Abstract
To ensure the safety of structure and prevent it from catastrophic failure, it is highly desirable to seek a reliable non-destructive inspection (NDI) for detecting the defects in structure. There are numerous NDI methods for detecting the cracks or defects, among them direct current potential drop (dc-pd) technique is useful and has been proved effective. By applying direct current to the electrical conductive structure and investigating the variation of potential distribution disturbed by the discontinuity of the structure, the defects in the structure can be identified. Because of its robust and direct application on the electrical conductive structure, test can be done without the making of specimen or using expensive electronic equipment and techniques, and it also can be used in real time monitoring of engineering structure. Thus it is worth systematically exploring further this technique.This paper has firstly established the finite element analysis model for the electrical potential drop based on the electrical field theory. Then a factor, called “defect influence factor ”, has also been devised to identify the defects in an attempt to filter out the interferences caused by the parameters such as the quantity of current, input types and places of current supply, and the geometry of structure except for the defects. The factor is the ratio of the electrical potential drop on a defective structure to that of an identical one without defects. By depicting the contours of the defect influence factor, the defects of the structure can be clearly shown. For dealing with the multiple defects or cracks detection, the work has extended the finite element alternating method developed by Chen et al. (1994) originally for heat conduction analysis of multiple cracks to dc-pd to calculate the electrical potential distribution. Several samples with single or multiple defects (such as circular holes, square holes, and slots) and cracks were undergone the experiment. With the set of defect detection procedure and defect identification criteria, the shape, size, number and location of defects have been accurately identified by depicting the contours of the defect influence factor. Good agreement between the computed results of finite element method/ finite element alternating method and experimental data shows the merits of this technique.Although dc-pd technique has advantages, there are disadvantageous constraints in it. It will cause measuring difficulties due to point-to-point contact measuring using two probes. In addition, the internal resistance of structure results in joules heating which may in turn make damage to the structure parts after the currents flow through the electrical conductive structure for a while. In order to overcome these measuring constraints and thermal damages, finding to use infrared to implement non-contact thermal measurement, and establish electro-thermal (ET) technique for defect detection is also an important part of this work.Quite similar to the considerations taken in dc-pd technique, a factor called “temperature defect influence factor ” is devised for the detection of structural defects. The factor, defined as the ratio of temperature rise of a defected structure to that of a perfect one, has the effect of filtering out the possible interferences on the detection coming from parameters such as the quantity of current, input types and places of current and the geometry of structure except for defects. By depicting the contours of the temperature defect influence factor, the defects or cracks in the structure can be clearly identified. To verify the effectiveness of the ET technique newly developed in this work, in addition to the establishment of ET finite element analysis model, an infrared thermography is used to measure temperature distribution. Samples used for detection test include two thin quad flat packages (TQFP) with pop corn and debonding defects and one stainless plate with crack. The results from the former experiments are then compared with x-ray data. The comparison shows positive usefulness of the ET technique developed in this work.The defect identification criteria ( and ) of this work are variable according to the structural material. When extension this present technique to apply in the industrial product inspection, a vast calculation of the or and classified the variation with different material for establishing a table of crack test specification will be necessary. The detection of defects with dc-pd and ET technique all deals with through-wall defects or cracks, except that in TQFP specimen. But in reality, the structural defects are present in partial through-wall cracks caused by delamination, impurity, and oxidation and corrosion. Also taken into consideration is the coming of micro, nano level structures design of products. These all require high resolution of defect detection which lies ahead for future work.