Poster: Demonstrations of a Fast 4D TEM System for UXO Characterization

Panel 1: NanoTEM - A System for Fast Multi-Channel TEM Data Acquisition

What is "Fast" TEM?

TEM systems measure a voltage related to the time rate of change of the secondary magnetic field at one or more time-gates after the transmitter current has been shut off. The term Fast TEM is a comparative term suggesting that time gates for transient measurements are earlier (hence faster) than those in conventional TEM measurements. The Zonge NanoTEM system can measure transients at times as early as 1 usec after cessation of the transmitter current, much earlier than competing systems. But the capability to measure transients at early times is only part of a fast TEM system. It is also necessary to generate secondary transients containing information at early times.

Fast Transmitter Shut-Off - We illustrate an ideal NanoTEM transmitter waveform in a figure on the right. The top figure on the far right is an oscilloscope trace of the current waveform for one of our NT-32 NanoTEM transmitters driving a 1-m (8T) loop. The current turn-off approximates a linear decay of current with time (blue curve). The time derivative of the magnetic induction field B (primary field) drives electromagnetic induction. The bottom figure shows the current waveform (lower trace) at a greatly expanded time scale (5 us/div). Also shown in the figure is the voltage output from one of the receiver coils (upper trace). The NanoTEM system for this particular antenna set measures the transient starting at approximately 11.8 usec after the initiation of current turn-off.

High-Speed Data Acquisition - The TEM receiver requires a bandwidth corresponding to that of the secondary transients generated by the transmitter. The NanoTEM transmitter generates transients with a bandwidth of several hundred KHz. The NanoTEM receiver consists of 3 high-speed digital acquisition channels each with a bandwidth of 400 KHz (800 kSamples/sec).

Panel 2: Demonstrations of a Fast 4D TEM System for UXO Characterization

Static Mode Interpretation

Static mode data are acquired with a static antenna array that is placed at a number of discrete locations (i.e., "9-Spots" at Blossom Point). The antenna position and attitude are carefully controlled and the data at each point have low noise because of our ability to stack many transients and thereby substantially reducing random noise. The nearby panel illustrates a) the field setup, b) the resulting data patch (9x3 transient components), and c) the resulting model-based polarizability plot showing the principal polarizability transients and their parametric representation. This example is for a Mk23 practice bomb (letter 'M' in the photo). The resulting 3 polarizability transients have characteristics typical of a UXO-like object:
1. The largest polarizability is along the axis of symmetry (U-axis)
2. The two minor polarizability transients are similar in shape and magnitude.
3. The slope of the power-law decay parameterization (pk) is greater than 0.7 suggesting a magnetically permeable body.

Dynamic Mode Interpretation

Data acquisition in the dynamic mode is significantly faster. However, because the antenna cart is continuously moving at a rate of 30-60m/min, the quality of the transients at each data point is not as good as that acquired in the static mode. None-the-less, when the signal level is high (as with relatively large and/or shallow targets), the dipole model interpretation methodology produces results that are remarkably consistent with those generated using static data. The figure sequence above again illustrates a) the field setup, b) a data patch showing the 3-component data (mid time composite gate), and c) the resulting model-based decomposition. Note the similarity in the polarizability transients between the statically acquired data and the dynamically acquired data.

DNT_Dipole (Model-Based Interpretation)

DNT_Dipole fits 3-component transient data in the vicinity of each of a list of targets designated within Oasis Montaj to a point dipole model having an anisotropic polarizability tensor. The geometry and mathematics involved are illustrated below. The model is widely used within the UXO community, providing estimates of the target polarizability along 3 principal axes. These polarizabilities are proportional to the TEM transients that one would measure if the target response were oriented so that its principal axes are in turn parallel with a uniform polarizing magnetic field. DNT_Dipole estimates the 3 principal polarizability transients (dpi/dt; i ={u,v,w}) using the multi-component transient data acquired by the NanoTEM system. DNT_Dipole also estimates the target position, and attitude angles (j=heading, q=pitch, and d=roll). The figures above illustrate DNT data flow from acquisition (LEFT), to processed local data patch (CENTER), and finally to the interpretation parameters (RIGHT). The figure sequences illustrate static acquisition (UPPER) and dynamic acquisition (LOWER). For the MK23 target at Blossom Point, both acquisition methods yield remarkably similar results.

Panel 3:
NRL Baseline Ordnance Classification Test Site, Blossom Point
Standardized UXO Technology Demonstration Site, Aberdeen Test Center