ), ${\mathrm{E.}}_{\mathrm{D.}}=1.6{\left(\frac{1-\mathrm{\&alpha;}}{\mathrm{Ku}}\right)}^{1/3}{\mathrm{Pe}}^{-2/3},$ ${\mathrm{E.}}_R=0.6\frac{1-\mathrm{\&alpha;}}{\mathrm{Ku}}\frac{R^2}{(1+R)},$ ${\mathrm{E.}}_I=\frac{J\&times;\mathrm{Stk}}{4K{u}^2},$ In the above formula, Ku=-0.5lnα-0.75+α-0.25α ² , $P_e=\frac{{\mathrm{Vd}}_f}{\mathrm{D.}},$ $\mathrm{D.}=\frac{\mathrm{\&sigma;}{C}_{c}T}{3\mathrm{\&pi;\&mu;}{d}_p},$ C _c is the Cunningham slip correction factor, σ is the Boltzmann constant; J=(29.6-28α ^0.62 )R-27.5R ^2.8 , $\mathrm{Stk}=\frac{{\mathrm{\&rho;}}_p{d}_p^2{C}_{c}V}{18\mathrm{\&mu;}{d}_f}.$

Furthermore, the absolute temperature T of the filtered fluid in step G is 300-1000 K, the thickness t of the filter material is 0-1000 μm, the head-on wind speed V1 is 0-2 m/s, V2 is 0-2 m/s, and V3 is 0-2 m/s. 2m/s, V4 is 0~2m/s, V5 is 0~2m/s, particle diameter _dp is 0~5μm.

3. Beneficial effect

Compared with the existing known technology, the technical solution provided by the invention has the following remarkable effects:

(1) A mimetic reconstruction method of a fiber filter medium structure of the present invention performs binarization processing on the scanning electron microscope image of the actual fiber filter medium, and through the determination of the optimum threshold, the fiber pixels and pore pixels are effectively separated , which laid a solid foundation for the subsequent accurate determination of the centerline of the fiber filter medium, the three-dimensional rotation radius di, and the three-dimensional rotation direction df. The mimic reconstruction model can realistically simulate the structure of the real fiber filter medium, overcoming the application limitations of traditional methods ;

(2) The performance calculation method of a kind of fiber filter medium structure of the present invention, based on real fiber filter medium image, input a series of relevant parameters, by the method for numerical simulation, fiber filter medium is studied, obtain the performance parameter of real fiber filter medium , does not need to conduct real experiments, can save a lot of time and manpower and material resources, and has high practical value.

Description of drawings

Fig. 1 is the scanning electron micrograph of real fiber filter medium of the present invention;

Fig. 2 is the image after the binary processing of the real fiber filter medium of the present invention;

Fig. 3 is the statistical histogram of fiber diameter of the present invention;

Fig. 4 is the porosity histogram of the present invention;

Fig. 5 is the fibrous filter medium structure mimic reconstruction figure of the present invention;

Fig. 6 is the curve diagram of pressure loss under different head-on wind speeds among the present invention;

Fig. 7 is a graph of filtration efficiency under different head wind speeds in the present invention.

Detailed ways

In order to further understand the content of the present invention, the present invention will be described in detail in conjunction with the accompanying drawings and embodiments.

Example 1

In conjunction with the accompanying drawings, a mimetic reconstruction method of a fiber filter medium structure of the present embodiment, the steps are:

Step 1. Scan the real fiber filter medium under a scanning electron microscope to obtain a scanning electron microscope image of the fiber filter medium. Refer to FIG. 1 for the scanning electron microscope image of the fiber filter medium obtained in this embodiment.

