CN104090033B - Coarse grain material FDTD ultrasound detection Building of Simulation Model method based on EBSD collection of illustrative plates - Google Patents

Abstract

A kind of coarse grain material FDTD ultrasound detection Building of Simulation Model method based on EBSD collection of illustrative plates, belongs to ultrasonic non-destructive inspection techniques field. After the steps include: to directly utilize the crystal orientation collection of illustrative plates of EBSD technical limit spacing coarse grain material, according to the actual die structure in macroscopic view metallograph, select the crystal grain in threshold angle definition collection of illustrative plates, and with the color that main orientation is corresponding, crystal grain is filled with, and obtained the image being made up of square pixels point by gray proces. Grain orientation is by Eulerian angles corresponding to gray valueΦ、Represent, in order to the elastic anisotropy stiffness matrix of quantitative Analysis crystal grain. Compared to conventional model, this model has grainiess and grain orientation describes accurately, operation efficiency advantages of higher, provides model basis for the defect quantitative in solution coarse grain material ultrasound detection, location, qualitative question. The present invention can also expand the foundation to other ultrasonic phantoms of elastic anisotropy polycrystalline material such as austenitic weld seam, biphase titanium alloy, has good prospect for promotion and application.

Description

Coarse grain material FDTD ultrasound detection Building of Simulation Model method based on EBSD collection of illustrative plates

Technical field

The present invention relates to a kind of coarse grain material FDTD ultrasound detection Building of Simulation Model method based on EBSD collection of illustrative plates, particularly relate to ultrasonic non-destructive inspection techniques field.

Background technology

The coarse grain materials such as austenitic stainless steel are widely used in the field such as nuclear power and chemical industry, and its safety receives much concern. Due to complicated coarse structure and grain orientation, this type of material has very strong elastic anisotropy, cause Ultrasonic NDT process occurs the phenomenons such as serious acoustic beam deflection, construct noise, signal distortion, it is difficult to accurately defect is positioned, qualitatively and quantitatively.

For solving the problems referred to above, researcher in this field attempts to set up corresponding ultrasound detection phantom, describes the reciprocal action between ultrasound wave and elastic anisotropy structure by simulation means. The model of early stage is to grow up from the basis of weld seam phantom, and modeling approach is mainly macroscopical metallograph of reference material, is multiple anisotropy unit by model partition. Meanwhile, according to column crystal grain growth feature, bond material is gained knowledge, and artificially arranges the crystal orientation of unit. Due to this semiempirical method, orientation and distribution to crystal grain is not accurately measured, and causes that the ultrasonic calculating of model and experimental result have very big-difference. Although the later stage introduces XRD (X-RayDiffraction, X-ray diffraction) to material local crystal orientation measured, have modified the orientation of model to a certain extent to arrange, but still cannot accurately record grain-oriented distribution characteristics, make the development of model receive restriction.

Within 2009, French atomic energy committee member develops a kind of phantom based on Thiessen polygon graphic-arts technique in CIVA business software. This model utilizes convex surface closed area that Thiessen polygon surrounds as simulation crystal grain, and assumes that crystal grain is elastic isotropy medium. Although this model can relatively accurately describe grainiess, but, the velocity of sound fluctuation arranged between elastic isotropy crystal grain by random function carrys out the anisotropic method of Equivalent Elasticity, do not account for the velocity of sound variation characteristic with ultrasonic propagation direction, cause having greater difference between the result of calculation of model and experimental result.

In recent years, EBSD EBSD is accepted by increasing investigation of materials scholar both at home and abroad. This technology can characterize the grain morphology of material again can the grain orientation of quantitative description material, it is believed that be the combination of SEM and XRD. 2009, the research worker of University of Birmingham utilized EBSD technology to scan austenitic weld seam cross section, obtains crystal orientation collection of illustrative plates. With reference to TuPu method, utilize CAD draw in the CIVA software model being made up of 157 elastic anisotropy unit. Finally, according to the crystal orientation of unit, it is assigned to corresponding stiffness matrix, completes the foundation of model. Compared to conventional model, the model of this data acquisition by experiment can describe grainiess and grain orientation more exactly.

