51st AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference<BR>18th AIAA 2010-3138

12 - 15 April 2010, Orlando, Florida

Model Based Definition

Clark Briggs1
ATA Engineering, San Diego, California, 92130

Gerald Brown2
The Boeing Company, Huntington Beach, California, 92647

David Siebenaler3
The Boeing Company, Seattle, Washington, 98108

James Faoro4
Raytheon Missile Systems, Tucson, Arizona, 85747

and

Sidney Rowe5
NASA Marshall Space Flight Center, Huntsville, Alabama, 35812

Model based definition of products is supported by increasingly capable commercial

software, but it has not seen wide adoption in the aerospace and defense industry. This paper
reviews the role of model based definition in the product development lifecycle, examines
currently available software, illustrates aspects of recent deployments, and sketches an
approach to formulating the business case for adoption. This paper has been developed by
the Design Engineering Technical Committee of the American Institute of Aeronautics and
Astronautics in its role of providing technology transfer to the industry.

Nomenclature
CAD = Computer-aided design
CAM = Computer-aided manufacturing
CMM = Coordinate measuring machine
CNC = Computer numerical controlled
COTS = Commercial off-the-shelf
DDMS = Design data management system
DSFT = Design system focus team
GD&T = Geometric dimensioning and tolerancing
MBD = Model based definition
MSFC = Marshall Space Flight Center
PDD = Product definition data
PDF = Portable document format
PDM = Product Data Management
PLM = Product lifecycle management
PTC = Parametric Technology Corporation
TDP = Technical data package

1
Vice President, Business Development, 11995 El Camino Real, Suite 200, and Associate Member.
2
Senior Engineer, Advanced Technology Concepts, 5301 Bolsa Av, H45N-E405, and Senior Member.
3
Senior Engineer, Virtual Eng. Advanced Technologies, 9725 East Marginal Way South, MC42-54.
4
Principal Mechanical Engineer, Analysis Dept, TU/M30/S21 9000 S. Rita Rd, and Senior Member.
5
Engineering Manager, ED32, and Member.
1
American Institute of Aeronautics and Astronautics

Copyright © 2010 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.
 I. Background
HE American Institute of Aeronautics and Astronautics4 (AIAA) Technical Committees bring together experts
T in their fields and give them the opportunity to exchange knowledge in service to their profession. The Design
Engineering Technical Committee (DETC) promotes the development and dissemination of design technologies to
assist the design engineer in creating and defining practical aerospace products.
Design engineering is a discipline that creates and transforms ideas and concepts into a product definition that
satisfies customer requirements. The role of the design engineer is to create, synthesize, iterate, and present design
solutions. The design engineer coordinates with engineering specialists and integrates their input to produce the
form, fit, and function documentation to completely define the product.
The Design Process subcommittee charter includes promoting the improvement of the entire design engineering
process, and the mission includes pursuing design integration as a systems-oriented approach to improving the
design engineering process. The members of the subcommittee prepared this white paper to support the understand-
ing and adoption of model based definition, which will lead to the eventual improvement of business performance of
our profession’s companies.

II. Purpose and Scope

Across the aerospace and defense industry, model based definition (MBD) is not a well understood term. It can
mean different things to different people in different contexts. The DETC Design Process Subcommittee intends to
clarify the meaning and provide examples both from vendors supplying software tools and from companies using
these tools.
This paper addresses the meaning of Model Based Definition, but simply stated, it will be the encapsulation of
the product-defining data in computer models. The most visible application in the industry is the move away from
basing manufacturing on two-dimensional (2-D) drawings.
Product definition data, at a minimum, will be that data required to manufacture the product. In a broader mean-
ing, it may also encompass supporting data such as reports and analyses.

