Tutorial Exercises and Case Studies

Design Tutorials

Module SESA2007/SESM2008 ; Engineering Design

James Scanlan; School of Engineering Sciences
Page 1
 Tutorial Exercises and Case Studies

Introduction
This module uses a number of small hypothetical design exercises to encourage
students to develop their understanding of the principles and techniques involved.
These exercises are then put into practice in a single case study towards the end of the
module where students are required to undertake a concept design study and
demonstrate how formal design principles /techniques covered in the labs can be
applied.

Module SESA2007/SESM2008 ; Engineering Design

James Scanlan; School of Engineering Sciences
Page 2
 Tutorial Exercises and Case Studies

Tutorial Exercise 1
Background
You have been asked by a high volume car manufacturer to design a car jack for
changing car wheels. This jack will be supplied by the manufacturer as a standard fit on
a range of models.
The jack will be fitted internally within the car. The scope of your design brief does not
cover tools for removing wheel trim or wheel nuts.
You would be able to request modifications to the form and geometry of jacking points
but for structural reasons it not likely that you would be able to change the position of
jacking points.

Objective
To think about, document and analyse customer requirements.
Key questions;
Who are the customers?
What are their objectives
What problems might they face?
Are there any minorities that you need to take into account?
What are the extreme scenarios that you need consider?

Remember:
You need to use language and terminology that the customer groups would
understand and recognise
Ideally the requirements should be un-biased and solution-neutral.
You will need to bias and normalise the relative importance of the customer
requirements (as discussed in the lecture).

Deliverables
Clear set of customer requirements (aim for at least 12 distinct requirements)
Construct a binary weighting matrix to establish relative importance.
Enter binary decisions.
Construct a spreadsheet that can invert the lower triangular part of the matrix
and summate.
Use the spreadsheet to generate a set of normalised relative importance scores.

Module SESA2007/SESM2008 ; Engineering Design

James Scanlan; School of Engineering Sciences
Page 3
 Tutorial Exercises and Case Studies

Tutorial Exercise 2
Objective
To select a design concept as a candidate to evaluate in further detail. To practice
ideation and filtering.

Procedure
The detail design phase of a project is expensive. In order to ensure that you invest
effort in the right concept you need to

generate a large range of candidate design concepts which adequately cover the
scope of the design space
document each concept as a distinct entity
Filter out concepts that clearly do not meet any hard constraints that you have
identified
Rank the remaining concepts
Select the most highly ranked concept

Techniques
An important consideration is the need to partition the concept analysis process:

Stage 1
The objective of the first stage is to generate ideas. The success of this stage is
measured in terms of the range, novelty and number of distinct ideas and concepts
generated. This requires a constructive attitude/ mindset/ atmosphere.

Stage 2
The second stage concerns selecting the best idea/concept in a structured and logical
(auditable!) way. This needs critical appraisal and requires a judgemental and
discriminative attitude/ mindset/ atmosphere.

In an industrial setting it is important to ensure that these two stages are divorced from
each other as they are not complimentary activities. In many companies the two
functions are undertaken by two separate parts of the organisation.

Module SESA2007/SESM2008 ; Engineering Design

James Scanlan; School of Engineering Sciences
Page 4
 Tutorial Exercises and Case Studies

Tutorial Exercise 3
Objective
To identify the key design parameters and, through a mapping process, identify the
relative importance of these parameters.

Procedure
You will start with the ranked list of customer requirements and the relative importance
factors that you generated in Tutorial 2.
You will need to think about the key design variables and list them. All of the design
variables need to be measurable and should be expressed in terms of a specific
engineering unit such as viscosity, mass, force etc.
You need to enter the customer requirements and design parameters into a spreadsheet
matrix.
You need to identify and quantify whether there is a relationship between each of the
requirement/design parameter pairs. Typically this is expressed as a variable between 0
and 1. To simplify this mapping process this is often quantised based on :strong (0.9),
medium (0.5) and weak (0.1) relationships.

Measure of design importance

Using this matrix of mappings you need to use the spreadsheet to generate a normalised
set of weightings for the set of design parameters.

