Kutz M. Handbook of materials selection
Подождите немного. Документ загружается.
475
CHAPTER 17
MANAGING MATERIALS DATA
Deborah Mies
MSC.Software Corporation
Santa Ana, California
1 HISTORY OF MATERIALS DATA
MANAGEMENT 476
2 IMPLEMENTING A MATERIALS
DATA MANAGEMENT SYSTEM 479
2.1 Planning 480
2.2 Implementation 480
2.3 Deployment and Support 484
3 CREATING THE DATABASE 484
3.1 Defining the Project Team 485
3.2 Defining the End User Data
Requirements 485
3.3 Defining the Functional
Requirements 486
3.4 Designing the Database 486
3.5 Developing a Prototype Database 490
3.6 Populating the Database 490
3.7 Building the Database 491
3.8 Customizing the User Interface 492
3.9 Qualifying the Database 492
4 COMMERCIAL DATABASE
MANAGEMENT SYSTEMS 494
5 MATERIALS DATA STANDARDS 497
6 SUMMARY 502
REFERENCES 502
Previous chapters have described the sources and usage of materials data in
product manufacturing organizations. This chapter discusses the process of man-
aging materials data. After examining its history, discussions will be provided
on the design and implementation of materials data management systems, the
design and creation of the underlying database(s), and the evolution and use of
standards for materials data reporting and exchange. A brief summary of com-
mercially available database management systems is provided. We will begin by
establishing the premise that the ability to manage the vast amount of materials
information used in manufacturing organizations is perceived as a tremendous
tool for increasing efficiency and profitability.
Materials data is used to support many different functions in a manufacturing
organization from the initial phases of design to the disposal requirements at the
end of a products’ life cycle. The most critical of these functions is the process
of analysis, used to predict the product’s performance in its end-use environment.
The completeness of the required data sets is crucial to most forms of mathe-
matical analysis required for engineering simulation. The pedigree or back-
ground data for the tested material property data, as well as the specific
properties required for the analysis must be readily available to the analyst.
These properties are either acquired from established sources or the result of
rigorous testing or a combination of the two. It has been estimated that the cost
HandbookofMaterialsSelection.EditedbyMyerKutz
Copyright Ó 2002 John Wiley & Sons, Inc., NewYork.
476 MANAGING MATERIALS DATA
of testing for a complete set of materials properties required by a designer for
a new application can be in excess of $1 million with an acquisition time of
greater than 5 years.
1
Additionally, it is estimated that engineers spend over half
their time acquiring and verifying the validity of this data. Effective management
of this data is paramount for organizations having functions that are materials
data intensive. It reduces the cost of materials data acquisition, improves the
usability and reusability of acquired data, and increases the efficiency of the
engineering organization.
Historically, the confidence in a new product design relied on large safety
margins and field service experience. The introduction of computerized design
and analysis systems offered the promise of selecting the right material for an
optimized design early in the product development cycle. This requires that
consistent and high-quality materials information be used throughout the design
to manufacturing process. In addition to the resulting cost reductions, the benefits
of effective data management are realized in terms of higher quality, decreased
warranty costs, shorter product development cycles, and traceable and auditable
systems.
2
So what does ‘‘materials data management’’ mean to an organization? It
means:
●
Making the materials properties acquired by an organization available to
all functions requiring that data within the organization in a consistent
and timely manner from an official, centralized source
●
Providing a means of controlling the sources of materials properties used
for a given function
●
Controlling access to specialized and proprietary information
●
Providing the available materials properties in a format that optimizes its
usability and reusability
1 HISTORY OF MATERIALS DATA MANAGEMENT
In many organizations, the engineer responsible for a specific analysis maintains
the information he requires. It is often found in old notebooks, jotted on scratch
pads, in textbooks, and other references known only to him. This data provides
little, if any, traceability and limited pedigree information. The values are often
just simply those that have ‘‘always been used.’’ If the engineer leaves the or-
ganization, the search for the required data begins again. Often the data is not
accessible to other individuals within the same company, and the acquisition
process is duplicated, sometimes resulting in the use of different values for
identical properties, producing inconsistencies in analysis of the same system or
structure. It has been estimated that 50% of an engineer’s time is spent searching
for and verifying materials property data.