III. The Move Away From Drawings

Two-dimensional drawings are the most widely used format for capturing and transmitting the design specifica-
tions for a product. Over the past fifty years, design engineering has moved from ink on vellum produced on the
drawing board, to electronic drafting using computer software, to three-dimensional (3-D) solid modeling. During
this period, the deliverable was the drawing. Even though recent digital advances allowed the drawing image to be
captured, routed, and stored in a computer, the design content could only be understood by visual inspection, study,
and thought. Manufacture, assembly, and test of the product require many subsequent operations that utilize the
product data in several different ways. Computer automation of the design to manufacture to operation flow requires
machine readable design data. The result has been to not make the drawings themselves machine readable but to
encapsulate the data in models and to use new software tools to access the design data in the models.
Drawings contain several broad categories of product definition data. The majority of the drawing area is typi-
cally 2-D views of the product shape with dimensions to define sizes and positional relations. The views typically
include the canonical front, top, and side views, but might also include a reference isometric view and multiple
detail views. The human reader who studies these views can form a mental understanding of the three-dimensional
shape, the part’s external interactions with neighboring components, and its internal details. Computer software that
replicates this 2-D to 3-D reconstruction has not been successful because of the tremendous complexity of the
human mental processes. On the contrary, computer-aided design software has changed the primary basis from 2-D
views to the 3-D solid model. The drawing is a derived product of the solid model in that particular legacy layout
recognizable from the days of the drawing board.
The parts list is another important element of the drawing. This details the components used in the product, the
quantity required, and other specifications. Interim electrification steps have resulted in managing the parts list in a
computer system independent of the drawing management system. These are sometimes called separate parts lists
and are essential to manufacturing activities such as configuration control, ordering, and inventory management.
Geometric dimensioning and tolerancing (GD&T) is defining data that quantifies the variability, accuracy, and
relationships of features and basic dimensions of the product. The complex symbology of GD&T augments a dimen-
sion and includes definition of datum planes for references. Nondimensional areas of the drawing, such as the title
block and notes, can also contain tolerance information.

2
American Institute of Aeronautics and Astronautics
 The body of drawings includes a text area for notes that are typically numbered and called out by reference to the
number at multiple locations on the drawing. The notes list can call out required processes to be invoked, clarifica-
tion of relationships with other parts, and manufacturing sequence instructions.
The drawing contains a title block and other text tables that define the history of the part, the approvals, the
identifying part number and title, and generally applicable default data such as the material and tolerances.
All of these types of data must be carried in the model format.
The desired end state that does not depend on the 2-D drawing goes by many names, but the most common is
"drawingless." Because of the many manufacturing processes that are dependent upon the drawing in its many
meanings, there is a series of intermediate or precursor states.
The prominence of computer numerical controlled (CNC) machining and computer-aided manufacturing (CAM)
methods for part fabrication leads many organizations to an early transformational state where the drawing is the
definition of record but the 3-D CAD model is the digital source used for CNC CAM. This automates the transfer of
the nominal product shape via the 3-D model and eliminates the manual reconstruction of the part geometry in the
CAM software. In this interim state, the dimensional inspection activity of quality assurance can also be based on
the 3-D CAD model. Coordinate measuring machine (CMM) programs can be based on the model, thereby avoiding
a similar reconstruction of the part geometry. Both inspection and part fabrication still depend on the drawing for
additional information such as tolerances and materials.
Limited dimension drawings are 2-D drawings that are supplemented with models. The drawing is in a digital
format such as portable document format (PDF) and displays a reduced set of dimensions. The product cannot be
fabricated from the reduced set, and the accompanying model must be used for fabrication. This interim state often
automates the transfer of design definition only for part CNC fabrication. The viewable drawing will still be the
source for notes, the parts list, tolerances, approvals, title, and history. The dimensions raised on the drawing can
serve various purposes depending upon the organization's practices.
A pro forma drawing is nothing more than a simple place holder in a legacy drawing processing system that
contains a simple four-view layout, identifying information such as part number and title, and, perhaps, abbreviated
approvals. This state supports a late stage process transformation where nearly all manufacturing steps depend on
the model-borne definition data, but certain legacy control systems still require a drawing image of record. This
might occur when a manufacturer had automated much of the migration to drawingless but the customer still
requires a drawing deliverable.

