The Eighth Annual University of Leeds Learning and Teaching Conference 1 Computational software and the learning cycle Malcolm Povey, Nick Parker and Mel.

The Eighth Annual University of Leeds Learning and Teaching Conference

1

Computational software and the learning cycle

Malcolm Povey, Nick Parker and Mel Holmes

School of Food Science and Nutrition

7th January 2011

Context and challenge

2

• Food science is multidisciplinary (biology, chemistry, physics, maths)

• Food science undergraduates have limited ability/interest in mathematics and physics

• Level of physics required is often a surprise to students and they often struggle to see the relevance of the theory

• The conventional approach to physics and maths based on progressive theory and highly simplified case studies is not ideal

• How do we engage and convey the principles effectively ?

Vehicle - Food 2045: Food Innovation and Design

Supported by a Academic Development Fund for Learning and Teaching 2008-2010

Learning outcomes

• To understand the physical principles underlying the analysis of food processing operations and their application to process design.

• To become proficient with an advanced modelling tool capable of supporting design and innovation for new and existing foods.

• To be able to gain a quantitative understanding of complex problems in food processing operations.

Teaching format•14 hours of lectures on the basics of heat processes•7 hours of problem classes•9 hours of computing classes (COMSOL Physics package)

Hierarchical thought – Kolb’s Learning cycle and Bloom’s Taxonomy

4

Addressing the Learning Cycle

5

1. Experience - Define real-world problem and develop intuitive theoretical model

2. Reflective observation - Discuss relationships between model and real situation and consider limitations

3. Abstract conceptualisation - Solve and interpret results, are they reasonable?

4. Active experimentation – Redefine model parameters, boundary conditions etc to improve results. Apply to new situations

COMSOL Multiphysics

6

• Powerful Industry-recognised solver

• Predefined multiphysics-application templates solve many common problem types. Fluid flow, heat transfer, structural mechanics and

electromagnetic analyses.

http://www.comsol.com/products/multiphysics/

COMSOL and Basic Theory

7

Comsol Modelling stages

8

The typical modelling steps include:

1. MODEL (Definition of the geometry: Draw, Draw mode)

2. PHYSICS (Definition of the equations, parameters of the matter, initial and boundary conditions: Physics)

3. MESHING

4. SOLVING

5. RESULTS (Postprocessing)

Worked Example – the Battered ChipExperience and observation

9

Appropriate Boundary Conditions also required

10

Post-processing – Concepts and interpretation

Variety of plots available, e.g.

•Surface

•Probe

•Cross-section

11

Demonstration

‘Open’ project – Extensional thought

12

Our suggestion: modify the battered chip - 50% students

50% produced new, innovative models

Student Feedback

13Very

challe

nging

A little

challe

nging

Too

easy

A little

eas

y

Neutra

l

Very sti

mulating

A little

stim

ulating

Very

borin

g

A little

bor

ing

Neutra

l

Challenging Stimulation

COMSOL

Lectures

Summary

14

Pro’s Con’s

User-friendly and visually engaging Could be over-whelming to some students

Large user resources Software is a ‘black box’

Learning cycle quickly negotiated By-passes fundamental material

Develops and encourages higher-order thinking

Easy to be complacent and avoid critical thinking

Real life problem solving with limited maths/physics/computing background

Not suited to visually impaired

Encourages innovation & exploration and active learning

Positive student feedback

Attain a transportable and recognised skill (CV – employment)

Expands knowledge and application of mathematical models (further studies)