Eddy Current Array Probes for Crack Detection and Sizing in Carbon ...

Eddy Current Array Probes

for Crack Detection and Sizing

in Carbon Steel Welds

Middle East Steel Conference

Dubai, UAE

October 21-23, 2014

Presented by:

Keith Sinclair

Director of Sales, India, Middle East & Africa

M. Sirois, S. Parmentier, N. Decourcelle, M. Grenier

Eddyfi®

National Seminar & Exhibition on Non-Destructive Evaluation, NDE 2014, Pune, December 4-6, 2014 (NDE-India 2014)

Vol.20 No.6 (June 2015) - The e-Journal of Nondestructive Testing - ISSN 1435-4934www.ndt.net/?id=17843

CONFIDENTIAL

Content

1. What are Eddy Current Arrays?

2. Why Carbon Steel Welds?

3. Development Objectives

4. ECA design: configuration, lift-off, sizing…

5. Our solution

6. Trials on Welded Specimen

7. Towards a Solution

8. Conclusions

2 © Eddyfi NDT, Inc., 2014

CONFIDENTIAL3

Slow raster scan

(single-probe,

conventional

ECT)

Conventional eddy current

testing of carbon steel

weld (single sensor/probe)

• Only basic signals available for interpretation (no imaging, no C-Scan)

• Limited data storage (traces are saved)

• Lots of manual scanning involved: time-consuming, dexterity required

• No reliable depth sizing possible

© Eddyfi NDT, Inc., 2014

1.What Are Eddy Current Arrays?

CONFIDENTIAL

• An eddy current array (ECA) is composed of several

individual coils, grouped together in one probe.

• The coils are excited in a specific sequence to eliminate

interference from mutual inductance (channel multiplexing).

• Arrays can be made flexible or shaped to any geometry.

• Software provides graphical display (2D and 3D C-scans).


4

Fast single-pass

(ECA)


CONFIDENTIAL

Typical benefits:

Reduced Inspection Time

Improved Flaw Detection (POD)

More Intuitive Interpretation

Full Inspection Records

5



CONFIDENTIAL

2.Why Carbon Steel Welds?

6

Abundant & Critical

cc cc


CONFIDENTIAL

Vulnerable & Challenging

7


cc

cccc


CONFIDENTIAL

• Penetrant Testing / Magnetic Particle Testing

• Surface preparation required, coating removal required

• Long: pre-clean, application, dwell-time, developer, evaluation, cleaning…

• On-site, immediate interpretation required

• Crude reporting and record-keeping capabilities (ruler & picture)

• Environmental concerns related to the use of chemicals

• No depth sizing possible

8



CONFIDENTIAL

• Detect surface-breaking cracks in CS joints

• Length and depth sizing

• Accommodate weld crown in the ”as is” condition

• Lift-off tolerance: coating/paint up to 2-3mm (0.080”-0.120”)

• Maximize coverage and speed

• Leverage advanced imaging capabilities (C-Scan, 2D/3D)

The solution:

A dedicated probe leveraging eddy current array (ECA)

technology for the examination of carbon steel welds.

3.Development Objectives


CONFIDENTIAL

3.Development Objectives

10

Lift

• Surprisingly, defects within a range

of 0.5mm to 15mm in depth were

easily distinguishable.

• However, defects signals curvature

could complicate depth sizing

using either vertical or horizontal

component of signals.

• As lift-off occurs during weld

scans, the operating point moves

along this hooked curve producing

significant phase changes.

…cannot be reached with a Transmit-Receive (TR)

configuration using pancake coils side-by-side:


CONFIDENTIAL

4.ECA Design:

11

A specific Transmit-Receive (TR) configuration

• Defects at 90 degrees with lift-off

• All defects have the same phase shift

• Lift-off “flatness”

25kHz 100kHz 50kHz


CONFIDENTIAL12

• Lift-off monitoring is possible:

lift-off curve using horizontal amplitude (50kHz)

3mm 2mm 1mm 0mm

4.ECA Design: Lift-Off Monitoring


CONFIDENTIAL © Eddyfi NDT, Inc., 201413

• By measuring the horizontal offset of the operating

point during scans, lift-off can be measured and

depth sizing then corrected

4.ECA Design: Lift-Off Monitoring

CONFIDENTIAL14

4.ECA Design: Depth Sizing

1mm

2.5mm

5mm

10mm

• Note: EDM notches used

during these tests were 1mm,

2.5mm, 5mm, 10mm,

12.5mm, 15mm and 20mm

deep, all 25mm in length.