Organizations for whom traceability is important often require ‘‘certified’’
data, which has been evaluated by experts and approved for a specific applica-
tion. These organizations have long recognized the value of the materials data
they use. Many organize the data into official handbooks that are distributed
throughout the organization, fulfilling the requirement for controlling the data
used in their analysis processes. However, the process of creating the handbooks
1 HISTORY OF MATERIALS DATA MANAGEMENT 477
is time-consuming and costly. Rocketdyne Division of Boeing North America
documented the process for adding data from a vendor for a new material into
its handbook
3
in the following manner:
1. Test data obtained from vendor was typed into a curve-fitting program.
2. A best-fit curve was constructed based on the test data.
3. Archived data and data from other sources for the same material and test
condition were retrieved to compare with the newly constructed curve.
4. If a difference existed between the new data and historical data for similar
materials, the difference was reconciled in one way or another, and the
new data was finalized.
5. A written report was generated for distribution to those who immediately
needed the data.
6. The report was archived in the files.
7. The newly constructed curve would eventually be entered in the hand-
book.
Since the hardcopy manual was difficult to update, this process may have
taken up to a year. Rocketdyne’s hardcopy materials properties manual consisted
of four volumes with over 1500 pages and 5000 curves. Revisions to this manual
involved a major effort in terms of time, as well as paper.
As usage of computers became more widely available, companies began to
realize the potential for tremendous cost savings through the optimization of
their processes. They envisioned the capability for systems that would allow
them to go beyond mere data storage within their organizations. The specialized
use of computers in storing materials data for access by designers and analysts
companywide provided new capabilities to search for and compare materials
based upon specified properties and allowed designers to consider new materials
options early in the design phase of a new product cycle. The possibilities for
shortening the design cycle, making products of higher quality, improving the
consistency of manufactured parts became a reality. Most of these improvements
were driven by the need to improve profitability, others by the need to provide
audit trails as required for ISO 9000 compliance. Response to these new re-
quirements has led to the development of numerous ‘‘materials databases’’ in
the past decade, either as commercially available products or in-house devel-
opments.
The introduction of product data management (PDM) systems to the market-
place provided additional tools for the integration of product data throughout an
organization. While slow to catch on, by 1999 the estimated market size for
PDM systems reached $1.6 billion. Its growth was driven by an increasing
awareness of the value of these systems in improving workflow, configuration
management, document and information integrity, and communication through-
out the enterprise.
4
Various specialized functions, including materials data
management systems, are now being linked with PDM systems. For the manu-
facturing community, PDM is a natural complement to computer-aided engi-
neering (CAE), by providing a repository for data that is accessed by CAE. The
integration of CAE and PDM enables the access, verification, and control for
478 MANAGING MATERIALS DATA
CAE input and output data.
5
‘‘Simulation data management’’ is vital to the in-
tegration of CAE into the product development process. The benefits are anal-
ogous to those of materials data management systems and include:
●
Improved access to authorized data sources
●
Higher throughput for analysts and designers, resulting in faster and more
reliable engineering decisions
●
Fewer simulation errors, increased repeatability of simulation methods,
with full traceability
●
Enhanced collaboration between design and simulation, within the enter-
prise and throughout the supply chain
The Internet, or World Wide Web, has been credited with the substantially
increased market for PDM-type systems. Being universal, inexpensive, accessi-
ble, and hardware-independent, web-enabled PDM formed a prototype for web-
enabled materials data management systems. Even in organizations where the
databases are not fully integrated, and legacy systems remain, web enablement
allows the data from multiple sources to be displayed on the screen at the same
time. This includes not only data from the PDM system but from the shop floor,
parts management systems, purchasing, finance, shipping, and legacy systems
as well. Once accessed, a web interface, with the look and feel of standard
reporting mechanisms, can be used to generate reports by simply printing the
screen. Most companies find that establishing a workable web browser interface
as a front end to their PDM system is quite simple, requiring minimal program-
ming expertise and technical support.
6
While the Internet provides access from or to vastly different sources, the
Intranet has enabled organizations to distribute materials information from a
centralized location to all its satellite offices and divisions. This bourgeoning
expertise has facilitated the distributed ‘‘build-design’’ teams in many global
enterprises that support the ‘‘design anywhere, build anywhere’’ business model.
