Czichos H., Saito T., Smith L.E. (Eds.) Handbook of Metrology and Testing

Подождите немного. Документ загружается.

908 Part D Materials Performance Testing

16.2.4 Digital Radiological Methods

Users of industrial radiology, who are interested to

move from classical ﬁlm radiography to digital indus-

trial radiology (DIR), should know which particular

ﬁeld they want to enter. Basically there exist unreg-

ulated and regulated areas (standards, guidelines and

others). Table 16.2 describes the four basic application

areas.

Before new methods and detectors will be de-

scribed, the advantages of traditional ﬁlm radiography

should be summarized. One of the major advantages

of x-ray ﬁlm radiography is its practicability. Important

x-ray ﬁlm properties are

•

they are dust- and waterproof packed;

•

they consist of ﬂexible and break-proof material;

•

the storage period after development is > 50 years

(expected shelf life > 500 years);

•

the readability is independent of technological de-

velopment (e.g. independent of data format).

Film Digitization

Film digitization systems can be classiﬁed by the sam-

pling technology (Table 16.3).

For example, digitization with a laser scanner pro-

ceeds as shown in Fig. 16.42. The ﬁlm passes a collec-

tion tube. A laser beam (wavelength about 680 nm, red)

with a ﬁxed diameter (e.g. 50 μm) scans the ﬁlm. The

diffuse transmitted light through the ﬁlm is integrated

by the collection tube and registered by a photo-

multiplier (PMT) on top of the collection tube (not

shown in Fig. 16.42). During the scan the folding mir-

ror deﬂects the laser beam and moves the spot along

a horizontal line on the ﬁlm. The ﬁlm is advanced with

a constant speed. The resulting voltage at the photo mul-

tiplier is proportional to the light intensity behind the

ﬁlm. After logarithmic ampliﬁcation a digitization with

Table 16.2 Application areas of digital industrial radiology (DIR)

Digital industrial radiology

Film replacement New industrial areas

Standards, Nonstandard Serial part Computed

regulations applications inspection tomography

Welding Wall thickness, Automated defect 3-D-casting

corrosion, erosion recognition (ADR) inspection

Casting Buildings, bridges Completeness test Ceramic composites, plastics

Electronics Plastics, composites Dimensional check Special applications

Food, tires, wood ...

Table 16.3 Classiﬁcation of digitization systems by tech-

nology

Principle Scanner type

Point-by-point digitization Laser scanner

Line-by-line digitization CCD line scanner

Array digitization CCD camera

12 bits yields grey values that are proportional to the

optical density of the ﬁlm.

The essential difference to other digitization principles

is the reversed optical alignment. The laser scanner il-

luminates with focused light and measures the diffuse

light intensity behind the ﬁlm. All other methods illu-

minate the whole area of the ﬁlm with diffuse light (the

ﬁlm is illuminated with a diffuser) and measure the light

intensity that passes the ﬁlm in one direction at each

spot (camera objective or human eye in classical ﬁlm

inspection).

Complementary metal--oxide--semiconductor(CMOS)

cameras, which generate a logarithmic output signal rel-

ative to the input light intensity, are also available. In

this case the digitized grey values will be proportional

to the ﬁlm density, and do not follow the exponential

characteristics as digitized with charge-coupled device

(CCD) chips. This is an advantage over CCD chips for

the digitization of radiographic ﬁlms, but the signal-to-

noise ratio for CMOS detectors is usually considerable

lower than for CCD chips.

The standard ISO/EN 14096-1/2 deﬁnes the qualiﬁ-

cation procedure and the minimum requirements for

ﬁlm digitizers in NDT. This is particularly impor-

tant for microradiography. NDT applications employ

x-ray energies of 50–12 000 keV. The standards require

a spatial pixel size of 15–250 μm depending on the en-

ergy. This corresponds to a required spatial resolution

of 16.7lp/mm (line pairs per millimeter) for energies

< 100 keV and e.g. 1 lp/mm for 1300 keV.

Part D 16.2

Performance Control 16.2 Industrial Radiology 909

On the basis of the image quality of ﬁlm radiogra-

phy the standard requires three quality classes: DA,DB

and DS. The user may select the testing class based on

the needs of the problem.

