1、Designation: E2627 10Standard Practice forDetermining Average Grain Size Using Electron BackscatterDiffraction (EBSD) in Fully Recrystallized PolycrystallineMaterials1This standard is issued under the fixed designation E2627; the number immediately following the designation indicates the year oforig
2、inal adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. Asuperscript epsilon () indicates an editorial change since the last revision or reapproval.1. Scope1.1 This practice is used to determine grain size frommeasurements
3、of grain areas from automated electron back-scatter diffraction (EBSD) scans of polycrystalline materials.1.2 The intent of this practice is to standardize operation ofan automated EBSD instrument to measure ASTM G directlyfrom crystal orientation. The guidelines and caveats of E112apply here, but t
4、he focus of this standard is on EBSD practice.1.3 This practice is only applicable to fully recrystallizedmaterials.1.4 This practice is applicable to any crystalline materialwhich produces EBSD patterns of sufficient quality that a highpercentage of the patterns can be reliably indexed usingautomat
5、ed indexing software.1.5 The practice is applicable to any type of grain structureor grain size distribution.1.6 The values stated in SI units are to be regarded asstandard. No other units of measurement are included in thisstandard.1.7 The values stated in inch-pound units are to be regardedas stan
6、dard. The values given in parentheses are mathematicalconversions to SI units that are provided for information onlyand are not considered standard.2. Referenced Documents2.1 ASTM Standards:2E7 Terminology Relating to MetallographyE112 Test Methods for Determining Average Grain SizeE766 Practice for
7、 Calibrating the Magnification of a Scan-ning Electron MicroscopeE1181 Test Methods for Characterizing Duplex Grain SizesE1382 Test Methods for Determining Average Grain SizeUsing Semiautomatic and Automatic Image Analysis3. Terminology3.1 Definitons:3.1.1 cleanupPost processing applied to EBSD scan
8、 datato reassign extraneous points in the scan grid to neighboringpoints. The extraneous points are assumed to arise fromnon-indexed or misindexed EBSD patterns.3.1.2 (crystallographic) orientationThe rotation requiredto bring the principle axes of a crystal into coincidence with theprinciple axes a
9、ssigned to a specimen. For example, in a rolledmaterial with cubic crystal symmetry, it is the set of rotationsrequired to bring the 100, 010 and 001 axes of the crystalinto coincidence with the rolling, transverse and normaldirections of the specimen. Orientations may be described interms of variou
10、s sets of angles, a matrix of direction cosines ora rotation vector.3.1.3 electron backscatter diffraction (EBSD).A crystal-line specimen is placed in a scanning electron microscope(SEM) at a high tilt angle (70). When a stationary electronbeam is positioned on a grain, the electrons are scattered i
11、n asmall volume (typically 30nm in the tilt direction, 10nm in thetransverse direction and 20 nm in depth for a field emission gunSEM and approximately an order of magnitude larger in thelateral directions for a tungsten filament SEM). Electrons thatsatisfy Braggs law are diffracted back out of the
12、specimen. Thediffracted electrons strike a phosphor screen (or alternatively aYAG crystal) placed in the chamber. The colliding electronsfluoresce the phosphor and produce a pattern. The pattern iscomposed of a set of intersecting bands (Kikuchi lines). Thesebands are indicative of the arrangement o
13、f crystal lattice planesin the diffracting crystal volume. Assuming the material is ofknown crystal structure, the orientation of the crystal within thediffracting volume can be determined.3.1.4 EBSD patternAn EBSD pattern is composed of aset of intersecting bands. The geometrical arrangement of the
14、sebands is indicative of the crystallographic orientation of thecrystal lattice within the diffraction volume.1This practice is under the jurisdiction of ASTM Committee E04 on Metallog-raphy and is the direct responsibility of Subcommittee E04.11 on X-Ray andElectron Metallography.Current edition ap
15、proved April 1, 2010. Published May 2010. DOI10.1520/E2627.2For referenced ASTM standards, visit the ASTM website, www.astm.org, orcontact ASTM Customer Service at serviceastm.org. For Annual Book of ASTMStandards volume information, refer to the standards Document Summary page onthe ASTM website.1C
16、opyright ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States.3.1.5 EBSD scanUnder computer control, the beam of theSEM is moved to a point on the specimen, an EBSD patterncaptured and indexed to determine the crystallographic orien-tation at the be
17、am location. This process is repeated for a set ofpoints lying on a regular grid.3.1.6 grainIn EBSD, grains have a specific meaning.They are defined as a group of similarly oriented neighboringpoints on the scan grid. The group is surrounded by a perimeterwhere misorientation across that perimeter e
