ASM Metals HandBook Vol. 17 - Nondestructive Evaluation and Quality Control

Подождите немного. Документ загружается.

Microwave instruments for nondestructive inspection based on Doppler techniques are used for industrial in-motion

applications, particularly in Europe. The Doppler principle is based on a frequency shift in a wave that is reflected from a

moving target. The amount of frequency shift is proportional to the velocity of the target.

References cited in this section

3. A. Harvey, Microwave Engineering, Academic Press, 1963

9. G.W. McDaniel and D.Z. Robinson, Thermal Imaging by Means of the Evapograph, Appl. Opt.,

Vol 1, May

1962, p 311

10.

P.H. Kock and H. Oertel, Microwave Thermography 6, Proc. IEEE, Vol 55 (No. 3), March 1967, p 416

11.

H. Heislmair et al., State of the Art of Solid-State and Tube Transmitters, Microwaves and R.F.,

April 1983,

p 119

12.

H. Bierman, Microwave Tubes Reach New Power and Efficiency Levels, Microwave J., Feb 1987, p 26-42

13.

K.D. Gilbert and J.B. Sorci, Microwave Supercomponents Fulfill Expectations, Microwave J.,

Nov 1983, p

14.

E.C. Niehenke, Advanced Systems Need Supercomponents, Microwave J., Nov 1983, p 24

15.

W. Ts

ai, R. Gray, and A. Graziano, The Design of Supercomponents: High Density MIC Modules,

Microwave J., Nov 1983, p 81

Microwave Inspection

William L. Rollwitz, Southwest Research Institute

Thickness Gaging

Thickness measurements can be made with microwave techniques on both metallic and nonmetallic materials. For metals,

two reflected waves are used, one from each side of the spectrum, as shown in Fig. 16. The measurement is made using

the standing wave technique. When the wave is incident on a metal (electrically conductive), most of the wave is

reflected; only a small amount is transmitted (refracted). The transmitted wave is highly attenuated in the metal within the

first skin depth. For nonmetallic materials (electrically nonconductive), the reflected wave is much smaller than the

incident wave, so that any standing wave that does develop does not have a large amplitude.

Fig. 16

Diagram of a reflectometer for determining metal thickness using microwave techniques. Arms A and B

differ in length by an integral number of half-wavelengths for detector output null.

As shown in Fig. 17, the distance between the minima of the standing wave is equal to one-half wavelength; that is, when

a detector antenna moves a distance of one-half wavelength along a standing wave, the amplitude changes. Therefore,

with a reflectometer (shown schematically in Fig. 16), the thickness can be measured by forming a standing wave in the

detector arm of a hybrid tee, a standard microwave device. The wave from the generator is split by the hybrid tee into two

waves of equal amplitude. Each of these waves travels down separate waveguides, is reflected from the metal surface, and

travels back through the waveguide to the hybrid tee. The hybrid tee then recombines the two reflected waves, directing

them into the detector. One arm of the system has been changed in length to allow the sample to be inserted. When the

difference between the respective roundtrip distances traveled by the two waves is one-half wavelength, the two waves

interfere destructively, and a minimal voltage is read on the detector. The frequency of the generator can be adjusted so

that a minimal voltage is obtained for any given thickness of specimen. The specimen thickness can then be calculated

because the wavelength is known and the distance to one side of the sample is measured or calibrated.

Fig. 17 Response of

a detector moved along the path of interfering waves for large amplitude difference and

amplitudes nearly equal

The reflectometer shown in Fig. 16 can also be adapted to measure the thickness of an electrically nonconductive material

if one side is coated with a thin film of a conductive material. The difference between the two arm lengths would then be

a distance equivalent to twice the thickness of the nonconductive material. If the velocity of the wave in the material and

the frequency are known, the thickness can be computed.

If the dielectric material whose thickness is being measured also has absorption, the reflected wave is reduced in

amplitude such that a smaller-amplitude standing wave is obtained. The smaller-amplitude standing wave decreased the

accuracy of the measurements of the position of the minimum. A comparison of the standing waves for low and high

absorption is given in Fig. 18.

Fig. 18

Relation of the amplitude of reflected microwaves and the ratio of material thickness to wavelength.

