Exercises Exerciseswithworkedsolutions Ma Uoc en

Exercises With
Worked Solutions
Professor Mike Ashby

Department of Engineering
University of Cambridge
M. F. Ashby, 2010
For reproduction guidance see back page
This compilation of Exercises with worked solutions is a part of a set of teaching resources created by Mike Ashby to help introduce students to materials,
processes and rational selection.
The Teaching Resources website aims to support teaching of materials-related courses in Design, Engineering and Science.
Resources come in various formats and are aimed primarily at undergraduate education. Some of them are open access.
M.F. Ashby 2010
www.grantadesign.com/education/resources
CES EduPack Exercises with Worked Solutions
th
Adapted from the Solution Manual to Materials selection in mechanical design, 4 edition,
Prof. Mike Ashby Engineering Department, University of Cambridge, CB2 1PZ.
Exercises with worked solutions
Contents
E1. Introduction to exercises
E2. Material evolution in products
E3. Devising concepts
E4. Using material properties
E5. Using material selection charts
E6. Translation: constraints and objectives
E7. Deriving and using material indices
E8. Multiple constraints and objectives
E9. Selecting material and shape
E10. Hybrid materials
E11. Selecting processes
E12. Materials and the environment
M.F. Ashby 2010

E1 Introduction to exercises
These exercises are designed to develop facility in selecting
materials, processes and shape, and in devising hybrid materials when
no monolithic material completely meets the design requirements.
Each exercise is accompanied by a worked solution. They are
organized into the twelve sections listed on the first page.
3.
A request for a selection based on one material index alone (such

as M = E 1 / 2 / ) is correctly answered by listing the subset of
materials that maximize this index. But a request for a selection of
materials for a component a wing spar, for instance (which is a
light, stiff beam, for which the index is M = E 1 / 2 / ) requires
The early exercises are easy. Those that follow lead the reader
through the use of material properties and simple solutions to
mechanics problems, drawing on data and results contained in the
CES EduPack and the booklet Useful Approximate Solutions to
Standard Problems (UASSP which can be found as an open access
resource on the teaching resource website.) ; the use of material
property charts; techniques for the translation of design requirement to
identify constraints and objectives; the derivation of indices, screening
and ranking, multi-objective optimization; coupled choice of material
and shape; devising hybrids; and the choice of materials to meet
environmental criteria.
Three important points.

1. Selection problems are open-ended and, generally, underspecified; there is seldom a single, correct answer. The proper
answer is a sensible translation of the design requirements into
material constraints and objectives, applied to give a short-list of
potential candidates with commentary suggesting what supporting
information would be needed to narrow the choice further.
2. The positioning of selection-lines on charts is a matter of
judgement. The goal is to place the lines such that they leave an
adequately large "short list" of candidates (aim for 4 or so), drawn,
if possible, from more than one class of material.
M.F. Ashby 2010
more: some materials with high E 1 / 2 / such as silicon carbide,

are unsuitable for obvious reasons. It is a poor answer that
ignores common sense and experience and fails to add further
constraints to incorporate them. Students should be encouraged
to discuss the implications of their selection and to suggest further
selection stages.
All the materials selection problems can be solved using the

Standard databases in the CES EduPack software, which is
particularly effective when multiple criteria and unusual indices are
involved. This version of the solution manual has been written such
that it can be completed without having to refer to the book Materials
th
Selection in Mechanical Design, 4 edition. The original solution
manual for the book can be found on the Granta Design website.
Hint: When using the CES EduPack, you can avoid manually
calculating the average values of material properties (they tend to
span a range) by going to ToolsOptionsNumbers and checking
the Display data ranges as average values checkbox.

E2 Material evolution in products
E2.1. Use Google to research the history and uses of one of the
following materials
Tin
Glass
Cement
Titanium
Carbon fiber
Present the result as a short report of about 100 - 200 words (roughly
half a page).
Specimen answer: tin. Tin (symbol Sn), a silver-white metal, has a
long history. It was traded in the civilisations of the Mediterranean as
early as 1500 BC (the Old Testament of the Christian bible contains
many references to it). Its importance at that time lay in its ability to
harden copper to give bronze (copper containing about 10% tin), the key
material for weapons, tools and statuary of the Bronze age (1500 BC
500 BC). Today tin is still used to make bronze, for solders and as a
corrosion resistant coating on steel sheet (tin plate ) for food and drink
containers a tinnie, to an Australian, is a can of beer. Plate glass is
made by floating molten glass on a bed of liquid tin (the Pilkington
process). Thin deposits of tin compounds on glass give transparent,
electrically conducting coatings used for frost-free windshields and for
panel lighting.
E2.2 Research the evolution of material use in
Writing implements (charcoal, lead (graphite), quill pens, steel

nib pens, gold plus osmium pens, ball points..)
Watering cans (wood galvanized iron polypropylene)
Bicycles (wood bamboo steel, aluminum, magnesium,

titanium CFRP)
Small boat building (wood aluminum GFRP)
Book binding (Wood leather cardboard vinyl)
M.F. Ashby 2010

E3
Devising concepts
These two examples illustrate the way in which concepts are

generated. The left-hand part of each diagram describes a physical
principle by which the need might be met; the right-hand part
elaborates, suggesting how the principle might be used.
Concept
Electric fan pulling air stream through

paper or cloth filter in portable unit
C1 Entrain in air stream
and filter
E3.1 Concepts and embodiments for dust removers. The need for
a device to remove household dust has many established
solutions. You can read about the various ways this problem has
been solved in the last 100 or more years by searching the
Internet.
Now is the time for more creative thinking. Devise as many
concepts to meet this need as you can. Nothing, at the concept
stage, is too far-fetched; decisions about practicality and cost
come later, at the detailed stage. List concepts and outline
embodiments as block diagrams like this:
C2 Entrain in air
stream,
Embodiment
Central pump and filter linked to

rooms by ducting
Central turbo-fan with centrifugal

dust collector linked to rooms by ducting
Centrifugal turbo-fan, injected water

spray to wash air stream, in portable unit
Electric fan pulling air stream through

paper or cloth filter in portable unit
C1 Entrain in air stream
and filter
Central pump and filter linked to

rooms by ducting
Centrifugal turbo-fan with surrounding

dust collector, in portable unit
C3 Entrain in air
stream,
Concept
Embodiment
Central turbo-fan , water spray to wash

air stream, linked to rooms by ducting
Axial fan drawing air stream between

charged plates, in portable unit
C4 Entrain in air
stream, trap
C5 Trap dust on
adhesive strip
Central fan with electrostatic collector,

linked to rooms by ducting
Reel-to-reel single sided adhesive tape

running over flexible pressure-pad
Answer. Design problems are open-ended; there are always

alternative solutions. Here are some for dust removers.
M.F. Ashby 2010

E3.2 Cooling power electronics. Microchips, particularly those for
power electronics, get hot. If they get too hot they cease to
function. The need: a scheme for removing heat from power
microchips. Devise concepts to meet the need and sketch an
embodiment of one of them, laying out your ideas in the way
suggested in exercise E3.1.
Embodiment
Concept
Massive, with sufficient heat capacity to absorb heat

over work cycle without significant increase in
C1 Conduction
Compact, requiring back-up by coupling to convection,
evaporation or radiation
Answer. Four working principles are listed below: thermal conduction,

convection by heat transfer to a fluid medium, evaporation
exploiting the latent heat of evaporation of a fluid, and radiation,
best achieved with a surface with high emissivity.
The best solutions may be found by combining two of these:
conduction coupled with convection (as in the sketch an oftenused combination) or radiation coupled with evaporation (a
possibility for short-life space structures).
Free convection not requiring fan or pump.

C2 Convection
Forced convection with fan or pump
Unconfined, such as a continuous spray of volatile fluid

C3 Evaporation
Confined, utilising heat-pipe technology
Radiation to ambient, using high emissivity coatings

C4 Radiation
Radiation to cooled surface, from high emissivity
surface to highly absorbent surface
M.F. Ashby 2010

E4
Using material properties
E4.3 A thick-walled tube has an inner radius ri = 10 mm and an outer
Useful Approximate Solutions to Standard Problems (UASSP) is a

booklet available to instructors and students via Grantas Teaching
Resources. This and the data in the CES EduPack can be used to
solve the questions in this chapter.
radius ro = 15 mm. It is made from polycarbonate, PC. What is the

maximum torque that the tube can carry without the onset of yield?
Retrieve the (mean) yield strength y of PC, the expression for the
E4.1 A cantilever beam has a length L = 50 mm, a width b = 5 mm and

a thickness t = 1 mm. It is made of an aluminum alloy. By how much
will the end deflect under and end-load of 5 N? Use data from the CES
EduPack for the (mean) value of Youngs modulus of aluminum alloys,
the equation for the elastic deflection of a cantilever from UASSP, 3
and for the second moment of a beam from UASSP, 2 to find out.
Answer. The torque at the onset of yield for a thick walled tube is
Answer. The deflection of a cantilever under an end load F is, from

UASSP, 3,
F L3
bt3
with
I =
3EI
12
The mean Youngs modulus E for aluminum is 75 GPa. Inserting the
data from the question results in an end deflection = 6.7 mm.
E4.2 A spring, wound from stainless steel wire with a wire diameter
d = 1mm, has n = 20 turns of radius R = 10 mm. How much will it
extend when loaded with a mass P of 1 kg? Assume the shear
modulus G of stainless steel to be 3/8 E where E is Youngs modulus
(from the data in the CES EduPack), and use the expression for the
extension of springs from UASSP, 6 to find out.
Answer. The extension u of a spring under a force F = Pg = 9.81 N
(here g is the acceleration due to gravity) is
u =
64 F R 3 n
Gd4
Youngs modulus for stainless steel is 200 GPa, so shear modulus
G 76 GPa. Inserting the data gives a deflection u = 10.4 mm.
M.F. Ashby 2010
torque at onset of yield from UASSP, 6 and that for the polar moment of a
thick walled tube from UASSP, 2 to find out.
Tf =
K y
2 ro
with
K =
( ro4 ri4 )
The mean yield strength of y of PC from the CES EduPack is 65 MPa.

Inserting the data from the question gives a torque at the onset of yield of
T f = 138 N.m.
E4.4 A round bar, 20 mm in diameter, has a shallow circumferential

notch with a depth c = 1 mm with a root radius r = 10 microns. The bar
is made of a low carbon steel with a yield strength of y = 250 MPa. It is
loaded axially with a nominal stress, nom (the axial load divided by the
un-notched area). At what value of nom will yield first commence at the
root of the notch? Use the stress concentration estimate of UASSP, 9 to
find out.
Answer. The stress concentration caused by notch of depth c and root
radius r is
1/ 2
max
c
= 1 +
with
2 for tension
nom
r
Yield first starts when max = y . Inserting the data from the question
gives a nominal stress for first yield of 11.9 MPa. Stress concentrations
can be very damaging in this example, a cyclic stress of only 12 MPa
will ultimately initiate a fatigue crack at the notch root.

E4.5 An acrylic (PMMA) window is clamped in a low carbon steel
frame at T = 20 C. The temperature falls to T = -20 C, putting the
window under tension because the thermal expansion coefficient of
PMMA is larger than that of steel. If the window was stress-free at
20C, what stress does it carry at -20 C? Use the result that the bi-axial
stress caused by a bi-axial strain difference is
E
=
1
where E is Youngs modulus for PMMA and Poissons ratio = 0.33 .
You will find data for expansion coefficients in Table A7, and for moduli
in UASSP, 5. Use mean values.
Answer. The strain difference caused by difference in thermal

expansion, , when the temperature changes by T is
= PMMA Low C steel T

From the CES EduPack,
PMMA = 117x 10-6/C
and
Low C steel = 12.3 x 10-6/C
-3
giving = 4.2 x 10 . The modulus of PMMA from the CES

EduPack is E = 3.0 GPa. The equation given in the question then
predicts a tensile stress in the window of = 19 MPa.
E4.6 The PMMA window described in Exercise 4.5 has a contained

crack of length 2a = 0.5 mm. If the maximum tensile stress that is
anticipated in the window is = 20 MPa, will the crack propagate?
Choose an appropriate equation for crack propagation from UASSP, 10
and data for the fracture toughness K 1c of PMMA from the CES
EduPack to calculate the length of crack that is just unstable under this
tensile stress.
M.F. Ashby 2010
Answer. The crack length is small compared with the width of the
window, so the appropriate choice of equation describing crack
instability is
C a K 1c
with
C = 1 .0
Inserting the data we find the length of the shortest crack that is just
unstable:
2
2K
2a = 1c = 2.1 mm, using (K 1c ) PMMA = 1.15 MPa.m1/2

Thus the 0.5mm crack will not propagate.
E4.7 A flywheel with a radius R = 200 mm is designed to spin up to

8000 rpm. It is proposed to make it out of cast iron, but the casting
shop can guarantee only that it will have no crack-like flaws greater
than 2a = 2 mm in length. Use the expression for the maximum
stress in a spinning disk in UASSP, 7, that for the stress intensity at a
small enclosed crack from UASSP, 10 and data for cast iron from the
CES EduPack to establish if the flywheel is safe. Take Poissons ratio
for cast iron to be 0.33.
Answer. The maximum tensile stress in a spinning disk is
(3 + ) 2 R 2 , and K =
max =
1
max a K 1c
8
for a contained crack. Here = 2 W / 60 radians/sec when W is
the rotational velocity in rpm. Inserting the data from the question and
3
1/2
the mean values for density = 7150 kg/m and K 1c = 38 MPa.m
from the CES EduPack, we find the maximum that the rotational
velocity that will just cause the cracks to propagate is 4350 radians/s,
or 41,600 rpm. The flywheel is safe.

E4.8 You wish to assess, approximately, the thermal conductivity of
polyethylene (PE). To do so you block one end of a PE pipe with a wall
thickness of x = 3 mm and diameter of 30 mm and fill it with boiling
water while clutching the outside with your other hand. You note that
the outer surface of the pipe first becomes appreciably hot at a time t
18 seconds after filling the inside with water. Use this information, plus
data for specific heat C p and density of PE, to estimate for PE.
and is the material density and C m the material cost per kg. Data
from the CES EduPack are assembled below, using the means of the
ranges..
How does your result compare with the listed value in the CES
EduPack?
Answer. The distance x that heat diffuses in a time t is
approximately
x =
2at
with
a =
Density
3
kg/m
Price
$/kg
Cost per unit

3
vol $/m
Cast iron
7150
0.66
4719
Nylon
1130
3.45
3900
Surprisingly, the nylon casing has a lower material cost than that made
of cast iron.
Cp
( a is the thermal diffusivity). Inserting the data from the question and
3
the mean values = 950 kg/m and C p = 1850 J/kg/K, we find
0.44 W/m.K. The result given in the CES EduPack for the thermal
conductivity of PE is 0.403 0.435 W/m.K.
E4.9 It is proposed to replace the cast iron casing of a power tool with
one with precisely the same dimension molded from nylon. Will the
material cost of the nylon casing be greater or less than that made of
cast iron? Use data from the CES EduPack to find out.
Answer. If the dimensions of the cast iron and nylon cases are the
same, the volume of material required to make them are equal. Thus the
cheaper option is the one with the lower material cost per unit volume
C v , where
and is the material density and C m the material cost per

kg. Data from Appendix A3 and A11 are assembled below,
using the means of the ranges..
cheaper option is the one with the lower material cost per unit volume
C v , where
Cv = Cm
and is the material density and C m the material cost per kg. Data
from the CES EduPack are assembled below, using the means of the
ranges..
Cv = Cm
M.F. Ashby 2010

E5. Using material selection charts
E5.2
The 20 exercises in this section involve the simple use of the charts
that can be created using the CES EduPack to find materials with given
property profiles. This involves first creating the chart, then applying the
appropriate box or line selection and reading off the materials that lie
on the appropriate side of the line. It is a good idea to present the
results as a table. You can copy this table from the CES EduPack.
Answer. There are only three materials with modulus E > 50 GPa
and density < 2000 kg/m3.
The software offers links to processes, allows a wider search by using

the Level 3 database, and gives access to supporting information via
the Search Web function.
E5.1
Material
Cast and Wrought
Magnesium alloys
CFRP carbonfiber reinforced

plastic
A component is at present made from a brass, a copper alloy.

