Encyclopedia of Structural Health Monitoring [engineering] - C. Boller, et al., (Wiley, 2009) WW.pdf - książki - Bio Med 10 - bibiariel

Chapter 1

Structural Health Monitoring — An

Introduction and Deﬁnitions

Christian Boller

Saarland University & Fraunhofer Institute for Non-Destructive Testing, Saarbrucken, Germany

(and formerly of The University of Shefﬁeld, Shefﬁeld, UK)

(although not expressed like this at the time) possibly

dates back much earlier than the 1980s when the

combination of words was created. Indeed, it may

date back to the origins of structural engineering.

Engineering structures are designed to be safe. The

difﬁculty one trading in this regard is the desire to

construct something for a speciﬁc purpose out of a

material of which one can never know enough in

terms of the material’s properties as well as the envi-

ronment the structure is going to operate in. We are

happy with the knowledge we can gain in this regard

and all we do not know needs to be covered by a

safety factor that we have to assume at best guess. The

less we know about the operational conditions of a

structure and the performance of materials and struc-

tures, the higher the safety factor will have to be. This

is the risk and dilemma structural engineering is in.

Engineering structures are designed to withstand

loads. These loads can be mechanical loads of a

static and/or dynamic nature. Loads can, however,

also be of an environmental nature such as temper-

ature, humidity or chemical, and again the structure

can be exposed to these loads in either a static or

even very short term and thus dynamic condition

such as a thermo shock. Knowledge of loads applied

to a structure has to come from experience. This

experience has been either gathered on similar struc-

tures in the past or from assumptions. The safest

1 Background

2 Loads Monitoring

3 Damage Monitoring

4 Sensor System Implementation

Strategies

5 SHM Potential Determination

6 Conclusions

Acknowledgments

References

Further Reading

BACKGROUND

Structural health monitoring (SHM) is a combination

of words that has emerged around the late 1980s.

It is very much associated with what we are accus-

tomed with in the classical medical and health sector,

combines it possibly with what we are accustomed to

in control (monitoring) and links it to what we are

truly considering here: engineering structures. SHM

Encyclopedia of Structural Health Monitoring . Edited by

Christian Boller, Fu-Kuo Chang and Yozo Fujino 2009

John Wiley & Sons, Ltd. ISBN: 978-0-470-05822-0.

Introduction

way to design a structure is to design it against an

ultimate design limit load, which is the maximum

load ever experienced with such a structure added by

a safety margin. Designing a structure against this

load, however, makes the structure heavy. Often, the

maximum load of a structure may just occur once in

the structure’s life, if ever at all. In that case one may

start to question the extreme safety built in, speciﬁ-

cally if the maximum load applied would not result

in any observable damage.

Loads applied to a structure are the reason for

structural deterioration and hence resulting damage.

This damage may be generated at a microscopic level

and may gradually progress until it becomes observ-

able and critical. Trading with this observability and

criticality is the art of damage tolerant design, which

has allowed structures to become lighter weight. The

way the damage accumulates is of a fairly random

nature, and scatter of a factor of 2 in operational life

is absolutely normal. This requires a careful means

and procedure of inspection at well-deﬁned intervals.

The booming development of sensing technology

in terms of sensors decreasing in size and cost, and

the combination with microprocessors with increasing

power and enhanced materials design and manu-

facturing in terms of functional materials or even

electronic textiles have opened avenues in merging

structural design and maintenance with those advan-

ced sensing, signal processing, and materials manu-

facturing technologies. Taking advantage of this

lateral integration is what SHM is about. The central

set of questions in this ﬁeld is therefore

Without compromising safety, could we make our

structures

•

combining the sensors through

– advanced microelectronics and possibly

– wireless technology with

–

advanced microprocessors and

–

advanced signal processing?

If one would try to give an answer to all these,

the answer could somehow result in SHM, and a

deﬁnition for SHM could possibly be the following.

SHM is the integration of sensing and possibly also

actuation devices to allow the loading and damaging

conditions of a structure to be recorded, analyzed,

localized, and predicted in a way that nondestruc-

tive testing (NDT ) becomes an integral part of the

structure and a material.

As a consequence, SHM requires to look at loads

as well as damage monitoring with respect to their

sensing and assessment algorithms and needs to get

those merged in a holistic process such that the

health of a structure can be accompanied during the

complete life cycle process of the structure consid-

ered. How this has emerged and could be further

achieved is described as the holistic process in the

following articles with further details to be found

throughout the wide range of articles being provided

throughout this encyclopedia.

