A method for radiographic determination of the thickness of objects; and a device for reducing scattered radiation, for radiographic diagnostic examination of objects

Abstract

The subject of the present invention concerns the non-destructive radiological determination and testing of the thickness of objects. The key principle of the procedure is that the film / detectors currently used for non-destructive material testing, in particular AgBr2-based X-ray films and GOS-based flat-panel detectors, can be used, after appropriate validation and calibration, for measuring material thicknesses beyond their state-of-the-art measurement range; in such a way that these film / detector devices - with computer extension - display and interpret radiographically displayable sections of the objects under examination, which sections are not influenced by the charge in the object under examination. Furthermore: The device according to the present invention application is a duplex Bucky or Lisholm grid designed for industrial radiography to reduce scattered radiation. This device may be supplemented with the so-called filters.

Description

[0001] A METHOD FOR RADIOGRAPHIC DETERMINATION OF THE THICKNESS OF OBJECTS; AND A DEVICE FOR REDUCING SCATTERED RADIATION, FOR RADIOGRAPHIC DIAGNOSTIC EXAMINATION OF OBJECTS

[0002] The subject of the invention is a non-destructive testing method for industrial practice and a set-up for its implementation, which allows to determine the material thickness of the object to be tested, in particular industrial pipes and pipelines, using imaging radiation, preferably X-ray / gamma radiation. One of the conditions for the applicability of the technique is that the device behind the object, which receives the test radiation, must also be capable of imaging. In accordance with this, here and in the following, the term 'radiation absorbing device' is preferably used to refer, by way of example, to X-ray film or an electronic (digital) imaging detector, and thus specifically to an X-ray cassette (so-called flat-panel 'detector'). In the process, radiation is emitted from an essentially point source (i.e. quasi-point source) through the object under examination to a device receiving said radiation, on which absorption measurements are made at a given image resolution. The object to be examined has an attenuation effect on the intensity of the imaging radiation emitted by the source and passing through it, which can be used for diagnostic imaging and material thickness measurement. Here and hereafter, the image resolution of the radiation absorbing device is given in pixels, where the set of pixels of the same size and shape considered covers substantially the entire incident radiation absorbing surface of the radiation absorbing device used.

[0003] The measurement of material thickness itself is a long-established industrial practice, e.g. for measuring plate thicknesses, coatings, etc. Suitable and applicable radiation absorbing devices for the detection of the measurement are various solid state detectors, scintillators [e.g. Li, Ge, Nal(TI)]; while the radiation sources are narrow spectrum X-ray / gamma-ray sources; e.g. Am241: 59.37 keV; Na22: 511 keV, 1280 keV; Co57:122 keV; Cs137: 662 keV.

[0004] The absorption effect on which the measurements are based is described by the well-known relation ln(l / l0) = -μx, where l0is the intensity of the radiation incident on the surface of the radiation-absorbing device without the interposition of an object ("empty" film / detector), which is essentially equal to the intensity of the radiation emitted by the radiation source; I is the intensity of the radiation passing through the object under test, μ is the linear radiation attenuation coefficient, and x is the measurable material thickness for a given pixel of the radiation absorbing device used. The μ values are material properties that depend on the type of radiation used for measurement and the material of the object under test, and are known from literature and publications. The technical problem is that the material thickness of tubes made of e.g. iron / steel (Fe) or its alloys - above a nominal tube diameter of about 250-300 mm and a wall thickness of about 15 mm - can only be measured with a high measurement uncertainty or not at all by means of radiation detectors that are also suitable for imaging. With smaller wall thicknesses, the nominal tube diameter may be larger; with higher irradiance, the material thickness of the object to be examined may be increased at the expense of image contrast.

[0005] The closest prior art to the above substance testing method is the international patent application PCT / HU2017 / 050030, "Method to radiographically determine geometrical parameters and / or substance state of an object under study", which is granted to the USA, Europe, Russia, China, India, South Korea and Brazil; which patent is industrially applicable but discusses a complex and complicated solution. According to the considered patent application, the material thickness of the object under investigation is calculated pixel by pixel, i.e. per pixel. The intersection curves corresponding to the material thicknesses thus calculated may, where appropriate, be used to construct the cross-sectional drawing of the object, be it a specially filled or unfilled pipe. In the calculation described here, the dose absorbed by the radiation detecting device - in the case of X-ray film, specifically the blackening of the film - is measured, from which the material thickness can be calculated and from which the said intersection curves can be derived.

[0006] Going further, "pixel" based material thickness measurements / calculations require the use of the aforementioned relation ln(l / I0) = -μx; this requires - in principle - two images: one with the object under test and one without the object under test (under unchanged exposure conditions). From the first image the value of I can be determined, the second one gives the value of I0. However, the determination of the I0value is problematic, especially for large material thicknesses, because at this thickness the l0measurement range becomes saturated for a given radiation absorbing device, and the I0values obtained by measurement become inaccurate. More specifically, the desired relation to determine the material thickness should be fulfilled over the whole operating range of the radiation absorbing device used to perform the measurement, specifically for e.g. Ir192 for Fe with a thickness of ~100 mm. However, currently known radiation absorbing devices can measure up to approximately 30-40 mm thickness. In addition, scattered radiation from the object under test and from the ambient field can significantly degrade the signal-to-noise ratio and distort the measurement signals. Experience has shown that radiation interceptors, once saturated beyond the measurable range when measuring I0, no longer give accurate measurements and, to the best of our knowledge, the results obtained in this case are not interpretable at all.

[0007] These - so far - unsolved technical problems have inspired experts to measure, for example, the wall thickness as a projection of the wall thickness - the wall thickness profile - (the "shadow" of the pipe wall according to the above formula) in pipe tests. For pipes, this is still the most common method of measuring wall thickness - which is relatively simple to apply - but it does not increase the range of measurement. (A typical and elegant example is the "Varex Imagine Corp." Patent US2022 / 0099599 A1 and the 2008 Development of protocols for corrosion and deposit evaluation in large diameter pipes by radiography guideline from Vienna, Austria). Furthermore, Misty I. Halt et al.'s "Radiographic modelling for NDE of subsea pipelines" [NDT&E International 86 (2017) pages 113-122] describes a simulation model for radiographic inspection of subsea pipelines. As a test object, an empty pipe of stepped design with an outer diameter of up to 322.6 mm and a wall thickness of 25.0 mm is used, using a flat-panel detector with a surface area of 40x40 cm. The parameters of the simulation model for the pipe under consideration are determined in a laboratory tank with side walls and are thus significantly affected by the scattering of the measurement radiation.

[0008] Marko Rakvin et al's "Evaluation of pipe wall thickness based on contrast measurement using computed radiography (CR)" [Procedia Engineering 69 (2014) pages 1216-1224] discusses the use of phosphor imaging panels in the context of the computed radiography (CR) material testing method with pixel-based intensity measurements. The authors have investigated in detail the saturation of the imaging panel using X-rays with energies of 175, 200 and 250 keV. By optimising the radiographic radiation, they tried to avoid saturation of the panel and thus inaccuracy of the resulting measurements.

[0009] "The use of radiography for thickness measurement and corrosion monitoring pipes" by K. Edalati et al [Int. Journ. of Pressure Vessels and Piping 83 (2016) pages 736-741] discusses the radiographic inspection of empty pipes with an outside diameter of up to 150 mm and a wall thickness of 15.0 mm using X-ray film, with particular reference to the dependence of the distance between the radiation source and the film relative to the position of the different points in the cross-section of the pipe under inspection.

[0010] According to the documents considered to define the state of the art, the above-mentioned problems are known, but the documents discussed here do not provide a substantive solution. The root cause and source of the problem is that the formula is in itself an idealised condition - which does not exist in reality - because the primary beam passing through the object under examination is not only attenuated but also scattered; this scatter is subtracted from and / or added to the measured intensity signal; i.e. the measured intensity is false. Furthermore, it gives the impression that, up to the limit of the measurable intensity, the cross-sectional material projection of the object under test can be described without distortion by attenuation of the primary radiation passing through the object under test; ignoring the well-known fact that, in the case of pipe tests, the charge in the pipe precludes any measurability and testability. In accordance with the foregoing, the present invention aims to develop a combined radiographic inspection method that, together with the determination of diagnostic status, aims to determine the thickness of the objects under inspection in such a way that these material thicknesses are approximately equal to or greater than the material thicknesses of known and practically used (radiographic) diagnostic tests; and, in the case of pipe tests, are charge neutral.

[0011] Furthermore, by using the process according to the invention, radiation absorbing devices commonly used in industry, in particular flat-panels and films, can be used above the blackening limit ("saturation") for I0. It should be noted that the value of I must remain measurable. The fact that a general solution for the application of the relation ln(l / l0) = -μx has not yet been found makes it necessary to investigate what the radiation absorbing device used to measure the thickness of the material, as a detector, measures in the first place. In this respect, film-type detector devices based on AgBr2films (not "phosphorus" based films) are used as an example, while for electron charge-coupled detectors, only those based on gadolinium oxysulphide (GOS) are used as an example. We have found that their measurement results differ from each other and therefore the detector devices used for the measurements need to be recalibrated.

[0012] In Appendix A, the determination of the half value thickness (HVT) of the Ir192 isotope source for Agfa D5(Pb) X-ray films for Fe and water is discussed in detail as an example. Using this example and generalising it to other isotopes or other types of detectors, the skilled person will obtain similar results. In addition, modifications of ln(l / l0) = -μx as the basic formula are discussed in detail, which, as will be seen later, are advantageous for the determination of material thicknesses according to the invention (e.g. flat plates and pipe wall thicknesses).

[0013] In "Appendix B" we raise the idea that the HVT value for the iron for Ir192 source differs from the theoretically expected HVT value for AgBr2based X-ray films. Going further, the HVT data for unalloyed and alloyed steels may differ.

[0014] In "Appendix C", we discuss, in particular, the validation of GOS-based flat-panels (FP) and the conclusions that can be drawn from the validation measurement data. Two important issues are highlighted, one of which is less frequent but not less serious; the electronic noise generated in the FP detector itself, which appears in the measurement signal; and how this noise in the measurement signal can be extracted from the measured signal. It is noted that the manufacturer's protocol includes a noise removal procedure, but this is not always applicable.

[0015] The other is a common problem phenomenon: the Bragg scattering in the object under study, as a disturbance that fundamentally affects the formula mentioned above. This is the scope of the filtering device according to the present invention, mainly for diagnostic investigations, which is a metallic, multi-layered bendable object (device) with a piano-parallel arrangement and / or a flat-faced Bucky (Lisholm) grating adapted for industrial applications to reduce scattered radiation and increase contrast ratios. APPENDIX A

[0016] As a preface, it is worth mentioning that, in general, the international literature does not include HVT data attributable to isotopes, because the film companies provide exposure diagrams or logarithmic measuring devices. At the present state of the art, exposure data are already available oonn so-called "smart" phones, e.g. https: / / apps.apple.com / us / app / ge-film-exposure-calculator / id877526566 .

[0017] Examples of isotope radiation sources are Co60, Se75 and Ir192, etc., and we also use X-ray and gamma irradiators (machines). The Agfa D5 (Pb) X-ray film and the irradiation source Ir192 were selected as a reference, with a distance of 500 mm between the film and the irradiation source (B&W film / irradiation source), with an activity of 740.0 GBq (20.0 Ci) for the Ir192 irradiation source, and with a blackening of Dens(ity) 2.5 for the exposure chosen, where Dens is given as Iog10 (the test is performed for Fe).

[0018] The HVT is the thickness at which the corresponding exposure time (Exp_T) increases quadratically with the linear increase of the HVT. (By means of machine calculations, the Exp_T time can be given to the nearest second, and the material thickness for Fe can be given to the nearest tenth of a mm up to 99.9 mm.)

[0019] For the Ir192 source - based on the data in the link above - the following exposure time data in units of seconds (s) are obtained for Dens 2.5 - i.e.: Iog10 = 316.228 (in which case the blackening of each n * HVT material thickness is Dens 2.5); and then we assign the attenuation ratios 2n=2,4, 8 256 to n * HVTs; and the I0and l0Dens values that can be assigned to these data:

[0020] Table 1 Based on the above, the HVT values for Fe with different radiation sources are Co60: 2.21 cm; Se75: 1.04 cm; Ir192: 1.46 cm.

[0021] In reality, we only expose to one material thickness (let's say 1*HVT in the example: 1.46 cm), then the Dens values of the so-called n*HVT thicknesses (where n = 1, 2, ..., 8) and the corresponding log 10 I intensities are as follows:

[0022] Table 2

[0023] As the doses are halved in this process, I0is taken to be two times 1*HVT I, i.e. : 632.455. Therefore I0= Dens 2,801 ; log 10: 632,455; and the material thickness of Fe is 0,00 cm (I Dens is given as log 10).

[0024] Of course, you can select other data; should be Exp_T (Dens 2.5) 3*HVT - marked as → in the table -, here results are I0= Dens 3,403; log 10: 2529,822, Fe = 4,38 cm.

[0025] Table 3

[0026] Since these data are already known, we can use the internationally accepted and known basic equation: ln(l / I0)=-μx; where l=I0*e-μx; I0= I / e-μx; x=(lnl-lnI0) / -μ . (1) where μ=ln2 / HVT; and HVT=ln2 / μ; ln2=0, 693147; μ(Fe)=0, 4747583; x=MX material measure (thickness) for decimator thickness: ln10 / μ; for thickness N: InN / μ.

[0027] Given that in industrial practice (pipe) dimensions are usually given in mm, cm and inches (also), we use HVT instead of μ, additionally - in our opinion - the form I0 / l is preferable to l / I0in equation (1) (as this refers to the attenuation ratio and in human diagnostics - in case of filters, grids - is the so-called exposure factor), then:

[0028] Based on our reference data for any FF distance and any Ir192 activity - in case of Agfa D5(Pb) film - the Exp_T can be written in second units in the following form:

[0029] Exp_T D5(Pb) = [(FFmm / 500)2 / (Ci / 20)]*2(MX / HVT)*171 , 59375 (7) where 171 ,59375 = 1*HVT Exp_T / 2 = I0Exp_T From the above:

[0030] If the production (real) blackening curve of the relevant film type is known - i.e. it can be stored in a database, all our calculation data will correspond to this data and will be of the expected accuracy. Importantly, for any given Exp_T time, an MX material thickness corresponding to Dens 2.5 may be assigned - irrespective of the material thickness of the sample under test - to which I0can be assigned, because I0can only and exclusively be assigned to Dens 2,5. It is evident that, under these circumstances, no reference etalon image is required and that I0 can be a virtual value over the range of measurability (see later and in Table 1). Note: in the case of film, Exp_T (7) is strictly linked to the value of Dens, because this is the only way to comply with the conditions of computability. For film, the Exp_T time — as a possibility of error — may occur in manually controlled isotope and short Exp_T time exposures, while for FP this problem does not apply. For FP, due to the quasi-straight characteristic of its blackening curve, Exp_T may be a variable: it is not linked to an expected Dens value; but it can - and should - be predetermined according to (4) - (6); see later.

[0031] APPENDIX B

[0032] We must highlight two problems.

[0033] (1) In case of AgBr2-based films, the HVT of the Ir192 source with respect to Fe is 1.46 cm - corresponding to an energy of nearly 1.0 MeV - whereas the imaging spectrum of the Ir192 source (0.206; 0.316; 0.468; 0.610 MeV) has a transmission energy of about half of this, ≈0.5 MeV, or slightly less; at which point a max. HVT ≈ 1.06 cm is expected. The discrepancy may be explained by the energy dependence of the AgBr2molecules - and other additives in the film. (This phenomenon is also encountered with other detectors, see later.) With full confidence, an impact counter could measure real data, but this is not feasible with a film / detector radiation capture device. We assume that the HVT value of the Ir192 radiation source for water, according to the already cited patent application based on the current state of the art is ≈ 9.82 cm at 1.0 MeV. [We should note that the HVT value of the Ir192 source for air is ≈ 40 m (!) at normal atmospheric pressure and temperature.]

[0034] Based on the above, we note that for films, the energy dependence of AgBr2on X-ray / gamma and electron emission (capture) is widely known and published; while the GOS detector and its total electron emission, as a characteristic factor (Kf) of the GOS detector, is less well known and published. Therefore, it is reasonable to assume that this characteristic factor transforms the intensity of I0and I of the X- ray / gamma (plus electron) radiation absorbed by the GOS detector that is generated in the GOS detector itself, and that this transformation also influences the HVT data. As the characteristic factor is also present in AgBr2, the imaging spectra of Co60, Se75 and Ir192 isotopes - and their intensities, we present the theoretical physical "official" Fe HVT values calculated for the transfer energies of these isotopes and for AgBr2(where Kf is the characteristic factor as the ratio of AgBr2HVT / "HVT".)

[0035] Table 5

[0036] Note: Table 5 is evidence that AgBr2deviates from the guiding HVT values of the theoretical physics tables known so far. Therefore, it is justified to validate the corresponding radiation source and the corresponding detector according to the invention. This evidence includes the so-called " Contrast Unsharpness" (Uf mm), which is specific to the energy of the radiation source (Agfa NDT):

[0037] Co60: Uf0,35 mm ≈ 1 ,25 MeV, HVT ≈ 1 ,65 cm. Se75: no data.

[0038] Ir192: Uf0,17 mm ≈ 0,44 MeV, HVT ≈ 0,95 cm.

[0039] The Ir192 data correlate with the aforementioned ≈ 0.5 MeV transition energy and the HVT≈ 1.06 cm that can be associated with it.

[0040] The different characteristics of AgBr2and GOS detectors are given by e.g. the (6) - GOS - and according to the state of the art the (4) - AgBr2formulas, which are given by Eqs: where A: radioactivity (Bq); Kγ: dosage constant; t: exposure time; DF: dose; I: distance (pixel focus); for units of measurement, see therein. The right-hand side of the equation is converted to Ci and Kγ is substituted. (Then the real value of Kγ according to R*m2 / ci*h is: Co60: 1,293;

[0041] Se75: 0,194; Ir192: 0,465.) where

[0042] Of course, the GOS and AgBr2detectors in these formulas are not bound to the corresponding formula; where FF: GOSγ at 1000 mm ( in a validated measurement) is considered constant.

[0043] From this, the equation (5) is true for GOS I0:

[0044] By knowing the real value of GOSγ (ballast, filter, collimator, etc.), the expected value of I0can be calculated; this is necessary to pre-plan a measurement.

[0045] (2) In practical life, iron - i.e. non-alloy steels, commonly known as carbon steels - is rarely encountered, but rather Fe alloys are typical. Strictly considering Fe alloys, the variation of their HVT can be predicted by multiplying the so-called Zeff effective atomic number and the density of the Fe alloys.

[0046] Te Zeffeffective atomic number can be calculated with the formula Where where kn is the % share of the nthcomponent and Zn is the atomic number of the nthcomponent. This means that for Fe, based on textbook data (i.e. Z=26 and density Q=7.87 g / cm3), Zeff * Q = 204.62 g / cm3.

[0047] However, for example, for an acid-resistant steel - AiSi 316 - with a composition (on a per unit volume basis) of Fe - 69%, Cr - 18%, Ni - 10% and Mo - 3%, and where Zeff = 27.203 and Q = 7.96 g / cm3, or their product is 216.54, then the HVT of the alloy decreases in the ratio of 204.62 / 2016.54; i.e: 14.60mm changes to 13.79mm.

[0048] APPENDIX C

[0049] To discuss the validation of the GOS-based DXR25U-W type flat-panel (FP), as described in the state of the art patent application, and the conclusions that can be drawn from them: our device has a 16 x 16 inch detector surface, in front of which a filter with a surface density of = 0.5 g / cm2 of Pb, Cu, Al - where Zeff; 33.8; Q: 7.0 g / cm3 - is placed as a front layer, and 1.0 mm PE and 2.0 mm Plexiglas are placed as a front layer to avoid external loads. 0.5 mm Pb has been fixed to the back of the device using PE cladding. An additional 2.0 mm Pb backplate was used during laboratory measurements. The scattered radiation from the Ir192 source was shielded with a 90-degree Cone W collimator, but was also tested without a collimator. (The data shown are always meant with a collimator, because this is preferred during in-service tests for radiation protection and measurement reasons; and the filter can be modified, see later).

[0050] Note: the feed (here: bild up) is of particular importance in establishing the so- called electron balance - and since this process has not been investigated - we will only mention that it is implemented by (e.g.) plexiglass sheets placed directly in front of the detector depending on the X-ray / gamma energy. Since we do not know the FP manufacturing solution (obviously there is a ballast of some thickness in front of the GOS detector), we have not used a separate bild-up ballast. It is important that the FP is always validated according to an operable variant; regardless of whether the bild up ballast is applied or not. The ballast we used is effectively a filter, partly to reduce scattered radiation and partly to increase the contrast enhancement ratio. For information the plexiglass thicknesses are: Co60:4, 0-5,0 mm; Se75:1, 5-2,0 mm; lr192:2,0-2,5 mm

[0051] (For dual photon accelerators: At 6 MV ≈15 mm, at 18 MV ≈30 mm).

[0052] Furthermore, we have defined a square of approximately 5.0 mm x 5.0 mm on the surface of the FP, which is selected for reference measurements; it is also the point of projection of the line of the FF perpendicular to the FP of the source. Let the value of FF be the minimum at the reference point, in which case I0is at its maximum; whereas for all other coordinates assignable to MX points, I0is less than this value. The fact that FF is not perpendicular to the incidence - and not at the minimum value - does not change the fact of the reference point; in this case, however, the maximum of I0may be at another pixel point.

[0053] The value of any I0shall always be calculated, in relation to the FF reference point, for the coordinate of the corresponding MX.

[0054] With I0measurements, we can conclude on the scattered radiation from the ambient field by representing the I0intensity passing through the reference point by a line parallel to the x-axis. In this case, FF=1 unit at the reference point; while the intensity of the other points on the line varies according to the law of distance. If scattered radiation is encountered, the I0intensity curve is correspondingly distorted in an increasing direction. According to our current knowledge, the I0itself is determined with the filter, including the applied ballast. The validation (calibration) of I0is not subject to a value without upper limitation, because then the value of I0must remain within the measurability range. With I0measurement, the perpendicularity of the incidence angle can be determined by marking measurement points on the x, y axes passing through the reference point, at the same distance from the reference point (e.g. 150 mm), where the intensity of each measurement point is approximately equal (± x ≈ ± y) This procedure can also be used to check the noiselessness of the FP (see later).

[0055] To reduce measurement errors, the size of the FF was set at 1000 mm (39.37 inches). The instrument blackening was set to increasing blackening in proportion to the increase in dose; because this has an inverse adjustability. The device can - in principle - resolve the blackening scale to 214= 16384 parts; but of this - in practice - about 14170 can be displayed (Dens 4.15; the FS50B film digitizer can measure up to Dens 4.7).

[0056] The recordings were set to three repetitions until N noise was detected. Exp_T can be set from 0.2 sec to -150 sec with a resolution of 0.1 sec.

[0057] It is worth noting that the FP manufacturer's GE program requires length / thickness data in inches (") and activity in Ci. If the isotope half-life of Ir192 (73.83 days) is given in 1771.92 hours, the expected I0 intensity at the selected reference point of the FP can be determined at any time.

[0058] To begin with, we now look at the verification of the noise-loaded FP device mentioned in the introduction. The validation essentially requires the "empty" detector surface, including the ballast already installed, for the I0 determination, while the ballast filter must be removed for the calibration according to the manufacturer's protocol. We have found that the N noise present in the S intensity signal is distinguishable from each other regardless of the manufacturer's calibration protocol procedure. We demonstrate two specific measurements where, first, the operating temperature of the FP device is partially stabilized; i.e., the device generates heat during operation and thus a relatively slow temperature increase occurs. The so-called blank screen was then irradiated, with exposure times increased every second, until we reached an S0intensity of about 14000. For each measurement, without repetition, the measurement procedure was restarted, which took about 40 sec + Exp_T. Plotting the measured and calculated data into a coordinate system, the time Exp_T sec was plotted on the x-axis, while the measured and calculated intensities S,N,I were plotted on the y-axis. (The S0intensity could be measured for 30 sec.) It was found that this SO intensity can be described by the so-called "linear regression" and Exp_T can be calculated up to 150 sec.