•

DS: the enhanced technique, which performs the

digitization with an insigniﬁcant reduction of

signal-to-noise-ratio and spatial resolution; applica-

tion ﬁeld: digital archiving of ﬁlms (digital storage);

•

DB: the enhanced technique, which permits some

reduction of image quality; application ﬁeld: digital

analysis of ﬁlms; ﬁlms have to be archived;

•

DA: the basic technique, which permits some re-

duction of image quality and further reduced spatial

resolution; application ﬁeld: digital analysis of

ﬁlms; ﬁlms have to be archived.

Image Processing

A digital image is physically not more than a data ﬁle

in a computer system linked to a program capable of

displaying its contents on a screen or printout dot by dot.

That means it consists of an array of individual image

points. These are called picture elements (pixel). The

pixels are in practice small rectangles (squares) showing

a certain color and brightness.

The image resolution depends on the number of pix-

els within a given area; the more the better (Fig. 16.43).

Radiography commonly deals only with grey-level im-

ages.

The numerical values of any pixels within an im-

age can be subjected to various kinds of calculations.

Each pixel is characterized by a grey (intensity) value,

which is proportional to its brightness. The most triv-

ial step is to adjust brightness and contrast. This allows

one to overcome the lower brightness of monitors com-

pared to commercial ﬁlm viewers. While the human eye

can only distinguish between some 60–120 grey values

(at a given adaptation, equivalent to 6–7 bits of infor-

mation) a digital image may contain up to 65 536 grey

levels (i. e. 16 bits). The contrast and brightness oper-

ations are essential to select the range of interest. The

gamma transformation alters the linearity of the trans-

fer from the original digital values listed in the image

ﬁle to its brightness displayed on the screen, taking into

account the brightness sensitive perception of human

eyes.

Multiple-point calculations take advantage from re-

lationships between adjacent pixels of the image (matrix

operations). They are based on a variety of so-called

ﬁlter algorithms that are supposed to extract the de-

sired features from the total image information. They

Folding mirror

Galvanometer

Laser

Collection

tube

Film transport

Fig. 16.42 Principle of a laser scanner (“LS 85”, Kodak, USA)

may eliminate unwanted large, patchy intensity ﬂuc-

tuations or noise masking other image details. Typical

ﬁlters are describes in Table 16.4. Processing of ra-

diographic images shows astonishing results that can

be found, together with ﬁlter recipes, in different text-

books [16.22,23].

Computed Radiography

and its Industrial Application

Phosphor imaging plates (IP) are image media for ﬁlm-

less radiography. The technique is also called computed

radiography (CR). IPs are routinely used in medicine

and biomedical autoradiography. Different systems are

Fig. 16.43a,b Radiographic image of a test weld BAM5,

(a) 1024× 512 pixels; (b) 64 × 32 pixels

Part D 16.2

910 Part D Materials Performance Testing

Table 16.4 Typical image-processing ﬁlters

Filter Advantage Disadvantage

Low-pass ﬁlter (smoothing) Increases the signal-to-noise ratio Reduces the spatial resolution

High-pass ﬁlter Increases the contrast of ﬁne details in Reduces the signal-to-noise ratio

relation to intensity changes in a wide range

Median ﬁlter Increases the signal-to-noise ratio, removes Reduces the spatial resolution

single peaks and outliers (single white or

black pixels such as salt and pepper distortions),

does not smooth edges

Edge ﬁlter Enhances edges or extracts edges to lines Increases the noise

available for NDT applications. Novel applications in

addition to ﬁlm radiography emerge taking advantage

of the higher sensitivity (shorter exposure time) and the

digital processing as well as the capability to analyze

digital radiographs with affordable computer systems.

A set of standards was published 2005 in Europe and

USA, which deﬁnes the classiﬁcation and practice of

CR systems [16.24].

IPs are handled nearly in the same way as radio-

graphic ﬁlms. After exposure, they are read by a laser

scanner producing a digital image instead of being de-

veloped like a ﬁlm (Fig. 16.44). Any remaining latent

image can be erased with a bright light source so that

the same IP can be recycled up to more than 1000 ×.