Step 2: Set the optimal threshold, perform binarization on the image obtained in Step 1, and extract the binary image of the fiber filter medium and pores, where the fiber pixel value is 1 and the pore pixel value is 0. The determination process of the optimal threshold is as follows:

a. Gray-scale stretching of the scanning electron microscope image of the fiber filter medium:

K=255/(b-a)

M=(-255*a)/(b-a)

Y=KX+M

In the above formula, a is the minimum grayscale value of the SEM image; b is the maximum grayscale value of the SEM image; X is the grayscale value of any pixel in the SEM image; Y is the grayscale stretch value corresponding to the pixel point X.

b. Determine the optimal threshold:

The stretched image is divided into two parts: background and target, the target is the fiber medium, and the background is the pores between the fiber medium. The greater the inter-class variance between the background and the target, the greater the difference between the two parts that make up the image. When part of the target is misclassified as the background or part of the background is misclassified as the target, the difference between the two parts will become smaller. Therefore, the split that maximizes the between-class variance means the smallest probability of misclassification. In this embodiment, it is assumed that the resolution of the image after the grayscale stretching in step a is M×N, the segmentation threshold of the target and the background is recorded as T _m , the ratio of the number of pixels in the target range to the entire image is ω0, and the target range The average gray value of the pixels in the background is μ0, the ratio of the number of pixels in the background to the entire image is ω1, the average gray value of the pixels in the background is μ1, and the variance between objects and backgrounds is denoted as g. The average value of the total gray level is μ. The number of pixels whose gray value is less than the threshold _Tm in the image is denoted as N0, and the number of pixels whose gray value is greater than the threshold _Tm is denoted as N1, then:

ω0＝N0/M×N (1)

ω1=N1/M×N (2)

N0+N1=M×N (3)

ω0+ω1=1 (4)

μ＝ω0*μ0+ω1*μ1 (5)

g＝ω0(μ0-μ)^2+ω1(μ1-μ)^2 (6)

Substituting formula (5) into formula (6), the equivalent formula is obtained:

g＝ω0ω1(μ0-μ1)^2 (7)

The traversal method is used to obtain the corresponding threshold T _m when the inter-class variance g is maximized, and this threshold is the optimal threshold.

It is worth noting that the binarization processing of the real fiber filter media image in this embodiment can be said to be the basis for the mimetic reconstruction of the entire fiber filter media structure. Whether the binarization processing can accurately distinguish between fiber pixels and pore pixels, Determines the accuracy of the mimic reconstruction model, and the key to binarization lies in the determination of the optimal threshold. The method for obtaining the optimal threshold value in this embodiment is especially suitable for images with smooth gray-scale transformations. Since the gray-scale transformation of the scanning electron microscope image of the fiber filter medium is smooth, the results obtained by using the method of this embodiment can fully meet the requirements of mimetic reconstruction after testing. Accuracy requirements, and the above method is simple, the amount of data processing is small, which greatly reduces the complexity of the mimic reconstruction process. After the optimal threshold is determined, extract the pixel gray value of the scanning electron microscope image of the real filter medium and compare it with the optimal threshold. Pixels whose degree value is higher than the threshold are classified as fiber pixels, and the pixel value is recorded as 1, and a binary image is obtained (see Figure 2).

Step 3: Perform fiber filter medium centerline detection on the binary image obtained in step 2 to obtain the fiber filter medium centerline ss. The specific process is: take any pixel point value of 1 in the binary image obtained in step 2 as the center and extend outward in 8 directions, and the 8 directions described in this embodiment are respectively 0° directions (parallel to the horizontal plane) Right), counterclockwise rotation 45° direction, 90° direction, 135° direction, 180° direction, 225° direction, 270° direction and 315° direction. Count the number of pixels with consecutive pixel values of 1 in 8 directions (for example, starting from the central pixel, count the number of pixels with consecutive pixel values of 1 in the 0° direction, and the pixel with a pixel value of 0 Then end the statistics), if the number of pixel points counted in any two diagonal directions of the pixel point is equal or the difference is 1, then keep this pixel point, otherwise discard it. The above operation is performed on all pixel points whose pixel value is 1, and the centerline ss of the fiber filter medium is obtained.

Step 4: For the pixels retained in step 3, compare the number of pixels with a continuous pixel value of 1 in the 4 diagonal directions, and use the diagonal direction with the least number of pixels as the three-dimensional rotation direction df. One-half of the number of pixels is used as the three-dimensional rotation radius di.