But, owing to elastic anisotropy element number affects the operation efficiency of this model, in order to shorten operation time, have to adopt the mode simplified model reducing element number, nonetheless, for the model of 157 unit, run and once need nonetheless remain for nearly 100 hours. Additionally, for having thick ordered orientation column crystal, this simplification is less on the impact of result of calculation. Different from weld seam, coarse grain material does not contain only the column crystal of ordered orientation, and possibly together with the equiax crystal of orientation random distribution, and crystallite dimension is relatively small. According to above-mentioned method for simplifying, can cause that grainiess and orientation distortion are serious, affect the computational accuracy of model. Therefore, the elastic anisotropy dividing elements of coarse grain material is crossed range request and is reduced grainiess as much as possible, to ensure computational accuracy. It is also desirable to take corresponding method to improve the computational efficiency of this high-precision model.

Summary of the invention

It is an object of the invention to provide a kind of coarse grain material FDTD ultrasound detection Building of Simulation Model method based on EBSD collection of illustrative plates. Compared to conventional model, this model should have grainiess and grain orientation describes the advantages such as accurate, model calculation is in hgher efficiency. The construct noise in coarse grain material ultrasound detection can also be simulated, contribute to the deep reciprocal action understood between ultrasound wave and anisotropic grain, provide model basis for the defect quantitative in solution coarse grain material, location, qualitative question. The present invention can also expand the foundation to other ultrasonic phantoms of elastic anisotropy polycrystalline material such as austenitic weld seam, biphase titanium alloy simultaneously, has good prospect for promotion and application.

The technical solution used in the present invention is: a kind of coarse grain material FDTD ultrasound detection Building of Simulation Model method based on EBSD collection of illustrative plates, comprises the steps:

(1) according to " GB/T19501-2004 EBSD analyze method general rule " national standard, sample cut, polish, machine glazed finish, the pretreatment such as destressing electrobrightening. Select 40 μm of scanning steps that sample to be tested is carried out EBSD analysis;

(2) utilize EBSD device that the whole cross section of sample is scanned, about 6mm every time², then according to the zonule collection of illustrative plates obtained is spliced by order, finally obtain the perfect crystal orientation collection of illustrative plates being spliced by multiple collection of illustrative plates;

(3) grainiess according to real material, utilizes channel5 to analyze software and adjusts the threshold angle in crystal orientation collection of illustrative plates, it is determined that the grain contours of model, it can be used as the elastic anisotropy unit of model;

(4) utilize channel5 to analyze software statistics and go out the color range value that pixel quantity in each unit is maximum, then unify to be this color range value by the color of pixels all in crystal grain, and write down the Eulerian angles that this color range value is corresponding;

(5) EBSD collection of illustrative plates is exported as the picture of JPG form, then opens in Photoshop software, be converted into gray level image, the color range value of crystal grain is arranged to corresponding gray value, derives with PCX picture format and preserve gray-scale map;

(6) Eulerian angles that crystal grain is corresponding are utilizedΦ、Try to achieve the direction cosine matrix R representing crystal orientation,

By 9 array element R in R₁₁To R₃₃Derive spin matrix R_D,

$R_D=\left(\begin{array}{cccccc}{R}_{11}^2& R_{12}^2& R_{13}^2& 2R_{11}{R}_{12}& 2R_{11}{R}_{13}& 2R_{12}{R}_{13}\\ R_{21}^2& R_{22}^2& R_{32}^2& 2R_{21}{R}_{22}& 2R_{21}{R}_{23}& 2R_{22}{R}_{23}\\ R_{31}^2& R_{32}^2& R_{33}^2& 2R_{31}{R}_{32}& 2R_{31}{R}_{33}& 2R_{32}{R}_{33}\\ R_{11}{R}_{21}& R_{12}{R}_{22}& R_{13}{R}_{23}& R_{11}{R}_{22}+R_{12}{R}_{21}& R_{11}{R}_{23}+R_{13}{R}_{21}& R_{12}{R}_{23}+R_{13}{R}_{21}\\ R_{11}{R}_{31}& R_{12}{R}_{32}& R_{13}{R}_{33}& R_{11}{R}_{32}+R_{11}{R}_{31}& R_{11}{R}_{33}+R_{13}{R}_{31}& R_{11}{R}_{33}+R_{13}{R}_{31}\\ R_{21}{R}_{31}& R_{22}{R}_{32}& R_{23}{R}_{33}& R_{21}{R}_{32}+R_{22}{R}_{31}& R_{21}{R}_{33}+R_{23}{R}_{31}& R_{22}{R}_{23}+R_{23}{R}_{31}\end{array}\right)---\left(2\right)$

