The invention relates generally to ultrasonic non-destructive testing (NDT) using the techniques of full-matrix capture (FMC) and total focusing method (TFM), and more particularly to a method of using time-of-flight of surface waves to determine the ultrasound velocity in a test object.
In a typical existing full matrix capture (FMC) acquisition, a plurality of pulser elements of one or more ultrasonic array probes are individually pulsed and an A-scan (a plot of echo response amplitude vs reception time) is received for each pulse at each of a plurality of receiver elements. The result of the acquisition is an N×M matrix of response A-scans, where M is the number of pulsers and N is the number of receivers. The existing FMC acquisition method enables full beam forming capabilities in post-processing for both emission and reception. Of all the possibilities enabled by FMC, one of the most promising aspects is the ability to provide optimized focusing at all of the imaging plane positions. This is referred to as Total Focusing Method (TFM).
In existing practice, TFM is applied by dividing the imaging volume into an array of voxels, and summing the response A-scans from the FMC matrix, with delays appropriate to the time-of-flight from each pulser via each voxel to each receiver. The term “voxel” is used herein to denote an elementary volume within the imaging volume, analogous to the term “pixel” as applied to two-dimensional images.
Since the calculations performed on FMC data to achieve a TFM image involve determining time-of-flight, the acoustic velocity of the relevant wave type in the test object must be known. Relevant waves types are shear waves (hereinafter referred to as S-waves) and longitudinal waves (hereinafter referred to as P-waves). A significant problem in FMC/TFM analysis is that the acoustic velocity in steel, for example, depends on the composition of the test object material, its thermal treatment and other factors that are not known when doing a non-destructive inspection. In fact, as shown in
In current practice, ultrasonic velocity measurement methods are mainly based on calculating the time necessary to reach a reflector at a given distance from the probe. However, all methods in current practice involve use of calibration blocks which by definition are not exactly the same material as the test object. Therefore the velocity measurements are necessarily imprecise with respect to the test object, and the quality of the resulting TFM images is significantly affected.
When using two probes in “pitch-catch” (P-C) mode for FMC-TFM imaging, the time-of-flight also depends on the distance between the probes. Therefore the TFM images will also sensitively depend on accurate knowledge of that distance. Typically the probe distance is maintained by means of a mechanical link whose length is adjustable. Therefore it is important to have a measurement method which can confirm that the length of adjustable link has not been inadvertently changed, and that the defined probe distance is effectively maintained throughout a lengthy series of inspections.
There therefore exists a need for a method of accurately determining the acoustic velocity in the test object during the inspection so that the TFM images resulting from the inspection will provide a reliable measure of the intensity of indications.
Furthermore, there also exists a need for a method of accurately determining the distance between probes during a P-C FMC-TFM inspection.
Accordingly, it is a general objective of the present disclosure to provide a real-time method of accurately determining the acoustic velocity in the test object.
It is further an objective of the present disclosure to provide a real-time method of determining the acoustic velocity which does not require presence of a flaw in the test object.
It is further an objective of the present disclosure to provide a real-time method of determining the velocities of both the S-wave and the P-wave in the test object.
It is further an objective of the present disclosure to provide a real-time method of determining the acoustic velocities in a steel test object either in the presence or absence of a weld cap.
It is further an objective of the present disclosure to provide a real-time method of determining the separation distance between two probes in a P-C FMC-TFM acquisition.
These objectives are achieved in the present disclosure by determining the times-of-flight of both the P-type surface wave and the Rayleigh surface wave generated at the critical angles between the wedge and the test object. Since the wedge properties are known, if the time-of-flight is determined, the associated acoustic velocity in the test object may also be determined. Knowing the acoustic velocities of P-waves and Rayleigh waves, the velocity of S-waves in the test object may be calculated. In addition, with knowledge of the acoustic velocities, the probe separation may be determined by time-of-flight measurements.
M number of emitting elements
N number of receiving elements
θC critical angle
θP, θR critical angles for P-waves, Rayleigh waves
V acoustic velocity
VP, VR, VS acoustic velocities for P-waves, Rayleigh waves, S-waves
VW acoustic velocity in receiving wedge
S probe separation
τ constant time-of flight
τP, τR constant times-of flight for P-waves, Rayleigh waves
τ0 reference constant time-of flight for S=0
δtn reception delay of nth receiving element
α angular offset of received wavefront
tAn time-of-flight from emitting element to nth receiving element
dn distance from 1st to nth receiving element
p pitch of receiving elements
For the purposes of the present disclosure, the following assumptions are made:
As shown in
As shown in
It should also be noted that surface wave 28 comprises a mixture of P-waves and Rayleigh waves, wherein the P-waves are P-polarized (vibrations parallel to the propagation direction) and the Rayleigh waves are a mixture of P- and S-polarizations (elliptical vibrations). The P-waves propagating in surface wave 28, also known as lateral waves, have the same velocity as bulk P-waves propagating in the volume of test object 10. In the present disclosure, use of the term “P-wave” will include both surface P-waves and bulk P-waves.
