384 Proceedings of the National Seminar & Exhibition on Non-Destructive Evaluation NDE 2009, December 10-12, 2009 Axial Higher Order Modes Cluster (A-HOMC) Guided Wave for Pipe Inspection. Chandrasekaran Jayaraman, Krishnamurthy C.V, Krishnan Balasubramaniam, R.Sai Kishore and Issac Anto Centre for Nondestructive Evaluation and Department of Mechanical Engineering, Indian Institute of Technology-Madras, Chennai, India Abstract This paper reports the potential of Higher Order Modes Cluster Guided Waves (HOMC-GW) which are highly non-dispersive for axial testing in pipe over a couple of meters. The HOMC-GW is a recently explored phenomena which is found to occur at very high frequency-thickness product i.e. 15 MHz.mm to 35 MHz.mm. 2D ABAQUS finite element simulations were used to decide the optimal wedge angle which facilitates the generation of the HOMC-GW. Experiments were carried out on a 6 inch diameter mild steel pipe with machined pin hole type defects of various diameter and depth. A commercial 2.25 MHz, 1 inch diameter transducer mounted on a machined acrylic wedge was used. The potential of axially propagating HOMC-GW on pipe like structures is being brought into light. Keywords: Wave propagation, wave simulation, Guided Waves, Higher Order Modes Cluster, Non dispersive waves, Axial guided waves 1. Introduction In-service degradation due to mechanisms such as corrosion at pipe supports is one of the leading causes of process piping failure. The support regions in pipes are very vulnerable to accelerated corrosion as these are sites for water logging, stress and contamination. The use of guided wave modes is potentially a very attractive solution to the problem of inspecting the embedded portions of structures. The ultrasonic guided waves travel long distances, depending on the frequency and mode characteristics of the wave and follow the contour of the structure in which they are propagating. Applications to problems like corrosion monitoring, pipeline wall thinning inspection, weld defect detection and such industrial problems are well known [4,5]. The solution for a guided wave traveling axially through pipe like structures could be approximated to that of a plate [6], in some cases. Hence, for analysis purposes, the disperse solution for a plate was taken into consideration. Table 1, constructed specifically at 16 MHz-mm product for a mild steel plate presents the various modes and associated angles of excitation through a Plexiglas wedge.to identify the relevant guided waves for specific applications, their dispersion characteristics and excitation modes need to be examined. Three standard ways of representing the steady state solution for guided waves are shown in Fig. 1. This solution is for a mild steel plate with the wave incident through Plexiglas. Dispersive modes (indicated by steep slopes in Fig. 1) are complex to work with as it involves sophisticated signal processing techniques to handle them. Non-dispersive modes (indicated by flat portions in Fig. 1) are very useful in many practical NDE applications. Figure 1b shows the group velocity curves for these modes. When phase velocities are difficult to measure in dispersive regimes, group velocities are measured.in Fig. 3, the region marked 3, lying between 15 MHz-mm product to 35 MHz-mm product is the higher order modes cluster of guided wave (HOMC- GW) regime and forms the subject of the present investigation. Modes in this regime are found to have small differences between their group velocities and small differences in the associated angles of excitation. Table 1, constructed specifically at 16 MHz-mm product presents the various modes and associated angles of excitation through Perspex. As the various modes that take part in the cluster move with nearly the same group velocity, with the cluster appearing to move as a distinct non-dispersive envelope. 2. Simulation of HOMC-GW A series of 2D plane strain finite element simulations were carried out to understand the behavior of HOMC-GW. The operating frequency was fixed at 2.25 MHz and ABAQUS was used to model the wedge angle and the Mild steel sample thickness to cover the span of region 3 marked in Fig. 1c. The element used for simulation was CPE4R [9]. HOMC-GW behavior was found to have optimal features at a wedge angle of 54º. These features involved occurrences of a nearly non-dispersive envelope, negligible out-of-plane displacement component and relatively constant crosssectional displacement patterns. Significantly the HOMC- GW clusters generated had group velocity distinct from the group velocities of the cluster member modes. Table 1 lists the various modes, their group velocities and the angle of excitation of these modes through Plexiglas at 16 MHz-mm product. Simulations were carried out for each of the higher order mode wedge angles listed in Table 1.
