Life-cycle energy optimisation for sustainable vehicle design

Full text

F2014-LWS-070
LIFE-CYCLE ENERGY OPTIMISATION FOR SUSTAINABLE VEHICLE DESIGN
1,3O’Reilly, Ciarán J*; 1,3Göransson, Peter; 2,3Potting, José; 1,4Cameron, Christopher J; 1,3Wennhage, Per
1KTH Aeronautical and Vehicle Engineering Department, Stockholm, Sweden
2KTH Environmental Strategies Research Division, Stockholm, Sweden
3The Centre for ECO2 Vehicle Design at KTH, Stockholm, Sweden
4Swerea SICOMP AB, Linköping, Sweden
KEYWORDS – Life-cycle energy; Vehicle design; Optimisation; Functional conflicts
ABSTRACT
A methodology is presented in this paper, in which the trade-offs in energy between vehicle production,
operational performance and end-of-life are formulated as a mathematical problem that may be optimised. This
methodology enables the consideration of the life-cycle environmental impact, through the proxy of life-cycle
energy, in the very first stages of transport vehicle design where it can be concurrently balanced with other
functionalities. The methodology is illustrated through a sandwich panel design case study. The optimisation
results for this case demonstrate that a design solution does exist, which meets functional requirements with a
minimum life-cycle energy cost. They also highlight that a pure lightweight design may result in a solution,
which is sub-optimal from a life cycle point-of-view.
INTRODUCTION
Efforts to reduce the environmental impact of transport vehicles have been primarily focussed on reducing
energy consumption during the use stage of the life cycle, as it considered as dominant over the production and
end-of-life stages [1]. These efforts have mostly followed three basic strategies that are illustrated with the help
of Figure 1 – improving transport efficiency (i.e. reduced vehicle movement, 𝑊
!"#$%&'"(), improving engine
efficiency (𝐸!"#$%&'%%'($/𝐸!"#$) or by switching to energy sources that have less environmental impact (e.g. bio-
based fuels or clean electricity, 𝐸!"#$) [2,3].
Figure 1: Vehicle use-stage energy flow.
These efforts, however, address only part of the energy-use picture for vehicles. During the use stage, significant
energy is required to overcome the dynamic losses (from drag, inertia, friction, etc.) of the vehicle itself, 𝐸!"##,!.
They account for approximately 50% of the fuel consumption for a typical mid-sized car over a standardised
drive cycle [3]. These energy demands are intrinsic to the vehicle design, as a function of its structure and shape.
The vehicle design, as such, also offers an important starting point for reducing energy requirements and so
ELoss,E
Veh i c le
Engine
EFuel
ETransmission
WTransport
ELoss,V

energy consumption [4]. The substitution of lightweight materials for heavier traditional ones is an example of
where this work has begun [5]. Additionally, the impact from the entire life cycle should be considered. The
environmental impacts from the production and the end-of-life of the vehicle are not altogether insignificant
[1,3,6,7]. Moreover, as most of the strategies to reduce use-stage energy consumption may involve moving to
some extent the impact to other life-cycle stages, care must be taken to ensure the overall life-cycle
environmental impact has been reduced. For example, a switch to electric motors presumes that both the motor
and the electricity are produced in a way that is less harmful than burning fuel in an IC engine, and the use of
lightweight materials presumes that the energy savings in the use stage are not countered by impact overheads in
the production and end-of-life stages. A more holistic life-cycle approach to the design of the vehicle structure
and shape may potentially enable a substantial reduction of the overall life-cycle environmental impacts from
transport vehicles.
The life-cycle environmental impact from vehicles may be assessed using Life-Cycle Assessment (LCA) [8-10].
LCA is often a complex time and data intensive task [11,12]. The heavy demands of performing a thorough LCA
do not lend themselves to LCA being used directly in the development of new concepts, which potentially could
have a large effect on the vehicle’s life-cycle environmental impacts [13]. In order to make a full assessment of
the product, the design must be complete (or nearing completion) and comparisons are typically made between a
small set of already existing designs (for example see [1]). This is useful in identifying bottlenecks in the current
design. However, at this point it is usually too costly (and still unclear) to make major changes to a design, as to
do so would involve reassessing a large number of degrees of freedom, that have been fixed up to that point [14].
As a result, improvements tend to be small-step and incremental, and are firmly based within the existing design
paradigm. Ideally, life-cycle impact information should be known at the very beginning of the design process
when a large solution set can be searched for an optimal solution.
To break this chicken or egg dilemma, it may be useful to consider life-cycle energy inventory as a way of
including life-cycle considerations in the early vehicle design. Life-cycle energy sums up the energy per
constituent material, including the contributions related to cradle-to-gate production and waste processing of
these materials. It has been demonstrated that life-cycle energy may be used as a proxy for the environmental
life-cycle impact of systems in which energy plays a dominant role, such as transport systems [15]. Additionally,
as energy is also a relevant parameter for the engineering of a vehicle’s operational performance, it is amenable
to direct investigation in both engineering and environmental disciplines. For example, the energy saved in
accelerating a lightweight1 load-carrying vehicle body, as a function of its structure, may be compared to the life-
cycle energy changes resultant from selecting different materials. An ability to directly assess these relationships,
and the complexity with which they interact with each other, at an early stage in the vehicle may lead to vehicle
architectures that are departures from the current design paradigm, and which may prove to have a significantly
lower overall life-cycle impact when finally assessed thoroughly by LCA. This is the core rationale for the
present research work.
In this paper a life-cycle energy optimisation (LCEO) methodology is presented, in which the trade-offs in
energy between vehicle production, operational performance (i.e. as related to the vehicle structure and shape)
and end-of-life are formulated as a mathematical problem that may be optimised. The methodology enables the
consideration of the life-cycle environmental impact of a transport vehicle, through the proxy of life-cycle
energy, in the very first stages of design where it can be concurrently balanced with vehicle functionalities. The
focus in this paper is on the LCEO methodology itself and thus the complexity of the vehicle object, its
functionalities, and the optimisation there of, need to be kept deliberately simple. The methodology is therefore
demonstrated through a case study in which a sandwich panel is designed. The life-cycle energy balance, which
includes the material quantities and the usage of the vehicle, is examined and optimised within functional
constraints.
A LIFE-CYCLE ENERGY OPTIMISATION METHODOLOGY
The optimisation methodology presented here (and introduced in more detail in [16]) integrates a number of
cross-functional requirements in a multi-disciplinary way and, perhaps most importantly, it opens the way for a
design framework based on multi-disciplinary optimisation that in principle could be performed at any given
level of complexity [17,18]. A function is defined here as a solution-independent value-creating operational
activity. The vehicle’s primary or root function is within a wider transportation system and is defined here as to
1 Light-weighting vehicles is primarily driven by the use of weight as a proxy for environmental impact.
However, this is a limited parameter as it does not deal with production and end-of-life considerations and it
does not influence all major use-stage losses either as aerodynamic drag is independent of weight.

carry and move a payload (driver/passenger/cargo) from one location to another (i.e. to transport). The vehicle
system’s function is subject to a large number of parameters such as time, safety, cost of its production and
operation, environmental impact, etc. Each parameter is constrained in some way, so to continue the previous
examples, low time, high safety, low cost, low environmental impact, etc. are required. Constraints arise at all
levels within the system and limit the set of feasible design solutions. This primary vehicle function is achieved
through a large set of nested sub-systems, each with their own functions and constraints, for example, an engine
for energy conversion, a chassis for structural support, etc.
More often than not, conflicts exist between the requirements and constraints for different sub-systems and their
functions. For example at a vehicle level, a vehicle’s load carrying function is proportional to the vehicle weight,
whilst its movement (i.e. its acceleration) is inversely proportional to weight. Designing a vehicle involves
compromising between cross-functional conflicting requirements to find an optimal solution. However, as sub-
functions are quite often designed in isolation, sub-functional optimisation may, through secondary trade-offs,
further exacerbate the negative influences of functional conflicts. For example, transporting a heavier vehicle
requires a larger, usually heavier, power unit. This in turn increases the overall weight further and so further
increases the power need, leading to an escalation of the problem. In this light, the introduction of new
technologies, such as electric motors and lightweight composite materials, may bring direct benefits for a
particular sub-function. Additionally, they may also have indirect repercussions for other vehicle sub-functions
over the vehicle life cycle. A cross-functional framework is employed here so as to address exactly these types of
trade-off dilemmas across all life-cycle stages.
The selection of materials (with different strengths, masses and embedded energies) to fulfil a number of
functions with conflicting constraints is framed as a life-cycle energy optimisation (LCEO) question. The result
of this optimisation not only meets the functional requirements but also has the minimum environmental impact.
Life-cycle or accumulated energy, 𝐸!, for a vehicle sums up the energy contributions from the cradle-to-gate
production (or embedded) 𝐸!, use for transportation 𝐸! and waste processing 𝐸!, and may be expressed as
𝐸!𝑋=𝐸!𝑋+𝐸!𝑋+𝐸!(𝑋)
(1)
where
𝐸!𝑋=𝐸!,!𝑚!
(2)
𝐸!𝑋=𝑊
!𝑁
(3)
𝐸!𝑋=𝐸!,!𝑚!
(4)
and 𝐸!,! is the production energy for each of the constituent materials, 𝑚! the actual contributed mass of each
constituent material, 𝑊
! is the energy or mechanical work required to move the vehicle according to a
prescribed drive cycle, 𝑁 is the number of such cycles during the entire use stage of a vehicle, and 𝐸!,! is the
end-of-life energy for each constituent material. 𝑋 is a set of functional parameters that are to be optimised.
Although the framework presented here includes the end-of-life energy, it is not included further in the present
discussion. The focus in this paper is thus limited to the production and use stages. The production energy is
determined using a cradle-to-gate energy inventory for each material.
The energy consumed during the use stage of the vehicles life cycle is very much dependent on the makeup of
the vehicle and how it is driven. To standardise the usage parameters, the energy required for a prescribed drive
cycle is modelled following [19] where
𝑊
!=𝑊
!+𝑊
!+𝑊
!
(5)
𝑊
!=1−𝑟𝑚𝑔𝑐!𝐶!
(6)
𝑊
!=𝑚𝐶
!
(7)
𝑊
!=0.5𝜌𝑐!𝐴𝐶!
(8)
with 𝑊
!, 𝑊
! and 𝑊
! being the energy required to overcome rolling resistance, inertial resistance to acceleration,
and aerodynamic drag respectively. 𝑟 is the fraction (in %) of kinetic energy regained during deceleration, m is
the mass, 𝑐! is the rolling resistance coefficient, 𝜌 is the air density, 𝑐! is the coefficient of drag and 𝐴 is the
frontal area of the vehicle. The remaining terms in these equations are dependent only on the chosen drive cycle,
not on the vehicle, and are given in terms of the sum of discrete work increments ∆𝑠 over the drive cycle with

𝐶!=∆𝑠!, 𝐶!=𝑣!
!∆𝑠! and 𝐶
!=𝑎!∆𝑠!, where 𝑣! and 𝑎! are the incremental velocity and acceleration. The
unknowns in Equations (6)-(8) may be determined to meet functional constraints via engineering models and
methods.
CASE STUDY: A SIMPLIFIED VEHICLE COMPONENT
In order to illustrate the LCEO methodology, it is applied here to a simplified vehicle component, more
specifically a sandwich panel, with structural loading and dynamic functional constraints. The sandwich panel is
composed of two fibre-reinforced laminate face sheets and a foam core, as illustrated in Figure 2, and is simply
supported along its edges. The composition and thickness of these layers is left to the optimiser to choose and an
unsymmetrical configuration with respect to these parameters may result. The full dimensions of the panel are
1.5 m by 1.7 m, but in the simulations performed a ¼ model with symmetry conditions applied in the planar
directions is used.
Figure 2: Schematic of ¼ model sandwich panel to be optimised with top and bottom facing sheets and a core.
To determine the thicknesses and material mixtures required to fulfil the functional constraints for the panel, a
number of parameters were chosen as design variables. While implementing thickness variables is trivial, using
material properties as variables required a method of parameterisation to some meaningful physical quantity. The
concept of hybridisation is employed to give a continuous representation of the properties of a mixture of
materials [20]. 𝑃 in Figure 2 denotes a material’s engineering property – so the Young’s modulus, density or
Poisson’s ratio – which is to be proportionally determined depending on the volume fraction 𝑉 selected. A list of
the variables used is shown in Table 1. For the fibre-reinforced laminates, it should be noted that 40% of the
laminate is Epoxy resin and they are all symmetric layups consisting of [0 +/-45 90] fibre stacks.
Table 1: List of optimisation variables. The 𝑖 subscript here denotes the top (𝑖=1) and bottom (𝑖=2) face sheet.
Face Sheets
𝑉
!,!"
Volume fraction of carbon fibre reinforced laminate (assuming 40% epoxy in the
laminate)
𝑉
!,!"
Volume fraction of glass fibre reinforced laminate (assuming 40% epoxy in the
laminate)
𝑡!
Thickness of the face sheet
Core
𝑉
!,!"#
Volume fraction of Polyethylene foam
𝑉
!,!"#
Volume fraction of Polyurethane foam
𝑉
!,!"#
Volume fraction of Polivinylchloride foam
𝑡!
Thickness of core
A number of materials were considered for possible use in this sandwich panel. These are listed in Table 2 along
with the engineering properties [17]. The list could be extended as required for other components. Carbon- and
glass-fibre laminates were considered for the facing sheets with PET, PUR and PVC considered for the core.

Table 2: Materials considered and their engineering properties.
Material
Young Modulus
[MPa]
Density
[kg / m3]
Poisson’s Ratio
[-]
Carbon fibre
150000
1850
-
Carbon fibre laminate
57379
1500
0.3
Glass fibre
40000
1940
-
Glass fibre laminate
18794
2520
0.3
Epoxy
3200
1150
0.3
PET
130
100
0.3
PUR
0.07
22
0.3
PVC
100
110
0.3
The central idea of this methodology is that in order to achieve a low life-cycle energy design, functional
requirements should be the controlling factor and optimisation tools should be used to tailor the vehicle design to
meet the needs required whilst minimising the life-cycle energy. For a vehicle component, a typical usage
pattern and a set of functional requirements are needed to set up a proper optimisation problem. The functional
requirements are set as structural load capacity and structural dynamic properties. Verification that these
constraints have been met is achieved via the integration of a finite element solver into the LCEO methodology.
From the large set of feasible solution that meet these constraints, the solution with the lowest life-cycle energy
may be arrived at by mathematical minimisation of the function in Equation (1) using an optimisation algorithm.
In the present case study this LCEO may be described as
min(𝐸!𝑋)=min(𝐸!𝑋+𝐸!𝑋)
(9)
subject to the constraints
𝑉
!−1≤0;!𝑗=1,𝑁
(10)
𝑑!
𝑑!"#,!
−1≤0;!𝑘=1,2
(11)
1−𝑓
!
𝑓
!"#,!
<0;!𝑙=1,2
(12)
where 𝑑 is the displacement of the sandwich panel and 𝑓 is the frequency of the normal mode, 𝑗 denotes an index
for the 𝑁 constituent materials considered, 𝑘 an index for the two displacement constraint considered and 𝑙 an
index for the two normal mode constraints considered. The set of optimisation variables is given by
𝑋=𝑉
!,!" 𝑉
!,!" 𝑉
!,!"# 𝑉
!,!"# 𝑉
!,!"# 𝑉
!,!" 𝑉
!,!"!!!!𝑡!𝑡!𝑡!
(13)
For the case study presented here, three specific analyses have been used to incorporate the assumed functional
requirements – linear elastic response to localised loading, linear elastic response to distributed pressure, normal
modes analysis. Figure 3 visualises these using the 1/4 symmetric model used within the paper. The first load
case is static pressure applied to a square area of approximately 100mm in the centre of the panel. The second
load case is a static pressure distributed over the entire top surface of the panel. The third and fourth functional
constraints are the frequencies arising out of a normal modes analysis, targeting the first and second natural
frequencies of the panel with the chosen boundary conditions.
Figure 3: Four functional load constraints used – localised unit loading, global unit loading, normal modes analysis 1st mode and 2nd mode
(from left to right).

The max displacement and frequencies selected for these constraints are 𝑑!"#,!=2.5×10!!!𝑚, 𝑑!"#,!=
2.5×10!!!𝑚, 𝑓
!"#,!=50!𝐻𝑧 and 𝑓
!"#,!=215!𝐻𝑧.
For each of the constituent materials used in the LCEO analysis, the embedded energy is required and the
inventory data used in this work are shown in Table 3. In this proof of concept study, some gross assumptions
are made related to the energy required to actually prepare the materials used for the production process itself.
Table 3: Cradle-to-gate inventory data for materials considered. The Japan Automobile Research Institute (JARI) provided this data to the
authors through a direct communication.
Material
𝐸! [MJ / kg]
Form of material
Carbon fibre
286
Glass fibre
30
Assembled roving
Epoxy
137.1
PET
69.4
Bottle grade
PUR
101.5
PVC
56.7
Figure 4: New European Drive Cycle (NEDC).
In the present work the New European Drive Cycle (NEDC), shown in Figure 4, has been employed. The drive
cycle constants may be evaluated as 𝐶!=11,013!𝑚, 𝐶!=3,989,639!𝑚!/𝑠! and 𝐶
!=1,227!𝑚!/𝑠!. This drive
cycle was repeated for a set of selected total life-cycle driving distances. These were 60,000 km, 180,000 km and
360,000 km. In the use-stage energy model, the following parameter values were chosen – the fraction of kinetic
energy regained during deceleration 𝑟=15%, the rolling resistance coefficient 𝑐!=0.01, the air density
𝜌=1.2, the coefficient of drag 𝑐!=0.3 and the frontal area of the component 𝐴=1!𝑚!.
0 200 400 600 800 1000 1200
0
20
40
60
80
100
120
140
time [s]
velo city [km/h]

RESULTS
The optimisation was performed for two different cost functions, one taking only the use-stage energy (USE,
𝐸!) into account and one based on the life-cycle energy (LCE, 𝐸!) as proposed in [16]. The results are shown in
Table 4.
Table 4: Results of use-stage energy and life-cycle energy optimisation for three driving distances.
