درباره بهره برداری از کوپلینگ حرارتی برای شناسایی اجزاء بیضوی در فلزات

A comparison between reported analytical results with experimental data of the magnetic flux density on cylindrical tin inclusions of elliptical cross-section embedded in a copper matrix under external thermal excitation is presented. By changing the aspect ratios b/a designated by e of the elliptical inclusions, a wide range of real situation such as slender inclusions can be simulated. The aspect ratio of the elliptical cylindrical inclusions varied from 0.50 to 3.250. A fairly modest 2.3 °C/cm temperature gradient in the specimen produced magnetic flux densities ranging from 2 to 100 μT at 2 mm lift-off distance between the tip of the magnetometer probe and the specimen. The experimental magnetic field distribution illustrated the potential for the non contacting thermoelectric technique to detect and characterize metallic inclusions of different geometries based on their magnetic signature. Preliminary results on a cylindrical hard alpha (TiN) inclusion embedded in Ti–6Al–4V matrix is also presented to demonstrate that the proposed method might be applicable to a wide range of alloys including high-strength, high-temperature engine materials.

Most metallic materials contain inclusions which can be either metallic or non-metallic. Inherent to elaboration process, they are distributed inside the materials. These inclusions have generally a higher melting point than the host metals. Such inclusions in alloys reduce mechanical properties, are detrimental to surface finish and increase porosity, as well as having a tendency to increase corrosion. Furthermore, they act as stress raisers and can cause premature failure of in-service components. There are many established methods of traditional NDE that are employed today to analyze the quality of new and in-service materials. The detection of inclusions is their prior task. From all of these, one can retain ultrasound, X-ray, eddy current or electrostatic conductivity [1]. Each of them has advantages and disadvantages. Successful results have been obtained using the previous mentioned techniques for non-metallic inclusions. However, metallic inclusions are a little bit more complex to detect due to the very similar physical properties they could have with the host metal. In this work, we consider an alternative detection technique, which potentially has the possibility to detect surface and subsurface metallic inclusions, i.e. the non-contacting thermoelectric power measurements. It has been demonstrated that the thermoelectric coupling in metallic materials can be exploited as a viable means of characterizing all types of imperfections and material defects such as inclusions, inhomogeneity, residual stresses, texture, fretting and segregations [2], [3], [4], [5], [6] and [7]. This nondestructive detection is carried out in an entirely non-contact way by using various types of magnetometers to sense the weak thermoelectric currents around the affected region when the specimen is subjected to an external temperature gradient. A schematic diagram of the thermoelectric measurement process in the presence of material imperfections is shown in Fig. 1. For this case, an external heating or cooling is applied to the specimen to produce a temperature gradient in the region to be tested. This creates a situation in which different points of the boundary between the host material and the imperfection are at different temperatures, therefore also at different thermoelectric potentials. This will produce opposite thermoelectric currents inside and outside the imperfection. The thermoelectric currents form local loops that run in opposite directions on the opposite sides of the imperfection relative to the prevailing heat flux. When the specimen is scanned with a highly sensitive magnetometer, the magnetic field of these thermoelectric currents can be detected even when the imperfection is buried below the surface few milimeters and the sensor is as far as a couple of centimeters from the specimen [8] and [9]. Full-size image (14 K) Fig. 1. Schematic diagram of noncontacting thermoelectric detection of material imperfections by magnetic sensing. Figure options Several authors have developed analytical models to predict the magnetic field produced by thermoelectric currents around inclusions of specific geometries such as a cylinder and a sphere respectively in a homogeneous host material under external thermal excitation [10] and [11]. In subsequent publications, some of these theoretical models were verified by experimental results [5] and [8]. This research work have provided a better understanding of the phenomenon and helped lay down the foundations for the development of this emerging NDE technique. However, real applications involve more complicated-geometry inclusions. Achieving this ambitious goal will require closely related theoretical and experimental efforts, the development of new, predictive analytical models, more sensitive experimental procedures, and, ultimately, increased probability of detection for small inclusions and weak material imperfections. In a recent article, Faidi and Nayfeh demonstrated the existence of the underlying physical phenomena by presenting a theoretical model capable of predicting the magnetic field produced by thermoelectric currents around cylindrical inclusions of elliptical cross-section under external excitation [12]. They investigated the shape and magnitude of the resulting thermoelectric signal with respect to the inclusion geometry. The effect of the orientation of the elliptical inclusion on the signal magnitude was also reported. Although the best experimental tool for such studies is undoubtedly a superconducting quantum interference device (SQUID)-based magnetometer, we managed to use a fluxgate magnetometer to provide experimental evidence of the theoretical predictions through the example of elliptical cylindrical tin inclusions by varying the aspect ratio e of the elliptical inclusions in a copper matrix at 2 mm lift-off distance between the sensor and the surface specimen. First, we are going to present a brief review of the analytical model of Refs [12] and [13], and then we will proceed by describing the experimental procedure and finally, discuss the experimental results and compare them to the analytical predictions.

We conducted an experimental research to verify a developed analytical model based on the magnetic sensing of thermoelectric currents produced by cylindrical inclusions of elliptical cross-section embedded in a copper matrix under the influence of an external temperature gradient. The experimental magnetic flux density data obtained from the magnetic field produced by different aspect ratios e of elliptical cylindrical tin inclusions in copper was in good qualitatively agreement with the proposed analytical model. Due to the difference in the thermoelectric properties of the host and the inclusion, the temperature and electric field distributions will be distorted accordingly to the geometry of the inclusion, therefore the thermoelectric signal produced by an inclusion will depend not only on its material properties and size, but also on its shape. Our results indicate that elongated inclusions can be best detected by aligning the externally enforced temperature gradient normal to the major dimension of the inclusion. Further experimental efforts are needed to verify these assumptions and to develop inversion methods capable of quantitatively evaluating the measured magnetic signals in terms of size and shape for inclusions of known physical properties.