NDE 2009, December 10-12,2009 385 Fig. 1 : Disperse plot for a Mild steel plate with Plexiglas incidence. (a) Phase velocity plot, (b) Group velocity plot, (c) Angle of incidence through Plexiglas. (1) Conventional guided wave regime. (2) Higher order modes guided wave regime. (3) Higher order modes cluster guided wave regime. Table 1 : Table shows the angle of incidence for the different modes at 16 MHz-mm along with the corresponding group and phase velocities for Mild steel plate. Sl.No Modes at Incidence Group Phase fd= 16 Mhz angle velocity velocity -mm Through m/s m/s Plexiglas in degrees 1 A1 56 3211 3287 2 S1 54 3111 3371 3 A2 51 2956 3516 4 S2 47 2754 3745 5 A0 65 3016 3016 6 S0 65 3016 3016 3. Experimenal verifications To experimentally verify the axial HOMC a 54 Plexiglas wedge was machined to match the pipes curvature. The experiments were conducted on 6 inch pipe Mild steel with wall thickness of 7.10 mm.the data presented in this paper was acquired using a conventional 2.25 MHz center frequency, 25.4 mm diameter transducer. A conventional ultrasonic system was used for measurements. This system is a standard ultrasonic pulser/receiver interfaced to the PC using a bus powered National Instruments NI 5133 based data acquisition card to acquire data using commercial PANAMETRICS (V104, 25.4 mm diameter). A custom developed application using standard tools from LABVIEW was used to acquire, store and analyze the data captured. To understand the axial HOMC better from inspection point of view, three configurations of experiments will be carried out. The first set of experiments was carried out to verify the velocity of the Axial HOMC formed. The second one was conducted to analyze the beam spread of the axial HOMC generated. The third set of experiment was conducted to verify the distance covered by the axial HOMC. 3.1 Experiment to verify the velocity of Axial HOMC Figure 2 shows the configuration used for testing the group velocity of the axial HOMC formed. A pulse echo experiment mode was used. The velocity computed was the averaged value of trials with respect to the end reflection from the pipes end. The wedge was moved successively further away from the free end at equal spaced distances and the results captured. The experimentally computed group velocity was 3133.12 m/s. Group velocity was calculated from the simulation data and was found to be 3135 m/s. This showed a good match
386 Chandrasekaran Jayaraman et al. : Proceedings of the National Seminar & Exhibition on Non-Destructive Evaluation Fig. 2 : Experimental configuration for velocity verification along with the comparison of group velocity values from experiment and simulation. between the simulated and the experimental values. It was observed that the velocity of the mode cluster was unique and did not match with any of the individual modes group velocity. 3.2 Experiment to determine the beam spread of the axial HOMC Experiments were conducted to verify the circumferential beam spread of the generated axial HOMC guided wave. The wedge with the transducer was kept in a fixed position and the signals recorded at many points along the curvature of the pipe along the excitation path of the wedge to find the beam spread of the axial HOMC. Figure 3 shows the configuration of the beam spread experiments with the points on which actual signal captured marked in red. From the experiments it was observed that the formed axial HOMC had most of its energy falling inside the wedge width and did not have a significant beam spread in the circumferential direction. Figure 4 shows the beam spread for a distance of 3 cm from the wedge to 23 cm from the wedge edge. The box marked covering the peak in Fig. 4 denotes the wedge foot print on the pipe. It was observed that most of the modes energy was circumferentially concentrated along the wedge foot print and the beam spread of the axial HOMC was relatively less when compared to the axial guided wave modes used in regime 1. Fig. 3 : Configuration of the beam spread experiment conducted
