Cost Optimization of Aircraft Structures MARKUS KAUFMANN Doctoral Thesis Stockholm, Sweden 2009

Cost Optimization of Aircraft Structures MARKUS KAUFMANN Doctoral Thesis Stockholm, Sweden 2009
Cost Optimization of Aircraft Structures
Doctoral Thesis
Stockholm, Sweden 2009
TRITA-AVE 2009-83
ISSN 1651-7660
ISBN 978-91-7415-500-6
Kungliga Tekniska Högskolan (KTH)
Department of Aeronautical and Vehicle Engineering
SE-100 44 Stockholm, Sweden
Akademisk avhandling som med tillstånd av Kungl Tekniska högskolan framlägges till
offentlig granskning för avläggande av teknologie doktorsexamen i lättkonstruktioner
fredagen den 11 december kl 10.15 i sal D2, Lindstedtsvägen 5, Kungliga Tekniska
högskolan, Stockholm.
© Markus Kaufmann, autumn 2009
Tryck: Universitetsservice US-AB
Optimization is intrinsically tied
to our desire to excel, whether we
are an athlete, artist or engineer.
Garret N. Vanderplaats [1]
The work presented in this thesis was carried out at the Department of Aeronautical and Vehicle Engineering at KTH. Funding was provided by the European
Framework Program 6, project ALCAS, AIP4-CT-2003-516092, and the Swedish
National Aeronautics Research Programme 4, project KEKS. The financial support
is gratefully acknowledged. Special thanks go to Dan Holm at Alfgam AB for the
use of Xopt and to Keith Garland and Andy Langridge at Galorath International for
the use of SEER. Finally, the collaboration with the people at Saab Aerostructures
is thankfully acknowledged.
”Why Sweden?” I’ve been asked many times during the last four years. Well, I’ll
never forget my professor’s words while he showed me corrugated sandwich structures
six years ago. ”Plastic’s fantastic!” – Dan, the idea of designing exceptional things
with an exceptional material was one of the reasons to follow your invitation and
pursue a PhD under your guidance. You and my co-advisor Malin Åkermo were a
constant source of enthusiasm, inspiration and motivation.
Another reason to return to Stockholm was the friendship with Peter & Frida,
Ylva, Staffan, Joakim & Sofie, Chris & Camilla, Johan & Lotta and Mättu & Camilla
(in order of appearance). Tack för alla oglömbara stunder vi hade tillsammans! A
warm thank you to my friends and colleagues at KTH, particularly to my in-house
buddies Chris and Mio, and to Ylva (who critically supported my endeavors to
improve the local language).
My brother was the most frequent flyer between Zurich and Stockholm, and
proofreader of so many things I’ve written during the last years. You’re a great guy
and always welcome to your ”second” family!
I would further like to thank my parents for the endless support they provided
during all the years. Thank you for your belief in me!
But first and foremost, I would like to thank Anneke, whose love and support
made this thesis possible. Soon a chapter of our life is over and I’m looking forward
to the continuation of our story – wherever it will be!
Markus Kaufmann
Stockholm in 2009
Composite structures can lower the weight of an airliner significantly. Due to the higher
process complexity and the high material cost, however, the low weight often comes with a
significant increase in production cost. The application of cost-effective design strategies is
one mean to meet this challenge.
In this thesis, a simplified form of direct operating cost is suggested as a comparative
value that in combination with multidisciplinary optimization enables the evaluation of a
design solution in terms of cost and weight. The proposed cost optimization framework
takes into account the manufacturing cost, the non-destructive testing cost and the lifetime
fuel consumption based on the weight of the aircraft, thus using a simplified version of the
direct operating cost as the objective function. The manufacturing cost can be estimated
by means of different techniques. For the proposed optimization framework, feature-based
parametric cost models prove to be most suitable.
Paper A contains a parametric study in which a skin/stringer panel is optimized for
a series of cost/weight ratios (weight penalties) and material configurations. The weight
penalty (defined as the specific lifetime fuel burn) is dependent on the fuel consumption of
the aircraft, the fuel price and the viewpoint of the optimizer. It is concluded that the ideal
choice of the design solution is neither low-cost nor low-weight but rather a combination
Paper B proposes the inclusion of non-destructive testing cost in the design process
of composite components, and the adjustment of the design strength of each laminate
according to inspection parameters. Hence, the scan pitch of the ultrasonic testing is
regarded as a variable, representing an index for the guaranteed material quality. It is
shown that the cost for non-destructive testing can be lowered if the quality level of the
laminate is assigned and adjusted in an early design stage.
In Paper C and Paper D the parameters of the manufacturing processes are upgraded
during the cost optimization of the component. In Paper C, the framework is extended by
the cost-efficient adaptation of parameters in order to reflect the situation when machining
an aluminum component. For different weight penalties, the spar thickness and stringer
geometry of the provided case study vary. In addition, another cutter is chosen with regard
to the modified shape of the stringer. In Paper D, the methodology is extended to the
draping of composite fabrics, thus optimizing not only the stacking layup, but also the
draping strategy itself. As in the previous cases, the design alters for different settings of
the weight penalty. In particular, one can see a distinct change in fiber layup between the
minimum weight and the minimum cost solution.
Paper E summarizes the work proposed in Papers A-D and provides a case study on
a C-spar component. Five material systems are used for this case study and compared
in terms of cost and weight. The case study shows the impact of the weight penalty, the
material cost and the labor rate on the choice of the material system. For low weight
penalties, for example, the aluminum spar is the most cost-effective solution. For high
weight penalties, the RTM system is favorable. The paper also discusses shortcomings
with the presented methodology and thereby opens up for future method developments.
This doctoral thesis is based on an introduction to the area of research and the
following appended papers:
Paper A
M. Kaufmann, D. Zenkert and P. Wennhage. Integrated cost/weight optimization
of aircraft structures. Accepted for publication in Structural and Multidisciplinary
Optimization, online version doi: 10.1007/s00158-009-0413-1.
Paper B
M. Kaufmann, D. Zenkert and C. Mattei. Cost optimization of composite aircraft
structures including variable laminate qualities. Composites Science and Technology,
68; 2748-2754, 2008.
Paper C
M. Kaufmann, T. Czumanski and D. Zenkert. Manufacturing process adaptation
for integrated cost/weight optimisation of aircraft structures. Plastics, Rubber and
Composites: Macromolecular Engineering, 38(2):162-166(5), 2009.
Paper D
M. Kaufmann, D. Zenkert and M. Åkermo. Cost/weight optimization of composite
prepreg structures for best draping strategy. Accepted for publication in Composites
Part A: Applied Science and Manufacturing, 2009.
Paper E
M. Kaufmann, D. Zenkert and M. Åkermo. Material selection for aircraft components
based on cost optimization. Manuscript submitted to Journal of Aircraft (Design
Parts of this thesis have also been presented as follows:
M. Kaufmann, D. Zenkert and P. Wennhage. Integrated cost/weight optimization
of composite skin/stringer elements. In Proceedings of the 16th International
Conference on Composite Materials (ICCM-16), Kyoto (Japan), 2007.
M. Kaufmann, T. Czumanski and D. Zenkert. Manufacturing Process Adaptation
for the Integrated Cost/Weight Optimization of Aircraft Structures. In Proceedings
of the 13th European Conference on Composite Materials (ECCM-13), Stockholm
(Sweden), 2008.
M. Kaufmann and D. Zenkert. Cost Optimization of Composite Aircraft Structures.
In Proceedings of the 22nd Nordic Seminar on Computational Mechanics (NSCM22),
Aalborg (Denmark), 2009.
M. Kaufmann. Integrated Cost/Weight Optimization of Composite Aircraft Structures. In Proceedings of the 9th Swedish Production Symposium (SPS’09), Gothenburg (Sweden), 2009.
I Introduction
1 Background
2 Structural Design of Aircrafts
2.1 Design Phases . . . . . . . . . .
2.2 System Integration . . . . . . .
2.3 Design of Composite Structures
2.4 Constraints and Allowables . .
2.5 Non-Destructive Testing . . . .
3 Design Optimization
3.1 Multiobjective Optimization . . . . . . . . . . . . . . . . . . . . . . . 17
3.2 Weight Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.3 Integrated Cost/Weight Optimization . . . . . . . . . . . . . . . . . 19
4 Cost and Cost Estimation
4.1 Life-Cycle Cost and Direct Operating Cost
4.2 Design and Cost . . . . . . . . . . . . . . .
4.3 Estimation of Manufacturing Cost . . . . .
4.4 Estimation of Non-Destructive Testing Cost
5 Forming of Composites
5.1 Models for Composite Forming . . . . . . . . . . . . . . . . . . . . . 29
5.2 Composite Modeler . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
6 Cost Optimization Framework
7 Conclusion
8 Future Work
II Appended papers
M. Kaufmann
Part I
1 Background
”In a switch that could make Airbus’s next jetliner more competitive with rival Boeing
Co.’s new 787 Dreamliner, the European plane maker plans to build the frame of
its planned A350 model from advanced composite materials instead of metal. The
lighter structure – similar to that of the Boeing plane – reduces fuel consumption,
increases a plane’s range and reduces wear on key parts such as landing gear. The
shift also cuts the need for costly maintenance inspections. [. . . ]”
This article, published by the Wall Street Journal on Saturday, September 15,
2007, summarizes the achievements in the field of aerospace structures over the last
couple of years. Material configurations are undergoing a change from metals to
composites, thus lowering the structural weight and avoiding fatigue and corrosion.
No cost aspects, however, are mentioned in the article. How much more does the
composite version cost compared to the metallic baseline? And how far could an
increase in manufacturing cost be motivated by the saving of weight?
Figure 1.1: Airbus A350
M. Kaufmann
The aircraft market’s competition forces the airline operators to save costs.
Hence, efforts are made to lower acquisition and operating costs. One possibility to
lower the latter is to reduce the fuel consumption of the aircraft. In Figure 1.2, the
development of the price for A1 jet fuel since the year 2000 is shown. As can be
seen, the fuel price quickly quadrupled in 2008 before returning to a level roughly
twice the price of 2000. Thus, the aircraft manufacturers are pushed even more to
design products with lower fuel consumption. But is there a way to balance these
low-weight and low-cost design objectives?
A first estimation of the impact of the structural weight on the lifetime fuel
consumption can be made by means of a simple fuel burn calculation. According to
Scandinavian Airlines, an airliner in the A330 class with 260 seats and a take-off
weight of 233 tonnes typically consumes 0.035 l/seat/km. Let us assume that the
average gross weight is about 200 tonnes and that the aircraft flies for 25 years, 300
days/year a range of 2·7000 km/day. Thus, the total flown distance in the life of
the plane is estimated to be 100 million km. With the above fuel consumption and
passenger utilization, the total life fuel consumption is 1 billion liters of jet fuel
or about 5000 liter per kilogram flight mass. At a fuel price of about e0.40/l (or
US$ 740/metric tonne), the lifetime fuel cost per kilogram gross weight is e2000/kg
This calculation is made for the average gross weight of an aircraft. The airframe
weight is only 20-30% of the gross weight. Thus, one can expect the monetary
impact of structural weight savings to be even higher, as any weight savings during
the very early design phases can be looped back through the whole aircraft design
process: any weight saving is also accompanied by the use of smaller engines or
smaller wings. As a consequence, the net savings in aircraft take-off weight is much
greater than the weight saved on the structure alone.
fuel price [US$ per metric tonne]
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Figure 1.2: Development of the price for A1 Jet fuel since 2000
(Data provided by Nordea Bank AB)
We could change the design objectives and consider a possible increase in revenue
instead of the fuel saving. Assuming an aircraft with one transatlantic return trip
a day, 300 days a year for 25 years and a ticket price of e500 would result in a
revenue of e3’750’000 per chair. If we further assume that 125 kg weight saving is
needed per additional passenger, a total of e30’000 could potentially be earned per
kilogram structure.
A lot of work has been performed in the field of pure weight optimization. The
application of a lifetime cost per kilogram airframe structure, however, provides us
with a tool to include cost and weight in the optimization of aircraft structures.
Gutowski et al. described the composite’s drawbacks already back in 1991 [2].
The cost of the composite raw material is roughly ten times the cost of aluminum.
The material cost, however, reflects only a small part of the total cost of the final
product. First, the labor cost to manufacture a composite component is much
higher than for a comparable metal part. Second, the production of composite
components brings a need for an extensive quality management and rigorous nondestructive testing. Third, the increased design complexity comes with an increase
in development cost.
These risks were also identified by Rais-Rohani and Dean [3]. According to
them, the high material, fabrication and tooling costs may make the use of highperformance materials cost-ineffective. Thus, the costs of raw materials, tooling,
fabrication, assembly, scrap, repair, certification and environmental factors should be
included in the design of composite structures. Sobieszczanski-Sobieski and Haftka
observed an absence of cost optimization frameworks for composite components [4].
Soutis pointed out that early composite designs were ”replicas of those that employed
metallic materials” from the 40s, as most composite parts were manufactured with
expensive hand-laid carbon epoxy prepregs using quasi-isotropic stacking sequences.
According to the authors ”the high material cost and man-hour-intensive laminate
production jeopardized their acceptance” [5].
In this thesis, a framework for the cost optimization of aircraft structures is
proposed. In Chapter 2, the three phases in the structural design of aircrafts are
presented. Chapter 3 introduces the field of multidisciplinary design optimization.
By means of a literature survey, existing cost/weight optimization frameworks
and their differences are presented. Chapter 4 deals with the definition of direct
operating cost and the estimation of manufacturing cost, followed by an overview
of composite forming in Chapter 5. Finally, an introduction to the proposed
optimization framework and the appended papers is given, followed by a short
discussion and the identification of possible future work.
2 Structural Design of Aircrafts
Design Phases
The design of aircrafts can roughly be divided into three design phases, the conceptual
design phase, the preliminary design phase and the detail design phase. They all
have a distinctive multidisciplinary character, the result being a tradeoff made by
all stakeholders.
Conceptual Design
In the conceptual design phase, numerous design alternatives are compared and
evaluated, based on cost/weight/passenger/range tradeoff studies. The result is an
initial aerodynamics and propulsion concept, including overall dimensions, weights
and global loads.
Preliminary Design
In the preliminary design phase, a global finite element model is built up from which
local loads and loading conditions are derived. An illustration of typical loads on
an airliner is given in Figure 2.1. As can be seen, aeroelastic loads such as tensile,
compressive and torsional loads in wing, fuselage and empennage represent only a
fraction of the load cases the structural engineers have to consider. Other loads
arise from the cabin pressure (hoop stress), bird strike or impact loads on the tail.
Very high local stresses can be found in the landing gear ribs, the sidestay fittings
and the pylon structure.
The task of the structural engineers is the design of the inner structure of the
aircraft. The design is constrained by the aerodynamic configuration. It has to
withstand all loads and should be as light as possible. Different levels of detail are
investigated in the preliminary design phase, see Figure 2.2. First, the structural
arrangement of the major parts, such as ribs and spars in the wing and lap joints and
butt joints in the fuselage are defined. Then, the structure is designed on a panel
level. The strength and stiffness of the structural members are defined and verified
by means of finite element models, while changes in the configuration (e.g. stiffener
distance, flange type, rib stiffeners, etc.) are still possible. According to Assler [6],
the design process is influenced by a variety of factors at this stage. Examples are
M. Kaufmann
Lower skin:
Bending and
Upper skin:
Longit. Tension
(transverse shear and torsion)
Hoop Tension
High Local Loads
Longit. Compression (bending)
Corrosion Resistance
Figure 2.1: Typical local loads on an airliner (by courtesy of H. Assler,
Airbus Deutschland GmbH)
Airworthiness regulations
Environmental considerations
General aircraft requirements (operational profile, maintenance, etc.)
Specific requirements for structural details
Available materials and technologies
Manufacturing capacities and capabilities
Non-destructive testing and investigation capabilities
Design cost
Detail Design
In the detail design phase, the structure is analyzed by means of high-fidelity
models, and the fabrication, tooling and assembly processes are defined. The result
is a detailed work breakdown structure including all structural parts, mounting,
bolts and rivets, clips, doors, brackets, etc. Each part has to fulfill its particular
requirements, based on structural failure, fatigue, corrosion resistance, lightning
strike, sealing, conduction, maintenance or testing.
System Integration
As the development of a new aircraft type involves great cost and – in turn – a great
financial risk, aircrafts are not designed, fabricated and assembled by one single
manufacturer anymore; the latter have been replaced by consortiums of system
integrators and suppliers. In Figure 2.3, the supply hierarchy in the commercial
aerospace industry is shown.
On top of the supply pyramid, a system integrator (e.g. Boeing or Airbus) is
responsible for the coordination, the overall design and the final assembly of the
Fuselage Section
Butt Joint
Structural Detail
Fuselage Panel
Figure 2.2: Levels of detail (by courtesy of H. Assler,
Airbus Deutschland GmbH)
aircraft. In a subordinate position, other members of the consortium take part in
the development of sub-systems (such as wing structures), its manufacturing, subassembly and delivery. Further down in the hierarchy are equipment and component
suppliers. They do not take part in the design process and are restricted to the
manufacturing and delivery of components. Estimated profit margins are the highest
in the top position of the pyramid and – according to Johansen et al. [7] – the higher
the position in the hierarchy, the greater the risk for that company in the overall
The embedding of a sub system supplier (e.g. Saab, Bombardier or Alenia)
occurs in the conceptual design phase, sometimes already within pre-development or
research projects. The benefits of this approach are the early exchange of knowledge
and experience, and a good integration of the otherwise widespread design teams.
As the risks are shared, each of the partners are under pressure to continuously
increase the efficiency in terms of low cost, low weight and producibility. The
main drawbacks mentioned in [7] are communication problems across the company
borders, such as cultural differences, hierarchical misunderstandings and delayed
System integrator
Sub system
Equipment supplier
Component supplier
Figure 2.3: Supply hierarchy in the commercial aerospace industry
M. Kaufmann
information flows. Further mentioned is the need for an extensive product lifecycle
management database (PLM). In the case of the A380, the PLM is used by more
than 5000 engineers and contains more than 3000 CAD drawings of about 150’000
Design of Composite Structures
As mentioned, the fuel consumption stands for a considerable part of the operating
cost of an airliner. As shown in Jacobsen [8], the fuel cost can be reduced by
developing more efficient engines, by minimizing the aerodynamic drag, by optimizing
the flight trajectory of the aircraft and by reducing its mass. The latter was the main
motive to change from metals to composite materials, and the portion of composite
materials of the total structural weight is continuously increasing in commercial
airliners, see Figure 2.4.
Composite Structural Weight [%]
1970 1975 1980 1985 1990 1995 2000 2005 2010
Figure 2.4: Portion of composite materials in Airbus aircrafts
(by courtesy of H. Assler, Airbus Deutschland GmbH)
The first application of composites to commercial aircrafts were radar domes
in 1940. Since then, composite materials have increasingly replaced their metallic
counterparts. In 1975, NASA developed a series of composite parts for research
purposes, and the elevators of the B727 and B737 and the vertical fin of the DC10
were redesigned. Secondary structures (e.g. the leading edge, trailing edge, flaps,
ailerons and rudder) were made of carbon fibers in the Boeing 777, and with the
center wing box of the A380, composites were used also for primary (load-carrying)
structures. The latter enabled weight savings of 1500 kg compared to the aluminum
baseline, see Marsh [9].
A comparison of the specific strengths and stiffnesses is presented in Figure 2.5.
As can be seen, the specific strength of carbon fiber reinforced plastic is much higher
than for aluminum or titanium, whereas the specific moduli are approximately the
same. A composite component that is designed for stiffness will therefore have a
higher safety factor against material failure than its metallic counterpart. This
characteristic accounts for the good fatigue behavior of composites, cutting down
the maintenance cost for the airlines. Another advantage is the possibility to tailor
composites specifically to a desired function. This is done by either adjusting the
fiber angle distribution or by unifying the geometry and thus reducing the number
of parts.