Use a Youngs modulus Density (E-) chart to suggest three
other metals that, in the same shape, would be stiffer. Stiffer
means a higher value of Youngs modulus.
E5.3
Answer. Metals that are stiffer than brass are listed in the table.
Material
Steels
Nickel alloys
Tungsten alloys
Use a Youngs modulus Density (E-) chart to identify

materials with both a modulus E > 50 GPa and a density <
3
2000 kg/m .
Comment
Magnesium is the lightest of all common
structural metals only beryllium is lighter,
but it is very expensive and its oxide is
toxic.
CFRP is both lighter and stiffer than
magnesium. That is one reason it is used
for competition cars and bikes.
Use a Youngs Modulus-Density (E-) chart to find (a) metals

that are stiffer and less dense than steels and (b) materials (not
just metals) that are both stiffer and less dense than steel.
Comment
The cheapest stiff, strong structural metal,
widely used.
More expensive than steel
Refractory (high-melting) and relatively
expensive
Answer. (a) No metals in Figure 4.3 are both stiffer and less dense
than steel, though nickel alloys come close. (b) Several ceramics
qualify: Boron carbide, B4C, silicon carbide, SiC, silicon nitride
Si3N4, aluminium nitride AlN and alumina Al203.
Material
Alumina Al203
Silicon nitride Si3N4
Boron carbide, B4C
Silicon carbide, SiC
Aluminium Nitride,
AlN
M.F. Ashby 2010
10
Comment
Alumina is the most widely used of all
technical ceramics (spark plugs, circuit
boardsO) All ceramics are brittle they
have low values of fracture toughness K1c
and toughness G1c .

E5.4
Use a E- chart to identify metals with both E > 100 GPa and
E/ > 0.02 GPa/(kg/m3).
E5.5
Use a E- chart to identify materials with both E > 100 GPa

1/3
and E 1 / 3 / > 0.003 (GPa) /(kg/m3). Remember that, on
taking logs, the index M = E 1 / 3 / becomes
Answer. The chart shows the selection. The metals that lie in the
search area are listed in the table.
Material
Steels
Nickel alloys
Titanium alloys
Log ( E ) = 3 Log ( ) + 3 Log ( M )

and that this plots as a line of slope 3 on the chart, passing
through the point E = 27 when = 1000 in the units on the chart.
Comment
Cheap, widely used. Stiff structural material.
More expensive than steel
Titanium alloys are very expensive.
Answer. The chart shows the selection. The materials that lie in the
search area are listed in the table. No metals survive.
Material
Comment
Carbon-fiber composites excel in stiffness
at low weight.
Boron carbide is exceptionally stiff, hard
and light; it is used for body armour.
CFRP
Boron carbide, B4C
E1/3/
= 0.003
E/
=
0.02
Search area
Search area
E = 100
E = 100
Slope 3
M.F. Ashby 2010
11

E5.6 Use a E- chart to establish whether woods have a higher
specific stiffness E/ than epoxies.
Answer. Parallel to the grain, woods have much higher specific

stiffness than epoxies. Perpendicular to the grain, woods have
about the same value as epoxies.
Material
Woods parallel to the grain
Woods transverse to the grain
Epoxies
E5.7
E/ (GPa/(kg/m ) x 10
8 - 29
0.7 4.0
1.8 2.5
when you want strength at low weight (landing gear of aircraft,

mountain bikes). Use a f / chart to decide.
Answer. The guide line for f / on the strength density chart of

Figure 4.4, if drawn through titanium alloys shows that they have a
much the higher specific strength than any steel, even though the
best steel are as strong..
Titanium alloys
Steels
f / (MPa/(kg/m3)) x 103
54 - 270
32 - 190
Answer. The strongest steels, titanium alloys (Ti-alloys) and carbon

fiber reinforced polymers (CFRP) meet these limits.
Do titanium alloys have a higher or lower specific strength

(strength/density, f / ) than the best steels? This is important
Material
Use a modulus-strength E f chart to find materials that

have E > 10 GPa and f 1000 MPa.
E 5.8
Material
High strength steels
High strength nickel alloys
Titanium alloys
CFRP
E5.9
All have f above 1000 MPa,

a very large value
Are the fracture toughness, K1c, of the common polymers

polycarbonate, ABS, or polystyrene larger or smaller than the
engineering ceramic alumina, Al2O3? Are their toughness
G1c = K12c / E larger or smaller? A K1c E chart will help.
Answer. Most polymers have a lower fracture toughness, K1c, than

alumina. Their toughness, G1c = K12c / E , however, are much
larger. Even the most brittle of polymers, polystyrene, has a
toughness G1c that is nearly ten times greater than that of
alumina. The values in the table are read using the K1c axis to
read off values of fracture toughness, and the G1c = K 12c / E
contours to read off values of toughness.
Material
Polycarbonate
ABS
Polystyrene
Alumina, Al2O3
M.F. Ashby 2010
Comment
12
1/2
K1c (MPa.m )
2.1 4.6
1.2 4.3
0.7 1.1
3.3 4.8
G1c (kJ/m )
0.5 - 8
1-8
0.25 0.7
0.04 0.07

1/2
E 5.10 Use a fracture toughness-modulus chart to find materials that

1/2
have a fracture toughness K 1c greater than 100 MPa.m and a
toughness G1c = K 12c / E (shown as contours on Figure 4.7)
3
greater than 10 kJ/m .
Answer. The table lists the results. Only metals have both high
fracture toughness K 1c and high toughness G1c . That is one
reason that they are used for pressure vessels (boilers, submarine
hulls, gas containers etc).
Material
High strength steels
High strength nickel alloys
polymers have, additionally, K 1c > 1 MPa.m . This makes them

attractive for snap-together parts that must flex without failing.
Leather has a particularly high value of resilience, K 1c / E .
E 5.12 One criterion for design of a safe pressure vessel is that it

should leak before it breaks: the leak can be detected and the
pressure released. This is achieved by designing the vessel to
tolerate a crack of length equal to the thickness t of the pressure
vessel wall, without failing by fast fracture. The safe pressure p is
then
Comment
All have high K 1c and high G1c
E5.11 The elastic deflection at fracture (the resilience) of an elasticbrittle solid is proportional to the failure strain, fr = fr / E ,
where fr is the stress that will cause a crack to propagate:
K
fr = 1c
c
Here K 1c is the fracture toughness and c is the length of the
longest crack the materials may contain. Thus
1 K 1c
fr =
c E
Materials that can deflect elastically without fracturing are
therefore those with large values of K 1c / E . Use a K1c E chart
to identify the class of materials with K 1c > 1 MPa.m
values of K 1c / E .
1/2
and high
Answer. Polymers and foams both have large K 1c / E allowing them

to flex without fracturing. They have much higher values K 1c / E
than metals or ceramics, and thus large fracture strains. Only
M.F. Ashby 2010
2
4 1 K Ic
R f
where f is the elastic limit, K 1c is the fracture toughness, R is

the vessel radius. The pressure is maximized by choosing the
material with the greatest value of
M =
K 12c
Use a K 1c f chart to identify three alloys that have particularly

high values of M.
Answer. Alloys with high values of K 12c / y , read from the chart, are
listed below
Material
Low alloy steels
Stainless steels
Copper alloys
Tin
Nickel alloys
13
Comment
Traditional material for pressure vessels.
Used for nuclear pressure vessels.
Small boilers are made of copper.
Tin is used in storage tanks for pharmaceutical
chemical solutions
Reactors for chemical engineering and turbine
combustion chambers are made of nickel based
alloys.

E5.13
A material is required for the blade of a rotary lawnmower. Cost is a consideration. For safety reasons, the designer
specified a minimum fracture toughness for the blade: it is K 1c >
1/2
30 MPa m . The other mechanical requirement is for high
hardness, H, to minimize blade wear. Hardness, in applications
like this one, is related to strength:
E5.14 Use a Thermal conductivity-Electrical resistivity ( e ) chart to

find three materials with high thermal conductivity, , and high electrical
resistivity, e .
Answer. Aluminum nitride is the best choice. The next best are
alumina and silicon carbide.
H 3 y
where f is the strength. Use a K 1c f chart to identify three
1/2
and the highest possible
materials that have K 1c > 30 MPa m
strength. To do this, position a " K 1c " selection line at 30 MPa
1/2
m
and then adjust a "strength" selection line such that it just
admits three candidates. Use a strength cost per volume chart
to rank your selection by material cost, hence making a final
selection.
Answer. Applying the property limit K 1c > 30 MPam1/2, and reading

of materials with f > 800 MPa, gives three groups of materials.
Chart 4.19 identifies the price of each of these per unit volume.
The steels are far cheaper than the other two.
Material
High-strength steels
Nickel alloys
Titanium alloys
Comment
Traditional material for blades
Meets the requirements, but more
expensive than steel
Meets the requirements, but MUCH more
expensive than steel.
M.F. Ashby 2010
Material
Comment
Aluminum nitride, AlN

Alumina, Al2O3
Silicon nitride, Si3N4
Favored material for heat sinks requiring

this combination of properties.
All meet the requirements
E5.15 The window through which the beam emerges from a highpowered laser must obviously be transparent to light. Even then, some
of the energy of the beam is absorbed in the window and can cause it to
heat and crack. This problem is minimized by choosing a window
material with a high thermal conductivity (to conduct the heat away)
and a low expansion coefficient (to reduce thermal strains), that is, by
seeking a window material with a high value of
M = /
Use a chart to identify the best material for an ultra-high powered
laser window.
14

Answer. The chart shows three transparent materials with high / :
soda glass, silica glass and alumina, which, in single crystal form
(sapphire) or ultra-fine grained form (Lucalox) is transparent and hard.
Diamond, not shown on the chart, has an exceptionally high value: it
has been used for ultra high-powered laser windows.
Material
Soda glass
Comment
M = 3 x 105 W/m; poor resistance to thermal
pulse
Silica glass
M = 2 x 106 W/m; much better than soda glass
Alumina (sapphire) M = 3 x 106 W/m; a little better than silica

glass
(Diamond)
M = 3 x 108 W/m; outstanding
E5.17 (a) Use a Youngs modulus-Relative cost (E Cv,R) chart to find

the cheapest materials with a modulus, E, greater than 100 GPa.
(b) Use a Strength-Relative cost ( f C R ) chart to find the cheapest
materials with a strength, f , above 100MPa.
Answer. (a) The two cheap classes of material that meet the
constraints are cast irons and carbon steels.
(b) Cast irons and steels are again the best choice. It is because
of their high stiffness and strength at low cost that they are so
widely used.
E5.16 Use a Maximum service temperature (Tmax) chart to find

0
polymers that can be used above 200 C.
Answer. The chart shows just two classes of polymer with maximum
0
service temperatures greater than 200 C. They are listed below.
Material
Comment
Polytetrafluorethylene, PTFE (Teflon) is used as non-stick
PTFE
coatings for cooking ware, easily surviving
the temperatures of baking and frying.
Silicone elastomers
Silicones are polymers with a Si-O-Si chain
structure instead of the C-C-C chain of
polyolefins. They are more stable than
carbon-based polymers, but expensive.
M.F. Ashby 2010
15

E6. Translation: constraints and objectives
Translation is the task of re-expressing design requirements in terms
that enable material and process selection. Tackle the exercises by
formulating the answers to the questions in this table. Dont try to
model the behavior at this point (that comes in later exercises). Just
think out what the component does, and list the constraints that this
imposes on material choice, including processing requirements.
Function
Constraints
What essential conditions must be

met?
Objective
What is to be maximized or
minimized?
Free
variables
What parameters of the problem is

the designer free to change?
What does component do?
course, minimizing mass and cost are both objectives, and then tradeoff methods are needed.
They come later. For now use judgement to choose the single most
important objective and make all others into constraints.
Two rules-of-thumb, useful in many translation exercises. Many
applications require sufficient fracture toughness for the component
can survive mishandling and accidental impact during service; a totally
brittle material (like un-toughened glass) is unsuitable. Then a
necessary constraint is that of adequate toughness. This is achieved
by requiring that the fracture toughness K 1c > 15 MPa.m 1 / 2 . Other
applications require some ductility, sufficient to allow stress
redistribution under loading points, and some ability to bend or shape
the material plastically. This is achieved by requiring that the (tensile)
ductility f > 2% .
The CES EduPack software can be used to impose the constraints
and to rank the survivors using the objective.
Here it is important to recognize the distinction between constraints and

objectives. As the table says, a constraint is an essential condition that
must be met, usually expressed as a limit on a material or process
attribute. An objective is an quantity for which an extremum (a
maximum or minimum) is sought, frequently cost, mass or volume, but
there are others, several of which appear in the exercises below. Take
the example of a bicycle frame. It must have a certain stiffness and
strength. If it is not stiff and strong enough it will not work, but it is
never required to have infinite stiffness or strength. Stiffness and
strength are therefore constraints that become limits on modulus,
elastic limit and shape. If the bicycle is for sprint racing, it should be as
light as possible if you could make it infinitely light, that would be best
of all. Minimizing mass, here, is the objective, perhaps with an upper
limit (a constraint) on cost. If instead it is a shopping bike to be sold
through supermarkets it should be as cheap as possible the cheaper
it is, the more will be sold. This time minimizing cost is the objective,
possible with an upper limit (a constraint) on mass. For most bikes, of
M.F. Ashby 2010
16

E6.1
A material is required for the windings of an electric air-furnace

o
capable of temperatures up to 1000 C. Think out what attributes a
material must have if it is to be made into windings and function
properly in a furnace. List the function and the constraints; set the
objective to minimize cost and the free variables to choice of
material.
Function
Constraints
Objective
Free variables
Answer. If the material is to be used as windings it must be able to be

drawn to wire and wound to a coil. It must conduct electricity and
o
be able to operate at 1000 C in air. The constraints are tabulated
below.
Function
Constraints
Objective
Free variables
High temperature furnace winding

Maximum service temperature, Tmax > 1000 C
Able to be rolled or drawn to wire
Good electrical conductor
Good resistance to oxidation at elevated
temperature
Minimize material cost
Choice of material
E6.2 A material is required to manufacture office scissors. Paper is an

abrasive material, and scissors sometimes encounter hard
obstacles like staples. List function and constraints; set the
objective to minimize cost and the free variables to choice of
material.
Answer. To resist abrasive wear the scissors must have blades of

high hardness. In cutting, they will sooner or later encounter a
staple or other hard obstruction that would chip a brittle blade
some toughness is required. These two parameters help reduce
wear, but there are other factors that influence it, so it is sensible
to specify good wear resistance. Finally, the scissors must be
formed if the handles are integral with the blades, they must be
forged or stamped from sheet, requiring the ability to be processed
in this way.
M.F. Ashby 2010
E6.3
Scissors
High hardness
Adequate toughness: K 1c > 15 MPa.m 1 / 2
Good wear resistance

Able to be forged
Choice of material
A material is required for a heat exchanger to extract heat

o
from geo-thermally heated, saline, water at 120 C (and thus
under pressure). List function and constraints; set the objective
to minimize cost and the free variables to choice of material.
Answer. The obvious constraints here are those on servicetemperature, corrosion resistance, the ability to conduct heat
well and strength. There are manufacturing constraints too: if
the heat exchanger is to be made from tubes or folded sheet, the
material must be available in these forms, and have sufficient
ductility to allow manufacture.
Function
Constraints
Heat exchanger
Good thermal conductor
Resistance to corrosion in salt water:
good/excellent
Sufficient strength to support the pressure of
the super-heated water
Ability to be rolled to sheet or tube
Adequate ductility to allow shaping, f > 2%
Objective
Free variables
17

Choice of material

E6.4
A C-clamp is required for the processing of electronic

o
components at temperatures up to 450 C. It is essential that the
clamp has as low a thermal inertia as possible so that it reaches
that temperature quickly, and it must not charge-up when exposed
to an electron beam. The time t it takes a component of thickness
x to reach thermal equilibrium when the temperature is suddenly
changed (a transient heat flow problem) is
x2
2a
Function
Constraints

Sufficient strength to carry clamping loads
without failure
Electrical conductor (to prevent charging)
Objective
Maximise the thermal diffusivity, a of the

material
Choice of material
Free variables
C-clamp of low thermal inertia
where the thermal diffusivity a = / C p and is the thermal

conductivity, the density and C p the specific heat.
List function, constraints and objective; set the free variables to
choice of material.
E6.5
A furnace is required to sinter powder-metal parts. It

o
operates continuously at 650 C while the parts are fed through
on a moving belt. You are asked to select a material for furnace
insulation to minimize heat loss and thus to make the furnace as
energy-efficient as possible. For reasons of space the
insulation is limited to a maximum thickness of x = 0.2 m. List
the function, constraints, objective and free variable.
Answer. This is a problem involving steady-state heat flow. The

heat lost by conduction per unit area of insulation per second, q,
is
q =
Figure E1.
Answer. If the C-clamp is to reach temperature quickly it must,

according to the equation given in the question, be made of a
material with as high a thermal diffusivity, a as possible. It carries
loads, so it must have adequate strength, and it must not charge
up, so it must be an electrical conductor.
M.F. Ashby 2010
T
x
where is the thermal conductivity, x the insulation thickness

and T the temperature difference between the interior of the
furnace and its surroundings. The aim is to minimize q, leading
(via the equation) to the objective of minimizing the thermal
conductivity of the insulation material. There are two
constraints: one on thickness, the other on service temperature.
18