IV. Application Across the Product Lifecycle

The long-scale temporal aspects of product development, utilization, and retirement are described in terms of the
lifecycle. Lifecycles begin when a need is recognized and a development activity is approved. The development
period is often several stages or phases of definition, design, and testing. A fabrication period covers production; an
operational period covers the usable life of the product; and retirement covers the product’s phase-out and disposal.
These vary by product and industry.
For parts of the aerospace industry, there are usually three design phases, which are conceptual design, prelimi-
nary design, and detail design. In other areas, there may be a concept development phase followed by an engineering
and manufacturing development phase. In industries where products are fielded and upgraded in a more incremental
fashion, there may be a single design phase followed by a field implementation and deployment phase.
The immediate applicability of MBD is in the hand-off from completion of design to the initiation of manufac-
ture. The MBD is, in fact, the encapsulation of the design information that is the input to fabrication. This is
reflected in the motivation to move from drawings to models as the product definition for fabrication.
Some consideration has been given to the role of MBD in earlier lifecycle phases. One example that is analogous
to manufacturing is rapid prototyping. It is common during projects to fabricate scale models, or even full-scale
models if the product size permits, from rapid prototyping fabrication processes. These take as inputs the same or
very similar digital electronic design data as production manufacturing processes.
Clearly, the details of the designs are not available in early lifecycle stages. Nonetheless, there is value in both
the concept and implementation of MBD in these early times. If the project can actually identify the product MBD
artifacts and maintain their continuity over time, there is considerable potential value that can be harvested in a ret-
rospective analysis. Other uses might include using the MBD as the state of the design at any given time, using the
MBD as the basis for building contract data packages for further design activities, and using the MBD as the data set
for other disciplines such as analysis.
By the time of release to manufacturing, the MBD dataset will contain solid nominal geometry, the part’s coor-
dinate systems, dimensions, tolerances, annotations, engineering notes, general notes, and material specifications.

3
American Institute of Aeronautics and Astronautics
Earlier in the project, much of this information is not available according to current practices. For example, the solid
models will commonly lack details and fastening solutions, the material specification may only consist of the gen-
eral family or series, and dimensions will lack tolerances. With better tools, some these elements that are more
important might be defined earlier in the process. For example, high precision assemblies might incorporate GD&T
designs early to facilitate performance modeling or cost estimation.
MBD has the capability to expedite final part inspection and verification, by utilizing the solid model data as
input criteria for the CMM used for measuring finished part dimensions. Automated verification of other data such
as surface finish material, plating quality, and surface roughness would be desirable as well. However, there are sev-
eral complicating factors to contend with such as the compound tolerances inherent in certain features and difficulty
of measuring specific part details. When MBD is not fully implemented in the above fashion, it becomes a major
challenge to convert the CMM “cloud of points” output into surfaces and other specific features that can be com-
pared to the model in a meaningful fashion.
The application of MBD is closely linked to the product lifecycle. For example, during the operational period of
the product, the complete system MBD could be used for the training of manufacturing and maintenance personnel
to familiarize them with the assembly and disassembly of the product. At the end of the product’s life, the informa-
tion included in the MBD could help the disposal engineer in deciding which components can be reused.

V. Current Embedded GD&T Examples

While embedded GD&T is a small part of MBD, it represents much of the current market vendors’ development
progress. The following screen shots are from the Siemens Product Lifecycle Management3 (PLM) CAD tool NX
and show a part contrived for the purposes of illustrating embedded GD&T in the Siemens product set.
NX is the primary design authoring environment where the designer collects and captures the MBD data. See
Figure 1.

Figure 1. GD&T in an example part.

4
American Institute of Aeronautics and Astronautics
 Downstream non-CAD consumers don’t need to be NX CAD users. The Siemens product set includes several
specialized tools for these disciplines. One example is NX CAM, which can use NX native CAD files including
embedded GD&T.
For others, the Siemens product set includes non-CAD viewers. This capability is based on the JT data format, as
opposed to the NX native CAD part file. In the Teamcenter PLM environment, the viewer is tightly integrated into
the PLM environment. Figure 2 shows the same part as in Figure 1 in the non-CAD visualization tool Teamcenter
Visualization.

Figure 2. Example part with embedded GD&T in the Teamcenter visualizer.

There are other consumers that aren’t in the managed Teamcenter environment, for a variety of business and IT
technology reasons. The JT2Go JT viewer is a free Siemens viewer for this purpose. Figure 3 shows the sample part
from Figure 1 in the JT2Go viewer.

5
American Institute of Aeronautics and Astronautics
 Figure 3. Example part in free viewer.

These Siemens non-CAD viewers provide the ability to section and measure the model. Some of the tools cap-
ture markup on added layers. When used in the managed Teamcenter environment, the PLM system provides ser-
vices such as access control and versioning.

VI. MBD as Provided by Dassault Systemes CATIA V5 CAD Software

Dassault Systemes1 has two levels of their product offering covering the functionality and capabilities needed to
support processes and methods which enable organizations to move away from two-dimensional (2-D) drawings
and/or 2-D CAD drawing datasets, and increase the use of three-dimensional (3-D) data as the master product
definition data (PDD) representation. These products are 3-D Functional Tolerancing and Annotation 1 (FT1) and
3-D Functional Tolerancing and Annotation 2 (FTA). This section will provide an overview of CATIA V5 concepts,
capabilities, and functionality that organizations can leverage to develop and release PDD and associated
engineering requirements for detailed parts, assemblies, and installations within CATIA V5’s 3-D CATPart and
CATProduct datasets. These capabilities are needed by organizations implementing MBD processes and methods
using CATIA V5.

