Combining multi-phase field simulation with neural network analysis to unravel thermomigration accelerated growth behavior of Cu6Sn5 IMC at cold side Cu–Sn interface

Abstract

Integration of data-driven approach with theory,computation and experiment.• Free energy density function introduced for thermomigration.• Heat of transport for Cu in Cu 6 Sn 5 phase determined using neural network analysis.• Cold side IMC grain growth modeled using multi-phase field method.

Entropy per rapidity in Pb-Pb central collisions using Thermal and Artificial neural network(ANN) models at LHC energies

“…In our work, best MSE and R values are obtained by using four hidden layers. The number of neurons in each hidden layer are (40,40,40,40), (20,20,10,20), (40,30,30,30), (40,40,40,40)…”

“…Artificial neural network model [27][28][29][30][31][32][33] is a machine learning technique most popular in high-energy physics community. In the last decade important physics results have been separated utilizing this model.…”

“…Best performance value and regression are obtained by using four hidden layers. The number of neurons in each hidden layer are (100, 100, 120, 120), (70, 90, 80, 80), (100, 80, 80, 70), (20,30,30,20), (30,20,40,40) Extrapolation of the observed transverse momentum spectra to p T = 0 is necessary to determine the entropy S. To achieve this, we fitted both the experimental and simulated p T spectra to two various functional models, Tsallis distribution [11,12] and the HRG model [13]. The aim of combining two models is to fit the entire p T curve.…”

Simulation for Cu Atom Diffusion Leading to Fluctuations in Solder Properties and Cu6Sn5 Growth during Multiple Reflows

“…In recent years, machine learning and data analytics have been recognized as techniques for the real-time quality monitoring of material processing experiments. For instance, the convolutional neural network (CNN) model was developed to detect Cu 6 Sn 5 IMC and bubbles in liquid solder [25]; the multi-phase field model was combined with machine learning to assess the growth of Cu 6 Sn 5 during thermomigration and electromigration process [26,27]; and finite element analysis and machine learning were utilized to predict the morphology of IMC in an Sn-x Ag-y Cu/Cu (SAC/Cu) system [28]. In a word, data-driven science has become a powerful tool to predict the growth behavior of IMC.…”

Формирование интерметаллида Cu-=SUB=-6-=/SUB=-Sn-=SUB=-5-=/SUB=- в тонких пленках Cu/Sn

“…Преимущества бессвинцовых припоев на основе сплавов Cu−Sn для окружающей среды и здоровья человека, побудили исследователей изучить свойства и поведение системы Cu−Sn [1]. В электронной аппаратуре при пайке с использованием припоев Cu−Sn, наблюдается интенсивное образование интерметаллида Cu 6 Sn 5 за счет реакции между Cu и Sn [2][3][4]. При миниатюризации электронных устройств, доля интерметаллического слоя по сравнению с общей толщиной паяного соединения увеличивается.…”