Note that the critical angle for a P-wave is different from the critical angle for a Rayleigh wave and therefore critical angle θc as shown in
where vS is the S-wave wave velocity in test object 10.
Hereinafter, the acoustic velocity of surface wave 28 in test object 10 will be designated v, where it is understood that v=vP for P-waves and v=vR for Rayleigh waves.
Still referring to
Hence, one of the novel and important aspects of the present disclosure is application to the determination of acoustic velocities in the test object of the above explained newly discovered concept of equal times-of-flight when the emitted and received beams are at the critical angle
Another novel and important aspect of the present disclosure is application of the above explained concept of equal times-of-flight together with the operation of “pitch-catch” mode and FMC acquisition to derive the P-wave and shear-wave velocities during an inspection operation. Determination of acoustic velocities in the actual test object during the inspection has many significant benefits. For example, complicated processes of determining sound velocity in a calibration block in existing practice are avoided. Accuracy of the inspection does not have to be compromised by using an assumed acoustic velocity, or by measuring acoustic velocity in a material which does not correspond exactly to the test object.
where vw is the velocity of P-waves in the wedge, which is known because the properties of the wedge material are known and produced with known tolerances. In equation (2) it is understood that v and θc may take values for either P- or Rayleigh surface waves.
Still referring to
where tDE is the time-of-flight for waves propagating between points D and E, and tDBC is the time-of-flight for waves propagating between points D, B and C. By applying Snell's law (equation (2)) and the geometry of similar triangles OBC and ODE, it can be shown that when beams 140 and 141 are at the critical angle θc
tDE=tDBC (5)
Since beam 140 and element (n) are representative of all beams received by all elements of probe 14, equation (5) shows that the time-of-flight is the same for all elements of probe 14. Therefore:
tAn=τ (6)
where tAn is the time-of-flight from point A to receiving element (n) and τ is a constant value of time-of-flight. Note that the constant time-of-flight τ has a value τP for P-waves and a value τR for Rayleigh waves.
A preferred method of data acquisition according to the present invention is to fire any element (m) of probe 12 and to acquire A-scans for all N elements of probe 14. In this case, probe 12 is the emitting probe, wedge 16 is the emitting wedge, probe 14 is the receiving probe and wedge 18 is the receiving wedge. By the principle of reversibility, an alternative method is to fire any element (n) of probe 14 and to acquire A-scans for all M elements of probe 12. In this case, probe 14 is the emitting probe, wedge 18 is the emitting wedge, probe 12 is the receiving probe and wedge 16 is the receiving wedge.
It will be shown below that the acquired A-scan data may be used to determine vP, vR, vS and probe separation S.
Still referring to
θc=θw−α (7)
where θw is the wedge angle of wedge 18. The applied delay δtn is given by:
where dn is the distance CE′ between element (1) and element (n).
By inserting equations (2) and (7) into equation (8), the following equation is obtained:
Equation (9) expresses the applied delay δtn for an element (n) at position dn as a function of the surface wave velocity v and known wedge properties vw and θw. According to the method of the present disclosure, a set of reception delays δtn is determined so that equation (6) is satisfied, meaning that the time-of-flight tAn is constant and angular offset α corresponds to reception at critical angle θc. Equation (9) may then be used to determine surface wave velocity v.
Note that for the particular case of wedge 18 being a flat wedge, where θw=0, equation (9) reduces to:
so that the measurement of the surface wave velocity v is independent of the wedge velocity vw. The negative sign in equation (10) indicates that element (1) is delayed with respect to element (n), unlike the case illustrated in
Note that, although in
Continuing to refer to
δtn=ε*n (11)
Noting that dn−dn-1=p, and inserting equation (11) into equation (9):
where p is the pitch of the elements of probe 14, and v=vP for P-waves or v=vR for Rayleigh waves.
Use of equation (12), together with the slope of delays determined from
Having determined vP and vR, the S-wave velocity vS in test object 10 may be calculated from equation (1).
Once the surface P-wave velocity vP and the appropriate reception delays δtn are known, the distance S between probes 12 and 14 may be determined by time-of-flight measurements. A reference acquisition with a single emitting element (m) on probe 12 is first performed with wedges 16 and 18 both in contact with test object 10, but with surfaces 16a and 18a of the respective wedges in contact so that separation distance S=0. The reference acquisition enables measurement of a reference propagation delay τ0 from emitting element (m) on probe 12 to any receiving element (n) on probe 14. Optionally, in order to validate that the wedge parameters are correct and applicable to both wedge 12 and wedge 14, a second reference acquisition may be performed with a single emitting element (n) on probe 14, measuring reference propagation delay τ0′ from emitting element (n) on probe 14 to any receiving element (m) on probe 12. If wedge parameters are valid, measurements of τ0 and τ0′ should be in close agreement, and an average value may be used in equation (13) below.
