health monitoring for structural “hot spots”: a ... · pdf fileboeing is a...

12/6/2007 | 1BOEING is a trademark of Boeing Management Company.Copyright © 2006 Boeing. All rights reserved.

Use or disclosure of data contained on this page is limited to the restrictions on the title page of this document

Health Monitoring for Structural “Hot Spots”: A Systems Approach

Mark DerrisoAdvanced Structures Branch

Air Vehicles Directorate Air Force Research Laboratory

This document does not contain technical data within the definition contained in the International Traffic in Arms Regulations (ITAR) and the Export Administration Regulations (EAR), as such is releasable by any

means to any person whether in the U.S. or abroad. The Export Compliance log number for this document is Export Approval # BOE111207-306 (assigned IAW PRO-4527, PRO 3439)

Eric HaugseMaterials and Structures Technology

Phantom WorksBoeing

12/6/2007 | 2

Engineering, Operations & Technology | Phantom Works

Copyright © 2006 Boeing. All rights reserved.Use or disclosure of data contained on this page is limited to the restrictions on the title page of this document

Presentation Outline

SHM Level Setting

Design Space

Example Applications

Design Framework

Closing Remarks

12/6/2007 | 3



Structural Health Management (SHM) Evolution

“Structural safety is an evolutionary accomplishment, and attention to design features is key to its achievement. Acquisition and review of service data and other firsthand information from customer airlines is necessary to promote safe and economic operation of the worldwide Boeing fleet…………This paper describes these structural health monitoring approaches”.

U. Goranson, Boeing Commercial Airplane Group, Key Note Speech, Stanford Workshop on Structural Health Monitoring, 1997.

12/6/2007 | 4



Current State: Schedule-Based Maintenance

Structural Design

Full- Scale Fatigue Testing

Problem Areas Identified

Front SparMain Spar

Rear SparClosure Spar

F-15 Wing

Conduit Hole(hot spot)

Inspection Schedule Constructed

12/6/2007 | 5



Schedule vs Condition

Schedule-based maintenanceInitially works wellHowever, over time req’ts change

Use vehicle systems longer than plannedUse for different missions than designedNew problem areas identified

Results in decreased availability, increased inspection times, and increased O&S costs

Condition-based maintenanceIncreases availability, increasesreliability, and decreases O&S costswhile maintaining vehicle safety

B-52

KC-135

12/6/2007 | 6



Health Monitoring for Structural “Hot Spots”: Design Space

Integrated Design Implementation• Reliability Based Design Criteria• Damage Prognostics• Unsupervised Learning• System Integration

Key technology paradigm shift is from scheduled, hands on inspections to in-situ monitoring, condition based maintenance, and design integration.

Mon

itorin

g C

over

age/

Mon

itorin

g D

esig

n C

ost

Maintenance Effort & Cost Based on Condition / Maintenance Effort & Cost Based on Schedule

Monitoring Development• Advanced NDI/ In-situ SHM• Improved POD/ Diagnostics• Virtual (model based) Sensing• Optimized Weight, Volume

Autonomous Local

Monitoring

Inspection Enhancement

Current StateSchedule Based

Maintenance

Broad Area Condition

BasedMaintenance/

Design Feedback•High Sortie Rate•Gate Availability

•Fleet Management•Maintenance Management

12/6/2007 | 7



2002 2003 2004 2005 2006 2007

Flight demonstration sensor installation

In-service data collectionCoupon and component

testing

Requirements development

with customer

SHM for Bonded Repairs (Mark Derriso, AFRL)

12/6/2007 | 8



Understanding Requirements from a Systems Implementation Standpoint: Damage Repair Decisions

Allowable Damage Range

DAMAGESIZE

STRUCTURALSTRENGTH

Strength with BVID

BVI

D

Allo

wab

le

dam

age

limit

Ultimate

Tem

p R

epai

r Li

mit

Example failure strain (strength) vs. damage size curve. This will change for location, layup, distance from design details, etc.

STRUCTURALSTRENGTH

TEMP R

EPAIR

Limit(fail safe)

SERVICE TIME

Damageoccurs

Act

ual

dam

age

size

Damage size prediction with error estimate

Temporary repair patches return structure to ultimate load carrying capability

Pristine

Temporary repair limit

Tem

pora

ry

repa

irsPe

rman

ent

repa

irs TEMP REPAIR RANGE

SHM system will provide geometry information

about the damage (e.g., location, area, depth, etc.)

SHM system will provide geometry information

about the damage (e.g., location, area, depth, etc.)

12/6/2007 | 9



Composite Impact Damage Estimation Where Damage Sizing is a Requirement

Impact

IML

Scattered Image

Volume

A-scan damage outline

Impact

IML

OMLdepth

diameterGoals: Obtain the best prediction of damage delamination area and maximum delamination depth

12/6/2007 | 10



5’x7’ Panel

Impact Damage Sizing, Comparison to NDI

10 20 30 40 50 60 7015

20

25

30

35

40

45

12 3 4 5

6

78

10

11 1213 14

15

1617

ALL DIMENSIONS IN INCHES

Green: C-scan, Blue: A-Scan, Red: SHM

12/6/2007 | 11



Lap Joint/ Repair Monitoring

Upper skin

Lower skinTear strap

Sealant

Rivets

StringerStringer clip

Upper skin

Lower skinTear strap

Sealant

Rivets

StringerStringer clip

12/6/2007 | 12



Phased Array approach helps meet requirements and cost targets for a linear application like a lap splice