As use and acceptance of these systems expand, additional applications have
surfaced in the area of business-to-business collaboration. The power of the
Internet allows companies to readily transact with each other and access each
other’s information, making collaboration a reality. The first wave of Internet
collaboration was in the supply chain, as companies worked with customers,
suppliers, and intermediaries to manage inventories. By simplifying transactions
and sharing information, companies continue to drive down costs, cut cycle
times, and reduce inventories. The cutting edge of supply-chain collaboration is
moving upstream from there. Intermediaries and suppliers are now doing col-
laborative planning and forecasting, creating a shared vision of what customers
will eventually buy, so that the right product can be in the right place at the
right time. Companies are also collaborating on product development. Instead
of a system maker and a component supplier developing their products separately
and then trying to make them fit, they use a shared design database to develop
everything together from the outset. In some industries, competitors are stan-
dardizing on common parts and practices by sharing information over the Inter-
net. Companies must be prepared to work with others—not out of a sense of
2 IMPLEMENTING A MATERIALS DATA MANAGEMENT SYSTEM 479
altruism, but for shared benefit.
7
Today, an array of ‘‘off-the-shelf’’ systems and
custom implementations are available, making collaborative PDM tools conven-
ient and economical for even small and medium-size companies.
2 IMPLEMENTING A MATERIALS DATA MANAGEMENT SYSTEM
Based on the previous discussion, it is apparent that the primary drivers for
materials data management in today’s industries are the requirements to:
●
Reduce the time to manufacture and the associated need for quick access
to approved data.
●
Disseminate or exchange data between different functions or organiza-
tions.
●
Increase product quality and consistency resulting in lower warranty costs.
●
Increase profitability by decreasing design, manufacturing, and testing
costs.
●
Improve traceability of data pedigree to substantiate product warranty
claims.
Therefore, a materials data management system must be more than just a com-
puterized version of a set of handbooks.
8
It must serve the material information
requirements of the entire product life cycle. These include:
●
Conceptual design, where materials information is required to screen can-
didate materials based on functional performance requirements and prior
experience with candidate materials used in similar products.
●
Design, where materials information is needed for specific (stock) mate-
rials, their grades and finishes, standard components and/or joining ma-
terials as specified in the engineering bill of materials. At this stage,
specific material objects are designed and analyzed for end-use environ-
ments including manufacturing, assembly, service, maintenance, and dis-
posal.
●
Manufacturing, where materials information is needed for specific mate-
rial object manufacturing processes, tooling, assembly fixtures, joining
materials, dunnage, and in-process materials consumed during manufac-
ture.
●
Operation, where materials information is needed for in-service operation,
including data for health monitoring, maintenance, repair, and other ma-
terial-specific operating information provided by the manufacturer to the
end-user.
●
Disposal, where materials information on composition, procedures for re-
cycling, and other environmental impact data are required at the end of
the product’s service life.
9
The success of a materials data management system lies in appreciating that
the end users at each step of the product life cycle must access, retrieve, and
manipulate approved materials data, and that the system structure, its user in-
480 MANAGING MATERIALS DATA
terface, and its data content must be designed to provide for this wide range of
usage at the outset. A well-designed and integrated materials data management
system should provide efficient archiving and distribution of knowledge, elimi-
nating duplicate materials testing, and optimizing the number of specified ma-
terials.
10
Most current configurations of materials data management systems use client-
server technology, incorporating a centralized database management system
(DBMS), software applications that enhance the DBMS, and web access. As
discussed earlier, the flow of data throughout the organization, exported in a
usable format to each function may be enhanced by PDM configurations that
interact with the materials data management system. Then the materials data
becomes integrated throughout the organization. Otherwise, web enablement can
provide easy access to the materials data management system by simple entry
of a URL in a web browser window. The web-enabled materials data manage-
ment system is the focus of the following discussions.
2.1 Planning
A successful implementation requires full management commitment and careful
planning, which includes the following elements:
●
Clearly identifying the vision for a successful implementation, stated in
terms of the company’s business objectives. The vision statement is used
to guide decisions for purchases and implementation and to define the
metrics for success.
●
Creating a project plan, resource allocation plan, and time allocation plan.
●
Risk analysis and mitigation throughout all phases of the implementation.
