Methodology for Assessing the Reliability of...

Methodology for Assessing theReliability of Nondestructive Inspections on

Wind Turbine Blades

Dennis RoachSandia National Labs

Infrastructure Assurance & NDI DepartmentFAA Airworthiness Assurance Center

Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company,for the United States Department of Energy’s National Nuclear Security Administration

under contract DE-AC04-94AL85000.

FAA Airworthiness Assurance Centerdproach@sandia.gov

Aircraft and Wind Energy Common “Bond” -Use of Composites and Health Monitoring

Composite Structures on Boeing 787 Aircraft

A380 Pressure Bulkhead

Common Goals: Improve Flaw Detection Performance inDetection Performance in Composite Structure in Order to Prevent Failures, Extend Blade Lifetime, Implement Condition-Based MaintenanceBased Maintenance

Composite Flaw Scenarios- Ply waviness - Lightning Strike- Ply waviness - Lightning Strike- Erosion - Impact Damage- Disbonds, Delamination - Voids

Composite Skin Disbondedf H b

Voids

Foreign Object DamageGround Handling Damage

from Honeycomb

Damage

Lightning Damage

Focused Inspections with Hand-Held Devices

WindshieldWindshield

3 ply lap joint3 ply lap joint

Wide Area and C-Scan Inspection Methods

Shearography

UltraImage Scanner

SAM System

PE Phased Array UT

Thermography

MAUS

UT Wheel Array

MAUSSystem

Shearography

Ultrasonic Wheel Array

Shearography (LTI) Image

Thermography (TWI) Image

SAM Image

MAUS Imageg

B737 Composite Horizontal Stabilizer

Objective: Assess rapid (wide area) NDI methods that could have applications to B787 production and/or in-service inspection.

Ultrasonic Scan

Thermography Scan

Determining Performance ofComposite Inspections

Composite Flaw Detection Experiment

Participation from over 25 airlines and maintenance depotsmaintenance depots

Industry-wide performance curves generated to quantify:• how well current inspection• how well current inspection

techniques are able to reliably find flaws in composite honeycomb structure

Experiment to Assess Flaw

Detection

• the degree of improvements possible through integrating more advanced NDI techniques and procedures.

Detection Performance• Statistically relevant and realistic

flaw profiles• Blind application of techniques

to study hits, misses, false calls, and flaw sizing

Experiment DesignSpecimen Types - modeled after range of construction found on

i ft S lid l i t b (12 20 24 32 li ) C t d daircraft; Solid laminate carbon (12, 20, 24, 32 plies); Contoured and tapered surfaces, Substructures – (stringers, ribs, spars); Bonded & sealed joints

Fl T t ti ti ll l t fl di t ib ti ith i iFlaw Types - statistically relevant flaw distribution with sizes ranging from 0.2 in.2 to 3 in.21) interply delaminations; disbonds2) substructure damage3) ki t tiff di b d

Low Energy Impact

00

45+ 45

3) skin-to-stiffener disbonds4) impact damage

0

0

0

-45

-45

0

9090

+ 45

Pyramid Pattern Matrix Crack from impact

P id P tt M t i C k f i t

Application of NDI - blind tests implemented in aircraft maintenance depots; guideline procedures provided; use of ref standards to set up equipmentof ref. standards to set up equipment

Expected Results - evaluate performance attributes1) accuracy & sensitivity (hits, misses, false calls, sizing)2) versatility, portability, complexity, inspection time (human factors)2) versatility, portability, complexity, inspection time (human factors)3) produce guideline documents to improve inspections4) introduce advanced NDI where warranted

Thick Laminate With Complex Taper

RINGER 2.00" X 2.00" X .125" THK

20 PLIES

12 PLIES

.50" STEPPER PLY Type I Specimen

30 PLIES

20 PLIES

.25" STEP PER PLY

4.75"

7.25"

12 PLIES3.50"

4.75"

14.00".25" STEP PER PLY

38.00"8.25"

.50" STEPPER PLY

19 00"

8.00"

2.00"

1.00"

7.25"19.007.25" P-1

E-1

I-1

W-2

BB-3

HH-3

X-2

T-2

PP-2

TT-2

F-2

M-2

R-2B-3

L-2

O 3

FF-3

W-2

Y-2

BB 2

U-3

Z-3

GG-2

L-1

LL-3

19.00"

O-3 BB-2L 1

38.00"

Thick Laminate With Complex Taper - Fabrication

Flaw templates - ensure proper location of flaws

Flawsinserted

Specimen Set - Flaw Detection in Solid Laminate Composites

Thickness Range:12 – 64 plies

Specimen Set - Flaw Detection in Solid Laminate Composites

Thickness Range:12 – 64 plies

Simple Tapers

Complex tapers

Substructure FlawsSubstructure Flaws

Curved Surfaces

Array of flaw types

NDI Ref. Stds.

Experiment Implementation at Airlines, 3rd Party Maintenance and Adv. NDI Organizations

Application of Advanced NDI Methods

Laser Ultrasonics

Phased Array UTPhased Array UT

Probability of Detection - Inspection Performance Over a Range of Test Specimen TypesOver a Range of Test Specimen Types

Flaws in Composite Honeycomb Construction

1

3 Ply Fiberglass

1

3 Ply Fiberglass 3 Ply Carbon

1

3 Ply Fiberglass 3 Ply Carbon 6 Ply Fiberglass

1

3 Ply Fiberglass 3 Ply Carbon 6 Ply Fiberglass 6 Ply Carbon

1

3 Ply Fiberglass 3 Ply Carbon 6 Ply Fiberglass 6 Ply Carbon 9 Ply Fiberglass

1

3 Ply Fiberglass 3 Ply Carbon 6 Ply Fiberglass 6 Ply Carbon 9 Ply Fiberglass 9 Ply Carbon

0 6

0.7

0.8

0.9

tect

ion

0 6

0.7

0.8

0.9

tect

ion

0 6

0.7

0.8

0.9

tect

ion

0 6

0.7

0.8

0.9

tect

ion

0 6

0.7

0.8

0.9

tect

ion

0 6

0.7

0.8

0.9

tect

ion

0.3

0.4

0.5

0.6

roba

bilit

y of

Det

0.3

0.4

0.5

0.6

roba

bilit

y of

Det

0.3

0.4

0.5

0.6

roba

bilit

y of

Det

0.3

0.4

0.5

0.6

roba

bilit

y of

Det

0.3

0.4

0.5

0.6

roba

bilit

y of

Det

0.3

0.4

0.5

0.6

roba

bilit

y of

Det

0

0.1

0.2

0 0 5 1 1 5 2 2 5 3

Pr

0

0.1

0.2

0 0 5 1 1 5 2 2 5 3

Pr

0

0.1

0.2

0 0 5 1 1 5 2 2 5 3

Pr

0

0.1

0.2

0 0 5 1 1 5 2 2 5 3

Pr

0

0.1

0.2

0 0 5 1 1 5 2 2 5 3

Pr

0

0.1

0.2

0 0 5 1 1 5 2 2 5 3

Pr

0 0.5 1 1.5 2 2.5 3

Flaw Size (Dia. in Inches)

0 0.5 1 1.5 2 2.5 3

Flaw Size (Dia. in Inches)

0 0.5 1 1.5 2 2.5 3

Flaw Size (Dia. in Inches)

0 0.5 1 1.5 2 2.5 3

Flaw Size (Dia. in Inches)

0 0.5 1 1.5 2 2.5 3

Flaw Size (Dia. in Inches)

0 0.5 1 1.5 2 2.5 3

Flaw Size (Dia. in Inches)

Inspection Improvements Via Advanced NDI Techniques

Comparison of Advanced Inspection Techniques withBest Conventional NDI Result on 9 Ply Carbon

Thermography MAUS IV Shearography CATT S.A.M. Wichitech DTH

0 8

0.9

1False Calls 744000 4.4

0.6

0.7

0.8

Det

ectio

n

0.3

0.4

0.5

Prob

abili

ty o

f

0

0.1

0.2

00 0.5 1 1.5 2 2.5 3

Flaw Size (Dia. in Inches)

Variation in Performance of a Device O th S t f I t

PoD Curve Comparisons - LFBT on 6 Ply FiberglassPoD Curve Comparisons - LFBT on 6 Ply Fiberglass with Cumulative Average

Over the Set of Inspectors

PoD Curve Comparisons - MIA on 6 Ply CarbonPoD Curve Comparisons - MIA on 6 Ply CarbonPoD Curve Comparisons - LFBT on 6 Ply Fiberglass

0 7

0.8

0.9

1

PoD Curve Comparisons - LFBT on 6 Ply Fiberglass with Cumulative Average

0 7

0.8

0.9

1

PoD Curve Comparisons - MIA on 6 Ply Carbon

0 7

0.8

0.9

1

MIA 1

PoD Curve Comparisons - MIA on 6 Ply Carbon with Cumulative Average

0 7

0.8

0.9

1

MIA-1MIA-2

0.3

0.4

0.5

0.6

0.7

Prob

abili

ty o

f Det

ectio

n

LFBT-1LFBT-2LFBT-3LFBT-4LFBT-5LFBT-6LFBT 70.3

0.4

0.5

0.6

0.7

Prob

abili

ty o

f Det

ectio

n

LFBT-1

LFBT-2

LFBT-3

LFBT-4

LFBT-5

LFBT-6

LFBT-7 0.3

0.4

0.5

0.6

0.7

Prob

abili

ty o

f Det

ectio

n MIA-1MIA-2MIA-3MIA-4MIA-5MIA-6MIA-7MIA-8MIA 90.3

0.4

0.5

0.6

0.7

Prob

abili

ty o

f Det

ectio

n MIA-3MIA-4MIA-5MIA-6MIA-7MIA-8MIA-9MIA-10

0

0.1

0.2

0 3

0 0.5 1 1.5 2 2.5 3

Flaw Size (Dia. in Inches)

LFBT-7LFBT-8LFBT-9LFBT-10LFBT-11

0

0.1

0.2

0.3

0 0.5 1 1.5 2 2.5 3

Flaw Size (Dia. in Inches)

LFBT-8

LFBT-9

LFBT-10

LFBT-11

Cum. Ave.0

0.1

0.2

0 0.5 1 1.5 2 2.5 3

Flaw Size (Dia. in Inches)

MIA-9MIA-10

0

0.1

0.2

0 0.5 1 1.5 2 2.5 3

Flaw Size (Dia. in Inches)

Cum. Ave.

