Applications of Motorized Rotational RFEC Probes … IMTT...34th Annual Review of Progress in QNDE Golden Colorado, July 23 – July 27, 2007 Innovative Materials Testing Technologies,

34th Annual Review of Progress in QNDEGolden Colorado, July 23 – July 27, 2007

Innovative Materials Testing Technologies, Inc.

Yushi Sun, Weiqiang Wan, Changhong Sun, Xinle Yang, Haiou Zhu & Tianhe Ouyang, IMTT 1

Applications of Motorized Rotational RFEC Probes in Thick

and Multilayer Structure Crack Detection


3141 W. Torreys Peak Drive, Superior, CO 80027Tel. 303 554 8000; Fax. 303 554 8001

(The Complete Version)




Challenges1.Requirement of deep penetration in multilayer

structure

2.Structure variations around a fastener –background noise comparable or greater than crack signal

3.Material property variations, such as permeability variation of a steel fastener.

4.Extremely weak crack signal submerged in background noise and structure variation signals




RFEC ProbeA solution for Deep Penetration

Indirect Coupling Path

Pickup UnitDrive Unit Direct Coupling Path

The probe blocks the direct coupling path. The energy released from the drive unit is forced to go along the indirect coupling path.

Therefore, the entire signal received by the pickup unit has passed the wall twice and carries the whole information about the wall condition.




Rotational RFEC ProbeMinimizing Noise from Fastener

FastenerPickup coil

Drive coil

Probe head

Assembly

Rotation Guide

Lens for centering




Motor-Controlled RotationEnsuring Repeatable Signal for Online

Signal Processing

Motor

Suction base

Motor controller

Vacuum in

Suction base

RFEC Probe




Computerized SSEC InstrumentEnabling Online Signal Processing

InstrumentSSEC

Motorized RFEC Probe

PC Monitor or Laptop




Purpose of This Study

1. Investigate influence of 1st & 2nd layer edges on 2nd layer crack detection while the 1st layer is thick

2. Investigate influence of steel fastener on 2nd

layer crack detection at different edge conditions

3. Improving 2nd layer crack detect-ability using motorized rotational RFEC probes with online signal processing and/or pattern recognition




Study # I

1st or 2nd layer edge effect, but crack is a little away from edge

Ferromagnetic or non- ferromagnetic fastener




Example 1Detection 2nd layer horizontally oriented cracks

C-130 center wing box standard with steel fasteners

1.0”

1.0”

1 2 3 4 5 6 7 8 9 10 11 12

.600 .500 .400 .300 .200 .100

.600 .500 .400 .300 .200 .100

1st Layer - 3:1 Triangle Notch

2nd Layer - Rectangular Notch

•

0.25

00.

140

1 2 3 4 5 6 7 8 9 10 11 12

1st layer –0.250” Aluminum

2nd layer –0.140” AluminumWidth = 1.000”

2nd layer EDM notches –0.100” - 0.600”




2040.600” notch

2080.200” notch

2090.100” notch

2060.400” notch

2070.300” notch

2050.500” notch

211No notch

212No notch

202No notch

203No notch

210No notch

Detection of 2nd Layer Notches—impedance plane (row data)

No EDM Notch

Increase of EDM Notch Size




Detection of 2nd Layer Notches—imaginary (row data)

No EDM Notch


211No notch

212No notch

202No notch

203No notch

210No notch

2040.600” notch

2080.200” notch

2090.100” notch

2060.400” notch

2070.300” notch

2050.500” notch




0.6” EDM0.5” EDM0.4” EDM

0.3” EDM0.2” EDM0.1” EDMNo EDM

Detection of 2nd Layer Notches—Comparison (1)Before signal processing

0.6” EDM0.5” EDM0.4” EDM0.3” EDM0.2” EDM0.1” EDMNo EDM




Detection of 2nd Layer Notches—Comparison (2)Before Signal Processing

SSEC Signal

0

50

100

150

200

250

1 2 3 4 5 6 7 8 9 10 11

#212

No

Cra

ck

#203

No

Cra

ck

#211

No

Cra

ck

#210

No

Cra

ck

#204

2nd

laye

r rig

ht 0

.600

”