18、xceeds a specifiedtolerance value.3.1.7 indexingThe process of identifying the crystallo-graphic orientation of the crystal lattice associated with anEBSD pattern generated by the interaction of the electron beamwith that lattice.3.1.8 misorientation The set of rotations (Euler angles)required to br
19、ing one crystal lattice into coincidence with asecond crystal lattice.3.1.9 misorientation toleranceIf the angular differencebetween two neighboring pixels is less than this tolerance valuethen they are assumed to belong to the same grain.3.1.10 orientation MapEach point in the scan grid isassigned
20、a color according to its orientation. This forms animage showing the microstructure.3.1.11 step size (D) The distance between neighboringpoints on the scan grid.4. Summary of Practice4.1 An EBSD scan is performed on a specimen, post-processing routines are applied to the scan data, and theindividual
21、 points of the scan are grouped into grains accordingto their orientation. Average grain size is determined from thefield average of grain areas based on the number of points inthe EBSD map and the step size.5. Significance and Use5.1 This practice provides a way to estimate the averagegrain size of
22、 polycrystalline materials. It is based on EBSDmeasurements of crystallographic orientation which are inher-ently quantitative in nature. This method has specific advan-tage over traditional optical grain size measurements in somematerials, where it is difficult to find appropriate metallo-graphic p
23、reparation procedures to adequately delineate grainboundaries.6. Apparatus6.1 An electron backscatter diffraction (EBSD) systemmounted on a Scanning Electron Microscope (SEM) is used.The EBSD system is constituted by a low-light sensitive videocamera (typically a charge-coupled device or CCD camera)
24、.The camera images a medium for detecting the diffractedelectrons such as a phosphor screen or YAG crystal. EBSDpatterns formed on the detecting medium are imaged using thecamera and transmitted to a computer.6.2 Software capable of reliably indexing an EBSD patternto determine the crystallographic
25、orientation from the EBSDpattern is needed. The computer and resident software shouldbe capable of rapid collection of orientation data from EBSDpatterns.6.3 Electronics and software to control the beam in the SEM(or the stage, or both) are required to collect orientation data atpoints on a regular
26、scan grid.7. Hazards7.1 There are no hazards specific to this test method.However, SEM operators should be familiar with safe SEMoperating procedures to prevent exposure to X rays and comingin contact with the high voltages inherent to SEMs. Careshould also be exercised in preparing specimens for EB
27、SD asis the case for specimen preparation for light and electronmicroscopy.8. Sampling and Test Specimens8.1 Specimens should be selected to represent averageconditions within a heat lot, treatment lot, or product, or toassess variations anticipated across or along a product orcomponent, depending o
28、n the nature of the material beingtested and the purpose of the study. Sampling location andfrequency can be based upon agreements between the manu-facturers and the users.8.2 Specimens should not be taken from areas affected byshearing, burning, or other processes that will alter the grainstructure
29、.8.3 The surface to be polished should be large enough inarea to permit measurement of at least five fields at the desiredmagnification.9. Specimen Preparation9.1 It is important to follow good metallographic prepara-tion procedures for successful EBSD work. EBSD is a surfacesensitive technique. The
30、 surface should be free from deforma-tion and have minimal topography. Careful mechanical polish-ing or electropolishing, or both should be performed onspecimen surfaces. However, as compared with preparingspecimens for grain size measurements based on opticalmicroscopy, the surface should not be et
31、ched or treated toproduce relief to delineate grain boundaries. This would bedisadvantageous for obtaining EBSD patterns that can beindexed. The grain boundaries are delineated by processing ofthe orientation measurement data.10. Preparation of Apparatus10.1 Good practices should be used in the oper
32、ation of boththe SEM and EBSD systems. Particular attention should begiven to geometric alignment of the specimen surface with theassumed measurement plane (typically 70 from the horizon-tal). Misalignments can easily arise from sample preparation ormounting of the sample. It is critical to mitigate
33、 such misalign-ment, as they affect both the accuracy of the orientationmeasurements as well as the accuracy of the beam position,particularly in the direction of tilt. Accurate measurements oforientation and position are critical for accurate characteriza-tion of grain size.11. Calibration and Stan
34、dardization11.1 The EBSD instrumentation should be properly cali-brated according to specific manufacturer instructions, includ-ing the EBSD imaging system magnification calibration perPractice E766 (SEM magnification calibration).E2627 10212. Procedure12.1 The first step is to perform an EBSD scan.
