Page Areas:



Current Submenu:

Additional Information:

Innovative Metrology 2017

Rückblick: Innovation Messtechnik 2015

Where to find us

Plan of our location where to find us.

HF-Building, Northern Wing, Ground-Floor ...  more of Where to find us (Titel)


Position Indication:

Content

Measurements of the magnetic field with a high spatial resolution via magnetoresistive sensors

Dipl.-Ing. Patrick Hölzl

In the last few years the number of non-destructive testing (NDT) applications, which are based on solving an inverse problem has increased significantly. Whenever one wants to draw conclusions from an observed physical quantity on a not directly observable, but underlying parameter, an inverse problem must be solved. This approach is essential for NDT applications.

Magnetic imaging is a NDT method which leads to an inverse problem if the underlying current density should be determined. The current density itself is a meaningful parameter, which either allows to determine the geometry or the conductivity of an electrical conductor.

To recognize tiny structures (in the range of a few µm), the magnetic field must be measured with a high spatial resolution. Usually Superconducting Quantum Interference Devices (SQUIDs) are used in this context. SQUIDs are able to measure very small deviations of the magnetic field (in the range of a few atto Tesla) with a high spatial resolution (a few µm). Due to the required refrigeration ( in the past liquid helium was used , today liquid nitrogen) the handling of SQUIDs is expensive.

Much cheaper and easier to use are magnetic field sensors based on the Giant Magneto Resistance (GMR) or the Tunnel Magneto Resistance (TMR) effect, which are also used in conventional hard disk drives. A major producer of magnetoresistive sensors for industrial applications is the company NVE. Adversely these state of the art sensors possess a rather poor spatial resolution (in the mm range); respectively the magnetic field cannot be directly determined by a measurement, as will be shown in the following example.

Figure 1: Left: Photography of the stripboard used as a test specimen to evaluate the spatial resolution of magnetoresistive sensors. Center: FEM simulation of the current density distribution inside the copper track. Right: Component of the magnetic field at different locations.

In Figure 1 the test specimen used to evaluate the spatial resolution of magnetoresistive sensors is shown. Due to the mounting holes a deflection of the current density occurs in front of and behind a hole, which leads to an increase of the magnetic field. Additionally a 0.3mm long crack was implemented into the track, to simulate a defect. The aim is to recognize the crack by measuring the magnetic field with a NVE AB001-02 gradiometer. The gradiometer measures the difference of the magnetic field at two measuring points in 0.5mm distance.

Figure 2: Left: Gradient of the field component Hy measured with the NVE AB001-02. Right: Spatial sensitivity of the used gradiometer, which was determined via a calibration measurement.

The measured gradient of the field component Hy in a height of 520µm above the stripboard is shown in Figure 2 on the left side. Based on the spatial sensitivity (also known as point spread function (PSF)) of the sensor, the original field component Hy can be determined from the measurement data via a deconvolution, the result is shown in Figure 3 on the left side. In both images the implemented crack is clearly visible. Due to the deconvolution the spatial resolution of the sensor can be increased up to a few µm, with sensitivity in the nano Tesla range.

Figure 3: Left: Original field component Hy calculated with the measurement data in Fig. 2. Right: Current density component Jx inside the copper track determined from the field component Hy.

 

In Figure 3 the distribution of the current density component Jx is shown on the right side. The current density can be obtained from the calculated field component Hy also via a deconvolution.

Keywords: non-destructive testing, magnetic imaging, GMR und TMR sensors, magnetometer, gradiometer, deconvolution, inverse problems