Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018
SAE TECHNICAL
PAPER SERIES 2003-01-2950
Measurement-Assisted Assembly Applications
on Airbus Final Assembly Lines
Benoit Marguet
EADS CCR
Bernard Ribere
AIRBUS
SAE Aerospace Automated Fastening Conference and Exhibition
2003 Aerospace Congress and Exhibition
Montreal, Canada
September 8-11, 2003
400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760 Web: www.sae.org
Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise,
without the prior written permission of SAE.
For permission and licensing requests contact:
SAE Permissions
400 Commonwealth Drive
Warrendale, PA 15096-0001-USA
Email: permissions@sae.org
Fax: 724-772-4891
Tel: 724-772-4028
For multiple print copies contact:
SAE Customer Service
Tel: 877-606-7323 (inside USA and Canada)
Tel: 724-776-4970 (outside USA)
Fax: 724-776-1615
Email: CustomerService@sae.org
ISSN 0148-7191
Copyright © 2003 Airbus France SAS. Reprinted by SAE International with permission.
Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE.
The author is solely responsible for the content of the paper. A process is available by which discussions
will be printed with the paper if it is published in SAE Transactions.
Persons wishing to submit papers to be considered for presentation or publication by SAE should send the
manuscript or a 300 word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.
Printed in USA
Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018
2003-01-2950
Measurement-Assisted Assembly Applications on
Airbus Final Assembly Lines
Benoit Marguet
EADS CCR
Bernard Ribere
AIRBUS
Copyright © 2003 Airbus France SAS. Reprinted by SAE International with permission.
ABSTRACT with theoretical values for the dimensions and the
geometry of the components to assemble together. If the
This paper describes the 'Measurement-Assisted components differ too widely from their theoretical
Assembly' activities led by the Final Assembly Line unit specifications, the fixed tool cannot assemble them
at Airbus - Toulouse. These activities are meant to without reworking the faulty parts or raising concessions.
eliminate some of the problems associated with the
conventional process of locating and positioning large
airframe sub-assemblies during the final assembly
process, like fuselage-to-fuselage or wing to fuselage
junctions. All these activities include laser or photogrammetry
subsystems, computer-aided measuring systems, Best- Hard Tooling Center Section
Fit optimization software, specific Graphical User
Interface. The combination of these technologies offers:
jigless assembly, faster assembly process, rework and
waste reduction and many more advantages.
Rear section
BACKGROUND
Fuselage-to-
Airbus have been designing and producing civil aircrafts fuselage junction
for over 25 years. Although during this time there have
been numerous improvements in aircraft design
methodologies and tools (i.e CAD deployment or
concurrent engineering method), the assembly process Fig. 1: Assembly process
on the Final Assembly Lines has largely remained
unchanged since the first A300. This situation is however MEASUREMENT-ASSISTED ASSEMBLY
not specific to Airbus but is shared with our competitors
[1]. Based on the above description, aircraft manufacturers
have recently been looking for a way to improve the
The 'conventional' assembly process uses large fixed airframe assembly process. One way to improve this
tools that ensure that each component (fuselage, wings, process is to replace the fixed tooling and to use the
etc.) is correctly positioned on the assembly stations (cf. Measurement-Assisted Assembly method (MAA). The
fig.1). Thus, the assembly junction quality is based on MAA method differs from the conventional assembly
the tooling quality. In order to ensure the quality, the tools process in two regards:
must be certified periodically to prove that they always
meet the specified tolerances for the assembly. This · Firstly, it relies on a measuring system (i.e. laser
periodic certification is time- and cost-consuming. tracker or radar) in order to locate the components to
Moreover, these fixed tools are generally designed and assemble in the station. The measuring system is
built for the purpose of assembling a single or a small set also used to identify the exact 3D geometry of all
of similar products. To reconfigure the assembly line for surfaces participating in the assembly process.
another aircraft family is therefore generally impossible · Secondly, an optimization software ('Best Fit'
or very difficult to achieve. Lastly, these tools are software) is used to compute the optimum position of
developed based on the nominal aircraft specifications, the components, taking into account their actual
Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018
geometry, the allowed dimensional tolerance and the · What geometric entities (surfaces, lines, points) are
theoretical shape. Once the Best Fit application has needed to optimize the component's position? How
found the best position for the components, a can they be identified?
comparison with the CAD model can be made to · What is the input data available (CAD model, GD&T
define the points outside the tolerance (see fig. 2). values, assembly process)?
