11
A Phased Array Technique for Crack Characterization Philippe BREDIF, Clarisse POIDEVIN, CEA/LIST Saclay France Olivier DUPOND, EDF R&D France Abstract. The last advances in ultrasonic phased array transducers and systems offer new possibilities in NDT. This paper presents a phased array technique developed in order to improve cracks sizing, especially when the crack is under stress. In that case, crack lips are compressed and diffraction echoes can be difficult to detect. A method based on a phased array technique was developed to improve the characterization of cracks. The principle of the method is based on the detection of small diffracting areas along the flaw. It is implemented with a large aperture transducer offering a high spatial resolution. The method consists in detecting the emerging part of the crack and examining the crack from the lower part to the top by moving the beam depth focusing. Adequate delay laws are computed to adapt depth focusing and to enhance signal to noise ratio on diffracted signals. Experimental results demonstrate that the use of a specific phased array transducer improves significantly the crack depth sizing accuracy, independently of the load applied on the crack. 1 Introduction Phased-array techniques offer a large flexibility in NDT inspection thanks to the application of amplitude and delay laws which allows controlling the beam features such as depth focusing and/or refracted angles. A method based on phased array techniques was developed in order to improve crack depth sizing. The motivation for such a development is that, according to previous basic studies, in the eventuality of closed cracks, tip diffraction echoes may be difficult to detect with standard transducers. This paper details the method of inspection developed to improve crack depth sizing. A very high-spatial resolution phased array transducer was developed. This transducer was used to study the influence of an external load on the detectability of the crack. Finally, high-spatial resolution benefits were experimentally evaluated by comparison with a conventional focused transducer. 2 Inspection method The method is based on the detection of small diffracting points. These small points result from a non perfect contact between the two surfaces of a crack even if it is closed. The detection of these points is performed using a focused transducer with high-spatial resolution. The characterization is done in two steps. First, we detect the base of the crack by focusing the energy at the backwall surface level. We then scan the ultrasonic beam along the crack until detection of the top of the crack. In the following example, a 10 mm partially closed crack was characterized. Four successive delay laws were applied to focus the energy at different depths: 30 mm ECNDT 2006 - Th.1.1.2 1

A Phased Array Technique for Crack Characterization · 2006. 9. 30. · A Phased Array Technique for Crack Characterization Philippe BREDIF, Clarisse POIDEVIN, CEA/LIST Saclay France

  • Upload
    others

  • View
    6

  • Download
    0

Embed Size (px)

Citation preview

Page 1: A Phased Array Technique for Crack Characterization · 2006. 9. 30. · A Phased Array Technique for Crack Characterization Philippe BREDIF, Clarisse POIDEVIN, CEA/LIST Saclay France

A Phased Array Technique for Crack Characterization

Philippe BREDIF, Clarisse POIDEVIN, CEA/LIST Saclay France Olivier DUPOND, EDF R&D France

Abstract. The last advances in ultrasonic phased array transducers and systems offer new possibilities in NDT. This paper presents a phased array technique developed in order to improve cracks sizing, especially when the crack is under stress. In that case, crack lips are compressed and diffraction echoes can be difficult to detect. A method based on a phased array technique was developed to improve the characterization of cracks. The principle of the method is based on the detection of small diffracting areas along the flaw. It is implemented with a large aperture transducer offering a high spatial resolution. The method consists in detecting the emerging part of the crack and examining the crack from the lower part to the top by moving the beam depth focusing. Adequate delay laws are computed to adapt depth focusing and to enhance signal to noise ratio on diffracted signals. Experimental results demonstrate that the use of a specific phased array transducer improves significantly the crack depth sizing accuracy, independently of the load applied on the crack.

1 Introduction

Phased-array techniques offer a large flexibility in NDT inspection thanks to the application of amplitude and delay laws which allows controlling the beam features such as depth focusing and/or refracted angles. A method based on phased array techniques was developed in order to improve crack depth sizing. The motivation for such a development is that, according to previous basic studies, in the eventuality of closed cracks, tip diffraction echoes may be difficult to detect with standard transducers. This paper details the method of inspection developed to improve crack depth sizing. A very high-spatial resolution phased array transducer was developed. This transducer was used to study the influence of an external load on the detectability of the crack. Finally, high-spatial resolution benefits were experimentally evaluated by comparison with a conventional focused transducer.

2 Inspection method

The method is based on the detection of small diffracting points. These small points result from a non perfect contact between the two surfaces of a crack even if it is closed. The detection of these points is performed using a focused transducer with high-spatial resolution. The characterization is done in two steps. First, we detect the base of the crack by focusing the energy at the backwall surface level. We then scan the ultrasonic beam along the crack until detection of the top of the crack. In the following example, a 10 mm partially closed crack was characterized. Four successive delay laws were applied to focus the energy at different depths: 30 mm

ECNDT 2006 - Th.1.1.2

1

Page 2: A Phased Array Technique for Crack Characterization · 2006. 9. 30. · A Phased Array Technique for Crack Characterization Philippe BREDIF, Clarisse POIDEVIN, CEA/LIST Saclay France

(backwall surface), 25, 20 and 15 mm. For the 4 laws, the phased-array probe generates 45° shear waves. Figure 1 illustrates the principle of inspection.

Figure 1.Principle of inspection

3 A wide aperture phased array design

A wide aperture phased-array probe with high-spatial resolution was designed using the CIVA software. This transducer radiates 45°-shear waves in steel and naturally focuses the energy at a depth of 25 mm. The active area is circular (Ø 100 mm) and composed of 121 elements working at the frequency of 4.5 MHz. The following figure describes the phased-array transducer.

Figure 2. Phased array description.

The wide aperture combined with the natural focusing of the probe provides a very high spatial resolution. The focal point, less than 1 mm in the focal area, can be moved along the direction of refraction using electronic delay laws (electronic focusing). A MultiX UT acquisition system (128 parallel US channel) is used to monitor the 121 elements of the probe. Figure 3 shows a CIVA simulation of the ultrasonic field radiated by this transducer and focused at 25 mm.

Phased array

Mechanical focusing

Sectorial elements

Top view

crack

Focal points

15

20

25

30

45° SW

zoom

Scan axis

Focal depthdisplacement

Backwall surface

2

Page 3: A Phased Array Technique for Crack Characterization · 2006. 9. 30. · A Phased Array Technique for Crack Characterization Philippe BREDIF, Clarisse POIDEVIN, CEA/LIST Saclay France

Figure 3. Ultrasonic field simulations at the nominal depth.

4 Experimental results on mock-up

Experimental results are shown in Figure 4 on a 10-mm high fatigue crack. For each position of focusing, the figure shows the simulated field radiated in the component and the corresponding experimental Bscan. When the beam is focused 30 mm deep, we only detect the corner echo at the base of the crack (a). When the focal point moves from the backwall surface to depths of 25 and 20 mm, diffraction echoes are detected along the crack (b and c). Finally, diffraction echoes disappear for a focal depth of 15 mm (d) for which the focal point is above the top of the crack. Detection of the diffraction at the top of the crack is maximal for a focal depth of 20 mm.

∅0.8 mm

Ultrasonic field transmitted in the incident plane

3 mm

25 mm in depth

Ultrasonic field transmitted in the plane perpendicular to the

incident plane

3

Page 4: A Phased Array Technique for Crack Characterization · 2006. 9. 30. · A Phased Array Technique for Crack Characterization Philippe BREDIF, Clarisse POIDEVIN, CEA/LIST Saclay France

Figure 4. Example of experimental results. a) Focusing at 30 mm in depth; b) Focusing at 25 mm in depth; c) Focusing at 20 mm in depth; d) Focusing at 15 mm in depth

4.1 Effects of an external load

4.1.1 Test description

Ultrasonic inspections were made on a block containing a fatigue crack. The block was positioned on a four-point bending rig to apply a compressive stress on the crack.

True BSCAN

Corner echo

Diffraction

5.5

Focusing at 25 mm in depth

SW45°

dept

h

Scan axis

25

Ultrasonic beam simulationb)

Focusing at 20 mm in depth

SW45°

dept

h

Scan axis

True BSCAN

Corner echo

Top diffraction

10

Ultrasonic beam simulation

20

c)

True BSCAN

Corner echo

No diffraction detected

Ultrasonic beam simulation

Focusing at 15 mm in depth

dept

h

Scan axis

15 SW45°

d)

Ultrasonic beam simulation

dept

h

Scan axis

No diffraction detected

Focusing at the backwall - 30 mm in depth

30

SW45°

True BSCAN

Corner echo

a)

crack

4

Page 5: A Phased Array Technique for Crack Characterization · 2006. 9. 30. · A Phased Array Technique for Crack Characterization Philippe BREDIF, Clarisse POIDEVIN, CEA/LIST Saclay France

Ultrasonic scans were made for three different loads: no load, a 90 kN load and a 180 kN load. Figure 5 describes the assembly test for loading.

Mock-up 1386-T216-M1 Material 304L l x L x h (mm3) 400 x 70 x 29 Nominal height of the crack (mm) 10 Opening at the base of the crack (µm) 50 Opening at the tip of the crack (µm) 0.5 Kmin (MPa√m) 17 Kmax (MPa√m) 28

Table 1: Mock-up description and cracking configuration

Figure 5. Assembly test for loading

4.1.2 Acquisition without load

Results obtained when no load is applied are shown in Figure 6. The transducer generates 45° shear waves and a delay law is applied to focus the energy 20 mm deep. It is important to note that the specimen already suffered several cycles of compression from previous experiments. The reference is the reflection echo of a 2 mm in diameter side-drilled hole located 20 mm from the backwall surface. Diffraction echoes are observed at the top of the crack (8.5 mm from the backwall surface). The mean amplitude of these echoes is 28 dB lower than the reference echo. The signal-to-noise ratio is 8 dB allowing good detection of the diffraction echoes.

Without external load

Screw-boltMock-up

0 0

Fatigue crack

Under load

Tightening bolts

Closed crack

4 points bending system

Δ Δ

Δ = -0.3 mm ~90 kNΔ = -0.6 mm ~180 kN

5

Page 6: A Phased Array Technique for Crack Characterization · 2006. 9. 30. · A Phased Array Technique for Crack Characterization Philippe BREDIF, Clarisse POIDEVIN, CEA/LIST Saclay France

Figure 6. Results of acquisition without load. Focusing depth at 20 mm in depth.

4.1.3 Acquisition under load - 90kN

The following figures show the results obtained for an applied load of 90 kN. Diffraction echoes are detected at the top of the crack, and also at intermediate positions at 6.5 and 7.5 mm from the backwall. The amplitudes of the top diffraction echo relatively to the reference are respectively -22 dB and -25 dB.

Figure 7. Results of acquisition obtained for a load of 90 kN. Focusing depth at 20 mm in depth.

4.1.4 Acquisition under load (180 kN)

Figure 8 shows the results obtained for an applied load of 180 kN. Diffraction echoes are detected at the top of the crack and 1.5 mm under the top. Their mean amplitudes are respectively -25 dB and -17 dB amplitude.

z

x

z

Tip diffraction

Corner

8.5

mmTip : -25 dB

Reference : side drilled hole ∅ 2 mm at 20 mm in depth

True Bscan (focusing at 9 mm from backwall)Dscan (focusing at 9 mm from backwall)y

Intermediate

6.5

mm

7.5

mm

Intermediate-22 dB

Intermediate-25 dB

y

Tip diffraction

Corner

Tip : -18 dB

z

x

z

True Bscan (focusing at 9 mm from backwall)Dscan (focusing at 9 mm from backwall)

8.5

mm

Reference : side drilled hole ∅ 2 mm at 20 mm in depth

Intermediate Intermediate-21 dB

7 m

m

zy

x

Dscancrack

zy

x

Dscancrack

6

Page 7: A Phased Array Technique for Crack Characterization · 2006. 9. 30. · A Phased Array Technique for Crack Characterization Philippe BREDIF, Clarisse POIDEVIN, CEA/LIST Saclay France

Figure 8. Results of acquisition obtained for a load of 180 kN. Focusing depth at 20 mm in depth.

To summarize these experiments, we observed that the amplitude of the diffraction echoes at the top of the crack decreases with the load. Simultaneously, the amplitude of diffraction echoes located 1.5 mm under the top increases and become greater than the top diffraction echoes. The following table contains the results obtained with the wide aperture transducer for three different loads.

Top diffraction

Height = 8.5 mm

Intermediate diffraction

Height = 7.5 mm

Intermediate diffraction

Height = 7 mm

Intermediate diffraction

Height = 6.5 mm External load

Amplitude * SNR Amplitude * SNR Amplitude * SNR Amplitude * SNR

0 kN -18 dB 18 dB - - -21 dB 15 dB - -

90 kN -25 dB 10 dB -22 dB 12 dB - - -25 dB 10 dB

180 kN -25 dB 10 dB - - -17 dB 18 dB - - (*)The reference is the reflection echo off a 2 mm in diameter side-drilled hole located 20 mm from the backwall surface.

Table 2: Amplitudes of diffraction echoes with the wide aperture transducer The following figure shows a micrograph from one side of the mockup. There is no geometrical features linked to the detection of intermediate echoes 1.5 mm under the top of the crack.