Source: Ref 3

The range over which the simple thickness meter shown in Fig. 16 can be used unambiguously is approximately one-

fourth of the wavelength of the microwaves in the material being measured. If the thickness varies more than one-fourth

wavelength, ambiguous results will be obtained. The nominal thickness of most materials can usually be estimated or is

known within manufacturing tolerances to within plus or minus one-eighth wavelength, so that the ambiguity can be

resolved.

A diagram of another simple microwave thickness gage is shown in Fig. 19. This single-side gage has a direct waveguide

to the testpiece, which must be backed by a thin metallic coating. The standing wave that is set up in the waveguide is

detected by a probe in the waveguide itself. If the frequency were fixed and the probe were movable, it would trace out a

curve similar to those shown in Fig. 18. In practice, however, the probe is fixed and the frequency is swept. A material of

known thickness is inserted, and the center frequency and frequency sweep width for the microwave generator are

adjusted until a single one-half wavelength standing wave is observed on the oscilloscope. The sawtooth output from the

microwave generator is proportional to the instantaneous frequency shift, so that horizontal position on the oscilloscope

display is proportional to frequency.

Fig. 19 Diagram of a single-side microwave thickness gage. See text for discussion.

Also observed on the oscilloscope is the output signal from the frequency meter, which occurs as a sharp peak when the

frequency of the oscillator passes through the frequency to which the meter is set. This absorption peak appears

superimposed on the standing wave as a pulse (Fig. 20). When the equipment is properly set, the peak is not seen except

at the null of the standing wave, as shown in Fig. 20(a). When the material of known thickness is replaced by one that is

thinner, the peak appears on the standing wave, as shown in Fig. 20(b). When the material is greater than the known

thickness, the display is as shown in Fig. 20(c). By observing the movement of the peak with respect to the standing

wave, the material can be calibrated in thickness relative to the known thickness.

Fig. 20 Scope display and standing wave relationship for a single-sided microwave thickness gage.

See also

Fig. 19.

The use of a microwave thickness gage in a variety of production operations has demonstrated its suitability for

nondestructive inspection and quality control of dielectric components. The data for the graph in Fig. 21(a) was obtained

from polyethylene coating on common aluminum foil wrapping material. Measurements were made on coatings of

thicknesses of 25, 38, 44, and 51 m (0.0010, 0.0015, 0.00175, and 0.0020 in.). The data for Fig. 21(b) were obtained

from a fiberglass-reinforcement resin plate used for fabricating boat hulls. The data for Fig. 21(c) were obtained from

aluminum-backed thick polyurethane foam used as a thermal insulation blanket for space vehicles. All the data in Fig. 21

were taken with a simple amplitude-sensitive reflectometer operating at 9.4 GHz. An antenna (a 25 mm, or 1 in., square

horn) was used to obtain the data in Fig. 21(b) and 21(c). The data in Fig. 21(a) were obtained with an antenna known as a

short-field probe (Ref 16).

Fig. 21 Readings of an amplitude-sensitive reflectometer as a funct

ion of film thickness for (a) polyethylene,

(b) fiberglass plate, and (c) polyurethane

References cited in this section

3. A. Harvey, Microwave Engineering, Academic Press, 1963

16.

R.J. Botsco, Nondestructive Testing of Plastic With Microwaves, Parts 1 and 2, Plast. Des. Process.,

Nov

and Dec 1968

Microwave Inspection

William L. Rollwitz, Southwest Research Institute

Detection of Discontinuities

Discontinuities such as cracks, voids, delaminations, separations, and inclusions predominantly reflect or scatter

electromagnetic waves. Wherever these types of flaws occur, there is a more or less sharp boundary between two

materials having markedly different velocities for electromagnetic waves. At these boundaries, which are usually thin

compared to the wavelength of electromagnetic radiation being used, the electromagnetic wave is reflected, refracted, or

scattered. The reflected or scattered radiation has appreciable amplitude only if the minimum dimension of the

discontinuity is larger than about one-half the wavelength of the incident radiation in the material being tested.

Porosity and regions of defective material such as departures from nominal composition do not produce strong reflection

or scattering. They do influence the attenuation and the velocity of the electromagnetic wave. When there is absorption,

the transmitted wave has an exponential decay with respect to the distance traveled.

Continuous-wave reflectometers are those in which the amplitude of the standing wave is measured by a detector

in a combination send-receive, slotted microwave coaxial transmission line. They can be used to detect discontinuities in

such components as glass-filament solid-rocket motor chambers using a frequency of 12.4 to 18.0 GHz (Ref 17).

Reflections from internal discontinuities change the amplitude of the standing wave measured by the detector and give a

change in the output. When the material is scanned and the reflectometer signal is recorded by intensity modulating a pen

recorder, the discontinuities can be observed as light or dark areas. This type of C-scan, when connected with level-

sensing equipment, can be used to record only variations above or below a certain level, thus showing only the

discontinuities. A diagram of a continuous-wave microwave system that incorporates a slotted coaxial transmission line is

shown in Fig. 22.

Fig. 22 Diagram of a continuous-wave microwave system used to detect discontinuities

The detection of voids depends on a change in the wave reflected into the antenna resulting from the added reflection

from the discontinuity caused by the void. The system shown in Fig. 22 can be modified for these measurements by

substituting a directional coupler for the slotted line. This coupler allows the output signal to be proportional only to the

amplitude of the reflected signal. The reflected signal from components such as a rocket-motor chamber consists of a

large reflection from the outer surface of the case plus smaller ones from the glass insulation and the insulation/propellant

interfaces. At any void or unbonded area, an additional reflection occurs that increases the overall reflection into the

antenna. A standing wave can be set up by the reflection in the material so that, at some frequencies, complete

cancellation occurs for reflections that are one-fourth wavelength apart. Because of this possibility, frequencies should be

carefully chosen. Several frequencies are used to ensure that no cancellation will occur.

In the frequency-modulated reflectometer, a linear sweep is made of the entire frequency range of two

generators: 1 to 2 GHz and 12.4 to 18 GHz (roughly the L band and the K

band, respectively, per Table 2). A separate

system is used for each band; each band is swept at a rate of 40 sweeps per second. Figure 23(a) shows the basic

components of the frequency-modulated reflectometer (Ref 18). A graph of the signal at the detector as a function of time

is shown in Fig. 23(b). The transmitted signal from the antenna is reflected back into the antenna and the detector from

targets 1 and 2. The reflected signals also vary in frequency, as do the transmitted signals, but at the detector they are

displaced in time, as shown by the dashed lines in Fig. 23(b). Therefore, at any instant of time, there are three signals at

the detector: the first from the transmitted wave, the second from target 1, and the third from target 2.

Fig. 23 Setup and signal response of a frequency-modulated microwave reflectometer. (a) Schema

tic showing

components of a frequency-

modulated microwave reflectometer. (b) Paths of transmitted and reflected signals

at detector in relation to frequency and time

Each of these signals is of a different frequency, corresponding to the different propagation times involved. The frequency

difference, f, is equal to the rate of sweep (in hertz) times the delay (in seconds). The phase of the reflected signal, ,

relative to the reference (transmitted) signal is 40 f /c. Both the frequency and the phase of the reflected signals are

proportional to the distance to the respective reflectors. Small changes in the position of the reflector cause relatively

large changes in phase, which in turn cause large changes in the shape of the signal obtained at the crystal detector. As the

reflector moves through a distance equal to phase shifts of 0°, 90°, 180°, and 270°, the output of the detector, e( f), is a

function of the frequency difference and changes in shape as shown by the curves in Fig. 24.

Fig. 24 Theoretical indications of reflections, e( f), produced by a frequency-

modulated microwave

reflectometer during phase shifts of 0°, 90°, 180°,

and 270° as functions of frequency shift of the modulation

The arrangement of equipment for a frequency-modulated microwave reflectometer using L-band frequencies is shown

schematically in Fig. 25. In this setup, separate antennas are used for sending and receiving. The detector senses both the

transmitted signal and the reflected signals. These are mixed in the detector to produce a difference frequency equal to the

reflected wave, f, minus the reference frequency, f

, generated by the swept oscillator. A separate beat frequency is

developed for each reflected signal. For two discrete reflectors, the output signal of the crystal detector consists of two

frequencies, one for each target.