A. The Problem
MBD is a method of developing and releasing PDD for detailed parts, assemblies, and installations within the 3-
D datasets of a CAD system i.e., CATIA V5. The objective of MBD is to provide complete product definition with-
out the use of 2-D drawings or dressed-up and annotated projected orthographic views derived from 3-D data. The
data elements that constitute an MBD dataset generally reside in a single file. For CATIA V5, these are CATParts
for detail parts and CATProducts for assemblies and installations. MBD requires the use of mono-detail CAD data-
sets for detail parts with associated annotations, notes, material specifications, etc.

6
American Institute of Aeronautics and Astronautics
 Detail part MBD datasets, CATParts, should contain the following elements, which collectively constitute the
master representation of the detail’s product definition: solid nominal geometry; the detail’s coordinate system or
systems; associated dimensions, tolerances, and annotations; engineering notes; general notes; and material require-
ments.
Assembly and installation MBD datasets, CATProducts, should contain the following elements, which collec-
tively constitute the master representation of the assembly’s or installation’s product definition: mono-detail solids,
instanced in the assembled or installed position and condition; associated dimensions, tolerances, and annotations;
engineering notes; general notes; material requirements; and a parts list.
The PDD and the MBD data must be accessible, viewable, and consumable — basically usable — by
downstream users and processes. Various methods and tools are used to produce detailed parts, assemblies, and
installations without the use of paper or electronic 2-D drawing media. Organizations implementing MBD processes
and methods within engineering must also deal with those organizations, functions, and processes that consume
PDD from engineering and/or depend on it as an input to their processes and deliverables. Moving engineering away
from 2-D drawings and/or 2-D CAD drawing datasets to MBD processes and methods demands that Manufacturing,
Supplier Management, and other organizations’ processes, methods, and tools must be integrated, or at least inter-
faced, with engineering’s MBD tools and deliverables.

B. CATIA V5’s Capabilities

Dassault Systemes’ CATIA V5 3-D Functional Tolerancing and Annotation products cover the definition and
management of tolerance specifications and annotations within the 3-D space of CATPart 3-D detail parts and
CATProduct assemblies and installations. CATIA V5 provides a comprehensive set of 3-D annotation: associative
dimensioning, tolerancing, geometric dimensioning and tolerancing (GD&T), annotations, text notes, flag-notes, and
other dress-up features within the CATIA V5 3-D datasets. The 3-D annotations are created and managed within a
branch of CATIA V5’s specification tree called an annotation set. This branch contains a hierarchy of sub-branches
for the different types of 3-D annotations.
The system allows the user to create and manage 3-D tolerance specifications and annotations directly on 3-D
parts, assemblies, or installations. The interface allows the user to select the surfaces or other elements to be anno-
tated or toleranced from the 3-D graphic area and to select from the available options, tolerance types, other modifi-
ers, etc. The system offers a choice of options consistent with the selected elements. The tolerance annotation is then
created, located and oriented, and displayed around the 3-D geometry. CATIA V5 supports Cut, Copy, Paste, and
Drag & Drop of tolerances and annotations typical of Microsoft Windows functionality.
Engineering and general notes can be created and viewed in 3-D space, similar to tolerances and annotations. Or,
string (text) parameters can be used for notes where the parameter’s value is the text of the note or note code.
Material specifications can be applied to the entire detail part, as with a metallic part, or to sub-elements of the
part in the case of a composite part.
Dassault Systemes has product offerings for the viewing of 3-D annotations and interpretation of annotations and
tolerances on specific areas of the PDD, or across the complete digital product.