• Until 12.5mm in depth

(2:1 length/depth ratio) sizing

remains good. However, for

ratios under 2:1 (rarely the

case for real cracks), EC

would tend to use the shortest

path and to bypass notches in

surface rather than « diving »

under it.


CONFIDENTIAL15

• Using the vertical component of the signals, depth

sizing is possible:



CONFIDENTIAL

• The idea here is to correct depth sizing depending on

lift-off variations.

• Based on the orthogonality between defects and lift off

signals, any horizontal offset might be considered as a

potential lift off causing a decrease in amplitude on a

given defect.

• From there, multiple depth curves need to be calibrated

with various lift-off values, i.e. 0mm, 1mm, 2mm and

3mm.

• Using these multiple depth curves, a 3D depth “plan”

could be plotted with depth vs. vertical amplitude

vs. lift-off.

16



CONFIDENTIAL17

4.ECA Design: Depth Sizing Correction

• Thus, using the following depth curves with lift-off at

0mm and 1mm…


CONFIDENTIAL18


• And these at 2mm and 3mm…


CONFIDENTIAL19

• We can plot a 3D depth curve vs. lift-off:



CONFIDENTIAL

• For example, here is a defect 5mm deep with 1mm of

lift-off:

20


Lift-off sizing Depth sizing


CONFIDENTIAL21

• One of the receiving coils is used to size the length :

This defect is 7.7 mm long and has been

sized at 7.5 mm using the encoder.

4.ECA Design: Length Sizing


CONFIDENTIAL22

Validating:

• EC behaviors around simulated defects

• Lift-off and impact on EC uniformity

• Aspect ratios of crack-like defects

• Coil optimization

Finite-Element Analysis (FEA) with COMSOL

0mm 1mm 5mm

10mm

4.ECA Design: Validation


CONFIDENTIAL23

Finite-Element Analysis (FEA) with COMSOL

4.ECA Design: Validation


CONFIDENTIAL25

5. Our Solution – “Sharck™”


Patent-pending

CONFIDENTIAL

• Single-V, as welded

• Toe cracks and HAZ crack

26

6.Trials on Welded Specimen

• Carbon Steel

• 9.5 mm (3/8”) thick


CONFIDENTIAL27



Flaw 1

Flaw 2

Flaw 3

Flaw 1

Flaw 2

Flaw 3

CONFIDENTIAL28


• First step: lift-off monitoring

→ 0.2mm thus, the 0mm lift-off curve must be selected

Sizing Flaw #1


CONFIDENTIAL29


• Second step: depth sizing with vertical amplitude

Sizing Flaw #1

→ 1.3mm depth (real depth is 1.25mm)


CONFIDENTIAL30


• Third step: length sizing

Sizing Flaw #1

→ 3.5mm length, real length is 3.2mm


CONFIDENTIAL31



CONFIDENTIAL

What’s next?During the upcoming months, a thorough validation of the

probe capabilities is taking place.

32

• Performance demonstrations, assessment of limitations

• New software tools including:

3D depth curves

New filtering processes

Special measurement cursors

Auto-calibration feature

And more!

7.Towards a Solution…


CONFIDENTIAL

8.Conclusions

• An eddy current array technique has been developed

for the detection and sizing of surface-breaking

cracks in carbon steel welds;

• A transmit-receive, multiplexed array probe provides

adequate signals and even allows detecting

transverse cracks;

• Results are promising, and capabilities in terms of

tolerance to paint/coating (lift-off), length and depth

sizing are being investigated;

• Additional work remains to thoroughly assess the

capabilities and limitations of ECA technology in the

context of carbon steel welds.


Keith Sinclair

Director of Sales - India, Middle East & Africa

[email protected]

www.eddyfi.com

Eddy Current Array Probes for Crack Detection and Sizing in Carbon ...

Documents

Transcript of Eddy Current Array Probes for Crack Detection and Sizing in Carbon ...