Specific strength σ/ρ
Carbon (QI)
10 15 20 25 30 35 40
Specific stiffness E/ρ
Figure 2.5: Specific strengths and stiffnesses of different metals and alloys, quasi-isotropic glass fiber reinforced plastic (Glass/QI)
and quasi-isotropic carbon fiber reinforced plastic (Carbon/QI)
Constraints and Allowables
The structure has to withstand anticipated external loads. For composites loaded in
tension, material failure might be the limiting constraint. For composites loaded
in compression, material failure and stability concerns form the topology of the
structure. The situations under which the integrity of the structure needs to be
proved are described in regulations published by the aviation authorities. Structural
constraints, for example, are based on the airworthiness requirements, defined in
JAR 25.613 Material Strength Properties and Design Values. This document is
released by the Joint Aviation Authorities1 , a European body representing the civil
aviation regulatory authorities of a number of European states. Similar regulations
exist also in the United States (FAR). Therefore it is often referred to the FAR/JAR
The determination of an allowable stress and strain limit is based on statistical
evaluation of specimen tests. A typical design strength, for instance, is the stress
level where at least 90% of the population pass with a confidence of 95%. These
M. Kaufmann
tests have to be performed for filled or unfilled holes, as stress concentrations around
fastener and bolt holes can be the cause for material failure. The test phase can
include up to 4000 coupons to generate complete data for a certification program,
see Niu [10].
A composite laminate can fail by different modes, e.g. by fiber failure, microbuckling, matrix failure or fiber/matrix debonding. Other effects are delaminations due
to pull-off loads, free-edge effects, poisson’s ratio mismatch or compressive buckling.
Therefore, most engineers (and aeronautical engineers in particular) tend to use
rather conservative failure criteria.
Unlike metal structures, composite structures are limited by strain and not by
stress concerns. The strain limits of composite laminates are only indirectly related
to the strain levels of the matrix or the fibers; it is rather a design strain based on
coupon tests. One of the limiting load cases of coupon tests is the compression after
impact test, simulating prior damage from tool drops and runway debris. There,
the remaining compressive failure strain of a damaged composite panel is evaluated.
Typically, values around 0.4% are obtained which is much lower than for unnotched
Albeit not being the most elaborate failure theory, the maximum strain criterion
is still widely used in the aerospace industry. The maximum strain criterion is given
as three independent sub-criteria
ε̂1,c < ε1 < ε̂1,t
ε̂2,c < ε2 < ε̂2,t
|γ12 | < γ̂12 .
The indices c and t denote compression and tension, respectively; 1 and 2 denote the
ply’s longitudinal and transversal direction, andˆdenotes the allowable strain value.
Apart from material failure, the design of composite laminates is governed
by a series of other rules. Examples given are the requirement for symmetrical
stacking, a minimum amount of 10% of each ply angle, or the location of fiber splits.
Aircraft engineers maintain a stacking sequence consisting of 0◦ , 90◦ and ±45◦ plies.
Although there exist stacking sequences that allow a significant weight reduction,
certification issues prevent their use yet.
Non-Destructive Testing
The Federal Aviation Administration’s regulations of airworthiness require the
quality assurance of each assembled part. Unlike metallic structures, composite
parts are fabricated in-situ and the grade of these structures is highly dependent on
process robustness and workmen skills. Typical manufacturing generated defects in
composites are voids, porosity, fiber misalignment, wrinkling, poor cure, resin-rich or
resin-poor areas, forgotten release papers and low-quality adhesive bonds. Therefore,
each composite part undergoes rigorous non-destructive testing (NDT) prior to the
For metallic and composite aircraft structures, NDT is also part of the damage
tolerance concept. Micro-cracks are basically tolerated under the condition that
the airliner is regularly checked for structural integrity (continuous monitoring and
sufficiently slow growth of cracks). In so-called D checks, complete overhauls at
six to ten years intervals, the paint is removed and cracks or delaminations are
sought. Apart from these regular checks, the integrity is also tested after bird strikes,
hard landings or similar incidents. All these inspections are very costly due to the
downtime of the aircraft.
Ultrasonic Methods
Due to the nature of composite structures, flaws can occur in monolithic structures
(porosity, delamination, cracks), the adhesive layers (debondings), or sandwich
cores (density irregularities, cracks). While flaws in the outer skin can be detected
with single-sided access, the underlaying defects often need through-transmission
scanning. Thick structures are generally more difficult to test than thin structures.
The most common method for the inspection of aircraft structures is ultrasonic
testing (UT). There, a transducer is passed over the area being tested. Ultrasonic
waves penetrate the structure, while the receiver records the reflected (pulse-echo
mode) or the transmitted (through-transmission mode) sound waves, see Figure
2.6. The screen on the diagnostic machine will show these results in the form of
amplitude and pulse readings, as well as the time of flight.
Figure 2.6: Ultrasonic test setup in through-transmission mode [11]
The presentation of the amplitude of the wave as a function of time (the so-called
A-scan) is sufficient for manual detection of flaws. Scanning along a given route
leads to the B-scan presentation with the in-depth position of the flaw as a function
of scan distance and time of flight. The C-scan in turn represents an areal defect
image of the scanned part by scanning a 2D-pattern, while the D-scan combines the
in-depth information of the B-scan with the C-scan. In Figure 2.7, the procedure
to obtain a C-scan is shown. The density of the scan pattern (separated by the
distance later referred to as the scan pitch) determines the size of the detectable
M. Kaufmann
The advantage of UT is the high sensitivity, permitting detection of extremely
small flaws. The penetrating power of UT is higher than for other methods, thus
allowing detection of flaws deeper in the part. Unfortunately, damping effects of
local inhomogeneities of composite structures (e.g. due to stacked prepregs) reduce
the reflected energy significantly. Thus, the resolution of a C-scan is very much
dependent on the material system, the porosity level and the thickness.
Thick structures are difficult to inspect due to high damping characteristics and
high structural noise. While air-coupled ultrasonic is possible in principle, this is
restricted to through-transmission mode. For pulse-echo testing, improvements are
still necessary, and water is therefore still common as the coupling medium between
the probe and the specimen.
Test Piece
Side View
Top View
C-Scan Presentation
Figure 2.7: Presentation of ultrasonic scan results as a C-scan [11]
Referring to the literature, the design of composite structures has not been
influenced by NDT aspects. In order to capture the full life cycle of a composite
component, however, NDT should play a role in an early design phase. Therefore, a
methodology was developed that included the parameters of the in-production and
in-service testing in the design process. Hence, the scan distance of the ultrasonic
C-scan was introduced as a variable in the design optimization. In a feature-based
model, the NDT cost was calculated from the scan distance (the scan pitch). Further,
the design allowables of the laminate were adapted, since the scan pitch had a direct
influence on the detectable flaw size. This methodology is presented in Paper B.
3 Design Optimization
Product design is a process where ideas are generated and screened, and concepts
are formulated, rephrased and rejected. This solution-finding process is iterative
and – in a wider sense – an optimization process. The former methods of trial and
testing, however, have been replaced more and more by abstract models. In the
field of aeronautics, such models can include flow models, cost models, structural
models, models of the material properties, or dynamic flight models. Often, several
models from different disciplines are necessary in order to represent the behavior and
characteristics accurately enough. Most of the costs of the final product are defined
in the conceptual design phase, and to neglect relevant design aspects in this phase
would be disadvantageous. Hence, the goal of concurrent aerospace engineering is to
gather knowledge from different disciplines by involving a multidisciplinary group
of engineers in the design process. Examples of involved areas are fluid mechanics,
statics and dynamics (engineering mechanics), mathematics, electrotechnology,
propulsion, control engineering, aircraft structures, materials science, production
engineering, aeroelasticity, avionics, risk and reliability, or noise control.
In recent years, attempts have been made to perform these design tasks simultaneously rather than sequentially. However, difficulties were encountered related
to the large amount of data that had to be shared, as well as cultural and communicative problems between members of different fields or with different backgrounds.
Another obstacle is often given by the company’s hierarchy: the information flow is
not provided, hierarchical structures do not promote concurrent engineering and
different departments are separated and distributed spatially.
One approach to incorporate the different disciplines in aircraft design into an
automated design environment is the use of multidisciplinary design optimization
(MDO). As shown in Figure 3.1, an MDO framework combines relevant design
disciplines and runs the analysis tools simultaneously. Feedback is given to an optimization algorithm that calculates a new design solution by means of mathematical
programming or stochastic search methods, and provides inputs for the next round
of evaluation. These steps are repeated until the objective function, e.g. the weight
of the part, is judged to be sufficiently minimized.
For an overview on MDO applications in the field of aerospace, it is referred to
the review article written by Sobieszczanski-Sobieski and Haftka [4]. They concluded
that most of the literature on multidisciplinary optimization covers the interaction
M. Kaufmann
Initial parametric
Update parametric
Analysis 1
Analysis 2
Analysis n
Form objective function
and constraints
optimization: find
new parameters
optimum model
Figure 3.1: Multidisciplinary design optimization
between aerodynamics and structures (see Raymer [12] and Bartholomew [13]) or
shape parametrization techniques (see Samareh [14]).
Multidisciplinary design optimization frameworks are continuously being developed, see Samareh and Bhatia [15] and Townsend et al. [16]. For implementations
like NASA’s FIDO1 project, requirements like an intuitive user interface, handling
of a large problem size and support of collaborative design aspects were important.
Salas and Townsend [17] described the requirements of such frameworks in detail.
A short review on cost considerations in multidisciplinary aircraft design was given
by Rais-Rohani and Dean [3]. In 1996, they proposed that costs of raw materials,
tooling, fabrication, assembly, scrap, repair, certification and environmental factors
should be included in the design of composite structures. They motivated their
reflection with former studies on the weight-to-cost relation of structures made of
advanced materials which showed that the high material, fabrication and tooling
costs may make the use of high performance materials cost-ineffective. According
to that article, material and fabrication costs of composite structures are the key
drivers and of comparable importance as the assembly and maintenance costs of
metal structures.
1 Framework
for Interdisciplinary Design Optimization
Multiobjective Optimization
The multiobjective optimization framework proposed in this work contains the
optimization of different, often contradictive goals, such as low-cost and low-weight.
Several ways to capture multiobjective design problems exist, see Marler and
Arora [18]. Two or more objectives should lead to an optimal geometry. Therefore,
they have to be incorporated mathematically into one objective function F . The
multiobjective design problem is given as
f1 (x)
 f2 (x) 
F (x) =  .  , n ≥ 2
 .. 
fn (x)
subject to
h(x) = 0,
g(x) ≤ 0,
x ≤ x ≤ x.
In our case, f1 and f2 represent the cost and the weight of a composite aircraft
part. Stability and failure criteria are represented by the inequality constraint g(x),
whereas x and x are lower and upper limits of the variable vector x. An equality
constraint h(x) is optional and not used throughout this work.
Goal Programming
One approach to solve the problem above is referred to as goal programming. Goal
programming uses one of the objectives as the objective function, whereas upper
values (goals) are set for the other objectives. For example, one could implement a
cost goal by setting an upper cost limit as a constraint to the optimization problem.
This problem can be formulated as
subject to
weight of a composite element
- prescribed load case
- maximum manufacturing cost.
On the other hand, one could do the reverse and optimize for cost only, while aiming
for a prescribed structural performance (weight goal). This is given as
subject to
cost of a composite element
- prescribed load case
- maximum weight.
Both formulations have the drawback that the target cost or the target weight has
to be defined in advanced; being a one-shot technique, the goal has a great influence
on the optimal solution, and a poor formulation could lead to inferior designs.
M. Kaufmann
Multilevel Programming
Sometimes, the objectives can be ordered hierarchically in terms of importance.
Hence, a top level objective function is defined and the set of points that minimize
this first level objective is sought. In a second step, the set is reduced to the points
that minimize the second level objective, and the method proceeds until the lowest
level objective has been minimized. An example for multilevel optimization of
aircraft structures is given by Gantois and Morris [19].
This work, however, deals with the tradeoff behavior of a combined cost/weight
optimization. It is then not possible to order the two objectives, as they are on the
same level and cannot be separated hierarchically. Thus, first weight, then cost or
first cost, then weight approaches would not make sense here.
Pareto Optimality
A topic closely related to multiobjective optimization is Pareto optimality. Imagine
a full search exploration of all possible designs of a structural part, e.g. the points
given in Figure 3.2. Each design solution is represented by a variable set x, a
manufacturing cost f1 and a weight f2 . The points that constitute combinations of
lowest cost and lowest weight are marked with cross symbols; they are called Pareto
A Pareto point is a point in the design space for which there is (a) no possible
design solution with a lower weight and the same manufacturing cost or (b) no
possible design with the same weight and a lower manufacturing cost. The curve that
connects all Pareto points is called the Pareto frontier. The Pareto frontier is of great
importance in multiobjective optimization, as it represents the tradeoff behavior
of the two objectives. A lot of optimization algorithms deal with the generation
of a complete Pareto frontier, thus providing a choice of possible solutions to the
optimization problem. Two classes of optimization algorithms developed to perform
this particular task are the Homotopy Techniques, see Watson and Haftka [20], and
the Normal-Boundary Intersection, see Das and Dennis [21] and Huang et al. [22].
Figure 3.2: Design solutions, the corresponding Pareto points x and
the Pareto frontier –.
Minimizing Weighted Sums
More promising is an approach called weighted sums where two or more objectives
are incorporated into one objective function and weighted by predefined parameters
αi . These parameters represent the tradeoff between the objectives and result in a
one-shot design solution. By varying αi , however, a range of solutions with different
cost/weight tradeoffs can be obtained. The objective function is given as
F (x) =
αi fi (x),
αi > 0,
i = 1, 2, . . . , n.
The approach of minimizing weighted sums is criticized by Das and Dennis [23],
as the generation of evenly distributed Pareto points fails. However, in the case of
a combined cost/weight optimization, one would use the parameters to establish
a relationship between the manufacturing cost and the structural weight, and the
generation of a Pareto frontier is not necessary. This approach is described in detail
below, where this cost/weight relationship is called weight penalty, representing the
”lifetime fuel burn cost per unit structural mass”.
Weight Optimization
A lot of work has been done within the field of weight optimization. Representatively,
the work of Kang and Kim [24] is mentioned, in which skin/stringer structures
similar to those in Papers A and B of this thesis were weight-optimized. As a second
example, Walker [25] studied the topology optimization of stiffened panels with
different stiffener configurations. His aim was to maximize the buckling load for a
given plate thickness by varying the ply angle of an angle-ply symmetric layup.
Integrated Cost/Weight Optimization
Research on cost optimization of aircraft structures started late, mainly due to the
lack of sophisticated cost models. In 1997, Sobieszczanski-Sobieski and Haftka wrote
”Very few instances can be found in which aerospace vehicle systems are optimized
for their total performance, including cost” [4]. Since then, a lot of progress has been
made. One of the earlier studies that integrated costing information into the design
process was performed by Geiger and Dilts [26]. In 1996, they presented a conceptual
model for an automated design-to-cost approach; the main aim was the provision of
a framework that helped the structural engineer with the decision-making process.
Heinmuller and Dilts applied this framework to the aerospace industry. They
explained the design-to-cost concept as a tradeoff between operational capability,
performance, schedule and cost. In 1997, they concluded that ”enabling automated
design-to-cost in a typical aerospace manufacturing company will be a difficult and
time consuming process”. The following ten years would confirm that they were
right [27].
M. Kaufmann
Minimum Life
Cycle Cost
Life Cycle Cost
Figure 3.3: Tradeoff between acquisition and operating costs, see [28].
At Boeing, an approach for the life-cycle design of aircraft concepts was taken
by Marx et al. [28]. Three material configurations for a wing of a high-speed civil
transport aircraft were considered. It was concluded that lower operating cost could
be achieved by a costly design with higher reliability (less maintenance, downtime,
etc.), and – vice versa – that the lowest acquisition cost does not always signify the
lowest life-cycle cost (LCC). As shown in Figure 3.3, the point depicted as Minimum
Life Cycle Cost would be the best alternative for both the manufacturer and the
airline. Marx et al. included R&D cost, manufacturing and sustaining costs, and
In 1999, Gantois presented a PhD thesis where manufacturing cost was taken into
account for the multidisciplinary optimization of airliners [19]. The objective function
was formed by weight and drag components, and a sub-level cost optimization was
implemented in order to achieve the lowest manufacturing cost possible once the
super-level goals were reached. The optimization was accomplished by a topological
optimization (number of ribs, number of stiffeners).
At Rolls-Royce, it was observed that the traditional separation of an organization
into a design and a cost department was a source of frustration and delay to the design
process. Therefore, the Design Analysis Tool for Unit-cost Modeling (DATUM)
project was launched in 2002. The aim was (a) an understanding of the current
costing tools available on the market, (b) the development of an own costing tool
that would support design decisions throughout the development process and (c) its
application to an optimization framework. The DATUM project was described by
Scanlan et al. [29].
Park et al. [30] pursued the optimization of structures considering mechanical
performance and manufacturing cost. For the design of a resin-transfer-molded part,
the stacking sequence of a composite plate was optimized in order to maximize
the stiffness. Simultaneously, the mold filling time was minimized by changing the
number and position of resin injection gates. They used a weighted sum approach
with the displacement and the filling time as the two objectives.
Edke and Chang published a paper entitled ”Shape optimization of heavy load
carrying components for structural performance and manufacturing cost” in 2006 [31].
They presented a cost optimization framework that minimized the machining cost
of an aluminum torque tube subject to stress constraints. A Sequential Quadratic
Programming method was chosen for the optimization of six design variables, and
the objective function was formed by the material cost, the machining cost and the
tooling cost. The machining times were estimated in a virtual machining model
in Pro/MFG. The part weight, however, was not included in the formulation of
objective function or constraints.
In 2007, Curran et al. proposed a method to include the manufacturing cost of a
metallic skin/stringer panel in Dassault’s V5 platform [32]. The manufacturing cost
was processed in MS Excel, whereas structural constraints were formed by means of
closed-form solutions provided by the Engineering Sciences Data Unit (ESDU) [33].
This enabled cost optimization of metallic structures, automatic update of a CAD
model and simulation of assembly processes.
Cost/Weight Objectives
As seen above, most research was performed on minimizing weight or manufacturing
cost while maintaining a given structural performance (goal programming). When
reduction of both cost and weight is sought, however, the two objectives have to be
incorporated into one objective function.
Preparatory work was done by Kassapoglou [34–36]. First, stiffened composite
panels were optimized separately for minimum cost Cmin and minimum weight
Wmin , and in a second step, the objective functions
F (x) = α1
F (x) =
C − Cmin
W − Wmin
+ α2
(C − Cmin )2
(W − Wmin )2
+ α22
were applied. Here, C and W represented the actual cost and weight in each
iteration, respectively. The idea was resumed by Kelly and Wang [37] and Wang et
al. [38]. They proposed a simplified objective function on the form
F (x) = C + 500 · W
where $500/kg represented the ratio between cost and weight. This methodology
was applied to the optimization of a closed box structure, an aileron and a Krueger
Curran et al. [39–44] developed a similar framework for the optimization of an
aluminum fuselage panel. Structural constraints were formed by the von Mises
criterion, and local and global buckling coefficients. The manufacturing cost consisted
of material, fabrication and assembly costs, and the objective function was given as
F (x) = α1 · C + fuel burn
or F (x) = α1 · C + α2 · W
M. Kaufmann
where α1 was 2 or 3.5 and α2 = $300/kg. Only metallic structures were considered,
presumably due to the readily available buckling data from ESDU.
A similar approach was used by Iqbal and Hansen [45] for the optimization of
welded truss structures. Apart from balancing manufacturing cost and weight, they
also proposed a tradeoff function between the weight and the structural compliance.
In an outlook they suggested how the combination of weight (or cost including
weight) and compliance could be integrated in the design of aircraft structures.
The approach of a weighted sum has some advantages when it comes to cost/
weight optimization. First, it can easily be implemented in an optimization framework. Second, the combination of cost and weight (weighted by parameters αi ) gains
an economical significance. In this thesis, Curran et al.’s approach was extended
and the component’s share of the direct operating cost was used as the objective
function. For details it is referred to Chapters 4 and 6.
4 Cost and Cost Estimation
Before turning to the field of cost estimation techniques, it is worth to introduce
the definitions and concepts of cost as given in Roskam [46].
The cost of an aircraft is the total amount of expenditures/resources needed to manufacture that aircraft.
The price of an aircraft is the amount of dollars paid by
customers, e.g. the operator.
Profit is Price minus Cost.
Depending on the role in the lifecycle of an aircraft programme, another viewpoint
is taken. A part supplier, for instance, might offer his product at the lowest possible
price in order to stay competitive. His aim is therefore to minimize the manufacturing
cost. The aircraft manufacturer (system integrator), on the other hand, needs to
provide an aircraft which has low design and manufacturing cost and is competitive
in terms of operating cost. The operator is interested in cost savings throughout
the lifetime of the aircraft, i.e. low acquisition, operating and disposal costs.
Life-Cycle Cost and Direct Operating Cost
The life-cycle cost analysis investigates the cost of a system or product over its
entire life span. According to Roskam [46], the analysis of a typical system includes
costs for
Planning and conceptual design
Preliminary design and systems integration
Detail design and development
Manufacturing and acquisition
Operation and support
Note that most of the cost impacts are defined in the earliest design phases, whereas
the bigger part of the expenditures occurs during the operation and support phase.
M. Kaufmann
The life-cycle cost (LCC) of an airliner program can be expressed as
LCC = Crdte + Cacq + Cops + Cdisp
Crdte = Research, Development, Technology and Evaluation
Cacq = Manufacturing and Acquisition
Cops = Operating Cost
Cdisp = Disposal Cost.
For details on LCC, it is referred to the articles on cost models and LCC cost
estimation written by Asiedu and Gu [47], Durairaj et al. [48] and Woodward [49].
The commercial aircraft operators are not particularly interested in LCC. For
them, the operating cost and the direct operating cost (DOC) are more important,
as they represent the cost of flying the aircraft. The latter can be described as
DOC = Cflt + Cmaint + Cdepr + Clnr + Cfin
Cflt = f (crew, fuel, insurance)
Cmaint = f (maintenance, repair, overhaul)
Cdepr = f (price, flight hours)
Clnr = f (landing and navigation fees, registry taxes)
Cfin = f (financing strategy)
where the indices flt, maint, depr, lnr and fin denote flight, maintenance, depreciation, landing, navigation and registry, and financing, respectively. For the sake of
competitiveness of an aircraft, it is desired to minimize the DOC given in Equation
(4.2). On the basis of aircraft components, however, it is difficult to model the DOC.
Crew and insurance costs and the financing strategy, for instance, are unaffected
by the structural design and could be excluded from the designer’s point of view.
Therefore, most of the cost optimization approaches shown in Section 3.3 use a
reduced form of the expression given above. Only the fuel cost (as a function of the
weight) and the acquisition cost are included in these models.
Here, we go a step further and include the testing and maintenance cost as well.
The items in Equation (4.2) examined in this work are marked in bold.
Design and Cost
Different design guidelines exist that integrate cost as an inherent element. These
guidelines can be classified in concepts such as design to cost [50], design for
cost [51, 52], design for manufacturability [53], design for assembly [54, 55], design
for manufacturability and assembly [56], the Hitachi assemblability evaluation
method [57] and integrated product and process development [58]. Two of them,
design to cost (DTC) and design for cost (DFC), require some explanation.
Design to cost can be regarded as goal optimization with a specified cost target
Cmax while the structural performance F (x) is maximized. Thus, the mathematical
formulation is given as
subject to
F (x)
gj (x) ≤ 0
j = 1, 2, . . . m
C(x) ≤ Cmax .
Design for cost, on the other hand, maintains a prescribed structural performance
Fmin while the cost C(x) is minimized. This optimization problem is given as
subject to
gj (x) ≤ 0
j = 1, 2, . . . m
F (x) ≥ Fmin .
Estimation of Manufacturing Cost
According to Niazi et al. [59], there is a distinction between qualitative and quantitative cost estimation techniques. Qualitative techniques estimate the cost based on
previously manufactured products and scale the manufacturing cost on the basis of
similarities, whereas quantitative techniques are based on design features, manufacturing processes and the material. As seen in Figure 4.1, both groups of estimation
techniques can be subdivided further. Two classes are highlighted: feature-based
and parametric cost estimation techniques.
An example of a feature-based representation is a brick with a hole. The hole
can be considered as a feature in the brick, closely connected to the manufacturing
process used to create it, e.g. by drilling. In the same manner, the manufacturing
cost can be estimated. Starting from raw material or semi-finished products, process
features are added and their costs estimated.
Parametric techniques, on the other hand, are based on cost relevant design
parameters, so-called cost drivers. The functions that relate the cost drivers to
cost are commonly termed cost estimation relationships (CERs). In the example
above, the brick would be represented by a cuboid and a cylinder, and a CER links
the design parameters (height, length, depth, hole diameter) to the manufacturing
cost [60].
Several commercial tools for the estimation of manufacturing cost are available.
Most popular among them are SEER by Galorath Inc.1 and the software package
by Price Systems2 . SEER-MFG (formerly known as SEER-DFM) was selected in
this work, since it combines a feature-based approach (manufacturing processes can
be added, removed and altered as features) with the advantages of a parametric
M. Kaufmann
Product Cost Estimation Techniques
Qualitative Techniques
Quantitative Techniques
Intuitive Techniques
Analogical Techniques
Decision Support
Neural Network Model
Analysis Model
(a) Qualitative cost estimation techniques
Product Cost Estimation Techniques
Qualitative Techniques
Quantitative Techniques
Parametric Techniques
Analytical Techniques
Cost Estimation
Cost Models
Cost Estimation
(b) Quantitative cost estimation techniques
Figure 4.1: Classification of cost estimation techniques (by courtesy of
Thomas Czumanski)
estimation. This provides a great level of detail and sensitivity for the optimization
with respect to cost. Here, the basics of SEER-MFG are described in brief.
SEER-MFG uses kinematic and semi-empirical algorithms to derive direct and
indirect labor and tooling times along with material and other expenses. In principle,
the total cost of a unit subjected to an operation is given as
Clabor + Cmaterial + Csetup + Ctooling
where Clabor
f (time, directLaborRate, partComplexity)
f (volume, density, price/kg)
f (setupTime, setupLaborRate, partComplexity)
f (operationType, produtionQty, toolingComplexity)
First, the time needed to perform the operation is estimated as function of
geometric parameters. Multiplied by the directLaborRate (which includes the cost
for the operator, the cost for the machine or facility and some overhead cost) the
labor cost is calculated. Similarly, the volume of the raw material is multiplied
with its density and the price/kg. The setup cost is based on the time estimated
to prepare each operation, whereas the total tooling cost (for molds, dies, etc.) is
divided by the production quantity.
Estimation of Non-Destructive Testing Cost
It is possible to estimate the cost of NDT within SEER-MFG. However, in order to
be more flexible regarding the calculation of in-production and in-service NDT, it
was decided to implement an own feature-based cost model. Therefore, a generic
database was developed in collaboration with Christophe Mattei from Linköping
University. The database provides the following features:
Thin flat laminate with access from both sides
Thin flat laminate with access from one side
Thick flat laminate with access from both sides
Thick flat laminate with access from one side
Adhesive bond
Each of these features requires input data in form of length, width, thickness and –
if applicable – radius. Further, the feature definition includes a scanning technique
(pulse-echo or through-transmission), a complexity index, the educational level of
the operator and, associated with the latter, a cost per hour or per scanned area.
Hence, the cost for each feature is estimated, and the total cost of non-destructive
testing is the sum of all feature costs. For more details it is referred to Paper B.
5 Forming of Composites
The material properties of a composite laminate are functions of the fiber orientations
of the single plies. These fiber orientations are strongly dependent on the component’s
geometry, the properties of fibers and matrix (e.g. the stiffness of a fiber bundle and
its orientation) and parameters, such as temperature, applied force and deformation
rate during the forming process.
Problems that may arise during manufacturing are voids, wrinkles and fiber
bridges, fiber misalignment, tolerance mismatches, radius thickening or thinning,
residual stresses, global shape distortions or spring-in effects. Thus, a lot of experience is needed to design components that are manufacturable.
In recent years, a lot of research has been performed in the field of forming
processes. This research has led to better understanding of forming processes and
the influence of process parameters, and to the development of simplified models for
the design of producible complex composite components. Several existing material
systems have been examined and ranked in terms of formability, and new material
systems developed in order to improve the manufacturing process. This chapter
gives a short overview of the mechanisms affecting forming of composites, draping
models and commercial software. For more detailed information it is referred to the
book ”Composites forming technologies”, edited by A.C. Long [61].
There are different composite manufacturing processes, such as hand-laid prepreg, RTM or vacuum infusion. They all share the complexity of forming a flat dry
preform or a prepreg into a 3D shape before curing. The kinematics of ply draping
are generally the same, which makes it possible to use fairly simplified models for
the estimation of the fiber angles. The existence of the matrix, however, affects the
friction between plies and the mechanisms in the ply.
Models for Composite Forming
During forming of composites, the fibers interact with each other and with the
tool. The forming mechanisms are basically the same for dry fiber preforms and
prepregs. The additional stiffness of the viscous prepreg matrix, however, raises
the complexity in terms of temperature and rate dependencies. In Table 5.1, the
different deformation mechanisms during forming of composites are shown.
M. Kaufmann
Intra-ply shear
Intra-ply strain
Inter-ply shear
Ply bending
- In-plane shear of the material
- Rotation and slip between warp
and weft yarns
- Picture frame test and bias extension test for characterization
- Rate and temperature dependency for prepregs
- Extension (or compression) parallel to tow directions
- Standard tensile test for characterization
- Initial stiffening due to crimp in
- Relative movement between layers, and between layers and tools
- Pressure, and (for prepregs) rate
and temperature dependency
- Bending of individual layers
- Rate and temperature dependency for prepregs
- Important for forming of structures with single curvature
Table 5.1: Deformation mechanisms during forming, see Long [61].
The orientation of the fibers, primarily the angle between warp and weft yarns,
governs a series of properties, such as structural properties, the coefficients of
thermal expansion, local fiber volume fractions and the permeability of the material.
Thus, it is the aim of most material models to model the fiber reorientation as a
function of the shape of the component and the process parameters. Several model
classifications exist, whereof kinematic models and mechanical models can be seen
as the two main groups.
Kinematic Models
The simplest draping model is the mapping algorithm based on research performed
in the 50s by Mack and Taylor [62] and Van West et al. [63], also known as the
Pin-Jointed Net model (PJN). This model is illustrated in Figure 5.1.
Figure 5.1: Pin-Jointed Net model
A raster of points is mapped to the tool surface by keeping the fiber segments
constant. The algorithm solves the kinematic relationships given as
(xij − xi−1,j )2 + (yij − yi−1,j )2 + (zij − zi−1,j )2 = a2
(xij − xi,j−1 )2 + (yij − yi,j−1 )2 + (zij − zi,j−1 )2 = b2
F (xij , yij , zij ) = 0
where x, y and z are point coordinates, a and b are the distances of the fiber
segments and F is the surface equation. The algorithm is applied point by point,
thus finding the coordinates of the new point (i, j) based on the two existing points
(i − 1, j) and (i, j − 1).
Examples for commercial design tools using a kinematic approach are PAMQUIKFORM1 , FiberSIM2 , Interactive Drape3 or Composite Modeler4 . None of
these tools account for yarn slippage, shear locking, yarn bending, or boundary
conditions such as blank holding forces. This is subject to to further research, see
Potter et al. [64–66], Wang et al. [67] and Wiggers [68].
Truss based models are slightly more complex than mapping algorithms as they
take into account the shear stiffness of the material. As can be seen in Figure
5.2, the model consists of truss elements which are connected by diagonal spring
elements. The trusses represent the fibers, whereas the springs are responsible for
the shear stiffness, see Nguyen et al. [69]. Little input and time is needed for the
calculation in FE, which is one of the big advantages compared to the mechanical
models described below. Main limitations are the missing matrix behavior, and the
neglect of inter-ply shear in the simulation of hot draped composite stacks.
M. Kaufmann
side truss
Figure 5.2: Unit cell of a truss model with one or two diagonal spring
Mechanical Models
The mechanical models are most complex, as they rely on non-linear elastic or
viscoplastic, and sometimes even bi-component material models representing the
elastic fibers, the viscous matrix and the friction between yarns. The calculation is
performed using non-linear and/or explicit FE code, thus being rather extensive in
computation time, see Boisse et al. [70] and Badel et al. [71]. Tool and laminate
geometries, material properties and appropriate friction laws have a great impact
on the output and should therefore be given appropriate attention. Examples of
commercial codes are AniForm5 and PAM-FORM6 .
Multi-Layered Models
Most of the model approaches above can only describe single layers or an assembly
of single layers. When draping a multi-layer material (such as hot draped prepreg
stacks), the inter-ply behavior is important. The simulation can be done in two
First, one can use a shell element for each layer of the stack, and implement
friction laws which take into account the interaction between the shell elements
while being formed. The total number of degrees of freedom (DOFs) and the number
of contact evaluations grow linearly with the number of stacked layers.
Alternatively, one can use only one through-the-thickness element and incorporate
the drape behavior (including slip, inter-ply and intra-ply shear) in this single layered
shell element. This method is advantageous regarding the computation time. On
the other hand, an extensive number of material parameters are needed to describe
the material behavior, in particular for the intra-ply behavior.
Kinematic and mechanical approaches provide tools which can simulate the composite
forming process with acceptable computational efforts. A limitation is certainly the
need for material properties over the full temperature range in the process, which
calls for an excessive experimental effort. Another challenge is the modeling of
interlaminar shear and tool/ply friction for which pressure, temperature, deformation
rate, fabric structure and orientation are recognized as governing parameters. The
underlying mechanisms, however, are subject to further research [72].
According to the literature, the use of draping models should not only allow for
a more detailed assessment of the draped fabric, but also for an optimization of the
draping strategy, see Guillermin [73]. Examples of possible areas of optimization are
maximum intra-ply shear deformations, fiber orientations, positions of darts and
splices, and location of the point of initial contact during draping. The optimization
of the forming of woven composites was performed by Skordos et al. [74], who
used a truss model to simulate the draping of a hemisphere and compared it with
experiments. The same model was used to reduce wrinkling in the subsequent
optimization of the holding force along the edge of the fabric. A genetic optimization
algorithm was used for this purpose.
In Paper D, Guillermin’s idea to optimize for best draping was implemented and
exemplified by means of a case study. Thus, the result of a draping strategy was
examined with objectives such as material consumption, total cut length prior to
draping, resulting fiber angles and material shear, see Figure 5.3. For this purpose,
the commercial draping tool Composite Modeler was used.
cut length
fiber angles
Figure 5.3: Structural and economical impact of the draping strategy
in terms of resulting fiber angles, material consumption,
cut length and drapability.
M. Kaufmann
Composite Modeler
Composite Modeler is a kinematic draping tool developed by Simulayt Inc. and
forms an add-on to ABAQUS/CAE. It maps unidirectional and woven composites
onto a previously modeled shell surface and stores the draped lamina information in
sections. Input parameters are shown in Figure 5.4 and contain:
Material type The type of the material (weave or tape) has a distinct influence
on the shear mechanism (i.e. scissoring or sliding mechanisms).
Seed point and seed curves The seed point is the location of the initial contact
between a draped ply and the underlying mold or ply stack. A seed curve constrains
warp or weft directions along one or two paths on the surface.
Reference angle The reference angle is the angle between the 1-axis of the ply
and a reference coordinate system.
Extension type The extension type governs the draping kinematics. Three
extension types are available in Composite Modeler:
The first extension yarn lies in a geodesic direction nearest
the principal axis.
The draping proceeds from a seed point minimizing shear
strain deformation energy.
The draping proceeds from a seed point minimizing the
maximum shear.
Other parameters Other parameters cover the maximum shear angle during the
draping of the fabric (also known as shear locking angle), warp/weft angles and
warp/weft ratios. In addition, the thickness and angle tolerances for the generation
of the ABAQUS sections can be defined.
seed point
1-axis of the draped ply
reference angle
draped 0°
fiber angle deviation
at control points
nominal 0°
Figure 5.4: Definition of seed point, reference coordinate system, reference angle φ, nominal fiber direction, draped fiber direction
and fiber angle deviation at control points.
Use of Composite Modeler
The procedure of using ABAQUS/CAE and Composite Modeler is the following:
mesh the part, call the draping plug-in, define layup, material and draping properties
(such as seed point location and reference angle), export the flat pattern, and map
the simulated composite properties back onto ABAQUS sections. During the draping
calculation, a PJN mesh as shown in Figure 5.1 is applied to the shell surface, and
the internal data points (such as fiber angles and thicknesses) are averaged to section
properties. A recently integrated feature allows access to the output data not only
on the screen, but also by means of a text file (.vfp). The data comprise:
Resulting fiber angles Stored in ABAQUS section definitions, shown on the
screen as a ply stack plot and saved in the .vfp file, the latter containing the internal
PJN data points.
Fabric shear Shown on the screen and written to the .vfp file. Colors depict
areas where the maximum fabric shear is reached and risk for wrinkling occurs (see
Figure 5.5).
Figure 5.5: Screen shot of the fabric shear modeled in Composite Modeler. The area with risk for wrinkling is emphasized with
Ply thickness The ply thickness is stored as part of the ABAQUS sections. In
addition, the internal data points of the PJN model are saved to the .vfp file.
Flat pattern The shape of the ply when undraped, see Figure 5.3. This shape
is saved as a drawing exchange format (.dxf) file which can be post-processed by
means of scripts in order to obtain ply area, scrap ratio and perimeter of the cut,
see Lang [75].
6 Cost Optimization Framework
The result of this thesis is a multiobjective optimization framework for aircraft
structures that incorporates cost and weight aspects into the objective function.
The framework is designed modularly in order to capture a variety of structures,
materials, processes and constraints.
The direct operating cost DOC (Equation (4.2)) was used as the basis and
simplified to the objective function seen below. Thus, only design-driving cost
aspects were considered to be part of the objective function. The optimization
problem was formulated as
subject to
DOC of an aircraft component
prescribed load case
xi < xi < xi , i = 1 . . . n,
with the direct operating cost given as the weighted sum
DOC = α1 Cman + α2 Cndt,prod + N α3 Cndt,serv + pW.
Cman is the manufacturing cost, Cndt,prod and Cndt,serv are non-destructive testing
costs for in-production and in-service inspection, p is a weight penalty (in e/kg)
and W is the weight of the structure. The parameters αi incorporate calibration
factors due to depreciation, overhead cost and other cost adjustments, and N is
the estimated number of regular inspections during the lifetime of the aircraft. The
final framework is illustrated in Figure 6.1.
In Curran’s and Kassapoglou’s work (see Section 3.3), closed-form solutions
provided the basis for the structural calculation. Here, it was proposed that an FE
tool (e.g. ABAQUS) would calculate the structural performance of the component.
Thus, the problem was independent from any limitations, such as geometries,
material models and boundary conditions. In addition, the setup was reduced to the
generation of the FE model and its parametrization. A major drawback, however,
was the computational effort that was necessary in order to generate the structural
The calculation of the structural constraints emerged to be the limiting factor
and a gradient-based method was chosen. The method of moving asymptotes (MMA)
was developed by Svanberg and first published in 1987, see [76–78]. This solver
M. Kaufmann
draping database
best ply
objective function
Figure 6.1: The proposed optimization framework
obtained the results from the different analysis blocks, i.e. SEER-MFG (for the
calculation of the manufacturing cost), ABAQUS/CAE, Composite Modeler and
the NDT model. Based on that feedback, the objective function, the constraints
and the update of the variables were computed.
The approach of a weight penalty p was introduced in the work done by Kelly
and Wang [37], Wang et al. [38] and Curran et al. [39]. The quantification of p,
however, is not trivial. The literature proposes values between e45/kg and e380/kg,
whereas own estimations, based on the fuel consumption of an A330 and today’s fuel
price, resulted in a weight penalty of approximately e2000/kg. A definite value for p
could not be given, as it depended on the viewpoint of the designer, the application
and the operational profile. Instead, it was concentrated on the effect of different
settings of the weight penalty p on the design. This was done as follows:
Paper A
A skin/stringer panel was optimized using a simplified version of the objective
function formed by the equation DOC = Cman + pW . The weight penalty p was
varied between 0 and ∞, thus capturing the whole spectrum between pure cost
optimization and pure weight optimization. The optimization was done for three
material configurations: an all-metal, a mixed and an all-composite configuration.
The optimal design solution was highly dependent on the weight penalty, and it
was shown that the ideal choice of the design solution was neither low-cost nor
low-weight but rather a combination thereof.