Utilize R_DThis structure stiffness matrix C is rotated, obtain the elastic stiffness Matrix C under this grain orientation ',

$C^{\′}=R_{D}C{R}_D^{-1}---\left(3\right)$

Wherein, C is 6 × 6 matrixes, is made up of 3 independent elastic constants,

$C=\left(\begin{array}{cccccc}{C}_{11}& C_{12}& C_{12}& 0& 0& 0\\ C_{12}& C_{11}& C_{12}& 0& 0& 0\\ C_{12}& C_{12}& C_{11}& 0& 0& 0\\ 0& 0& 0& C_{44}& 0& 0\\ 0& 0& 0& 0& C_{44}& 0\\ 0& 0& 0& 0& 0& C_{44}\end{array}\right)---\left(4\right)$

(7) by ultrasonic for mode input to FDTD numerical simulation program, the lower boundary of model is set to solid-vacuum interface, its excess-three border is set to infinite boundary, select Gaussian pulse matching waveform as sound source, it is placed on the diverse location of model, carry out numerical computations, finally analog result and experimental result are contrasted.

The invention have the advantages that: the FDTD ultrasound detection phantom utilizing EBSD experimental data to build, solve the problem that conventional model is difficult to accurate description grainiess and elastic anisotropy, also improve computational efficiency problem, and can calculate and actually detected similar ultrasonic time domain waveform, contribute to deep understanding ultrasound wave propagation characteristic in coarse grain material, provide reliable model basis for quantitative, the location of defect, qualitative analysis.

Accompanying drawing explanation

Fig. 1 is the EBSD crystal orientation collection of illustrative plates of Z3CN20-09M axle-radial section.

Fig. 2 is that grain contours is affected (a) 0 degree, (b) 5 degree, (c) 10 degree, (d) 15 degree, (e) 20 degree, (f) 30 degree by different threshold angle.

Fig. 3 is macroscopical metallograph of Z3CN20-09M.

Fig. 4 is the grain contours that 20 degree of threshold angle are determined.

Fig. 5 is the gray-scale map after gray proces.

Fig. 6 is the position of Analogue probe.

Fig. 7 is ultrasonic time domain waveform and the FFT spectrum curve of Analogue probe.

Fig. 8 is simulation and the experimental result contrast of ultrasonic time-domain signal.

Detailed description of the invention

Based on the coarse grain material FDTD ultrasound detection Building of Simulation Model method of EBSD collection of illustrative plates, the pressurized-water reactor nuclear power plant Z3CN20-09M main pipeline material being 96mm for thickness, comprise the following steps that

(1) using the Z3CN20-09M of 96mm thickness as object of study, cutting 96mm × 12mm × 2mm sample along pipeline axle-radial direction, then sample is cut into 4 24mm × 12mm × 2mm, carry out EBSD analysis, remainder is used as ultrasonic tesint time domain waveform. According to " GB/T19501-2004 EBSD analyze method general rule " national standard, sample is polished, machine glazed finish, the pretreatment such as destressing electrobrightening. Select 40 μm of scanning steps that sample to be tested is carried out EBSD analysis;

(2) as a kind of microscopic measurement instrument, EBSD can only scan 6mm every time²Region, it is necessary to be sequentially carried out scanning in order.4 specimen surfaces amount to the EBSD collection of illustrative plates obtaining 271 zonules, utilize channel5 to analyze software and it is spliced, finally obtain the EBSD orientation collection of illustrative plates being spliced into by 271 pictures, as shown in Figure 1.