Step 5. Taking the pixel point on the center line obtained in Step 3 as the base point of rotation, and using the 3D rotation radius di and 3D rotation direction df obtained in Step 4 as rotation parameters, three-dimensional rotation is performed to obtain a mimic reconstruction image of the fiber filter medium. The specific process is: select a pixel point on the center line, take the pixel point as the center, take di as the rotation radius, and df as the rotation direction to perform three-dimensional rotation to obtain a circular surface with the pixel point as the center of the circle. All the pixel points perform the above operations, and the three-dimensional reconstructed image of the fiber filter medium can be obtained (refer to FIG. 5 ).

It is worth noting that, since the above-mentioned 3D reconstruction process calculates a large amount of data and takes a long time, in order to reduce the model construction time and the amount of data processing, this embodiment adopts an equidistant step between the reconstructed circular surfaces to superimpose the circular surfaces. That is, one pixel point is extracted at equal intervals of 5 to 15 pixels on the center line of the extracted fiber filter medium for three-dimensional circular surface reconstruction, and the rest of the pixel points are superimposed according to the reconstructed circular surface. The general process is:

Use the planar Cartesian coordinate system to divide the scanning electron microscope image of the fiber filter medium into m parts in the X direction, and divide it into n parts in the Y direction, and draw a straight line parallel to the coordinate axis from each division point, and divide the scanning electron microscope image into m×n A small rectangle to calculate the function value of the dot. Then import the previously obtained 3D rotation radius di, 3D rotation direction df, and centerline pixel coordinate information into the grid that has been drawn. Finally, the overall 3D reconstruction model is obtained by reconstructing the pixels extracted at equal intervals. . In order to reduce program running time, save computer memory, and reduce computer configuration requirements, the optimal number of grid divisions is n/15~n/5 and m/15~m/5. In this embodiment, the grid X interval is divided into previous n/10 parts, and the Y interval is divided into previous m/10 parts. This embodiment has determined the best interval division through multiple experiments. Through the above processing, data processing The volume is reduced by 100 times compared with the original one, the memory occupied by the processing data is reduced from 5G to 80M in the traditional method, the model construction time is greatly shortened, and the constructed fiber medium structure can fully meet the requirements of mimic reconstruction.

In the method for calculating the performance of a fiber filter medium structure in this embodiment, steps A to D are the same as steps 1 to 4 in the process of mimetic reconstruction, and will not be repeated here.

Step E, perform inversion processing on the binary image obtained in step B, the fiber pixel value is 0, the pore pixel value is 1, and the number of pixel points with a pixel value of 1 is counted, denoted as a. The number of pixels of the real fiber filter medium image obtained in step A is counted, denoted as b, and the porosity a/b of the fiber filter medium image is obtained. Count the number of regions containing only pixels with a pixel value of 1 to obtain a fiber pore histogram, as shown in Figure 4, where the abscissa in Figure 4 represents the number of pixels with a pixel value of 1 contained in the region, The ordinate represents the number of regions containing x pixels, and the number of regions is represented by a logarithm with base 10.

Step F, perform statistics on the three-dimensional radius of rotation di obtained in step D, and obtain a statistical diagram of fiber diameter (see FIG. 3 ) and an average fiber diameter d _f .

Step G, substituting the average fiber diameter d _f obtained in step F, the absolute temperature of the filter fluid, the thickness of the filter material t, the head-on wind speed V and the diameter of the particles d _p , and using the empirical formula to obtain the different wind speeds of the fiber filter medium in the dusty state and clean state Filtration efficiency curve and pressure loss curve. The empirical formula used in this embodiment is specifically analyzed as follows:

Empirical formula for pressure loss:

Pressure loss is an important parameter used to characterize filter performance, and its size is related to the energy consumption of the entire filtration system, so pressure loss is of great significance for energy saving. According to Darcy's law, it is a function of air viscosity coefficient μ, filter thickness t, head wind speed V, fiber diameter _df and dimensionless pressure loss f(α):

$\mathrm{<span\; class="notranslate">\; \&Delta;p</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>\frac{\mathrm{<span\; class="notranslate">\; \&mu;tV</span>}}{{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; f</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}$

In the formula, f(α) is a dimensionless pressure loss, which is only a function of packing density or solid volume fraction (SVF) α, and the dimensionless pressure loss has different expressions based on different theories form.