[0059] Obviously, the measured signal (mS intensity) cannot be a value higher than 14170; however, S0, I0can be determined up to 150 sec without upper limit (I).

[0060] Following the same procedure, but without the use of a radiation source (for 30 sec.), we measured the noise N, where we also found that, as long as the FP temperature does not vary significantly, it can be determined by "linear regression". Therefore, S0-N = l0, where the "linear regression" of I0is also obvious (linear regression: lin.reg.).

[0061] Subsequently, Exp_T time was defined as a series of numbers whose members are doubles of each other, since 10-12 measurement points are sufficient. Thus, we have selected the time intervals 2, 4, 8, 16, 32, 64, 128; 5, 10, 20, 40, 80; 9, 18, 36, 72, 144; 15, 30, 60, 120; and 25, 50, 100 sec (Exp_T) x times; as well as 1 and 150 seconds.

[0062] In an ideal case, only the N noise-free Soy ≈ I0y would be measurable, but as we will see, there is always a residual noise - component B; therefore, it is not recommended to bypass this verification procedure. We tabulate the measured values of S0y over a time interval of 2 - 30 sec, its linear regression between 1 - 150 sec; its linear regression Ny between 1 - 150 sec; and its linear regression I0y between 1 - 150 sec;

[0063] In this case the Ir192 radiation source activity is 34.324 Ci; FF: 1000 mm.

[0064] (S, N, I0) y = Component A * x + Component B

[0065] Sy (±0,3858 %) = 404,3662 * x + 1695,7908

[0066] Ny (±0,0130 %) = 17,1850 * x + 1632,2257 I0y (±0,7312 %) = 387,1797 * x + 63,5838 | .3O '™'|"’

[0067] Table 6a

[0068] If (Exp_T) x = y - B / A; and mS is the intensity measured at the object under study, then based on mSy intensity that can be assigned to (Exp_T) x: mSy (Exp_T) x = ly (Exp_T) x I0 / I = lin.reg. I0y / ly = I0y / A * mSyx + B = (6) = (11)

[0069] Subsequently, calibration was performed according to the manufacturer's recommendation - using an X-ray irradiator - i.e. "zeroing" N noise; and it was found that at Exp_T 30, 60 and 120 sec, no N noise was measurable. Going back to our Ir192 isotope studies, N noise was slowly "coming back" - and because it was still building up - it could not be determined by linear regression. S0y (as a gross signal) was of course measurable in the time interval 2 - 32 sec and could be determined by linear regression between 1 and 150 sec; where the activity of the radiation source was 29.01 Ci.

[0070] (S) y Component A * x + Component B S0y (±0,2767 %) = 409,4507 * x + 29,0441

[0071] Table 6b

[0072] It is also evident that - for a known activity - the irradiation dose DF can be entered i.e. in this case it is possible to determine the I0associated with any (Ir192) activity isotope - by linear regression; simplified according to (9).

[0073] With this preliminary work, specific Fe samples were allocated to validate the thickness of the material. These can effectively be considered as Fe etalons (nonalloy carbon steels). We had - as we considered the HVT to be 14.6 mm - 6 flat etalons of 14.60 mm and 2 flat etalons of 20.00 mm, as well as 1 " staggered " etalon.

[0074] The 14.60 mm etalons were approximately 140 x 140 mm flats; the 20.00 mm etalons were 100 x 240 mm. The staggered etalon was also (100 x 240) x 1-20 mm, where the step level difference is 1 mm and the step width is 12 mm. The thickness of the etalons were measured using a (digital) caliper and an ultrasonic wall thickness gauge.

[0075] We found that the individual error in the thickness of all the etalons is less than ±0.05 mm. If we disregard the N noise discussed above, we obtained measurement signals up to 127.6 mm in total, excluding the stepped etalon, but we did not "validate" measurements above 87.6 mm because an intense Bragg scattering (disturbance) was started from the sidewalls of the flat etalons, which would have distorted the measurement, i.e., the intensity increased by the scattering signals would then be displayed. For the large objects we examined - by increasing the flat etalons to about 300 x 300 mm - the signal can still be measured at a material thickness of MX 140-160 mm; where the accuracy of the measurement is about ±4.0% of the material thickness MX (for film it is about ±2.0%). Then the intensity of the outgoing signal is about 1.0% of the intensity of the incoming l0primary radiation; i.e. » ln100 / μ, where for D5 (Pb) film the MX material thickness is 97.0 mm; while for FP (for 19.1 mm HVT, see later) it is 126.2 mm. The measuring - reference - point of the etalons should be at least 2-4 MX distance from the side walls of the etalons. If the reference point of the I measurements is placed at an appropriately selected point on these flat etalons, signals due to Bragg scattering - of wall origin - at this point can no longer be measured (see Figures 1-7). In our opinion, we require 16 pieces of 10.0 mm flat etalons and one piece of the example "staggered" etalon for validation.

[0076] By investigating the possibility of narrowing the beam of the outgoing primary radiation (the collimator is designed for this purpose too) or masking it, the Bragg scattering problem could be improved, a concept known mainly for human diagnostic irradiators. To investigate this, we have made two pieces of approx. (140 x 140) x 6 mm lead plates, with a "window" of about 50 x 50 mm in the middle. These tools essentially form a frame to be placed in front of the Fe etalons to be measured, on the side facing the radiation source.

[0077] We have also created two more lead sheets of approx. (140 x 70) x 2 mm - which are not really frames - but we can narrow the "window" to approx. 50 x 10 mm. This is marked with a + sign. Measurements have confirmed that the frame does not only eliminate the scattering from the side walls of the etalons, but also the reduces the natural scattering that occurs in the objects in the covered area; see Table 7.

[0078] The objects under investigation are flat panels, where the aforementioned Pb frames were used for the flat panels, with 0, 1, 2, + labels as appropriate.

[0079] Using the data from Table 6.b: Ir192 activity 29,01 Ci FF 1000 mm, Exp_Tx = 32 sec., and S0y = 13116.

[0080] Table 7

[0081] Considering these problems, it is obvious that, for example, the hole depth of drilled etalons appears to be deeper due to Bragg scattering of wall origin. In case of perpendicular sidewalls, only and exclusively extensive flat plates can be authenticating etalons. As a consequence, for example, the "Step / hole Type IQI" stepped drilled etalon is suitable for image quality validation up to a few mm thickness; whereas the drilled plate "ASTM plaque type penetrometer" is fully adequate.

[0082] Some of these certification flat plate etalons provide a material thickness MX relative to the flat plate object under investigation, where the HVT may be influenced by the Zeffeffective index; the guideline HVT is: 19,1 ±1,0 mm.

[0083] Continuing with the concept of the noise-free validation of flat plates, the Fe HVT can be derived up to about 7*HVT, if e.g. I0 intensity is calibrated below 28340 (i.e. the first measured value is then less than 14170) and one half-scale FFee plate per measurement is applied; then we can produce e.g. 28160→ 14080→ 7040→ 3520→ 1760→ 880→ 440→ 220; or 25600→ 12800→ 6400→ 3200→ 1600→ 800→ 400→ 200 absorption values (it is not recommended to go below 200, because blackening is no longer linear). It is also worthwhile to produce decimation thicknesses e.g. 25600→ 2560→ 256. With further increasing I0, the absorbance values can be detected up to the final limit of measurability.

[0084] When measuring liquids, it is preferable to have the FP in a horizontal position, where the vessel should have a surface area of at least 400x400 mm and should not stress the FP. If the FF is 1000 mm, then the height and thickness of the liquid, increasing by increments of 50 mm, can be measured up to about 800 mm. To the best of our present knowledge, we believe that for both D5 (Pb) X-ray film and FP, the MX_Fe material thickness that can be measured with certainty is 7*HVT. For D5 (Pb) X-ray film, this is 102.2 mm; and this is consistent with international and manufacturer guidelines. This can be increased up to 8*HVT with product development (increased Dens). (For comparison: Co60 ≈ 7.4 inch, 8.5 HVT; Se75 ≈ 1.6 inch, 3.9 HVT; Ir192 ≈ 4.0 inch, 6.9 HVT are the upper limits for radiographic - imaging - studies).

[0085] Note: Tables 6. a and 6.b demonstrate that S0-N=I0is a feasible reality, where S0, I0may exceed the measurability range and I can be calculated from the known data of Sy; respectively, Sy is suitable for directional measurement at low N noise intensities. Whereas Table 7 demonstrates that at FP (without frame) the HVT is already above the limit of ~Fe14.6 mm, which is - among other reasons - why the pixel measurements so far were not feasible, and the GOS detector does not measure the HVT data as a constant value; given that different exposure times have different geometry-dependent (Bragg) scattering values plus different N noise values, the HVT data that can be associated with them will fluctuate.

[0086] (In our example N noise is still negligible).

[0087] It is essential to significantly reduce Bragg scattering for measurements. The question is what is considered to be a sufficient reduction and how this is to be achieved.

[0088] In industrial and laboratory practice, e.g. for pipe testing, the largest MX material section associated with the pipe wall - and the pipe wall arc itself - is a Bragg reflecting surface; and certain alloys and crystal structures increase the intensity of Bragg scattering significantly; where interferences in the Bragg scattering bandwidth are - in practice - not computable, but the Bragg scattering fading path length can be determined. (This path length should be greater than the material thickness ≈ In MX already mentioned.) In fact, the sum of the radiation intensity is the sum of the primary radiation, Rayleigh, photo, Compton scattering, and scattering corresponding to the pair creation above 1 MeV energy, and Bragg scattering, as well as backscattering from the environment. Scattering interference phenomena can also occur in the AgBr2 / GOS film / detector material itself, which is omitted in the present case, i.e. it is fulfilled that lΣ= In = I + Rayleighσ + fotoσ + Comptonσ + Braggσ + environmentσ (+ paircreationσ).

[0089] In our opinion, a significant part of these scatterings, excluding environmental backscattering, is carried by the Bragg scattering; which we do not consider to be meaningfully predictable: lΣ / (Braggσ >1 ,0) = I

[0090] These scatters and reflections distort the intensity curves of the tube (cross- sectional) that can be generated and displayed based on the computer-evaluated data of the FP; i.e. the increased intensity shows this tube section as thinner; because, as mentioned above, the Bragg scattering inside the tube is increased, while the tube arc scatters the incoming Bragg scattering partly outwards (and partly inwards), so that the primary radiation intensity is reduced, i.e. the tube arc appears thicker. This is essentially an X-ray optical curve phenomenon.

[0091] To explore and resolve this question, laboratory and in-service tests were carried out using an Ir192 source with the pipelines tested in service and with an M load ( steam pressure: 46 bar, operating temperature: 270°C). The table below shows s as the wall thickness; while Femax is the largest material section running in the pipe wall.

[0092]

[0093] Table 8

[0094] Based on the historical experience, the conditions and the technical data, we have concluded the following: the problem is essentially caused by a dramatic deterioration in contrast enhancement ratio and selectivity. Nevertheless, pipes and pipelines can preferably be defined as the sum of wall thickness 2s at the nominal diameter, regardless of diameter, wall thickness and M load (see later). The current filter of the FP device is adequate and sufficient for this operational procedure, but possible improvement options will be discussed in more detail.

[0095] According to the present state of the science, the wall thickness s can only be measured - by means of projection mapping - up to the currently known and commonly used extent of about Femax 100 mm; because above this value the Bragg scattering that occurs will distort the contrast, and the typical minimum intensity is not exact. During our investigations, the measured intensity curve of the wall thickness s could not be determined by a uniform mathematical modelling, because due to the scattering effects caused by the pipe wall, a characteristic curve could be fitted to a smaller or larger section of the intensity curve; e.g. the running intensity curve shows continuously varying - partly increasing - HVT data (see later). This latter fact is worrying in itself. Now turning to the device under the invention: it is questionable whether it is a viable way to filter out Bragg's scattering in the investigation process? We have concluded that our current diagnostic filter is no longer suitable for this purpose. A possible solution would be to supplement an optimised filter with the so-called Bucky grid or grids, i.e. this solution would allow to increase the desired contrast enhancement ratio and selectivity. Furthermore, it is questionable whether the measured intensity curve of the wall thickness s then correlates with the actual (mechanical) intersection curve; i.e. can it be mathematically simulated, modelled and calculated? Furthermore, the measurement uncertainty is increased if scattered radiation from the surrounding environment falls on the film / detector, because then the measurement result is distorted, especially in the case of FP, therefore care must be taken to reduce the scattered radiation too; and to always use the same type of collimator, because the collimator affects the intensity of the radiation falling on the film / detector.

[0096] For the object under inspection, collimated radiation is recommended primarily to reduce Bragg scattering; additionally, a pre-filter applied to the FP is recommended. In principle, it would be possible and allowable to use the so-called Bucky and / or Lisholm grids, which have the disadvantage of being manufactured for a fixed FF distance, making their industrial applicability challenging; furthermore, they operate in the energy range around 0.1 MeV and are consequently quite transparent for Ir192, i.e. the device itself is thin.

[0097] The filter and the Bucky grid could be used both separately and together. The Bragg scattering itself, as a perturbation state called In(interference), is present both in the object under inspection and in the film detector itself. This perturbation generates electron transport and, in some cases, secondary X-ray / gamma radiation; as a consequence, according to our present knowledge, the perturbation In cannot be extracted from the measured mSy (gross) signal: (mSy-N) + In = ly.

[0098] Here we can intervene to solve the problem in two ways: a) first and foremost, in addition to the collimator, a diagnostic filter is used. The present filter device is far from optimal, but in general, at the lower limit, it is sufficient and adequate for diagnostic tests and is also suitable for measuring the double wall thickness in the actual tube diameter, regardless of the M load in the tube. This filter can be permanently and stably mounted on the FP as a front plate. It should be noted that for a simplified filter, which already improves the image quality, we recommend a Fe plate of about 0.5 mm, potentially between two Pb plates of about 27.0 μm (at this thickness, rolled Pb film is already approximately gold-priced, but 27 μm thickness, as a reinforcement for Agfa films, is mass-produced). There is no advantage in further increasing the thickness of either Pb or Fe, because image quality does not improve, but rather deteriorates. The typical Pb filter recommended by the scientific literature - in the X-ray range - is 0.25-0.50 mm up to 100-200 kV; 0.6-1.0 mm up to 400 kV; we accept this but do not recommend it; nor do we recommend Cu filters.

[0099] The purpose of the filter is to attenuate scattered radiation away from the direction of the primary radiation to a greater extent than attenuation of the primary radiation, which improves the contrast enhancement ratio and selectivity; as well as significantly improving the visibility of the IQI wire-frame image quality indicator.

[0100] For this purpose, rolled metal foils with FCC and BCC crystallite lattice structures (preferably in thicknesses of 10-50 μm) are suitable; with particular regard to the fact that they are more advantageous from an X-ray optical point of view than the same thickness (metal) steamed and / or vaporized process. In simplified terms, this is due to the microsponge structure of the steamed and / or vaporized structure, which increases the scattering centres significantly and without advantage. The crystal structure of the rolled films is positively modified by the rolling; and increasing the number of layers in the films increases the X-ray optical capabilities.

[0101] Furthermore, a possible filter is a multilayer device consisting of rolled films; where the preferred material components are Pb, W, Ta, Mo, Nb, (Cu), Ni, Fe, Cr, V, Al. The devices according to the invention are energy-optimized, so that in particular the filter types are optimized according to μm thickness, surface density mg / cm2, Zeff's index, and density g / cm3. It is also reasonable to implement these filters in the X-ray range for Al, plastics and human diagnostics. For all filter devices, the rolled metal films used are 12.5 μm; except for Al, which is 10 μm and Pb, which is 27 μm. Their operating range: 100-2500 mg / cm2, density (for plastics, heavy metals) 3.2- 10.0 g / cm3; where each filtering device consists of at least two different components and of min 16 pcs max 256 pieces of rolled film.

[0102] The energy range of the filters in this example and the recommended order of their elements:

[0103] Pb-AI-4Ni-AI-4Ni-AI-4Ni-AI4Ni-AI-4Ni-AI-4Ni-2AI-4Ni-AI-4Ni-AI-4Ni-AI-4Ni-AI-Pb.

[0104] 54 pieces; 674,0 μm; 538,69 mg / cm2; Zeff: 55,40; 7,99 g / cm3A1 (0,40-0,8 MeV)

[0105] W-AI-W-AI-6Fe-AI-6Fe-AI-6Fe-AI-Mo-AI-Mo-2AI-Mo-AI-Mo-AI-6Fe-AI-6Fe-AI-6Fe-AI-

[0106] W-AI-W.

[0107] 59 pieces; 702,5 μm; 563,67 mg / cm2; Zeff: 56,82; 8,02 g / cm3A2 (0,40-0,8

[0108] MeV)

[0109] Ta-AI-Ta-AI-4Ni-AI-4Ni-AI-4Ni-AI-4Ni-AI-Mo-AI-Mo-2AI-Ta-2AI-Mo-AI-Mo-AI-4Ni-AI- 4Ni-AI-4Ni-AI-4Ni-AI-Ta-AI-Ta.

[0110] 59 pieces; 692,5 μm; 559,45 mg / cm2; Zeff: 54,70; 8,08 g / cm3A3 (0,40-0,8

[0111] MeV)

[0112] Mo-AI-Mo-AI-Mo-AI-2Ni-Fe-2Ni-Fe-2Ni-Fe-2Ni-Fe-AI-Mo-AI-W-AI-Mo-AI-Fe-2Ni-Fe-

[0113] 2Ni-Fe-2Ni-Fe-2Ni-AI-Mo-AI-Mo-AI-Mo

[0114] 43 pieces; 512,5 μm; 410,02 mg / cm2; Zeff: 44,51; 8,00 g / cm3Bl (0,20-0,4 MeV)

[0115] Mo-AI-Mo-AI-Mo-AI-5Ni-Fe-5Ni-Fe-AI-Mo-AI-Ta-AI-Mo-AI-Fe-5Ni-Fe-5Ni-AI-Mo-AI- Mo-AI-Mo

[0116] 43 pieces; 512,5 μm; 411 ,80 mg / cm2; Zeff: 42,95; 8,03 g / cm3B2 (0,20-0,4

[0117] MeV)

[0118] Fe-AI-Fe-AI-Fe-AI-Fe-AI-Fe-AI-Fe-AI-Fe-AI-Fe-AI-Fe-AI-Fe-AI-Mo-2AI-Mo-16AI-Mo- 2AI-Mo-AI-Fe-AI-Fe-AI-Fe-AI-Fe-AI-Fe-AI-Fe-AI-Fe-AI-Fe-AI-Fe-AI-Fe

[0119] 64 pieces; 700,0 μm; 355,85 mg / cm2; Zeff: 30,75; 5,08 g / cm3C1 (0,10-0,2 MeV) Fe-4AI-Fe-4AI-Fe-4AI-Fe-4AI-Fe-4AI-Fe-4AI-Fe-4AI-Fe-4AI-Fe-4AI-Fe-4AI-Fe-4AI-

[0120] Fe-4AI-Fe-4AI-Fe

[0121] 66 pieces; 695,0 μm; 278,12 mg / cm2; Zeff. 23,13; 4,00 g / cm3C2 (0,10-0,2 MeV)

[0122] Table 10

[0123] Note: measurements were performed on Agfa D5 (Pb) films; with 12, 24, 36 mm expanded flat plates; through 100 and 300 mm water thickness; 235 kV, 4.5 mA X- ray; and Ir192 radiation sources. For industrial and therapeutic accelerators (preferably 6-8 MV; i.e. 2-3 MeV), the appropriate filter types can be multiplied to increase the surface density to the desired size. E.g. for Ir192, Cs137, a single multiplication is needed in the energy range; Co60, a double multiplication in the energy range and a three to fourfold multiplication at 2-3 Mev. Obviously there is a limit where these filters are no longer sufficient. At this point, the option to progress further is to use the so-called Bucky grid(s). b) The Bucky-grid itself is a flat, non-bendable - square or rectangular - flat plate device; where on the centre line of a carrier plate - in a mirror-symmetrical arrangement - heavy metal (e.g. Pb) lamella strips of approximately 1 ,0 mm height are placed parallel to each other as close as possible to each other, but with an intermediate spacing. Furthermore, these lamellae are oriented at a predetermined FF distance, at which FF distance the radiation source can move linearly. This Bucky grid is opened in the direction of the primary radiation; whereas radiation other than the primary radiation (scattered) is absorbed in the lamella plates. Given that the Bucky grid is - as intended - manufactured for an energy range around 0.1 MeV; we would therefore need to multiply this device. Our opinion is that in the energy range 0.4-0.8 MeV, a grid with a thickness of 5-10 mm corresponding to the total thickness of the current 4-5 Bucky grids would be needed. It is advantageous - and we could do so - to add filters to the Bucky grids in arbitrary groupings.

[0124] To explore this problem, we have ordered three 300 x 400 mm Bucky grids from South Korea (JPI Healthcare Co. Ltd.) with a grid line of 230 lamellae / inch; grid frequency (ratio) 12:1 and FF 100 cm. It was observed that parallel stacking of the grids resulted in interference fringes, and that two grids could be stacked without interference by rotating them 90 degrees. We also concluded that these duplicated, perpendicularly stacked new types would need to be about 5-10 mm thick to reduce the intensity of the primary radiation by about half (exposure factor: 2.0). We also concluded that the stacked Fe flat plates were measurable up to 232 mm thickness. This could be extended to a plate thickness of about 10" (about 254 mm) using an Ir192 radiation source, or up to about 15" (about 381 mm) using a Co60 radiation source. Then the wall thickness s for the longest material section in the tube for Ir192 and Co60 sources is:

[0125] At 50" tube diameter (1270,0 mm) s wall thickness = 12,8 mm, and 29,2 mm;

[0126] At 40" tube diameter (1016,0 mm) s wall thickness = 15,6 mm, and 38,0 mm;

[0127] At 32" tube diameter (812,8 mm) s wall thickness = 19,7 mm, and 50,4 mm; At 25" tube diameter (635,0 mm) s wall thickness = 25,6 mm, and 63,5 mm; At 20" tube diameter (508,0 mm) s wall thickness = 32,8 mm, and 86,0 mm; At 16" tube diameter (406,4 mm) s wall thickness = 43,0 mm, illetve 132,5 mm;

[0128] At 12" tube diameter (304,8 mm) s wall thickness = 87,2 mm, and total tube diameter.

[0129] Table 11

[0130] The above-mentioned manufacturing unit is ready to produce a new type of grid - for industrial use - with a total plate thickness of about 7 mm, a grid frequency of 27:1 per plate (i.e. 54:1 for the device) and a gridline of 150 lamellae / inch; respectively, the intersection of the two centerlines on the device marks a fixed FF point Thus, we have a duplex structure (406.4 x 406.4) x 7.0 mm grid, with a grid mass of 5720 g; its density is 4.94 g / cm3; its surface density is 3.46 g / cm2. The width of the grid (Pb) lamellae (coil width) is about 3 mm (i.e. 6 mm). Note: The present human and veterinary grids have a lamella width of about 1 mm. It would be useful to manufacture industrial grids in duplex versions with a lamella width of 1,0 mm, 2,0 mm, 3,0 mm or 4,0 mm.

[0131] There is no doubt that this new type of grid and filter combination can expand and enhance weld inspection and condition determination in an unprecedented way. In our opinion, these tests can be performed without any pipe diameter limitation for gas pipelines, up to 25 inch pipe diameter for steam pipelines, and up to 16 inch pipe diameter for water and oil pipelines, with the qualification of welds. The filter and grid can also increase the indicative wall thickness characteristics of the measured intensity curve of the pipes, i.e., extend the range of measurability; but it does not and cannot give a uniformizable reading, not limited in principle to the range of measurability.

[0132] In principle, the examined tube object is an X-ray optical curve mirror. Here, the scattering of primary radiation is much more complex than in the case of extended flat plates. As is known, the scattered radiation that is directed towards the film / detector can be more or less neutralized by filters and grids; but the scattered radiation that is inherently cross-scattered or back-scattered due to the tube geometry cannot t?e measured and cannot be determined. Thus, the primary radiation on the film / detector is not proportional to the attributable material thickness; ergo, the HVT cannot be constant.