( )( ) ( )( )

Large Variation in Device Performance Small Variation in Device Performance

Conclusions from CompositeNDI Performance Studies

90% POD is not achieved for 1” dia Flaws

• How are we doing? – Flaw Detection with Conventional NDI

90% POD is not achieved for 1 dia. FlawsPOD can improve as inspection time per area increases (up to a limit)Human factors issues (time, attention to detail, proper deployment)Coverage of large areas was randomg gComplex equipment produced large data spreadTop performing NDI methods were identified (reliability, repeatability, ease of use)

I t i fl d t ti d f 66% t 72%• How can advanced NDI help? – Flaw Detection with More Sophisticated NDI

Improvement in flaw detection ranged from 66% to 72%Automated deployment & data presentation/analysis reduces many human factors concerns (100% coverage; flaw recognition on images)Allow for more rapid inspectionsAllow for more rapid inspections

Math, glorious Math!Math, Glorious Math or…..How do we calculate Damage Tolerance ??How do we calculate Damage Tolerance ??

• Fatigue guy: Fantastic! I’ve calculated K to within 0.5%!

• Stress guy: Great! I’ve calculated stress to within 10%!

Loads guy: WOOHOO! I think I got the sign right!• Loads guy: WOOHOO! I think I got the sign right!

• NDI guy: You want me to find a crack how small??

You as the Inspector –pPerformance Assessment

You will be given three seconds to study a piece of artwork. Then one box will highlight.

You must decide if the box has changed color.

Did the selected square change color?

NOOCHANGE

Did the selected square change color?

YESSCHANGE

Did the selected square change color?

NOOCHANGE

Did the selected square change color?

YESSCHANGE

How did you do?How did you do?

• Did you try to game the system?– FACT: inspectors do better where they expect to find flaws

• Did you have the same level of concentration from the beginning to the end?

– There is significant research on how fatigue and difficulty of tasks affect performance.

Methodology for Assessing the Reliability ofMethodology for Assessing the Reliability ofNondestructive Inspections on Wind Turbine Blades

Dennis RoachSandia National Laboratories

FAA Airworthiness Assurance Center

As aircraft operators respond to calls for the ensured airworthiness of global airline fleets, the implementation of advanced airworthiness assurance technology is of growing importance. The development and application of new Nondestructive Inspection (NDI) techniques, the use of advanced materials, and general improvements in maintenance and repair practices need to keep pace with the growing understanding of aircraft structural aging phenomena. In 1991 the Federal Aviation Administration (FAA) established the Airworthiness Assurance Center (AANC) t S di N ti l L b t i Th AANC k ith i ft f t ilit d t d i li(AANC) at Sandia National Laboratories. The AANC works with aircraft manufacturers, military depots, and airlines to foster new technologies associated with nondestructive inspection (NDI), structural mechanics, repair methods, flight controls, fire safety, and crashworthiness. This presentation describes the structured methodology that the AANC has developed to quantify the performance of NDI techniques. This methodology may be useful to the wind energy industry since many of the inspection challenges, requirements and materials are similar to those encountered in the aviation industry. The aircraft industry continues to increase its use of composite materials,

t t th i th f i i l t t l l t Th t d t l d hi h t th tmost noteworthy in the arena of principle structural elements. The extreme damage tolerance and high strength-to-weight ratio of composites have motivated designers to expand the role of fiberglass and carbon graphite in aircraft structures. This has placed greater emphasis on the development of improved NDI methods that are more reliable and sensitive than conventional NDI. The AANC has been pursuing this goal via a host of studies on inspection of composite structures. Tap testing, which uses a human-detected change in acoustic response to locate flaws, and more sophisticated nondestructive inspection methods such as ultrasonics or thermography, have been applied to

i i b f li ti t d t t id di b d d d l i ti i dh i l b d d itan increasing number of applications to detect voids, disbonds, and delaminations in adhesively bonded composite aircraft parts. Low frequency bond testing and mechanical impedance analysis tests are often used to inspect thicker laminates. Probability of Detection experiments are underway to assess the performance of both conventional and advanced NDI techniques. Industry-wide performance curves are being produced to establish: 1) how well current inspection techniques are able to reliably find flaws in composite honeycomb structure, and 2) the degree of improvements possible through the integration of more advanced NDI techniques and procedures.