LOADS MONITORING

Design of engineering structures is at least based

on static strength. This is the way structures have

been designed since the past mainly by applying a

test load that exceeds the maximum operational load

by a safety factor to be deﬁned or by designing

the structure against the expected maximum load

times the safety factor, or possibly both. The problem

of fatigue in structures became apparent with the

upcoming railway industry in the second half of the

nineteenth century. August Wohler [1] was possibly

the ﬁrst to determine that components—and in his

case railway axles—would fracture at loads much

lower than the ultimate tensile load if the components

were exposed to a repetitive loading. Wohler further-

more determined that the number of cycles to fracture

is related to the level of the repetitive load being

applied. This resulted in the effect of materials fatigue

to be established and has been mainly described in

the form of a fatigue–life curve also often denom-

inated as S–N or Wohler curve. S–N curves were

better available

•

lighter weight

•

more cost efﬁcient and

•

more reliable

by making sensors (and possibly also actuators) to

become an integral part of the structure?

What about

•

looking at advanced cheap sensors, which are

continuously becoming

–

smaller

–

lighter and

–

cheaper?

•

making the sensors an integral part of the struc-

tural component?

An Introduction and Deﬁnitions

and are therefore used to determine the allowable

loads of railway axles compared to static loads, which

were now named fatigue loads. This principle is well

applicable because fatigue loads on railway axles are

fairly constant over the life cycle and hence can

be considered constant amplitude. Wohler’s principle

was shortly expanded to other railway applications

such as frames or boogies of railway engines and

carriages as well as even railway steel bridges.

Constant amplitude loading is not the way struc-

tures are conventionally loaded while operating in

service. Most of them are more loaded in accor-

dance to a randomized nature where small loads

follow a high load or vice versa such as the different

time domain signals shown in Figure 1. Ernst Gaßner

[2] was possibly the ﬁrst to recognize in the 1930s

that service loading of structures had to be treated

different from constant amplitude loading or in other

words constant amplitude loading to be a speciﬁc

condition within the frame of in-service loading.

Fatigue–life curves for in-service loading were there-

fore deﬁned in terms of the maximum load applied

in the in-service spectrum versus the number of

loading cycles sustained until fracture. Palmgren [3]

from Sweden published his damage accumulation rule

(published later in English by Miner [4]) seeming

initially trivial for constant amplitude loading, which

however became not only more attractive to be

applied for randomized in-service loading but also an

issue for wide scientiﬁc discussion until even today.

Parallel to fatigue analysis, the other wide area

of fracture mechanics was initiated during the early

twentieth century, mainly driven by people such as

Grifﬁth [5] initially. Fracture mechanics was further

developed since that time and speciﬁcally during

WW2, which led to an enhanced understanding of

the fatigue process. Furthermore, the combination of

fatigue loading and fracture mechanics allowed the

new principle of damage tolerant design to be estab-

lished. Damage tolerant design as opposed to safe

life design, which is shown in Figure 2, is a principle

that allows damages such as cracks to be present in a

structure as long as the overall integrity of the struc-

ture is not compromised. This is achieved by either

monitoring slow crack growth through well-deﬁned

inspection intervals or building in load redundancy

in a way such that when a component is due to

fracture another component is still able to take over

the load without compromising the overall struc-

ture itself. Considering damage tolerance in struc-

tures has allowed those structures to become much

more lighter weight when compared to the traditional

safe life design. Damage tolerant design has there-

fore become mainly the standard for designing civil

aviation metallic structures nowadays, with military

and even jet engine structures to gradually move

towards those design principles as well. However,

there are also other areas that have taken advantage

of the damage tolerance principle such as marine and

offshore structures.

Rear axle of a car

Pressure in a condensation chamber of a reactor

Stresses on a car wheel

Momentum generated by the engine of a milling

machine

M D

Bending moment at the stub axle of a car

M B

Acceleration in the center of gravity of a fighter

airplane

Pressure in a pipeline

Acceleration in the center of gravity of a

transport aircraft

n z

Time

Figure 1. Time domain signals for different in-service random loads.

Introduction

Fatigue design

How much

potential left?