Paper B
A skin/stringer panel was optimized using the objective function given in Equation (6.2). Further, the design strength of each laminate was adjusted according
to the parameters of non-destructive testing. One of the parameters, the scan
pitch, was a representative value for the guaranteed laminate quality. It was shown
that – similar to the results of Paper A – the optimum laminate quality was again
dependent on the weight penalty. The designs of the investigated skin/stringer
panels were mainly governed by fulfilling the buckling constraint. As a consequence,
the design strength could be lowered by adjusting the scan pitch of the ultrasonic
testing, reducing the cost of NDT by 35-54% and the component’s direct operating
cost by 4-14%.
Paper C
The results of Paper A and Paper B showed that the actual cost could even be
lower than the estimations using prescribed process parameters. Thus, the suboptimization of machining and other process parameters was necessary in order to
estimate the lowest manufacturing cost in each iteration. A framework for the suboptimization of machining parameters was proposed, minimizing the manufacturing
cost in each iteration by the adaptation of manufacturing parameters. The framework
extension was added to the existing implementation and tested on the center wing
box rear spar of an airliner. Three optimizations were performed, and a low cost, a
low weight and an intermediate design solution were found. The difference between
the low cost and the low weight solutions was 4.4% in manufacturing cost and 9.7%
in weight. Based on these optimizations, the effect of the parameter adaptation
module was analyzed.
Paper D
The optimization framework was enhanced by a kinematic draping simulation which
allowed the fiber angles to be simulated more realistically. First, a draping knowledge
database was generated in which combinations of seed points and reference angles
were evaluated in terms of fiber angle deviation, scrap, ultrasonic cuts and material
shear. Second, the solver picked the best sets of plies during the subsequent
optimization. The methodology was tested by means of a curved C-spar which was
designed using plain weave and unidirectional prepreg. It was shown how different
objectives during the generation of the draping database led to different design
solutions. No non-destructive testing cost was included in this work.
Paper E
The total optimization framework as shown in Figure 6.1 was applied to the design
of a curved C-spar. The case study included five material systems: aircraft grade
aluminum, two non-crimp fabrics and two types of prepreg. The results were
M. Kaufmann
compared in relation to each other and it was shown that (depending on the estimated
fuel burn share of the component) a different material system was favorable when
optimizing for operating cost. In addition, several what-if scenarios were examined in
which changes in labor rates, material costs and production quantities were studied.
7 Conclusion
The proposed optimization framework is an approach to include more of the total
life-cycle of the aircraft into the design process than done today. Depending on the
viewpoint of the designer, cost and weight are traded differently. Thus, a simplified
form of the direct operating cost is used to evaluate the merit of a design solution. A
weight penalty is assigned to balance the cost and weight objectives. In combination
with the proposed NDT and draping tools, the literature’s cost/weight optimization
concepts are brought forward towards a more holistic view of cost-efficient design.
In each of the case studies, different design solutions were favorable in dependency
of the value of the weight penalty. Thus, the fiber angle distribution, the thicknesses
and the stringer configuration changed. It was shown that the engineer should not
perform cost or weight optimization alone, but rather a combination thereof.
The cost and the structural model generated good estimates and provided
sufficient sensitivity for the shape and size optimization. Under the condition that
the component was modeled with appropriate elements, a fast convergence rate was
achieved. A drawback of the current form of the framework, however, was the time
needed to parameterize the models. More work would be necessary to bridge the
gap from a CAD model to an applicable optimization setup.
8 Future Work
A lot of aspects in the design of aircraft structures were implemented in this work.
Nevertheless, there are four fields where future work is proposed.
First, more work is needed to capture the total life-cycle of an airliner already
in its design process. Little primary structure made of composite material is flying
today, as both the A350 and Boeing’s 787 are still in prototype stages. Therefore,
little experience of ageing carbon fiber wings and fuselages exists, and to foresee
the number and thoroughness of inspections is delicate. An enhanced maintenance
model, repairability and end-of-life aspects could be integrated into the objective
Second, it is suggested to enhance the framework by probabilistic design methods. The structural performance, for example, should be robust to manufacturing
tolerances, such as angle or thickness deviations, porosities or irregularities in the
material properties. Thus, the use of a reliability-based optimization framework
would enable the design of robust structures.
Third, the NDT cost and the NDT strength reduction models could be improved.
The difference between in-production and in-service testing could be more elaborate
by applying different scanning techniques and overhead adjustments. Another
suggestion is the application of the strength reduction as function of the stacking
sequence, material properties and manufacturing technique. In addition, a stiffness
reduction due to porosity might be included in the structural model. Beyond that,
the inspection interval could be adapted to the stress level and the structural function
of each feature. A probabilistic damage model could be included to capture the
possibility of failure and repair for each structural member.
Finally, one could work on developments for the optimization of the draping
strategy. In the current state, a restricted set of seed points is included for the
generation of the draping knowledge database. One could imagine enhancements of
the methodology where all points of the component’s surface are considered. The
use of a response surface method could provide that functionality.
G. N. Vanderplaats. Structural optimization for statics, dynamics and beyond. Journal
of the Brazilian Society of Mechanical Sciences and Engineering, 28(3):316–322, 2006.
T. Gutowski, R. Henderson, and C. Shipp. Manufacturing costs for advanced composites aerospace parts. SAMPE Journal, 27(3):37–43, 1991.
M. Rais-Rohani and E. B. Dean. Toward manufacturing and cost considerations in multidisciplinary aircraft design. In Collection of Technical Papers AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference, volume 4, pages 2602–2612, 1996.
J. Sobieszczanski-Sobieski and R. T. Haftka. Multidisciplinary aerospace design
optimization: survey of recent developments. Structural and Multidisciplinary Optimization, 14:1–23, 1997.
C. Soutis. Fibre reinforced composites in aircraft construction. Progress in Aerospace
Sciences, 41(2):143–151, 2005.
H. Assler. Design of aircraft structures under special consideration of NDT. In Jörg
Völker, editor, Proceedings of the 9th European NDT Conference (ECNDT), Berlin,
K. Johansen, M. Comstock, and M. Winroth. Coordination in collaborative manufacturing mega-networks: A case study. Journal of Engineering and Technology
Management - JET-M, 22(3):226–244, 2005.
M. Jacobsen. On improving efficiency of flight using optimization. PhD thesis,
Kungliga Tekniska Högskolan (KTH), Trita-AVE 2009:45, 2009.
G. Marsh. Composites lift off in primary aerostructures. Reinforced Plastics, 48(24):22–27, 2004.
[10] M. C. Niu. Airframe Structural Design. Hong Kong Conmilit Press Ltd, Hong Kong,
second edition edition, 1996.
[11] NDT Resource Center. Data presentation. retrieved October 24, 2009, from EquipmentTrans/DataPres.htm, (created on December 19, 2007).
[12] D. Raymer. Enhancing Aircraft Conceptual Design using Multidisciplinary Optimization. PhD thesis, KTH, Department of Aeronautics, 2002. 91-7283-259-2.
M. Kaufmann
[13] P. Bartholomew. The role of MDO within aerospace design and progress towards an
MDO capability. In Collection of Technical Papers - AIAA/USAF/NASA/ISSMO
Symposium on Multidisciplinary Analysis and Optimization, AIAA-1998-4705, pages
10–31, 1998.
[14] J. A. Samareh. A survey of shape parameterization techniques. In CEAS/AIAA/
ICASE/NASA Langley International Forum on Aeroelasticity and Structural Dynamics,
Williamsburg, VA, June 1999.
[15] J. A. Samareh and K. G. Bhatia. A unified approach to modeling multidisciplinary
interactions. In Proceedings of the 8th AIAA/USAF/NASA/ISSMO Symposium on
Multidisciplinary Analysis and Optimization, AIAA-2000-4074, 2000.
[16] J. C. Townsend, J. A. Samareh, R. P. Weston, and W. E. Zorumski. Integration of a
CAD system into an MDO framework. Technical report, NASA Langley Technical
Report Server, 1998.
[17] A. Salas and J. C. Townsend. Framework requirements for MDO application development. In Collection of Technical Papers - AIAA/USAF/NASA/ISSMO Symposium
on Multidisciplinary Analysis and Optimization, 1998, AIAA-1998-4740, 1998.
[18] R. T. Marler and J. S. Arora. Survey of multi-objective optimization methods for
engineering. Structural and Multidisciplinary Optimization, 26(6):369–395, 2004.
[19] K. Gantois and A. J. Morris. The multi-disciplinary design of a large-scale civil
aircraft wing taking account of manufacturing costs. Structural and Multidisciplinary
Optimization, 28:31–46, 2004.
[20] L. T. Watson and R. T. Haftka. Modern homotopy methods in optimization. Computer
Methods in Applied Mechanics and Engineering, 74(3):289–305, 1989.
[21] I. Das and J. E. Dennis. Normal-boundary intersection: a new method for generating
Pareto optimal points in multicriteria optimization problems. SIAM Journal on
Optimization, 8(3):631–657, 1998.
[22] C.-H. Huang, J. Galuski, and C. L. Bloebaum. Multi-objective pareto concurrent
subspace optimization for multidisciplinary design. AIAA Journal, 45(8):1894–1906,
[23] I. Das and J. E. Dennis. Closer look at drawbacks of minimizing weighted sums of
objectives for pareto set generation in multicriteria optimization problems. Structural
Optimization, 14(1):63–69, 1997.
[24] J.-H. Kang and C.-G. Kim. Minimum-weight design of compressively loaded composite
plates and stiffened panels for postbuckling strength by genetic algorithm. Proceedings
of the 14th International Conference on Composite Materials, 2003.
[25] M. Walker. The effect of stiffeners on the optimal ply orientation and buckling load of
rectangular laminated plates. Computers and Structures, 80(27-30):2229–2239, 2002.
[26] T. S. Geiger and D. M. Dilts. Automated design-to-cost: Integrating costing into the
design decision. Computer Aided Design, 28(6-7):423–438, 1996.
[27] B. Heinmuller and D. M. Dilts. Automated design-to-cost: Application in the aerospace industry. In Proceedings - Annual Meeting of the Decision Sciences Institute,
volume 3, pages 1227–1229, 1997.
[28] W. J. Marx, D. N. Mavris, and D. P. Schrage. A knowledge-based system integrated
with numerical analysis tools for aircraft life-cycle design. Artificial Intelligence for
Engineering Design, Analysis and Manufacturing, 12:211–229, 1998.