[0133] Therefore, the neat and sympathetic idea that the absorption of the ml intensity measured at the pipe under test would be suitable for measuring the wall thickness of the pipe wall (see extensive flat plates) is essentially untrue.

[0134] To conclude the "Appendix" chapters, we return to the Bragg scattering phenomenon described for the Flat etalons, pointing out the seriousness of this problem. With reference to the attached drawings - which, according to the invention, do not include filters, strainers and / or Bucky gratings, Bucky gratings, where

[0135] - Figure 1 shows a top view of the relative arrangement of a flat-panel detector device and a etalon to be measured, preferably made of iron,

[0136] - Figure 2 shows a side view of the detector device ("flat-panel") and the iron etalon to be measured with the typical beam paths of the radiation source in the A and B ranges,

[0137] - Figure 3 shows the intensity present on the surface of the detector device,

[0138] - Figure 4 shows a top view of the other arrangement of the detector device ("flatpanel") and the etalon to be measured,

[0139] - Figure 5 shows a side view of the detector device ("flat-panel") and the iron etalon to be measured with the typical beam paths of the radiation source as shown in Figure 4,

[0140] - Figure 6 shows the intensity present on the surface of the detector device in the arrangement shown in Figure 4,

[0141] - Figure 7 illustrates the exposure FFmm data that can be attributed to Table 8, and - Figure 8 shows the theoretical physics; and HVTmm data of AgBr2 / GOS detectors by MXFemm material thickness for Ir192 radiation source; where these HVT data are significantly different from each other.

[0142] - Figures 9 a, b, c, d, e, f, g, h (typical diagrams) are part of the invention training to be detaild later.

[0143] - 10. a, b are device (typical) diagrams of the radiation source, test object (pipe) double Bucky or Lisholm grids with multilayer filter and film detector position (the first Bucky grid is in separate top view).

[0144] In the following, we interpret the Ir192 radiation paths with respect to the Fe etalon and the occurring wall originating Bragg scattering in two possible ways in relation to Figures 1-8. The size of the etalons have been increased to 300 x 450 mm in order to have a measurable section in the evaluation in which no wall originating Bragg scattering can be demonstrated to occur. Figures 1-8 are to scale, except for the intensity curves.

[0145] Reference symbols used in the diagrams:

[0146] 1 - FP detector surface; in our case 400 x 400 mm.

[0147] 2 - Fe etalon, object to be examined in general; specifically 10 pieces of stacked

[0148] 300 x 450 mm Fe etalon

[0149] (layering is not marked in the diagrams)

[0150] 3 - γ radiation source (Ir192 isotope) in position 1 (see later)

[0151] 4 - measured intensity (ml, mlsC) curve (not to scale) to be displayed on the display surface of the detector device

[0152] 5 - theoretical physical "HVT" ≈ 10,0 mm

[0153] 6 -HVT of AgBr2detector 14,6 mm

[0154] 7 - HVT of GOS detector averaged linearly over ≈ 20.0 mm, and according to the invention instruction:

[0155] 8 - the pipe object to be inspected

[0156] 9 - mechanical (pipe) intersection curve MIC

[0157] 10 - virtual intensity curve vIsC

[0158] 11 - HVT character curve HVT CnC

[0159] 12 - I0character curve I0CnC

[0160] 13 - duplex Bucky (Lisholm) grid

[0161] 14 - grid (Pb) lamella 15 - grid centerline (centerline) 16 -filter.

[0162] Note: Figure 1 shows the top view of 1 FP and the 2 Fe etalon to be measured, as well as the characteristic beam paths of the 3 γ radiation source, where these beam paths appear in 2 Fe etalon and 1 FP in the A and B regions of the diagram. Figure 3 shows the 4 (I) intensity, which is different in the A and B regions.

[0163] Figures 4, 5, 6 can be equivalent to Figures 1, 2, 3, with the exception that the 2 Fe etalon is rotated in such a way that the intensity 4 (I) in the range A is increased. Interpretation: the source 3 γ is placed above the reference point Q2 (specifically, the geometric center) of FP 1. The FF distance is selected, as shown in Figure 7, in such a way that the Exp_T times required for the same dose increase quadratically as a function of these distances. Figures 2, 5 and 7 show that this FF distance is 353.5 mm; in this case it is 1 unit of Exp_T. Further increasing the FF distance by 500.0 mm multiplies by 2, by 707.1 mm multiplies by 4. and by 1000.0 mm multiplies by 8 the Exp_T1 time required to achieve the same dose (blackening). As shown in Figures 2, 5, the Exp_T1 time was chosen. An example overview table is given for three possible Exp_T time variants, but other times - and FF distances - are also possible.

[0164] Table 12

[0165] The detector surface of FP 1 is divided into "A" and "B" regions as shown in Figures 1-6, where in region A we want to visualize the wall originating Bragg scattering by means of the 4 (I) intensity curves; while in region B the wall originating Bragg scattering is not present. Figure 2 shows that the c side of the 2 Fe etalon is " scattering in ", but the scattering path length completely attenuates this. In all cases, the validated HVT reference point is Q2. The reference values for the reference points 1, 2 and 3 are In Figures 1, 2, 3, and 3, the sides a, b, d of the 2 Fe etalon are the scattering surfaces, which is clearly visible in the A field of the intensity curve of 4 (I) in Figure 3 [y1 axis represents the intensity in the case of the 4 (I) intensity curve, and x1 axis represents the measured length of the 2 Fe etalon, where the tested object - the 2 Fe etalon - and the reference point Q2 are in the same x1 , y1 coordinate system].

[0166] One could move the 3 γ radiation source - with unchanged FF distance - closer to the side of the 2 Fe frontal at several positions, in which case the intensity of the wall-originated Bragg scattering will increase. In Figs. 4, 5, 6, the 2 Fe etalon is rotated as in Fig. 4, then the intensity of the wall originating Bragg scattering from sides a, d increases.

[0167] It is crucial that the reference point Q2 remains the point free of wall originating Bragg scattering during verification. As mentioned above, for verification purposes, the 2 Fe etalons are measured e.g. in the range 5-165 mm, and then a table similar to Figure 8 can be constructed showing the HVT data as a function of MX material thickness. During this procedure, the corresponding HVT is determined for each etalon plate applied according to Equation (3). If 1 FP is in a horizontal position, it is necessary to ensure that it is unloaded. If 1 FP is in a vertical position, then the 2 Fe etalons can be supported in a simpler way, but the supports must not be scattering objects. It is worth mentioning here that if the flat etalons, aligned with the Bucky grid, were formatted with side walls set at a fixed FF distance to a truncated cone format, then the wall-originated Bragg scattering would no longer occur.

[0168] From the above observations, it is also evident that the internationally accepted literature (see - prior state of the art patent application - descriptive section "Health Physics Resources - UCRL-501741: Compilation of x-Ray cross Sections". ) - the HVT half-life "constants" attributable to the relevant radiation sources are highly geometry and detector dependent, as well as dependent on the prelays and Bucky gratings applied to the detector; it is therefore reasonable and required to recalibrate the X-ray and / or gamma-ray radiation sources attributable to the relevant detector to the materials of the most important objects under investigation.

[0169] It is fundamentally important to analyse in detail aluminium and its alloys, titanium and its alloys, iron and its alloys, copper and its alloys; and water and various oils, gases and steams, plastics. It is also reasonable and required to know the calculability of the intensity increase due to the scattering of X-ray and / or gamma rays from these materials and the geometric position of these materials; and to exclude the occurrence of these scatterings.

[0170] Moving on to the measurement of material thicknesses, since the objective is mainly to test pipes and pipelines, but other objects may also be tested, it is necessary to know the geometric parameters of the object to be tested; the available film / detector, the available radiation source and the relevant radiation physics characteristics. (The MX, HVT relationship that can be assigned to the film / detector and measured by the film / detector.) In the case of pipes, it is necessary to know the actual diameter of the pipe, the nominal wall thickness, the alloy of the pipe material, as this may modify the HVT data; see relation (10). Not an exclusive requirement, but it is advantageous to know the material, pressure, temperature - and, if available, the external thermal insulation characteristics - of the charge in the pipe. In order to assign to any pixel of the film / detector the appropriate material section of the tube under test - with engineering accuracy - it is necessary to know the relative geometric position of the film / detector - tube - source; and the so-called reference point defined by the angle of incidence, preferably the inclination, of the source FF on the film / detector (Q2 in the figures.) For particularly large diameter tubes this is not simple. For this purpose, we also use an additional auxiliary frame that holds FP and the radiation source in the predetermined position. The frame itself - perpendicular to the pipe axis - lies on top of the pipe under test or on top of the pipe insulation, and in this frame the FP can be moved in such a way that the pipe can be scanned over its full diameter. The movement of the FP in the frame can be measured to an accuracy of ± 0.1 mm, and so can the position of the radiation source. It is required that the frame can be stably fixed onto the pipe; that it can be easily and quickly moved between positions; and that it can, in principle, be rotated around the pipe circumference.

[0171] Regardless of filters and grids, the relative positions of the source-object- film / detector should be determined before each object is examined. This can be achieved in a three-dimensional coordinate system by setting the z-axis perpendicular to the point Q1 in the coordinate system x1 , y1 of the film / detector with origin Q1 (and the radiation source can be moved along the z-axis) and thus creating a spatial geometry. This creates a plane x1', y1' with origin Q1' parallel to the coordinate planes x1, y1 with origin Q1 such that the focal point of the radiation source is in this plane x1', y1'; and the object under examination - the part that can be imaged by the radiation source - is located between the parallel planes x1, y1 - x1', y1', where x1', y1' is obviously not limited in size - i.e. the coordinates can also be negative numbers - and the distance FF is the section Q2, Q2'. This effectively means that, for example, the actual diameter of the pipe under investigation may be larger than that of the corresponding film / detector, i.e. the total cross-section of the pipe cannot be represented by a single image. In case of a film, the evaluation of the film bent on the pipe cylinder (cylinder cleavage) remains flat in the x1 , y1 coordinate coordinate plane. In this case, it is important to know the coordinates of the perpendicular of incidence (TFF) of the radiation source in the plane of the film / detector; even if the perpendicular of incidence TFF does not touch the surface of the film / detector. Obviously, it is also required to know the radiation source pixel coordinates (PFF), where the TFF section is 1 unit of the reference distance, so that the angle subtended by the TFF and PFF lines can be given if necessary.

[0172] In the case of both film and FP, the projection point Q2 can be determined using a tubular device that can be fixed to the collimator and removed; and a test shot can be taken where only a relatively small diameter circular spot is illuminated; then the centre of this spot is Q2; and Q2' is the focal point of the radiation source; ( in case of FP - on a straight section of tube - the solution already mentioned and used for l0verification is possible, i.e. to select the intensity x,y passing through the reference point Q2, which must have a "regular" curvature and must pass through the reference point Q2). For film, a Dens base value must be selected (our choice is Dens 2.5), to which a calculator-determinable time is exponentiated, see equation (7). It is emphasized that other film types and / or other Dens values and other radiation sources can be chosen! The calculator exposure time according to (7) and the associated Dens value are mandatory! By fulfilling these conditions as boundary conditions (see later) MX is fulfilled according to (8).

[0173] With the exposure applied to the film, the I0and I values are automatically generated, from which the MX material section can be defined pixel by pixel. It is both fundamental and typical that I0can be calculated if the boundary conditions are met; I can be calculated if it is within the measurement limit and can be measured! For AgBr2-based films (see Tables 2, 3) log 10 Dens=l. In other words: I can be replaced by log10 Dens in the given relations; e.g. (4) = (2)-(5). This is also true - in principle - for FP, but there the intensity of I appears directly, here - preferably - we can use relations (2), (4).

[0174] For FP, all tests start with the definition of the I0value at the reference point - and from this, calculated for any pixel point - up to an exposure time of max. 150 sec, based on the expected exposure parameters. Since we know that the pipe diameter HVT data are variable, we assume at a first reading that we know the HVT value associated with the actual diameter of the pipe under test (wall thickness 2s and load M), measured by the film / detector with filter and / or Bucky grid, and the reference thicknesses; therefore, this measurement is primarily limited to 2s.