Sanity checks
Within the mapping matrix are there any empty rows? What might this suggest?
Are they any empty columns? What might this suggest?
Rank the design parameters by weighting and try and rationalise this. Are these
priorities sensible? Do they accord with common sense and judgement? You cannot
cheat by directly altering these but you can review the matrix entries.

Module SESA2007/SESM2008 ; Engineering Design

James Scanlan; School of Engineering Sciences
Page 5
 Tutorial Exercises and Case Studies

Tutorial Exercise 4

Objective
To understand the effect of conflicts within a design mapping matrix.
To be able to perform a first-order approximation of non-linear mapping functions and
to use these to be able to undertake simple automated design search (optimisation).

Scenario;
A new aircraft is being designed. A part of the design search space concerns tyre
selection and a single design parameter namely the hardness of the elastomer selected
for the tyre.
Elastomer hardness is measured using the Shore scales A and D.
(http://www.ptli.com/testlopedia/tests/DurometerShore-d2240.asp)

Customer requirements
This matrix shows that we are considering three customer requirements;
How long tyres last
Grip
Puncture resistance
The relative importance of each of these is given in the matrix shown towards the end
of this section.

Module SESA2007/SESM2008 ; Engineering Design

James Scanlan; School of Engineering Sciences
Page 6
 Tutorial Exercises and Case Studies

Relationship Equations
The three default characteristic mapping functions are given below;

1
Max = 1 pv
np
2

1
Min = 1 np

pv
2


1
opt = 2

1 + pv np

tol
where
pv = parameter value
np = neutral point
tol = tolerance
Procedure
Copy the partly filled in matrix into a spreadsheet.

You will notice that there are all three types of relationship in this example. This
suggests that;
For tyre life elastomer harness needs to be maximised
For grip elastomer harness needs to be minimised
For puncture resistance elastomer harness needs to be optimised around a value
of 80 with a neutral point tolerance of 10.

Module SESA2007/SESM2008 ; Engineering Design

James Scanlan; School of Engineering Sciences
Page 7
 Tutorial Exercises and Case Studies

Use the equations to calculate the merit against each of the relationships

Take a weighted mean of the individual merit scores using the relative importance of
the customer requirements. This represents the overall design merit.

The overall design merit should range in value from 0 to 1. The higher the value the
better as this represents the degree to which customer requirements are satisfied.
Because this example involves a three-way trade-off, a compromise has to be reached
in satisfying conflicting customer requirements. An overall design merit of 1 is not
possible.

Design Search
The design search involves finding the value of harness (Value (pv)) that maximises the
overall design merit. You will notice that the matrix includes constraints (U cstrnt, L
cstrnt) giving the upper and lower bounds for the elastomer hardness. Your search
space should not go beyond these constraints as they represent, for example, the
maximum and minimum values that suppliers can achieve.
Microsoft Excel has an add-in that includes a simple optimisation algorithm (shown
below).

This allows an objective function, constraints and parameters to be specified in order to

find a maximum or minimum.
Use solver to find the optimum value for the aircraft tyre elastomer
hardness.

Module SESA2007/SESM2008 ; Engineering Design

James Scanlan; School of Engineering Sciences
Page 8
 Tutorial Exercises and Case Studies

Partly filled in design matrix

Overall Design
Merit Elastomer hardness (Shore A)

0
U L

Value (pv) cstrnt cstrnt

40 120 40

ec iph pi tn
io
na sn hs tir
tr oit no pt ec e
op al it eg na tir m
m
I er lae ra re e de
fo rf t/ lo M th
eiv th o la gi
ta ep rt T ew
le gi yT ue
R e N
Customer reqmt W
How long tyres last 0.3 1 max 95 0.00

Grip 0.5 1 min 40 0.00

Puncture

resistance 0.2 1 opt 80 10 0.00

Module SESA2007/SESM2008 ; Engineering Design

James Scanlan; School of Engineering Sciences
Page 9
 Tutorial Exercises and Case Studies

Tutorial Exercise 5

Objective
To integrate the design parameter mapping, conflict analysis and optimisation using a
single design as an illustration.
Given a set of customer requirements and number of design variables the objective is to
maximise the design merit (goodness of the design).

Customer Requirements
The customer requirements you have been given are.