#205

2nd

laye

r rig

ht 0

.500

”

#206

2bd

laye

r rig

ht 0

.400

”

#207

2nd

laye

r rig

ht 0

.300

”

#208

2nd

laye

r rig

ht 0

.200

”

#209

2nd

laye

r rig

ht 0

.100

”

#202

No

Cra

ck




2040.600” notch

2080.200” notch

2090.100” notch

2060.400” notch

2070.300” notch

2050.500” notch

211No notch

212No notch

202No notch

203No notch

210No notch

No EDM Notch


Detection of 2nd Layer NotcheProcessed Using Ellipse Fitting




-10

-5

0

5

10

15

20

25

30

35

1 2 3 4 5 6 7 8 9 10 11

#212

No

Not

ch

#203

No

Not

ch

#211

No

Not

ch

#210

No

Not

ch

#204

0.6

00”

Not

ch

#205

0.5

00”

Not

ch

#206

0.4

00”

Not

ch

#207

0.3

00”

Not

ch

#208

0.2

00”

Not

ch

#209

0.1

00”

Not

ch

#202

No

Not

ch

Detection of 2nd Layer Notches - After Ellipse FittingAll Cracks Detected




Example 2Detection 2nd layer vertically oriented cracks

with titanium fasteners+Thicker 2nd layer +Close 1st edge +Triangle EDM notches

A AB BC CD D0B C D

E EF FG H H1

F G H0

1st layer - 0.250” Aluminum

2nd layer - 0.310” Aluminum

0.65

0”




2nd Layer Vertical Notches, Titanium Fastener

0.200”X0.267” 0.220”X0.293” 0.220”X0.293” 0.300”X0.313”0.170”X0.227”0.150”X0.200”0.133”X0.173”0.100”X0.133”

No EDM Notch





Signal Rotation Angle - All Cracks detected

0 5 1 0 1 5-2 0

0

2 0

4 0

6 0

8 0

f a s te n e rs

angl

e (d

egre

e) S N R a t C : 3 .4ED M : 0 .1 5 0 " ( W ) x 0 .2 0 0 " ( D )

R o w : T U

A -

0.10

0”(L

)X0.

133”

(D)

B -

0.13

3”(L

)X0.

173”

(D)

C -

0.15

0”(L

)X0.

200”

(D)

D -

0.17

0”(L

)X0.

227”

(D)

E -0

.200

”(L)

X0.2

67”(

D)

F -0

.220

”(L)

X0.2

93”(

D)

G -

0.25

0”(L

)X0.

313”

(D)

H -

0.30

0”(L

)X0.

313”

(D)

AB

–N

o ED

M N

otch

D0

–N

o ED

M N

otch

EF –

No

EDM

Not

ch

FG –

No

EDM

Not

ch

GH

–N

o ED

M N

otch

BC

–N

o ED

M N

otch

CD

–N

o ED

M N

otch

H0

–N

o ED

M N

otch





with steel fasteners

A AB BC CD D0B C D

E EF FG H H1F G H0

1st layer

2nd layer






2nd Layer Vertical Notches, Steel Fastener

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mv

Imag

inar

y m

v

Impedance Plane

Angle Threshold:10

Fastener:h4s

Angle: 7

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mv

Imag

inar

y m

v

Impedance Plane

Angle Threshold:10

Fastener:g4s

Angle: -3

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mv

Imag

inar

y m

v

Impedance Plane

Angle Threshold:10

Fastener:f4s

Angle: 5

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mvIm

agin

ary

mv

Impedance Plane

Angle Threshold:10

Fastener:e4s

Angle: 8

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mv

Imag

inar

y m

v

Impedance Plane

Angle Threshold:10

Fastener:c4s

Angle: 2

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mv

Imag

inar

y m

vImpedance Plane

Angle Threshold:10

Fastener:b4s

Angle: -1

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mv

Imag

inar

y m

v

Impedance Plane

Angle Threshold:10

Fastener:a4s

Angle: 2

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mv

Imag

inar

y m

v

Impedance Plane

Angle Threshold:10

Fastener:h2s

Angle: 49

crack !!

