Center For NDE Research Articles

TARGET – Development of Submersible ROV System for BMN Inspection

Most PWRs have penetrations in the RPV lower heads for in-core nuclear instrumentation. These penetrations generally are made of nickel-based Inconel Alloy 600. Weld materials are typically Alloy 82/182. Operating conditions of PWR plants are causing nickel-based alloys cracking through a process called primary water stress corrosion cracking (PWSCC). In 2003, the licensee for the South Texas Project Unit 1 (STP-1) identified apparent boron deposits on the lower RPV head near two bottom mounted nozzles (BMNs). The NRC issued Bulletin 2003-02 to obtain information on licensee inspection activities and inspection plans for the RPV lower head. EPRI issued MRP-206 report that provides inspection and evaluation guidelines for BMNs for PWR plants, including guidelines for periodic bare metal visual examination for evidence of primary coolant leakage, or periodic non-visual nondestructive examinations for indications of service-induced cracking. The non-visual inspections (ultrasonic testing examination) may detect service-induced degradation before through-wall cracking, leakage, circumferential cracking below the bottom of the J-groove weld, release of loose parts, or incipient boric acid wastage of the low-alloy steel reactor vessel lower head material occurs. Therefore, periodic examinations will adequately manage potential for cracking by PWSCC and preserve structural integrity. INETEC developed TARGET system for BMNs inspection, consisted of submersible ROV and specially designed probe, composed of several UT probes. UT system and technique to detect, length and depth size the service-induced degradation in the BMN volume material is developed. The EPRI NDE Center performed a technical review and validated INETEC’s ultrasonic examination technique for BMNs. Aforementioned validation was done according to requirements defined by: 1) MRP-206, 2) MRP-411.

Engineering studies for fluorescent penetrant inspection with a focus on developer application methods

Fluorescent penetrant inspection (FPI) is the most widely used global inspection method, playing a particularly important role in aviation. Given the contribution that reliable implementation of the FPI process can make to flight safety, the US Federal Aviation Administration (FAA) has funded a programme to assess the FPI parameters and their role in effective detection of typical flaws. Iowa State University's Center for NDE (CNDE) has led a team that includes many industry partners from the aviation industry during this six-year programme. The industry partners have provided guidance and prioritisation input, and in many cases access to internal data, samples or use of facilities to support the programme. The focus of the programme has been on quantitative assessment of FPI performance using a combination of data gathering methods, which has included indication luminance as measured using a photometer, UV-A indication appearance using a fluorescence stereomicroscope, and in some cases the more traditional probability of detection (POD) study. Numerous studies have been completed, although results of developer studies will be the focus of this paper. Choice of indication development parameters starts with the selection of the developer form. All four forms have been evaluated during this programme: dry powder, water-soluble, water-suspendible, and nonaqueous wet developers. The results of the developer studies are presented below along with key 'lessons learned'.

Development of a diagnostic expert system for eddy current data analysis using applied artificial intelligence methods

A diagnostic expert system that integrates database management methods, artificial neural networks, and decision-making using fuzzy logic has been developed for the automation of steam generator eddy current test (ECT) data analysis. The new system, known as EDDYAI, considers the following key issues: (1) digital eddy current test data calibration, compression, and representation; (2) development of robust neural networks with low probability of misclassification for flaw depth estimation; (3) flaw detection using fuzzy logic; (4) development of an expert system for database management, compilation of a trained neural network library, and a decision module; and (5) evaluation of the integrated approach using eddy current data. The implementation to field test data includes the selection of proper feature vectors for ECT data analysis, development of a methodology for large eddy current database management, artificial neural networks for flaw depth estimation, and a fuzzy logic decision algorithm for flaw detection. A large eddy current inspection database from the Electric Power Research Institute NDE Center is being utilized in this research towards the development of an expert system for steam generator tube diagnosis. The integration of ECT data pre-processing as part of the data management, fuzzy logic flaw detection technique, and tube defect parameter estimation using artificial neural networks are the fundamental contributions of this research.

A novel method for characterizing contact-mode shear wave transducers in immersion

The dominating shear motion of a contact-mode shear wave transducer is often accompanied by a weak parasitic longitudinal mode due to its finite size and constrained boundary conditions. The amplitude and phase of this out-of-plane vibration may be exploited for determining the shear polarization direction and for detecting any asymmetry or distortion of the shear transducer’s vibration pattern. In practice, the contact-mode shear probe is immersed in water and energized by an rf pulser to serve as the transmitter. A C-scan is then made of the face of the shear transducer using a focused longitudinal wave transducer as the receiver. The resulting image shows the out-of-plane longitudinal vibration of the shear transducer. For commercial contact-mode shear probes, this pattern usually consists of two crescent-shaped regions of higher vibration amplitude but of opposite phase. A line bisecting the two crescents and passing through their centers defines the shear polarization direction. Finite-element modeling of the shear displacement of a circular disk with constrained boundary confirmed the experimentally observed behavior. This approach provides a simple method for characterizing commercial shear wave transducers. [Work supported by Center for NDE, Iowa State University.]

