Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
MEASURING LAYER THICKNESS USING BACKSCATTERING OF HIGH ENERGY PHOTONS
Document Type and Number:
WIPO Patent Application WO/1996/007077
Kind Code:
A1
Abstract:
The present invention provides a method and apparatus for non destructive, in situ measuring thicknesses of layers on substrates. The method and device uses a probe of a radioactive source and a photodetector mounted behind the source for detection of backscattered photons. In one aspect the method is used to measure the thickness of paint deposited onto galvanized steel such as used in vehicles. A source containing radioactive 109Cd producing high energy photons of energy 23 and 25 keV is spaced from the painted surface so the photons impinge on the painted substrate. The intensity of photons backscattered by Compton scattering in the paint layer is proportional to the mass density of the paint to give a direct measurement of the paint thickness. The photons penetrating through to the substrate are absorbed within the substrate. In another aspect the method is utilized for measuring ice thickness on airfoils. The probe is mounted on the interior of the airfoil and the source is 241Am producing 60 keV photons which penetrate through the airfoil substrate and are backscattered within the ice layer and back through the airfoil substrate to the detector. The shape and density of the source holder in addition to the geometrical arrangement of the source and detector with respect to the airfoil substrate are used to block photons backscattered in the airfoil substrate thereby favoring scattering in the ice layer over that in the aluminum.

Inventors:
MacKENZIE, Innes, K.
Application Number:
PCT/CA1995/000498
Publication Date:
March 07, 1996
Filing Date:
August 30, 1995
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
UNIVERSITY OF GUELPH.
International Classes:
B64D15/20; G01B15/02; (IPC1-7): G01B15/02; B64D15/20
Foreign References:
EP0380226A11990-08-01
GB1385279A1975-02-26
Other References:
BATES J R ET AL: "GAMMA BACKSCATTER THICKNESS MEASUREMENT FOR CONTROL OF MULTIPLE- STRIP RUBBER CALENDERS", PROCEEDINGS OF THE ANNUAL CONFERENCE OF ELECTRICAL ENGINEERING PROBLEMS IN THE RUBBER AND PLASTICS INDUSTRIES, AKRON, APR. 15 - 16, 1991, no. CONF. 43, 15 April 1991 (1991-04-15), INSTITUTE OF ELECTRICAL AND ELECTRONICS ENGINEERS, pages 73 - 75, XP000299130
Download PDF:
Claims:
THEREFORE WHAT IS CLAIMED IS:
1. A method for measuring thickness of a layer on a surface of a substrate, comprising; a) directing primary photons produced from a suitable radioactive source toward a substrate surface being probed for the presence thereon of a layer; b) measuring an intensity of secondary photons backscattered from the layer using a photodetection means, shielding the photodetection means from primary photons from the source and photons backscattered in the substrate; and c) determining the thickness of the layer from the intensity of backscattered photons.
2. The method according to claim 1 wherein the step of directing primary photons includes adjusting the position of the radioactive source in the holder to control the collimation of the beam of primary photons.
3. The apparatus according to claim 2 wherein the layer is comprised substantially, but not exclusively, of atomic elements having atomic numbers less than the atomic numbers of the elements making up the substrate.
4. The method according to claim 3 wherein said layer is a paint layer disposed on a surface of a metal substrate.
5. The method according to claim 4 wherein the metal substrate is selected from the group consisting of aluminum, titanium, iron, copper, zinc and combinations thereof, the radioactive source producing primary photons having energies in the range from about 14 keV to about 25 keV, the source being located so that the surface on which the layer is disposed is in opposing relation to the source.
6. The method according to claim 5 wherein said substrate is an iron based substrate, the radioactive source being selected from the group consisting of "Co, 241Am and 109Cd.
7. The method according to claim 6 wherein said iron based substrate is galvanized steel and the radioactive source is 1∞Cd.
8. The method according to claim 7 wherein said radioactive source of 109Cd has a source strength in the range from about 0.3 millicuries to about 30 millicuries.
9. The method according to claim 3 wherein said layer is an ice which can form on an outer surface of the substrate.
10. The method according to claim 9 wherein the source is adjacent to an inner surface of the substrate, the radioactive source producing primary photons having sufficient energy to penetrate through the substrate into the ice layer on the outer surface, and wherein at least some of the photons backscattered in the ice layer penetrate back through the substrate to the photodetection means.
11. The method according to claim 10 wherein the substrate is an airfoil of an aircraft, the aircraft having a flight deck.
12. The method according to claim 11 including the step of probing for ice at a plurality of locations on the airfoil and displaying the ice thickness on the flight deck.
13. The method according to claim 12 wherein the radioactive source is 241Am producing primary photons of energy 60 keV.
14. The method according to claim 13 wherein the 2 1Am source has a radioactive source strength in the range of from about 0.8 microcuries to about 1 millicurie.
15. An apparatus for measuring thickness of a layer on a surface of a substrate, comprising; a) a holder having a front surface, a back and a cavity in the front surface; b) a photodetector positioned with respect to the source to measure the intensity of photons backscattered from layer on a surface of a substrate; and c) a radioactive source located in the cavity, the holder having an effective density and shape to direct a beam of primary photons from the radioactive source out of the cavity to irradiate the layer and to substantially block primary photons from the source and photons backscattered in the substrate but not photons backscattered from the layer from impinging on the photodetector.
16. The apparatus according to claim 15 wherein the holder includes means for adjusting the position of the radioactive source in the cavity to control the collimation of the beam of primary photons.
17. The apparatus according to claim 16 wherein the layer is comprised substantially but not exclusively of atomic elements having atomic numbers less than the atomic numbers of the elements making up the substrate.
18. The apparatus according to claim 17 wherein said photodetector includes a Nal(TI) Xray scintillator coupled with a photomultiplier detector.
19. The apparatus according to claim 17 wherein the holder is constructed of a metal selected from the group consisting of molybdenum, gold, platinum, lead, silver, tantalum and tungsten.
20. The apparatus according to claim 17 wherein the layer is a paint layer disposed on a metal substrate, the radioactive source being 109Cd.
21. The apparatus according to claim 20 wherein said substrate is a metal body of a vehicle undergoing a painting process, the holder being positioned with the front surface spaced from said metal body.
22. The apparatus according to claim 21 wherein said radioactive source of 10βCd has a radioactive source strength in the range of from about 0.3 millicuries to about 30 millicuries.
23. The apparatus according to claim 22 wherein said metal body is galvanized steel.
24. The apparatus according to claim 17 wherein the layer is ice disposed on a substrate, the radioactive source being 2 1Am.
25. The apparatus according to claim 24 wherein said substrate is an airfoil of an aircraft, the airfoil having an inner surface and an outer surface, the ice being disposed on the outer surface, the holder being positioned with the front surface adjacent to the inner surface.
26. The apparatus according to claim 25 wherein the 241Am source has a radioactive source strength in the range of from about 0.8 microcuries to about 1 millicurie.
27. The apparatus according to claim 26 wherein a source holder and associated radioactive source and photodetector forms a probe, including a plurality of probes attached to various locations on the interior surface of the airfoil, each probe providing an output indicative of an ice thickness at that location on the exterior of the airfoil.
28. The apparatus according to claim 27 including means for displaying the calculated the ice thicknesses at said various locations.
29. A method for measuring the thickness of an ice layer forming on a surface of a substrate, comprising; a) locating a suitable radioactive source adjacent to an inner surface of the substrate and directing primary photons through the substrate to an outer surface of the substrate on which an ice layer can form; b) measuring an intensity of secondary photons backscattered from the ice layer using a photodetection means, shielding the photodetection means from primary photons from the radioactive source and photons backscattered in the substrate; and c) determining the thickness of the ice layer from the intensity of photons backscattered from the ice layer.
30. The method according to claim 29 wherein the substrate is an airfoil of an aircraft, the aircraft having a flight deck.
31. The method according to claim 30 wherein the radioactive source is 241Am producing primary photons of energy 60 keV and the airfoil is an aluminum based metal.
32. The apparatus according to claim 31 wherein the 241Am source has a radioactive source strength in the range of from about 0.8 microcuries to about 1 millicurie.
33. The method according to claim 31 including measuring ice thickness at a plurality of locations on the airfoil and displaying the ice thicknesses on the flight deck.
Description:
MEASURING LAYER THICKNESS USING BACKSCATTERING OF HIGH ENERGY PROTONS

FIELD OF THE INVENTION

The present invention relates to an in situ and non-destructive method and device for measuring the thickness of layers on substrates using backscattering of high energy photons.

BACKGROUND OF THE INVENTION

The ability to measure, nondestructively and in situ the thickness of growing thin films is very advantageous in many industrial applications. For example, it is important to be able to monitor the thickness of paint being sprayed on cars, trucks or aircraft during production. The costs associated with painting vehicles, particularly in assembly line production is quite significant so that applying too thick a paint layer has serious economic repercussions. Alternatively, if there is too thin a paint layer this may result in the vehicle having to be repainted.

Post production painting of vehicles generally involves applying three distinct layers comprising a primer coating or layer applied directly to the metal substrate, a base coating containing the pigment applied on top of the primer coating and a clear coating applied on top of the base coating. The total thickness of these layers is about 0.05 mm to 0.10 mm with about half of the total thickness being due to the top clear coat. It is preferable that each layer be of uniform thickness and manufacturers are particularly concerned about controlling the thickness of the base coat; however the base coat is the thinnest layer (about 0.01 mm) which makes it very difficult to control its thickness.

There are several known ways of estimating the average thickness of the paint layers. One way is to simply weigh the paint used to cover a certain area and, knowing the mean density of the paint, calculate the average thickness which is generally expressed in units of mg/cm 2 , known as the "areal density." Disadvantages of this and similar techniques is it is not an in situ technique, it is very labour intensive and does not give any information

about the uniformity of the layers.

At present there is no single, reliable, economic method for accurate, in-situ and nondestructive monitoring of paint thickness as it is being applied to substrates. X-ray backscattering is one method which shows promise as a technique for estimating film thicknesses; however, this technique has severe limitations. A simplified model used in considering backscattering of x- rays from a paint layer on a metal backing is based on two assumptions: 1) the x-rays interact with the paint layer only by the mechanism of Compton scattering and because the total Compton scattering cross-section is almost exactly proportional to the mass, the backscattered x-ray intensity should be proportional to the mass/unit area of the paint layer over a broad range of x-ray energies; and 2) that x-rays penetrating through to the steel backing or substrate are fully attenuated or absorbed in the substrate and not back scattered. The assumption that the intensity of the backscattered x-rays from the paint is almost exactly proportional to the paint thickness over a broad range of x-ray energies usually holds because the paint layer is so thin and comprised of elements of low atomic number. The model breaks down generally because of the assumption that the metal panel is a perfect absorber over a broad range of energies. This will be more fully discussed below but this drawback has severely limited the application of x-ray backscattering as a viable in situ technique.