[0175] When measuring the selected reference point 2s, we should check whether the x,y intensity curves (intersection curves) passing through the reference point Q2 show any local defects, because the analysis of the so-called intersection curves will show whether local defects are possible or not. It is possible that the nominal data given for the reference point 2s is not real data, this is not apparent from the measurement on the reference point alone. (The reference measurement is only taken at a fault- free location where the selected reference point is preferably on the x-axis of the diameter of the pipe under test in the 3h-9hdirection; because the wear of the pipe wall is more uniform here than on the y-axis in the 6h-12hdirection. ) Going further, one solution is to pre-measure and determine the pipe section to be tested - and its M load - for HVT (and the Bragg scattering that occurs in this pipe section). Then, for the measurable range of the same pipe section - preferably at the actual diameter - the measured intensity sum is reduced by the ratio of the Bragg scattering that is determined separately:

[0176] (Insulation can be added to the lzand denominator of the formula; in addition, the HVT data of the M load in the pipe must be validated against the pipe parameters.)

[0177] From all this, it follows that the iron and M components are selectable:

[0178] Iron component:

[0179] According to (12), (13) and (14) above, the M load in the pipe can be eliminated; and then the (cross-sectional) intensity curves of the pipe that can be displayed on the FP computer will show real data. In principle, three images can be created: individually for each of the iron and M load components; or as a summation of these two components.

[0180] If the calculated value of the iron component I is greater than lΣ / (Braggσ>1,0)=l, then MX_Fe is less than the nominal data; i.e. the pipe is worn, thinned; and if this worn (thinner) part is filled by e.g. water; then this wear can be given with an accuracy of about 10%. Note that - if the water is e.g. in an Fe-based pipe - then the material of the pipe may change the HVT data of the water. In our opinion, the gamma spectra travelling in water (in M charge) are partially altered, "hardened", if they pass through the double pipe wall. In case of a steam pipeline, an empirical formula or relation can be developed by preliminary tests; but the water equivalent of steam can be defined from the nominal data and the specific enthalpy - specific entropy; and temperature - and specific entropy diagrams known from thermodynamics, knowing the steam pressure and temperature. It is essential to note that MX_Fe may also be defined by other methods according to the invention for a pipeline with a charge, see later.

[0181] If the object under test is radiophysically homogeneous - i.e. it has no load - then I can be set to the middle of the measurability range - to a predefined value. (In fact, I defines a Dehs value; therefore we denote it by Densl). In the actual photograph, the measured value of I (denoted ml), when different from the predetermined value of Densl, is then indicative of a weakening ratio (for WR, Fe: FeWR) from which the degree of error can be concluded; where for the nominal - reference - MX thickness, Densl = ml.

[0182] In principle, FPExp_T is selectable, but it is preferable to relate it to two predictable data: Densl or l0: If the measured value of I (ml) is different from the given value of Densl, a modified MX thickness is generated according to (2). I0and the measured I (ml) are applied to the modified material thickness MX, then the applicable HVT is determined by interpolation from the HVT table in Table 5, but in batch and detail; and the determination of MX is recalculated using these modified factors.

[0183] For FP, it is essential that I is predefined and fixed - i.e.: Densl - where the first calculated material thickness MX is based on nominal data; and the measurement result may therefore differ. In this case, a modification procedure (mod.) is applied, which is in fact an iteration based on feedback, e.g. the initial (nominal) material thickness MX and the attributed HVT data, then the resulting measured MX is now mod1 MX and the attributed HVT is mod1 HVT. In our opinion, in industrial practice, a maximum of mod2 MX and mod2 HVT is required and then only the measurement inaccuracy ±1,0%. Of course, there may be additional inaccuracies from geometry and environmental backscatter, which in total can reach - as already mentioned above - ±4.0%. If there is a charge M in the examined pipe, then the object is already radiophysically inhomogeneous - however, if the corresponding HVT data of the load M are known - Densl is still - preferably - defined for MX_Fe, such that the expected ml remains within the measurability range. If we know the weakening rate (MWR) of the M load in the pipe, then I0 must also be "weakened", i.e. reduced, by the same rate, according to the weakening rate of the load (So the load already shows the Densl calibrated to MX_Fe as ml according to the weakening rate. ) Recommendation: according to the expected combined attenuation ratio (ΣWR) of the M load and MX_Fe in the pipe, Densl can be increased beyond the measurability range; respectively, a minimum ml can be defined; then Densl = ml min. * MWR; for a radiophysically homogeneous body (Fe) (i.e. when the object can be described as consisting of a single material), the multiplication factor is FeWR.

[0184] From all this we can conclude: The localisation of the defect point in the pipe can be defined by changing the position of the radiation source and / or by sectional scans.

[0185] In this case, the industrial Bucky grid is not applicable because, when changing the position, the exposure is repeated without changing the position of the film / detector, and the radiation source is moving along the pipe axis (and / or constituent), which displacement causes the projection - mapping - position of the defect location on the film / detector to change; and this change geometrically defines the defect location. In the so-called sectional scans, the radiation source is moved along the circumference of the pipe, following the principle of the welding inspection. If the tests are performed in multiple sections, one following the other on the pipe cladding, we can reduce the wall reflections and backscattering. Each wall thickness measurement may also be a welding inspection where appropriate. Following the principle of the welding inspection, the wall thicknesses of the pipe cross-section can be defined and a diagnostic inspection of the welding joint can be carried out (this is advantageous for at least two reasons: firstly, the load may attack the material transition zone of the weld during operation, and secondly, cracks may develop in the transition zone itself.)

[0186] It is also advantageous and possible to perform diagnostic condition assessment on the pipe section under investigation, e.g. detection of debris, sediment, wear, etc. For the investigation of debris and sediment, a sideview should also be provided. In case of weld joint inspections, gas and slag inclusions, fusion defects can be well detected; while (capillary) cracks in the transition zone are less detectable. For the analysis and evaluation of the films, a film digitizer such as the FS50B is indeed recommended, since - according to the present invention - the desired accuracy cannot be achieved with simple blackening meters; and the processing of the film can only be automated!

[0187] It is important to note that during this procedure, two types of I intensities and three types of I0intensities can occur: on the one hand, Densl as the predefined "blackening" value and ml, i.e. the intensity actually measured; on the other hand, the I0intensity in the validation procedure ("blank" film / detector), the intensity in this validation procedure should be aI0; furthermore, e.g. the calculated I0, calculated from Densl according to (5), let this be cl0; and finally the reduced I0, i.e. rl0, as already mentioned. The validation aI0can contain two types of intensity: according to the Exp_T time (aI0Exp_T) and according to the max Exp_T time (aI0max), for FP it is max 150 s (if we were unable to attribute a specific intensity to the measured time unit, we determine the missing data by linear regression). From these data, Densl; as well as cl0, rl0and aI0max may exceed the range of measurability; while ml and aI0Exp_T do not.

[0188] In operational practice, the MWR attenuation rate of the M load in the pipe is not always predeterminable, but it can be measured. During the procedure, the known nominal data are given according to the following criteria: Consider given FF; aI0Exp_T, and aI0max; the pipe parameters; the actual diameter; the nominal wall thickness - from which 2s and the assignable HVT_Fe - and the attenuation ratio of the first measurement (reference) point:

[0189] (If no measured data are available, then for 2s effective diameter HVT_Fe should be 19.1 mm). Choose a value of Densl such that ml is - obviously - measurable, then cI0= Densl "*" FeWR; Exp_T should be according to formula (9). Perform the measurement, then Densl / ml = ( load) MWR; where cI0 / MWR = rI0; and this rI0will be the constant for further measurements (until the load parameters change). From this point on, at each measurement point, rl0is a constant value and ml can vary according to the measurement results. The serial number of the reference ml is denoted by 0ml and the other measurement points are numbered in ascending order (see Table 11 for the necessary relationships). This procedure is load-neutral and does not require the knowledge of the characteristics of the load; i.e. steam and oil pipelines are tested by the same procedure. It is essential that the charge itself and the pipe itself ( independently) are radiophysically homogeneous; therefore, by determining the s point of the selected 2s reference point, we can produce the validated 2s reference points which are valid for the rest of the entire identical pipe section. Determination of the wall thickness s is the next chapter of the invention, see later. (Note: the air attenuation ratio cannot be detected when testing an empty pipe). Before we move forward, we provide the main relationships - and their interpretation - so far, as well as the so-called iron equivalent (MX_Fe), which gives the thickness of iron that the object under test would be equivalent to. Here we can also use the so-called modus calculations (mod1); and the maximum of the material section that can be pulled tangentially to the inner pipe wall (MX_Femax, Femax).

[0190] DF radiation dose (for time Exp_T).

[0191] Densl : predefined, dedicated "blackening" value, (e.g. Dens2.5 for films) flexible for FP, not limited to an upper value.

[0192] Note: from an invention point of view, an object is radiophysically homogeneous if its effective atomic number Z (10) can be defined and the attenuation ratio (MWR) of the M load in the pipe is known; if the attenuation ratio of the M load in the pipe is not known - i.e. can only be defined by measurement - the object is considered to be radiophysically inhomogeneous. s : secundum, or the pipewall thickness s’ : enlarged projection of the pipe wall

[0193] TFF : the incident angle of the radiation source on the film / detector plane PFF : incidence line, position of the radiation source per pixel ml : the measured intensity ("blackening"); for a radiophysically homogeneous object of nominal thickness Densl = ml. mlsC : measured intensity curve vIsC : virtual intensity curve

[0194] MIC : mechanical intersection curve; MlCmax(imum) HVTChC : HVT character curve; HVTChCmax(imum) I0ChC : I0character curve

[0195] WR : weakening ratio; originally I0 / I, according to (6)

[0196] FeWR : weakening ratio of iron; as above (6)

[0197] MWR : weakening ratio of the load in the pipe. ΣWR : the combined weakening ratio of all components of the object.

[0198] MX_Fe : the iron equivalent, which gives the thickness of iron that the object under test would be equivalent to.

[0199] MX_Femax: the largest material section that can be drawn tangentially to the inner pipe wall, see at relation (2). al0Exp_T : the validation intensity of the "empty" FP irradiated over Exp_T time. aI0max : max time of FP radiation intensity (150 s). cI0: Intensity calculated from Densl based on (5). rI0: the reduced intensity of the above cI0, which is the ratio of the calculated intensity (cl0) to the MWR attenuation ratio of the M load in the Pipe.

[0200] From all these, the following equations are possible: io

[0201] Table 13

[0202] It is important that for both FP and film, cI0, rI0and DensI can exceed the measurability scale, while ml, by definition, cannot.

[0203] In industrial practice, the measurement of wall thickness 2s is related to the inspection of welds and / or the determination of the condition of the pipe under test. Here, knowledge of the 2s reference thickness verification - gauge - is essential. This can be an already known data or a validation etalon test. The validation etalon itself can only and exclusively be an object which is radiophysically - and thus also for Bragg scattering - equivalent to the tested pipe; i.e. the ml intensity measured in the etalon is quasi equivalent to the ml intensity of 2s. The test should be performed on an empty pipe, where the use of a filter and / or Bucky grid is allowed. Furthermore, it is specifically required to measure the actual s wall thickness (inservice). Based on what we know so far, it is obvious that the wall thickness s can only be measured as far as the maximum Femax intensity of the material section that can be drawn to the inner wall of the pipe - tangentially. This value, with filter and Bucky grid, is about 10", or about 254 mm. It is also obvious that the curve of the tube wall - due to X-ray optical reasons - scatters a significant part of the primary radiation out of the measuring range, therefore - based on our current knowledge - no meaningful HVT data can be generated.

[0204] In the course of our investigations, we found that the mechanical intersection curve (MIC) of the empty pipe under investigation - i.e. the encircling curve of the path lengths of the radii travelling in the pipe - can be easily mapped and displayed. As a working hypothesis, Appendix (1)(4) assumes that the primary radiation HVT through the pipe object under investigation is a constant value; then, based on the attenuation of this primary radiation, we can construct a calculated intensity I; where, if 2s = HVT, then the intensity curve I is equivalent to the mechanical intersection curve MIC; in all other cases, it is mathematically representable as a correlation to it. Furthermore: as it is well known that primary radiation, from the moment it enters the object under investigation, generates secondary and scattered radiation - which produces additional I intensity - where this secondary intensity also depends on the volume of the irradiated object (Table 7). As a consequence, these radiations are concentrated or superposed at a pixel point in an indeterminate way; i.e. the actual measured ml intensity curve cannot be determined.

