Search

Determination of thermal dissipation capability and thermomechanical behavior of electronic components is becoming a relevant topic in view of the fact that thermal management can strongly affect reliability and lifetime of the component. The level of miniaturization, which is achieved in current electronics, requires that no-contact analysis must be privileged in view of the fact that any contact with the chip can significantly alter the heat capacity of the component and to introduce a systematic bias in the measurements. Moreover, the high complexity of electronics components suggests that full-field methods should be adopted in order to take into account about differences in terms of thermomechanical response in different areas of the same sample. In this article, an original optical set-up including a speckle interferometer and a projection moiré branch was developed; this system comprises two different laser sources emitting at two different wavelengths. Recording by a color camera allows detecting simultaneously the speckle pattern and the projected fringes. The whole information is successively separated in the post-processing stage and this allows obtaining in-plane and out-of-plane displacement fields. The system demonstrated its capability to determine dynamic response of the LM1084 analyzed component and, moreover, to detect the presence of a functional damage.

Among the many available methods for measuring residual stresses, the hole-drilling method is surely one of the most popular methods. It is a semidestructive method consisting in drilling a very small hole into the specimen. This produces residual stress relaxation around the hole that causes strains also to change. Strain on the surface can be measured by strain gages or by optical techniques like electronic speckle pattern interferometry. Holes can be drilled at different depths, so that a stress profile can be determined using the stress–strain relationships. However, due to the mathematics of the problem, small measurement errors on the strain can imply big errors in the calculated stress. Accuracy of the final results is strongly dependent on the measurement procedure and on the algorithms adopted to calculate stresses. In this study, the influence of strain measurement errors is evaluated with respect to different measurement procedures. Two kinds of algorithms, namely, integral and power series, are compared also with respect to the use of two different commercial systems both implementing the cited methods. Moreover, the effects of a proper choice of depth increments and also of the number of steps are investigated. Effects of implementation of regularization algorithm are discussed as well as the influence of the drilling depth. Finally, the ability of the analyzed methods in discriminating two different strain signal sources is investigated.