Phased Array Design to Meet Application Requirements

12/6/2007 | 13



Prototype Flight Installation for Proof of Concept

Phased Array Design Implementation

12/6/2007 | 14



Hard Landing Assessment Example

GOALS - Reduce aircraft schedule interrupts by:1. Reducing number of falsely reported hard landings2. Aiding the maintenance process

PHASE 2 APPROACH• Pilot “by feel” initiated inspection• Flight parameters and sensor information

used to predict detailed load information in critical structural areas

• Recommend maintenance action• Aid maintenance procedure

AUTOMATEDLOAD PREDICTION

PHASE 1 APPROACH• Pilot “by feel” initiated inspection• Maintenance page provides flight

parameters at touchdown• Lookup table used to determine

maintenance action

MAINTENANCE PAGE

787 HEAVY/HARD LANDING MAINTENANCE PAGE

DATE XX XXX XXGROSS WEIGHT XXX.XCG XX.X

MAIN GEAR:UTC XX:XX:XX.XPRIOR TO TOUCH DOWN:

CG SINK RATE XXX.XPITCH ANGLE XXX.XCRAB ANGLE XXX.XROLL ATTITUDE XXX.XBODY ROLL RATE XXX.XCG NORM ACCEL XXX.X

TOUCHDOWN:PEAK CG NORM ACCEL XXX.X

787 Maintenance ManualGW < MLW+10K

Sink < 5 Sink <8Pitch Angle <10 <10Crab Angle <5 <3Roll Angle <4 <2Roll Rate <8 <4Initial CG NZ >.9 >.95Peak CG NZ <1.5 <2.0

CURRENT APPROACH• Pilot “by feel” initiated inspection• Limited flight data usage with AMM• Large number of false positives

787 Maintenance ManualGW < MLW

Roll Angle <2Peak CG NZ <1.9

MAINTENANCE MANUAL

INSPECTION

LOADSLOAD

PREDICTIONMODEL

FLIGHTPARAMETERS

SENSORDATA

LOADSLOAD

PREDICTIONMODEL

FLIGHTPARAMETERS

SENSORDATA

LOADPREDICTION

MODEL

FLIGHTPARAMETERS

SENSORDATA

LANDING EVENT

12/6/2007 | 15



0 0.5 1 1.5 20

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

H.L. Sim: -BS307

N.N

. Pre

dict

ion:

-BS

307

damage indicator location

Sample Hard Landing Trade Results

• Composite plot of predicted versus simulated (i.e., truth) normalized damage indicators.• Adjust threshold to eliminate false negatives, but at a cost of increasing false positives.• Adding physical sensors increases performance to a point.

0 0.5 1 1.5 20

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

H.L. Sim: -BS307

N.N

. Pre

dict

ion:

-BS

307

falsepositive

falsenegative

normallanding

hardlanding

12/6/2007 | 16



Load Monitoring for Condition Based Maintenance

Critical Crack Length(Allowable Damage)

Detectable Crack Length

Damage ToleranceDurability

Flights, N

N = Inspection IntervalCra

ck L

engt

h, L

Crack Initiation Crack Propagation

N N N

Damage Detection

PeriodN N

Inspection program

Critical Crack Length(Allowable Damage)

Detectable Crack Length

Damage ToleranceDurability

Flights, N

N = Inspection IntervalCra

ck L

engt

h, L

Crack Initiation Crack Propagation

N N N

Damage Detection

PeriodN N

Inspection program

Operational load data can be used to determine inspection thresholds and intervals based on durability and damage tolerance methods.Approach must show significant maintenance cost improvement overscheduled inspection approach, while maintaining or improving reliability.This is a key component of an overall systems, condition based maintenance approach.

12/6/2007 | 17



Hot Spot Monitoring and Design Framework (AFRL)

Understand the structure

Develop system level SHM

requirements

Develop candidate SHM system designs

Compare each design to the requirements

Remove the design from

consideration

Structural deign

Develop potential benefits for

applying SHM

Understand available

technologies and methods

Understand the costs of the

SHM systems

Does the design meet the

requirements?

Implement the lowest cost SHM

system design

Yes No

12/6/2007 | 18



Closing Comments

Implementation requires clear understanding of benefits and requirements.

Understanding of impact to overall design and design criteria is critical to understanding the implementation time frame.

A design framework that allows SHM systems to be designed in the context of the overall system (structures, systems, support) is critical to implementation success.

12/6/2007 | 20



SHM End-to-End Flow: Damage Detection Example

SensorsDataAcquisitionHardware

A1

A2

An

VisualInspection

NDE/NDI

Other Data(reports of damage

events, fleet history …)

Data Fusion, Integration& Damage

Determination

StructuralTransformation

(to actual damageand failure mode

of Structure)

Structuralanalysis

InternalLoads

Margins

RepairRequirements

Data Collection & Processing

Sensors Information Processing toDefine Current Structural State

DecisionMaking &

Distribution

PrognosisLoad, environments

for future flightsDamage Progression

(if no repair)Continued Airworthiness

Assurance

A1, A2, An = Multiple algorithms for SHM information processing

health monitoring for structural “hot spots”: a ... · pdf fileboeing is a...

Documents