Another example of the importance of being able to monitor in situ film thickness relates to measurement of ice buildup on aircraft. Buildup of ice layers on aircraft wings or other materials has been and continues to be a cause of aircraft disasters. During or after takeoff of the aircraft the added weight of the ice, which can be very significant, as well as the accompanying change in aerodynamic flow patterns over the airflow surfaces can cause crashes. Preventative procedures such as de-icing the aircraft typically are carried out when the aircraft is near the hangar after which the aircraft taxis to the end of the runway for takeoff. During this period ice can again build up on the aircraft depending on the distance the aircraft must taxi and the severity of

the weather conditions.

One current method of measuring ice thickness on an airfoil uses microwave electromagnetic radiation. The microwave radiation is used to monitor the thickness and dielectric constant of the growing layer from which the composition is calculated. United States Patent Nos. Pat. No. 4,054,255 and 4,688,185 issued to Magenheim and Magenheim et al. respectively disclose using a dielectric layer affixed to the wing surface as a surface waveguide into which a low power microwave signal is directed. The impedance and reflection properties of the waveguide change as ice builds up on the waveguide and this change is measured and related to the buildup of ice.

Drawbacks to microwave monitoring systems are the expense of the power supplies and the need for sophisticated software for handling the data. Microwave monitoring systems necessitate cutting holes in the wings of the aircraft or otherwise modifying the wings to include waveguide elements which increase installation costs, disturb the flow pattern over the air foil and may reduce structural strength.

Another known method for monitoring ice build-up involves the use of internal reflection to measure ice thickness. United States Patent No. 4,797,660 issued to Rein Jr. teaches use of internal reflection of EM using a prism mounted to the wing surface. A light source and detector are positioned to impinge light onto the exposed surface of the prism and a detector measures internally reflected light from the exposed surface with the reflected intensity being a function of the buildup on the exposed prism surface. United States Patent No. 5,296,853 issued to Federow et al. is directed to a laser ice detector comprising a light source, light detector and temperature sensor with the light source and detector embedded in a plastic housing mounted flush with the surface of the wing. The system is designed to give total internal reflection when ice is absent from the plastic surface. The presence of ice on the plastic is accompanied by loss of total internal reflection. United States Patent No. 4,797,660 issued to Michoud et al. discloses an ice thickness measuring technique for aircraft using internal reflection of light. The device is designed to discriminate against water and ice

with for example falling rain acting to modulate the light signal received by the detector in a characteristic manner thereby distinguishing it over the signal observed to due to ice buildup. As with microwave techniques, a drawback to internal reflection is the need for modification of the airfoil surface. For safety considerations there is required a rapid and accurate method of measuring the build-up of ice on aircraft in flight. Ice build-up occurs predominantly on the ground and at low altitudes with little build-up occurring at normally high cruising altitudes for jet aircraft. However, the steady increase in air traffic unaccompanied by construction of more airports has resulted in the practise of "stacking up" low priority flights before giving clearance to land. This is particularly the case in inclement weather and at relatively low altitudes, conditions most conducive to icing. Therefore, there is a need for a rapid, accurate, economic, in situ and nondestructive method of measuring the thickness of growing films on substrates.

SUMMARY OF THE INVENTION The present invention provides a non-destructive, in-situ method of measuring thickness of layers comprising elements with low atomic numbers coated on disposed on a substrate during or after deposition on the substrate. The method comprises the steps of directing primary photons produced from a suitable radioactive source toward a substrate surface onto which a layer can be disposed, the layer being comprised of elements with atomic numbers less than the atomic numbers of the elements of the substrate. The intensity of secondary photons backscattered from the layer disposed on the substrate surface is measured using a photodetection means. The thickness of the layer is determined from the intensity of backscattered photons. There is provided means for adjusting the position of the source in the holder for controlling the collimation of the primary photon beam.

Where the method is used for measuring the thickness of a paint layer coating on a steel substrate, the radioactive source is preferably 109 Cd producing primary photons of energy of 22 and 25 keV, the holder being located so that the surface on which the paint layer is deposited is in opposing

relation to the holder.

Where the method is used for measuring the thickness of an ice layer coating an outer surface of the skin of an airfoil, there can be no obstructions placed on the outer airfoil surface. Therefore, the source holder is located sufficiently close to the inner surface of the airfoil to provide a beam of primary photons sufficiently collimated so that photons backscattered in the aluminum substrate are substantially attenuated by the holder. The radioactive source is preferably 241 Am producing primary photons of energy 60 keV having sufficient energy to penetrate the aluminum substrate to be backscattered from the ice layer back through the aluminum to impinge on the detector.

The present invention provides an apparatus for measuring thickness of layers on a substrate during or after the layer is formed on the substrate. The apparatus comprises a holder having a front surface, a back and a cavity in the front surface. A photodetector is positioned with respect to the source to measure the intensity of photons backscattered from the layer and the apparatus is provided with a radioactive source located in the cavity. The holder has an effective density and shape to direct a beam of primary photons from the radioactive source out of the cavity for irradiating the layer and to substantially block primary photons from the source and photons backscattered in the substrate but not photons backscattered from the layer from impinging on the photodetector. The holder is provided with means for adjusting the position of the source in the holder for controlling the collimation of the beam of primary photons.