Optimization software
Based on this data, the second task for the MAA
application is to define how to optimize the position of
each component relative to the others. Behind this
question, we find:
· The criteria to optimize (distance between specific
points, compliance with shape tolerance, etc.),
· The type of optimization algorithm (Least Squares
Method, Minimum Sum of Deviation, Tolerance
envelope, etc.).
· The constraints associated with jig displacement (i.e:
locked translation and/or rotation).
System architecture
Finally, a complete system architecture has to be
Fig. 2: Best Fit application defined, taking into account:
· The Graphical User Interface
Benefits
· The standard for data exchange (i.e: CAD model,
point coordinates, etc).
It is commonly assumed that, compared to a
'conventional' assembly process [2], [3], using the MAA · The type and the format of the database.
method produces the following benefits: Once all these questions have been analyzed and solved,
the MAA application can be made. This application
· Drastically reduce the cost of tooling and jigs.
generally runs as shown in the flowcharts below:
· Produce a more repeatable assembly junction.
· Improve junction tolerance achievement and thus
overall product quality.
· Allow a large range of station flexibility.
· Reduce the amount of man-hours required for the
component marry-up process. This reduction
includes time for tool setup, component positioning
and savings in move iterations to locate the
component.
All these benefits lead to a positive and very quick R.O.I
despite the cost of the hardware and software needed for
the MAA application. This is particularly true in the case
of new product development with new assembly lines, as
for the A380.
MAA PROCESS
Although the different suppliers have now reached a Fig.3: MAA flowchart
suitable maturity in measuring systems and software, the
industrial deployment of the MAA application into an SOME AIRBUS APPLICATIONS IN FAL
existing or a new assembly line is still a difficult task to
perform. Based on the Airbus work-sharing system, only two
A number of different issues have to be carefully major assembly junctions are performed on Airbus Final
analyzed and decided in order to succeed from the start. Assembly Lines: fuselage-to-fuselage junction and wing-
Measuring system to-fuselage junction. For both of these junctions, Airbus
has studied and developed a specific Measurement-
The first task to perform concerns the measuring system Assisted Assembly application.
itself. This task deals with the following questions:
· What are the Datum Reference Frames (global and Fuselage-to-fuselage junction
local) for the components and for the station? How
can they be identified? The MAA application for fuselage-to-fuselage junction is
already developed in Airbus FAL. The main challenge of
Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018
this deployment is to perform MAA without any major optimum position is computed by means of two specific
modification of the assembly lines. algorithms. The first one transforms the position of all
points from their local coordinate frame (transportation
In this application, the fuselage sections are measured
cart coordinate frame) to the station coordinate frame.
before being loaded onto the station jigs. This
This transformation is not only a displacement matrix
identification is made with the sections on their
from one coordinate frame to another, as it takes into
transportation trolley, using a laser tracker (see fig. 4). A
account the structural distortion of the section when it is
set of data is measured on the fuselage circumference,
loaded onto the station. The second algorithm optimizes
as well as a set of data on the seat rails and the four
the position of one section relative to the other, taking
locating points at the handling points. A fixed set of
into account the assembly rules used by the operators.
points on the shop floor is also measured in order to be
able to relocate the laser station when it changes its
Once the software has defined the optimum position of
position.
the fuselage sections, a complete report is edited. This
report gives all necessary information for the operators to
set up the jig position in order to achieve the optimum
junction between the two fuselage sections (fig. 6).