NDE 2009, December 10-12, 2009 387 Fig. 4 : Experimental results showing the beam spread of the Axial HOMC formed. B - wedge foot print on the pipe 4. Conclusions The phenomena of HOMC were found to be valid for axial guided waves in pipes. There has been no report in literature of the usage of axial HOMC for pipe inspection. Axial HOMC in pipes, similar to HOMC in plates occurs at high frequency-thickness (15 MHz-mm to 40 MHz-mm). The plate approximation during the modeling and analysis in order to design the wedge provided reasonably good prediction. The circumferential beam spread of the axial HOMC was found to be narrow and had significant existence only within the foot print of the wedge that was used to generate the wave. This observation is advantageous as the circumferential sizing and resolution characteristics axial HOMC offers will be better. It was demonstrated that the axial HOMC can travel up to a range of 2 m in a pulse echo mode. The better performance of a 54 wedge (Optimal HOMC) over a 56 (Sub-optimal HOMC) wedge was demonstrated through simulated A-scans. The simulated velocity value matched well with the experimental velocity value and it was observed that the velocity of the axial HOMC guided wave was unique and different from the group velocity of the individual mode members at the considered frequency-thickness (16 MHz-mm). Axial HOMC guided wave appears to have several attractive features for medium range NDE applications like corrosion detection welded support of pipes, corrosion detection in suspended pipe support, inaccessible regions in nuclear installations and other common application like lamp post corrosion monitoring. Similar to the HOMC in plates, the axial HOMC in pipes offer (i) tighter envelope that improves the temporal resolution (ii) shorter wavelength that improves the spatial resolution, (iii) The vanishing surface displacements of the out-of-plane component that is insensitive to surface loading, and (iv) sub-surface defect detectability. References 1. Jim Britton, Corrosion at pipe support, Causes and solutions, 2003 2. Meeker T R and Meitzler A H, Guided Wave propagation Elongated Cylinders and Plates, Physical acoustics, 1 Part A, 1964, pp. 111-167. 3. Zemmanek J Jr,, An Experimental and Theoretical Investigation of Elastic Wave Propagation in a Cylinder,\ The JASA, 52(1) (part 2) (1972) pp. 265-283 4. Wang W D, Applications of guided wave technique in the petrochemical industry, in Review of Progress in QNDE, 18A, op. cit. (1998) pp. 277-284. 5. USNRC, An Approach for Plant Specific, Risk- Informed Decision Making. In- service Inspection of Piping, U.S. Nuclear Regulatory Commission, Draft Regulatory Guide DG-1063 (1997).
388 Chandrasekaran Jayaraman et al. : Proceedings of the National Seminar & Exhibition on Non-Destructive Evaluation 6. Wei Luo, Xiaoliang Zhao and Rose J L, A Guided Wave Plate Experiment for a Pipe, Journal of Pressure Vessel Technology 2005 by ASME AUGUST 2005, Vol. 127 / 345. 7. Shivaraj K, Balasubramaniam K, Krishnamurthy C V and Wadhwan R, ASME Trans. J. Pressure Vessel Technology, (2007). 8. Satyarnarayan L, Chandrasekaran J, Bruce Maxfield and Krishnan Balasubramaniam, Circumferential higher order guided wave modes for the detection and sizing of cracks and pinholes in pipe support regions, NDT&E International, 41 (2008) 32 43. 9. ABAQUS 6.5.1. ABAQUS Analysis users manual, Section 22.2.3 2D solid element library. 10. Moser F, Application of finite element methods to study transient wave propagation in elastic wave guides. Diplomarbeit Thesis, Institute A of Mechanics, University of Stuttgart, 1997. 11. Friedrich Mosera L, Laurence J. Jacobsa, Jianmin Qub, Modeling elastic wave propagation in waveguides with the finite element method, NDT&E International, 32 (1999) 225 234.