BRIEF DESCRIPTION OF THE DRAWINGS

The method of measuring thickness in inhomogeneous layered systems according to the present invention will now be described, by way of example only, reference being had to the accompanying drawings, in which:

Figure 1 is a longitudinal section of a radioactive source, holder and detector for measuring thicknesses of layers on substrates according to the present invention;

Figure 2a illustrates the backscattered x-ray energy spectra for the

K-shell x-rays of silver from a copper target;

Figure 2b illustrates the backscattered x-ray energy spectra for the K-shell x-rays of silver from low density polyethylene; and

Figure 3 shows the ratio of the reflection of silver K-shell x-rays from thick specimens of various materials to that from a thick block of low density polyethylene;

Figure 4 is a plot of relative x-ray albedos versus x-ray energy and atomic number Z;

Figure 5 is a longitudinal section of a probe including radioactive source, holder and detector for measuring thickness of paint layers on substrates according to the present invention;

Figure 6 illustrates the use of a plurality of the probes of Figure 5 to measure the thickness of paint layers applied to a vehicle on an assembly line; Figure 7 is a plot illustrating the variation of count rate versus time due to solvent evaporation from the freshly spray painted surface;

Figure 8 is a longitudinal section of a radioactive source, holder and detector for measuring the buildup of ice on an airfoil substrate according to the present invention; Figure 9 illustrates the dependence of intensity of γ-ray backscattering on the thickness of the ice layer built-up on an aluminum alloy used for aircraft wings for three different aluminum alloy thicknesses;

Figure 10 shows the measured intensity of backscattered 60 keV γ-rays as a function of the thickness of plastic pseudo ice overlying 5052 aluminum alloy of 0.762 mm thickness in plot (a) and 0.559 mm in plot (b) wherein each layer of plastic corresponds to 0.147 mm of ice;

Figure 11 shows the measured intensity of backscattered 60 keV gamma-rays as a function of the thickness of plastic backing overlying 6061 aluminum alloy of 2.223 mm thickness wherein each layer of plastic corresponds to 0.147 mm of ice; and

Figure 12 illustrates the use of a plurality of the devices of Figure 8 to measure the thickness of ice on the airfoil of an aircraft.

DETAILED DESCRIPTION OF THE INVENTION A) BASIC CONFIGURATION OF SOURCE-DETECTOR-TARGET SYSTEM

The basic design and geometric arrangement of an axially or cylindrically symmetric detector-source geometry constructed in accordance with an aspect of the present invention will be discussed first followed by descriptions of preferred embodiments for the two specific applications of measurement of paint thickness on metal panels such as vehicles and monitoring of ice thickness on the airfoil of aircraft. The preferred embodiments of this invention illustrated in the drawings are not intended to be exhaustive or to limit the invention to the precise form disclosed so that the applications cited are exemplary in nature and are not intended to limit the scope of the invention. The particular applications disclosed herein are chosen to describe the principles of the invention and its applicable and practical use to thereby enable others skilled in the art to best utilize the invention. Referring to Figure 1 , a longitudinal section of a detector-source- target arrangement constructed in accordance with the present invention is shown at 20. A scintillation detector 22 includes a thin (1.0 mm) Nal(TI) scintillator 24 housed in an aluminum cylinder (not shown) of 5.08 cm external diameter and 15.24 cm in length which also houses a photomultiplier 26. A protective covering 28 such as mylar extends across the scintillator. A lead shielding 30 is provided around the sides of detector 22 to minimize multiple scattering from nearby objects.

A source holder 32 is provided with a longitudinal cavity 34 extending partly therethrough for holding a radioactive source 38. Holder 32 is shown as being tubular with a radius R, and cavity 34 defines a detector axis

36. Holder 32, also referred to as an absorber post, is fabricated of a sufficiently thick and dense material so that primary radiation from source 38 is blocked or absorbed before hitting detector 22 below the source.

Radioactive source 38 is preferably a commercially available sealed source of x-rays or γ-rays typically 3.0 mm in length and diameter.

Source 38 sits on a threaded stud 42 threadable movable in cavity 34. Source 38 sits at an adjustable depth Z^ below the top surface 40 of holder 32. The

geometry and structure of holder 32 are such that with a source 38 in the holder, the primary radiation moves upwards in a cone whose half angle is adjustable by the depth Z The area of a target 46 (shown as two contiguous layers 48 and 50) spaced a distance Z 2 from surface 40 irradiated by the source is determined by both the half angle and the spacing Z-.. Target 46 is therefore located with respect to the source 38 and detector 22 array so that it intercepts the cone of primary photons emanating from the source which may interact with the target in several ways to produce a variety of secondary radiation. The diameter of source holder 32 may vary from about 5 mm to about 22 mm and the holder may be fabricated of gold or other suitable high density material depending on the application. For example, platinum, tungsten, silver, molybdenum, lead and tantalum may all be used as materials for the source holder. The detector assembly may optionally include an iris 54 defining an aperture 56 and having an inner radius R 2 symmetrically disposed with respect to source holder 32. Iris 54 is formed of a material which acts to absorb x-rays and γ-rays. Therefore, the backscattered photons can reach detector 22 only by passing through the annulus defined by radius R 1 of the source holder and R 2 of iris 54. Holder 32 blocks primary radiation from the source impinging on the detector. Iris 54 is optional since holder 32 is preferably made of a material having an effective density and shape to substantially block photons from the source from impinging on the photodetector and so is not required for some applications described herein.

The variables of the detector-source-target system include the dimensions R v R 2 , Z Z , the presence or absence of iris 54 and the choice of radioactive source 38. There are in principle several radioactive sources which may be used with the choice of radioactive source being dependent on the specific application. The preferred radioactive sources used in the present invention include 241 Am and 109 Cd and 57 Co. Further details of the detector- source configuration are found in Innes K. MacKenzie, An Axially Symmetric

Gamma-Ray Backscatter System For Dumond Spectrometry, Nuclear Instruments and Methods in Physics Research A299 (1990) 377-381.