Light weight
Good impact resistance
Good visibility
Low noise
Easy to put on/remove
Comfortable

Module SESA2007/SESM2008 ; Engineering Design

James Scanlan; School of Engineering Sciences
Page 10
 Tutorial Exercises and Case Studies

Design parameters
The design parameters and constraints are

Shell thickness (mm) upper constraint 8mm, lower constraint 1mm.

Liner thickness (mm) upper constraint 40mm, lower constraint 5mm.
Liner density (Kg/m2) upper constraint 10 Kg/m2, lower constraint 0.2 Kg/m2.
Visor area (cm2) upper constraint 200 cm2, lower constraint 100 cm2.

Procedure
Carry out a binary weighting matrix to establish normalised, biased,
non-linear weightings.
Determine strength of relationships between customer requirements and
design parameters
Determine relationship mapping function between customer requirement
and design parameter (Max, Min, Opt)
Calibrate relationship function by identifying neutral points and neutral
band tolerance (for opt functions).
Calculate the weighted, normalised, sum of the mapping functions. This
represents the overall design Merit and should range from 0 (a design
that fails to satisfy customer requirements to any degree) to 1 (a perfect
design).

Module SESA2007/SESM2008 ; Engineering Design

James Scanlan; School of Engineering Sciences
Page 11
 Tutorial Exercises and Case Studies

Tutorial Exercise 6
Objective
The objective of this exercise is to plot the value vector for a design showing the two
orthogonal dimensions of cost and design merit. A hierarchical cost model will be
developed which is able to react to changes in the design parameters.

Software
You will be using a tool called DecisionPro (Vanguard). This enables hierarchical
calculations to be structured and presented in a logic tree(see below).

total material per part

Page
Unevaluated
material cost
Unevaluated
material cost rate
Undefined

material preparation cost

Page
Unevaluated

forming process cost

Undefined

metal removal cost

Page
Unevaluated

process cost heat treatment cost

Page
Unevaluated Unevaluated

Surface treatment cost

Page
Unevaluated
Generic Part cost
Unevaluated assembly costs
Page
Unevaluated

Quality assurance costs

Page
Unevaluated

transportation
Page
logistics cost Unevaluated
Unevaluated
packing
Page
Unevaluated

tooling design
Page
Unevaluated

non_recurring costs process planning

Unevaluated Undefined

margin CNC programming

Undefined Undefined

Procedure
Using the design parameters given in tutorial 5 as variables (assume any other relevant
design variables are fixed) create a cost model for your motorcycle helmet.
You need to:
Construct a simple BOM (Bill of Material or Product Data-Structure) tree
Approximate (guess) the cost factors such as raw material, process costs (at this
stage it does not matter too much if you get these wrong. What is more
important is to get the model structure correct)
Create expressions (and simple geometrical calculations) that cause the cost
model to vary if any of the given design parameters vary.

Module SESA2007/SESM2008 ; Engineering Design

James Scanlan; School of Engineering Sciences
Page 12
 Tutorial Exercises and Case Studies

Stochastic analysis
DecisionPro has a very sophisticated set of tools to allow sensitivity and uncertainty
analysis to be undertaken. For uncertain inputs you might want to put in a stochastic
variable to represent the confidence intervals associated with state of knowledge of a
value in the cost model.

DecisionPro can then be used to undertake sensitivity analysis and to sample from
stochastic inputs to provide a deeper understanding of the cost uncertainties for your
model.

Value Vector
If you plot cost against merit this allows best value solutions (measured by the
inverse derivative of the cost/merit relationship) to be identified.
Ultimately the combination of cost and merit-based design models can be used to
optimise for value.

Cost C

Merit
Module SESA2007/SESM2008 ; Engineering Design
James Scanlan; School of Engineering Sciences
Page 13
 Tutorial Exercises and Case Studies

Tutorial groups and teams

In general students are required to work individually on the tutorial exercises. However,
is may prove useful to work in small groups to undertake some of the exercise.
All the exercises are individually assessed and therefore students need to ensure that all
work is properly documented in individual log books

Each student must have copies of all relevant computer-based work which they will be
required to show to tutors and for which they need to demonstrate understanding.

For the final case study students will be working in small teams.

Module SESA2007/SESM2008 ; Engineering Design

James Scanlan; School of Engineering Sciences
Page 14