Failure Analysis of Diode Components of Electronic Relay Packages Via Thorough Microstructural Characterization

Diode components used in the electronic relay packages of avionics of a weapon system had been found corroded. Two sets of the complete assembly of the damaged diode parts consisting of diode body, plain washer, spring washer and mounting nuts were analysed for finding out the root causes of failures. Detailed metallographic characterization of the failed materials was carried out using optical microscope, scanning electron microscope (SEM), electron dispersive spectroscopy (EDS) analysis within SEM and microhardness study. Inadequate electroless nickel (EN) plating along with significant amount of porosity and presence of MnS type of stringers affected the corrosion resistance of the diode body tremendously for both the sets. Inadequate electroplated Zn layer was observed on spring washer, plain washer and nut, wherein both uniform and pitting corrosion were detected. Important recommendations included: (i) thicker (more than 25 µm) EN coating with as-specified porosity level for diode body, (ii) replacement of electroplated zinc layers with galvanized zinc (of higher thickness) layer on all the fasteners, (iii) precaution in minimizing formation of galvanic couplings observed in the present system, like that between EN-coated diode body and zinc coated fasteners, between mild steel and stainless steel fasteners etc.

Anti-Noise Simulation of the Hysteresis Mathematic Model Using Computationally-Efficient GA

Hysteresis is a particular feature of a wide range of physical systems and devices such as electromagnetic fields, mechanical stress–strain elements and electronic relay circuits. The extended Bouc-Wen model is one of the most widely used hysteresis models in mechanics. It has the capability to emulate the behavior of a wide class of hysteretic systems. However multi parameters have plagued its further application because the capability of computer and algorithm available currently can not meet the need completely. Thus to exploit an effective parallel algorithm is very essential. This paper is committed to propose a novel Genetic Algorithm (GA) so as to identify the parameters of the Bouc-Wen model with noise disturbance efficiently and accurately. Finally a large amount of noise-involved experimental data obtained from a real MR damper is employed to verify the proposed approach has the capability to estimate the satisfactory parameters of the Bouc-Wen model. Also suggested are the implications of the present study on other nonlinear hysteretic models or other complex mathematical models.