C. How CATIA V5 Solves the Problem

The system provides basic syntactic and semantic creation and checking of geometric tolerances. Syntactic geo-
metric tolerances and annotations are created using the provided nonsemantic functions and are created without any
checking or verification by the system. These non-semantic tolerances and annotations are created based entirely on
user input. Although nonsemantic tolerances are 3-D and displayed in the 3-D space concurrently with the 3-D
geometry, they are typical of annotations on 2-D drawings. With semantic creation and checking, the validity of
dimensioning and tolerancing specifications are defined by the international standard being used – ISO, ANSI, JIS.
CATIA V5 is fully compliant with the latest revision of these major international standards. The creation of correct
tolerances and annotations is aided by the Semantic Tolerance Advisor function. Semantic geometric tolerances are
defined and validated with respect to the active CATPart or CATProduct. The user can set system options for the
default tolerancing standard. Defaults can also be set for angular, linear, and geometrical tolerances.
The Text function is used to create 3-D engineering and general notes, which are positioned and displayed
around the 3-D geometry. Parameters of the type String can also be used as notes. These parameters are organized
and displayed in the CATIA V5 specification tree. When using a string parameter as a note, it is the value of the
parameter that is the text of the note or note code. The parameters can be placed in the Parameter branch of the
specification tree or any Geometrical Set in the tree.
Within the 3-D space, tolerances, and annotations are located and oriented in what is referred to as an annotation
plane. Annotation planes are system- or user-defined planar references used to help organize and position tolerance
7
American Institute of Aeronautics and Astronautics
specifications and annotations in the 3-D space of the CATPart or CATProduct. Annotation planes are defined and
displayed around the 3-D geometry.
Material specifications can be applied from a catalog of materials to the entire detail part, as with a metallic part,
or to sub-elements of the part in the case of a composite part. When a material is applied, a Material parameter is
automatically created in the specification tree. A basic material catalog is provided. Custom material catalogs can
also be developed and employed. Inertia characteristics, based on the applied material definition, can be calculated
and retained. These include density, mass, center of gravity, inertia matrix, and moments.
The system provides tools to create and manage “captures,” which filter the show/no show display of tolerances,
annotations, and/or annotation planes according to a specific context. The user can define which elements are in the
capture and specify a point of view by associating a camera to the capture. Captures, which are managed in the
specification tree, provide easy access and navigation to preconfigured views for downstream users who don’t
necessarily have the skills to navigate the 3-D graphic area.

D. Conclusion
CATIA V5’s semantic creation and checking capability ensures the validity of dimension and tolerance specifi-
cations based on selected international standards. Organizations can set up system options for the default angular,
linear, and geometrical tolerances. In order to provide integration between engineering and manufacturing, Dassault
Systemes provides integration between CATIA and DELMIA through their DELMIA DMU DIMENSIONING
AND TOLERANCING REVIEW 2 (MTR) product. This product offers comprehensive tools for interpretation of
annotations and tolerances. Beyond the scope of this CATIA and DELMIA integration, organizations will need to
identify other integration and interface problems, issues, and risks and develop solutions. These solutions will not
only access and utilize MBD data, but they will also enable downstream organizations and processes to efficiently
utilize CATIA V5 3-D MBD.

VII. MBD as Provided by Parametric Technology Corporation's Pro-E CAD

Software
Parametric Technology Corporation's2 (PTC) Pro-E solid modeling software is a common package utilized
within the aerospace industry. One of their customers is utilizing the current version of Pro-E to realize the model
based design implementation capable of several maturity levels. The conventional design to manufacturing process
relies on the design creation of solid models and engineering drawings derived from those models. These drawings
are the basis for the bidding, fabrication, and inspection steps. In the drawing creation and revision phases, as well as
in the fabrication documentation phases, a great deal of manual work is carried out, which is inefficient. The discon-
nect between the model and drawing occurs when design changes or “hanging paper” is allowed to accumulate onto
a complete drawing prior to direct incorporation. A longtime goal of model based design is to eliminate the 2-D
drawing step and its associated “hanging paper” and related errors. The key is to organize all the data into the 3-D
model data in a meaningful manner that the supply base can interpret. By supplying free or low-cost model reading
software, Pro-E model data and associated embedded notes, tolerances, contractual information, etc. can be inter-
preted by the suppliers. MBD provides more benefit within the Pro-E environment when compatible PDM systems
are used, viewers use models created to standard rules, and customers are in agreement with this methodology.
The most basic question that arises is how to evaluate a model based on its complexity. For example a simple
part might only have notes, key and critical dimensions, and profile tolerances. For more complex parts, additional
explicit dimensioning, cross sections, detail views, and fabrication specifications would also be embedded in the
model data. For one PTC implementation, the journey from drawings to a fully automated technical data package
(TDP) across an enterprise is defined by six steps of increasing maturity. These increasing maturity design process
models imply changes in the tasks of design, detailing, check, manufacturing and enterprise utilization.
At the larger scale of “big” models that span large systems, the MBD implementation has to be compatible
across a variety of software and hardware covering the planning, designing, fabrication, and sustainment phases of a
product’s lifecycle. The direct impact to the solid model is that its embedded data requirements grow over the life of
the product. The types of embedded data could include either documentation or links to documents that verify fac-
tors such as the design analysis, revision history, trade studies, failure analysis history, etc.
One PTC customer’s experience with MBD demonstrated significantly reduced nonrecurring design costs,
reduced first article costs, reduced TDP changes, shorter design cycle times, and reduced support and recurring
costs. Associated benefits included a reduced learning curve and a more flexible work force, as well as concurrent
design and assembly verification. Product costs were reduced by defining and validating tolerance management in
factory, assembly, and quality processes. The business assessment of the implemented MBD processes highlighted

