Inspection Geometry Research Articles

MCNP calculations for container inspection with tagged neutrons

We are developing an innovative tagged neutrons inspection system (TNIS) for cargo containers: the system will allow us to assay the chemical composition of suspect objects, previously identified by a standard X-ray radiography. The operation of the system is extensively being simulated by using the MCNP Monte Carlo code to study different inspection geometries, cargo loads and hidden threat materials. Preliminary simulations evaluating the signal and the signal over background ratio expected as a function of the system parameters are presented. The results for a selection of cases are briefly discussed and demonstrate that the system can operate successfully in different filling conditions.

Detection and Sizing of Defects in Control Rod Drive Mechanism Penetrations Using Eddy Current and Ultrasonics

Over the last two years, concern has been generated about the capabilities of performing nondestructive evaluation (NDE) of the closure-head penetrations in nuclear-reactor pressure vessels. These penetrations are primarily for instrumentation and control rod drive mechanisms (CRDMs) and are usually thick-walled Inconel tubes, which are shrink-fttted into the steel closure head. The penetrations are then welded between the outside surface of the penetration and the inside surface of the closure head. Stress corrosion cracks initiating at the inner surface of the penetration have been reported at several plants. Through-wall cracks in the CRDM penetration or CRDM weld could lead to loss of coolant in the reactor vessel. The CRDM penetration presents a complex inspection geometry for conventional NDE techniques. A thermal sleeve, through which pass the mechanical linkages for operating the control rods, is inserted into the penetration in such a way that only a small annulus (nominally 3 mm) exists between the thermal sleeve and inside surface of the penetration. Ultrasonic (UT) and eddy current testing (ET) techniques that could be used to provide defect detection and sizing capability were investigated. NDE techniques that could be applied inside and outside the annulus were investigated, but the major goal was to provide high defect-detection sensitivity without requiring removal of the thermal sleeve. As a result of this investigation, both ET and UT techniques for the CRDM penetrations have been developed and evaluated. Long, thin probes were designed to fit into the annulus to carry both eddy current coils and irrigated ultrasonic transducers into the region of interest. The eddy current probes were used primarily to detect cracks in the penetration while the ultrasonic transducers were used to provide an estimate of the remaining wall thickness. This paper describes the ET and UT techniques, the probes developed, and the results obtained using these probes and techniques on CRDM penetration mock-ups.

Surface roughness and the ultrasonic detection of subsurface scatterers

The influence of surface roughness on the detection of subsurface scatterers using phase-coherent, ultrasonic immersion transducers is reported. Experiments were conducted with phase-coherent transducers, immersed in a liquid bath, that insonified samples with roughened surfaces. Measurements were made of: (1) the signal-to-noise ratio for a subsurface void when viewed through a rough surface; (2) the frequency-dependent loss (between 2 and 20 MHz) that was induced in the transmission and reflection coefficients of the coherent beam for flat aluminum plates with root-mean-square roughness varying between 5 and 50 μm; and (3) the effects of surface roughness on the material noise. The following results were found. First, surface roughness greatly degraded the signal-to-noise ratio (S/N) for some typical inspection geometries. Second, after appropriate normalization, the transmission and reflection coefficients for the coherent beam were found to be nearly universal functions of the angle of incidence, except near the critical angles. Large reductions (several orders of magnitude) in the phase-coherent signal S resulted for some typical inspection setups. The noise, on the other hand, remained nearly unchanged for the same setups. The degradation in S/N was attributed to the randomization of the phase of the coherently transmitted and reflected beams. Both the sensitivity of the signal and the relative insensitivity of the backscatter noise are explained in terms of the phase-screen approximation. This simple approximation is also used to explain the nearly universal form of the normalized transmission and reflection coefficients, and to predict that the backscatter noise may be either slightly enhanced or slightly decreased by surface roughness.

Model for the ultrasonic inspection of rough defects

A model for studying the effects of defect surface roughness on ultrasonic signals is described. The model contains many aspects of a real inspection system, giving predictions of interest to practical ultrasonic non-destructive testing. Acoustic Kirchhoff theory is used in a full time-dependent calculation, to give A, B or C scan information for arbitrary inspection geometries. This paper presents comparisons of the model predictions and experimental results. Comparison with time-harmonic pulse-echo measurements from a corrugated surface shows agreement to within 2 dB for average scan amplitudes. Comparison is also made with time-dependent pulse-echo measurement of the variation of average signal amplitude with angle of incidence, for surfaces ranging from smooth to very rough. Agreement between the experimental results and the model predictions is typically to within 3 dB, except for very rough surfaces. Model predictions are then presented for 45° tandem inspection and 60° pulse-echo inspection of defects of various orientations. The model shows how roughness decreases the detectability of well-oriented defects but enhances the detectability of defects which are mis-oriented by more than ≈ 20°. The effects of roughness on sizing techniques are also briefly discussed. It is shown that amplitude measurements for the dB drop technique may require careful interpretation when surface roughness is present. Time-based sizing techniques may also be affected, since roughness can lead to the loss of distinct diffracted signals from which to make timing measurements.