B) MEASUREMENT OF PAINT THICKNESS ON SUBSTRATES USING X- RAY BACKSCATTERING

A drawback to previous attempts to use x-ray backscattering for measuring paint thickness on for example vehicles on an assembly line using the soft 6 kiloelectron volts (keV) x-rays from 55 Fe was strong absorption of the x-rays in the air gap between the probe and the paint layer. Sway of vehicles on an assembly line is unavoidable and so a minimum air gap of about 2.5 cm preferably exists between the probe and the surface being painted to avoid physical contact between the probe and panel being painted.

A more fundamental limitation of x-ray backscattering using low energy x-rays relates to the fact that typically Rayleigh scattering from the metal substrate has been assumed negligible and therefore ignored when interpreting the data. The inventor however has determined that Rayleigh scattering contributes significantly to the backscattered intensity when using soft x-ray photons of energy 6 keV produced by 55 Fe. As discussed above, the simplest model used in considering backscattering of x-rays from a paint layer on a metal backing is based on the assumption that the x-rays are reflected by the paint layer only and x-rays penetrating through to the steel backing or substrate are fully attenuated or absorbed and not backscattered. In this model the x-rays interact with the paint layer only by the mechanism of Compton backscattering.

While the assumption that the intensity of backscattered x-rays from the paint is almost exactly proportional to the paint thickness over a broad range of x-ray energies is valid, the model breaks down with the assumption that the metal backing is a perfect absorber. Using steel as an example, it is always true that absorption has to compete with scattering in the steel. A good approximation to Compton scattering per atom is that it is proportional to the atomic number Z whereas photoelectric absorption is proportional to Z 5 so the ratio is approximately proportional to Z A . Hence steel (iron) is a vastly better absorber than paint which comprises primarily carbon, hydrogen and other low-

Z elements.

However, there is another important consideration. That is that the photoelectric absorption cross section varies roughly as the inverse 7/2 power

of the x-ray energy. To illustrate the importance of this, consider the mass attenuation coefficients as a function of x-ray energy for copper given in Table I. Table I compares x-ray mass attenuation coefficients as a function of x-ray energies for several metals. At x-ray energies above 14 keV the mass attenuation coefficient varies fairly slowly with energy. That is because attenuation due to the photoelectric effect is weak in this regime and much of the attenuation is caused by the slowly varying Compton effect. At energies from about 8.95 to 13 keV the changes are very rapid and this corresponds to the region where photoelectric absorption dominates. At about 6 keV the attenuation in copper is about 23 times as great as at 34.92 keV. A similar situation applies to iron (not shown) so that, when considering only photoelectric absorption and Compton scattering one is led to the conclusion that in order to achieve the ideal of high absorption in the metal substrate very low x-ray energies are required. This prior art model ignores the effect of Rayleigh scattering which in certain energy ranges may effectively compete or dominate photoelectric absoφtion and/or Compton scattering. At low energies such as 6 keV poor contrast between paint and steel is obtained due to the backscattered intensity being dominated by Rayleigh scattering instead of the expected Compton scattering.

The backscattered energy spectra for the K x-rays of Ag (22 and 25 keV) incident on copper are shown in Figure 2(a), and for polyethylene in Figure 2(b). The intensity vs. energy spectra for polyethylene (CH 2 ) n , shows only two peaks A1 and A2 for polyethylene which are due to Compton scattering proving that it is an almost ideal Compton scatterer at 22 and 25 kev.

When the same measurement is made on copper there are two additional Rayleigh peaks B1 and B2 that are more intense than the Compton peaks A1 and A2.

The dominance of Rayleigh scattering increases rapidly at lower x-ray energies. Therefore, for a paint film on a steel metal substrate and using an 55 Fe source of x-rays of energy of approximately 6 keV, Compton scattering from the steel will be considered insignificant compared to Rayleigh scattering

from the steel and so the contributions to the backscattered x-ray intensity are Compton scattering from the paint and Rayleigh scattering from the steel. Therefore the contrast between the backscattering from the steel substrate and the paint is destroyed and most of the intensity of the backscattered photons comes from the steel even with very thick paint layers so that low energy soft x-rays cannot be used reliably to measure paint thickness on steel substrates. The method of the present invention is based on the fact that in order to measure the thickness of materials of low atomic numbers coated on substrates with high atomic numbers, one must use high energy photons in the appropriate energy range such that Compton scattering from the layers comprised substantially of elements of low atomic number Z competes successfully with the total of Compton and Rayleigh scattering from the substrate with high atomic number. The choice of options is very limited because there are few radioactive sources having a reasonable half-life that produce primary photons of suitable energy.

The influence of the underlying substrate material is best understood with reference to Figure 3 which shows the ratio of the reflection of silver K x-rays (22 and 25 keV) from very thick specimens of various materials to that from a thick block of low-density polyethylene. Hereinafter this relative reflectivity ratio is referred to as the albedo (whiteness). Paint is a low-Z mixture with an albedo close to 1.0. An ideal backing material would exhibit an albedo of zero but it is clear that this ideal is unattainable. The lowest albedo value for the K x-rays of silver is about 0.025 is obtained using a zinc substrate; i.e. zinc reflects about 1/40th as much as polyethylene. Iron exhibits an albedo slightly higher than zinc but the difference is significant as will be further discussed below.

