Digital Radiography Systems Techniques and Performance Evaluation for Space Applications

Transcription

1 172 Digital Radiography Systems Techniques and Performance Evaluation for Space Applications V.N. Misale, S. Ravi and R. Narayan Proceedings of the National Seminar & Exhibition on Non-Destructive Evaluation Liquid Propulsion Systems Centre (LPSC) Indian Space Research Organization (ISRO) Bangalore NDE 2009, December 10-12, 2009 Abstract The x-ray radiography is a mandatory NDT method for qualification of propulsion system components and assemblies to ensure structural integrity with zero defects. The major development in conventional film radiography is digitization of x-ray films, followed by radiographic image processing and quantitative evaluation to enable accept/ reject decisions. An LS85 Kodak film digitizer is used for digitization and analysis. Performance evaluation of this system with reference to current standards and a case study for estimating depth of defect are discussed. The next major development is film less amorphous Silicon based flat panel system with pulsed x ray source The Fox Rayzor flat panel system is being used for various space applications. This performance evaluation of this system and two case studies to measure the gap between the vanes and hemisphere and Electron beam impingement analysis are discussed in this paper. 1. Introduction Liquid propulsion systems are the main entities of our space program. The propulsion systems consists of propellant tanks, pressurant tanks, various components like filter assemblies, pyro valves, latch valves, venturi, plumb line systems with transition joints, reducers, welds, engines, thrusters, etc. The propulsion system elements are realized in all welded condition. And X ray radiography is a major method to ensure that welds are free from internal defects. Other NDT techniques like ultrasonic testing, penetrant testing, eddy current testing, holography, are also used as supplementary methods. Propulsion systems with zero defects are used in both launch vehicles and satellites. Once launched space systems are not amenable for any repair or rework. Further the effects of defective components can be catastrophic leading to mission failure. Hence it is necessary to ensure zero defects at various stages of realization of the systems. X ray radiographic NDT plays a vital role in this direction and it is one of the mandatory tests for qualification of the systems. Using X ray radiography, the discontinuities are characterized and analyzed to determine if the observed discontinuities are acceptable. If they are not acceptable, they are repaired and rechecked. If they are beyond repair, the components are rejected. Traditionally x-ray film radiography with fine grain x-ray films and mini focus x-ray machines were good enough to detect the unacceptable defects in propellant tanks. Recently the film radiography is being supplemented with modern digital radiography methods for defect characterization, better coverage, improved detection capabilities, faster inspection, better archiving and loss-less image transfer for teleinterpretation. These new methods being used are x-ray-film digitization and image processing techniques using Kodak LS85 film digitizer and amorphous silicon detector based VIDISCO Fox-Rayzor flat panel imaging system for near real time image capture, image processing, archiving and lossless image transfer for tele-interpretation. In this paper we discuss performance evaluation of these systems and few case studies as applied to our propellant tanks. 2. X ray Film digitization system The purpose of x-ray film digitization is to capture the film radiographic image and convert it into digital image. Digitization helps in digital archiving, quantitative evaluation, image processing, automatic image evaluation, (remote) image transfer and production of reference catalogues for flaw Fig. 1 : Principle of a laser scanner (Kodak LS85) and photograph

