3D Computed Tomography (CT) Its Application to Aerospace Industry C. Muralidhar, M. P. Subramanian, V. Ravi Shankar and G. Chandrasekhar Directorate of Non Destructive Evaluation, Defence Research & Development Laboratory 497 Non-Destructive Evaluation 2016
Kanchanbagh, Hyderabad 500 058, India E-mail: dr_c_muralidhar@rediffmail.com, mp.subramanian@gmail.com, ravishankar.vunnam@gmail.com, golichandrasekhar@gmail.com. Abstract Computed Tomography (CT) generates a thin cross-sectional image (slice) of an object. CT images are free from overlying and underlying areas of the object and are highly sensitive to small density differences (<1%) between structures. CT systems developed with Linear Detector Array (LDA) generate image slices of the test object and its 3D volume created by stacking the slices one over the other, whereas CT system developed with Flat Panel Detector (FPD) can directly generate 3D volume from 2D X-ray projections using the Cone Beam Reconstruction Algorithm. We have developed 3D CT system with 450 kv X-ray source, Flat Panel detector array, Mechanical manipulator and cone beam reconstruction algorithm to handle objects of 1000 mm diameter and 2000 kg weight for first time in India. The paper discusses the capabilities of 3D Computed Tomography system with FPD for NDE of Aerospace components such as End Dish. An End dish is made up of an aluminum alloy, bonded inside with thermal insulation (liner) and fitted to rocket motor. The End dish is provided with boss at two locations. Two holes are drilled through thermal insulation and metal casing into the boss. This hole that passes through thermal insulation and metal casing should maintain single axis after bonding. Any deviation from the hole axis and lack of bond integrity between casing and liner results in abnormal performance of rocket motor during its functional test. These holes drilled in oblique plane are difficult to inspect by Metrology and also difficult to measure deviations in hole axis. In such scenario, 3D CT is employed to obtain 3D image of End dish to verify bond integrity and also to measure deviations in hole axis. 1. Introduction Non Destructive Evaluation (NDE) deals with the evaluation of structural integrity of hardware without affecting its functionality and useful lifetime. In general X-ray Radiography is extensively used for evaluation of defects and internal details of an object under test, but Radiography compresses 3D information of an object into a 2D image and collinear defects are difficult to find. Computed Tomography (CT) generates a thin cross-sectional (slice) image of an object and the image represents point-by-point distribution of linear attenuation coefficients of the object. CT images are free from overlying and underlying areas of the object and are highly sensitive to small density differences (<1%) between structures [1]. X-ray detectors used in Computed Tomography systems can be either Linear Detector Array (LDA) or 2D Flat Panel Detector (FPD). CT system with LDA generates cross-sectional images of the test object that can be stacked one over the other to create 3D volume of the object. On the other hand CT system with FPD can directly generate 3D volume from 2D X-ray projections using the Cone Beam Reconstruction Algorithm. This 3D volume can be sliced in arbitrary planes to get the extent of internal features and measurements in 3 dimensions. W. C. Scarfe et.al. [2] explains what cone beam CT is and how it works and J. Noel et. al. [3] points out the advantages of CT in 3D scanning of Industrial Parts. The paper discusses the advantages of using 3D Computed Tomography (CT) with FPD that has been developed with 450 kv X-ray system, Flat Panel detector, Mechanical manipulator and cone beam reconstruction algorithm to handle objects of 1000 mm diameter and 2000 kg weight for first time in India. 3D CT has been employed on End dish, a mission critical Aerospace component of complex shape and size. An End dish is made up of an aluminum alloy, bonded on one side with thermal insulation, fitted to rocket motor for single shot application. The End dish is provided with boss (shown in fig 2) at two locations through which two holes (one hole at each boss) are drilled at an oblique angle initially into the metal casing and later into the liner bonded on other side of the casing to facilitate for testing and confirm its design 498 Non-Destructive Evaluation 2016