Figure 4 illustrates a plot of relative x-ray albedos versus x-ray energy and atomic number Z which the inventor has discovered is most useful for selecting the best x-ray source for a given task. The best results are obtained for the largest contrast (i.e. ratio of albedos) between the low-Z paint and the backing material. For measuring paint layer thickness on galvanized steel the preferred sources will produce primary photons in the energy range

of about 14 keV to about 25 keV with Figure 4 showing that a 109 Cd source of the Ag K-shell primary photons of energy of 22 keV and 25 keV, is a most preferred source. For measuring the thickness of paint layers on steel panels, these Ag K-shell photons at 22 keV and 25 keV produce better contrast than the N p L photons from a 241 Am source having an energy of 17 keV, the 14 keV photons from a 57 Co source and substantially better contrast than the soft 6 keV x-rays from M Fe.

Referring to Figure 5, the method for measuring thickness of a paint layer 60 on a substrate 62 involves positioning a probe 58, comprising detector 22 and source holder 32 containing a small x-ray source 38, at a distance of about 1 cm (or further depending on the anticipated sway) from panel 62 and measuring the intensity of the backscattered x-rays. Since paint layer 60 is much thinner than substrate 62 and because it is comprised of elements of low atomic number, the intensity of the x-rays backscattered by the paint will be proportional to the thickness of the paint.

The use of the 22 and 25 keV primary photons from 109 Cd as part of the paint thickness monitoring device is also very advantageous because these photons are attenuated to a much lesser extent by air than the 6 keV photons of 55 Fe. This is particularly problematic on production lines where side- to-side swaying of the vehicle occurs as it moves along the production line so that the probe 58 must be spaced at least 1 cm from the vehicle depending on the amount of sway.

The data for different colored paint layers on steel panels are summarized in Table II at the end of the description which displays the backscattered x-ray intensity from several General Motors panels each painted with a different color. The instrumentation used to obtain the results disclosed herein was produced by Ludlum Measurements Inc. (Sweetwater, Texas, USA) and the radioactive source used was a 109 Cd source of Ag K x-rays (22 and 25 keV) having a strength of about 0.3 millicuries which produces a counting rate on steel panels of only about 500 cps. The last column in Table II was obtained by an ancillary measurement using a clear plastic sheet of thickness 0.10 mm having an areal density of 13.2 mg/cm 2 adhered to a bare steel panel. This

gave a ratio of 1.406 and served as a calibration for the paint layers.

The results in Table II are noteworthy for two reasons. First, the accuracy is about 2% for the estimate of areal density. Secondly, this required 100 seconds using the weak x-ray source but those skilled in the art will appreciate that it would for example require only 10 seconds with a stronger source of strength of 3 millicuries and only about 1 second with a 30 millicuries source and a fast counting system. As the source strength decreases the precision decreases (roughly as the square root of the inverse source strength). The operative upper limit of the source strength is determined by the counting rate the detection system can tolerate. Use of detectors with high counting efficiency allows weak, non-hazardous sources to be used.

In order to more fully illustrate the effect of substrate composition on backscattering intensity studies were conducted using a paint coating of 10 mg/cm 2 on a) an aluminum panel; b) a steel panel; and c) a zinc panel. The electronic system employed can count at 5,000 counts/sec (cps) and the source strength is chosen to provide that rate. A system for aluminum would use a much weaker source than one for steel, and one for zinc would use a stronger source than for steel. Of the total counts, the fraction, f, contributed by the paint is (a) 0.057 for aluminum, (b) 0.258 for steel and (c) 0.304 for zinc. If the counting rate is 5,000 cps for a time of t seconds the counts due to the paint is given by:

{5,000t - 5,000(1 -f)} = 5,000 (f)(t) and the % of error in this count is:

100 x (5.000t + 5.000(1 -fffl* 5,000 (f)(t)

If it is desirable to measure the intensity to a precision of 5% the counting times are (a) 48 seconds for the aluminum panel, (b) 2.1 seconds for the steel panel and (c) 1.5 seconds for the zinc panel. To summarize, the background scattering from the panel interferes with the measurement in two ways. It forces the use of a weaker source in order to avoid saturating the counting system and secondly, it introduces statistical errors that force longer counting times to achieve a given precision.

The fairly modest goal of 5% precision requires about a minute of counting on Al panels and 1% precision would require about 25 minutes.

The situation is much better with steel panels. In this case, a precision of 5% is obtained in about 2 seconds of counting and 1 % precision in 1 minute. These figures can be reduced by about 30% by using galvanized iron (almost equivalent to zinc) if the time is a critical factor. Therefore, the present method is highly advantageous for assembly line painting of vehicles comprising a substrate of galvanized iron, shown in Figure 5 as a zinc layer 61 on an ferrous based substrate 62. If it is desired to use the present counting system near to its design limits then stronger x-ray sources may be employed. Furthermore, the present counting system is an extremely simple, conservative design. It is well within the capabilities of currently available instrumentation to operate at a counting rate of 50,000 cps. Hence, where speed of analysis is essential, such as when the thickness monitoring system forms part of a feedback loop in a system controlling the painting operations, the system is readily adaptable to count at 100 times the rate used in acquiring the data for Table II. In other words, the counts acquired in 100 seconds can be obtained in 1 second. There would be substantially no sacrifice in the accuracy of the data but the high- speed instrumentation could cost several times as much as the low-speed system.

Referring to Figure 6, there is shown generally at 70 a plurality of probes 58 mounted on a frame 72 forming part of an assembly line for painting a vehicle 74. The output of each probe 58 is input into a computer 76 by wires 78 which may be used for monitoring paint thickness during painting and which may be used as a feedback element for controlling the painting process. In order to avoid the problem of the probes being coated by paint during the painting operation they may be disposed along the painting line between painting stations and suitably shielded. Alternatively, the probes could be positioned within the painting stations in retractable, shielded enclosures so that the painting operation can be interrupted and the probes moved into position and unshielded to measure the paint thickness.