2 NDE 2009, December 10-12, evaluation. The principle of point-by-point digitization is employed in our system. Fig.1 shows the principle of our laser scanner. 2.1 Digitization process The x-ray film is moved in front of a collection tube. A laser beam (wavelength ~680 nm, red) with a fixed diameter (50 µm) passes through the film. The collection tube integrates the diffuse transmitted light through the film and a photo multiplier (PMT) registers the same with voltage proportional to the light intensity behind the film. During the scan the folding mirror moves the laser beam along a horizontal line on the film. The output after logarithmic amplification and digitization with 12 bits yield grey values that are proportional to the optical density of the film. The scanning speed is 75 lines per second. The laser scanner illuminates the film with focused light and measures the diffuse light intensity behind the film. Whereas the other methods of digitization namely line by line digitization using CCD scanners and array digitization by CMOS cameras illuminate film with diffuse light and measure the light intensity that passes the film in one direction. Laser scanner principle has two main advantages: The quantity of collected light is considerably higher. The laser focuses the whole light intensity on the point to measure. The radiation passing the film is collected over a spatial angle as great as possible in the detector. Scattered light from regions with low optical density passing to regions of high optical density, thereby distorting the measurement, is nearly avoided by point wise illumination. 2.2 Film digitizer experiments An EPRI reference radiograph (8"x10") as per ASTM E (1) shown in Fig. 2 was scanned for the evaluation of our digitization systems to evaluate basic parameters of our film digitization system. The reference radiograph has following six types of targets. 1. Spatial resolution targets, consisting of converging line pairs in a range of lp/mm. 2. Density contrast sensitivity targets, block targets of 1 cm² with ÄD =0.05 at D=2 and ÄD=0.10 at a darker background of density D=3.5, 3. Stepped density targets, a series of 13 Nos of 1cm 2 blocks with density between D=0.5 and 4.5: 4.5, 4.02, 4.0, 3.5, 3.02, 3.0, 2.5, 2.02, 2.0, 1.5, 1.02, 1.0 and 0.5, 4. Sharp edge targets to ensure unsharpness < 10 micron 5. Linearity targets, these targets provide a geometrical measurement scale in horizontal and vertical direction of 25.4 mm units, 6. Parallel line pair target: a parallel line pair gage with a resolution between 0.5 and 20 lp/mm. Fig Results : EPRI Reference radiograph for digitizer Evaluation (not to scale) 1. Spatial resolution achieved 4.0 lp/mm 2. Density contrast sensitivity achieved D =0.02 at D=2 and ÄD=0.1 at a darker background of D=3.5, 3. Linear response of [Grey values (GV)] to optical densities (OD) is evident in the range 0.0 to 3.5 D. and from D the response needs correction factor nearly proportional to optical density. See table 1 4. Linearity is uniform and equal in both horizontal and vertical directions. 5. Working ranges of optical densities: 0-4; 6. When the optical densities are more than 4.0 clipping occurs. And digitizer cannot be used. 2.4 Performance of digitizer The European standard CEN TC 138 (2)Non-destructive testing - Qualification of radiographic film digitization systems has 2 parts namely - Part 1: Definitions, quantitative measurements of image quality parameters, standard reference film and qualitative control. And Part 2: Minimum requirements. In this section we discus performance of our LS85 Kodak Digitizer with reference to the standard. 1. Standard requires that digitizer must be able to reach the density of 4 or 5, without increase the image noise by its own detector noise. Our scanner is able to cover the density range up to 4.0D with linear response is evident up to density level 3.5D and from 3.5 D to 4.0 D minor correction is needed. The correction factor is as follows

3 174 Misale et al. : Proceedings of the National Seminar & Exhibition on Non-Destructive Evaluation Table 1 : Optical density Vs Grey level values Optical Expected Observed Correction density grey value grey value needed NIL NIL NIL NIL NIL NIL NIL NIL NIL NIL NIL ~+50D ~+50D clipping True Grey value = Observed grey value 0<D<3.5 = Observed grey value + 50 D 3.5d D< Further, the NDT X-ray films have a high signal-to-noise ratio (SNR) and high spatial resolution. Keeping in view the wide range of NDT applications, in terms of the grain sizes of silver halide crystals, (1 µm-100µm) and X- ray energies between 40 and kev requirements for spatial resolution is limited to the inherent unsharpness, which is caused by interaction of X-rays with the screen-film system. The unsharpness values range between 30 and 1000 µm, depending on the energy and the screen film system. The inherent unsharpness values determined independently for 200keV with D4 films were found to be ~80µm. and hence the minimum requirement for resolution is 4.0 lp/mm, which was achieved in our digitizer. 3. Next requirement is minimum digital resolution requirement >10 bit. Our system is has 12 bit resolution. 4. Of the three quality classes namely DS - the enhanced technique, which performs the digitization with an insignificant reduction of signal-to-noise-ratio and spatial resolution, DB -the enhanced technique, which permits some reduction of image quality and DA -the basic technique, which permits some reduction of image quality and further reduced spatial resolution. Our system is found to be of class DS while working up to Optical density level 3.5 and that of class DB from density level 3.5 to 4.0D. Digitized images archived have not shown any deterioration since last 5 years thus validating the film digitization and computer based archiving, handling processing, evaluation. 5. As far as image transfer is concerned, KODAK LS85 digitizer is fully compliant with ASTM E 2339, Standard Practice for Digital Imaging and Communication in Nondestructive Evaluation (DICONDE), 2.5 Case Study: Digital evaluation of Electron Beam (EB) welds of a propellant tank. In a propellant tank closure joint joining two cylindrical shells made of titanium alloy Ti6AL4V by means of electron beam weld, a discontinuity named under bead undercut was observed and it was not accessible for visual inspection and repair. Defect characterization in terms of size and depth was necessary for acceptance decisions. The joint configuration is a butt weld with bottom lip configuration and full penetration EB weld. Figure 3 shows cross sectional view of joint. Fig. 3 : Cross sectional view of EB joint in a propellant tank