(3) utilize channel5 to analyze software and determine the grain contours in original orientation collection of illustrative plates, and then the crystal grain of acquisition phantom is as elastic anisotropy unit. According to " GB/T19501-2004 EBSD analyzes method general rule " national standard and " ISO24173 EBSD Determination of Orientation general rule " international standard, in EBSD collection of illustrative plates, crystal grain is defined as the misorientation set lower than the contiguous zone of a threshold angle. Such as, threshold angle is 10 degree, namely characterizes the oriented adjacent within the angle difference 10 degree of crystal orientation and can be regarded as a crystal grain. As the important parameter determining grain boundary and die locations, the selection of threshold angle is very crucial. As shown in Figure 2, when threshold angle is less than 10 degree, small-sized crystal grain in white edge, it is filled with, increase along with threshold angle, small-size grains is fewer and feweri, and when threshold angle reaches 30 degree, crystal grain maximum in white edge has been same crystal grain with adjacent grain mergin. Therefore, threshold angle is more little, and the crystal grain quantity in model is more many, and average crystal grain diameter is more little. On the contrary, when threshold angle is very big, can cause originally be not an orientation grain mergin together, cause that the crystal grain quantity of model is less than material actual die quantity. By contrasting macroscopic view metallograph Fig. 3, it has been found that when threshold angle is 20 degree, model pattern is consistent with the macrostructure of material, as shown in Figure 4.

(4) after determining the grain contours line of model, the orientation in grain contours is unified, make each crystal grain only have unique crystal orientation. Utilize channel5 to analyze software statistics and go out one group of Eulerian angles that pixel number in each grain contours is maximum, correspondingly, the color of crystal grain is modified as the color corresponding to main Eulerian angles, finally obtains the EBSD collection of illustrative plates being made up of 10 orientations or color;

(5) collection of illustrative plates is exported as the picture of 2400 × 300 pixel compositions, opens in Photoshop software, 10 colors in artwork are arranged to corresponding gray value, derive with PCX picture format and preserve gray-scale map, as shown in Figure 5. Eulerian angles corresponding to each gray scale refer to such as table 1. Model is made up of 2400 × 300 different gray scale square net pixels, and size of mesh opening is 40 μm, corresponding with scanning step. Each pixel is made up of a gray value, meanwhile, each gray value respectively corresponding one group represent grain-oriented Eulerian anglesΦ,

Color corresponding to table 1 different orientation region, gray value and Eulerian angles

(6) Eulerian angles that crystal grain is corresponding are utilizedΦ、Try to achieve the direction cosine matrix R representing crystal orientation, by 9 array element R in R₁₁To R₃₃Spin matrix R can be derived_D, utilize R_DThis structure stiffness matrix C is rotated, obtain the elastic stiffness Matrix C under this grain orientation ', wherein, C is 6 × 6 matrixes, is made up of 3 independent elastic constants.

(7) by ultrasonic for mode input to Fdtd Method numerical simulation program, upper, left and right for infinite boundary, lower end is solid-vacuum boundary. It is 25pixel/mm that mesh width is selected, and corresponding with EBSD figure spectral resolution, whole model is made up of 2400 × 300 pixels. Select the ultrasonic loading of list array element of 2mm aperture, respectively Analogue probe is placed on model center position, center 2mm to the right, center 4mm to the right, center 2mm to the left, 4mm position to the left, center, as shown in Figure 6.Select 1MHz Gaussian pulse matching waveform as sound source, waveform and the waveform fitting to 201 compositions according to actual sound source. In 1MHz matching waveform below figure 7 shown in (a), in its spectrum signature such as Fig. 7 after fast Fourier transform (fastFouriertransform, FFT), (b) dominant frequency is 1.04MHz. Finally, utilize Fdtd Method solving wave equations accurately to calculate ultrasound wave transmission on each propagation position and between anisotropic grain, reflection, refraction behavior, obtain the time domain waveform such as Fig. 8. Meanwhile, utilizing the time domain waveform of the Z3CN20-09M sample of 1MHz probe collection thickness 96mm, then this measured waveform and analog waveform are contrasted, as shown in Figure 8, two kinds of waveforms are substantially identical. And compared to CIVA software, the calculating time of FDTD simulated program does not rely on the anisotropy element number of model, runs and once only needs several minutes.