[29] J. Scanlan, A. Rao, C. Bru, P. Hale, and R. Marsh. Datum project: Cost estimating
environment for support of aerospace design decision making. Journal of Aircraft,
43(4):1022–1028, 2006.
[30] C. H. Park, W. I. Lee, and A. Vautrin. Simultaneous optimization of composite
structures considering mechanical performance and manufacturing cost. Composite
Structures, 65(1):117–127, 2004.
[31] M. S. Edke and K. H. Chang. Shape optimization of heavy load carrying components
for structural performance and manufacturing cost. Structural and Multidisciplinary
Optimization, 31(5):344–354, 2006.
[32] R. Curran, G. Gomis, S. Castagne, J. Butterfield, T. Edgar, C. Higgins, and C. McKeever. Integrated digital design for manufacture for reduced life cycle cost. International
Journal of Production Economics, 109(1-2):27–40, 2007.
[33] Engineering Sciences Data Unit. Local buckling of compression panels with unflanged
integral stiffeners. ESDU International, Item 70003, 1970.
[34] C. Kassapoglou. Simultaneous cost and weight minimization of composite-stiffened
panels under compression and shear. Composites Part A: Applied Science and Manufacturing, 28(5):419–435, 1997.
[35] C. Kassapoglou. Minimum cost and weight design of fuselage frames. part a: Design
constraints and manufacturing process characteristics. Composites Part A: Applied
Science and Manufacturing, 30(7):887–894, 1999.
[36] C. Kassapoglou. Minimum cost and weight design of fuselage frames. part b: Cost
considerations, optimization, and results. Composites Part A: Applied Science and
Manufacturing, 30(7):895–904, 1999.
[37] D. Kelly, K. Wang, and S. Dutton. A guided tradeoff for cost and weight
for generating optimal conceptual designs. In Collection of Technical Papers AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference, volume 2, pages 894–904, 2003.
[38] K. Wang, D. Kelly, and S. Dutton. Multi-objective optimisation of composite aerospace
structures. Composite Structures, 57(1):141–148, 2002.
[39] R. Curran, A. Rothwell, and S. Castagne.
A numerical method for costweight optimisation of stringer-skin panels. In Collection of Technical Papers AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference, volume 7, pages 5262–5277, 2004.
M. Kaufmann
[40] M. Price, R. Curran, A. Murphy, S. Raghunathan, J. Early, S. Castagne, E. Benard,
and B. Crawford. Integrating design, manufacturing and cost for trade-offs on aircraft
configurations. In Collection of Technical Papers - 6th AIAA Aviation Technology,
Integration, and Operations Conference, volume 1, pages 373–383, 2006.
[41] S. Castagne, R. Curran, M. Price, P. Collopy, A. Murphy, and S. Raghunathan. Value
driven design optimization of aircraft structures. In Collection of Technical Papers 6th AIAA Aviation Technology, Integration, and Operations Conference, volume 1,
pages 604–611, 2006.
[42] R. Curran, S. Castagne, J. Early, M. Price, S. Raghunathan, J. Butterfield, and
A. Gibson. Aircraft cost modelling using the genetic causal technique within a systems
engineering approach. Aeronautical Journal, 111(1121):409–420, 2007.
[43] S. Castagne, R. Curran, A. Rothwell, M. Price, E. Benard, and S. Raghunathan. A
generic tool for cost estimating in aircraft design. Research in Engineering Design,
18(4):149–162, 2008.
[44] S. Castagne, R. Curran, and P. Collopy. Implementation of value-driven optimisation for the design of aircraft fuselage panels. International Journal of Production
Economics, 117(2):381–388, 2009.
[45] A. Iqbal and J. S. Hansen. Cost-based, integrated design optimization. Structural and
Multidisciplinary Optimization, 32(6):447–461, 2006.
[46] J. Roskam. Airplane Design Part VIII: Airplane Cost Estimation Design Development
and Manufacturing and Operating. Darcorporation, 1990.
[47] Y. Asiedu and P. Gu. Product life cycle cost analysis: state of the art review.
International Journal of Production Research, 36(4):883–908, 1998.
[48] S. K. Durairaj, S. K. Ong, A. Y. C. Nee, and R. B. H. Tan. Evaluation of life cycle
cost analysis methodologies. Corporate Environmental Strategy, 9(1):30–39, 2002.
[49] D. G. Woodward. Life cycle costing - theory, information acquisition and application.
International Journal of Project Management, 15(6):335–344, 1997.
[50] J. V. Michaels and W. P. Wood. Design to cost. John Wiley and Sons Inc., 1989.
[51] E. B. Dean and R. Unal. Elements of designing for cost. In Proceedings of the AIAA
1992 Aerospace Design Conference, Irvine, CA, February 1992.
[52] R. Curran, S. Raghunathan, and M. Price. Review of aerospace engineering cost
modelling: The genetic causal approach. Progress in Aerospace Sciences, 40(8):487–534,
[53] D. M. Anderson. Design for Manufacturability. CIM Press, 1990.
[54] P. Dewhurst and G. Boothroyd. Design for assembly: automatic assembly. Machine
Design, 56(2):87–92, 1984.
[55] P. Dewhurst and G. Boothroyd. Design for assembly: robots. Machine Design,
56(4):72–76, 1984.
[56] G. Boothroyd. Product design for manufacture and assembly. Computer Aided Design,
26(7):505–520, 1994.
[57] S. Miyakawa and T. Ohashi. The hitachi new assemblability evaluation method
(AEM). In Proceedings of the Ontario Conference on Product Design for Assembly,
Newport, RI, USA, 1986.
[58] J. Wilson et al. Integrated Product and Process Development. The National Center
for Advanced Technologies, 1994.
[59] A. Niazi, J. S. Dai, S. Balabani, and L. Seneviratne. Product cost estimation:
Technique classification and methodology review. Journal of Manufacturing Science
and Engineering, Transactions of the ASME, 128(2):563–575, 2006.
[60] Y. Wei and P. J. Egbelu. A framework for estimating manufacturing cost from
geometric design data. International Journal of Computer Integrated Manufacturing,
13(1):50–63, 2000.
[61] A. C. Long. Composites forming technologies. Woodhead Publishing Ltd, 2007.
[62] C. Mack and H. M. Taylor. The fitting of woven cloth to surfaces. Journal of the
Textile Institute, (47):477–488, 1956.
[63] B. P. Van West, R. B. Pipes, and M. Keefe. A simulation of the draping of bidirectional
fabrics over arbitrary surfaces. Journal of the Textile Institute, 81(4):448–460, 1990.
[64] K. Potter. Bias extension measurements on cross-plied unidirectional prepreg. Composites Part A: Applied Science and Manufacturing, 33(1):63–73, 2002.
[65] K. Potter. Beyond the pin-jointed net: Maximising the deformability of aligned continuous fibre reinforcements. Composites Part A: Applied Science and Manufacturing,
33(5):677–686, 2002.
[66] K. Potter. In-plane and out-of-plane deformation properties of unidirectional preimpregnated reinforcement. Composites Part A: Applied Science and Manufacturing,
33A(11):1469–77, 2002.
[67] J. Wang, R. Paton, and J. R. Page. Draping of woven fabric preforms and prepregs for
production of polymer composite components. Composites Part A: Applied Science
and Manufacturing, 30(6):757–765, 1999.
[68] J. Wiggers. Analysis of Textile Deformation during Preforming for Liquid Composite
Moulding. PhD thesis, University of Nottingham, 2007.
[69] M. Nguyen, I. Herszberg, and R. Paton. The shear properties of woven carbon fabric.
Composite Structures, 47(1):767–779, 1999.
[70] P. Boisse, B. Zouari, and J.-L. Daniel. Importance of in-plane shear rigidity in finite
element analyses of woven fabric composite preforming. Composites Part A: Applied
Science and Manufacturing, 37(12):2201–2212, 2006.
[71] P. Badel, E. Vidal-Salle, and P. Boisse. Computational determination of in-plane shear
mechanical behaviour of textile composite reinforcements. Computational Materials
Science, 40(4):439–448, 2007.
[72] Y. Larberg. Deformability of unidirectional prepreg materials. Licentiate Thesis,
Kungliga Tekniska Högskolan (KTH), Trita-AVE 2009-32, 2009.
M. Kaufmann
[73] O. Guillermin. Computer-aided design and manufacturing. ASM Handbook Volume
21, Composites, pages 366–372, 2001.
[74] A. A. Skordos, C. Monroy Aceves, and M. P. F. Sutcliffe. Drape optimisation in
woven composite manufacturing. In Proceedings of the 5th International Conference
on Inverse Problems in Engineering: Theory and Practice, 2005.
[75] A. Lang. Numerical simulation of draping and cutting of CFRP prepregs. Bachelor
thesis 08-026, Eidgenössische Technische Hochschule Zürich (ETHZ) and Kungliga
Tekniska Högskolan (KTH), 2008.
[76] K. Svanberg. Method of moving asymptotes – a new method for structural optimization.
Journal for Numerical Methods in Engineering, 24(2):359–373, 1987.
[77] K. Svanberg. A globally convergent version of MMA without linesearch. In Proceedings
of the First World Congress of Structural and Multidisciplinary Optimization, Goslar
(Germany), pages 9–16, May/June 1995.
[78] M. Bruyneel, P. Duysinx, and C. Fleury. A family of MMA approximations for
strutural optimization. Structural and Multidisciplinary Optimization, 2002.
Division of work between authors
Paper A
Kaufmann was responsible for the implementation and the numerical experiments.
The analysis of the results was performed jointly by the authors. The paper was
written by Kaufmann with support from Zenkert.
Paper B
Mattei proposed the cost model. The implementation, the experiments and the
analysis of the results were performed by Kaufmann. The paper was written by
Kaufmann with support from Zenkert.
Paper C
Czumanski implemented the framework under Kaufmann’s supervision. Kaufmann
and Czumanski jointly carried out the case study. The paper was written by
Kaufmann with support from Czumanski and Zenkert.
Paper D
Kaufmann carried out the implementation of the framework, the numerical experiments and the analysis of the results. The paper was written by Kaufmann with
support from Zenkert and Åkermo.
Paper E
Kaufmann carried out the numerical experiments and the analysis of the results.
The paper was written by Kaufmann with support from Zenkert and Åkermo.
Part II
Appended papers
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF