10/16/2009

ANDE Course - Phased Array Ultrasonics

October 08, 2009

C.V.Krishnamurthy

Overview

• Phased Array probe

• Beam forming
• Array probe configurations
– Linear
– Matrix
– Circular
– Sectorial-annular

• Probe modeling

1
 10/16/2009

Array Probe
A
e
A – aperture
e – element width
g – gap
L p – pitch
L – element length

p g
• An array is basically a large single element transducer,
which has been subdivided by cutting it into small segments

• Typical element sizes are from 0.02 inches to 0.1 inches,

although custom sizes are available.

Why not an array of large elements?

A large probe will give a
good flat coverage, but its
small beam angle limits
its "visibility".

Recall that the 6 dB beam

divergence is given by

sin   0.5 
A

Aperture 6 dB beam width

6.35 mm (0.25 inch) 4.52
12. 7 mm (0.5 inch) 2.26 
25.4 mm (1.0 inch) 1.13

2
 10/16/2009

Why divide the probe into small elements?

A small element has a much larger
beam divergence angle, and it is this
large angle which opens up the
useful features of arrays such as
dynamic focusing and beam
steering.

Another feature of small elements is

their energy transfer efficiency -
smaller elements take less energy to
excite and are more efficient
receivers due to the lower mass to
be energized.

A B C
Beam divergence is also a function
of frequency, lower frequencies will
give more divergence than higher.

Typical element sizes/frequencies for industrial applications

are 1mm wide for 2.5 MHz and 0.5 mm wide for 5 MHz

Overlapping beams using

Phased Array Elements

Small Flaw

3
 10/16/2009

Phased Array Beamforming

Beamforming requires precise pulsing and time delays.

Receiving is the reverse of pulsing.

Array Probe Head

The array head module includes

16, 32, 64, or 128 elements
(dependant on array type)

A separate pulser and

preamplifier for each element,
together with a multiplexer, which
connects up to 8, 16, or 32
elements to create a
virtual probe.

The output from the virtual probe

is connected to 8, 16, or 32 coaxial
wires,which connect the module
to the main evaluation electronics.

4
 10/16/2009

Typical Parameters

• Max No. of elements in system The pulser fire signals can be

256 delayed from 0 to 2.5 s, in
• Max No. of elements to fire as steps of 2.5 ns.
one group 32
• Pulser Voltage 50V fixed spike The returned RF echoes from
• Amplifier Bandwidth 0.25 - 20 each channel are amplified
MHz +/- 10 dB in steps of 0.1 dB
• Max PRF 20 kHz and are digitized at 50 MHz.
• Pre amplifier gain fixed 6 dB
• Digitization resolution 50 MHz The digitized echoes are
delayed from 0 to 2.5 s in
steps of 2.5 ns. This process is
entirely digital.

Typical Array Probe Types

A Linear array is a series of transducer elements
aligned in a single housing, typically a rectangular
single element that has been segmented into smaller
individual elements.

A Curved array is similar to the linear array

with the elements curved to produce a desired
beam shape or conform to the geometry of the
part under test.

An Annular array is a series of

concentric ring elements contained
in a single housing. Imagine a single
round element being divided into
multiple individual concentric rings.

5
 10/16/2009

Other Array Configurations

Electronic Linear Scan

6
 10/16/2009

Electronic Focusing

Electronic Steering and Focusing

7
 10/16/2009

Uniform Arrays - I

Uniform Arrays - II

8
 10/16/2009

Uniform Arrays - III

Uniform Arrays - IV

9
 10/16/2009

Array Beam Characteristics - I

A2
Near-field to far-field distance N
4

6 dB angular beam divergence sin   0.5 
A
F
Focusing power (when focusing S
option is used) N
F
Beam dimension at focal
distance (in steering plane)
d st   
A

A is the dimension of the active aperture

Recap: Depth of Field

10
 10/16/2009

Strong and Weak Focusing: Example

Linear array probe pitch p = 1 mm,
frequency f = 5 MHz
Number of
active 10 16 32
elements
Active
Aperture 10 16 32
(mm)
N (mm) 84 216 865
F (mm) 84 84 84
S 0.99 0.39 0.10
d (in mm)
at F
2.49 1.55 0.78

Calculations for water medium

(v = 1480 m/s)

Array Beam Characteristics - II

• Beam width (main beam) 

sin   0.5  A
determined by active A
aperture A e


• Steering width determined sin  st  0.5  L
by element width e e

• Angular position of lobes  p g

determined by frequency f  lobe  
and pitch p:
p

11
 10/16/2009

Focusing using Phased Array - I

Focal depth: 4 mm Focal depth: 8 mm

Focusing using Phased Array - II

Focal depth: 12 mm Focal depth: 16 mm

12
 10/16/2009

How many elements are needed?

(Beam Formation – I)

Simulations at 5 MHz in Al

How many elements are needed?

(Beam Formation – II)

Simulations at 5 MHz in Al

13
 10/16/2009

15 MHz / 6mm x 0.2mm / 0.3mm spacing / 1.5” focus

Beam quality
1
1
0.9 8 Elements
0.9

0.8
0.8

0.7 0.7 Grating Lobes

0.6 0.6

0.5 0.5

0.4 Focus 0.4

0.3 0.3
0.2 0.2
0.1
8 Elements 0.1
0
10 20 30 40 50 60 70 80 90 0
-30 -20 -10 0 10 20 30
1
1
0.9
0.9 16 Elements Narrower
0.8
0.8 Main Lobe
0.7
0.7
0.6
0.6
0.5
0.5
0.4
0.4
0.3 Focus
0.3
0.2
0.2
0.1
16 Elements 0.1
0
10 20 30 40 50 60 70 80 90 0
-30 -20 -10 0 10 20 30

On Axis (mm) Off Axis (mm)

Effect of Steering Angle on Directivity

N = 16, d = /2

14
 10/16/2009

Directivity for small d/

Grating Lobes
For an N-element array, inter-element spacing d,
time-delay between adjacent elements , the
steering angle is given by
 c 
 s  sin 1  
 d 

Directivity is a product of the directivity of

discrete line sources H2( ), and the directivity of
a single element H1( )

H ( )  H1 ( )  H 2 ( )

Example plot for N=16,

For sufficiently small e/, H( )  H2( ) c = 5850 m/s, f = 2.3 MHz,
and is given by e = /100, d = 2/(1+3) and
s = 30

15
 10/16/2009

Directivity for large d/

Avoiding Grating Lobes

Maximum inter-element
spacing without producing
grating lobes

Note: When N is large, for a 60 sector scan corresponding to

maximum steering angles 30, even an inter-element spacing of
2/3 does not produce grating lobes.

Maximum steerable angle

given the inter-element
spacing d, and the
number of elements N

16
 10/16/2009

Modeling a 1-D Phased Array

64 rectangular elements
Centre Frequency: 7.5 MHz
Overall lateral dimension: 19.1 mm
(Near –field distance in Steel is about 115 mm)
Medium: Steel

Normal Incidence Beam Steered to 39º Focused On-axis Focused & Steered to 39º
0 0 0 0 1
Axial Distance (mm)

20 20
20 20
40 40
40 40
60 60

80 60 80 60

100 100
80 80
120 120 0
-60 -40 -20 0 20 40 60 -60 -40 -20 0 20 40 60 -60 -40 -20 0 20 40 60 -60 -40 -20 0 20 40 60
Lateral Distance (mm)

Steering and Focusing

- Beam Quality in the Far Field of the Array

17
 10/16/2009

Steering and Focusing

- Beam Quality in the Near Field of the Array

Regular and Random 2D - Arrays

18
 10/16/2009

Random and Optimized 2D - Arrays

Array probes on Wedges

19
 10/16/2009

Focal law for Wedge - I

Refraction
point X axis or Scan axis

Interface
Law scan offset

Depth

Refracted Angle

The calculator searches the Snell point. It considers the center of the active aperture
(from elements 2 to 7 in this example). Then, the X, Z point of the focal point is
determined. The wedge delay is calculated and the focal law is offset accordingly.

Focal law for Wedge - II

X axis or Scan axis

Interface
Time

Sound path (time)

Focal point (X,Z)
In wedge
In material Element number

20
 10/16/2009

Examples of PAUT Applications

• Relative Arrival Time Technique (RATT)

• Absolute Arrival Time Technique (AATT)
• Linear scan - Cruciform case
• Dynamic depth focusing
• Sectorial scan
• Synthetic Aperture Focusing Technique
• Advantages of phased array
• Limitations

Tandem Scans

Phased arrays allow for dynamic scanning using the tandem technique.
Separate array groups are defined as transmit and receive "virtual
probes" and scanned to cover the test area.
This technique can be used for testing weldments in thick sections.

21
 10/16/2009

Dynamic Depth Focusing

Uses the basic focusing techniques, but sequentially

focuses at various depths to cover the thickness of the part
to be tested.
Useful with linear arrays for a line scan effect, or annular
arrays to give a point focus effect

Dynamic Depth Focusing

Phased array with specified Phased array with

focal depth dynamic depth focusing

DDF is useful for inspecting thick components in a single pulse.

The beam is refocused electronically on its return.

22
 10/16/2009

Tube Inspection

Segment arrays for

large pipes

Rotating water
system segment
arrays for mid-size
pipes

Rotating water system

encircling arrays for
small size tubes.

Relative Arrival Time Technique - Principle

h  (CD  AB) cos

Ultrasound path between crack tip and corner trap signals

for a surface crack.

23
 10/16/2009

RATT – Notches in Large MS Pipes

Actual Estimated Estimated
Depth (mm) from from
Simulations Experiment
3 3.1 2.9

5 4.8 4.7

7 7.2 7.4

Comparison of the simulated and experimental

estimated notch sizes obtained by RATT on 10-mm
thick mild-steel pipe sample at 45 angle of incidence

RATT – Inspection Angle

Comparison of simulated and experimental B-scan images of 7-mm bottom surface crack
obtained for the various angles of incidence. (a) 35, (b) 45 and (c) 55 angle inspections

24
 10/16/2009

Example of Signal Processing in RATT

L. Satyanarayan et al., Inverse method for detection and sizing of cracks in thin sections ..., Theor.
Appl. Fract. Mech. (2008)

AATT - Principle

h  UT2 cos  2  UT1 cos1

25
 10/16/2009

AATT Examples

Estimated crack length: 4.9 mm (5 mm)

Angle of Inspection: 49

Estimated crack length: 6.4 mm (6.3 mm)

Top tip Inspection Angle: 74.5
Corner Inspection Angle: 42.5

Example: Monitoring Crack Growth

PA

A
B
C
Steel specimen
Schematic
Snapshot

Scanned Images
10
A 10
20
B A
30
B
20

C
40

30
C
50

60
40

70

50
80

90

Experiment 60 Simulation
10 15 20 25 30 35 40 10 15 20 25 30 35 40

Nondestructive PA results compared with that of

Applications in the Fatigue Crack Growth Studies Destructive Dye penetrant test
in Large Nuclear Components - BARC

26
 10/16/2009

S-scan
S-scans are stacked A-scans

Detection of four side-drilled holes (SDHs)

(a) Sectorial scanning, (b) S-scan view using 30

Examples of S-scan

Turbine Blade Root

Turbine Welded Rotor

27
 10/16/2009

Volumetric Weld Coverage - I

TOFD is sensitive to all defects

including volumetric defects

TOFD has dead zones near surfaces

PE complements TOFD

The combination covers 100% of the

weld volume

Alternative to RT (ASME CC2235,

AWS)

Volumetric Weld Coverage - II

TOFD
PE 45 SW

PE 60 SW

28
 10/16/2009

Volumetric Weld Coverage - Data Visualisation

Probe Movement

45-SW 60 SW TOFD 60-SW 45-SW

Synthetic aperture focusing

with Phased Array
Individual A- Scans
Time

Small Flaw

29
 10/16/2009

Example of SAFT
Element by element pulse-echo from a 1.5 mm dia SDH in Al
100 100 100
100