According to the structural characteristics of the fiber filter medium, the dimensionless resistance expression of the pressure loss of the filter can be calculated according to the pressure drop of a single cylindrical fiber:

$\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 16</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 0.5</span>}\mathrm{<span\; class="notranslate">\; ln</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 0.5</span>}\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}$

The dimensionless pressure loss expression assumes that each cylinder is surrounded by a coaxial cylinder with radius b (distance between cylinder centers is 2b), that is, the influence of surrounding fibers is taken into account, and the shear stress on the surface of the cylinder is assumed to be zero.

At the same time, a dimensionless drag expression is provided assuming zero curl on the surface of an outer cylinder surrounding and coaxial with the fiber:

$\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 16</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}{\mathrm{<span\; class="notranslate">\; Ku</span>}}$

Among them, Ku=-0.5lnα-0.75+α-0.25α ² is the hydraulic coefficient of Kuwahara.

For the case where the SVF is in the range of 0.6%-30%, and the fiber diameter df is between 1.6-80μm, the dimensionless pressure loss expression can be used:

f(α)=64α ^3/2 (1+56α ³ ).

One of the above dimensionless pressure loss functions can be selected during calculation.

Filtration efficiency empirical formula:

Based on the Kuwabara shell model, if the flow form and fiber network structure are known, the filter filtration efficiency E can be calculated from the single fiber filter material filtration efficiency (Single Fiber Efficiency, SFE, EΣ). Filter Efficiency:

$\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; (</span>\frac{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 4</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}{\mathrm{<span\; class="notranslate">\; E.</span>}}_{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; t</span>}}{\mathrm{<span\; class="notranslate">\; \&pi;</span>}{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span>$

where α is the solid volume fraction (SVF), t is the thickness of the fiber layer, and _df is the fiber diameter.

The total SFE can be obtained by combining the SFE of each mechanism alone:

E _Σ ＝1-(1-E _R )(1-E _I )(1-E _D )

where E _D is the SFE due to Brownian diffusion, E _R is the SFE due to interception, and E _I is the SFE due to inertial collision.

(1) SFE caused by Brownian diffusion

Diffusion motion plays an important role in the capture efficiency of small particles, and the fluid has the ability to equalize the thermal energy distribution of the gas. When there are suspended particles in the gas, they can be balanced. Therefore, the suspended particles also get heat energy from the air. Brownian motion is caused by the equilibrium exchange of heat between gas molecules and suspended particles. Based on this fact, the particles do random diffusion (Brownian diffusion), and due to this movement, the chances of the particles colliding with the fibers increase, resulting in capture. The quantification of this motion by the diffusion coefficient indicates that single-fiber trapping efficiency is related to Brownian motion.

The SFE caused by Brownian diffusion in this embodiment is:

${\mathrm{<span\; class="notranslate">\; E.</span>}}_{\mathrm{<span\; class="notranslate">\; D.</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1.6</span>}{<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}{\mathrm{<span\; class="notranslate">\; Ku</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 3</span>}}{\mathrm{<span\; class="notranslate">\; Pe</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 3</span>}}$

In the above formula, Ku=-0.5lnα-0.75+α-0.25α ² ,

in:

_dp is particle diameter

T is the absolute temperature of the fluid

μ is the viscosity of air

V is the wind speed

C _c is the Cunningham slip correction factor which can be expressed as $C_c=1+{\mathrm{kn}}_p(1.257+0.4e^{-1.1/{\mathrm{kn}}_p})$

σ is the Boltzmann constant σ=1.38×10-23(m2kg s-2k-1)

(2) SFE caused by interception

Particles are captured by this mechanism when the particle size is large enough for Brownian diffusion to be negligible but not large enough for inertial collisions to occur. In this case, it is feasible to simplify the particles. Therefore, particles captured by interception can be interpreted as induced by airflow.