Fig. 6: position to achieve
In addition, a chart gives the deviation of the points from
their theoretical values (see. fig. 7). This report allows full
traceability of the junction process.
Fig. 4: MAA on fuselage-to-fuselage
The measuring process is automatically driven by a
dedicated software script. The Graphical User Interface
has been specifically designed for Airbus purposes, for
use by any operator on the shop-floor (see fig. 5).
Fig. 7: point deviations
Wing-to-fuselage junction
After the success of the MAA application on the
fuselage-to-fuselage junction, another MAA application
has been studied on the wing-to-fuselage junction. The
goal of this application is similar to the previous one. A
Fig. 5: Measuring system GUI measuring system allows us to identify the actual
geometry of the junction surfaces (upper, lower, right and
Once the measuring process is complete, the software left sides) for the center section and the wing part. The
computes the optimum position of one fuselage section dataset is then computed by a 'best fit' software in order
relative to the other and translates the best fuselage to define the optimum position of the wing relative to the
position into small displacements of the jig heads. The center section. Finally, this optimum position of the wing
Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Sunday, August 12, 2018
is given to a control system, which drives the positioner These two activities are now complete, allowing Airbus to
jigs to the right location (see fig. 8). deploy an industrial MAA application for any wing-to-
fuselage junction.
Rt
Datum points
Datum points
Measu for the section
for the wing
for se red surface
ction
junct s
Mea ion
su
for w red su
ing rf
junc aces
tion
(1)
Rv Measuring device
(1)
Measuring
software
Wing positioners
(2)
(3)
Best Fit
Positionner controlers
Software
Fig.9: Best Fit software for wing-to-fuselage junction
CONCLUSION
Measurement-Assisted Assembly is today the standard
Fig. 8 : MAA application for wing-to-fuselage junction for the final assembly process of Airbus aircraft. This
application helps an operator to achieve the best quality
However, compared to the fuselage-to-fuselage junction, on aircraft junctions. It is also a way to drastically cut
this MAA application differ in three regards: back assembly time by reducing the time for tool setup
and the time to locate the components. This process has
· The shapes of the junction surfaces to identify are been studied and deployed successfully on the Final
much more complex to measure. The surface to Assembly Lines of Airbus for the fuselage-to-fuselage
identify is a 3D free form, the dimensions are wider and wing-to-fuselage junctions.
and the tolerances to achieve are very tight.
· The geometric and dimensional tolerances to comply ACKNOWLEDGMENTS
with on the junction are specific to the wing-to-
fuselage junction. The tolerance envelope is a Leica Geosystems
diamond taking into account both the translation and
rotation deviation of the junction. This specific
tolerance envelope is not applied by commercial REFERENCES
optimization software.
· The optimization process has to take into account 3 1. Muske S., Salisbury D., Salerno R., Calkins J., 1999,
items at the same time (the center section and the 747 Data Management System Development and
two wings). Commercial optimization software does Implementation, CMSC Conference.
not perform optimization with more than 2 elements. 2. Rüscher O., Mayländer H., 2001, Automated Alignment
and Marry–up of Aircraft Fuselage Sections with a Final
To answer the first point, a study of new measuring Assembly Line, SAE paper No. 2001-01-2570.
systems was made (see fig. 9). The study included a 3. Williams G., Chalupa E., Rahhal S., 2000, Automated
new laser radar and tracker system and new portable
Positioning and Alignment Systems, SAE paper
coordinate-measuring machines. This study identified the
No. 2000-01-3014.
best system to achieve a fast and accurate measuring
process of large distances and high tolerances. CONTACT
At the same time, a specific software program was Benoît Marguet, Ph.D., benoit.marguet@eads.net
developed in order to compute a best-fit algorithm Research engineer, EADS CCR
capable of finding the optimum position within a
diamond-shaped tolerance envelope and of dealing with Bernard Ribere, bernard.ribere@airbus.com
N elements (N>2). Head of Manufacturing Research and Development,
Manufacturing Unit FAL Long Range, Airbus