An IP consists of a ﬂexible polymer support which is

Expose

Reuse

High sensitivity

computed radiography

Phosphor

imaging plates

Read

High dynamics

Medium resolution

Reusable

Data output

Laser

PMT

Erase

Fig. 16.44 Principle of application of phosphor imaging plates (IP)

for computed radiography (CR). The plate is exposed in a lightproof

cassette. The scanner reads the IP with a red laser beam. All ex-

posed areas emit stimulated blue light. The photomultiplier collects

the blue light through a blue ﬁlter and converts it into an electri-

cal signal. The signal intensity values are converted to digital grey

values and stored in a digital image ﬁle. Remnants of the image

on the IPs are erased by intensive light. They can be used usually

100–1000 × if they are handled carefully

coated with the sensitive layer which is sealed with

a thin transparent protective layer. The sensitive layer

of the most common systems consists of a mixture of

BaFBr doped with Eu

. X-ray or gamma-ray quanta

result in an avalanche of charge carriers i. e. electrons

and holes in the crystal lattice [16.25].

These charge carriers may be trapped at impurity

sites i. e. electrons at a halogen vacancy (F-center) or

holes at an interstitial Br

molecule (H-center). Red

laser light (600–700 nm) excites electrons trapped in

aBr

−

vacancy (F

-center) to a higher state from which

they may tunnel and recombine with a nearby trapped

hole. Transfer of the recombination energy excites

a nearby located Eu

ion. Upon return to its ground

state this Eu

ion emits a blue photon (390 nm). This

process is described as photostimulated luminescence

(PSL).

The advantages of IP technology are

•

linearity with radiation dose

•

high dynamic range, up to 10

•

sensitivity higher than with ﬁlm

•

reusable for 1000 cycles

•

no chemical darkroom processing

•

capability for direct image processing.

The disadvantages are

•

limited spatial resolution, except in new high-

deﬁnition systems

•

high sensitivity in the low-energy range

•

sensitive to scattered radiation.

The available systems of phosphor imaging plates

and corresponding laser scanners cover radiation dose

differences of up to 10

. This feature reduces the

number of exposures for objects with a high wall-

thickness difference. It also compensates partly for

incorrectly calculated exposure conditions. The number

of so-called test exposures is reduced. The IP reader is

typically separated from the inspection site. CR is based

Part D 16.2

Performance Control 16.2 Industrial Radiology 911

Fig. 16.45 Computer-based inspection of an insulated pipe

by projection radiography. The wall thickness is measured

by image processing in the marked areas

on ﬂexible IPs [16.26]. High-deﬁnition CR systems can

provide a spatial resolution of better than 25 μm. This

is sufﬁcient for weld and casting inspection of small

components at lower x-ray voltages as well as large

components.

Measurement of Pipe Wall Thickness for the

Evaluation of Corrosion, Erosion and Deposit

A typical application of the CR technology is radio-

graphic corrosion inspection in the chemical industry.

Figure 16.45 presents a typical example with a ther-

mally insulated pipeline. The insulation is covered

with an aluminum envelop. The radiographic inspec-

tion can be performed without removing the insulation.

This is a considerable advantage relative to the other

known methods. Radiographic pipe inspection for cor-

rosion and wall thickness measurement is a major

NDT technique for predictive maintenance. CR is

also more and more applied for inspection of valves

and armatures for functionality check and deposit

search.

X-rays

Line driver

Contacts

Switch

Scintillator

Amplifier,

multiplexer,

ADC

photodiode

Pixel matrix

Fig. 16.46 Scheme of a ﬂat-panel detector: The scintillator converts

x- or γ -rays into light, which is detected by the photodiodes. They

are read out by thin-ﬁlm transistors (TFT) on the basis of amorphous

silicon, which is resistant to radiation

Radiography with Digital Detector Arrays (DDA)

Flat-Panel Detectors. Two types of DDAs, also called

ﬂat-panel detectors, are available on the market. The

ﬁrst design (Fig. 16.46) is based on a photodiode ma-

trix connected to thin-ﬁlm transistors (TFT) [16.27].

These components are manufactured of amorphous sil-

icon and they are resistant against ionizing radiation.

Alternatively to the amorphous silicon panels CMOS

arrays can be applied. The photodiodes are charged by

light which is generated by a scintillator converting the

incoming x- or gamma rays. This scintillator can be

a polycrystalline layer that causes some additional un-

sharpness by light scattering or a directed crystalline

system which acts like a face plate with lower un-

sharpness due to inhibited light scattering (Fig. 16.47).