8
American Institute of Aeronautics and Astronautics
both risks and opportunities; however, the opportunities were found to far outweigh the risks. Risks included factors
such as workforce resistance to adopt new methods, new technology resource needs, additional training, supplier
matching capability to fabricate parts with MBD, long-term data retention, and delivery method changes based on
moving industry standards. Preservation of the legal markings on the model and viewable version of the model is a
critical need to satisfy. The benefits included dramatic schedule reduction in time to market, reduced error rate
throughout the lifecycle, reduction of data duplication and vendor overhead costs, as well as alignment with future
Department of Defense strategy.
The maturity levels from 0 to 6 feature increasing dependency on the model as master concept. Level 0 is the
previous state of pure 2-D drawings and no models. In Level 1, the part is described by a 2-D drawing, which is
related to a 3-D model dataset (neutral format deliverable). Level 2 is the 2-D drawing with the 3-D model in its
native CAD format. Level 3 is the first stage lacking drawings with only a 3-D annotated model depending on light-
weight viewers for downstream utilization. Level 4 is the same as level 3 with the addition of data management
focused purely on models with some human input required. Level 5 is built on level 4 with automated PDM capa-
bility. Level 6 is full 3-D MBD with automated PDM and on-demand enterprise access. An intermediate level seeing
increasing use is limited dimension drawings. These must be augmented with the 3-D model data in order to fabri-
cate the part since only the critical feature tolerances, datums, notes, and legal information are presented on the
drawing.

VIII. The MSFC Foundation for Model Based Definition

In September 2007, the Engineering Directorate at the Marshall Space Flight Center (MSFC) created the Design
System Focus Team (DSFT). MSFC was responsible for the in-house design and development of the Ares 1 Upper
Stage, and the Engineering Directorate was preparing to deploy a new electronic configuration management and
data management system with the design data management system (DDMS) based upon a commercial off-the-shelf
(COTS) product data management (PDM) system. The DSFT was to establish standardized CAD practices and a
new data lifecycle for design data. Of special interest here, the design teams were to implement MBD in support of
the Upper Stage manufacturing contract. It is noted that this MBD does use partially dimensioned drawings for aux-
iliary information to the model.
To support implementation of DDMS, MBD, and a new data lifecycle, the design groups needed to work
together in the definition and implementation of standard practices. Critical design activities that were targeted
included configuration management of CAD models, drawings, and integration of the models into top-level
assemblies.
By April 2008, the director of engineering implemented the first recommendations by the DSFT:

• Begin migration to a new modular product architecture (MPA) based on explicit identification and
management of interfaces between product modules.

• Begin use of Work Group Approval (WGA) process for configuration control of CAD files prior to
formal release.

• Use a Designer Checklist that implements the new CAD standards.

• Implement the policy “Model is the Master” where manufacturing would be accomplished to the released
CAD model.

A Pathfinder was commissioned for the Upper Stage common bulkhead to exercise these new practices. The
Pathfinder ran for seven months and was completed in August 2008. A significant number of issues were identified,
and the DSFT worked diligently to close all 21 of the priority 1 items.
A second Pathfinder activity aimed at releasing a smaller assembly was based on the ullage settling motor
assembly to demonstrate the improved practices. Four problems were identified and corrected through modifications
to training and the Designer Checklist. The resulting standards and practices were documented in MSFC-STD-3528
computer-aided design standard, the Designer Checklist, a set of 29 desktop instructions, and the physical interface
matrix, which identified specific Upper Stage controlled interfaces.

9
American Institute of Aeronautics and Astronautics
A. Work Group Approval Process
The design data lifecycle implemented several new release states to be used prior to formal release that allowed
the models to move through a flow of progressive maturity. These use the read and write capabilities within DDMS
to set the access controls as defined by the lifecycle states.
While designers are initially working on the CAD models, the DDMS state is In-Work. The files are not seen by
others until the designer feels comfortable with the maturity of the files. When the designer approves the models to
be seen by others in the design team, the state in DDMS is set to Design. In this state, the models are stable and set
with Read Only permissions. The designer uses the Designer Checklist to ensure that the file is free from errors and
compliant with the CAD standard. Most of the design iterations occur between In-Work and Design. The commu-
nity at large would recognize that whenever a file was in the Design state, it was available for reference, but the user
would know that it was still in development. Adoption of the In-Work and Design states during the Pathfinder pro-
vided a significant stability to the upper-level assemblies.
The next state in DDMS in the progressive maturity is Workgroup Approved (WGA). Approval of promotion to
this state requires branch chief approval and implements a “prerelease” level of maturity needed for final analysis or
formal review, particularly a critical design review. As a result, a shorter review time for final release would be
needed since the detail analysis and review had already occurred.
These new design data lifecycle additions lead up to the traditional formal release. The designer initiates a
change request to start the approval cycle, and the model and drawing enter the DDMS state Under Review. Elec-
tronic approvals are collected from Checking, Designer, Materials, Producibility, Stress, Safety and Mission Assur-
ance, Branch Chief, and the next higher assembly. When the model is processed through the release desk, the
DDMS state is Released.

B. Lessons Learned
The DSFT identified some seventeen lessons learned as outcomes of the standards development, pathfinder
deployments, and initial application to the Upper Stage design completion. Due to space limitations, all cannot be
presented here, but the following are a few of the high-value examples.
1. Quickly identify the current state of the environment.
The Pathfinder test case brought out the extent of the problems through trial use with a real assembly. In some
cases, these were unknown to the design and configuration management community. Running the Pathfinder 1 and
Pathfinder 2 exercises set the stage for development of the CAD standard, best practices, and the WGA design data
lifecycle.
2. Identify lifecycle states within the design process.
A clear definition of each state and a clear understanding by the user community of the intention and criteria for
each state are critical to ensuring that models are ready to be shared by the community, particularly the assembly
engineering group.
3. Develop a CAD standard.
All design organizations must follow the same standard for developing models and drawings. The CAD standard
is a directive on how models are created within the frame of a specific CAD platform. It is critical that every
designer abide by this standard and that there be no exceptions in the mandatory use of the standard for any project
or program.
4. Ensure that first-level supervisors have buy-in and acceptance of the solutions, and ensure senior
management communicates their acceptance and direction to follow the solutions.
The first-level managers supply the personnel to accomplish the design activities. The DSFT met with the
managers, requested recommendations, and worked through the impact of the DSFT decisions on the supervisors’
work commitments. Communications need to be open and free of hidden agendas to make sure these supervisors
know the status of the implementation on a regular basis.
Prior to issuing a directive, official Engineering Directorate memoranda were distributed and then discussed in
designers’ forums. This gave everyone an opportunity to impact both the directive and how it would be
implemented.
5. Make sure senior management is informed regularly of progress and problems.
Formal reports were made to the management council, and weekly working meetings were held to address
needed corrective actions. Accomplishments, upcoming events, and broad issues were discussed at the tri-weekly
management stand-up meetings.
6. Make sure that there is agreement on when files can be used by the broader design community.
Encourage the use of a schedule of major design state dates published by the assembly engineering organization
for all design activities, but developed in conjunction with the design branch chiefs. This provides real milestones
10
American Institute of Aeronautics and Astronautics
for the push to the Design state. In return, the designer accepts the responsibility of pushing the models to the
Design state as often as possible prior to the due date to aid the assembly engineers.

IX. The Business Case

Any discussion of the business aspects of such a significant process change will necessarily summarize the cost
impacts in terms of reduced operational costs offset by investment costs. There are additional benefits that are often
not expressed in financial terms, although it is possible that they could be. Examples include reduced schedule risk,
reduced product performance risk, and reduced calendar time in design and fabrication.
The financial impacts will also be viewed differently under varying production rate, product model similarity
and reuse, and degree of customization. In an attempt to establish an initial common basis, this section will summa-
rize those issues that are largely independent of these factors.
All the data required to define the product are currently captured and available to downstream consumers, such
as manufacturing, although these data are actually captured and distributed in a wide variety of systems and formats.
One of the primary benefits of MBD is to consolidate this information in a single electronic source.
One widely understood benefit of MBD is a significant reduction in manually reproduced data. For example,
CNC programmers and machinists are often reconstructing the part geometry in their CAM tool and encoding tool
selection and finishes based upon their reading of the tolerancing shown on the face of the drawing. This can be a
source of errors from simply misreading a number to misunderstanding the GD&T design. More recently, the
drawing might be augmented with a CAD solid model which, subject to difficulties in translation and import, can
reduce the subsequent errors. A related risk can arise if the configuration control of the drawing and its matching
solid model rely on manual practices such as requiring the consumer to match the model to the drawing because the
drawing and the model travelled different paths.
With tools that fully support MBD, these risks are significantly reduced. When the CNC programmer checks out
the design model, it contains all the geometry and GD&T as constructed by the design engineer. When loaded into a
CAM tool which recognizes both the geometry and the tolerances, the CAM tool can provide machine tool and path
selections that meet the design intent. Thus, there are fewer data and process errors and significantly reduced CAM
programming time to reach first tape out.

A. Estimating Cost Savings

Many organizations are satisfied with an estimate of potential savings that can be based on a few typical, but
relevant, scenarios. For example, the CNC programming group might provide typical time estimates for a current
process that recreates the geometry and, perhaps with some hands-on trials, estimates the reduced time when given
the solid model. They might also estimate the time saved from automating the GD&T implementation, perhaps
focusing on the less-skilled portion of the workforce.
The time saving per programming activity can be significant when considered across the number of parts manu-
factured. Although recovering from errors occurs less frequently, the cost of recovery can be quite a bit higher. It
might be acceptable to examine recent errors and the costs of recovery and then estimate the occurrence rate for the
near future. When the cause of the error is eliminated by the automation of the new process, these estimated costs
are avoided and can be used in the business case.
An example might be a scrapped part caused by the long time it takes to reconstruct the part geometry. Perhaps
the fabrication team had to start quite early and the design team might have moved the design forward in the interim.
The two teams may well have reviewed the design prior to committing to fabrication, but it can easily happen that a
visual inspection misses a design detail that was changed. This might not be discovered until quite a bit later, after
part inspection, when the assembly won’t fit together. A similar failure can occur if the programmer interprets the
GD&T design in such a way that the assembly doesn’t fit together.
Other opportunities for savings might occur if outside part fabrication shops are used. This business practice
leads to a more arms-length working relationship, at times, such that questions of interpretation that arise during
fabrication do not get reconciled with design. Often, such clarifications lead to fabrication changes that invalidate
the contract bid or extend the delivery period. MBD can reduce the occurrence of design issues and, when imple-
mentation is accompanied with supplier electronic access, better communications can lead to quicker responses.
Armed with a variety of such examples, the cost estimate can be based on a generally accepted rate of occur-
rence. Experience with this approach can be satisfactory in many organizational engineering cultures, and the mag-
nitude of the savings is often significant enough to fund the initial deployment stages.

11
American Institute of Aeronautics and Astronautics
B. Estimated Implementation Costs
The cost of implementation can include new software for designers and users such as fab, as well as new IT
hardware systems for configuration control and management of the MBD models now shared across new
organizations.
Training will be a part of implementation and should cover using new software capabilities as well as using new
business practices. Depending upon organizational practices, supporting consultants might be used to assist the
workforce during adoption and initial deployment.
Labor costs for the technical team might require funding, depending upon the organization’s practice.

C. Deployment
Getting MBD into use can involve significant new investments in IT systems, a healthy dose of organizational
behavior and practice change, and changes to business operations and contracts with suppliers. As a result, adoption
can require extended time frames; three to five years is not uncommon.
As is commonly recommended in such efforts, a phased deployment might be based upon incremental steps such
as technology exploration, proof of concept, pilot deployment, and operational deployment. Large organizations
with multiple development sites might roll out the pilot and operation deployments by location. Other organizations
with multiple program or product lines that are largely independent might roll out by program line.

X. Conclusion
Aerospace companies are in the midst of many improvements to the product development and delivery process,
and the replacement of 2-D drawings with model based definition is a challenging current example. MBD is widely
supported by design software, and tools for use in manufacturing and quality assurance are becoming available.
With study, investment, and exploration, companies can gain the promised benefits of lower technical risks and
shorter development times.

References
1
Dassault Systemes, http://www.3ds.com.
2
Parametric Technology Corporation, http://www.ptc.com.
3
Siemens PLM Software, http://www.plm.automation.siemens.com.
4
American Institute of Aeronautics and Astronautics, http://www.aiaa.org.

12
American Institute of Aeronautics and Astronautics