National Seminar & Exhibition on Non-Destructive Evaluation, NDE 2014, Pune, December 4-6, 2014 (NDE-India 2014) Vol.20 No.6 (June 2015) - The e-journal of Nondestructive Testing - ISSN 1435-4934 www.ndt.net/?id=17822 Performance Evaluation of 3-Axis Scanner Automated For Industrial Gamma- Ray Computed Tomography Parag Walinjkar a, Lakshminaryana Y, B S Singh, R V Acharya, Umesh Kumar, and Ashutosh Dash Isotope Production & Applications Division, Bhabha Atomic Research Centre, Mumbai-85, INDIA E-mail: a pbw@barc.gov.in Abstract: The paper presents performance evaluation of automated 3-axis scanner consists of controller for mechanical manipulator and data acquisition unit for gamma-ray based Computed Tomography (CT). Data acquisition with synchronized mechanical motions has been achieved using an USB based controller. The system can scan cylindrical specimen up to 200 mm in diameter and 300 mm of height. A stainless steel prototype nuclear waste drum has been used as a specimen. The evaluation was carried out with 12mCi collimated source of Cs-137 and Φ 1x1 inch NaI(Tl) detector. Data was acquired in parallel geometry with 36 projections over 180 degrees. Each projection has 20 samples with data acquisition time of 1sec per sample. Attenuation map of the drum cross-section has been reconstructed with statistical optimization algorithm using information on geometrical parameters. The CT image obtained is in good agreement with actual specimen with specific regard to shape, size and linear attenuation coefficient within the circle of reconstruction. The results show satisfactory performance of the scanner. Keywords: Computed Tomography (CT), projection, Data Acquisition (DAQ) system, statistical optimization algorithm. Introduction Radioisotope and radiation applications in industry form an important part of using nuclear technology for development and societal benefits. The Isotope Production & Applications Division (IP&AD) of Bhabha Atomic Research Centre has made pioneering contributions to promote application of gamma sources for Non Destructive Testing and Examination (NDT&E) in industry. The Division has been active and instrumental in research and development activities pertaining to advance radiation based imaging for almost two decades. Industrial computed tomography (ICT) systems based on gamma rays and x-rays of first, second and third generation modalities have been developed and demonstrated earlier [1-3]. All the gamma-ray transmission tomography systems use either one or more collimated point sources and detector/s placed on the opposite side of an object under investigation [4]. These systems can be broadly classified into two types; one using single source with scanner and other using multiple sources without scanner. The major drawback with gamma-ray tomography systems is the relatively high cost, particularly for higher speed and higher resolution [4,5]. Hence for imaging of static objects or processes, the first generation systems using single source and a detector is preferred because of operational simplicity and lower cost. IP&AD has developed one such system using USB based ICT-controller based on narrow beam scanning technique [6]. The tomographic image generated is mainly depends on the attenuation of gamma ray through the specimen. Since the scanning becomes unavoidable part of the system the quality of image largely depends on the performance of the scanner. Therefore, performance of ICTscanner is experimentally evaluated using specimens to provide true information. Materials & Method The indigenously developed automated 3-axis scanner is a standalone system consists of a 3-axis mechanical manipulator, its controller, gamma detector and a PC as a data acquisition unit for
gamma-ray based Computed Tomography (CT). The block diagram of the ICT scanner is shown in the Fig. 1 below. Fig. 1: Block diagram of ICT scanner In-house developed control software for ICT-scanner facilitate user to input required scanning parameters i.e. angle of projection (Ɵ), step size (L) and dwell time as shown in Fig. 2 below. Dwell time is the time of data acquisition at each position. Based on the inputs, the total number of projections and samples are calculated and displayed. The specimen is moved in linear, vertical and angular direction to take respective positions and gamma attenuation data (counts) is automatically acquired once Start Acquisition button is pressed. The current value of sample and projection is updated at every position. The present status of the scanner is indicated by the respective indicator by turning it green. It automatically stores data of all positions required for generating tomographic image of an object as per parallel beam scanning technique [6]. Experimental Fig. 2: Snap shot of ICT-scanner software Performance evaluation of automated 3-axis ICT scanner was carried out using experimental data. The experimental arrangement to evaluate the performance of 3-axis ICT scanner is shown in Fig. 3 below. It is basically geometrical arrangement of the combination of gamma source, detector, a specimen (object under test) and the method adopted for acquiring the data for the requisite number of projections. The source detector arrangement is such that the detector is always in line with the beam of gamma source. It can scan cylindrical specimen up to 200 mm in diameter
