1、Designation: E 2001 98 (Reapproved 2003)Standard Guide forResonant Ultrasound Spectroscopy for Defect Detection inBoth Metallic and Non-metallic Parts1This standard is issued under the fixed designation E 2001; the number immediately following the designation indicates the year oforiginal adoption o
2、r, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. Asuperscript epsilon (e) indicates an editorial change since the last revision or reapproval.1. Scope1.1 This guide describes a procedure for detecting defects inmetallic and non-met
3、allic parts using the resonant ultrasoundspectroscopy method. The procedure is intended for use withinstruments capable of exciting and recording whole bodyresonant states within parts which exhibit acoustical or ultra-sonic ringing. It is used to distinguish acceptable parts fromthose containing de
4、fects, such as cracks, voids, chips, densitydefects, tempering changes, and dimensional variations that areclosely correlated with the elastic properties of the material.1.2 This standard does not purport to address all of thesafety concerns, if any, associated with its use. It is theresponsibility
5、of the user of this standard to establish appro-priate safety and health practices and determine the applica-bility of regulatory limitations prior to use.2. Referenced Documents2.1 ASTM Standards:E 1316 Terminology for Nondestructive Examinations23. Terminology3.1 DefinitionsThe definitions of term
6、s relating to con-ventional ultrasonics can be found in Terminology E 1316.3.2 Definitions of Terms Specific to This Standard:3.2.1 resonant ultrasonic spectroscopy (RUS), na nonde-structive examination method, which employs resonant ultra-sound methodology for the detection and assessment of varia-
7、tions and mechanical properties of a test object. In thisprocedure, whereby a rigid part is caused to resonate, theresonances are compared to a previously defined resonancepattern. Based on this comparison the part is judged to be eitheracceptable or unacceptable.4. Summary of the Technology (1)34.1
8、 Introduction:4.1.1 In addition to its basic research applications in phys-ics, materials science, and geophysics, Resonant UltrasoundSpectroscopy (RUS) has been used successfully as an appliednondestructive testing tool. Resonant ultrasound spectroscopyin commercial, nondestructive testing has a fe
9、w recognizablenames including, RUS Nondestructive Testing, Acoustic Reso-nance Spectroscopy (ARS), and Resonant Inspection. Earlyreferences to this body of science often are termed the “sweptsine method.” It was not until 1990 (2) that the name ResonantUltrasound Spectroscopy appeared, but the two t
10、echniques aresynonymous. RUS based techniques are becoming commonlyused in the manufacture of steel, ceramic, and sintered metalparts. In these situations, a part is vibrated mechanically, anddefects are detected based on changes in the pattern ofvibrational resonances or variations from theoretical
11、ly calcu-lated or empirically acceptable spectra. RUS measures allresonances, in a defined range, of the part rather than scanningfor individual defects. In a single measurement, RUS-basedtechniques potentially can test for numerous defects includingcracks and dimensional variations. Since the RUS m
12、easure-ment yields a whole body response, it is often difficult todiscriminate between defect types, that is, cracks or otherdiscontinuities. Nevertheless, on certain types of parts, it canbe accurate, fast, inexpensive and require no human judgment,making 100 % inspection possible in selected circu
13、mstances.Many theoretical texts (3) discuss the relationship betweenresonances and elastic constants and include the specificapplication of RUS to the determination of elastic constants(4). The technology received a quantum increase in attentionwhen Migliori published a review article, including the
14、 requi-site inexpensive electronic designs and procedures from whichmaterials properties could be measured quickly and accurately(5). The most recent applications include studies in ultrasonicattenuation, modulus determinations, thermodynamic proper-ties, structural phase transitions, superconductin
15、g transitions,magnetic transitions, and the electronic properties of solids. Acompendium of these applications may be found in theMigliori (1) text. Resonant ultrasound spectroscopy also founduse in the study of the elastic properties of the Apollo moonrocks (6).4.1.2 This guide is intended to provi
16、de a practical introduc-tion to RUS-based nondestructive test (NDT), highlightingsuccessful applications and outlining failures, limitations, and1This guide is under the jurisdiction of ASTM Committee E07 on Nondestruc-tive Testing and is the direct responsibility of Subcommittee E07.06 on Ultrasoni
17、cMethod.Current edition approved July 10, 2003. Published September 2003. Originallyapproved in 1998. Last previous edition approved in 1998 as E 2001 - 98.2Annual Book of ASTM Standards, Vol 03.03.3The boldface numbers in parentheses refer to the list of references at the end ofthis guide.1Copyrigh
18、t ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States.potential weaknesses. Vibrational resonances are consideredfrom the perspective of defect detection in 4.2. In 4.3 and 4.4,a review of some of the types of RUS measurements arepresented. In 4.5,
19、 some example implementations and configu-rations of RUS systems and their applications are presented.Finally, the guide concludes with a discussion of constraints,which limit the effectiveness of RUS.4.2 Mode Shapes and Defects:4.2.1 Resonant ultrasound spectroscopy/NDT techniques,operate by drivin
20、g a part at given frequencies and measuring itsmechanical response (Fig. 1 contains a schematic for the RUSapparatus). The process proceeds in small frequency steps oversome previously determined region of interest. During such asweep, the drive frequency typically brackets a resonance.When the exci
21、tation frequency is not matched to one of theparts resonance frequencies, very little energy is coupled tothe part; that is, there is essentially no vibration. At resonance,however, the energy delivered to the part is coupled generatingmuch larger vibrations. A parts resonance frequencies aredetermi
22、ned by its dimensions (to include the shape andgeometry) and by the density and the elastic constants of thematerial. The required frequency window for a scan dependson the size of the part, its mechanical rigidity, and the size ofthe defect being sought.4.2.2 Vibrational resonances produce a wide r
23、ange of dis-tortions. These distortions include shapes, which bend andtwist. It is known that increasing the length of a cylinder willlower some resonant frequencies. Similarly, reducing thestiffness, that is, reducing the relevant elastic constant, lowersthe associated resonant frequency for most m
24、odes; thus, for agiven part, the resonant frequencies are measures of stiffness,and knowledge of the mode shape helps to determine whatqualities of the part affect those frequencies. If a defect, such asa crack, is introduced into a region under strain, it will reducethe effective stiffness, that is
25、, the parts resistance to deforma-tion, and will shift downward the frequency of resonant modesthat introduce strain at the crack. This is one basis for detectingdefects with RUS-based techniques.4.2.3 The torsional modes represent a twisting of a cylinderabout its axis. These resonances are easily
26、identified becausetheir frequencies remain constant for fixed length, independentof diameter. A crack will reduce the ability of the part to resisttwisting, thereby reducing the effective stiffness, and thus, thefrequency of a torsional mode. A large defect can be detectedreadily by its effect on th
27、e first few modes; however, smallerdefects have much more subtle effects on stiffness, andtherefore, require higher frequencies (high-order modes) to bedetected. Detection of very small defects may require using thefrequency corresponding to the fiftieth, or even higher, mode.Some modes do not produ
28、ce strain in the end of the cylinder,therefore, they cannot detect end defects. To detect this type ofdefect, a more complex mode is required, the description ofwhich is beyond the scope of this specification. A defect in theend will reduce the effective stiffness for this type of mode, andthus, wil
29、l shift downward the frequency of the resonance. Ingeneral, it must be remembered that most modes will exhibitcomplex motions, and for highly symmetric objects, can belinear combinations of several degenerate modes, as discussedin 4.3.2.4.3 General Approaches to RUS/NDT:4.3.1 Test Evaluation Methods
30、 (1)Once a fingerprint hasbeen established, for conforming parts, numerous algorithmscan be employed to either accept or reject the part. Forexample, if a frequency 650 Hz can be identified for allconforming parts, the detection of a peak outside of thisboundary condition will cause the computer cod
31、e to signal a“test reject” condition. The code, rather than the inspector,makes the accept/reject decision. The following sections willexpand on some of these sorting criteria.4.3.2 Frequency Shifts:4.3.2.1 Resonant ultrasound spectroscopy measurementsgenerally produce strains (even on resonance) th
32、at are wellwithin the elastic limit of the materials under test, that is, theatomic displacements are small in keeping with the “nonde-structive” aspect of the testing. If strains are applied above theelastic limit, a crack will tend to propagate, causing a mechani-cal failure. Note that certain imp
33、ortant engineering properties,for example, the onset of plastic deformation, yield strength,etc., generally are not derivable from low-strain elastic prop-erties. Sensitivity of the elastic properties of an object to thepresence of a crack depends on the stiffness and geometry ofthe sample under tes