The next generation (second type of DDAs) of ﬂat

panels is based on a photoconductors like amorphous

selenium [8] or CdTe on a multi-microelectrode plate,

which is also read out by TFTs(Fig.16.48b). Di-

rect converting photodiode systems (Fig. 16.48a) are

not used in NDT yet, due to their low quantum efﬁ-

ciency.

DDAs are suitable for in-house and in-ﬁeld appli-

cations. In-ﬁeld applications are characterized by harsh

environmental conditions in some areas, which implies

the risk of hardware damage and restricts the mobile

application of DDAs. Alternatively, digital data may be

obtained by ﬁlm digitization or directly by the applica-

tion of CR.

Line Detectors. The classical concept of NDT with

line detectors is based on a ﬁxed radiation source,

moving objects and a ﬁxed line camera. This is

Part D 16.2

912 Part D Materials Performance Testing

X-rays

a) b)

Phosphor

Photodiode

Row select

Visible Photons

Photodiode

Row select

Visible Photons

Bias

Output

Bias

Output

Cesium

iodide

X-rays

Fig. 16.47a,b Principle of amorphous-silicon ﬂat panels with ﬂuorescence screens. (a) Additional unsharpness is gener-

ated in the phosphor layer due to light scattering. (b) Needle crystals of CsI on the surface of the photodiodes improve

the spatial resolution because the crystals conduct the light to the photodiodes like ﬁber light conductors

X-rays

a) b)

Diode

Row select

Electrons Electrons

Row select

Bias

Output

Bias

Output

Contact plate

Photo-

conductor

Collector

plate

X-rays

Fig. 16.48a,b Principle of direct converting ﬂat panels with amorphous silicon thin-ﬁlm transistor arrays for read out.

There is no light scattering process involved. The spatial resolution is determined by the pixel size of the detector

array. (a) Photodiodes convert directly the x-ray photons to electrons. This technique is suitable for low-energy ap-

plications. (b) A semiconductor (e.g. amorphous selenium or CdTe) is located on microelectrodes in a strong electrical

ﬁeld. Radiation generates charges, which can be stored in microcapacitors

the typical concept for baggage, car and truck in-

spection. Line detectors are available with a resolu-

tion of 0.25–50 mm. The most common principle is

the combination of scintillator and photodiodes. The

scintillator is selected in accordance to the energy

range.

Mechanized X-ray Inspection of Welds. High-reso-

lution lines have been introduced for weld inspec-

tion [16.28]. Figure 16.49 shows a mechanized x-ray

inspection system, which is based on an x-ray tube,

a manipulation system and a line camera. The cam-

era consists of an integrated circuit with about 1200

photodiodes. A GdOS scintillator screen is coupled to

the diodes. New applications take advantage of TDI

(time-delayed integration) technology to speed up the

scan [16.29]. Several hundred lines are used in parallel.

The signal is transferred from line to line synchronously

Fig. 16.49 View of the line scanner. The x-ray tube and

camera are mounted on a manipulator for mechanized in-

spection of pipes

Part D 16.2

Performance Control 16.2 Industrial Radiology 913

with the movement of the object or the scanning sys-

tem. The information is integrated on the chip and the

speed enhancement corresponds to the number of lines

used.

Specialized tomographic routines were developed

to reconstruct a three-dimensional (3-D) image of

the weld [16.28]. This method is very sensitive to

cracks and lack of fusion. The depths and shape of

these defects can be reconstructed and measured. Fig-

ure 16.50 shows the image of a reconstructed crack in an

austenitic girth weld in comparison to a cross-sectional

metallography.

Automated Evaluation of Digital X-ray Images:

Serial Inspection in Automotive Industry

Fast digital x-ray inspection systems are used in the

serial examination of industrial products since this

technique is capable of detecting ﬂaws rather dif-

ferent in their nature such as cracks, inclusions or

shrinkage. They enable a ﬂexible adjustment of the

beam direction and of the inspection perspective as

well as online viewing of the radioscopic image to

cover a broad range of different ﬂaws. This eco-

nomic and reliable technique has become of essential

signiﬁcance for different applications. The conﬁgura-

tion of such systems is schematically represented in

Fig. 16.51.