7 m

m

z

x

z

Intermediate-17 dB

Reference : side drilled hole ∅ 2 mm at 20 mm in depth

True Bscan (focusing at 9 mm from backwall)Dscan (focusing at 9 mm from backwall)y

8.5

mm

Tip : -25 dBTip diffraction

8.5

mm

Tip : -25 dBTip diffraction

Corner

Intermediate

7

Page 8: A Phased Array Technique for Crack Characterization · 2006. 9. 30. · A Phased Array Technique for Crack Characterization Philippe BREDIF, Clarisse POIDEVIN, CEA/LIST Saclay France

8 mm

1.5 mm

x 9 x 50

Figure 9. Micrograph from one side of the mock-up

4.2 High-spatial resolution benefits evaluation

To determine the benefits of a large aperture, we compare the performances of the transducer presented in this paper to a smaller aperture (∅ 50 mm). The smaller aperture is obtained by using only 25 central elements in transmission and reception. The following figures illustrate the element configurations, the ultrasonic field in the plan of incidence and perpendicular to that plan for both apertures.

8

Page 9: A Phased Array Technique for Crack Characterization · 2006. 9. 30. · A Phased Array Technique for Crack Characterization Philippe BREDIF, Clarisse POIDEVIN, CEA/LIST Saclay France

∅100 ∅50

∅ 0.8

3.525

39.5

4.5

18

∅ 100 ∅ 50

∅ 1.5

0 5 50

5

00

5

Axe T45° Axe T45°

0 35 0 35x (mm )

z (m

m)

x (mm)

x’ (mm) x’ (mm)

y’(m

m)

y’(m

m)

Figure 10. Civa simulations of the ultrasonic field radiated in the plan of incidence and in the plan perpendicular to the plan of incidence. On the left side: 100 mm aperture; on the right side: 50 mm aperture.

Experiments were made using the two different apertures, results are presented in the following figures. We notice diffraction echoes at the top of the crack and intermediate echoes for both aperture, however the amplitudes are better for the 100 mm aperture. The following table summarized all the results.

x

z

Axe T45°

Axe T45°

y’x’

Calcul dans le plan incident

Calcul dans le plan perpendiculaire à l’axe

T45°Simulation in the plan of

incidence

Simulation in the plan perpendicular to the plan of

incidence

9

Page 10: A Phased Array Technique for Crack Characterization · 2006. 9. 30. · A Phased Array Technique for Crack Characterization Philippe BREDIF, Clarisse POIDEVIN, CEA/LIST Saclay France

Top diffraction

Height = 8.5 mm

Intermediate diffraction

Height = 7 mm

Amplitude* SNR Amplitude* SNR

Observations

Aperture ∅ 100 -25 dB 10 dB -17 dB 18 dB 2e cycle

load = 180 kN

Aperture ∅ 50 -33 dB 6 dB -24 dB 15 dB 2e cycle

load = 180 kN Table 3: Amplitudes of diffraction echoes with the 100-mm aperture and with the 50-mm aperture.

The segmentation analysis shows that the diffraction echo at the top of the crack is continuously detected along the extension of the crack with the large aperture. With the smaller aperture, the detection is discontinuous. The intermediary echo located 7 mm from the backwall surface is detected in both cases.

Tip : -25 dB

corner

Intermediate : -17 dB

z

y Dscan

7 mm 8.5 mm

Segmented Dscan

z

y

Tip : -33 dB

corner

Intermediate : -24 dB

Dscan

Segmented Dscan

7 mm 8.5 mm

y

y

zz

Figure 11. Acquisition result under 180 kN load. Left side: obtained with the 100 mm aperture; right side : 50 mm aperture

We see that the biggest aperture allows to detect diffracting points that were invisible with the smaller aperture. The detection of partially closed cracks is greatly improved using a wide aperture phased-array transducer.

5 Conclusion

Previous studies showed that it could be difficult to detect diffraction echoes from the tip of a closed crack using conventional transducers. A higher spatial resolution is required to improve the detection of such signals. This paper describes a method based on phased-array techniques and large aperture transducer to improve crack depth sizing.

10

Page 11: A Phased Array Technique for Crack Characterization · 2006. 9. 30. · A Phased Array Technique for Crack Characterization Philippe BREDIF, Clarisse POIDEVIN, CEA/LIST Saclay France

The principle of the method relies on scanning the focused ultrasonic beam along the crack by applying adequate delay laws. Comparison with a conventional focused transducer has shown that crack depth sizing is greatly improved by combining high-spatial resolution with electronic scanning. The results presented in this paper demonstrate the efficiency of the technique to evaluate crack depth sizing.

11