Function
Constraints
Insulation for energy-efficient furnace

Insulation thickness x 0.2 m
Objective
Minimize thermal conductivity
Free variables
Choice of material
E6.6 Ultra-precise bearings that allow a rocking motion make use of

knife-edges or pivots. As the bearing rocks, it rolls, translating
sideways by a distance that depends on the radius of contact.
The further it rolls, the less precise is its positioning, so the smaller
the radius of contact R the better. But the smaller the radius of
contact, the greater is the contact pressure (F/A). If this exceeds
the hardness H of either face of the bearing, it will be damaged.
Elastic deformation is bad too: it flattens the contact, increasing
the contact area and the roll.
A rocking bearing is required to operate in a micro-chip fabrication
o
unit using fluorine gas at 100 C, followed by e-beam processing
requiring that all structural parts of the equipment can be earthed
to prevent stray charges. Translate the requirements into material
selection criteria, listing function, constraints, objective and free
variable.
Answer. The objective is to minimize the radius of curvature at the

contact. This is limited by the hardness H and modulus E of the
materials. The objective is met by seeking materials with:
maximum hardness H to enable the smallest possible area of
contact without damage, and maximum modulus E to minimize
elastic flattening of the contact. In a later exercise we return to
this problem of two objectives, treating it by trade-off methods.
Here we choose maximizing hardness as the primary objective,
since deficiency here results in a damaged bearing. We then
treat modulus as a constraint, together with the other obvious
constraints suggested by the design requirements.
Function
Constraints
Objective
Free variables
Rocking bearing
Good electrical conductor
Good resistance to fluorine gas
High modulus, E
Maximize hardness H of bearing faces
Choice of material
E6.7 The standard CD (Jewel case) cracks easily and, if broken, can
scratch the CD. Jewel cases are made of injection molded
polystyrene, chosen because it is transparent, cheap and easy to
mold. A material is sought to make CD cases that do not crack so
easily. The case must still be transparent, able to be injection
molded, and able to compete with polystyrene in cost.
Answer. The question expresses constraints on transparency,
moldablity and fracture toughness (it must be greater than that of
polystyrene). Given these, the cheapest material is the best
choice.
Function
Constraints
Objective
Free variables
Figure E2
M.F. Ashby 2010
19
Improved CD case
Optically transparent
Fracture toughness greater than that of
polystyrene
Able to be injection molded
Choice of material

E6.8 A storage heater captures heat over a period of time, then
releases it, usually to an air stream, when required. Those for
domestic heating store solar energy or energy from cheap off-peak
electricity and release it slowly during the cold part of the day.
Those for research release the heat to a supersonic air stream to
test system behavior in supersonic flight. What is a good material
for the core of a compact storage material capable of
o
temperatures up to 120 C?
Answer. When a material is heated from room temperature To to a
working temperature T, it absorbs heat Q per unit volume where
Q = C p ( T To )
and C p is the specific heat of the material of the core (in J/kg.C)
3
and is its density (in kg/m ). Thus the most compact storage
heater is one made from a material with a high C p . Even a
small storage heater contains a considerable quantity of core (that
is why they are heavy), so it is probable that an objective will be to
minimize its cost per unit volume. If, however, space were critical,
maximizing C p might become the objective.
Function
Core for compact storage heater
High heat capacity per unit volume C

p
Objective
Minimize material cost per unit volume
Free variables
Choice of material
Constraints
M.F. Ashby 2010
20

E7. Deriving and using material indices
The exercises in this section give practice in deriving indices.
(a) Start each by listing function, constraints, objectives and free
variables; without having those straight, you will get in a mess.
Then write down an equation for the objective. Consider whether it
contains a free variable other than material choice; if it does,
identify the constraint that limits it, substitute, and read off the
material index.
(b) If the CES EduPack software is available, use it to apply the
constraints and rank the survivors using the index (start with the
Level 2 database). Are the results sensible? If not, what constraint
has been overlooked or incorrectly formulated?
E7.1 Aperture grills for cathode ray tubes (Figure E3). There are
two types of cathode ray tube (CRT). In the older technology,
colour separation is achieved by using a shadow mask: a thin
metal plate with a grid of holes that allow only the correct beam to
strike a red, green or blue phosphor. A shadow mask can heat up
and distort at high brightness levels (doming), causing the beams
to miss their targets, and giving a blotchy image.
To avoid this, shadow masks are made of Invar, a nickel alloy

with a near-zero expansion coefficient between room temperature
and 150C. It is a consequence of shadow-mask technology that
the glass screen of the CRT curves inward on all four edges,
increasing the probability of reflected glare.
Sonys Trinitron technology overcame this problem and allowed
greater brightness by replacing the shadow mask by an aperture
grill of fine vertical wires, each about 200 m in thickness, that
allows the intended beam to strike either the red, the green or the
blue phosphor to create the image. The glass face of the
Trinitron tube was curved in one plane only, reducing glare.
The wires of the aperture grill are tightly stretched, so that they
remain taut even when hot it is this tension that allows the
greater brightness. What index guides the choice of material to
make them? The table summarises the requirements.
Function
Constraints
Objective
Free variables
Aperture grill for CRT
Wire thickness and spacing specified
Must carry pre-tension without failure
Electrically conducting to prevent charging
Able to be drawn to wire
Maximize permitted temperature rise without

loss of tension
Choice of material
Answer.
The model. A thin, taut wire slackens and sags when the strain
due to thermal expansion,
th = T
exceeds the elastic strain caused by the pre-tension,
pt =
.
E
Figure E3
M.F. Ashby 2010
21

Here is the thermal expansion coefficient of the wire, T the
temperature rise caused by the electron beams that strike it,
the tensile pre-stress in the wire and E its modulus. We wish to
maximize the brightness, and thus T .
The tension is limited by the elastic limit of the wire, f . Inserting
this and writing
pt th
The final choice, using this database, is stainless steel or

tungsten. If the selection is repeated using the more detailed
Level 3 database, carbon fiber emerges as a potential candidate.
Carbon fibre of the desired diameter is available; it conducts well,
both electrically and thermally, it has high strength and best of
all it has almost zero thermal expansion along the fiber
direction.
as the condition that the wire remains taut gives
T =
f
.
E
The result could hardly be simpler. To maximize the brightness,

maximize
M =
f
.
E
There is a second requirement.

The wires must conduct,
otherwise they would charge up, distorting the image. We
therefore require, also, that the material be a good electrical
conductor and that it is capable of being rolled or drawn to wire.
The selection. Applying the constraints listed in the table above
to the CES EduPack Level 1 or 2 database and ranking the
survivors by the index M leads to the selection listed below.
Material
High Carbon steel
Commercially pure
Titanium
Tungsten alloys
Nickel-based alloys
Comment
Carbon steel is ferro-magnetic, so will interact
with the scanning magnetic fields reject.
Extracting titanium from its oxide is difficult
making it an expensive option
A logical choice tungsten has a high melting
point and is routinely produced as fine wire
Many nickel alloys are weakly ferromagnetic
reject for the same reason as carbon steel
M.F. Ashby 2010
E7.2 Material indices for elastic beams with differing constraints

(Figure E4). Start each of the four parts of this problem by listing
the function, the objective and the constraints. You will need the
equations for the deflection of a cantilever beam with a square
cross-section t x t, given in UASSP, 3. The two that matter are
that for the deflection of a beam of length L under an end load
F:
F L3
3 EI
and that for the deflection of a beam under a distributed load f
per unit length:
1 f L4
8 EI
where I = t 4 / 12 . For a self-loaded beam f = A g where is

the density of the material of the beam, A its cross-sectional area
and g the acceleration due to gravity.
(a) Show that the best material for a cantilever beam of given length
L and given (i.e. fixed) square cross-section (t x t) that will
deflect least under a given end load F is that with the largest
value of the index M = E, where E is Young's modulus (neglect
self-weight). (Figure E4a.)
(b) Show that the best material choice for a cantilever beam of given
length L and with a given section (t x t) that will deflect least
under its own self-weight is that with the largest value of M =
E/, where is the density. (Figure E4b.)
22

(c) Show that the material index for the lightest cantilever beam of
length L and square section (not given, i.e., the area is a free
variable) that will not deflect by more than under its own weight is
M = E / 2 . (Figure E4c.)
(d) Show that the lightest cantilever beam of length L and square
section (area free) that will not deflect by more than under an
end load F is that made of the material with the largest value of
M = E 1 / 2 / (neglect self weight). (Figure E4d.)
Answer.
The model. The point of this problem is that the material index
depends on the mode of loading, on the geometric constraints
and on the design objective.
(a) The table lists the design requirements for part (a) of the
problem.
Function
End-loaded cantilever beam
Constraints
Length L specified
Section t x t specified
End load F specified
Objective
Minimize the deflection,
Free variables
Choice of material only
The objective function is an equation for the deflection of the

beam. An end-load F produces a deflection of
F L3
3 EI
where E is the modulus of the beam material and I = t 4 / 12

is the second moment of the area, so that the deflection
becomes
= 4
F L3 1

t4 E
The magnitude of the load F and the dimensions L and t are

all given. The deflection is minimized by maximizing
M1 = E
(b) The design requirements for part (b) are listed below
Figure E4
M.F. Ashby 2010
Function
Self-loaded cantilever beam
Constraints
Length L specified
Section t x t specified
Objective
Minimize the deflection,
Free variables
Choice of material only
23

The beam carries a distributed load, f per unit length, where
f = gt2
where is the density of the beam material and g is the
acceleration due to gravity. Such a load produces a deflection
(UASSP, 3)
3 f L4
=
2 E t4
3 g L4
=

2t2 E
Substituting for t (the free variable) from the second equation

into the first, gives
m =
3 g L5 2
2 E
The quantities L and are given. The mass is minimized by

maximizing
M3 =
(the objective function). As before, t and L are given. The

deflection is minimized by maximizing
M2 =
(d) The design requirements for part (d) are listed below
Function
End-loaded cantilever beam
Constraints
Length L specified
End-load F specified
Maximum deflection, , specified
(c) The design requirements for part (c) are listed below
Function
Self-loaded cantilever beam
Objective
Minimize the mass, m
Constraints
Length L specified
Maximum deflection, , specified
Free variables
Choice of material
Objective
Minimize the mass, m
Section area A = t 2
Free variables
Choice of material
Section area A = t 2
The section is square, but the dimension t is free. The objective

function is
m = t2L
The beam deflects under its own weight but now the section can
be varied to reduce the weight provided the deflection does not
exceed , as in the figure. The objective function (the quantity to
be minimized) is the mass m of the beam
The deflection is, as in part (a)
= 4
F L3 1

t4 E
Using this to eliminate the free variable, t, gives
m = t2 L
F L5
m = 2

M.F. Ashby 2010
24
1/ 2
1/ 2
E

The quantities F , and L are given. The mass is minimized by
maximizing
M4 =
E1/ 2
From a selection standpoint, M3 and M4 are equivalent.

The selection. Applying the three indices to the CES EduPack
Level 1 or 2 database gives the top-ranked candidates listed
below.
Index
Material choice
Metals: tungsten alloys, nickel alloys, steels.
Ceramics: SiC, Si3N4, B4C, Al2O3 and AlN, but of
course all are brittle.
Metals: aluminum, magnesium, nickel and
titanium alloys and steels all have almost the
same value of E /
High M1 = E
High M 2 =
High M 3 =
High M 4 =
2
E
1/ 2
E7.3 Material index for a light, strong beam (Figure E5). In

stiffness-limited applications, it is elastic deflection that is the
active constraint: it limits performance. In strength-limited
applications, deflection is acceptable provided the component
does not fail; strength is the active constraint. Derive the material
index for selecting materials for a beam of length L, specified
strength and minimum weight. For simplicity, assume the beam to
have a solid square cross-section t x t. You will need the equation
for the failure load of a beam (UASSP, 4). It is
Ff =
where y m is the distance between the neutral axis of the beam

and its outer filament and I = t 4 / 12 = A 2 / 12 is the second
moment of the cross-section. The table itemizes the design
requirements.
Composites: CFRP
All technical ceramics
Metals: aluminum and magnesium alloys superior
to all other metals.
Composites: CFRP excels
Ceramics: SiC, Si3N4, B4C, Al2O3 and AlN
Figure E5
Function
Constraints
Objective
Free variables
M.F. Ashby 2010
I f
ym L
25
Beam
Length L is specified
Beam must support a bending load F

without yield or fracture
Minimize the mass of the beam
Cross-section area, A
Choice of material

Answer.
The model. The objective is to minimize the mass, giving the
objective function
m = AL
Inverting the equation given in the question leads to an expression

for the area A that will support the design load F:
F L
A = 6
f
2/3
Substituting A into the objective function gives the mass of the

beam that will just support the load F f :

m = (6 F )2/3 L5 / 3
2/3
f
The mass is minimized by selecting materials with the largest
values of the index
M=
The selection. Applying the constraint on K 1c and ranking by the

index M using the CES EduPack Level 1 or 2 database gives the
top-ranked candidates listed below.
Material
CFRP
Metals: titanium,
aluminum and
magnesium alloys
Comment
Exceptionally good, mainly because of
its very low density.
Here the light alloys out-perform steel
E7.4 Material index for a cheap, stiff column (Figure E6).

In the last two exercises the objective has been that of minimizing
weight. There are many others. In the selection of a material for a
spring, the objective is that of maximizing the elastic energy it can
store. In seeking materials for thermal-efficient insulation for a
furnace, the best are those with the lowest thermal conductivity and
heat capacity. And most common of all is the wish to minimize
cost. So here is an example involving cost.
2/3
f
If the cantilever is part of a mechanical system it is important that it

have sufficient fracture toughness to survive accidental impact
loads. For this we add the requirement of adequate toughness:
Columns support compressive loads: the legs of a table; the pillars

of the Parthenon. Derive the index for selecting materials for the
cheapest cylindrical column of specified height, H, that will safely
support a load F without buckling elastically. You will need the
equation for the load Fcrit at which a slender column buckles. It is
Fcrit =
K 1c > 15 MPa .m 1 / 2
n2 2 E I
H2
where n is a constant that depends on the end constraints and
I = r 4 / 4 = A 2 / 4 is the second moment of area of the

column(see UASSPfor both). The table lists the requirements.
M.F. Ashby 2010
26
Function
Constraints
Objective
Free variables
The selection. The loading here is compressive, so brittle

materials are viable candidates. Since some are also very cheap,
they dominate the selection. Applying the index M to the CES
EduPack Level 1 or 2 database gives the top-ranked candidates
listed below.
Cylindrical column
Length L is specified
Column must support a

compressive load F without
buckling
Minimize the material cost

of the column
Cross-section area, A
Choice of material
Material
Ceramics: brick,
concrete, cement and
stone
Wood
Carbon steel, cast iron
Figure E6
Answer.
The model. A slender column uses less material than a fat one,
and thus is cheaper; but it must not be so slender that it will buckle
under the design load, F. The objective function is the cost
C = A H Cm
*
Where C m is the cost/kg of the material of the column. It will

buckle elastically if F exceeds the Euler load, Fcrit , given in the
question. Eliminating A between the two equations, using the
definition of I, gives:
4F
C
1/2
C
H 2 m
E 1/2
Comment
The low cost and fairly high modulus
makes these the top-ranked candidates.
Exceptional stiffness parallel to the grain,
and cheap.
Steel out-performs all other metals when
strength at low cost is sought.
E7.5 Indices for stiff plates and shells (Figure E7). Aircraft and
space structures make use of plates and shells. The index
depends on the configuration. Here you are asked to derive the
material index for
(a) a circular plate of radius a carrying a central load W with a
prescribed stiffness S = W / and of minimum mass, and
(b) a hemispherical shell of radius a carrying a central load W with
a prescribed stiffness S = W / and of minimum mass, as shown
in the figures.
The material cost of the column is minimized by choosing

materials with the largest value of the index
M=
E1 2
Cm
Cm is the cost/kg of the processed material, here, the material in the form of a
circular rod or column.
M.F. Ashby 2010
27

Use the two results listed below for the mid-point deflection of a
plate or spherical shell under a load W applied over a small
central, circular area.
Circular plate:
3 Wa
3 +
( 1 2 )
4 E t3
1 +
Wa
Et2
( 1
in which A 0.35 is a constant. Here E is Youngs modulus, t is

the thickness of the plate or shell and is Poissons ratio.
Poissons ratio is almost the same for all structural materials and
can be treated as a constant. The table summarizes the
requirements.
Constraints
Objective
Free variables
3 S a2
m = a2
Hemispherical shell = A
Function
Inserting this into the equation for the mass gives
Stiff circular plate, or

Stiff hemispherical shell
Stiffness S under central load W specified
Radius a of plate or shell specified
Minimize the mass of the plate or shell
Plate or shell thickness, t
Choice of material
Answer.
(a) The plate. The objective is to minimize the mass, m
3 S a2
t =
4 E

1 / 3 f ( )
E
The lightest plate is that made from a material with a large value of
the index
M1 =
E1/ 3
(a) The hemispherical shell The objective again is to minimize the

mass, m
m = 2 a2 t
Inverting the second equation in the question gives, for the shell,
Sa
t = A
1/ 2
f 2 ( )
where f 2 ( ) as before is a function of , and thus a constant.