Performance evaluation of the Golay code pulse compression technique

The Center for NDE, Iowa State University, has developed a laboratory prototype Golay code pulse compression system and tested it on a variety of materials. The performance of the system was evaluated in terms of signal-to-noise ratio enhancement (SNRE), resolution, and computation speed. The system's error sources also were discussed. The Golay code pulse compression was simulated on a computer and demonstrated the effective noise suppression. In addition, an equivalent pulse of the Golay code (delta-like pulse) was derived theoretically using a simple ultrasonic inspection model, which demonstrated its equivalence on the output correlated signal. Overall, the pulse compression technique extended the detection range for a given peak power and considerably reduced the system'swhite noise, hence providing enhanced signal-to-noise ratios (SNRs). An average of 30 dB improvement in SNR was obtained from highly energy-absorbent materials such as rubber, plastics, corks (insulation materials), and thick composites using the Golay codes of up to 512 bits. However, the technique did not effectively reduce coherent scattering noises from the coarse grain boundaries in cast stainless steels, Inconel weld metal, and material lay-ups in thin composites. Furthermore, it was found that, depending upon the system's hardware capabilities, the overall performance could be degraded considerably.

Performance Evaluation of the Golay Code Pulse Compression Technique

The Center for NDE, Iowa State University, has developed a laboratory prototype Golay code pulse compression system and tested it on a variety of materials. The performance of the system was evaluated in terms of signal-to-noise ratio enhancement (SNRE), resolution, and computation speed. The system's error sources also were discussed. The Golay code pulse compression was simulated on a computer and demonstrated the effective noise suppression. In addition, an equivalent pulse of the Golay code (delta-like pulse) was derived theoretically using a simple ultrasonic inspection model, which demonstrated its equivalence on the output correlated signal. Overall, the pulse compression technique extended the detection range for a given peak power and considerably reduced the system'swhite noise, hence providing enhanced signal-to-noise ratios (SNRs). An average of 30 dB improvement in SNR was obtained from highly energy-absorbent materials such as rubber, plastics, corks (insulation materials), and thick composites using the Golay codes of up to 512 bits. However, the technique did not effectively reduce coherent scattering noises from the coarse grain boundaries in cast stainless steels, Inconel weld metal, and material lay-ups in thin composites. Furthermore, it was found that, depending upon the system's hardware capabilities, the overall performance could be degraded considerably.

Interaction of shear wave polarization and composite laminate layup: Experiment and modeling

The interaction of shear wave polarization and the fiber directions in a composite laminate makes shear waves a sensitive probe for evaluating ply orientation and sequence in a layup. The transmitted shear wave signal, with the transmitting and receiving transducers perpendicular to each other, can easily distinguish a [0/+45/90/−45]6 from a [0/+45/90/−45]3S, and detect an error in the form of a single misoriented ply: the 12th ply in [0/90/+45/−45]3S mistakenly placed at +45° instead of at −45°. Shear waves can therefore detect subtle but realistic layup anomalies in composite laminates. In this work, a model is developed for the propagation of shear waves through a laminate with arbitary ply orientation and sequence. Experimental results and model predictions are compared, and the model is used for identifying layup anomalies that are detectable with shear waves. Potential applications of this new method in nondestructive evaluation of composite laminates will be discussed. [Work supported by NSF I/U Center for NDE.]

Simultaneous measurement of ultrasonic velocity and sample thickness and its application to NDE of composites.

Ultrasonic velocity is widely used for nondestructive evaluation of material properties. In cases where the thickness of the test piece is not precisely known or varies spatially, it is desirable to be able to determine the ultrasonic velocity and the sample thickness simultaneously. In immersion tests, the velocity and the thickness may be determined simultaneously by measuring a number of times-of-flight [cf. Kuo etal., J. Acoust. Soc. Am. 88, 1679–1682 (1990)]. In this work the ‘‘simultaneous’’ method was tested in four different configurations of pulse echo and through transmission and was applied to the characterization of porosity in graphite polyimide and graphite–epoxy composites. In addition, the velocity-thickness method was also implemented with water squirters. [Work supported by Center for NDE, Iowa State University.]

Radiated fields determined from moments of surface velocity

A quick and efficient method is derived for calculating the field radiated by a finite but otherwise arbitrary velocity profile. The method calculates the pressure field radiated into the half‐space (z > 0) from the normal component of velocity on the boundary plane (z = 0). The derivation starts from the Rayleigh integral and utilizes a power series expansion of the kernel in the radial coordinate of the source point. As a result, a formulation is obtained that contains separate terms for the source and for propagation. The source terms represent the moments of the velocity profile. The propagation terms can be obtained from a recurrence relation starting from certain special cases for which analytical expressions are known. Convergence and stability of the method will be discussed, and examples of field calculations for a few simple profiles will be given. [Work supported by Center for NDE at Iowa State University.]

Surface motion reconstruction based on half‐space field data

A finite but otherwise arbitrary normal velocity profile at the boundary of a half‐space is expanded in a set of orthonormal functions. The radiation problem is then formulated in terms of the moments of this profile. (Details of this formulation are given in the preceding abstract.) The radiation problem is discretized to obtain a matrix formulation and a singular value decomposition method is used to estimate the velocity profile from measurements of the radiated field. Additional constraints are imposed upon the derivatives of the profile in order to overcome the ill‐posed nature of the inverse source problem. The accuracy of the method will be discussed in terms of both the positioning of the field locations (nearfield versus farfield) and the weighting of constraints. Examples of velocity profile reconstruction will be given for a few simple radiators. [Work supported by Center for NDE at Iowa State University.]

The role of higher‐order acoustic nonlinearities in NDE

The acoustoelastic method is a well‐established way to obtain third‐order elastic constants in bulk materials. Modifications are needed to extend this technique to obtain higher orders, e.g., fourth‐order elastic constants, or to increase accuracy to adopt the technique for very thin layers, e.g., adhesive bonds. These higher‐order nonlinearities are separated from the stronger first‐order nonlinearities (associated with the third‐order elastic constants) by the appropriate choice of polarization and harmonic analysis of dynamic acoustoelastic results. It has been shown that inclusion of the fourth‐order elastic constants will lead to strength‐related mechanical properties of the adhesive bond. [This work was supported by the Center for NDE operated by the Ames Laboratory, USDOE for the Air Force Aeronautical Laboratories/Materials Laboratory under Contract No. SC‐89‐168 with Iowa State University.]

Fatigue Fracture Mechanics Modeling and Structural Integrity Assessment of Offshore Welded Tubular Joints

Evaluation of the structural integrity of offshore structures requires information on the reliability of nondestructive testing, the accuracy of fatigue crack growth modeling and other data. The University College London Underwater NDE Centre has been set up to provide information on the effectiveness and reliability of different nondestructive testing methods. To achieve this aim, a large library of cracked specimens will be assembled. In the preliminary phase of producing this library, a series of large-scale welded tubular joints were fatigue tested and the crack growth was fully monitored with the ACPD technique. This paper will describe briefly the background to the crack library and present the data obtained from fatigue tests. It will also describe a new model for fatigue crack growth prediction in tubular joints using fracture mechanics. This model allows the prediction of the size effect noted previously in the stress/life curves for tubular joints.

Scatter in ultrasonic wave attenuation measurements caused by transducers

The scatter in attenuation of ultrasonic waves in aluminum was investigated with five broadband, nonfocused, immersion transducers with center frequencies of 10 and 15 MHz. Three different experimental techniques were used: two types of multiple echo techniques and a multiple thickness technique. The objective was to investigate the repeatability of the measurement for the same transducer and same method, the bias towards a particular method, and the consistency of the measurements across transducers. No bias toward a particular method was found and repeatability was good. However, the results were found to be transducer dependent. Efforts have currently begun to image the transducers in order to perform their individual characterization. This characterization will be used instead of modeling transducers as rigid pistons. The effect of transducer characterization on the consistency in attenuation measurement will be discussed. [Work supported by Center for NDE, Ames, IA.]

Tennessee valley authority experience with field application of advanced UT system

An automated four-channel ultrasonic inspection system, the IntraSpect/98 was developed by Amdata Inc. and TVA engineering with TVA funding. Procedures for field application are developed and evaluated by the joint effort between TVA and the EPRI NDE Center. This paper describes the successful application of the IntraSpect/98 to the detection and sizing of IGSCC in stainless piping, examination of weld overlay, field examination of creviced safe-ends of reactor pressure vessel and measurement of wall thinning of fossil plant steam piping. The system was also applied to technique development for generator retaining ring and cast stainless steel pipe examination. The application of IntraSpect will be expanded to heavy section examination.

Underwater NDE centre is opened in London

Underwater NDE centre is opened in London

Transducer modeling using axial profile fitting techniques

Axial profiles of three piezoelectric, nonfocused, immersion pulse-echo transducers were measured using a small spherical reflector and an automatic scanning device. The scanning distance range included the first and second maxima. Measured data were compared with rigid disk radiation theory. The frequency-dependent active radius was used as a parameter in the curve fitting. In addition, three different approaches were used in the curve fitting. These included (1) fitting the entire range of the scanning distances, (2) fitting the first maximum, and (3) fitting the second maximum. The results of fitting the first maximum agree extremely well with the entire curve fitting, while the results obtained from the second maximum contain more errors. The results of modeling the transducers will be discussed in terms of the active radii and spatial sensitivity functions. In general, good agreement between the theoretical and experimental results was obtained. [Work supported by Center for NDE, Ames, Iowa.]

Weld crack characterization on offshore structures using AC potential difference and ultrasonics

To assure the integrity of offshore structures greater reliability and accuracy in the detection and sizing of surface-breaking defects in fatigue-affected welded nodes is being demanded. Accurate defect measurements can be used with fracture mechanics based models to assess fatigue crack growth rate and the remaining life of the joint. This paper describes the application of the AC potential difference method to the inspection of offshore structures and also discusses improved ultrasonic techniques developed for this inspection problem. A ‘library’ of tubular welded joints of typical geometries and containing fatigue damage is being collected at the new Underwater NDE Centre established at University College London. Initial trails using some of these specimens to assess detection and crack sizing capability are discussed.