and 300 mm of height. Here the specimen is placed on the 3-axis manipulator and moved keeping the source and detector at fixed position as shown in Fig. 3 below. Performance tests were carried out with a 12mCi, Cs-137 Collimated gamma source and NaI(Tl) detector of size Φ 1x1 inch. Detector γ Source Specimen Manipulator Fig. 3: The hardware design and actual photograph of ICT-scanner The ICT scanner initially tested for its mechanical operation with different sets of input parameters i.e. projection angle and step size. To find its overall performance for gamma ray tomography, the input parameters are optimized as projection angle of 5 degree, step length of 5mm and dwell time of 2sec. It set the scanning parameters as 36 numbers of projections, 40 samples per projection and 2 sec of data acquisition time per sample. Two types of preliminary experiments were performed with these scanning parameters. These experiments were carried out to analyze the operational performance of the ICT-scanner. Initially the scanning was done without any object between the source and detector i.e. for air. Then the experiments were carried out with actual objects. In first experiment a plastic bottle filled with water was scanned and in second experiment a stainless steel prototype drum with Aluminum rod and Perspex rod placed inside it was scanned with 10mm step size and 1sec dwell time. The reliability of automatic data acquisition methodology was tested by processing the acquired data, to generate a computer tomographic image of the drum cross-section, using image reconstruction algorithm known as iterative statistical optimization algorithm. This algorithm is based on the Poisson measurement statistics. Before applying the reconstruction algorithm, the projection data p is processed using Eq. 1 below I p = 0j j ln( ) I (1) j Where, I j and I 0j are the counts recorded by detector at position j of the acquired data taken with and without the object, respectively [7,8]. The reconstruction software has been developed in-house to meet the specific requirements. Result & Discussion The initial tests carried out for scanning operations shows that ICT scanner scans the specimen as per user inputs correctly. During air scan it is observed that the automatically collected data set matches with the calculated data set for the set input parameters. The data of air scan plotted as counts v/s sample position for few projections is shown in Fig. 4 below. It shows small variations within the acceptable limit of statistical error. The measurement accuracy calculated, as shown in the Table 1, from the air scan data found better than 97%. This ensures the overall performance of a gamma data acquisition system of ICT scanner is good.
Average counts Table 1: Measurement accuracy of scanner Minimum Maximum 3σ counts counts counts 8293 8088 8674 238 2.9% ACCURACY (error) Fig. 4: Counts plotted against sample position The experiments carried out with objects were with regard to position, shape, size and linear attenuation coefficient within the circle of reconstruction. The results from image reconstruction of first test specimen are shown in Fig. 5 below. The line profile clearly shows the attenuation of gamma ray in the object at specific projection angle. The thin plastic container of the bottle and water inside is clearly visible in the CT image as yellow colour ring around red colour that represents water. Fig. 5: Line profile and CT images of water bottle as test object The results from image reconstruction of second test specimen of a proto type steel drum containing Al and Perspex rod are shown in Fig. 6 below. It is observed that the drum boundary and two objects inside the drum are clearly visible in the CT image. It is also observed that the CT image shows different colour for both these objects indicating difference of density but the shape of the objects is not clearly visible. The square nature seen in the CT image is because of less number of samples per projection. The contour image was generated, for boundary, which clearly shows the circular shape of the objects inside the drum.
It is observed that in all the scanning took approximately 6 hrs to complete a full scan during the tests. Though this arrangement took longer scanning time than other techniques the minimal hardware and lowest cost motivates the use of this scanner. Fig. 6: CT image and contour image of the drum containing Al and Perspex rods Conclusion The CT image obtained is in good agreement with actual specimen with specific regard to shape, size and linear attenuation coefficient within the circle of reconstruction. Variations in shape and size of test specimens have been visualized. The results show performance of scanner is satisfactory. Further work is in progress to improve scanning methodology to reduce the scanning time. Acknowledgments Authors acknowledge Dr K L Ramakumar, Director, RC&IG, BARC and Dr A Dash, Head, IP&AD, BARC for constant encouragement and support during this work. References [1] Umesh Kumar, et al, "Development of an Industrial Isotopic Tomographic Imaging System CITIS-I", Report BARC/1993/E/019, 1993. [2] Umesh Kumar et al, "Prototype gamma-ray computed tomographic imaging system for industrial applications", INSIGHT, Vol. 42, No. 10, 2000. [3] Umesh Kumar et al, "Behaviour of reconstructed attenuation values with varying X-ray tube voltage in an experimental third-generation industrial CT system using linear detector array", Nucl. Inst. and Meth. in Physics Research-A, Vol. 490, Isuue 1-2, pp. 379-391, 2002. [4] G A Johansen, Nuclear tomography methods in industry, Nuclear Physics A 752, pp: 696c 705c, 2005. [5] IAEA-TECDOC 1589, Industrial Process Gamma Tomography, Final Report of a coordinated Research Project 2003-2007, Austria, IAEA, 2008. [6] Parag Walinjkar, et al, Development of Indigenous USB based ICT-Controller for Industrial Computed Tomography Scanner, proc. of National Symposium on Advances in Control & Instrumentation SACI-2014, 24-26 Nov., Mumbai, pp. 27-30, 2014. [7] Jongbum Kim, Sunghee Jung, Jinho Moon, Gyuseong Cho. Industrial gamma-ray tomographic scan method for large scale industrial plants, Nuclear Instruments and Methods in Physics Research A, 640: 139 150, 2011. [8] Z. Islami Rad, R. Gholipour Peyvandi, and R.Heshmati, Motion detection in CT images with a novel fast technique, Instruments and Experimental Techniques, Vol. 56, No.3, pp. 276-282, 2013.