●
Planning and managing deployment and closure of each project and sub-
project.
2.2 Implementation
Implementation of a materials data management system is a customization of
standard software development practices. The four phases in implementing a
new software solution include analysis, design, implementation, and testing. The
following sections provide information regarding the application of this process
to the design of a materials data management system within a manufacturing
organization. Implementation and testing, therefore, is followed by deployment
of the software throughout the organization and system support.
Analysis
The initial discovery phase provides the backbone for the entire implementation.
Its goal is to understand the organizations’ compelling needs and/or primary
business drivers for implementation of materials data management, how it should
function within the organization, and how the system should be deployed. State-
ments similar to the following represent typical business drivers:
●
To reduce cost and process cycle time by a factor of 10
●
To improve communication, integration, and flexibility
2 IMPLEMENTING A MATERIALS DATA MANAGEMENT SYSTEM 481
●
To promote innovation and understanding of the use of new materials
●
To improve productivity by running all operations from a single materials
database
●
To eliminate redundant materials testing by making data available to all
analysis functions
This initial phase must first define and document the company’s product de-
velopment process and how materials data is currently used in that process. The
use of detailed process charts for the mechanical design and analysis of the
product can clearly identify the data flow requirements and allow for standard-
ized methodologies.
11
All end users of the system and/or end-use applications
with which the system must be integrated should be identified during this pro-
cess. This may include the following materials-related functions:
●
Product design
●
Engineering analysis
●
Materials engineering
●
Manufacturing specification
●
Materials ordering or bill of materials
●
Other applications, such as PDM or collaborative PDM systems
●
Functions at satellite offices
For each function, the following should be identified:
●
The current data sources
●
The types of data to be stored in the database
●
How the data is retrieved and viewed
●
The form the data must be in to be read by the individual applications
●
How data it is to be reported or extracted from an application
●
What standards and terminology must be utilized by the system
●
Security or controlled access requirements
●
Versioning control requirements
This preliminary requirements gathering provides an understanding of what
schemas, databases, user interface requirements, queries, and other customiza-
tions may be required within the end user’s environment. It also provides an
indication of the internal and external resource requirements. The project may
be further promoted by assessing, near term and long term, the cost of inaction.
Deliverables: Preliminary findings document, vision statement, and the data
flow diagram.
Design
The system design phase must produce a detailed set of requirements for the
system. Software requirements include the definitions of the user interface soft-
ware, the DBMS, and any software required for versioning or controlled access
482 MANAGING MATERIALS DATA
to the data. Hardware requirements should include preferred platforms for both
client and server. From this, an initial materials information solution can be
proposed and a return on investment (ROI) analysis provided.
Selecting a DBMS requires understanding your organizational requirements.
While databases are traditionally designed to simplify the development of data-
intensive applications, a carefully selected database management system can
eliminate development altogether, by providing enough functionality that the end
user can use the tools immediately to enter, examine, and process application
data.
12
Traditional database requirements include:
●
The ability to store a large volume of structured data
●
Capability to retrieve data from storage via a simple query mechanism
●
Distributed capability, allowing data storage on more than one location or
machine
●
An applications package, providing easy data modeling capability
●
A crash recovery system, providing the ability to bring the database back
to a consistent state automatically
●
Data independence, allowing data changes without modifying existing ap-
plication programs
●
Reasonable performance requirements
Additionally, the DBMS used to store materials property data must have the
capability to store and handle complex data types, including numerical values
with units, footnotes, and numerical precision, graphical data, images, and doc-
uments.
The user interface software for a web-enabled materials data management
system must typically provide the following capabilities, most of which will be
identified in the discovery phase:
●
Graphical displays for comparison of materials properties
●
Search capabilities based on product names (character strings) or prop-
erties (numerical values)
●
Interaction with data pre- and postprocessing applications, such as Excel
or other proprietary software
●
Ability to add to or delete data from the database
●
Ability to extract or translate required data sets to identified applications,
in required formats
●
Ability to utilize existing standards for data modeling
●
Ability to convert units between standardized unit systems, ensuring con-
sistency in an analysis and compatibility with analysis or postprocessing
software
●
Customizability of the user interface, using common web interface meth-
ods and technology such as HTML and/or Java programming.
