Files in this item



application/pdfLU-DISSERTATION-2016.pdf (24MB)
(no description provided)PDF


Title:Multiphysics modeling and simulation for large-scale integrated circuits
Author(s):Lu, Tianjian
Director of Research:Jin, Jianming
Doctoral Committee Chair(s):Jin, Jianming
Doctoral Committee Member(s):Schutt-Ainé, José E.; Godddard, Lynford L.; Geubelle, Philippe H.
Department / Program:Electrical & Computer Eng
Discipline:Electrical & Computer Engr
Degree Granting Institution:University of Illinois at Urbana-Champaign
Integrated circuits
Heat transfer
Conjugate heat transfer
Thermal stress
Microchannel cooling
Coupled electrical-thermal-mechanical
Finite element method
Domain decomposition
Parallel computing
Abstract:This dissertation is a process of seeking solutions to two important and challenging problems related to the design of modern integrated circuits (ICs): the ever increasing couplings among the multiphysics and the large problem size arising from the escalating complexity of the designs. A multiphysics-based computer-aided design methodology is proposed and realized to address multiple aspects of a design simultaneously, which include electromagnetics, heat transfer, fluid dynamics, and structure mechanics. The multiphysics simulation is based on the finite element method for its unmatched capabilities in handling complicate geometries and material properties. The capability of the multiphysics simulation is demonstrated through its applications in a variety of important problems, including the static and dynamic IR-drop analyses of power distribution networks, the thermal-ware high-frequency characterization of through-silicon-via structures, the full-wave electromagnetic analysis of high-power RF/microwave circuits, the modeling and analysis of three-dimensional ICs with integrated microchannel cooling, the characterization of micro- and nanoscale electrical-mechanical systems, and the modeling of decoupling capacitor derating in the power integrity simulations. To perform the large-scale analysis in a highly efficient manner, a domain decomposition scheme, parallel computing, and an adaptive time-stepping scheme are incorporated into the proposed multiphysics simulation. Significant reduction in computation time is achieved through the two numerical schemes and the parallel computing with multiple processors.
Issue Date:2016-11-21
Rights Information:Copyright 2016 Tianjian Lu
Date Available in IDEALS:2017-03-01
Date Deposited:2016-12

This item appears in the following Collection(s)

Item Statistics