Claims (1)

1. the coarse grain material FDTD ultrasound detection Building of Simulation Model method based on EBSD collection of illustrative plates, it is characterised in that comprise the steps:

(1) according to " GB/T19501-2004 EBSD analyzes method general rule " national standard, sample cut, polish, machine glazed finish, destressing electrobrightening pretreatment, select 40 μm of scanning steps that sample to be tested is carried out EBSD analysis;

(2) utilize EBSD device that the whole cross section of sample is scanned, each 6mm², then according to the zonule collection of illustrative plates obtained is spliced by order, finally obtain the perfect crystal orientation collection of illustrative plates being spliced into by multiple collection of illustrative plates;

(3) grainiess according to real material, utilizes channel5 to analyze software and adjusts the threshold angle in crystal orientation collection of illustrative plates, it is determined that the grain contours of model, it can be used as the elastic anisotropy unit of model;

(4) utilize channel5 to analyze software statistics and go out the color range value that pixel quantity in each unit is maximum, then unify to be this color range value by the color of pixels all in crystal grain, and write down the Eulerian angles that this color range value is corresponding;

(5) EBSD collection of illustrative plates is exported as the picture of JPG form, then opens in Photoshop software, be converted into gray level image, the color range value of crystal grain is arranged to corresponding gray value, derives with PCX picture format and preserve gray-scale map;

(6) Eulerian angles that crystal grain is corresponding are utilizedΦ、Try to achieve the direction cosine matrix R representing crystal orientation,

By 9 array element R in R₁₁To R₃₃Derive spin matrix R_D,

$R_D=\left(\begin{array}{cccccc}{R}_{11}^2& R_{12}^2& R_{13}^2& 2R_{11}{R}_{12}& 2R_{11}{R}_{13}& 2R_{12}{R}_{13}\\ R_{21}^2& R_{\mathrm{22}}^2& R_{\mathrm{32}}^2& 2R_{21}{R}_{22}& 2R_{21}{R}_{23}& 2R_{22}{R}_{23}\\ R_{31}^2& R_{\mathrm{32}}^2& R_{\mathrm{33}}^2& 2R_{31}{R}_{32}& 2R_{31}{R}_{33}& 2R_{32}{R}_{33}\\ R_{11}{R}_{21}& R_{12}{R}_{22}& R_{13}{R}_{23}& R_{11}{R}_{22}+R_{12}{R}_{21}& R_{11}{R}_{23}+R_{13}{R}_{21}& R_{12}{R}_{23}+R_{13}{R}_{21}\\ R_{11}{R}_{31}& R_{12}{R}_{32}& R_{13}{R}_{33}& R_{11}{R}_{32}+R_{11}{R}_{31}& R_{11}{R}_{33}+R_{13}{R}_{31}& R_{11}{R}_{33}+R_{13}{R}_{31}\\ R_{21}{R}_{31}& R_{22}{R}_{32}& R_{23}{R}_{33}& R_{21}{R}_{32}+R_{22}{R}_{31}& R_{21}{R}_{33}+R_{23}{R}_{31}& R_{22}{R}_{33}+R_{23}{R}_{31}\end{array}\right)---\left(2\right)$

Utilize R_DThis structure stiffness matrix C is rotated, obtain the elastic stiffness Matrix C under this grain orientation ',

$C^{\′}=R_D{\mathrm{CR}}_D^{-1}---\left(3\right)$

Wherein, C is 6 × 6 matrixes, is made up of 3 independent elastic constants,

$C=\left(\begin{array}{cccccc}{C}_{11}& C_{12}& C_{12}& 0& 0& 0\\ C_{12}& C_{11}& C_{12}& 0& 0& 0\\ C_{12}& C_{12}& C_{11}& 0& 0& 0\\ 0& 0& 0& C_{44}& 0& 0\\ 0& 0& 0& 0& C_{44}& 0\\ 0& 0& 0& 0& 0& C_{44}\end{array}\right)---\left(4\right)$

(7) by ultrasonic for mode input to FDTD numerical simulation program, the lower boundary of model is set to solid-vacuum interface, its excess-three border is set to infinite boundary, select Gaussian pulse matching waveform as sound source, it is placed on the diverse location of model, carry out numerical computations, finally analog result and experimental result are contrasted.