Inserting this into the equation for the mass gives

m = 2 a 2 ( A S a )1 / 2
f ( )
E1/ 2
m = a2 t
where is the density of the material of which the plate is made.
The thickness t is free, but must be sufficient to meet the
constraint on stiffness. Inverting the first equation in the question
gives, for the plate,
1/ 3
The lightest shell is that made from a material with a large value of
the index
M2 =
1/ 3
f 1 ( )
E1/ 2
where f 1 ( ) is simply a function of , and thus a constant.
M.F. Ashby 2010
28

The index for the shell differs from that for the plate, requiring a
different choice of material. This is because the flat plate, when loaded,
deforms by bending. The hemispherical shell, by contrast, carries
membrane stresses (tension and compression in the plane of the shell
wall), and because of this is much stiffer. Singly-curved shells behave
like the plate, doubly-curved shells like the hemisphere.
The selection. Applying the three indices to the CES EduPack
Level 1 or 2 database gives the top-ranked candidates listed
below.
Index
High M 1 =
High M 2 =
Material choice
E1/ 3
1/ 2
Natural materials: wood and plywood

Composites: CFRP
Metals: aluminum and magnesium alloys
superior to all other metals.
Composites: CFRP
Ceramics: SiC, Si3N4, B4C and AlN
E7.6 The C-clamp in more detail (Figure E8). Exercise E4.4

introduced the C-clamp for processing of electronic components.
The clamp has a square cross-section of width x and given depth
b. It is essential that the clamp have low thermal inertia so that it
reaches temperature quickly. The time t it takes a component of
thickness x to reach thermal equilibrium when the temperature is
suddenly changed (a transient heat flow problem) is
x2
t
2a
where the thermal diffusivity a = / C p and is the thermal
conductivity, the density and C p the specific heat.
M.F. Ashby 2010
Figure E8
The time to reach thermal equilibrium is reduced by making the
section x thinner, but it must not be so thin that it fails in service.
Use this constraint to eliminate x in the equation above, thereby
deriving a material index for the clamp. Use the fact that the
clamping force F creates a moment on the body of the clamp of
M = F L , and that the peak stress in the body is given by
xM
2 I
where I = b x 3 / 12 is the second moment of area of the body.

The table summarizes the requirements.
Function
Constraints
Objective
Free variables
29
C-clamp of low thermal inertia

Depth b specified
Must carry clamping load F without failure
Minimize time to reach thermal equilibrium
Width of clamp body, x
Choice of material

Answer. The clamp will fail if the stress in it exceeds its elastic
limit f . Equating the peak stress to f and solving for x gives
1/ 2
FL
x = 6
b f
Inserting this into the equation for the time to reach equilibrium
gives
F L 1
t =3
b a f
The time is minimized by choosing materials with large values of
the index
E7.7 Springs for trucks. In vehicle suspension design it is desirable

to minimize the mass of all components. You have been asked to
select a material and dimensions for a light spring to replace the
steel leaf-spring of an existing truck suspension. The existing
leaf-spring is a beam, shown schematically in the figure. The
new spring must have the same length L and stiffness S as the
existing one, and must deflect through a maximum safe
displacement max without failure. The width b and thickness t are
free variables.
M = a f .
Additional constraints on modulus E > 50 GPa (to ensure that the
clamp is sufficiently stiff) on fracture toughness K 1c > 18 MPa.m 1 / 2
(to guard against accidental impact) and on formability will, in
practice, be needed.
The selection. Applying the constraint on K 1c and formability,

and ranking by the index M using the CES EduPack Level 1 or 2
database gives the top-ranked candidates listed below.
Material
Aluminum alloys
Copper alloys
Figure E9
Derive a material index for the selection of a material for this

application. Note that this is a problem with two free variables: b
and t; and there are two constraints, one on safe deflection max and
the other on stiffness S. Use the two constraints to fix free
variables. The table catalogs the requirements.
Comment
The obvious candidate good thermal
conductor, adequately stiff and strong, and easy
to work.
Here the high thermal diffusivity of copper is
dominating the selection.
Function
Constraints
Leaf spring for truck

Length L specified
Stiffness S specified
Maximum displacement max specified
Objective
Free variables
Minimize the mass

Spring thickness t
Spring width b
Choice of material
M.F. Ashby 2010
30

You will need the equation for the mid-point deflection of an elastic
beam of length L loaded in three-point bending by a central load F:
Equations (2) and (3) can now be solved for t and b, and these
substituted back into (1). The result is
E
2
m = 9 S max
L
2f
1 F L3
=
48 E I
and that for the deflection at which failure occurs
The mass of the spring is minimized by maximizing the index
2
1f L
max =
6 tE
M =
where I is the second moment of area; for a beam of rectangular

section, I = b t 3 / 12 and E and f are the modulus and failure
stress of the material of the beam. (See UASSP.)
2f
E
Additional constraints on fracture toughness K 1c > 15 MPa.m 1 / 2

(to guard against accidental impact) and on formability will, in
practice, be needed.
The selection. Applying the constraint on K 1c and formability,
and ranking by the index M using the CES EduPack Level 1 or 2
database gives the top-ranked candidates listed below.
Answer.
The model. The objective function the quantity to be minimized
is the mass m of the spring:
m = bt L
(1)
Material
Elastomers (rubber)
Comment
Oops! we have missed a constraint here.
Elastomers excel as light springs, but the
constraint on thickness t and depth b in this
application translates via equation (2) into an
additional constraint on modulus:
E > SL3 / 4bt 3 .
Titanium alloys
An expensive solution, but one that is lighter

than steel.
CFRP makes exceptionally good light springs.
The standard solution, but one that is heavier
than the others above.
where is its density. The length L is fixed. The dimensions b

and t are free. There are two constraints. The first is a required
stiffness, S. From the first equation given in the question
S =
= 48
EI
L3
= 4
E bt3
L3
(2)
The second constraint is that of a maximum allowable

displacement max without damage to the spring, given in the
question as
2
1f L
max =
6 tE
CFRP
High carbon steel
(3)
M.F. Ashby 2010
31

E7.8 Fin for a rocket (Figure E10). A tube-launched rocket has
stabilizing fins at its rear. During launch the fins experience hot
o
gas at T g = 1700 C for a time t = 0.3 seconds. It is important that
the fins survive launch without surface melting. Suggest a
material index for selecting a material for the fins. The table
summarizes the requirements.
If the heating time is small compared with the characteristic time

for heat to diffuse through the fin, a quasi steady-state exists in
which the surface temperature adjusts itself such that the heat
entering from the gas is equal to that diffusing inwards by
conduction. This second is equal to
q =
(Ts Ti )
x
where is the thermal conductivity, Ti is the temperature of the

(cold) interior of the fin, and x is a characteristic heat-diffusion
length. When the heating time is short (as here) the thermal front,
after a time t, has penetrated a distance
x (2a t )1 / 2
Figure E10
where a = / C p is the thermal diffusivity. Substituting this

Function
Constraints
Objective
Free
variables
High heat-transfer rocket fins

All dimensions specified
Must not suffer surface melting during
exposure to gas at 1700 C for 0.3 sec.
Minimize the surface temperature rise
during firing
Maximize the melting point of the material
Choice of material
This is tricky. Heat enters the surface of the fin by transfer from
the gas. If the heat transfer coefficient is h, the heat flux per unit
area is
q = h ( T g Ts )
where Ts is the surface temperature of the fin the critical
quantity we wish to minimize. Heat is diffuses into the fin surface
by thermal conduction.
M.F. Ashby 2010
value of x in the previous equation gives
q = C p 1/ 2
(Ts Ti )
2t
where is the density and C p the specific heat of the material

of the fin.
Proceed by equating the two equations for q, solving for the
surface temperature Ts to give the objective function. Read off
the combination of properties that minimizes Ts ; it is the index for
the problem.
The selection is made by seeking materials with large values of
the index and with a high melting point, Tm . If the CES software
is available, make a chart with these two as axes and identify
materials with high values of the index that also have high melting
points.
32

Answer. Equating the equations and solving for Ts gives
Ts =
h Tg
t +
C p Ti
t +
Cp
When t = 0 the surface temperature T s = Ti and the fin is completely

cold. When t is large, T s = T g and the surface temperature is equal to
the gas temperature. For a given t, T s is minimized by maximizing
C p . The first index is therefore
M1 = C p
and index that often appears in problems involving transient heat flow.
Melting is also made less likely by choosing a material with a high
melting point Tm . The second index is therefore
M 2 = Tm
The selection. The figure shows the two indices, created using the
CES EduPack Level 2 database. The top-ranked candidates are listed
below.
Material
Silicon carbide, SiC
Copper alloys
Aluminum alloys
Aluminum Silicon
carbide MMC
Comment
Silicon carbide out-performs all metals except
tungsten, and is much lighter. If used, its
brittleness would have to be reckoned with..
The exceptional thermal conductivity of copper
is dominating here it is able to conduct heat
away from the surface quickly, limiting the
surface heating.
An attractive choice, since Al-alloys are also
light
This metal matrix composite has almost the
thermal conductivity of aluminum and is stiffer
and stronger.
M.F. Ashby 2010
33

E8 Multiple constraints and objectives
Over-constrained problems are normal in materials selection. Often it
is just a case of applying each constraint in turn, retaining only those
solutions that meet them all. But when constraints are used to
eliminate free variables in an objective function, the active constraint
method must be used. The first three exercises in this section illustrate
problems with multiple constraints.
The remaining two concern multiple objectives and trade-off methods.
When a problem has two objectives minimizing both mass m and
cost C of a component, for instance a conflict arises: the cheapest
solution is not the lightest and vice versa. The best combination is
sought by constructing a trade-off plot using mass as one axis, and
cost as the other. The lower envelope of the points on this plot defines
the trade-off surface. The solutions that offer the best compromise lie
on this surface. To get further we need a penalty function. Define the
penalty function
E8.1 Multiple constraints: a light, stiff, strong tie (Figure E11).

A tie, of length L loaded in tension, is to support a load F, at
minimum weight without failing (implying a constraint on strength)
or extending elastically by more than (implying a constraint on
stiffness, F / ). The table summarises the requirements.
Figure E11
Function
Tie rod
Constraints
Must not fail by yielding under force F

Must have specified stiffness, F/
Length L and axial load F specified
Minimize mass m
Section area A
Choice of material
Z = C + m
where is an exchange constant describing the penalty associated

with unit increase in mass, or, equivalently, the value associated with a
unit decrease. The best solutions are found where the line defined by
this equation is tangential to the trade-off surface. (Remember that
objectives must be expressed in a form such that a minimum is sought;
then a low value of Z is desirable, a high one is not.)
When a substitute is sought for an existing material it is better to work
with ratios. Then the penalty function becomes
C
m
+ *
Co
mo
in which the subscript o means properties of the existing material and
the asterisk * on Z* and * is a reminder that both are now
*
dimensionless. The relative exchange constant measures the
fractional gain in value for a given fractional gain in performance.
Z* =
M.F. Ashby 2010
Objective
Free variables
(a) Establish two performance equations for the mass, one for
each constraint, from which two material indices and one coupling
equation linking them are derived. Show that the two indices are
M1 =
and
M2 =
and that a minimum is sought for both.

(b) Use the CES EduPack to produce a graph which has the
indices as axes, to identify candidate materials for the tie (i)
-3
-2
when /L = 10 and (ii) when /L = 10 , like in Figure E12.
34
Figure E12
Answer. The derivation of performance equations and the indices they contain is laid out here:
Objective
Constraints
Performance equation
Index
Substitute for A
Stiffness constraint
E A
L
F
m1 = L2
E
M1 =

m2 = L F
y
M2 =
(1)
m= AL
Substitute for A
Strength constraint F = y A
M.F. Ashby 2010
35
(2)

(The symbols have their usual meanings: A = area, L= length,
= density, F/ =stiffness, E = Youngs modulus, y = yield
Coupling
condition
Material choice
Comment
strength or elastic limit.)
L/ = 100
Ceramics: boron carbide,

silicon carbide
L/ = 1000
Composites: CFRP; after

that, Ti, Al and Mg alloys
These materials are

available as fibers as well
as bulk.
If ductility and toughness
are also required, the
metals are the best choice.
The coupling equation is found by equating m1 to m2, giving
= L
defining the coupling constant Cc = L/. The chart below shows

the positions of the coupling line when L/ = 100 and when L/ =
3
10 (corresponding to the required values of /L in the question)
and the materials that are the best choice for each.
Coupling
lines
The use of ceramics for a tie, which must carry tension, is

normally ruled out by their low fracture toughness even a small
flaw can lead to brittle failure. But in the form of fibers both boron
carbide and silicon carbide are used as reinforcement in
composites, where they are loaded in tension, and their stiffness
and strength at low weight are exploited.
The CES EduPack software allows the construction of charts
with axes that are combinations of properties, like those of / E
and / y shown here, and the application of a selection box to
identify the optimum choice of material.
= 10 - 3
L
= 10 -2
L
M.F. Ashby 2010
36

E8.2 Multiple constraints: a light, safe, pressure vessel (Figure
E13) When a pressure vessel has to be mobile; its weight
becomes important. Aircraft bodies, rocket casings and liquidnatural gas containers are examples; they must be light, and at the
same time they must be safe, and that means that they must not
fail by yielding or by fast fracture. What are the best materials for
their construction? The table summarizes the requirements.
where p , the pressure difference across this wall, is fixed by the

design. The first constraint is that the vessel should not yield
that is, that the tensile stress in the wall should not exceed y. The
second is that it should not fail by fast fracture; this requires that
the wall-stress be less than K 1c / c , where K 1c is the fracture
toughness of the material of which the pressure vessel is made
and c is the length of the longest crack that the wall might contain.
Use each of these in turn to eliminate t in the equation for m; use
the results to identify two material indices
M1 =
and
M2 =
K 1c
and a coupling relation between them. It contains the crack length,

c.
Figure E13
Function
Constraints
Pressure vessel
Must not fail by yielding
Must not fail by fast fracture.
Diameter 2R and pressure

difference p specified
Objective
Free variables
(b) Use the EduPack to produce a graph with the two material
indices as axes, like in Figure E14. The coupling equation
expresses the relationship between M1 and M2 and therefore Log
M1 and Log M2 and can be plotted as a straight line on the Log
M1Log M2 chart. Determine the gradient and intercept of this line
and plot it for first c= 5mm and then c= 5 m. Identify the lightest
candidate materials for the vessel for each case. M1 and M2 need
to be minimised to find the lightest material.
Minimize mass m
Wall thickness, t
Choice of material
(a) Write, first, a performance equation for the mass m of the

pressure vessel. Assume, for simplicity, that it is spherical, of
specified radius R, and that the wall thickness, t (the free variable)
is small compared with R. Then the tensile stress in the wall is
p R
2t
M.F. Ashby 2010
37
Figure E14
M.F. Ashby 2010
38