Distance [km]
60,000
180,000
360,000
Optimised function
USE
LCE
USE
LCE
USE
LCE
𝑉
!,!" [%]
96
23
90
25
100
33
𝑉
!,!" [%]
4
77
10
75
0
67
𝑉
!,!" [%]
100
72
100
73
100
75
𝑉
!,!" [%]
0
28
0
27
0
25
𝑉
!,!"# [%]
11
14
11
12
11
11
𝑉
!,!"# [%]
89
86
89
88
89
89
𝑉
!,!"# [%]
0
0
0
0
0
0
𝑡! [mm]
0.16
0.19
0.16
0.15
0.13
0.13
𝑡! [mm]
0.28
0.19
0.35
0.19
0.34
0.18
𝑡! [mm]
40.8
48.8
38.8
51.5
40.1
52.7
𝑚!"#$% [kg]
4.9
5.8
5.0
5.6
4.9
5.5
𝐸! [MJ]
678
559
716
555
708
557
𝐸! [MJ]
221
232
667
690
1328
1371
𝐸! (=𝐸!+𝐸!) [MJ]
899
791
1383
1245
2036
1928
DISCUSSION
In this paper, the methodology presented is applied in a proof-of-concept case study, which looks at the low-
energy design of a vehicle sub-functional unit, here a sandwich panel. This is a perhaps overly simplified
application. Nevertheless, it serves to illustrate the difference between choosing life-cycle energy and the use-
stage energy as an optimisation cost function. Based on the results in Table 4, some quite interesting
observations may be made already for this simplified vehicle component case study.
From the results it is quite clear that the USE solution tends to a minimum mass and, as could have been
expected, the lowest mass is found for the long-distance design. This is achieved through using a large portion of
carbon-fibre-based laminate in both face sheets. The thickness of the core is larger for the LCE solutions
compared to the USE solutions. While this is an efficient way to save mass (consistent with general sandwich
theory), the interesting outcome here is that it opens up for an increased use of the low-energy heavier glass-fibre
laminate in the face sheets. The embedded energy is between 15-20 % higher for the USE solutions compared to
the LCE solutions. Interestingly, the LCE solution is in all cases the heavier of the two, thus emphasising the fact
that minimising mass in order to save energy used for driving might lead to a sub-optimal design from a holistic
life cycle point-of-view.
Two distinct features emerge when analysing the results in Table 4. The total energy consumed is in all cases
lower when optimising for the LCE. In this respect the methodology has been successful in reducing the total
energy. The actual size of the reduction depends on the distance driven, with a 5-10 % reduction for longer
distances. To reach a low-energy design in the LCE cases, the core thickness is slightly increased with the
distance travelled.
The embedded or internal energy is remarkably stable in all usage profiles. The core composition is similar in all
cases investigated, with a large portion of low-density PUR foam blended with the stiffer PVC to obtain a
sufficiently high stiffness to uphold the functionality constraints of the panel. In all cases investigated the PET
foam is eliminated from the design, despite having a significantly lower embedded energy than the PUR, as its
engineering properties are less suited to the design constraints.

The methodology appears to be very promising. Further work is required to test the robustness of solutions for
more complex design cases, with greater numbers of functional constraints and with end-of-life considerations
included.
CONCLUSIONS
In this paper a life-cycle energy optimisation methodology has been presented and applied to the design of a
sandwich panel for use in a vehicle. The design has been optimised for use-stage energy and for life-cycle energy
with some marked differences in the outcomes. The life-cycle energy optimisation is successful in achieving a
design, which meets the functional constraints and requires less energy over its life cycle. The results
demonstrate that minimising mass in order to save energy used for driving might result in a sub-optimal design
from a holistic life cycle point-of-view.
ACKNOWLEDGEMENTS
The authors would like to acknowledge the contribution of the Japan Automobile Research Institute (JARI) who
provided the production energy inventory data. In particular, Shigeru Handa, Atsushi Funazaki and Tetsuya
Suzuki were central to this collaboration.
Additionally, the authors would like to thank Bombardier Transportation, Scania CV and Volvo AB for their
contribution to this work through participation in the Centre for ECO2 Vehicle Design.
REFERENCES
[1] Schweimer GW and Levin M. Life Cycle Inventory for the Golf A4. Research, Environment and Transport,
Volkswagen AG, Wolfsburg, Germany, 2000.
[2] EEA. Europe’s environment, The fourth assessment. Office for Official Publications of the European
Communities, Luxembourg, 2007, ISBN 978-92-9167-932-4.
[3] Sweeting WJ and Winfield PH Future transportation: Lifetime considerations and framework for
sustainability assessment. Energy Policy, 2012, 51: 927-938.