The SFE due to interception is:

${\mathrm{<span\; class="notranslate">\; E.</span>}}_{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0.6</span>}\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}{\mathrm{<span\; class="notranslate">\; Ku</span>}}\frac{{\mathrm{<span\; class="notranslate">\; R</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span>}$

In the formula, It is the ratio of particle diameter to fiber diameter.

(3) SFE caused by inertial collision

If the particle size in the dust-laden airflow is greater than 1 μm and the particle mass is large, when the airflow flows through the fiber layer, due to the inertia, some particles will deviate from the flow line and collide with the fiber and be trapped.

The expression of SFE due to inertial collision, E _I is determined by the value of Stk, The SFE caused by the inertial collision in this embodiment is:

${\mathrm{<span\; class="notranslate">\; E.</span>}}_{\mathrm{<span\; class="notranslate">\; I</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; J</span>}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; Stk</span>}}{\mathrm{<span\; class="notranslate">\; 4</span>}{\mathrm{<span\; class="notranslate">\; Ku</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}$

This embodiment is based on the real fiber filter medium image, input a series of related parameters, and study the fiber filter medium by numerical simulation method, the input parameters of this embodiment filter fluid absolute temperature T is 400K, the air viscosity coefficient μ is the filter fluid absolute temperature The function of T, μ=[-2.3×10 ^-6 (T-273.15) ² +0.00485(T-273.15)+1.72]×10 ^-5 , the thickness of the filter material t is 3μm, the head wind speed V1 is 0.1m/s, V2 is 0.2m/s, V3 is 0.5m/s, V4 is 1.0m/s, V5 is 2.0m/s, and the particle diameter _dp is 2μm. Bringing the above parameters into the pressure loss empirical formula and filtration efficiency empirical formula can obtain the filtration efficiency value and pressure loss value of the fiber filter media under different face wind speeds, and draw the fiber filtration efficiency with the face wind speed as the abscissa and the pressure loss as the ordinate. The medium pressure loss curve is shown in Figure 6. The filter efficiency curve of the fiber filter medium is drawn with the diameter of the filtered particles as the abscissa and the filter efficiency as the ordinate, as shown in Figure 7.

It is worth noting that in the actual experiment, the absolute temperature T of the input parameter value of the filtered fluid is 300-1000 ° C, the thickness of the filter material t is 0-1000 μm, the head-on wind speed V1 is 0-2 m/s, and V2 is 0-2 m/s , V3 is 0~2m/s, V4 is 0~2m/s, V5 is 0~2m/s, particle diameter _dp is 0~5μm. Different pressure loss curves and filtration efficiency curves can be obtained by inputting different values, which will not be repeated here.

The mimetic reconstruction of a fiber filter medium structure and its performance calculation method described in Example 1, binary processing is performed on the scanning electron microscope image of the actual fiber filter medium, and through the determination of the optimal threshold, the fiber pixels and pore pixels The points are effectively separated, laying a solid foundation for the accurate determination of the centerline of the subsequent fiber filter media, the three-dimensional rotation radius di, and the three-dimensional rotation direction df. application limitations. At the same time, based on the image of the real fiber filter medium, input a series of relevant parameters, study the fiber filter medium through numerical simulation, and obtain the performance parameters of the real fiber filter medium, without the need for real experiments, which can save a lot of time and manpower and material resources , has high practical value.

The above schematically describes the present invention and its implementations, but the description is not restrictive. What is shown in the drawings is only one of the implementations of the present invention, and is not actually limited thereto. Therefore, if a person of ordinary skill in the art is inspired by it, and without departing from the inventive concept of the present invention, he or she devises solutions and embodiments similar to the technical solution without creativity, all of which shall belong to the protection scope of the present invention.

Claims (9)

1. a mimicry method for reconstructing for fiber filter media structure, the steps include:

Step 1, obtain the image of true fiber filter medium;

Step 2, optimal threshold is set, step 1 gained image is carried out to binary conversion treatment, extract the bianry image of fiber filter media and hole, fiber pixel point value is 1, and hole pixel point value is 0;

Step 3, step 2 gained bianry image is carried out to fiber filter media center line detecting, obtain fiber filter media center line ss;

Step 4, carry out the contrast of the four pairs of angular direction taking pixel on step 3 gained center line as basic point, obtain three-dimensional rotation radius di and the three-dimensional rotation direction df of each pixel on center line;

Step 5, taking pixel on step 3 gained center line as rotation basic point, taking step 4 gained three-dimensional rotation radius di and three-dimensional rotation direction df as rotation parameter, carry out three-dimensional rotation, obtain fiber filter media mimicryization and rebuild image.

2. the mimicry method for reconstructing of a kind of fiber filter media structure according to claim 1, is characterized in that: in step 1, true fiber filter medium is scanned under scanning electron microscope, obtain the scanning electron microscope image of fiber filter media.

3. the mimicry method for reconstructing of a kind of fiber filter media structure according to claim 2, is characterized in that: described in step 2, the deterministic process of optimal threshold is as follows:

A, fiber filter media scanning electron microscope (SEM) photograph is carried out to gray scale stretching:

K＝255/(b-a)

M＝(-255*a)/(b-a)

Y＝KX+M

In above formula, a is scanning electron microscope image minimum gradation value; B is scanning electron microscope image maximum gradation value; X is any pixel gray-scale value of scanning electron microscope image; Y is the gray scale tension values that pixel X is corresponding;

B, determine optimal threshold:

If the ratio that in target zone, pixel number accounts for entire image in image after step a gray scale stretches is ω 0, in target zone, pixel average gray is μ 0, within the scope of background, to account for the ratio of entire image be ω 1 to pixel number, within the scope of background, pixel average gray is μ 1, and the segmentation threshold of target and background is designated as T _m, the inter-class variance of target and background is designated as g:

g＝ω0ω1(μ0-μ1)^2

Corresponding threshold value T while adopting the method for traversal to obtain making inter-class variance g maximum _m, this threshold value is described optimal threshold.

4. according to the mimicry method for reconstructing of a kind of fiber filter media structure described in claim 2 or 3, it is characterized in that: the method that obtains fiber filter media center line ss in step 3 is: pixel point value 8 directions that stretch out centered by 1 pixel in step 2 gained bianry image, 45 ° of intervals such as 8 described directions, add up the pixel number that in 8 directions, contiguous pixels point value is 1, as as described in the pixel number of any two diagonal angle directional statistics of pixel all equate or to differ be 1, retain this pixel, otherwise give up; The pixel that is 1 to all pixel point values is all carried out aforesaid operations, obtains fiber filter media center line ss.

5. the mimicry method for reconstructing of a kind of fiber filter media structure according to claim 4, it is characterized in that: described in step 4, four pairs of angular direction contrasts are specially: the pixel that step 3 is retained carries out the pixel number contrast that 4 pairs of angular direction contiguous pixels point values are 1 again, using minimum pixel number to angular direction as three-dimensional rotation direction df, this to 1/2nd of pixel number on angular direction as three-dimensional rotation radius di.

6. a Calculation Methods for Performance for fiber filter media structure, the steps include:

Steps A, obtain the image of true fiber filter medium;

Step B, optimal threshold is set, steps A gained image is carried out to binary conversion treatment, extract the bianry image of fiber filter media and hole, fiber pixel point value is 1, and hole pixel point value is 0;

Step C, step B gained bianry image is carried out to fiber filter media center line detecting, obtain fiber filter media center line ss;

Step D, carry out the contrast of the four pairs of angular direction taking pixel on step C gained center line as basic point, obtain three-dimensional rotation radius di and the three-dimensional rotation direction df of each pixel on center line;

Step e, step B gained bianry image is carried out to anti-phase processing, fiber pixel point value is 0, and hole pixel point value is 1, and the pixel number that statistical pixel point value is 1, is designated as a; The true fiber filter medium image slices vegetarian refreshments number that statistic procedure A obtains, is designated as b, obtains fiber filter media image porosity a/b; Statistical pixel point value is 1 pixel region number, obtains fiber hole histogram;

Step F, step D gained three-dimensional rotation radius di is added up, obtain fibre diameter statistical graph and average fibre diameter d _f;

Step G, substitution step F gained average fibre diameter d _fand filtration fluid absolute temperature T, filtrate thickness t, face velocity V and particle diameter d _p, application experience formula draws filtration efficiency curve and the pressure loss curve of fiber filter media under different wind speed.