Answer. The objective function is the mass of the pressure vessel:
m = 4 R 2t
The tensile stress in the wall of a thin-walled pressure vessel (UASSP, 11) is
p R
2t
Equating this first to the yield strength y , then to the fracture strength K 1c / c and substituting for t in
the objective function leads to the performance equations and indices laid out below.
Objective
Constraints
Yield constraint
pR
2 t
m1 = 2 p.R 3
y
Index
M1 =
(1)
Substitute for t
m = 4 R 2t
Fracture constraint
pR
2 t
K 1c

m 2 = 2 p R 3 ( c )1 / 2
K 1c
M2 =
(2)
K 1c
Substitute for t
The coupling equation is found by equating m1 to m2, giving a relationship between M1 and M2:
M 1 = ( c)1/2 M 2
On the (logarithmic) graph of the material indices, the coupling line therefore is log M1 = log ( c)
1/2
+ log M2
1/2
The position of the coupling line depends on the detection limit, c1 for cracks, through the term ( c) .
M.F. Ashby 2010
39

The figure shows the appropriate chart with two coupling lines, one
for c = 5 mm and the other for c = 5m.
Coupling line,
c = 5 microns
The resulting selection is summarised in the table.
Coupling condition
Crack length
c 5 mm
Material choice
from Figure E14
Titanium alloys
Aluminium alloys
Steels
( c = 0.125 )
Crack length
c 5 m
( c = 3.96 x 10 3 )
CFRP
Silicon carbide
Silicon nitride
Alumina
Coupling line,
c = 5 mm
Comment
These are the standard
materials for pressure
vessels. Steels appear,
despite their high density,
because their toughness
and strength are so high
Ceramics, potentially, are
attractive structural
materials, but the difficulty
of fabricating and
maintaining them with no
flaws greater than 5 m is
enormous
Note: using the EduPack will reveal a wider range of candidate

materials along the coupling lines.
In large engineering structures it is difficult to ensure that there are

no cracks of length greater than 1 mm; then the tough engineering
alloys based on steel, aluminum and titanium are the safe choice. In
the field of MEMS (micro electro-mechanical systems), in which films
of micron-thickness are deposited on substrates, etched to shapes
and then loaded in various ways, it is possible even with brittle
ceramics to make components with no flaws greater than 1 m in
size. In this regime, the second selection given above has
relevance.
M.F. Ashby 2010
40

E8.3 A cheap column that must not buckle
or crush (Figure E15). The best choice of
material for a light strong column depends
on its aspect ratio: the ratio of its height H
to its diameter D. This is because short, fat
columns fail by crushing; tall slender
columns buckle instead. Derive two
performance equations for the material cost
of a column of solid circular section and
specified height H, designed to support a
load F large compared to its self-load, one
using the constraints that the column must
not crush, the other that it must not buckle.
The table summarizes the needs.
Function
Constraints
Objective
Free variables
C
M 1 = m
c
and
C
M2 = m
E 1/2
where C m is the material cost per kg, the material density, c

its crushing strength and E its modulus.
(b) Data for six possible candidates for the column are listed in
5
the Table. Use these to identify candidate materials when F = 10
N and H = 3m. Ceramics are admissible here, because they have
high strength in compression.
Data for candidate materials for the column
Figure E15
Column
Must not fail by compressive crushing
Must not buckle
Height H and compressive load F specified.
Minimize material cost C
Diameter D
Choice of material
(a) Proceed as follows

(1) Write down an expression for the material cost of the column its
mass times its cost per unit mass, Cm.
(2) Express the two constraints as equations, and use them to
substitute for the free variable, D, to find the cost of the column
that will just support the load without failing by either mechanism
Material
Density
3
(kg/m )
Cost/kg
Cm ($/kg)
Modulus
E (MPa)
Wood (spruce)
Brick
Granite
Poured concrete
Cast iron
Structural steel
Al-alloy 6061
700
2100
2600
2300
7150
7850
2700
0.5
0.35
0.6
0.08
0.25
0.4
1.2
10,000
22,000
20,000
20,000
130,000
210,000
69,000
Compression
strength
c (MPa)
25
95
150
13
200
300
150
(c) Use the EduPack to produce a graph with the two indices as
axes. Express M2 in terms of M1 and plot coupling lines for
5
selecting materials for a column with F = 10 N and H = 3m (the
3
same conditions as above), and for a second column with F = 10
N and H = 20m.
(3) Identify the material indices M1 and M2 that enter the two equations
for the mass, showing that they are
M.F. Ashby 2010
41

Answer. This, and exercises E 8.1 and E 8.2 illustrate the method of
solving over-constrained problems. This one concerns materials for a
light column with circular section which must neither buckle nor crush
under a design load F.
The cost, C, is to be minimised
C =
D 2 H Cm
where D is the diameter (the free variable) and H the height of the
column, Cm is the cost per kg of the material and is its density. The
column must not crush, requiring that
4F
D2
where c is the compressive strength. Nor must it buckle:
2 EI
H2
The right-hand side is the Euler buckling load in which E is Youngs

modulus. The second moment of area for a circular column is
Figure E16
I=
4
D
64
The subsequent steps in the derivation of performance equations are

laid out on the next page:
M.F. Ashby 2010
42

Objective
Constraint
Substitute for D
Crushing constraint
C =
Ff
D2
4
C
C1 = F H m
c
(1)
D 2 H Cm
Substitute for D
Buckling constraint F
2 EI
H2
3 D4 E
64 H 2
C2 =
1/2
C
F 1/2 H 2 m
E 1/2
(2)
C
The first performance equation contains the index M 1 = m , the second, the index
c
C
~
M 2 = m . This is a min-max problem: we seek the material with the lowest (min) cost C
1/2
E
which itself is the larger (max) of C1 and C2. The two performance equations are evaluated in
~
the Table, which also lists C = max ( C 1 , C 2 ). for a column of height H = 3m, carrying a load F =
5
10 N. The cheapest choice is concrete.
Material
Wood (spruce)
Brick
Granite
Poured concrete
Cast iron
Structural steel
Al-alloy 6061
Density
3
(kg/m )
Cost/kg
Cm ($/kg)
Modulus
E (MPa)
700
2100
2600
2300
7150
7850
2700
0.5
0.35
0.6
0.08
0.25
0.4
1.2
10,000
22,000
20,000
20,000
130,000
210,000
69,000
M.F. Ashby 2010
Compressio
n
strength
c (MPa)
25
95
150
13
200
300
150
43
C1
$
C2
$
~
C
4.2
2.3
3.1
4.3
2.6
3.0
6.5
11.2
16.1
35.0
4.7
16.1
21.8
39.5
11.2
16.1
35.0
4.7
16.1
21.8
39.5

The coupling equation is found by equating C1 to C2 giving
M2 =
1/2
2
F

H2
1/ 2
M1
log
M 2 = log M 1 + log
1/2
2
F
2
H
1/ 2
It contains the structural loading coefficient F/H . Two positions for the coupling line are shown, one
2
2
5
2
corresponding to a low value of F/H = 0.011 MN/m (F = 10 N, H = 3 m) and to a high one F/H =
2
7
2.5 MN/m (F = 10 N, H = 2 m), with associated solutions. Remember that, since E and c are
measured in MPa, the load F must be expressed in units of MN. The selection is listed in the table.
Couplin
g
Large F/H
Small F/H
M.F. Ashby 2010
44

E8.4 An air cylinder for a truck (Figure E17). Trucks rely on
compressed air for braking and other power-actuated systems. The air
is stored in one or a cluster of cylindrical pressure tanks like that shown
here (length L, diameter 2R, hemispherical ends). Most are made of
low-carbon steel, and they are heavy. The task: to explore the
potential of alternative materials for lighter air tanks, recognizing the
there must be a trade-off between mass and cost if it is too
expensive, the truck owner will not want it even if it is lighter. The table
summarizes the design requirements.
Function
Constraints
Objectives
Free variables
Air cylinder for truck
Must not fail by yielding
Diameter 2R and length L specified, so the

ratio Q = 2R/L is fixed.
Minimize mass m
Minimize material cost C
Wall thickness, t
Choice of material
where is the density, y the yield strength and Cm the cost per kg
of the material, and the subscript o indicates values for mild
steel.
(b) Explore the trade-off between relative cost and relative mass,
considering the replacement of a mild steel tank with one made,
first, of low alloy steel, and, second, one made of filament-wound
CFRP, using the material properties in the table below. Define a
relative penalty function
Z* = *
m
C
+
mo
Co
where is a relative exchange constant, and evaluate Z for =

*
1 and for = 100.
Material
Mild steel
Low alloy steel
CFRP
Density
3
(kg/m )
7850
7850
1550
Yield strength
c (MPa)
314
775
760
Price per /kg

Cm ($/kg)
0.66
0.85
42.1
(c) Use the EduPack to produce a graph with axes of m/mo and C/Co,
like the one in Figure E18 below. Mild steel (here labelled Low
carbon steel) lies at the co-ordinates (1,1).
*
Sketch a trade-off surface and plot contours of Z that are

*
*
approximately tangent to the trade-off surface for = 1 and for
= 100. What selections do these suggest?
Figure E17
(a) Show that the mass and material cost of the tank relative to one
made of low-carbon steel are given by

m
=
y
mo
y ,o
and
y ,o
C
= Cm
y C m ,o o
Co
M.F. Ashby 2010
45

m =

2R 2 L p ( 1 + Q )
y
The material cost C is simply the mass m times the cost per kg of the
material, Cm, giving
C
C = C m m = 2R 2 L p ( 1 + Q ) m
y
from which the mass and cost relative to that of a low-carbon steel
(subscript o) tank are

m
=
y
mo
y ,o
y ,o
C
= Cm
y C m ,o o
Co
and
(b) To get further we need a penalty function:. The relative penalty

function
Z* = *
Figure E18
Answer. (a) The mass m of the tank is
m = 2R L t + 4 R 2 t = 2R L t ( 1 + Q )
where Q, the aspect ratio 2R/L, is fixed by the design requirements. The
stress in the wall of the tank caused by the pressure p must not exceed y ,
is the yield strength of the material of the tank wall, meaning that
pR
y
t
Substituting for t, the free variable, gives
M.F. Ashby 2010
m
C
+
mo
Co
This is evaluated for Low alloy steel and for CFRP in the table below, for
* = 1 (meaning that weight carries a low cost premium) Low alloy
steel has by far the lowest Z*. But when it is evaluated for * = 100
(meaning that weight carriers a large cost premium), CFRP has the
lowest Z*.
(c) The figure shows the trade-off surface. Materials on or near this
surface have attractive combinations of mass and cost. Several are
better low-carbon steel. Two contours of Z* that just touch the trade-off
line are shown, one for * = 1, the other for * = 100 they are curved
because of the logarithmic axes.
The first, for * = 1 identifies higher strength steels as good choices.
This is because their higher strength allows a thinner tank wall. The
contour for * = 100 touches near CFRP, aluminum and magnesium
alloys if weight saving is very highly valued, these become attractive
solutions.
46

Material
Mild steel
Low alloy steel
CFRP
Density
3
(kg/m )
7850
7850
1550
Yield strength
c (MPa)
314
775
760
Z* with * = 1
Price per /kg

Cm ($/kg)
0.66
0.85
42.1
Z,
=1
2
1.03
5.2
*
Z,
= 100
101
45.6
13.4
*
Z* with * = 100
M.F. Ashby 2010
47

E8.5 Insulating walls for freezers
(Figure E19). Freezers and
refrigerated trucks have panelwalls that provide thermal
insulation, and at the same
time are stiff, strong and light
(stiffness to suppress vibration,
strength to tolerate rough
usage). To achieve this the
panels are usually of sandwich
construction, with two skins of
steel, aluminum or GFRP
(providing the strength)
separated by, and bonded to,
a low density insulating core.
In choosing the core we seek
to minimize thermal
conductivity, , and at the
same time to maximize
stiffness, because this allows
thinner steel faces, and thus a
lighter panel, while still
maintaining the overall panel
stiffness. The table
summarizes the design
requirements.
Use the CES EduPack to plot the thermal conductivity of

foams against their elastic compliance 1 / E (the reciprocal of
their Youngs moduli E, since we must express the objectives in a
form that requires minimization). The numbers in brackets are
3
the densities of the foams in Mg/m . The foams with the lowest
thermal conductivity are the least stiff; the stiffest have the
highest conductivity. Explain the reasoning you would use to
select a foam for the truck panel using a penalty function.
Metal foams
C Ceramic
c foams
Figure E19
Polymer
foams
Function
Foam for panel-wall insulation
Constraint
Panel wall thickness specified.
Objectives
Minimize foam thermal conductivity,
Maximize foam stiffness, meaning Youngs

modulus, E
Free variables
Choice of material
Figure E20
Answer. The steps in making a reasoned choice are as follows.
M.F. Ashby 2010
Ceramic foams are brittle. This probably rules them out for the truck
body because is exposed to impact loads. But in other applications
ceramic foams particularly glass foams are viable.
48

(1) Sketch the trade-off surface: the low vs. low 1/E
envelope of the data, as shown below. The foams that lie on
or near the surface are a better choice than those far from it.
This already eliminates a large number of foams and identifies
the family from which a choice should be made. Note that
most metal foams are not a good choice; only if the highest
stiffness is wanted is the metal foam Aluminium-SiC (1.0) an
attractive choice.
(2) If it is desired to go further, it is necessary to construct a

penalty function:
1
Z = 1 + 2
E
Z is to be minimized, so 1 is a measure of the value
associated with reducing heat flow; 2 a measure of the value
associated with reducing core compliance. Rearranging the
equations gives
CCeramic
e foams
r
Metal foams
Penalty line
for 1/2 = 100
Polymer
foams
Penalty line
for 1/2 = 0.1
2 1

1 E
If the axes were linear, this equation would be that of a family

of straight, parallel, lines on the vs. 1/E diagram, of slope
2 / 1 , each line corresponding to a value of Z / 1 . In fact
the scales are logarithmic, and that leads instead to a set of
curved lines. One such line is sketched below for values
2 / 1 = 0.01 (meaning that thermal insulation is considered
very important, and stiffness less important) and for 2 / 1 =
100 (meaning the opposite). The foam nearest the point at
which the penalty lines are tangent to the trade-off surface is
the best choice. In the first example PVC foam with a density
3
of about 0.1 Mg/m is the best choice, but in the second a
ceramic or even a metal foam is a better choice.
M.F. Ashby 2010
49

E9. Selecting material and shape
The examples in this section relate to the analysis of material and
shape of Chapters 9 and 10 of Materials Selection for Mechanical
rd
Design 3 Edition. They do not use the CES EduPack, but help
develop important knowledge of the influence of shape on material
selection, and so have been left in here for completeness. They cover
the derivation of shape factors, of indices that combine material and
shape, and the use of the 4-quadrant chart arrays to explore material
and shape combinations.
E9.1 Shape factors for tubes (Figure E21).