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mv

Imag

inar

y m

v

Impedance Plane

Angle Threshold:10

Fastener:g2s

Angle: 28

crack !!

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mv

Imag

inar

y m

v

Impedance Plane

Angle Threshold:10

Fastener:f2s

Angle: 24

crack !!

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mv

Imag

inar

y m

v

Impedance Plane

Angle Threshold:10

Fastener:e2s

Angle: 24

crack !!

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mv

Imag

inar

y m

v

Impedance Plane

Angle Threshold:10

Fastener:d2s

Angle: 18

crack !!

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mv

Imag

inar

y m

v

Impedance Plane

Angle Threshold:10

Fastener:c2s

Angle: 19

crack !!

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mv

Imag

inar

y m

v

Impedance Plane

Angle Threshold:10

Fastener:b2s

Angle: -1

-60 -40 -20 0 20 40 60-60

-40

-20

0

20

40

60

Real mv

Imag

inar

y m

v

Impedance Plane

Angle Threshold:10

Fastener:a2s

Angle: 4

No EDM Notch


Greater variation in the size of impedance planed




Signal Rotation AngleDetect-ability: Notch ≥ 0.150”

0.10

0”(L

)X0.

133”

(D)

0.13

3”(L

)X0.

173”

(D)

0.15

0”(L

)X0.

200”

(D)

0.17

0”(L

)X0.

227”

(D)

0.20

0”(L

)X0.

267”

(D)

0.22

0”(L

)X0.

293”

(D) 0.

250”

(L)X

0.31

3”(D

)

No

EDM

Not

ch

No

EDM

Not

ch

No

EDM

Not

ch

No

EDM

Not

ch

No

EDM

Not

ch

No

EDM

Not

ch

No

EDM

Not

ch 0.30

0”(L

)X0.

313”

(D)




Study # 2

2nd Layer Crack near 2nd Layer EdgeNon-ferromagnetic Fastener




Example 3Detection 2nd layer vertical cracks

with Ti fasteners, crack close to 2nd layer edge

E EF FG GH H0F G H

Upper Row1st layer - 0.250” Aluminum


A AB BC CD D0B C D

Upper Row



0.75

0”




No EDM Notch


0.200”X0.267” 0.220”X0.293” 0.220”X0.293” 0.300”X0.313”0.170”X0.227”0.150”X0.200”0.133”X0.173”0.100”X0.133”

Detection of 2nd Layer Notchesimpedance plane & ellipse fitting




0 5 1 0 1 5-2 0

0

2 0

4 0

6 0

8 0

f a s te n e rs

angl

e (d

egre

e) S N R a t C : 3 .5 3ED M : 0 .1 5 0 " ( W ) x 0 .2 0 0 " ( D )

R o w : T L

A -

0.10

0”(L

)X0.

133”

(D)

B -

0.13

3”(L

)X0.

173”

(D)

C -

0.15

0”(L

)X0.

200”

(D)

D -

0.17

0”(L

)X0.

227”

(D)

E -0

.200

”(L)

X0.2

67”(

D)

F -0

.220

”(L)

X0.2

93”(

D)

G -

0.25

0”(L

)X0.

313”

(D)

H -

0.30

0”(L

)X0.

313”

(D)

AB

–N

o ED

M N

otch

D0

–N

o ED

M N

otch

EF –

No

EDM

Not

ch

FG –

No

EDM

Not

ch

GH

–N

o ED

M N

otch

BC

–N

o ED

M N

otch

CD

–N

o ED

M N

otch

H0

–N

o ED

M N

otch

* These are post-processing data. Real-time processing is also working, but more verification is needed.