adequacy. These holes drilled in oblique plane are difficult to inspect by Metrology and also difficult to measure any deviation from hole axis. 3D CT results have been discussed in terms of reduction in scanning time, features viewed, defects identified, measured and visualized their extent in 3 dimensions, as compared to 3D CT image generated through slices using LDA. 2. Experimental Procedure DRDL has indigenously developed X-ray 3D Computed Tomography (CT) system first of its kind using 450 kv X-ray source, 2D Flat Panel Detector (FPD), a 6-axes mechanical object manipulator with cone beam Feldkamp reconstruction algorithm to handle objects of 1000 mm diameter and 2000 kg weight. The resolution of CT system is 300 µm. An End dish made up of an aluminum alloy, bonded inside with thermal insulation was scanned, analyzed for holes drilled at an oblique angle first into metal casing subsequently into liner later to observe deviation from hole axis and for debonds between casing and liner. 3. Results and Discussion Fig. 1 3D CT system with FPD Fig. 2. End dish Fig. 1 shows X-ray 3D CT system developed (with an X ray source on right side and FPD on the left side). Fig. 2 shows an End dish. Figs. 3(a) show plane showing hole from metal to the liner, (b) top view, (c) side view, (d) 3D model of the End dish. It is clear from figs. 3(a, b, c, d) that the details of bond integrity between casing & liner and any deviation from hole axis (hole axis skewness) are very clearly seen from CT data. The time taken for generating the full 3D model of the End dish is about 45 minutes. This is very faster compared to 3D model generated by stacking the 2D slices generated using Linear Detector Array (LDA). The 3D model can be cut at an angle to view the internal details and to get dimensions including inaccessible areas. 499 Non-Destructive Evaluation 2016
Fig. 3. 3D model and different sectional views (a) plane showing hole from metal to the liner (b) top view (c) side view (d) 3D model 500 Non-Destructive Evaluation 2016
Fig. 4. CT image reveals bonded and Debond regions. Density profiles shows sharp fall at the debond region Fig. 5. Density profile for Debond region Fig. 6. Density profile for good region Fig. 4 shows Density profiles for bonded and debond regions. A debond is observed between casing & liner and the density profile plotted shows sharp fall in density value at the debond location corresponding to the air gap between casing & liner. Whereas the density profile for good bond region (fig. 4 right bottom) shows the smooth transition in density from metal to liner. The application of density profile in differentiating debond region from good region is illustrated with another example in fig. 5 & fig. 6. Bonded regions do not have any sharp fall in density as explained earlier. It is also possible to obtain the extent of debond in 3 dimensions which is very essential in defining the acceptance criteria of the End dish. The ability to visualize the internal features along with their dimensions of End dish suggests 3D CT a superior method as compared to other NDE methods. The bond integrity between casing and liner could be thoroughly evaluated beyond doubt which may otherwise results in abnormal performance of rocket motor during its functional test. Fig. 3(a) shows the holes drilled through thermal insulation and metal casing into the boss. This hole passing through thermal insulation and metal casing should maintain a single axis. As these holes are drilled in oblique plane, it is difficult to be inspected by metrology methods and to measure the deviation in the hole axis. Any deviation in the hole axis could be clearly seen by taking sectional view of the 3D model along the oblique plane in which the hole is drilled. This type of information is not possible to obtain through non-contact and nondestructively from conventional NDE methods. It is also possible to get dimensions of internal features including inaccessible areas, which is a unique feature of 3D CT. It is difficult to obtain this kind of information from X-ray Radiography because the bond integrity between casing & liner of a complex shaped article such as an end dish will get overlapped as radiography compresses 3D information into 2D. The contour of the dish is not amenable to orient the dish in a direction to show the interface between metal and liner in the radiograph. On the other hand, Ultrasonic Dry coupling through transmission can be used only at 501 Non-Destructive Evaluation 2016
few locations wherever the rubber tipped probes match the contour of the dish. 100% inspection is not possible with UT and do not obtain any information regarding hole axis skewness. It is clearly demonstrated from the above that 3D CT being non-contact method and free from overlying & underlying areas of End dish, is a powerful NDE method in evaluating both bond integrity and hole axis skewness as compared to RT &UT. 4. Conclusion 3D CT system is successfully used to verify the bond integrity between casing and liner as well as to ensure the minimum wall thickness of hole and hole axis skewness. Thus 3D CT is a powerful inspection tool in evaluating Aerospace mission critical component having complex shape such as an End dish. 5. References 1. Michael J. Dennis, in ASM Handbook Vol. 17, įnon-destructive Evaluation and Quality ωontrol, ASM International (1λλ7) 7η4-811. 2. William C. Scarfe, Allan G. Farman, What is Cone-Beam CT and How Does it Work?, The Dental Clinics of North America, 2008 Elsevier Inc. 3. Julien Noel, North Star Imaging Inc, Advantages of CT in 3D Scanning of Industrial Parts, 3D Scanning Technologies Magazine, 2008 502 Non-Destructive Evaluation 2016