(a) Evaluate the shape factor Be for stiffness-limited design in
bending of a square box section of outer edge-length h = 100mm
and wall thickness t = 3mm. Is this shape more efficient than one
made of the same material in the form of a tube of diameter 2r =
100mm and wall thickness t = 3.82 mm (giving it the same mass
per unit length, m/L)? Treat both as thin-walled shapes.
f
(b) Make the same comparison for the shape factor B for
strength-limited design.
Answer.
(a) The shape-efficiency factor for elastic bending for the square
box section is
1h
Be =
= 16.7
2 t
That for the tube is
3 r
Be =
= 12.5
t
The box is more efficient than the tube of the same m/L, and both
are much stiffer in bending than a solid square section of the same
area (and thus mass per unit length).
(b) The shape-efficiency factor for bending failure for the square
box section is
h
Bf =
= 5.8
t
That for the tube is
3
r
= 4.33
Bf =
2 t
The box is more efficient than the tube, and both are roughly 5
times stronger in bending than a solid square section of the same
area (and thus mass per unit length).
Figure E21
M.F. Ashby 2010
50

E9.2 Deriving shape factors for stiffness-limited design (Figure
E22). Derive the expression for the shape-efficiency factor Be for
stiffness-limited design for a beam loaded in bending with each of
the three sections listed below. Do not assume that the thin-wall
approximations is valid.
(a) a closed circular tube of outer radius 5t and wall thickness t,
(b) a channel section of thickness t, overall flange width 5t and
overall depth 10 t, bent about its major axis; and
(b) The second moment of area, I, of the channel (central figure)

about the axis shown as a dotted line is
I =
(10t )3 .5t
12
(8t )3 .4t
12
= 246 t 4
Its cross section area is

A = 18 t 2
from which
12 I
Be =
= 9.11
A2
(c) a box section of wall thickness t, and height and width h1 = 10t.
(c) The second moment of area, I, of the square box-section
(right-hand figure) about the axis shown as a dotted line is
h1 / 2
I =2
Answer.
(a) The shape factor for elastic bending of a tubular beam (lefthand figure) is defined by
12 I
Be =
A2
where I is the second moment of area and A is the area. The
second moment of area, I, of the tube about the axis shown as a
dotted line is
(5 t )4
1 4
h 1 h 42
12
h2 / 2
y 2 h2 dy
2 3
th
3 1
and
Figure E22
I =
y 2 h1 dy 2
(4 t )4
= 92.25 t 4
Its cross section area is
A = (5 t )2 (4 t )2 = 9 t 2
from which
Be =
12 I
A2
A = h12 h22 4 h1 t
Assembling these results, and inserting h1 = 10t gives
1 h1
Be =
=5
2 t
= 4.35
M.F. Ashby 2010
51

E9.3 Deriving shape factors for strength-limited design.
f
(a), (b) and (c) Determine the shape-efficiency factor B for

strength limited design in bending, for the same three sections
shown as Figure 29 of Exercise E9.2. You will need the results
for I for the sections derived in E9.2.
(d) A beam of length L, loaded in bending, must support a
specified bending moment M without failing and be as light as
possible.. Show that to minimize the mass of the beam per unit
length, m/L, one should select a material and a section-shape to
maximize the quantity
f
B f
M =
from which
6Z
B =
A3 / 2
= 3.87
(c) The second moment, Z, of the box-section about the axis

shown as a dotted line is
I
Z =
10 t
f
B =
from which
6Z
= 4.47
A3 / 2
2/3
where f is the failure stress and the density of the material of

f
the beam, and B is the shape-efficiency factor for failure in

bending.
Answer.
(a) The section modulus, Z, of the tube about the axis shown as
a dotted line is
I
I
Z =
=
= 18.45 t 3
y max
r
(here y max is the distance of the outer surface from the neutral
axis),from which
6Z
f
B =
= 2.31
A3 / 2
(b) The second moment, Z, of the channel about the axis shown
as a dotted line is
I
Z =
= 49.2 t 3
5t
M.F. Ashby 2010
(d) The beam of length L, loaded in bending, must support a

specified load F without failing and be as light as possible. Its
mass m is
(1)
m = AL
where A is the area of its cross section. Failure occurs if the load
exceeds the moment
M = Z f
(2)
where Z is the section modulus . Replacing Z by the shapef
factor B = 6 Z / A 3 / 2 gives
M =
Bf A 3 / 2
(3)
6
Substituting this into equation (1) for the mass gives
m = (6 M )2 / 3 L
2/3
f
B f
The best material-and-shape combination is that with the
greatest value of the index
M =
52
(4)
f 2 / 3
B f

E9.4 Determining shape factors from stiffness data. The elastic
shape factor measures the gain in stiffness by shaping, relative
to a solid square section of the same area. Shape factors can
be determined by experiment. The mass m of a beam of
length L and prescribed bending stiffness S B with a section of
efficiency Be is
L5 / 2
1
/
2
e E
where C1 is a constant that depends on the distribution of load

on the beam. Inverting the equation gives
12 L5 S 2
B
Be =
C m2 E
1
Thus if the bending stiffness S B , mass m and length L are

measured, and the modulus E and density of the material of
12 S B
m =
C1
1/ 2
( )
the beam are known, Be can be calculated.
(b) A steel truss bridge shown in Figure E23 has a span L and is simply
supported at both ends. It weighs m tonnes. As a rule of thumb,
bridges are designed with a stiffness SB such that the central deflection
of a span under its self-weight is less than 1/300 of the length L (thus
2
S B 300 mg / L where g is the acceleration due to gravity, 9.81 m/s ).
Use this information to calculate the minimum shape factor Be of the
three steel truss bridge spans listed in the table. Take the density of
3
steel to be 7900 kg/m and its modulus E to be 205 GPa. The constant
C1 = 384/5 = 76.8 for uniformly distributed load (UASSP, 3).
Bridge and construction date*

Royal Albert bridge, Tamar, Saltash UK (1857)
Carquinez Strait bridge, California (1927)
Chesapeake Bay bridge, Maryland USA (1952)
Span L
(m)
139
132
146
Mass m
(tonnes)
1060
650
850
*Data from the Bridges Handbook.
(a) Calculate the shape factor Be from the following

experimental data, measured on an aluminum alloy beam
loaded in 3-point bending (for which C1 = 48 see UASSP,
3) using the data shown in the table
Attribute
Beam stiffness SB
Mass/unit length
m/L
Beam length L
Beam material
Material density
Material modulus E
Figure E23
Value
5
7.2 x 10 N/m
1 kg/m
1m
6061 aluminum alloy
3
2670 kg/m
69 GPa
M.F. Ashby 2010
53

Answer.
(a) Inserting the data given in the question into the equation for
Be given in the example gives
(a) A beam, loaded in bending, must support a specified bending

moment M* without failing and be as light as possible. Section
shape is a variable, and failure here means the first onset of
plasticity. Derive the material index. The table summarizes the
requirements.
Be = 18.6
(b) The shape factor is given in terms of m, L and SB by the
equation given in the question. Substituting S B 300 mg / L
gives
3600 L4 g 2
C1 m E
Inserting C1 = 384/5 (UASSP, 3) and the data from the table gives
the results shown below.
Be =
Bridge and construction date*
F = mg
6
(10 N)
Royal Albert bridge, Tamar,
Saltash UK (1857)
Carquinez Strait bridge,
California (1927)
E9.5 Deriving indices for bending and torsion.
A = m / L
2
Be
(m )
10.6
0.97
49.3
6.5
0.62
65.4
8.5
0.74
74.8
Function
Constraints
Objective
Free variables
The results show that structured structures (trusses, rib-stiffened

structures) can offer very large shape efficiencies. Note how the
efficiency has grown over time.
M.F. Ashby 2010
Light weight beam

Specified failure moment M*
Length L specified
Minimum mass m
Choice of material
Section shape and scale
(b) A shaft of length L, loaded in torsion, must support a specified

torque T* without failing and be as cheap as possible. Section
shape is a variable and failure again means the first onset of
plasticity. Derive the material index. The table summarizes the
requirements.
Function
Constraints
Chesapeake Bay bridge,

Maryland USA (1952)
Objective
Free variables
54
Cheap shaft
Specified failure torque T*
Length L specified
Minimum material cost C
Choice of material

Answer.
(a) The mass m of a beam of length L and section area A is
(1)
m = A L
where is the density of the material of which the beam is made.
Failure occurs if the bending moment exceeds the value
M = Z f
(2)
f
Replacing Z by the shape-factor B for bending of a beam

gives
(3)
f A3 / 2
6 B
Substituting this into equation (1) for the mass of the beam gives
2
/
3
m = 6 M*
L
2/ 3
f
(4)
B f
The best material-and-shape combination is that with the greatest

value of the index
M =
Replacing Q by the shape-factor T
for torsion gives
(7)
f A3 / 2
4.8 T
Substituting this into equation (5) for the material cost of the shaft
gives
2
/
3
C
m
m = 4 .8 T *
L
2/ 3
f
(8)
T f
Approximating f by f / 2 , we find that the material-and-shape
T =
combination with the lowest associated material cost is that with

the greatest value of the index
f 2 / 3
f
T
M4 =
Cm
f 2 / 3
B f
M3 =
(b) The shaft, loaded in torsion, must support a specified load F

without failing and be as light as possible. The material cost C of
a shaft of length L and section area A is
C = A L Cm
(5)
where Cm is the cost per kg and is the density of the material
of which the shaft is made. Failure occurs if the torque T
exceeds the moment
T = Q f
(6)
where f is the shear stress at which plasticity first starts.
M.F. Ashby 2010
55

E9.6 Use of the four segment chart for stiffness-limited design
(a) Use the 4-segment chart for stiffness-limited design on the
next page to compare the mass per unit length, m/L, of a section
5
2
with EI = 10 Nm made from
(i) structural steel with a shape factor Be of 20, modulus E = 210
3
GPa and density = 7900 kg/m
(ii) carbon fiber reinforced plastic with a shape factor Be of 10,
3
modulus E = 70 GPa and density = 1600 kg/m , and
(iii) structural timber with a shape factor Be of 2, modulus E = 9
GPa and density = 520 kg/m
3.
The schematic Figure E24 illustrates the method.
Answer.
(a) The method is shown in the question. The results appear on
the chart on the following page. The CFRP beam with m/L of
about 4 kg/m is significantly (factor about 2) lighter than one made
of steel or softwood.
E * / * with E * = E / Be and
(b) Direct calculation of
* = / Be is shown in the table.

Material
Steel
CFRP
Softwood
E* = E / Be
2
* = / Be
3
(N/m )
(kg/m )
9
10.5 x 10
9
7.0 x 10
9
4.5 x 10
395
160
260
E* / *
2 1/2
(N/m ) /(kg/m )
259
523
258
The higher the value of E* / * the lighter is the beam. The

ranking is the same as that arrived at in Part (a).
Figure E24
(b) Show, by direct calculation, that the conclusions of part (a) are
consistent with the idea that to minimize mass for a given stiffness
one should maximize
E * / * with E * = E / Be and * = / Be .
M.F. Ashby 2010
56
M.F. Ashby 2010
57
M.F. Ashby 2010
58

E9.7 Use of the four segment chart for strength
(a) Use the 4-segment chart for strength-limited design on the
next page to compare the mass per unit length, m/L, of a section
4
with Z f = 10 Nm (where Z is the section modulus) made from
f
(i) mild steel with a shape factor B of 10, strength f = 200

3
MPa and density = 7900 kg/m
f
(ii) 6061 grade aluminum alloy with a shape factor B of 3,

3
strength f = 200 MPa and density = 2700 kg/m , and
Answer.
(a) The method is shown in the question. The results appear on
the chart on the next page. The steel section is the heaviest
( m / L 9 kg/m). The titanium alloy is the lightest ( m / L 2.5
kg/m).
f
(b) Direct calculation of B f
2/3
/ gives the values shown in
the table.
f
(iii) a titanium alloy with a shape factor B of 10, , strength f =
3
480 MPa and density = 4420 kg/m .

The schematic Figure E25 illustrates the method.
Material
Steel
6061 Al alloy
Ti alloy
Figure
(MPa)
(kg/m )
200
200
480
7900
2700
4420
f
The higher the value of B f
2/3
Bf
f
B f
2/3
10
3
10
2/3
/
3
(MPa) /(kg/m )
-2
2 x 10
-2
2.6 x 10
-2
6.4 x 10
/ the lighter is the beam.
The ranking is the same.
(b) Show, by direct calculation, that the conclusions of part (a)

are consistent with the idea that to minimize mass for a given
stiffness one should maximize
f
B f
2/3
M.F. Ashby 2010
59
M.F. Ashby 2010
60
M.F. Ashby 2010
61

E9.8 A light weight display stand (Figure E26). The figure
shows a concept for a lightweight display stand. The stalk must
support a mass m of 100 kg, to be placed on its upper surface at a
height h, without failing by elastic buckling. It is to be made of
stock tubing and must be as light as possible. Use the methods of
Chapter 11 to derive a material index for the tubular material of the
stand that meets these requirements, and that includes the shape
of the section, described by the shape factor
Be =
Material
Aluminum alloys
Steel
Copper alloys
Polycarbonate (PC)
Various woods
Modulus E Tube radius r

(GPa)
69
25 mm
210
30 mm
120
20 mm
3.0
20 mm
7 - 12
40 mm
Wall thickness/tube
radius, t/r
0.07 to 0.25
0.045 to 0.1
0.075 to 0.1
0.15 to 0.3
Solid circular sections
only
12 I
Answer. The objective function is the mass. The mass m of the
tubular stalk is
A2
where I is the second moment of area and A

is the section area. The table summarizes
the requirements.
.
m = Ah
where is the density and A the section area of the tube. The
constraint is that the stalk should not buckle. The buckling load
for a slender column of height h is
Fcr =
Figure E26
Function
Light weight column
Constraints
Specified buckling load F

Height h specified
Objective
Minimum mass m
Free variables
Choice of material
n2 2 E I
h2
e 2
n2 2 E B A
12
h2
(2)
where E is Youngs modulus, I is the second moment of area of

the tube and n is a constant that depends on the end constraint
(n = 1/2 in this instance UASSP, 5). Substituting for A from
equation (2) into equation (1) gives
m =
2 3 h2 1/2
F
e
(E B )1/2
(3)
The best choice of tubing for the stalk is that with the greatest
value of
M=
Cylindrical tubing is available from stock in the following materials

and sizes. Use this information and the material index to identify
the best stock material for the column of the stand.
M.F. Ashby 2010
(1)
62
( E Be )1/2

The shape factor for a thin-walled tube is calculated from the
bending moment of inertia, Ixx (UASSP, 2):
3
Ixx = r t
(4)
where r is the tube radius and t the wall thickness; the area A of the
section is 2 rt.
Its shape factor is
Be =
12 I
A2
3 r
(5)
(Note that, for a solid section, Be = 0.95 1 ). The selection can

be made by evaluating and comparing the quantity
M = ( E Be )1/2 / for each of the materials and sections, as

shown.
Material
Aluminum alloys
Steel
Copper alloys
Polycarbonate (PC)
Various woods
E
(GPa)
(kg/m3)
Be
M
1/2
3
(GPa /Mg/m )
71
210
120
3
8-15
2700
7900
8900
1200
500 - 800
4 - 14
10 - 21
10 - 13
3.3 - 6.7
1
6.2 - 11.7
5.8 - 8.2
3.9 4.4
2.6 3.7
3.5 7.7
Aluminum offers the highest index and thus the lightest stalk.
Steel is the next best choice, then wood.
M.F. Ashby 2010
63
E9.9 Microscopic shape: tube arrays (Figure E27). Calculate the

e
gain in bending efficiency, B
, when a solid is formed into
small, thin-walled tubes of radius r and wall thickness t that

are then assembled and bonded into a large array, part of which
is shown in the figure. Let the solid of which the tubes are made
have modulus Es and density s . Express the result in terms
of r and t.
Answer. The unit cell of the structure is shown as a hexagon on the

right-hand part of the figure. The area As of solid in the unit cell,
equal to its relative density / s , is
As = 6 r t
and the relative density of the array, / s , is
As
t
=
= 1.8
2
s
r
6 tan( 60 ) r
If an initially solid, square section, beam of edge length bs is
formed into an array like that of the figure, with an outer, square,
profile, the edge-length of this new square is
1/ 2

b = s
bs

The apparent modulus of the tube array parallel to the axis of the
tubes is

E s
E =
s
and the second moment of area of the tube array is
2

b4
bs
I =
=
12
s
Thus the bending efficiency is
Figure E27
Be =
M.F. Ashby 2010
64
EI
r
= s = 0.56
Es I s
E9.10 The structural efficiency of foamed panels (Figure E28).

Calculate the change in structural efficiency for both bending
stiffness and strength when a solid flat panel of unit area and
thickness t is foamed to give a foam panel of unit area and
thickness h, at constant mass. The modulus E and strength
f of foams scale with relative density / s as

E =
s
Es
and

f =
s
Answer. The panel, initially solid and of thickness t, is foamed to a

height h at constant area. The density falls from s to
= s
t
h
and the modulus and strength fall from Es to f ,s
3/ 2
f ,s
where E, f and are the modulus, strength and density of the

foam and Es , f ,s and s those of the solid panel.
t
E = Es
h
3/ 2
f = f ,s
h
and
The bending stiffness per unit area changes from Ss to S, so

that
Be =
S
So
EI
Es I s
h
t
1
( / s )
(using I s = t 3 / 12 and I = h 3 / 12 per unit width). Foaming

imparts a shape-efficiency factor for stiffness equal to the
reciprocal of the relative density.
Following the same procedure for strength (using
Z s = t 2 / 12 and I = h 2 / 6 per unit width) gives
Bf =
Figure E28
M.F. Ashby 2010
Z f
Z s f ,s
h
t
( / s )1 / 2
The shape-efficiency factor for strength is equal to the reciprocal

of the square root of the relative density.
65

E10. Hybrid materials
The examples in this section relate to the design of hybrid materials
described in Chapters 11 and 12. The first three involve the use of
bounds for evaluating the potential of composite systems. The fourth is
an example of hybrid design to fill holes in property space. Then comes
one involving a sandwich panel. The next three make use of the
charts for natural materials. The last is a challenge: to explore the
potential of hybridizing two very different materials.
E10.1 Concepts for light, stiff composites.