200 200
200 200

300 300
300 300

400 400
400
400

RAW
500 500

Image
500
500

600 600
600
600

10 20 30 40 50 60 10 20 30 40 50 60
10 20 30 40 50 60
10 20 30 40 50 60

100
100 100 100

200
200 200 200

300
300 300 300

400
400 400 400

500
500
SAFT 500 500

600
Image
600 600 600

10 20 30 40 50 60
10 20 30 40 50 60 10 20 30 40 50 60 10 20 30 40 50 60

X: 45
1 1 Y: 0.9681
X: 27 X: 27 X: 36 1
Y: 1 Y: 1 X: 45
Y: 0.9146 X: 43 X: 25 X: 35 1
Y: 1 Y: 1 Y: 0.9315
Y: 0.9681 X: 35 X: 49
0.9 0.9 X: 44 X: 27 Y: 1 Y: 1
Y: 0.9255 Y: 0.9589
X: 24 X: 34 0.9 X: 43
X: 47
X: 29 X: 40 Y: 0.9505
Y: 0.8902 Y: 0.8171 Y: 0.9178 0.9
Y: 0.8049 Y: 0.8049 X: 36
Y: 0.9146
0.8 0.8 X: 42 X: 22
Y: 0.8298 0.8 Y: 0.8764
X: 35
0.8
Y: 0.7397
X: 40 X: 29
0.7 0.7 Y: 0.6809 Y: 0.6986
X: 32
Y: 0.6489 0.7 X: 22
Y: 0.6438
0.7

0.6
0.6 X: 27 X: 53
0.6 Y: 0.5802 Y: 0.5802
0.6
X: 36
X: 34
Y: 0.5309
0.5 Y: 0.5054 X: 18
0.5 Y: 0.4938
0.5
0.5

0.4 X: 31
0.4 X: 35
Y: 0.3422 X: 33 0.4
Y: 0.3466 0.4
Y: 0.3148
X: 36
0.3 Y: 0.2977 X: 34
0.3 X: 37 X: 37
Y: 0.254
0.3 0.3 Y: 0.2826
Y: 0.2653
X: 30
Y: 0.2619
0.2
0.2 0.2
X: 29 0.2
Y: 0.115 X: 34
X: 40 X: 32
Y: 0.1011 X: 40
Y: 0.112 X: 39 Y: 0.1241
0.1 X: 33 X: 32
Y: 0.1107
Y: 0.09162
0.1 Y: 0.0715 0.1 Y: 0.07272 0.1

0
0 10 20 30 40 50 60 70
0 0 0
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

13 mm deep SDH 19 mm deep SDH 26 mm deep SDH 45 mm deep SDH

Advantages of Phased Arrays

• Inspection Speed • Good coverage (Multiple scan
– Real-time images
options)
• Flexibility • Real-time images
• Wedge-based applications
• Variety of probe types
• POD ( many angles and imaging)
• Similar to conventional procedures in
TOFD
• Access to remote areas
• Lends to Image processing
• Analysis Tools techniques
• Simulations aid quantitative
• Reporting assessment
• Allows for new probe designs

30
 10/16/2009

Advantages of Digital Recording

• Permanent record of all collected data

• Various signal processing options
• Enables data to be compared throughout the service life
of a component
• Re-analysis of raw data, at any time
• Variety of visual displays available

All the usual ultrasonic limitations

• Coupling
• Frequency, attenuation etc.
• Acoustic impedance mismatch requirements
• Dead zones

31
 10/16/2009

ASME Codes

• Phased arrays specifically accepted • ASME CC 2235 (CIT) allows use of

as Computerized Imaging UT instead of RT for wall thickness
Techniques > 12.7mm

• Code cases for manual S-scans and • TOFD, PA and TOFD/PE techniques
E-scans first submitted Feb 2006. allowed
Now approved.
• Phased arrays specifically approved
• Code cases for encoded E-scans for ASME CC
and S-scans expected for August
2006. Now Approved

• Mandatory phased array appendix

being drafted concurrently. Expect
approval of appendix in a few years

• Phased arrays currently being

approved through Performance
Demonstration approaches, e.g.
Article 14 and ASME code case
2235

32