7. the Calculation Methods for Performance of a kind of fiber filter media structure according to claim 6, is characterized in that: described in step B, the deterministic process of optimal threshold is as follows:

A, fiber filter media scanning electron microscope (SEM) photograph is carried out to gray scale stretching:

K＝255/(b-a)

M＝(-255*a)/(b-a)

Y＝KX+M

In above formula, a is scanning electron microscope image minimum gradation value; B is scanning electron microscope image maximum gradation value; X is any pixel gray-scale value of scanning electron microscope image; Y is the gray scale tension values that pixel X is corresponding;

B, determine optimal threshold:

If the ratio that in target zone, pixel number accounts for entire image in image after step a gray scale stretches is ω 0, in target zone, pixel average gray is μ 0, within the scope of background, to account for the ratio of entire image be ω 1 to pixel number, within the scope of background, pixel average gray is μ 1, and the segmentation threshold of target and background is designated as T _m, the inter-class variance of target and background is designated as g:

g＝ω0ω1(μ0-μ1)^2

Corresponding threshold value T while adopting the method for traversal to obtain making inter-class variance g maximum _m, this threshold value is described optimal threshold.

8. the Calculation Methods for Performance of a kind of fiber filter media structure according to claim 7, is characterized in that: the experimental formula described in step G is:

Pressure loss experimental formula:

$\mathrm{\&Delta;p}=f\left(\mathrm{\&alpha;}\right)\frac{\mathrm{\&mu;tV}}{{d_f}^2}$

Wherein: $f\left(\mathrm{\&alpha;}\right)=\frac{16\mathrm{\&alpha;}}{-0.5\ln \mathrm{\&alpha;}-0.5\frac{1-{\mathrm{\&alpha;}}^2}{1+{\mathrm{\&alpha;}}^2}};$ $f\left(\mathrm{\&alpha;}\right)=\frac{16\mathrm{\&alpha;}}{\mathrm{Ku}},$ Ku=-0.5ln α-0.75+ α-0.25 α ²; Or f (α)=64 α ^3/2(1+56 α ³), μ is air coefficient of viscosity, for filtering the function of fluid absolute temperature T, α is fractional solid volume;

Filtration efficiency experimental formula:

$E=1-\exp \left(\frac{-4\mathrm{\&alpha;}{E}_{\mathrm{\&Sigma;}}t}{\mathrm{\&pi;}{d}_f(1-\mathrm{\&alpha;})}\right)$

Wherein, E _Σ=1-(1-E _r) (1-E _i) (1-E _d), $E_D=1.6{\left(\frac{1-\mathrm{\&alpha;}}{\mathrm{Ku}}\right)}^{1/3}{\mathrm{Pe}}^{-2/3},$ $E_R=0.6\frac{1-\mathrm{\&alpha;}}{\mathrm{Ku}}\frac{R^2}{(1+R)},$ in above formula, Ku=-0.5ln α-0.75+ α-0.25 α ², c _cfor the Cunningham's skink slip correction factor, σ is Boltzmann constant; j=(29.6-28 α ^0.62) R-27.5R ^2.8, $\mathrm{Stk}=\frac{{\mathrm{\&rho;}}_p{d}_p^2{C}_{c}V}{18\mathrm{\&mu;}{d}_f}.$

9. the Calculation Methods for Performance of a kind of fiber filter media structure according to claim 8, it is characterized in that: described in step G, filtering fluid absolute temperature T is 300～1000K, filtrate thickness t is 0～1000 μ m, face velocity V1 is that 0～2m/s, V2 are that 0～2m/s, V3 are that 0～2m/s, V4 are that 0～2m/s, V5 are 0～2m/s, particle diameter d _pbe 0～5 μ m.