The object, irradiated from one side by x-rays,

causes a radioscopic transmission image in the detec-

tor plane via central projection. The relation between

the focus–detector distance (FDD) and the focus–object

distance (FOD) determines the geometrical magniﬁ-

cation of the image. An image converter such as an

x-ray image intensiﬁer, a ﬂuorescence screen or a dig-

ital detector array (DDA), also called a ﬂat-panel

Attenuation image

FDD

Digital signal

Test sample

Falt panel

detector

Image processing system

X-ray tube

FOD

Monitor

Monitor image

Data acquisition and

digital image

processing/evaluation

Fig. 16.51 Schematic set up of a digital industrial radiology system

Digital – Metallographic sectioning

Metallographic

cross sections

Tomographic

cross sections

Measured values:

Result

Etched

Austenitic weld sample

Not etched

5mm

Metallography –8.1 mm

Planartomography –7.7 mm

Fig. 16.50 Reconstructed crack in an austenitic weldment com-

pared to the metallographic cross sections

detector, converts the x-ray image to a digital im-

age.

Light alloy castings are widely used, especially

in automotive manufacturing. Due to imperfections

of the casting process, these components are prone

to material defects (e.g. shrinkage cavities, inclu-

sions). These parts are frequently used in safety-

relevant applications, such as steering gears, wheels

and increasingly wheel suspension components. These

parts have to undergo a 100% x-ray inspection for

safety.

A fully automated x-ray inspection system for

unattended inspection can guarantee objective and re-

producible defect detection (see example in Fig. 16.52).

The decision whether to accept or to reject a speci-

men is carried out according to the user’s acceptance

Part D 16.2

914 Part D Materials Performance Testing

Fig. 16.52 Defect detection with an automated system. Left:origi-

nal image; right: detected ﬂaw

speciﬁcation. These systems are known as automated

defect-recognition (ADR) units.

Automated x-ray inspection is also used for

a novel ﬁeld of application: the check of complete-

ness and function. For example, when the presence

and deformation of parts have to be checked. Fur-

A surface of a liquid

indicates chemical

warface agents

Fig. 16.53 Inspection of a grenade of World War I, ﬁlled with chem-

ical warfare. The chemical agent can be detected, by visualization

of the liquid-level line inside the grenade

Fig. 16.54a,b Radiograph of a shrapnel grenade: (a) original im-

age, and

(b) high-pass-ﬁltered image

thermore distances and spatial dimensions can be

examined.

16.2.5 Applications of Radiology

for Public Safety and Security

Public safety and security is yet another application

ﬁeld of digital industrial radiology. Since explosives

entail a considerable threat it is a central issue of

all security measures to detect them in time. There

exist a variety of different methods for explosive detec-

tion [16.30].

Inspection of Unexploded Ordnance (UXO)

Grenades containing warfare substances are identiﬁed

by the presence of a liquid surface visible in the respec-

tive radiograph (Fig. 16.53). They have to be inspected

to assign them to the appropriate way of safe disposal.

To detect a putative liquid ﬁlling, the suspected ammu-

nition is radiographically inspected in a tilted position.

If this shows the surface of the liquid, then the grenade

is ﬁlled with a warfare substance (such as e.g. mustard

gas or phosgene). Image processing reveals further de-

tails. Figure 16.54 shows how the digital radiograph

of a shrapnel grenade was enhanced by image pro-

cessing.

Dual-Energy Radiography for Baggage Control

The dual-energy technique is widespread in modern

baggage-inspection equipment. It provides additional

information about the chemical composition of the in-

spected material. The technique takes advantage of the

fact that radiation of different energy is absorbed dif-

ferently by the various elements. The absorption is

a function of the atomic number of the absorbing ele-

ments. As a consequence, radiographs taken at a lower

and a higher x-ray energy level are not identical. The

ratio between the logarithmic grey values of such a pair

of images represents the chemical composition of the

absorbing material and is independent of its layer thick-

ness. Figure 16.55 shows a dual x-ray tube set and

a typical dual-energy radiograph of a bag. This tech-

nique provides the operator with color-coded images,

where the color indicates the chemical composition

of the inspected specimen. This can provide hints of

the presence of explosives or contraband. Advanced

versions of inspection units include an additional spec-

trometer to measure the spectrum of the scattered

radiation. This method is rather sensitive and allows the

detection and speciﬁcation of various chemical com-

pounds within the object.

Part D 16.2

Performance Control 16.3 Computerized Tomography – Application to Organic Materials 915

Fig. 16.55 Dual x-ray tube assembly for dual-energy inspection of baggage and typical image of a bag. The atomic num-

ber of the components in the object are usually represented by the color: blue – high atomic numbers (e.g. metals); green

– medium atomic numbers (e.g. glass, polyvinyl chloride); yellow to brown – low atomic numbers (e.g. carbon hydrates)

16.3 Computerized Tomography – Application to Organic Materials

For tasks of materials performance control, tomog-

raphy with x-rays supported by computers, in short

computerized tomography (CT), is gradually spread-

ing in industry. Tomography, born in the medical ﬁeld,

makes it possible to show 0.1% differences in the den-

sity of materials. Moreover, the tomographic section of

an object under study, which is reconstructed in three-

dimensional space, is not affected by the rest of the

volume exposed to irradiation, and the geometry is well

deﬁned. The section-by-section reconstruction makes it

possible to achieve a complete exploration of the object.

For materials classes like polymers, such as natural rub-

ber (NR) or styrene butyl rubber (SBR), and composite

materials, such as polymer matrix composites, this

method is efﬁcient, gives high resolution and reliable

results. However, industrial application of a medical

scanner, as presented here, is applicable only for low-

density materials (4 < g/cm

). The application of CT

to materials of higher density is discussed in Sect.16.4.

We can discern four application ﬁelds for computer

tomography

1. nondestructive testing of big components;

2. nondestructive of small pieces;

3. materials study;

4. microtomography.

In this section, we will present the method and the ter-

minology within the scope in which the medical scanner

is commonly used. Then, we will deal with examples

chosen to share the performances of this technique for

industrial applications.

16.3.1 Principles of X-ray Tomography

When x-ray tomography is used, a collimated beam

goes through the tested object before being received by

a row of detectors opposite the x-ray source. The ob-

ject is then rotated through an angle of 180

◦

.During

the rotation, the attenuation of the intensity of the x-ray

beam is measured in a ﬁnite number of angular incre-

ments. Data acquisition is achieved by a computer that

carries out the reconstruction of the object with the help

of a suitable algorithm.

A x-ray CT is essentially based on the measure-

ment of the different absorption coefﬁcient of the

materials crossed by a penetrating beam such as x- or γ -

rays [16.31–39]. The CT image gives linear attenuation

coefﬁcients in the slice for each point, which depend on

the physical density of the material, its effective atomic

number and x-ray beam energy. The attenuation in the

intensity of a single x-ray beam is then deﬁned by the

equation of Lambert–Beer as the following;

I = I

exp(−μL) , (16.23)

where I

is the incident intensity, I is the transmitted

intensity, μ is the linear attenuation coefﬁcient and L is

the thickness of the absorber.

Part D 16.3

916 Part D Materials Performance Testing

Thus, a tomographic system measures the attenu-

ation of intensity from different angles to determine

cross-sectional conﬁguration with the aid of a comput-

erized reconstruction. The general view of the medical

scanner (x-ray CT) equipped with a mechanical test de-

vice – electrical jack plus carbon-ﬁber tube and also the

scanning system – is shown schematically in Fig. 16.56.

It consists of an x-ray source, a collimator, a detector

array and a computer system with data-storage media.

When going through a material, a beam of x-ray

undergoes an absorption that depends on the three fol-

lowing parameters.

1. The nature of the materials, or more exactly the re-

spective densities of the elements that make up the

material that is put in the path of the radiation (the

linear absorption coefﬁcient);

2. The thickness of each of these elements;

3. The incident intensity of the radiation.

The attenuation of the radiation through an object

complies with the Lambert–Beer law, which governs

every absorption phenomenon

I = I

μL

. (16.24)

The medical scanner is able to measure the x-ray atten-

uation in Hounsﬁeld units or tomography density even

if it does not give local mass density. To obtain this,

it is necessary to make a calibration versus a reference

which is water. The tomography density (DT) is a rel-

ative measure of the attenuation coefﬁcient with regard

to the water. For this reason, we need to know the value

given to water in the tomography scale in Hounsﬁeld

units (H). Basically with a constant X photon energy,

Sample

X-ray source

Collimator

Testpiece

rotation

Linear x-ray

detector

array

CT Image

Fig. 16.56 Medical scanner equipped with a mechanical test device (electrical jack plus carbon ﬁber tube) and schematic

of computed tomography system

the DT depends of the attenuation coefﬁcient by a theo-

rical relation where K is a constant equal to 1000

D =

μ −μ

. (16.25)

This relation of conversion between the attenuation

coefﬁcient and the tomographic density is based on wa-

ter attenuation coefﬁcient value, which is 1.8cm

−1

73 keV, corresponding to density zero in the Hounsﬁeld

tomography scale.

This conversion is done in two steps.

•

An intermediate scale is chosen. In this new scale,

for the water, there is a relation between K and C of

K =Cμ

(16.26)

and the attenuation coefﬁcient for a given material

is given by

int

μ. (16.27)

The tomographic density is linear with regard to the

intermediate attenuation μ

int

and the water density

must be equal to zero.

•

Make a translation in order to fulﬁl the condition

DT =μ

int

−K , (16.28)

so that

DT =

μ −K =

μ −μ

K . (16.29)

Calibration has been done with several materials in

order to have an empirical relation average DT of ma-

terials and respective attenuation coefﬁcient. Table 16.5

presents some results.

Part D 16.3

Performance Control 16.3 Computerized Tomography – Application to Organic Materials 917

Table 16.5 Attenuation coefﬁcients and tomographic den-

sities of some materials obtained with a medical scanner

Material μ(cm

−1

) DT (H)

Polyethylene 0.172 −72

Water 0.191 −0.4

Nylon 0.210 90

Polyester 0.217 139

Araldite 0.219 147

Rubber 18160/52 0.224 165

Delrin 0.262 349

Ebonite 0.288 434

Rubber 19199/48 0.340 729

Teﬂon 0.340 729

Conventional radiography makes it possible to show

defects bound up with variations of density compared

with the material as a whole. Thus, under most cir-

cumstances, inclusions, cracks, porosities, etc. can be

shown.

Nevertheless, radiography takes into account every-

thing the beam meets while going through the material

and gives a plane projection of them. That is why cracks

perpendicular to the beam are liable to escape detection.

Tomography breaks free from the drawbacks of

x-radiography by exploring the object with a 180

◦

ro-

tation, thanks to three-dimensional reconstruction, so it

is possible to know the absorption density of any point

within a given volume.

The units of absorption density are those used in

medicine. The following references are given:

1000 H (Hounsﬁeld) units Density of dried bone

0 Density of water

−1000 H units Density of air

This scale is suitable for composite materials whose

H density is about a few hundred units because of their

nature. The standardization carried out to associate the

medical unit with the density of the bulk Δ

gives for

one H unit: Δ

=1×10

−3

with a 120-kV x-ray source.

This implies that the medical scanner makes it pos-

sible to show variations of volumetric density of about

a thousandth.

The highest Z material that can be evaluated with

a medical scanner is aluminum.

The objectives of this paragraph are to present the

performances of x-ray tomography in order

•

to detect internal geometric defects,

•

to perceive the evolution of mechanical, physical,

chemical damage,

•

to display homogeneous variation related to pro-

cessing,

•

to observe differences in texture, chemical concen-

tration at the meso- and nanoscale.

16.3.2 Detection of Macroscopic Defects

in Materials

Using a medical scanner, metal matrix composites pro-

cessed by squeeze casting have studied. The ﬁrst step

was to test the alumina preforms, which are disks with

85 mm diameter and 15 mm thickness. In some disks,

spherical defects were introduced during the process-

ing. Figure 16.57 presents the results of the tomographic

examination. The average Hounsﬁeld density of the ce-

ramic preform is about −215. However, there is a large

difference of x-ray attenuation between the center of the

specimen and the edge, where the density is a maxi-

mum of 118. In fact, the density is lower in the center of

the specimen because the volume fraction of alumina is

smaller than near the edge. Thus, it is possible to check

the processing of the preform with tomographic obser-

vation as well as the repartition of alumina platelets

which is very difﬁcult using regular experimental meth-

ods. Of course, it is very easy to detect defects in

ceramic preforms, as is shown in Fig. 16.57.

Fig. 16.57 Defects in ceramic preform

Part D 16.3