Fig. 25 Schematic of components of a frequency-modulated microwave reflectometer using L-

band

frequencies. Dimensions given in inches. Source: Ref 18

To determine the number of reflected objects and their locations, the signal from the detector passes through a cross-

correlation spectrum analyzer. The readout of the signal processor is a plot of the amplitude of the detector signal as a

function of frequency. To accomplish this, the signal is cut into many segments of equal duration. If the spacing of these

segments corresponds to one of the frequency components of the signal, an output results. The spacing of these segments

is varied by the long, saw-tooth voltage, A, so that all the possible frequencies in the signal are swept. Therefore, the

output contains a voltage for each of the frequency components of the signal. Each component is displayed as a dot on an

oscilloscope. The vertical displacement of this dot is proportional to the amplitude reached by the integrator for that

particular frequency in the integration time. The K-band equipment is similar to that shown in Fig. 25 for the L band

except that a waveguide is used rather than a coaxial cable.

Frequency-modulated reflectometers typically operate in the range of frequency bands given in Table 3. Sources are now

available to frequencies beyond 200 GHz. These would produce resolution to 1 mm (0.04 in.). Application of these higher

frequencies to inspection and measurement has been limited by the relatively high price of the microwave sources.

Table 3 Frequency bands of frequency-modulated reflectometers

Resolution in air

Resolution in plastics

(a)

Frequency, GHz

Type

mm in. mm

in.

0.5-1.0 Coaxial cable

300 12 150

5.9

1.0-2.0 Coaxial cable

150 5.9 75

3.0

2.0-4.0 Coaxial cable

75 3.0 37

1.5

4.0-8.0 Coaxial cable

37 1.5 18

0.71

8.0-12.4 Waveguide 34 1.3 17

0.67

12.4-18.0 Waveguide 28 1.1 14

0.55

18.0-26.0 Waveguide 19 0.75 10

0.39

26.0-40.0 Waveguide 11 0.43 5.5 0.22

(a)

This assumes a refractive index of 2.

Applications have been used in penetrating rock, soil, and thick solid propellant samples in the low bands; concrete in the

medium bands; and plastics and composite materials such as filament-wound glass, Kevlar, and honeycomb in the higher

bands.

References cited in this section

17.

M.W. Standart, A.D. Lucian, T.E. Eckert, and B.L. Lamb, "Development of Microwave NDT In

spection

Techniques for Large Solid Propellant Rocket Motors," NAS7-544, Final Report 1117, Aerojet-

General

Corporation, June 1969

18.

A.D. Lucian and R.W. Cribbs, The Development of Microwave NDT Technology for the Inspection of

Nonmetallic Materials and Composites, in

Proceedings of the Sixth Symposium on Nondestructive

Evaluation of Aerospace and Weapons Systems Components and Materials,

Western Periodicals Co., 1967,

p 199-232

Microwave Inspection

William L. Rollwitz, Southwest Research Institute

Microwave Detection of Surface Cracks In Metals

When an electromagnetic wave is incident on a metallic surface that has slits or cracks, the metallic surface reradiates

(reflects) a signal because of induced current. Under the proper conditions, the reflected wave combines with the incident

wave to produce a standing wave. The reflection from a surface without a crack is different from that surface with a crack

and depends on the direction of the incident wave polarization relative to the crack. When the crack is long and narrow

and the electric field is perpendicular to the length of the crack, the reflected wave (and therefore any standing wave) is

affected by the presence of the crack. The amount of change is used to determine the size and depth of the crack.

Standing Wave System. The most sensitive means of detecting the small crack-related changes in the standing wave

is to use the standing wave in a resonant cavity (Ref 19). Such a resonant system is also sensitive to variables other than

surface cracks. However, a nonresonant standing wave system can be used. One such system is shown in Fig. 26. The

detection head with the test specimen forming one end of the system is shown in Fig. 26(a). The excitation is fed into two

slots, while the receiver is fed from two other slots. A diagram of the microwave circuit is shown in Fig. 26(b). The test

specimen is mounted so that it can be rotated.