The present method may be used to probe paint thickness during the application process before it has dried as well as being used as a probe for the dried paint. If the probe is used as part of a feedback element in a painting process, then a calibration must be used to compensate for evaporation of solvents from the paint after it is applied to the substrate. Paints applied using spraying include a solvent carrier which comprise low-Z elements. The solvent component could exceed 50% of the total volume of the paint mixture. The solvent will also contribute to the backscattered intensity so that each mixture being used would have to be characterized. In other words the x-ray measurement of paint thickness disclosed herein does not distinguish between paint and solvent per se since both are comprised substantially of elements of low atomic number, rather, the method measures the total areal density of matter deposited onto the substrate surface.

Figure 7 shows the decrease in counting rate versus time after spraying for one particular commercially available spray paint. The initial rate of decay is very rapid and is followed by an approximately linear decay for about 2 hours, corresponding to a linear loss of solvent from the layer. After about 16 hours the readings become very stable suggesting solvent evaporation has substantially ceased. The drying rate (solvent evaporation rate) depends on ambient temperature and air flow over the painted surface.

Therefore, in order to use the method for monitoring paint thickness shortly after application to the substrate, the effect on drying rate of composition of the spray paint, ambient temperature, rate of airflow over the substrate and the like is required. The method of measuring paint thickness is most advantageous when the paint is comprised substantially of low-Z atomic elements. However, the method is still advantageous with paints having some high-Z atomic elements present. For example, some paints include titanium dioxide powder. The thickness of these types of paint can be measured using the present technique as long as the impact of the higher atomic element components is accounted for in the calibration procedure.

Referring again to Figures 3 and 4, those skilled in the art will

appreciate that when Ag K x-rays produced by 109 Cd are used, zinc exhibits the lowest albedo and hence the best contrast. The thickness of paint layers on other substrates such as ungalvanized steel, Fe, Ti, Si and Al and the like may also be measured using the present method but will exhibit lower contrast than with galvanized steel having a zinc coating so that weaker sources and longer counting times are required.

C) MEASUREMENT OF ICE THICKNESS ON AIRCRAFT USING GAMMA- RAY BACKSCATTERING

The above described method for measuring paint thicknesses on metal may be used for measuring ice thickness on aircraft but using higher energy γ-rays instead of x-rays. One or more devices containing a γ-ray source are installed on the inside of the leading edges along the aircraft wing or tail section with the device containing a detector to measure γ-ray backscattering from ice forming on the wing. The fixed installations may be adapted to produce data continuously on the status of ice forming on the wings of the aircraft both on the ground and in the air.

Referring to Figure 8, a probe shown generally at 90 is provided for measuring ice thickness of an ice layer 92 on an airfoil 94 having an inner surface 96. Airfoil means aircraft parts with curved or flat surfaces such as the wings or rudder responsible for keeping the aircraft aloft during flight. To use the method of the present invention specifically to measure of ice thickness on an airfoil or leading edges of the aircraft, the source and detector must not be located on the exterior of the airfoil for aerodynamic considerations. This necessitates locating detector 22 and source holder 32' containing a radioactive source 98 on the interior of the airfoil so that the probe monitors the buildup of ice on the exterior airfoil surface. Aluminum or aluminum based alloys are currently the preferred material of construction of airfoils. The position of source 98 can be adjusted in cavity 34 by raising or lowering threaded stud 42. A radioactive source producing primary photons with sufficient energy to penetrate the aluminum is required and the geometry of the source holder is advantageously used to block photons backscattered by

the aluminum from reaching the detector while photons backscattered in the substrate are blocked by holder 32' from impinging on detector 22, see broken lines in Figure 8. Probe 90 preferably comprises an 241 Am radioactive source which produces γ-rays of energy 60 keV. The backscattered photons, which are reduced in energy to about 48 keV, are also energetic enough to penetrate back through the aluminum to impinge on detector 22. Holder 32' has an effective density and shape to substantially block or attenuate primary photons directly from source 98 and backscattered from the aluminum from impinging on photodetector 22. Probe 90 abuts against inner surface 96 of airfoil 94 with surface 40 of holder 32' preferably abutting the inner surface to ensure most of the photons backscattered in the aluminum are blocked or attenuated in source holder 32' while photons backscattered in ice layer 92 reach detector 22. Brackets 102 or other attachment means may be used to secure probe 90 to the interior of the airfoil. A tubular or cylindrically shaped source holders made of gold, tantalum and molybdenum and the like having a radius in the range from about 3.5 mm to about 7 mm provide suitable results.

Figure 9 shows the results of backscattering of 60 keV γ-rays from a flat sheet of aluminum placed 3.6 mm from the γ-ray source. With a 2 1 Am source strength of 200 microcuries (Isotope Developments Laboratory), the backscattered intensity from the aluminum was 139 kilocounts per minute

(kcpm) for 1.016 mm aluminum; 112 kcpm for 0.762 mm thick aluminum and 76 kcpm for 0.47 mm thick aluminum. Ice was simulated by placing thin sheets of plastic on the side of the aluminum opposite the side on which the source was located. The readings were converted to show ice on the x- and y-axis of Figure 9. The intensity of backscattering is expressed as a percentage increase above the thickness reading for aluminum alone. The increase is linear in the thickness of plastic for each thickness of aluminum. Radioactive source strengths in the range from about 0.8 microcuries to about 1 millicurie are preferred but as with the paint thickness measurement, the higher source strengths provide faster measurement times.