Figure E29 is a chart for exploring stiff composites with light alloy
or polymer matrices. A construction like that allows the potential
of any given matrix-reinforcement combination to be assessed.
Four matrix materials are shown, highlighted in red. The materials
shown in gray are available as fibers (f), whiskers (w) or particles
(p). The criteria of excellence (the indices E / , E 1 / 2 / and
E 1 / 3 / for light, stiff structures) are shown; they increase in

value towards the top left. Use the chart to compare the
performance of a titanium-matrix composite reinforced with (a)
zirconium carbide, ZrC, (b) Saffil alumina fibers and (c) Nicalon
silicon carbide fibers. Keep it simple: use equations 11.1 11.2
to calculate the density and upper and lower bounds for the
modulus at a volume fraction of f = 0.5 and plot these points.
Then sketch arcs of circles from the matrix to the reinforcement to
pass through them. In making your judgement, assume that f =
0.5 is the maximum practical reinforcement level.
M.F. Ashby 2010
Figure E29
66
Answer. The adjacent figure shows the result. The guide lines have
been moved to pass through titanium, allowing the performance
of each composite system to be assessed. A vertical, broken,
line through each of the bound-envelopes marks the f = 0.5
point. The Ti-ZrC composite system offers little or no gain in
performance because the composite trajectory lies almost
parallel to the guide-lines. By contrast, Ti-Saffil composites offer
gains. Best of the lot is Ti-Nicalon. These composites have a
trajectory that lies almost normal to the guide-lines, offering the
greatest increase in all three criteria of excellence.
E10.2 Use the chart of Figure E29 to explore the relative potential of
Magnesium E-glass-fiber composites and Magnesium
Beryllium composites for light, stiff structures.
Answer. The adjacent figure shows the result. Mg E-glass offers

almost no gain in performance. Mg Be, if it could be made,
offers very considerable gains.
M.F. Ashby 2010
67

E10.3 Concepts for composites with tailored thermal properties.
Figure E30 is a chart for exploring the design of composites with
desired combinations of thermal conductivity and expansion, using
light alloy or polymer matrices. A construction like that allows the
potential of any given matrix-reinforcement combination to be
assessed. One criterion of excellence (the index for materials to
minimize thermal distortion, / ) is shown; it increases in value
towards the bottom right. Use the chart to compare the
performance of a magnesium AZ63 alloy-matrix composite
reinforced with (a) alloy steel fibers, (b) silicon carbide fibers, SiC
(f) and (c) diamond-structured carbon particles. Use the equations
below to calculate the upper and lower bounds for and at a
volume fraction of f = 0.5 and plot these points. Then sketch
curves linking matrix to reinforcement to pass through the
outermost of the points. In making your judgement, assume that f
= 0.5 is the maximum practical reinforcement level.
Thermal Expansion Coefficient
~u = fr (1 + vr ) + (1 f )m(1 + vm) L[ fvr + (1 f )vm]
~L =
Errf + Emm (1 f )
Erf + Em (1 f )
Thermal Conductivity
u = fr + (1 f )m
Figure E30
~
r + 2 m 2 f ( m r )
L = m
r + 2 m + f ( m r )
Subscript u: upper bound; subscript L: lower bound
M.F. Ashby 2010
68
Answer. The figure shows the result. Mg-alloy steel composites

perform less well than magnesium itself. By contrast, Mg-SiC
composites offer gains (and are currently available). Even
bigger gains lie with Mg-diamond composites. These are not as
outrageous as they might sound: particulate industrial diamond is
produced in large quantities for cutting tools, rock drills, grinding
and polishing. Using it for a reinforcement in a composite for
very high performance applications is perfectly practical.
M.F. Ashby 2010
69

E10.4 Sandwich panels. An aircraft quality sandwich panel has
the characteristics listed in the table.
(a) Use the data below to calculate the equivalent density,
flexural modulus and strength (use the plastic section
modulus).
Panel
(b) Plot these on a modulus / density and a modulus /

strength chart. Do the flexural properties of the panel lie
in a region of property space not filled by monolithic
materials?
Data for glass fiber / aluminum honeycomb sandwich panel
Face material
0.38 mm glass fiber / epoxy
Core material
3.2 mm cell, 97.8 kg/m , 5052 Alu honeycomb
Panel weight per unit area, ma
2.65 kg/m
Panel length, L
510 mm
Panel width, b
51 mm
Panel thickness, d
10.0 mm
Flexural stiffness EI
67 N.m
Failure moment M f
160 Nm
Panel
Answer.
m
(a) The equivalent density ~ is ~ = a = 265 kg / m 3
d
~ ~ 12 EI
The equivalent modulus E is E =
= 15.8 GPa
bd3
4M f
t
= 125 MPa
The equivalent strength f is ~ f =
bd2
(b) See opposite. Yes, the flexural properties of the panel lie in an
area of property space not occupied by monolithic materials.
M.F. Ashby 2010
70

E10.5 Natural hybrids that are light, stiff and strong.
For this project use the CES EduPack Level 3 Database, and under subset
select All materials (instead of All bulk materials). Then create a custom
subset which includes natural materials (under Hybrids), natural fibers and
particulates, and the materials featured in the question.
(a) Plot aluminum alloys, steels, CFRP and GFRP onto a E chart for
the custom subset of materials you created above, where E is
Youngs modulus and is the density. How do they compare, using
the flexural stiffness index E 1 / 2 / as criterion of excellence?
Answer. The comparisons are shown in two the figures on the next
page.
(a) The comparison for stiffness, using the criterion of
excellence, E 1 / 2 / , is highlighted. The nearer a material lies
to this line, the better is its performance. Natural materials are
extraordinarily efficient by this measure. CFRP is comparable
with the best of them, but the other man-made materials
compare poorly with those of nature.
(b) The comparison for strength, using criterion of excellence,
2f / 3 / , is highlighted. Again, natural materials excel. CFRP

(b) Do the same thing for strength, using the flexural strength index
alone competes with wood, bamboo and palm.
2f / 3 / (where f is the failure stress) as criteria of excellence.

The table lists the necessary data.
Material
Youngs modulus
E (GPa)
Density
3
(kg/m )
Strength,
f (MPa)
Aluminum alloy
74
2700
335
Steel
210
7850
700
CFRP
100
1550
760
GFRP
21
1850
182
M.F. Ashby 2010
71
Steel
CFRP
CFRP
Steel
Al-alloy
Al-alloy
GFRP
GFRP
M.F. Ashby 2010
72
E10.6 Natural hybrids that act as springs.

CFRP
For this project use the CES EduPack Level 3 Database, and under
subset select All materials (instead of All bulk materials). Then
create a custom subset which includes natural materials (under
Hybrids), natural fibers and particulates, and the materials featured in
the question.
Spring steel
Cu 2% Be
The table lists the moduli and strengths of spring materials. Plot these
onto a copy of the E f chart for the materials in the custom subset
you created above, and compare their energy-storing performance with
that of natural materials, using the 2f / E as the criterion of choice.
Here f is the failure stress, E is Youngs modulus and is the
density.
Material
Modulus
(GPa)
Strength
(MPa)
Density
3
(kg/m )
Spring steel
206
1100
7850
Copper-2%Beryllium
130
980
8250
CFRP filament wound
68
760
1580
Answer. The comparison is shown on the adjacent figure. The

criterion of excellence, 2f / E , is highlighted. The nearer a
material lies to this line, the better is its performance. The best
man-made spring materials have a performance that is
comparable with that of antler, but falls short of that offered by
silks of various types. If the criterion for light springs, 2f / E is
used instead, the natural materials rank even more highly.
M.F. Ashby 2010
73

E10.7 Finding a substitute for bone. Find an engineering material
that most closely resembles the longitudinal strength of compact bone
(cortical bone (L)) in its strength/weight ( f / ) characteristics by
plotting data for this material onto a f chart for engineering
materials. Here f is the failure stress and is the density.
Answer. The modulus and density of compact bone match closely

those of magnesium and glass fiber reinforced epoxy, GFRP.
These materials work well for external braces. For
prostheses, bio-compatibility is important. This leads to the
used of cobalt and titanium-based alloys even though the
match of mechanical properties is very poor. Efforts are
underway to develop a bio-compatible
polyethylene/hydroxyapatite composites with a near perfect
match of properties. Hydroxyapatite is a natural ceramic that
is found in real bone it appears on the charts for natural
materials.
Cortical
bone (L)
E10.8
Creativity: what could you do with X? The CES EduPack
can produce a wide variety of materials property charts. These
can be used as the starting point for what ifO? exercises. As a
challenge, use any chart or combination of charts to explore
what might be possible by hybridizing any pair of the materials
listed below, in any configuration you care to choose.
(a) Cement
(b) Wood
(c) Polypropylene
(d) Steel
(e) Copper.
M.F. Ashby 2010
74

Answer. (a) The function and constraints are listed in the table.
Answer.
E11. Selecting processes

The exercises of this section use the process selection feature of the
CES EduPack.
Each exercise has two parts, labeled (a) and (b). The first involves
translation (i.e. converting a description of needs and wants into a list
of go no-go criteria . The second uses the CES EduPack to select
appropriate processes for the respective function. (Use the Tree
function to match processes to their appropriate material classes).
E11.1 Elevator control quadrant (Figure E31). The quadrant

sketched here is part of the control system for the wing-elevator of
a commercial aircraft. It is to be made of a light alloy (aluminum or
magnesium) with the shape shown in the figure. It weighs about 5
kg. The minimum section thickness is 5 mm, and apart from the
bearing surfaces the requirements on surface finish and
precision are not strict: surface finish 10 m and precision 0.5
mm. The bearing surfaces require a surface finish 1 m and a
precision 0.05 mm. A production run of 100 200 is planned.
(a) Itemize the function and constraints, leave the objective blank
and enter Choice of process for the free variable.
(b) Use the CES EduPack
to apply the constraints and
identify in more detail the
viable processes.
The processes surviving the

successive applications of
the constraints on the
quadrant:
(a) The shape-efficiency factor for elastic bending for the
Shape
elevator
section
(from an
Table
9.3)quadrant
is
Function
square box
Constraints
Material
1 hclass: aluminum or magnesium alloy
e
= 163D-solid
.7
Shape
B = 2 class:
t
Mass: 5 kg
That for the tube
is Minimum section: 5 mm
e = 3 r =0.5
Tolerance:
12.mm,
5 (0.05mm on bearing
B
surfaces) t
Surface than
roughness:
10 m
(1 m
on bearing
The box is more efficient
the tube
of the
same
m/L, and
surfaces)
both are much stiffer in bending than a solid square section of
Planned batch size of 100 - 200

the same area (and
thus mass per unit length).
Objective
-Free variables
Choice of process
(b) The shape-efficiency factor for bending failure for the

square box section (from Table 9.3) is
(b) These constraints are applied in the CES EduPack. The

h to those shown in the first
f
material constraint limits the
Bselection
=
= 5.8
t constraint eliminates two,
column of the next table. The shape
leaving
in the
That
for those
the tube
is second column. All of these can meet the
constraints on size and section, which are not restrictive. The
3 on the
r body are not met by
tolerance and roughness constraints
= 4.33
Bf =
2 if tthey have been oversand casting, though it is worth asking
specified.
Those
on the
bearing
surfaces
more
The
box is more
efficient
than
the tube,
and bothare
aremuch
roughly
5 times
restrictive
only
machining
from
solid
achieves
them.
The
stronger in bending than a solid square section of the same area
(and thus
answer
herelength).
is not to reject the others, but to add a finishing
mass
per unit
step, shown at the bottom of Mass range and Tolerance charts.
The result is the list shown in the third column. Economics are
important here the Economic batch-size chart suggests that
only the three listed in the last column are economic at a batch
size of 100 200.
The conclusion: explore investment casting (with additional
machining steps) and numerically controlled machining from
solid.
Figure E31
M.F. Ashby 2010
75
Processes
passing the
material
constraint.
Sand casting
Die casting
Processes that
survive the
shape
constraint
Sand casting
Die casting
Investment
casting
Investment
casting
Low pressure
casting
Low pressure
casting
Forging
Forging
Extrusion
Sheet forming
Powder
methods
Machine from
solid
FAILS
FAILS
Powder
methods
Machine from
solid
Processes
meeting the
tolerance
constraint
FAILS
Die casting plus
machining
Investment
casting plus
machining
Low pressure
casting plus
machining
Forging plus
machining
Powder
methods
Machine from
solid
Processes that
are economic
at a batch of
100 - 200
FAILS
Investment
casting plus
machining
FAILS
FAILS
E11.2 Casing for an electric plug (Figure E32). The electric plug
is perhaps the commonest of electrical products. It has a
number of components, each performing one or more
functions. The most obvious are the casing and the pins,
though there are many more (connectors, a cable clamp,
fasteners, and, in some plugs, a fuse). The task is to
investigate processes for shaping the two-part insulating
casing, the thinnest part of which is 2 mm thick. Each part
weighs about 30 grams and is to be made in a single step from
a thermoplastic or thermosetting polymer with a planned batch
4
6
size of 5 x 10 2 x 10 . The required tolerance of 0.3 mm
and surface roughness of 1 m must be achieved without
using secondary operations.
(a) Itemize the function and constraints, leave the objective
blank and enter Choice of process for the free variable.
FAILS
Machine from
solid
(b) Use the CES EduPack to identify possible processes to

make the casing
Figure E32
M.F. Ashby 2010
76

Answer. (a) The table lists the function and constraints.
Function
Constraints
Objective
Free variables
Shape an electric plug casing
Material class: thermoplastic or thermosetting

polymer
Shape class: 3D-solid
Mass: 0.03 kg
Minimum section: 2 mm
Tolerance: 0.3 mm
Surface roughness: 1m
4
6
Planned batch size of 5 x10 2 x10

-
Choice of process
(b) Using CES EduPack to apply the constraints gives the

selection shown in the table below.
Selection using CES
EduPack
Injection molding
Compression molding
Comment
The selection is identical with that derived
from the charts, but the software also
delivers data files containing supporting
information for each process.
E11.3 Ceramic valves for taps (Figure E33). Few things are more
irritating than a dripping tap. Taps drip because the rubber
washer is worn or the brass seat is pitted by corrosion, or both.
Ceramics have good wear resistance, and they have excellent
corrosion resistance in both pure and salt water. Many
household taps now use ceramic valves.
The sketch shows how they work. A ceramic valve consists of
two disks mounted one above the other, spring-loaded so that
their faces are in contact. Each disk has a diameter of 20 mm, a
thickness of 3 mm and weighs about 10 grams. In order to seal
well, the mating surfaces of the two disks must be flat and
smooth, requiring high levels of precision and surface finish;
typically tolerance < 0.02 mm and surface roughness < 0.1 m.
The outer face of each has a slot that registers it, and allows the
o
upper disc to be rotated through 90 (1/4 turn). In the off
position the holes in the upper disc are blanked off by the solid
part of the lower one; in the on position the holes are aligned. A
5
6
production run of 10 10 is envisaged.
(a) List the function and constraints, leave the objective blank
and enter Choice of process for the free variable.
(b) Use the CES EduPack to identify possible
processes to make the casing.
Figure E33
M.F. Ashby 2010
77

Function
Constraints
Objective
Free variables
Shape a ceramic valve

Material class: technical ceramic
Shape class: prismatic
Mass: 0.01 kg
Tolerance: 0.02 mm
Surface roughness: 0.1m
5
6
Planned batch size of 10 10
Choice of process
(b) Applying the constraints as shown into the EduPack,

eliminating processes that fail a constraint, leaves only one
candidate: powder methods. The tolerance and roughness
requirements for the mating faces require a subsequent grinding,
lapping or polishing step.
M.F. Ashby 2010
78
E11.4 Shaping plastic bottles

(Figure E34). Plastic bottles
are used to contain fluids as
various as milk and engine
oil. A typical polyethylene
bottle weighs about 30 grams
and has a wall thickness of
about 0.8 mm. The shape is
3-D hollow. The batch size is
large (1,000,000 bottles).
What process could be used
to make them?
(b) Applying the constraints and eliminating processes that fail a

constraint leaves only two candidates: blow molding and injection
molding.
Figure E34
(a) List the function and
constraints, leave the
objective blank and enter Choice of process for the free variable.
(b) Use the CES EduPack to identify possible processes to make

the casing.