13
●
Versioning control or provision for a formal update process to ensure that
a user can verify use of the most current data
2 IMPLEMENTING A MATERIALS DATA MANAGEMENT SYSTEM 483
●
Secured or controlled access to subsets of stored data. Compilations of
engineering knowledge, such as proprietary materials property data, rep-
resent a substantial investment for manufacturing companies, access to
which must be carefully controlled. This is particularly challenging in
global markets, where manufacturers may be partners on one project and
competitors in another.
13
When dealing with the Internet or Intranet and
PDM, both PDM security access and transmission security must be im-
plemented. Outside a company’s firewalls, transmissions can be secured
by encrypting files in the vault prior to transmission. Within companies,
usage can be secured using standard secured access processes.
4
●
Integration with other systems via PDM, application programmatic inter-
face (API) or integrated client technology.
Deliverable: Preliminary statement of requirements.
Implementation
The implementation process is generally scheduled over a period of time, phas-
ing in the database building process with implementation by the various func-
tions requiring access to the data. This process can be facilitated logically by
the use of the data flow diagram constructed in the initial design phase. It is
often helpful to identify low-risk, high-return pilot implementation opportunities.
The tasks for this phase include:
●
Definition of implementation deliverables and milestones
●
Identification of process redesign necessary prior to deployment
●
Development of a timeline to achieve a total implementation (the imple-
mentation plan), including both pilot and fan-out
●
Need for documentation and training
●
Development of an executive summary based on results for reference by
management
●
Estimates of the return on investment (ROI), based on the implementation
plan
Deliverables: ROI Projections, requirements document (user and high-level
technical requirements).
Testing
This phase includes the installation of hardware and software for a specified
portion of the organization or a specific function. To provide for a thorough
evaluation of the software, a prototype database must be designed, built, and
installed. The user interface must be designed or customized, and training ma-
terials created. Extensive testing must be performed on this initial concept of
the solution. Usage, performance, support requirements, user issues, and data
issues should be documented. From this, modifications to the software will be
identified and implemented. Resource requirements can be refined, and the de-
tails for fan-out can be specified and adjustments made as required. ROI pro-
jections can be revised accordingly.
484 MANAGING MATERIALS DATA
Success at this stage can be measured by the ability of a team to quickly
integrate the system into their day-to-day activities, documented time savings in
retrieving data, exporting data to specified applications, comparing materials,
and meeting reporting requirements.
Deliverables: Design document, and a functional application to be scaled
into a production application in the next phase.
2.3 Deployment and Support
At this phase, upgraded software is installed at the pilot installation site, verified,
and evaluated. Customization of the user interface is refined. Training should be
extended to include the remaining user and administrative community. Database
production procedures should be developed and implemented. Database building
is expected to continue, but the schema design should be finalized based on user
experience.
If the upgraded implementation is satisfactory, fan-out is continued based on
the adjusted implementation plans. In the case of web-enabled access, the ap-
plication software can be installed on the designated servers and accessed from
client machines. The web browser interface should be refined based up the most
up-to-date customized version of the software application. The requirements for
ongoing support should be defined.
Success is measured by centralized access to approved materials data for all
specified engineering and manufacturing functions. The majority of end users
should feel comfortable with the revised materials processes. Materials-related
activities should exhibit increased efficiency.
Deliverables: Customer training/administration/user manual(s) and produc-
tion application with complete production data integration.
3 CREATING THE DATABASE
The following is an overview of the process used to design and create a materials
database that best meets the specifications of the end users. It is not the intent
of this section to provide an exhaustive study of database design and imple-
mentation but only a general understanding of the scope of the tasks for creating
a database useful to a manufacturing organization. There are many excellent
resources available on this topic, from extremely technical to hands-on resource
guides. In most instances, the reference guides provided with DBMS software
adequately describe the technical details of database design required to physi-
cally create a database for use with that application.
The success of any project, especially that of creating a materials database
from numerous sources of data, requires careful planning and the commitment
and organization provided by a project team. The approach outlined in this
section requires defining the database in the context of the individual end-use
applications to assure that all end-users have as complete a data set as possible.
The requirements for each end-use application may then be merged to create
one database. After the database is compiled, the interface can be customized
to optimize use of the database by the various end users.