Wedges 16 and 18 are then moved apart by distance S and acquisition measurements are repeated, measuring propagation delay tAn from a single emitting element (m) on probe 12 to any receiving element (n) on probe 14, and optionally a propagation delay tAn′ from a single emitting element (n) on probe 14 to any receiving element (m) on probe 12. If wedge parameters are valid, measurements of tAn and tAn′ should be in close agreement, and an average value may be used to derive the constant time-of-flight value τP in equation (6) applied to P-waves. The separation S may then be determined from the following equation:
S=νP*(τP−τ0) (13)
Note that equation (13) assumes that vp does not change significantly in the time between measurements of τP and τ0.
Note that an alternative method of data acquisition according to the present disclosure is to sequentially fire multiple emitting elements (m) in probe 12. Noting that critical angle θc must be the same irrespective of the emitting element (m), and acquiring data at all elements (n) for transmissions at each of elements (m), receiving delay δtn may be determined for each of the emitting elements (m) in a manner analogous to the method described above in connection with
Step 202, the step of performing a reference acquisition, further comprises the steps shown in
Step 206, the step of measuring the P-wave velocity, further comprises the steps shown in
Step 210, the step of measuring the Rayleigh velocity, further comprises the steps shown in
The known parameters in known parameter unit 13 comprise the number N of the receiving elements of the receiving probe, the pitch p, the wedge angle θw′ of wedge 16, the wedge angle θw of receiving wedge 18, and the wedge velocity Vw.
P-wave velocity unit 4a comprises a P-wave region of interest estimator 42a for estimating regions of interest for P-waves in the A-scans, a P-wave delay finding unit 44a for finding a set of delays δtn|P that best matches the condition tAn=τP for P-waves, and a P-wave velocity calculator 46a for calculating the P-wave velocity in test object 10 based on equation (12) and the measured delays. Output from P-wave velocity unit 4a is vP, the P-wave velocity in test object 10.
Rayleigh velocity unit 4b comprises a Rayleigh region of interest estimator 42b for estimating regions of interest for Rayleigh waves in the A-scans, a Rayleigh delay finding unit 44b for finding a set of delays δtn|R that best matches the condition tAn=τR for Rayleigh waves, and a Rayleigh velocity calculator 46b for calculating the Rayleigh velocity in test object 10 based on equation (12) and the measured delays. Output from Rayleigh velocity unit 4b is vR, the Rayleigh velocity in test object 10.
S-wave velocity calculator 60 receives input of vP from P-wave velocity unit 4a and vR from Rayleigh velocity unit 4b, and using equation (1) calculates vS, the S-wave velocity in test object 10.
Probe separation calculator 50 receives input of vP and τP from P-wave velocity unit 4a, and reference propagation delay τ0 from reference acquisition unit 6, and using equation (13) calculates probe separation S between wedges 16 and 18.
As shown in
Zero separation A-scan acquisition unit 70 is configured to fire a selected element of probe 12 and to acquire A-scans from each element (n) of probe 14. Reference region of interest estimator 72 is configured to estimate regions of interest for P-waves in the A-scans, and reference delay finding unit 74 is configured to find a set of delays δtn|reference that best matches the condition that time-of-flight tAn=τreference for P-waves at zero separation, where τreference is a constant equal to τ0, the reference propagation delay. The reference propagation delay to is output from reference acquisition unit 6 to probe separation calculator 50. Probe separation calculator 50 operates when probe separation is non-zero, and calculates the value of probe separation S according to equation (13).
Note that the outputs of probe separation and velocity measurement system 1 are the acoustic velocities in test object 10 necessary for TFM time-of-flight calculations using FMC data. These velocity values pertain to the actual material condition of test object 10 and therefore provide more accurate and reliable TFM imaging than prior art methods in which velocity is measured in a calibration block, rather than in test object 10 itself. Probe separation and velocity measurement system 1 also provides an actual measurement of probe separation for the FMC acquisition, probe separation being another sensitive determining factor for time-of-flight calculations.
It should be noted that use of pitch-catch mode with two probes is an embodiment of the invention, but that other embodiments including use of a single probe comprising both emitting and receiving elements are also within the scope of the present disclosure.
Note also that, although the method of
Although the present invention has been described in relation to particular embodiments thereof, it can be appreciated that various designs can be conceived based on the teachings of the present disclosure, and all are within the scope of the present disclosure.
Number | Name | Date | Kind |
---|---|---|---|
4462257 | Gerhart | Jul 1984 | A |
4524621 | Yamanaka | Jun 1985 | A |
5974889 | Trantow | Nov 1999 | A |
5987991 | Trantow | Nov 1999 | A |
7237438 | Umbach | Jul 2007 | B1 |
20110179873 | Schubert | Jul 2011 | A1 |
Number | Date | Country | |
---|---|---|---|
20180284069 A1 | Oct 2018 | US |