[0205] The measured intensity curve (mlsC) could "in principle" correlate with the MIC intersection curve (for small diameters and thin wall thicknesses this is true). However - as stated earlier - the amplification of the scattering relationship would rule this out; i.e., the component I of relations (1-6) in the Appendix, which until now has been considered as the measured actual intensity, is in fact only one component of the measured intensity ml in the above relation (18).

[0206] It follows from the present invention that three of the four variables (I0; I, more correctly ml; MX, and HVT) occurring in the formulas in Appendix (1-6) occur in the pipe wall under investigation. In principle, I0, ml and MX; where MX is the length of an intersection passing through the pipe wall under test. The lengths of these intersections can be determined pixel by pixel, in s' projection; and, according to Appendix (3), the HVTs that can be derived from them, as well as the HVT ChC character curve of these HVTs.

[0207] Interpreted in the symbolic sense, the mechanical intersection curve of the empty pipe MIC - mapped from the nominal diameter - is a W format, and this W format should be also the format of the intensity curve measured by the mlsC; and this W is transformed into U format. Thus, the format transition W→U characterizes the mlsC measured intensity curve; while, in contrast, the HVT ChC character curve, which is the pixel by pixel envelope curve of the HVTs that can be attributed to the MIC intersection curve, remains - according to the invention - always in the W→ W format. Here two solutions are possible, as a first reading: a) " Book-in " (Bt.) is essentially based on real known data (actual diameter, actual wall thickness, etc.), i.e. we book reference etalons. There is no other initial step than to book in (Bt.) HVT ChC data by mapping the pipe diameters (5-50 inches) with varying wall thickness per pipe at the nominal diameter location (because other positions can be transformed from there). We do not recommend staggering the factory wall thickness of the pipe to be booked in, because then the scattered radiation patterns will appear differently on the film / detector. The length of the reference pipe should exceed the dimensions of the testing film / detector.

[0208] It is also recommended to take these pictures from 500, 707, and 1000 mm FF distance as shown in Figures 9 a, b, c. It is essential to note that here, according to the invention and the teachings, both the maximum of the MIC and the maximum of the HVT ChC are at the same distance from the outer pipe wall. Then the equality and / or deviation of the nominal wall thickness for pipes of the same diameter can be accurately measured for the actual wall thickness of the reference pipe. This equal distance is not affected by the M load in the pipe or the filter grid device. The maximum of the HVT ChC intensity may vary; but its projection position on the x~axis does not.

[0209] Now we can understand Figures 9 (a-h) showing the invention teaching. Figures a, b, c show examples of possible positions of the gamma source - pipe - FP (in figure 9 c, the TFF incidence perpendicular of the gamma source does not touch the FP surface). In all three figures, the projected position (s') of at least one s wall of the pipe can be displayed on the FP detector. It is sufficient if the cross-sectional projection of the pipe can be mapped to the x-axis based on the measured intensity, i.e. the projection of the inner pipe wall to the x-axis can be measurably represented on this intensity curve.

[0210] First, the spatial position of the radiation source - object - FP with respect to each other has to be positioned and determined in a way that the mlsC intensity curve of the projection image of at least one wall of the tested pipe can be displayed on the FP x axis. With the first low intensity image, the outer surface of the tube is mapped on the x-axis so that it can be seen through the insulation, and the coordinates of this location are recorded (this image is not diagnostically evaluable and also does not contain the required mlsC intensity curve). Subsequently, with increased exposure time, the full mlsC intensity curve including the position of Femax is displayed (This image is also not intended to be diagnostically evaluable, e.g. the low intensity image displayed in the previous image may be "burned in"; i.e. the outer edge of the pipe is no longer visible).

[0211] We now have a measurable section on the x-axis, which also contains the magnified projection image (s') of the examined pipe wall s. From this, for wall thickness s, we first give the incidence perpendicular FF (TFF) of the radiation source to FP and the coordinate of this perpendicular to the x axis. It is neither a condition nor an exclusion if this TFF incidence perpendicular does not touch FP. (The x coordinate must be given regardless, see Figure 9c). As the next step, we specify the line connecting the point s72 of the radiation source, which connects the selected pixel point to the radiation source (PFF).

[0212] Furthermore, the geometric position of the wall thickness s in the pipe wall is given as shown in Figures 9 a, f, where the line s points to the geometric centre of the pipe and the line PFF through the position s / 2 from the source - extending to s' / 2 - is perpendicular to the wall thickness s. Ide még jön egy mondat. The magnification (Ma) can be defined as s' / 2 radiation source section divided by s / 2 radiation source section. We also consider the mapping angle a subtended by the TFF line and the PFF line; then:

[0213] For this procedure, the tangential mapping of the pipe wall with a perpendicular incidence is not required. Optimal mapping is when the TFF perpendicular to the incidence passes through s / 2 → s' / 2 points, because then Cos a is 0. the measurement of the wall-thickness profile s is derived - trigonometrically - from the measured data of the actual pipe diameter (2r) and the tangent to the inner wall of this pipe under test (MX_Femax), where MX is the length of one material section per wall-thickness section s; where s actually extends to the limit of the tangent to the inner diameter of the pipe as the largest material section MX_Femax

[0214] These MX material sections are imaged from the quasi-point radiation source; therefore, they are not parallel lines, but this does not affect the accuracy of the measurement. Furthermore, when Femax is significantly absent from the mlsC measured intensity curve (Figure 9 d) we can display the corresponding HVT ChC character curve, and the MIC mechanical intersection curve, and the corresponding mlsC measured intensity curve for the pipe under test (Figure 9 f). In order to illustrate this, we need to see the position of l0relative to the mlsC measured intensity curve (Figure 9 e).

[0215] Note: as discussed earlier, the measured intensity curve of mlsC must be within the measurability range, but I0can exceed it without upper limit

[0216] (Figure 9e). In fact, figures d-e are complementary.

[0217] Knowing this data, the vIsC virtual intensity curve (Figure 9g) can be defined, which correlates - possibly matches - the MIC mechanical intersection curve; and thus the projection s' on the x-axis will accurately show any deviation from the nominal wall thickness s'. b) The second - and more preferable - option is to measure the wall thickness directly without the aforementioned booking in (the so-called booking in is undoubtedly preferable). Here, all data except the wall thickness are known. To do this, first of all, we must clarify the evolution of the MIC mechanical intersection curve and the HVT ChC character curve; e.g. according to the positions in Figures 9 a, b, c; from different FF distances (e.g. 500, 707 and 1000 mm); and according to the specific test concerned. For the analysis and modelling of the problem, we must have at least one (error-free) measurement pipe - e.g. 08" 203.2 mm * 20.0 mm, where Femax 121.06 mm - with a measured intensity curve (mlsC) of clearly U format. The longitudinal axis of the pipe should be parallel to the y-axis, and the radiation source (typical size: 1 ,0 mm * 3,0 mm) also parallel; to reduce the so-called geometric dullness. The incidence inclination of the radiation source should be 5 - 10 mm above the x-axis; because this way the mapping sections of the tube - parallel to the x-axis - are in a uniform position; and thus - on our FP device - we can map about 2000 lines. Note that the planar surfaces of the pipe ring cut per pixel row are not parallel to each other; i.e., the side facing the radiation source is the thinner; even though s' pixel area shows this to be parallel. From the image positions and data, we first determine the outer edge of the pipe from the y-axis - x-axis is the mm distance - then the mechanical intersection curve for the selected pixel set, the I0ChC characteristic curve, and the measured intensity curve mlsC; and the Cos a that can be determined per pixel.

[0218] Based on these, the Ma magnification and s' value can be determined. For example, let Ma magnification be 1.25 and Cos a 30° = 0.86602; then s' = 28.8675 mm (s = 20.00 mm).

[0219] Since our current FP device has a pixel resolution of 0.2 mm, s' projection can be broken down (measured) into 144.33 steps; i.e., in practical terms, 144-145 steps.

[0220] As a rule of thumb, s<s'; which means that a single s intersection is 0.13856 mm thick. Of course, we don't need to stop after step 145, because we should go at least 5-10 mm further; possibly up to the actual pipe diameter. Obviously, we are measuring real data from the outer edge of the pipe wall to the Femax projection, therefore no reference etalon is required for this procedure.

[0221] Knowing the outer edge of the pipe wall, the HVT ChC curve can be defined by determining at least 20-25 mm pixels of HVT pixel by pixel starting from the measurability range on the x-axis - which data are then mapped to the y-axis in alignment with the MIC curve - and using these to establish the equation describing the HVT ChC curve, which then represents the pipe under test as a solid cylinder. (Note: measurements of HVT of less than the actual wall thickness s, but at least 20- 25 mm pixels, are sufficient to establish this equation, which will be valid for the whole section of pipe under test with unchanged parameters. The equations of the above mentioned character curves are determined at the beginning of the test, e.g. according to the program "Origin").

[0222] In the programming, for modelling purposes, we increase the real wall thickness data; i.e., assuming the MIC mechanical intersection curve to be a solid cylinder body as previously, the MX intersections are given in increasing order pixel by pixel according to relation (20). Then this curve represents the actual pipe diameter on the y-axis; and the curve itself is written according to a logarithmic regression, where the starting point 0 of the curve is given on the xy-axis. Let us assume that we can measure the intensity of ml on our modeling calibration pipe from the 25th pixel, i.e. 5 mm, from the outer pipe wall s' (wall thickness s 3.464 mm). Then, knowing I0, ml and MX (52,607 mm), which are real data, we determine HVT in ascending order pixel by pixel up to the nominal wall thickness s' 28,8 mm, e.g. 5,0; 5,2; 5.4; 5,6 28,4; 28,6; up to 28,8 mm. Subsequently, a HVT ChC character curve is determined from the measured HVTs at a wall thickness section of s' 5.0-10.0 or 5.0-15.0 mm. Beyond s' 28.8 mm (s 20.0 mm) wall thickness, MX is no longer a real data - while I0and ml are - and therefore (at s 20.2 mm) at s' 30 mm the resulting HVTs no longer match the HVT ChC curve, i.e. the wall thickness s of our modeled calibration pipe ends here.

[0223] It follows that, for the nominal thickness s' of the tested pipe wall, the geometric position Femax of the MIC mechanical curve is compared with the geometric position of the corresponding HVT ChC maximum, and then the MIC mechanical intersection parameters are redesigned (modified) - i.e. the nominal wall thickness s, i.e. the position of Femax, is changed - until the geometric positions of Femax and the HVT ChC maximum are equal.

[0224] Therefore, according to the recognition and teaching of the present invention - in the case of pipe tests - HVT cannot be constant; - i.e. the MIC is linear in its crosssection - but it follows a character curve (ChC) whose maximum and geometrical position of the MIC mechanical cross-section curve Femax on the x-axis, i.e. its distance measured from the relevant outer wall, is always equal to the determinable distance of the maximum of the HVT ChC. (Obviously, this is also true for a solid cylinder object.) According to our experience - for relatively "thin-walled" pipes - mlsC measured intensity curve - can correlate with MIC mechanical intersection curve; then we can also measure wall thickness; however, it is certain that with the wall thickness measured in this way - due to the X-ray / gamma scattering conditions that have already been revealed and referred to, the image contrast that enables the wall thickness to be measured disappears with the increase of the pipe thickness (and / or the increase of the pipe diameter). After this - knowing I0, nl and HVT, since these are real data occurring in the pipe wall - according to Appendix (2), we can determine MX material cross-section thicknesses per pixel along the entire length of the examined pipe section, which also clearly mark the position of wall thickness s on the MIC mechanical cross-section curve. Given that the measurement uncertainty of I0and Ml data is approx. 0.5% (i.e. approx. 1% overall), the deviations exceeding this can be considered errors. This only means that there is a deviation from the reference value in the MX material thickness of the measured pixel, but it does not directly localize the location of the defect. (Please note that capillary cracks do not cause measurable changes in MX material thickness; while, for example, root defects, slag and gas inclusions will be measurable.)