Study # 3

2nd Layer Crack near 2nd Layer Edge

Ferromagnetic Fastener




Example 3Detection 2nd layer vertically aligned cracks

with steel fasteners & crack close to 2nd layer edge

Challenges:1.Fastener head shortens drive - pickup coil

distance – possible less penetration

2.No-crack fastener signal may vary from one fastener to other due to ferromagnetism.

3.2nd layer crack is on the same layer and same side with 2nd layer edge and crack signal is much smaller than edge effect.





with steel fasteners, crack close to 2nd layer edgeCase 1 – 0.190” thick 1st layer

E EF FG GH H0F G H



A AB BC CD D0B C D

Upper Row



0.75

0”




Signals from Fastener HolesMajor difference between no-crack and crack signals is on the top half of

the impedance planes – when pickup coil passing the 2nd layer edge

No EDM Notch




Fastener

Pickup coil

Drive coil

2nd layer edge

2nd layer notch

Limited Area, A, where Crack Signal Can Be Observed

A




Apply Ellipse Fitting only on Area A of the impedance planes



Yushi Sun, Weiqiang Wan, Changhong Sun, Xinle Yang, Haiou Zhu & Tianhe Ouyang, IMTT 32C

D no ED

M

AB

no EDM

BC

no EDM

D0 no ED

M

EF no EDM

FG no ED

M

H0 no ED

M

A 0.1”

EDM

B 0.13 ED

M

C 0.15”

EDM

D 0.17”

EDM

E 0.2”ED

M

F 0.23”ED

M

G 0.25”

EDM

H 0.3”ED

M

GH

no EDM






with steel fasteners, crack close to 2nd layer edgeCase 2 – 0.250” thick 1st layer

E EF FG GH H0F G H



A AB BC CD D0B C D

Upper Row



0.75

0”




Signals from Fastener Holes without Notch

No EDM Notch

Signal varies in its size and shape




Signals from Fastener Holes with Notch

0.100” notch 0.130” notch 0.150” notch 0.170” notch

0.200” notch 0.230” notch 0.250” notch 0.300” notch

Increase of EDM Notch SizeSignal changes in irregular way




Fastener

Pickup coil

Drive coil

2nd layer edge

2nd layer notch

Limited Area where Crack Signal Can Be Observed

A




Signals from Fastener Holes without Notch

No EDM Notch



Yushi Sun, Weiqiang Wan, Changhong Sun, Xinle Yang, Haiou Zhu & Tianhe Ouyang, IMTT 38Information

Signals from Fastener Holes with NotchIncrease of EDM Notch Size




-100

-80

-60

-40

-20

0

20

40

60

80

100

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

CD

no EDM

AB

no EDM

BC

no EDM

D0 no ED

M

EF no EDM

FG no ED

M

H0 no ED

M

A 0.1”

EDM

B 0.13 ED

M

C 0.15”

EDM

D 0.17”

EDM

E 0.2”ED

M

F 0.23”ED

M

G 0.25”

EDM

H 0.3”ED

MSignal Rotation Angle

Detect-ability: Notch ≥ 0.170”




Pattern Recognition – An Alternative Approach

A couple of pattern recognition algorithms have been tried to

identify cracks located at different layers and locations in

two three-layer 0.690” thick aluminum samples. The samples

are provided by Sandia Labs and Cessna Aircraft. The first

test results are encouraging. However, a large number of

samples and a huge amount of work are needed for the

training the scheme. The project is now pending.




Conclusions1. Motorization of rotational RFEC probe provides

highly repeatable signal that enables the use of online signal processing and pattern recognition.

2. As a consequence motorized rotational probe is capable of detecting some cracks that are undetectable using regular manual probes.

3. Selection of appropriate signal processing algorithm improves overall sensitivity and reliability of RFEC rotational probes in cracks detection in thick multilayer structure with complex geometry.

Documents

Applications of Motorized Rotational RFEC Probes … IMTT...34th Annual Review of Progress in QNDE Golden Colorado, July 23 – July 27, 2007 Innovative Materials Testing Technologies,