34、t. This concept is expanded upon under4.4.3.4.3.2.2 Fig. 2 shows an example of the resonance spectrumfor a conical ceramic part. Several specific types of modes arepresent in this scan, and their relative shifts could be used todetect defects as discussed above; however, the complexity issuch that,
35、for NDT purposes, some selections must be made sothat only a portion of such a large amount of information isused. For simple part geometries, the mode type and frequencycan be calculated, and selection of diagnostic modes can bebased on these results. For complex geometries, empiricalapproaches hav
36、e been developed to identify efficiently diag-nostic modes for specific defects. In this process, a technicianmeasures the spectra for a batch of known good and bad parts.The spectra are compared to identify diagnostic modes whoseshift correlates with the presence of the defect. The key is toisolate
37、 a few resonances, which differ from one another, whenknown defects are present in the faulty parts.4.3.3 Peak SplittingOne of the techniques employed foraxially symmetric parts is identified in texts on basic waveFIG. 1 Schematic of the Essential Electronic Building Blocks toEmploy RUS in a Manufac
38、turing EnvironmentE 2001 98 (2003)2physics (7). Some test procedures are based on simple fre-quency changes while others include the recognition thatsymmetry is broken when a defect is present in a homoge-neous, isotropic symmetrical part. These techniques employsplitting of degeneracies or simply “
39、splitting.” A cylinderactually has two degenerate bending modes, both orthogonal toits axis. The bending stiffness for both of these modes, andtherefore their resonance frequency, is proportional to thediameter of the cylinder. Because the part is symmetric, bothmodes have the same stiffness, and th
40、erefore, the same fre-quency (the modes are said to be degenerate and appear to bea single resonance). When the symmetry is broken by a chip,however, the effective diameter is reduced for one of theorthogonal modes. This increases the frequency for that mode,so both modes are seen. In addition, a cr
41、ack or inclusion affectsthe symmetry. This splitting of the resonances is illustrated inFig. 3, which shows spectra for a good part and two defectiveparts. The part is a steel cylinder. Fig. 3 also demonstrates auseful feature of this particular technique, that is, the size of thesplitting is propor
42、tional to the size of the defect. It is importantto recognize that not all resonance peaks are degenerate. Puretorsional modes, for example, are not degenerate, so theycannot be used for splitting.4.3.4 Phase Information and Peak Splittings:4.3.4.1 In practice, the same empirical approach describedf
43、or frequency shifts is used to identify diagnostic modes whoseFIG. 2 Typical Broad-Spectrum ScanFIG. 3 Shown is a Bending Mode Within a Resonance Spectrum of an Acceptable Steel Cylinder (a), One With a Small Defect (b), andOne With a Large Crack (c).E 2001 98 (2003)3splittings correlate with the si
44、ze of a defect of interest. Thesensitivity of this type of measurement is enhanced by theinterference, which occurs between closely-spaced peaks. Thedestructive interference develops into a visible spectral split-ting which would not be noticeable with the amplitude spec-trum (the real and quadratur
45、e components add to form theamplitude response). Most commercial systems function rea-sonably well without this attribute, but the problem can beexacerbated when the material exhibits resonance line widthswhich are greater than 1 % of the frequency. Under suchcircumstances, it may be impossible to d
46、etect splittings withoutphase information.4.3.4.2 Degenerate modes all have the same phase at low-symmetry points; therefore, if one mode is shifted slightlydestructive interference occurs between them, showing up as asplitting if the sample rigidity is sufficient. The frequencydifference between th
47、e two resulting peaks increases in directproportion to the defect size. This does not hold for accidentaldegeneracies, which are modes that by coincidence rather thanby symmetry have the same frequency. The actual shift infrequency is no more than would be expected for an isolatedmode, but the inter
48、ference enhances the visibility.4.3.5 Dimensional Measurements:4.3.5.1 Industrial sorting of parts often occurs in two sec-tors: defect detection and dimensional inspection. Ceramicplants often spend more resources on the latter than crack andchip investigations. It is a relatively simple matter to
49、use RUStechniques to measure physical parameters, such as weight,density, and dimensions. In practice, one measures all thephysical attributes possible. For a ring, these would includeweight, thickness, outer diameter, and inner diameter. It isimperative to use “good”, that is, as free of defects as possible,parts in this study. One measures a suitable number of thelowest resonances and either plots each resonance frequency asa function of each parameter or uses the correlation feature ofstandard spreadsheet programs. The best results are obtainedwhen singlets (single resonances