[0225] Based on the data so far, the wall thickness s and 2s of the same pipe can also be measured in the case of M load (in this case, the ml intensity curve shifts towards smaller values). It is advantageous if we have a 2s reference point, because - according to the philosophy already described earlier - we match its l0to the 2s HVT according to Appendix 5, i.e. we eliminate M load (Fig. 9 h). The reference point 2s can be created without an external etalon from the tested s wall thicknesses. It is our well-founded opinion that a virtual image of the body (object) to be examined can be created on the computer of the FP and the film digitizer - even before the actual examination; which can show all pre-programmed deviations compared to the actual test. Essentially, in addition to the selected exposure parameters, the activity of the radiation source must also be known, then the material thickness s and 2s of the object to be examined can be calculated item by item. On this virtual X-ray image, the body (object) to be examined can also be displayed, to any cross- sectional plane of which a selected HVT constant can be assigned (e.g. one of 12.7 mm, 14.6 mm, 19.1 mm) and the derived intensity curve I according to the basic formula (1) ln(l / l0) = - μx; which can be compared to the actual ml intensity curve on the real recording. The present inventive solution solves the set tasks in a way that the objectives of the invention are fulfilled if they are used as intended with any radiation source (e.g. Ir192, Cs137, Co60 isotopes, or a 6-8 MV betatron accelerator), any investigated object (e.g. 40-50 inch high-pressure oil and gas transmission lines), and any radiation-receiving device (film / detector).

[0226] In light of the above, the inventive solution solves the set technical task, moreover, the inventive solution is independent of the type of detector. In the present patent application - from the invention teaching point of view - there are at least ten essential novelties compared to all prior knowledge (including the patent application according to the state of the art):

[0227] (I) The basic formula (1) ln(l / l0) = - μx represents the attenuation of primary aw2t radiation. Furthermore, the I0that can be derived from this is the only real data, and I - which is the intensity of the primary radiation passing through the examined object - is not equal to the sum of the total primary and secondary radiations appearing in the measured intensity.

[0228] (II) The AgBr2 / GOS film / detectors also have different Kf characteristics and do not match the theoretical physical data either; furthermore, HVT is not constant.

[0229] (III) The exposure (summarized) parameters - and the relevant projection points - in each case define a MX material thickness section in the pipe wall, and a measurable section can be selected on the x axis, which determines the actual wall thickness s with the knowledge of the actual pipe diameter (2r) - and the relevant HVT ChC -which if we know HVT ChC is smaller than the actual wall thickness s, but can be determined from min. 20-25 mm pixel points; where, in case of pipe inspection, the procedure is load-neutral;

[0230] (IV) The verification etalons should be appropriate for the object being tested in terms of radiation physics and Bragg scattering.

[0231] (V) In the case of FP, the external scattered radiation can be visualized by measuring al0.

[0232] (VI) In the case of FP, the measurement accuracy can be increased with mode calculations up to signal / noise discriminability - plus geometric uncertainties.

[0233] (VII) Preferably, in the case of AgBr2films - according to relations (1)-(5) - log 10 Densl is defined and can be calculated. The computability of Densl according to (1), (4) is generally true - also in the case of FP; and here the value of Densl can exceed the measurability scale - without an upper limit.

[0234] (VIII) A virtual image of the body or object to be examined (in the case of FP and e.g. on the computer of the other mentioned film digitizer) can already be created before the actual examination.

[0235] (IX) If FP is loaded with N electronic noise, this N noise can be extracted from the measured signal using a regression procedure.

[0236] (X) According to the teaching of the present invention a significant part of the Bragg scattering generated in the examined object can be excluded from the radiation reaching the film / detector if the device is used with a filter and duplex Bucky grid.

[0237] We also note that the execution of the procedure according to the invention can be automated, i.e. the calibration / authentication step of the method to create a virtual image of the examined object, and then the step of the method to test / determine the thickness can preferably be carried out using a microcontroller or a properly configured / programmed computer.

Claims

AMENDED CLAIMS received by the International Bureau on 03 April 2025 (03.04.2025)1. A method for radiographic inspection and determination of material thickness of objects, where the object [2] to be inspected has mathematically definable geometrical and radiophysical parameters, wherein the inspection method is based on releasing imaging X-rays or gamma radiation from a quasi point source [3] - the source [3] and the object [2] - of radiation [2], and a radiation absorbing device [1] in a fixed position relative to each other, and this radiation is absorbed by a radiation absorbing device [1] - - which radiation absorbing device [1] may include a scattered radiation reducing filtering device [4] and / or a Bucky or Lisholm grid [5] which may be applied in front of the device - arranged on the opposite side of the object [2] to the radiation source [3], where the amount of primary radiation absorbed and transmitted at any given point of the radiation absorbing device [1] is measurable by the radiation absorbing device and which, as the technique currently stands, is the material thickness x=MX of the object under test:furthermore, this equation (1) can be described in a different reading and relation:where p=ln2 / HVT; respectively HVT=ln2 / μ; ln2=0,693147; x=MX material gauge (thickness) for decimating thickness: In10 / μ; for N-feed thickness: InN / μ; where Io is the intensity of the primary radiation from the radiation source [3] to the radiation absorbing device [1], without the insertion of the object to be tested, I is the intensity of the absorbed primary radiation passing through the object under test [2], μ is the linear attenuation coefficient of the imaging radiation and the object material, and x = MX is the radiophysically homogeneous material thickness of the object under test, and In is the logarithm function e, where e is the Euler number; HVT: HalfValue Thickness: where the primary radiation entering the object under test [2] - and essentially equivalent to the intensity of l0radiation - can, by law, generate Rayleigh, photo, Compton and Bragg scattering - as well as pair creation - from the moment it enters the object under test; furthermore the irradiated volume and geometry of the object under test further modify the resulting secondary radiation and Bragg scattering, according to Table 7; and the sum of these radiations - the superposition - is represented by the signal of the measured intensity, ml, according to the equation ml = Iprimer + iRayleigh + Ifoto + Icompton + I Bragg + (extern (+ Ipairproduction) ,From all this, we can derive the invention and teaching that the components of formula (18) cannot be determined in advance, therefore p and HVT cannot be constant - according to the literature - and that the intensity I according to (1-6) is not real data for the material thickness x = MX, because (18) ml is not the same as the intensity I (1-6) resulting from the attenuation of the primary radiation entering the object; then the test procedure for determining the material thickness MX of the object under test is characterised by(18) ml + (1-6) I = Iprimer; hence (4) ml is not a predictable value; furthermore HVT is not constant, hence the boundary condition is that in order to determine the component MX from the four main components Io, ml, MX, HVT according to (1-6), one needs to know Io and ml, and the reference thickness MX mapped by the radiation source [3] and geometrically associated with it; and based on this knowledge, the corresponding HVT according to [3]; i.e: the deviation from this reference thickness MX is measured during the test; noting that if the mapping position of the source [3] to the material thickness MX to be measured remains constant during the test, then HVT is considered constant; if these data change pixel by pixel; then the change of ml must still be measured, as long as the change of Io, MX, HVT remains calculable; thus, the formulas (1-6) according to the present invention method, with the exception of (4), are valid for calculations; furthermore, during the test:- in the first step of the procedure, the l0component of equations (1-6) are validated on the basis of equation (1-6) - or derived equations - by emitting radiation directly from the radiation source [3] to the selected Q2 reference point of the surface of the radiation absorbing device [1], i.e. withoutinterference of the object, with essentially perpendicular incidence l0, up to the upper limit of the measurement range but with different Exp_T exposure times, exposing the Q2 reference point of the radiation absorbing device to different Exp_T doses; then, for these DF doses of different Exp_T durations, we determine, in our case by linear regression and using the engineering knowledge required, for example with the ‘Origin programme’, the l0intensities that can be assigned to any pixel point of the radiation receiving device for a max. Exp_T duration and for any Exp_T duration and any FF distance, then l0is not limited from above beyond the measurement scale; and then measure these l0intensities to see if they do not contain (18) External backscatter; furthermore, these values are stored as reference data, where these Io reference data can be determined repeatedly, as required, at any pixel point of the radiation receiving device during the lifetime of the source; furthermore in the second step of the testing procedure, a reference area of known MX thickness is selected on the object to be tested - or an equivalent etalon in terms of radiophysics and Bragg scattering (18), which for MX reference area plates are extended parallel mirror symmetric surfaces; for pipes [8], the sums of the two opposite s wall thicknesses 2s of the tube diameters, regardless whether this pipe [8] is empty or is loaded with a load M; where the material of the pipe to be tested, its diameter and the material of its load M are not factors; only the reference material thickness 2s MX of the pipe wall is used; then, according to the parameters defined in the first step, we define l0at the chosen Exp_T time, then we measure the intensity ml at the reference thickness MX of the object, i.e. we determine the lo / ml = ΣWR total weakening rate, and from this data we determine the reference MX (3) HVT value; then, knowing all this in the third step of the testing procedure - on all chosen measuring points - the material thickness 2s MX of the test object is determined using the formula in relation (2)in the fourth step, correction of the measured 2s MX material thickness is carried out.a) If the measured material thickness 2s MX is larger than the reference material thickness, no further actions are required; b) if the measured material thickness is thinner than the 2s MX reference thickness - and this thinning is filled by load M - then for this missing material thickness, it must be determined what 2s MX material thickness would correspond to load M, and this equivalent value must be subtracted from the measured 2s MX material thickness.

2. The procedure according to claim 1 , which aims at realizing the nominal wall thickness s, given the actual pipe diameter, in a way that all necessary parameters except the actual wall thickness s are known - i.e: I0, ml and the material thicknesses MX that can be mapped per pixel row in the tube wall as calculable material sections - where 2r is the actual tube diameter, s is the part of the tube radius r, where s extends, in its full extent, from the outer edge of the tube wall to the inner wall of this tube up to the tangent MX_Femax, which is also the maximum material thickness MX that can be mapped per pixel on the y-axis by the radiation source of the tube under test; and also any MX material measure associated with any thickness of this wall thickness s, which may extend from the outer edge of this tube to MX_Femax material measure MX, where s is any thickness and any assignable MX material measures’= s * Ma * 1 / Cos a ; s = (s7Ma) * Cos a ; s’ / s=Ma * 1 / Cos a ). (19) where s is the projection of the wall thickness appearing on the x-axis, Ma is the magnification rate, which is the ratio of the section length of the s / 2-radiation source to the section of the s / 2-radiation source; and Cos a is the angle subtended by the perpendicular of incidence TFF and the line from the source, passing perpendicularly through the point s / 2, to the point of projection s / 2 PFF, which gives the projection position s' of the nominal wall thickness s on the x-axis; to which we assign sMX material sections according to the number of pixels in the projection s’ - i.e., we form MX material sections from the nominal wall thickness s as many as the number of pixels in the projection s’ - is characterised by redesign of the predefined range MIC ± of the reference nominal wall thickness s by mapping the geometric position of the outer edge of the pipe wall to be tested to thex-axis with an underexposed image, and then mapping the position Femax of the pipe wall to be tested with a second image so that the measured intensity ml can be measured; further, we determine the geometric position on the x-axis from which, given Io, ml and MX per pixel, the actual wall thickness s is smaller than, but at least 20-25 pixels from the outer wall - as here l0, ml and MX are real data - we determine all HVT-s (3) - which are varying values and cannot be calculated in advance - then given these HVTs, we determine the covering curve for these HVTs in the form of a ChC characteristic curve, which curve then represents the tube under test as a solid cylinder body by applying the MIC mechanical section data, assuming a solid cylinder body, to determine the HVTs in ascending order until the HVTs deviate significantly from the HVT ChC curve; i.e. i.e. the fracture point s’ of the HVT ChC characteristic curve determines the projection of the wall thickness on the x-axis, and this procedure can be further mapped from pixel row to pixel row; where this procedure is neutral for the load M; and the necessary MIC and HVT ChC characteristic curve calculations can be constructed using the 'Origin' program mentioned above.

3. For radiographic diagnostic examination of objects with claim 1 device, two superimposed Bucky or Lisholm grids are required in order to reduce the scattered radiation and increase the contrast enhancement ratio of the tested object placed on top of each other between the tested object [2], and the radiation absorbing device [1], such that the arrangement of the two Bucky or Lisholm gratings is parallel to the radiation-absorbing device , characterised by that the centreline of the second grating is preferably rotated ninety degrees relative to the first grating; whereby the intersection of the two centerlines on said device becomes the projection point of the radiation source, perpendicular to which the radiation source is located at the distance FF specified by the manufacturer.

Images