In addition to aluminum, airfoils may also be constructed from titanium or carbon based composites. For titanium airfoil of substantially the

same areal density as the aluminum airfoil it is contemplated that the same sensitivity to build-up of ice will be achieved using the probe of Figure 8. The inventor contemplates that for airfoil materials The present method can be used in situations in which the substrate is a laminar structure comprising more than one material as long as both materials can be penetrated by the γ-rays and backscattered secondary photons.

The present method for detecting ice buildup on metal surfaces does not depend per se on any properties of ice; it merely detects the additional low-Z material adhering to the outer surface of the wing. Therefore, those skilled in the art will appreciate that the sensors used on aircraft must be strategically located on the airfoil to give the best results. In addition, the backscatter intensity is sensitive to changes in geometry so that small distortions due to aerodynamic forces may cause a change in background signal for an aircraft on the ground and in the air so that airborne calibration may be required. A plurality of probes are preferably used since the ice in many circumstances may not form a uniform continuous layer across the airfoil. With measurement times of seconds (depending on the source strength) the present method is an in situ technique so that the ice layer can be detected and its thickness determined while it is forming as well as after it has formed. Referring to Figure 10, in another set of studies, the source and detector assembly was fixed 4 mm from two sheets of 5052 aluminum alloy, (a) 0.762 mm and (b) 0.559 mm. The intensity of backscattered γ-rays was measured as layers of plastic were firmly secured to the aluminum. Each plastic layer had a thickness of 0.089 mm, equivalent to 0.147 mm of ice. The backscattering results are displayed in Figure 10.

Figure 11 shows the backscatter intensity of 60 keV photons versus as a function of thickness of the plastic sheets (emulating ice) on a sheet of 6061 aluminum alloy with a thickness of 2.223 mm. In this configuration the source/detector array was bolted directly to the aluminum substrate metal in order to eliminate relative motion between the source and substrate.

In another study a uniform layer of water was frozen on one side

of a sheet of 5052 aluminum alloy of thickness 0.559 mm. The ice thickness was calculated by weighing the aluminum substrate with and without ice and measuring the surface area. The ice thickness was then measured using the present method. The summary of the results are: Calculated Ice Thickness Measured Thickness % Error

51.0 mils 53.4 mils 4.7%

30.9 mils 33.1 mils 7.1%

Figure 12 illustrates an aircraft 110 in which a plurality of probes 90 are disposed along the interior of the leading edges of the wings 112 and tail stabilizers 114. Outputs from each probe 90 are input into a microprocessor

116 located on the flight deck 118 for displaying the outputs to the flight crew. The method disclosed herein of monitoring ice thickness on aircraft is advantageous over prior art methods because it does not require drilling holes in the wing thereby avoiding aerodynamic problems associated with interference of air flow over the wing and weakening of the airfoil. Further, the use of the weak radioactive sources provides a very rapid measurement while at the same time it avoids the need for expensive RF or microwave generators.

The method of measuring thickness of layers in inhomogeneous layered systems disclosed herein has been described in the broadest sense with alternative embodiments for measuring paint thickness on steel panels and ice thickness on aluminum panels being illustrated and described. Those skilled in the art will appreciate that central to each application is the use of hindrance to selectively sample backscattered photons to ensure photons backscattered from the layer of interest are probed. This use of hindrance can include a) absorption by the specimen; 2) blocking or shadowing by the radioactive source holder; 3) collimation of the primary photon beam by varying the position of the source in the source holder; 4) choice of radioactive source and hence primary photon energy to increase absorption by the specimen or to decrease absorption by the substrate on which the layer being probed is located; and 5) choice of substrate.

The method disclosed herein is very advantageous for detecting the presence, and measuring the thickness, of layers on substrates using a

simple photodetector and commercially available counting electronics. The thickness of the layer can be measured very rapidly with the measurement time dependent on the source strength. Minor modifications to the source holder allows control of the sensitivity as a function of depth in the inhomogeneous layered target. Thus, minor modifications enable in situ thickness measurements in such diverse applications as paint layers on substrates (source/detector in front of the substrate) to ice on airfoil surfaces (source/detector behind the substrate) to be realized.

Therefore, while the present invention has been described and illustrated with respect to the preferred embodiments for measuring the thickness of paint layers and ice on airfoil surfaces, it will be appreciated that numerous variations of these embodiments may be made depending on the application without departing from the scope of the invention as described herein.

TABLE I

Energy (keV) Mass attenuation coefficients (cm 2 g "1 )

Copper (Z=29)

6.00 131.0

7.00 73.6

8.00 50.3

8.59 40.7

8.75 40.3

8.83 39.5

8.95 306.0

9.00 290.0

9.12 278.0

9.30 262.0

10.00 206.0

11.00 159.0

13.20 98.5

14.23 83.8

15.40 67.6

16.62 53.4

18.20 41.9

20.39 30.5

22.34 24.1

24.89 17.6

27.99 12.8

34.92 6.3

39.86 5.6

TABLE II

PANEL COLOR INTENSITY RATIO TOTAL PAINT

NO. c/1 OOsec w.r.t. bare THICKNESS panel mg/cm 2

1 bare 37,413 1.000 0

2 black 50,926 1.361 11.7 ± .08

3 white 53,823 1.439 14.3 ± .021

4 maroon 50,403 1.347 11.3 ± .018

5 red 53,905 1.441 14.3 ± .021

6 sandy 50,608 1.353 11.5 ± .018