[4] Knittel CR. Automobiles on Steroids: Product Attribute Trade-Offs and Technological Progress in the
Automobile Sector. American Economic Review, 2011, 7: 3368-3399.
[5] Ermolaeva NS, Castro NBG, Kandachar PV. Materials selection for an automotive structure by integrating
structural optimization with environmental impact assessment. Materials and Design, 2004, 25: 689-698.
[6] Pehnt, M. Assessing Future Energy and Transport Systems: The Case of Fuel Cells. Part 1: Environmental
Performance. Int J LCA, 2003, 8 (6): 365-378.
[7] Lane B. Life Cycle Assessment of Vehicle Fuels and Technologies. London Borough of Camden, London,
UK, 2005.
[8] Consoli F, Alen D, Boustead I, Fava J, Frankling W, Jensen AA, de Oude N, Parrish R, Postlewaite D,
Quay B, Siégun J and Vigon B. (eds.). Guidelines for life cycle assessment: A code of practice. SETAC–
Europe/North America, Brussels Belgium, 1993.
[9] Fava J, Consoli F, Denison R, Dickson K, Mohin T and Vigon B. (eds.). A conceptual framework for life
cycle impact assessment. SETAC-Europe/North America, Brussels, Belgium, 1993.
[10] ISO 14044. Environmental management – Life Cycle Assessment – Requirements and Guidelines. Geneva,
Switzerland: International Organization for Standardization (ISO), 2006.
[11] Schmidt W-P, Dahlqvist E, Finkbeiner M, Krinke S, Lazzari S, Oschmann D, Pichon S and Thiel C. Life
cycle assessment of lightweight and end-of-life scenarios for generic compact class passenger vehicles. Int
J LCA, 2004, 9(6): 405-416.
[12] Munoz I, Rieradevall J, Domènech X and Gazulla C. Using LCA to assess Eco-design in the automotive
sector. Int J LCA, 2006, 11(5): 323-334.
[13] Delucchi MA. Estimating the Climate Impact of Transportation Fuels: Moving Beyond Conventional
Lifecycle Analysis Toward Integrated Modeling Systems Scenario Analysis. Journal of the Washington
Academy of Sciences, 2013, 99(3): 43-66.
[14] Kauffman J, Lee K-M, Lindahl M and Sundin E. Product Design Considerations for Improved Integrated
Product/Service Offerings. Springer Netherlands, 2013.
[15] Huijbregts MAJ, Rombouts LJA, Hellweg S, Frischknecht R, Hendriks AJ, van de Meent D, Ragas AMJ,
Reijnders L, and Struijs J. Is Cumulative Fossil Energy Demand a Useful Indicator for the Environmental
Performance of Products? Environ Sci Technol, 2006, 40(3): 641-648.

[16] Göransson P, Cameron CJ, O’Reilly CJ, Wennhage P, Potting J and Cerin P. Optimising life-cycle energy
for sustainable vehicle design. Submitted to Materials and Design, 2014.
[17] Cameron CJ, Lind E, Wennhage P, and Göransson P. Proposal of a methodology for multidisciplinary
design of multifunctional vehicle structures including an acoustic sensitivity study. Int J Vehicle Structures
Systems, 2009, 1: 1-13.
[18] Cameron CJ, Lind Nordgren E, Wennhage P and Göransson P. On the Balancing of Structural and
Acoustic Performance of a Sandwich Panel based on Topology, Property, and Size Optimization. Journal
of Sound and Vibration, 2014, DOI:10.1016/j.jsv.2014.01.025
[19] Koffler C and Rohde-Brandenburger K. On the calculation of fuel savings through lightweight design in
automotive life cycle assessments. Int J LCA, 2010, 15: 128-135.
[20] Ashby MF and Bréchet YJM. Designing hybrid materials. Acta Materialia, 2003, 51: 5801-5821.