35、 Somea-priori knowledge of the estimated grain size is helpful inselecting an appropriate scan area and step size. A map shouldbe constructed from the scan data to insure that sensible resultsare obtained. In particular, the map should be inspected to seeif any grains appear speckled. If the automat
36、ed EBSD softwareis having difficulty indexing the EBSD patterns, the resultingorientation maps will appear “speckled”. Indexing problemsarise when the patterns are too weak for the band detectionalgorithms to accurately detect the bands in the patterns, if thesystem is poorly calibrated, if the crys
37、tal structure data isincorrect, or if the configuration parameters for the indexingalgorithms are not optimized. The “speckling” arises becauseincorrect orientations are obtained from individual patterns.Operators should following good practice as recommended bytheir EBSD vendors to mitigate such pr
38、oblems.12.2 Once the scan data have been collected, a cleanuproutine should be applied to the data to assimilate anynon-indexed or misindexed points into the surrounding neigh-borhood grains. The number of points modified by the cleanupprocess should be monitored. No more than 10% of the pointsshoul
39、d be modified in the process.12.3 Points of similar orientation should be grouped to-gether to form grains. A misorientation tolerance value of 5 isrecommended to define the orientation similarity, althoughother misorientation values could be specified in producer/purchaser agreements. After grain g
40、rouping, each point in thescan should be associated with a grain. Each grain group isidentified as a grain. At least 50 whole grains should beobservable in each scan area.12.4 Set up an EBSD scan so that the average grain containsabout 500 points. Count the number of scan points containedwithin each
41、 whole grain, Pi, within the scanned field, for anumber of fields (n), until at least 500 grains have beenmeasured. Exclude grains with point counts less than 100. Foreach grain with Piexceeding 100 calculate the correspondingarea, Ai. For scan data measured on a regular square grid, thearea of a gr
42、ain, Ai, is equal to the number of points in the grainmultiplied by the square of the step size:Ai5 PiD2(1)For data collected on a regular hexagonal grid, the area of agrain, Ai, is equal to the number of points in the grainmultiplied by the square of the step size, D2multiplied by thesquare root of
43、 three divided by two:Ai5=32PiD2(2)Store the areas of each grain in memory or record to a file.Calculate the mean area, A, for N grains measured in true areaunits (m2or mm2):12.5 A histogram of the frequency of the grain areas can beconstructed to determine or illustrate the uniformity of thegrain a
44、reas and to detect and analyze duplex grain sizeconditions. The analytical method is described in Test MethodsE1181, Appendix X2.A51n(i5lNAi(3)12.6 Calculate the standard deviation, s, of the measuredgrain areas, Ai:s 5F1N2 1(i5lNAi A!2G12(4)13. Interpretation of Results13.1 Follow the methodology s
45、et out in Section 14 of TestMethods E1382 for grain area measurements, namely: after thedesired number of grains, N, have been measured, calculate themean value of the measurement as described in 13.4 and itsdeviation as described 13.5. The ASTM grain size number, G,is then calculated according to t
46、he following:G 5 3.3223 log A! 2.995 Ain mm2(5)G 5 3.3223 log A106!# 2.995 Ain m2Round off the value of G to the nearest tenth unit.13.2 Determine the 95 % confidence intervals, 95%CI, ofeach measurement in accordance with:95% CI 56t s=N(6)(6)Where:t = 1.960 assuming N $ 50013.3 Determine the percen
47、t relative accuracy, %RA,ofthemeasurement by dividing the 95%CI value by the mean grainarea and multiplying by 100:%RA 595%CIA100 (7)13.4 For specimens with non-equiaxed grain structuresmeasure the mean grain area using measurements on longitu-dinal (l), transverse (t) and planar (p) oriented surfac
48、es. Thepooled mean grain area, Apooled, for the specimen can beobtained from the three mean area values for the three principleplanes (longitudinal, transverse, and planar) according to:Apooled5 Al At Ap!13 (8)Where l, t, and p subscripts indicate the test plane. The grainsize number can then be cal
49、culated based on Eq 3 using Apooled.Calculate the pooled standard deviation, spooled, according toEq 9 (Test Methods E1382, equation (A1.8).)spooled5 SN11!sl21 Nt1!st21 Np1! sp2Nl1! 1 Nt1! 1 Np1!D12(9)Determine the 95%CI and %RA based on the pooled standarddeviation and the pooled mean grain area of the three mea-surements according to equation Eq 6 and Eq 7.13.5 For a duplex grain size distribution, the analysis isconducted as described in Appendix X2 of Test MethodsE1181.14. Report14.1 The report should documen