Function
Constraints
Objective
Free variables
Shape a polyethylene bottle

Material class: Polyethylene (or thermoplastic)
Shape class: 3-D hollow
Mass: 0.02 0.04 kg
Minimum section: 0.7 - 1 mm
Tolerance: 1 mm
Surface roughness: 10 m
6
Planned batch size of > 10
Choice of process
M.F. Ashby 2010
79
E11.5 Car hood (bonnet) (Figure E35). As weight-saving assumes

greater importance in automobile design, the replacement of steel
parts with polymer-composite substitutes becomes increasingly
attractive. Weight can be saved by replacing a steel hood with
one made from a thermosetting composites. The weight of the
hood depends on the car model: a typical composite hood weighs
is 8 - 10 kg. The shape is a dished-sheet and the requirements on
tolerance and roughness are 1 mm and 2 m, respectively. A
production run of 100,000 is envisaged.
(a) List the function and constraints, leave the objective blank and
enter Choice of process for the free variable.
(b) Use the CES EduPack to identify possible processes to make
the casing.

Function
Shape a car hood
Constraints
Material class: Thermosetting composite

Shape class: Dished sheet
Mass: 8 10 kg
Tolerance: 1 mm
Surface roughness: 2 m
Planned batch size of 100,000
Objective
Free variables
Choice of process
(b) Applying the constraints and eliminating processes that fail a

constraint leaves resin transfer molding as the prime choice.
Figure E35
M.F. Ashby 2010
80

E11.6 Selecting joining processes.
(a) Use CES EduPack to select a joining process to meet the
following requirements.
Function
Constraints
Objective
Free variables
Create a permanent butt joint between steel

plates
Material class: carbon steel
Joint geometry: butt joint
Section thickness: 8 mm
Permanent
Watertight
-
Choice of process
(b) Use CES EduPack to select a joining process to meet the

following requirements.
Function
Constraints
Objective
Free variables
Answer. (a) Applying CES EduPack to the problem gives the

selection shown in the below.
Selection using CES EduPack
Level 1 or 2
Brazing
Friction welding
Friction-stir welding
Gas metal arc (MIG) welding
Gas-tungsten arc (TIG) welding
Manual metal arc (MMA) welding
Power (electron, laser) beam welding
Comment
There are many ways to

make a permanent watertight
joint in steel.
(b) Applying CES EduPack to the problem gives the selection

shown in the below.
Level 1 or 2
Snap fit
Threaded fasteners
Flexible adhesives
Rigid adhesives
Comment
All are practical
Create a watertight, demountable lap joint

between glass and polymer
Material class: glass and polymers
Joint geometry: lap joint
Section thickness: 4 mm
Demountable
Watertight
-
Choice of process
M.F. Ashby 2010
81

E11.7 Selecting surface treatment processes.
(a) Use CES EduPack to select a surface treatment process to meet

the following requirements.
Function
Constraints
Objective
Free variables
Increase the surface hardness and wear

resistance of a high carbon steel component
Material class: carbon steel
Purpose of treatment: increase surface

hardness and wear resistance
-
Choice of process
(b) Use CES EduPack to select a surface treatment process to

meet the following requirements.
Function
Constraints
Objective
Free variables
Apply color and pattern to the curved

surface of a polymer molding
Material class: thermoplastic
Purpose of treatment: aesthetics, color
Curved surface coverage: good or very

good
-
Choice of process
M.F. Ashby 2010
Answer. (a) Applying CES EduPack to the problem gives the

selection shown in the below.

Level 1 or 2
Carburizing and carbonitriding
Nitriding
Induction and flame hardening
Laser surface hardening
Electro plating
Vapor metallizing
Comment
There are many ways to

increase the surface hardness
of carbon steel.
(b) Applying CES EduPack to the problem gives the selection

shown in the below.

Level 1 or 2
Cubic printing
Pad printing
Electroless plating
Electrophoretic and Autophoretic
painting
Organic-solvent based painting
Vitreous enameling
Water-based painting
82
Comment
All are practical

E12. Materials and the environment
E12.1. Use the CES EduPack to produce a E H p chart to find the
polymer with a modulus E greater than 1 GPa and the lowest
embodied energy per unit volume.
Material
Cast iron
Polypropylene
Embodied energy*
MJ/kg
CO2*
kg/kg
Embodied energy
MJ/chair
CO2
kg/chair
17.3
1.02
147
8.7
95
2.7
152
4.3
* Mean values from the data in the CES EduPack
Answer. The construction is shown opposite. The polymer with a

modulus greater than 1 GPa and the lowest embodied energy is
polypropylene, PP.
E12.2. A maker of polypropylene (PP) garden furniture is concerned

that the competition is stealing part of his market by claiming that the
traditional material for garden furniture, cast iron, is much less energy
and CO2 intensive than the PP. A typical PP chair weighs 1.6 kg; one
made of cast iron weighs 8.5 kg. Use the data in the CES EduPack to
find out who is right are the differences significant if the data for
embodied energy are only accurate to +- 20%? If the PP chair lasts 5
years and the cast iron chair lasts 25 years, does the conclusion
change?
Answer. The table lists mean values of embodied energy and carbon
footprint, per kg, for the two materials. The last two columns show the
values per chair. If the difference in lifetime is ignored, the two chairs
do not differ significantly in embodied energy, but they do in carbon
release. If the longer life of the cast iron chair is recognized by dividing
the values by the life in years (to give energy and carbon per chairyear) the cast iron chair wins easily.
M.F. Ashby 2010
The Modulus Embodied energy chart: the one for stiffness at

minimum embodied energy.
83
E12.3 Identical casings for a power tool could be die-cast in

aluminum or molded in ABS or polyester GFRP. Use the CES
EduPack to produce a embodied-energy per unit volume bar-chart
to decide which choice minimizes the material embodied energy,
assuming the same volume of material is used for each casing.
3
Answer. The embodied energy /m , H p , (where H p is the energy

per kg and the density) bar chart shows that ABS has the lower
embodied energy per unit volume. For a casing using the same
volume of material, ABS requires only one fifth of the energy of
aluminum.
Material
Embodied energy/m ,
H p , MJ/m3
ABS
1 x 105
GFRP
2.4x 105
Aluminum
5.5x 105
E12.4 Disposable knives and forks (Figure E37). Disposable knives
and forks are ordered by an environmentally-conscious pizzahouse. The shape of each (and thus the length, width and profile)
are fixed, but the thickness is free: it is chosen to give enough
bending-stiffness to cut and impale the pizza without excessive
flexure. The pizzeria-proprietor wishes to enhance the greenness
of his image by minimizing the energy-content of his throw-away
tableware, which could be molded from polystyrene (PS) or
stamped from aluminum sheet.
Establish an appropriate material index for selecting materials for

energy-economic forks. Model the eating implement as a beam of
fixed length L and width w, but with a thickness t that is free,
loaded in bending, as in the figure. The objective-function is the
volume of material in the fork times its energy content, H p , per
unit volume ( H p is the embodied energy per kg, and the
density). The limit on flexure imposes a stiffness constraint
(UASSP, 3). Use this information to develop the index.
Flexure, in cutlery, is an inconvenience. Failure whether by
plastic deformation or by fracture is more serious: it causes lossof-function; it might even cause hunger. Repeat the analysis,
deriving an index when a required strength is the constraint.
Answer. The objective is for the energy content, mHp, to be

minimized. The free variable is the thickness t of the shaft of the
fork; all other dimensions are fixed. There are two alternative
constraints, first, that the fork should not flex too much, second,
that it should not fail.
Function
Constraints
Objective
Free variables
Environmentally friendly disposable forks

Length L specified
Width b specified
Stiffness S specified, or
Failure load F is specified
Minimize the material energy-content
Shaft thickness t
Choice of material
Figure E37
M.F. Ashby 2010
84

The resulting indices are
M1 =
E1/ 2
H p
and
M2 =
y2 / 3
H p
The selection can be implemented using the CES EduPack software,

by making a chart with the stiffness index as one axis and the strength
index as the other. The materials that best meet both criteria lie at the
top right
E12.5. Show that the index for selecting materials for a strong panel
with the dimensions shown in Figure E.38, loaded in bending, with
minimum embodied energy content is that with the largest value of
Answer. The objective function for the energy of the panel, H , is the
volume Lbh times the embodied energy of the material per unit volume,
Hp :
H
Its failure load F must be at least F * , where

I y
F = C2
F*
hL
Here C 2 is a constant that depends only on the distribution of the loads
and I is the second moment of area, which, for a rectangular section, is
I =
1y / 2
H p
where H p is the embodied energy of the material, its density and
y its yield strength. To do so, place a constraint on failure load F

Iy
hL
b h3
12
We can reduce the energy H by reducing h , but only so far that the
stiffness constraint is still met. Combining the last two equations and
solving for h gives
12 F * L
h
C2b y
requiring that it exceed a chosen value F * where
F = C2
= bhL H p
> F*
1. / 2
Using this to eliminate h in the objective function gives
where C2 is a constant and I is the second moment of area of the

panel
b h3
I =
12 .
12 F * L3 b
C2
1/ 2
Hp
1y / 2
The quantities F * , L , b and C 2 are all specified; the only freedom of

choice left is that of the material. The best materials for a strong panel
with the lowest embodied energy are those with the smallest values of
Mp =
Hp
1y / 2
Figure E38
M.F. Ashby 2010
85

E12.6. The makers of a small electric car wish to make bumpers out of
a molded thermoplastic. Which index is the one to guide this
selection if the aim is to maximize the range for a given battery
storage capacity? Plot it on a Strength / Density chart, and make
a selection.
Answer. The use-energy of any vehicle, be it gasoline or electric

powered, increases with the mass of the vehicle. The bumper
contributes to this mass. Maximizing range for a given battery
capacity means minimizing mass. The function of the bumper is to
sustain bending loads. The required index is
M =
y2 / 3
This is plotted on the strength density chart below, picking up the

slope of 1.5 from the guide line at the lower right. The selection line
is position such that a few polymers remain above it they are the
best choice. They are:
Polycarbonate, PC
Polyamide (nylon) PA and
Polyetheretherketone PEEK
In practice bumpers are made of blends of Polycarbonate with other

polymers such as Polypropylene or Polyamide, or of fiber reinforced
thermosets such as Polyester. PEEK is too expensive for
applications such as this.
M.F. Ashby 2010
86

E12.7. Energy-efficient floor joists . Floor joists are beams loaded in
bending. They can be made of wood, of steel, or of steel-reinforced
concrete, with the shape factors listed below. For a given bending
stiffness and strength, which of these carries the lowest productionenergy burden? The relevant data are listed.
Answer. (a) The index for an energy-efficient beam with a specified

stiffness, modified to include shape, is
(a) Start with stiffness. Derive the material index for stiffnesslimited, shaped beams of minimum mass. Adapt this to make
the index for stiffness-limited, shaped beams of embodied
energy by multiplying density by embodied energy /kg, H p .
where E is the modulus, the density, Hp the energy content
M1 =
Material
(b) Repeat the procedure, this time for strength, creating the
appropriate index for strength-limited shaped beams at
minimum energy content by adapting equation 9.28.
What do you conclude about the relative energy-penalty of design

with wood and with steel?
Modulus
E GPa
Strength
f MPa
Energy Hp
MJ/kg
Be
Soft wood
Reinforced
concrete
Steel
700
10
40
7.5
1.4
2900
35
10
1.4
7900
210
200
30
15
10
35
210
H p
3
(GJ/m )
5.25
14.5
238
Be
M1
1/2
3
(GPa /GJ/m )
2
2
15
0.86
0.57
0.23
(b) The index for an energy efficient beam of prescribed strength

is, f
f
B f
M2 =
H p
2/ 3
Where B is the shape factor for failure in bending. The next

table shows that, as before, wood is the most efficient material.
Material
Wood
Concrete
Steel
M.F. Ashby 2010
E
(GPa)
Wood
Concrete
Steel
Density
kg/m3
H p
per unit weight, and Be is the shape factor for stiffness controlled
design. The first Table shows values of M1 for the wood, the
concrete and the shaped steel beam. Wood wins.
Use the modified index to rank the three beams.
Material
( Be E)1/2
87
f
(MPa)
40
10
200
H p
3
(GJ/m )
5.25
14.5
238
Be
M2
2/3
3
(MPa /GJ/m )
1.4
1.4
4
2.8
0.4
0.36
Author
Open Access Resources include:
Professor Mike Ashby

University of Cambridge, Granta Design Ltd.
www.grantadesign.com
www.eng.cam.ac.uk
Reproduction
These resources are copyright Professor Mike Ashby.
You can reproduce these resources in order to use
them with students, provided you have purchased
access to Grantas Teaching Resources. Please make
sure that Mike Ashby and Granta Design are credited
on any reproductions. You cannot use these resources
for any commercial purpose.

You can register for a user name and password for these
resources here:
www.grantadesign.com/education/resources/
Other Resources Available:

Accuracy
Interactive Case Studies

Getting Started Guides
Materials Property Charts
Engineering Data Booklets
19 PowerPoint lecture units

Exercises with worked solutions
Recorded webinars
Posters
White Papers
Solution Manuals
We try hard to make sure these resources are of a high

quality. If you have any suggestions for improvements,
please contact us by email at
teachingresources@grantadesign.com.
M. F. Ashby, 2010
Grantas Teaching Resources website aims to support teaching of materials-related courses in Engineering, Science and Design.
The resources come in various formats and are aimed at different levels of student.
This resource is part of a set of resources created by Professor Mike Ashby to help introduce materials and materials selection to students.
The website also contains other resources donated by faculty at the 700+ universities and colleges worldwide using Grantas CES EduPack.
The teaching resource website contains both resources that require the use of CES EduPack and those that dont.
M.F. Ashby 2010
88

Exercises Exerciseswithworkedsolutions Ma Uoc en

Uploaded by

Copyright:

Available Formats

Exercises Exerciseswithworkedsolutions Ma Uoc en

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Exercises Exerciseswithworkedsolutions Ma Uoc en

Uploaded by

Copyright:

Available Formats

What are the different types of exercises covered in the document?

What are the different types of exercises covered in the document?

What are some of the important points about solving materials selection problems?

What are some of the important points about solving materials selection problems?

Exercises With

Professor Mike Ashby

M.F. Ashby 2010

CES EduPack Exercises with Worked Solutions

Exercises with worked solutions

M.F. Ashby 2010

CES EduPack Exercises with Worked Solutions

A request for a selection based on one material index alone (such

Three important points.

M.F. Ashby 2010

more: some materials with high E 1 / 2 / such as silicon carbide,

All the materials selection problems can be solved using the

CES EduPack Exercises with Worked Solutions

Writing implements (charcoal, lead (graphite), quill pens, steel

Watering cans (wood galvanized iron polypropylene)

Bicycles (wood bamboo steel, aluminum, magnesium,

Small boat building (wood aluminum GFRP)

Book binding (Wood leather cardboard vinyl)

M.F. Ashby 2010

CES EduPack Exercises with Worked Solutions

These two examples illustrate the way in which concepts are

Electric fan pulling air stream through

Central pump and filter linked to

Central turbo-fan with centrifugal

Centrifugal turbo-fan, injected water

Electric fan pulling air stream through

Central pump and filter linked to

Centrifugal turbo-fan with surrounding

Central turbo-fan , water spray to wash

Axial fan drawing air stream between

Central fan with electrostatic collector,

Reel-to-reel single sided adhesive tape

Answer. Design problems are open-ended; there are always

M.F. Ashby 2010

CES EduPack Exercises with Worked Solutions

Massive, with sufficient heat capacity to absorb heat

Answer. Four working principles are listed below: thermal conduction,

Free convection not requiring fan or pump.

Unconfined, such as a continuous spray of volatile fluid

Radiation to ambient, using high emissivity coatings

M.F. Ashby 2010

CES EduPack Exercises with Worked Solutions

Using material properties

E4.3 A thick-walled tube has an inner radius ri = 10 mm and an outer

Useful Approximate Solutions to Standard Problems (UASSP) is a

radius ro = 15 mm. It is made from polycarbonate, PC. What is the

E4.1 A cantilever beam has a length L = 50 mm, a width b = 5 mm and

Answer. The deflection of a cantilever under an end load F is, from

M.F. Ashby 2010

The mean yield strength of y of PC from the CES EduPack is 65 MPa.

E4.4 A round bar, 20 mm in diameter, has a shallow circumferential

CES EduPack Exercises with Worked Solutions

Answer. The strain difference caused by difference in thermal

= PMMA Low C steel T

PMMA = 117x 10-6/C