Safe life design

Damage tolerant design

Slow crack

propagation

Fail safe

Multiple

load path

Crack stopper

Figure 2. Principles in engineering structural design.

Aircraft structures are hence designed to be either

safe life or damage tolerant for a service loading

spectrum to be deﬁned during the design of the

aircraft. Safe life means that the structure is not

allowed to fracture for the loading spectrum deﬁned

and up to a given service life. Irrespective of its

condition, the component will have to be replaced by

a new component after its service life achievement.

Damage tolerance is however the alternative to safe

life design where a crack is allowed to grow in the

structure over a signiﬁcantly long period of time such

that it can be detected during that period safely and

does not lead to a critical hazard.

The ﬁrst aircraft to possibly ever seriously consider

damage tolerant design was the De Havilland Comet

designed and built in the early 1950s, which was

ﬁrst ﬂown in 1952. This aircraft was possibly the

ever most challenging aircraft built in civil aviation

history. The design did not just include damage

tolerance but also a pressurized fuselage as well as

jet engines integrated into the wing. This extreme

innovation push however led to some serious crashes

in 1953/1954, which resulted from a crack having

generated from one of the corner of the windows

where the crack had initiated due to underestimation

of the window corners’ notch combined with the

continuous pressurization cycles during each ﬂight.

The consequences of this accident were immediate

and led to

•

the major airframe fatigue test (MAFT) to be

introduced where a full-scale aircraft structure is

tested under in-service loading conditions on the

ground and this being ahead of the so-called ﬂeet

leader (the aircraft in a ﬂeet having the most ﬂight

cycles accumulated) in terms of ﬂight cycles; and

•

operational loads monitoring devices to be consid-

ered for aircraft in general.

While the former two consequences can be attribu-

ted to structural design, the latter can be clearly

attributed to SHM. Although nobody was using the

expression of SHM at the time, it was clearly shown

that a more precise knowledge of operational loads

would be highly essential. This led to the UK Royal

Air Force to design a mechanical device that was inte-

grated around the center of gravity of ﬁghter airplanes

in the late 1950s and was based on a set of accelerom-

eters being set at different acceleration thresholds

(Figure 3). Whenever the aircraft would exceed one

of the thresholds, the counter would count those as

a measure of exceedances. Hence the numbers being

recorded by the different accelerometers would repre-

sent the actual load spectrum of the aircraft at the time

of recording. The following formula for calculating a

ﬂight (or better fatigue) index was introduced

Flight index (F I )

K 2 ×

S 1 ×

( 2 . 31 A

0 . 03 B

0 . 001 C

0 . 001 D

0 . 28 E

•

the shape of the windows being near to rectan-

gular at the time to be changed to the more ovoid

shape we have nowadays;

3 . 43 F

10 . 36 G

18 . 63 H

1 . 16 WL)

( 1 )

An Introduction and Deﬁnitions

where K 2 ,S 1 ,and WL denote the mission coef-

ﬁcient, stores conﬁguration coefﬁcient and landing

coefﬁcient, respectively, while A to H represents the

readings from the different accelerometers.

Gaßner, who had introduced the eight-step block

program sequence as a ﬁrst step in in-service load

sequence analysis [6], continued to determine in-

service load sequences for automobiles where a load

cycle counting system was developed as shown in

Figure 4. Tests performed with this system on various

types of roads resulted in a variety of load spectra as

shown in Figure 5. This further led to the develop-

ment of a variety of standard load sequences such as

the Gauß and linear distribution spectrum for random

loading [7].

Further development was ongoing into a variety of

random load sequences such as for transport aircraft

(Transport WIng STandard (TWIST)) [9], ﬁghter

aircraft (FALSTAFF) [10], helicopters (HELIX

FELIX) [11], wind energy structures (WISPER) [12],

and possibly much more. Figure 6 shows the TWIST

random load sequence for the wing attachment of

a transport aircraft as an example. It can be seen

that this load sequence is not just composed of a

randomization of loads but also of a randomization

of different ﬂight types.

Figure 3. Acceleration exceedance monitoring system

introduced in the Royal Air Force in the 1950s.

3 .

. 10

Figure 4. Gaßner’s loads monitoring system used in the late 1950s and early 1960s for monitoring road transport load

sequences [8].

Encyclopedia of Structural Health Monitoring [engineering] - C. Boller, et al., (Wiley, 2009) WW.pdf

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: