DATA-DRIVEN PREDICTION MODEL OF COMPONENTS CHANGE IN SURFACE MOUNT TECHNOLOGY DURING REFLOW CYCLE

Authors

  • Hara Prasad Tripathy , Priyabrata Pattanaik , Sushanta Kumar Kamilla

Abstract

In surface mount technology (SMT), installed components are subject to motion during
the reflow cycle on soldered pads. This capacity is known as self-alignment, and is the product of
molten solder paste's fluid dynamic behaviour. This capability is crucial in SMT, since
inaccurate self-alignment causes defects such as overhanging, tombstoning, etc., whereas on the
other side it can allow components to be perfectly self-assembled on or near the desire location.
The goal of this study is to develop a machine learning model that predicts movement of the
components in xx and yy-directions as well as rotation during reflow. The analysis consists of
two steps: (1) experimental data are analysed to reveal the relationships between self-alignment
and various variables, including component geometry, pad geometry, etc. (2) advanced machine
learning prediction models are used to predict the distance and direction of components shifting
using support vector regression (SVR), neural network (NN), and random regressi forest. As a
result, RFR can predict components shifting with an average fitness of 99%, 99%, and 96% and
an average prediction error of 13.47 (μmm), 12.02 (μmm), and 1.52 (deg.) for component shifts
in xx, yy, and rotational directions respectively. This enhancement provides the future capability
in the pick-and-place system to refine the parameters to monitor the best placement position and
reduce the intrinsic defects caused by the self-alignment.

Downloads

Download data is not yet available.

Downloads

Published

2020-12-01

How to Cite

Hara Prasad Tripathy , Priyabrata Pattanaik , Sushanta Kumar Kamilla. (2020). DATA-DRIVEN PREDICTION MODEL OF COMPONENTS CHANGE IN SURFACE MOUNT TECHNOLOGY DURING REFLOW CYCLE. PalArch’s Journal of Archaeology of Egypt / Egyptology, 17(6), 5098 - 5106. Retrieved from http://mail.palarch.nl/index.php/jae/article/view/1772