4 NDE 2009, December 10-12, Fig. 4 : Electron beam weld of a propellant tank Amorphous silicon FPD Pixel Layout, single pixel & circuit. Fig. 5 : Magnified view of under bead undercut and its grey value line profile By scanning the x-ray film, digital image was obtained by LS85 Kodak digitizer. This digital image is shown in figure 4. Magnified view shows the presence of the under bead undercut. The size in terms of breadth and width could be measured easily from magnified image. To estimate depth dimension, we made a Ti6AL4V step wedge of known step thick nesses and its radiograph along with the discontinuity is shown in Fig. 5. Calibrated Grey value line profiles of thickness steps were compared with the defect line profile and an estimate of depth of defect was made. The defect depth (0.3mm) was well within operating margins and the tank was cleared for flight and flown successfully. Fig. 6 : Vidisco Fox Rayzor FPD Imaging 3. Digital flat panel radiographic imaging system Digital flat panel radiography is a form of x-ray imaging, where digital flat panel x ray sensors are used instead of x ray film. to record the X-ray image and make it available as a digital file that can be presented for interpretation and saved as an x-ray record. The advantages of DR over film include immediate image preview and availability, a wider dynamic range and ability to apply special image processing techniques that enhance overall display of the image. In this paper we discus Amorphous Silicon (a-si) based FPD imaging sysstem perforamance and case studies on NDE of Propellant tanks. 3.1 Amorphous Silicon (a-si) type of FPD have a-si detectors, positioned just below Gadolinium Oxysulfide (Gd2O2S) scintillator that converts X-ray to light. The light is then channeled through the a-si photodiode layer where it is converted to a digital output signal. The digital signal is then read out by Thin Film Transistors (TFT s) The image data file is sent to a computer for display, processing and interpretation. Figure 5 shows pixel layout individual pixel and circuit Fig. 7 : Mechanism of amorphous System silicon x ray detection and Gadox Scintillator representation(3). Figure 6 shows a photograph of Vidisco Fox Rayzor FPD Imaging system and figure 7 shows the mechanism of x ray detection in amorphous silicon detector. 3.2 Case Study of Propellant tank burn through and repair During the realization of a titanium propellant tank, electron beam weld was abruptly stopped due to instantaneous variation in power supply. The weld showed presence a deep open cavity. Possibility of burn through could not be ruled out, even though depth of open cavity was visually measured to be ~ 2mm against wall thickness of 5.5 mm. The defect was rewelded to fill the cavity. And the

5 176 Misale et al. : Proceedings of the National Seminar & Exhibition on Non-Destructive Evaluation Fig. 8 : EB weld defect in propellant tank, which was resolved as holes in guide cone and intermediate part, and weld is free from defects. while the gap was uniform in 3 vanes (~3 mm) and the other vane was touching the hemisphere. Both are shown in Fig Conclusions Fig. 9 : Left image shows vane gap and right image shows no gap. weld surface appeared to be satisfactory. However Normal radiographic exposure was as shown in figure 8. The defect appeared as burn through and was not in acceptable zone. This called for serious investigation and Vidisco Fox Rayzor FPD system was put to use. The digital radiography scanning with FPD was carried out at small angle and the defect began to resolve into 2 overlapping pores one darker image and other as lighter image. And when the angle was nearly tangential, total separation of images was seen and it was found that the weld was free from defect but the EB impingement has caused holes in guide plate of 0.6mm thickness and intermediate bottom of 2.0mm thickness. The hole sizes were characterized and effect of the holes on the function of the tank were investigated. The FPD system was found to be very effective. 3.3 Case study II: Propellant tank vane Assembly The surface tension type propellant tanks have vane assembly as apart of propellant management system. The vane assemblies are made of thin commercial pure titanium sheets. Normally a template is used during assembly to ensure certain minimum gap between vane and hemisphere. In one of the propellant tanks, during vibration testing variation in natural frequency of the internals was noticed and further investigations were called for to verify that gap exists. Vidisco Flat panel digital X ray system was put to use. Minimum gap between vane and hemispheres were imaged as a tangential projection all along the length of vanes and the same was repeated for all 4 vanes. It was observed that The digital radiography has supplimented x ray film radiography to a greater extent. The X ray film digitizers have given a new dimension for information extraction from x ray film. They help in archiving the image, they provide grey value levels which are linearly proportional to optical densities to enable quantitative characterization of discontinuities and depth evaluation. A case study of underbead undercut proves the point. They do not deteriorate the image quality and maintain acceptable quality levels. Image processing functions do helpin interpretation.our LS85 Kodak digitizer function is in accordance with current standards. One of the limitation was non linearity in film densitiesgreater than 3.5 and clipping effect at 4.0 D. The Vidisco Fox rayzor amorphous silicon flat panel imaging system is very versatile, amazingly quicker, and provides acceptable radiography quality and two case studies have given a new insight into the way we can do nondestructive evaluation of Propulsion systems. 5. Acknowledgements Authors are greatful to LPSC Management, DD SRQA, DD LPSC(B) and Director LPSC for constant encouragement in preparing this paper and kind permission to present this paper in National Seminar NDE 2009 at Tiruchirapalli Dec References 1. ASTM, Annual Book of ASTM standards section 3 Vol 3.03 Oct U Zscherpel and BAM Berlin May 2000 Vol 05, No05 3. P R Vaidya Flat panel detectors in Indusrrial radiography BARC Mumbai 4. Operating Manual Kodak LS85 X-Ray Film Digitizer 5. Operating Manaual Vidisco Fox Rayzor system