Multiscale modelling and analysis of process systems

1
Multiscale modelling and analysis of process
systems

• First year review
• Gordon Ingram
• Supervisor A/Prof. Ian Cameron

2
outline

• Background
• Definition, Aims, Justification, Literature
• Research plan
• Research progress
• Summary

3
definition

• A multiscale model is
• a composite mathematical model
• formed by combing partial models that describe
  phenomena at different characteristic length
  scales

A multiscale model is a composite model
combines different length scales
4
aims

• Formulate a theory of multiscale modelling
• Demonstrate the theory with two case studies
• Develop some MS CAPM software

5
justification

• Importance of a multiscale approach
• Design process
• Product specifications
• Expanding frontiers
• Need for a strategy
• Many applications
• But, little analysis

6
literature review

• General systematic modelling strategy
• But, multiscale aspects?
• Three key tasks

7
literature review

• The tasks unresolved issues?
• Choosing length scales to include
• Developing models at each scale
• Linking the partial models together
• Frameworks
• A classification scheme
• Properties of the framework classes

8
literature review
Framework classes (Adapted from Pantelides,
2001)
9
research plan
Multiscale modelling theory
Case studies
MS CAPM tool
10
research plan theory

• Formulation and pre-solution analysis
• Graph-based structural techniques

Theory
Regime map
11
research plan case studies

• Develop and compare models
• Packed bed catalytic reactor
• Why?
• Drum granulator
• Why?

12
research plan CAPM tool

• Identify repetitive tasks
• May include
• Graph representation
• Graph transformations
• Metric evaluation
• Links with other packages?

13
research progress

• Desired metrics
• Packed bed catalytic reactor
• Graph representation
• Graph transformations
• Granulator
• Length scales

14
research progress reactor
Froment and Bischoff (1990)
15
Known algebraic
Unknown algebraic
Inputs
Boundary
State
Outputs
km
uo
Co/p(z)
C(z)
Cz0
Co
a
To/p(z)
?f
T(z)
Tz0
Tw
cpf
To
U
CprRp
D
Macro
h
?b
Tp
Micro
?Hr
16
Known algebraic
Unknown algebraic
Inputs
Boundary
State
Outputs
uo
Co/p(z)
C(z)
Cz0
Co
To/p(z)
T(z)
Tz0
Tw
To
CprRp
Macro
Tp
Micro
17
Inputs
Models
Outputs
uo
Co/p(z)
Reactor model
Co
To/p(z)
Tw
CprRp
C(z)
To
Tp
T(z)
Macro
Pellet model
Micro
18
research progress - granulator
Characteristic time (s)
Characteristic length (m)
19
summary

• Increasingly multiscale approach
• Many models, but little guidance
• Some emerging practices
• Qualitative comparisons
• Three key, interrelated tasks
• Focus on model-linking frameworks
• Structure ? Graph theory metrics
• Two case studies

20
acknowledgments

• A/Prof. Ian Cameron
• Prof. Suresh Bhatia, Prof. Katalin Hangos, Dr Fu
  Yang Wang, A/Prof. Victor Rudolph
• Nicoleta Balliu, Hans Wildeboer, Steve McGahey,
  Qin Li, Bogdan Balliu

21
(No Transcript)