AdvancedTraining_12_Compact_Component_Modelling.pdf

Compact Component Modelling

Introduction
- Compact Model Topologies
- Deriving Compact Models
– The Computational Cold Plate Test
– The DELPHI Approach
- Compact Models in FLOPACK

The Traditional Approach
θja and θjc
- The θja and θjc approaches lump all heat paths together as one - use with caution.
- θja and θjc are environmentally dependent.
- Inaccuracies in predicting junction temperatures can be as high as 100%!

Compact Models
- A Compact Model seeks to capture the thermal behavior of the package accurately at pre-determined (critical) points
– junction
– case
– etc.
- .... by using a reduced set of parameters to represent the package
– These parameters need not be geometric
- The most popular approaches use some sort of thermal resistance network representation

Topologies
- Two-resistance network
– Simplest topology
– Easy to extract
– Easy to implement in most tools
– Relatively inaccurate (~ 30%) for absolute results
– Often sufficiently accurate for trends/parametric studies

Deriving Compact Models
- Several methods proposed
- We shall consider two
– The “Computational Cold Plate Test”
– The DELPHI Approach
- Computational Cold Plate Test
– Rjc = (Tj - Te)/P
– Rjb = (Tj - Te)/P
Tj = Junction Temperature
Te = Temperature of Isothermal Surface
P = Package Power
- How accurate is this method?
– Because of the “unrealistic” nature of the heat flux path lines in the two simulations, the resistances derived will tend to under predict the junction temperature
– This could be as much as 50%!
- Recommendation
– Use Computational Cold Plate Test only to get ball park estimates of junction temperature
– Useful for predicting trends (parametrics)
– For greater accuracy, use detailed models or more complex compact models (where available) www.resheji.com

The DELPHI Approach
- What was DELPHI?
- Project that proposed new methodologies for creating and validating component computational models
- Ultimate Goal
– To enable component manufacturers to supply validated compact thermal models of their parts to end-users
- Results were
– Detailed model understanding of some package types
– 2 experimental systems
– Double Cold Plate and
– Submerged Double Jet Impingement
– Complex compact model networks for some package types
– A methodology to tie these together

Implementing Nodes & Resistances in FLOTHERM

Implementing the 2 Resistor Model
- All power is dissipated in Junction block
- Moderate accuracy (20 - 30%) for most components but will predict trends correctly; easy to tweak.....

Arbitrary Resistance Networks
- Most components need more complex networks, especially when heat spreading within the component is significant (PBGA, PQFP …)
- Often involve “Shunt” resistors

Compact Models in FLOPACK
- 2-Resistor Compact Models
– Available through the FLOPACK web site for all package types
– Rjt and Rjb data can be measured by manufacturers
- Star Compact Models
– Available through the FLOPACK web site for leaded packages
– Easy to set up
– Accuracy often same as 2-resistor models
- Complex Compact Models
– Maximum accuracy; some in use
– Available through the FLOPACK web site for leaded packages
- Compact Model SmartPart in Version 3.1 of FLOTHERM
– Embedded resistor network solver

===============================

AdvancedTraining_1_Introduction.pdf
AdvancedTraining_2_Governing_Equations.pdf
AdvancedTraining_3_Turbulence_Modelling.pdf
AdvancedTraining_4_Boundary_Layers_and_Heat.pdf
AdvancedTraining_5_Validation_Cases.pdf
AdvancedTraining_6_Numerical_Solution.pdf

标签： 点击： 评论: