Heat generation transport in micro and sub- micro scale in electronic packaging
Energy exchange takes place in extremely small dimension and time scale in the process of micro-electronic packaging. For fast heating response. Fourier conduction law is inadequate to explain the phenomena. Thus. compensate the Fourier law, named as Non-Fourier law. Non-Fourier law, based on tw...
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2003
|
| Subjects: | |
| Online Access: | http://eprints.usm.my/58319/ http://eprints.usm.my/58319/1/OOI%20CHUN%20KEANG24.pdf |
| Summary: | Energy exchange takes place in extremely small dimension and time scale in the process
of micro-electronic packaging. For fast heating response. Fourier conduction law is
inadequate to explain the phenomena. Thus.
compensate the Fourier law, named as Non-Fourier law. Non-Fourier law, based on
two-phase-lag model has introduced two
from classical Fourier heat conduction equation when applied to rapid heating process.
These assumptions are finite thermal wave propagation speeds and time of equilibrium
between electron and lattice. From previous research on dual phase-lag model, different
governing equations have to employ for different boundary conditions, but with a
proposed two phase-lag model only a single governing equation is adequate. These
phase lags are the phase lag for temperature gradient (xT) and heat flux (xq). A finite
element method and Runge-Kutta method are applied in the development of threelower
temperature values as compared with one-dimensional and two-dimensional
model. The application of two phase-lag model to very-large-scale-integrated (VLSI)
interconnect thermal analysis, illustrates that circuit open failure occurs at current pulse
of 300ns. An implementation of Asymptotic Waveform Evaluation (AWE) scheme in
first and second order ordinary differential equation shows a break through as compared
with conventional methods. This advanced, powerful and efficient scheme shows
excellent results compared with Runge-Kutta method, central difference method and ANSYS* 5.4, and is several orders faster. |
|---|