[iwar] [fc:Digital.Bloodlines.Make.JSF.A.Different.Breed]

From: Fred Cohen (fc@all.net)
Date: 2002-01-29 07:23:03
Next message: Fred Cohen: "[iwar] [fc:More.Online.Security.Woes.For.FBI's.Data.Firm]"
Previous message: Fred Cohen: "[iwar] [fc:Killed.At.Their.Keyboards]"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Message-Id: <200201291523.g0TFN3s21231@red.all.net>
To: iwar@onelist.com (Information Warfare Mailing List)
Organization: I'm not allowed to say
From: Fred Cohen <fc@all.net>
Mailing-List: list iwar@yahoogroups.com; contact iwar-owner@yahoogroups.com
Precedence: bulk
Date: Tue, 29 Jan 2002 07:23:03 -0800 (PST)
Subject: [iwar] [fc:Digital.Bloodlines.Make.JSF.A.Different.Breed]
Reply-To: iwar@yahoogroups.com
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

Aviation Week &amp; Space Technology
January 28, 2002
Digital Bloodlines Make JSF A Different Breed
Software allows Lockheed Martin's Joint Strike Fighter team to manage it as
a 'product' that will receive a lifetime of upgrades 
By Michael Mecham, Fort Worth 
Martin McLaughlin was discussing the wonders of high-speed milling machines
recently, pointing out perfectly aligned filets and unblemished machined
surfaces as he stood next to a wing box jig here at the World War II-era
factory where Lockheed Martin Aeronautics Co. will assemble the F-35 Joint
Strike Fighter. 
McLaughlin, a senior Northrop Grumman manager who is the JSF team leader for
manufacturing integration, is tackling the challenge of how to assemble
aircraft for three different types of customer as if they were all the same.
Pointing to the jig, he notes that the same bill of materials will be used
to produce wings for the conventual takeoff and landing (Ctol) aircraft for
the U.S. Air Force, the carrier version (CV) for the U.S. Navy and the
short-takeoff, vertical landing version (Stovl) for the U.S. Marines and the
U.K. Royal Air Force and Navy. 
But the three wing boxes don't look the same to a milling machine. Beneath
their skins each design will reflect the different demands of its user--the
more punishing landing requirements of the Navy's carrier version requires
thicker inner surfaces, for instance, than land operations by the Air Force.

The JSF is the biggest procurement program in Pentagon history, but rates of
production among its various customers are undecided. What is known is that
Lockheed Martin Aeronautics Co. (LMAC) and teammates BAE Systems and
Northrop Grumman are unlikely to have the luxury of producing a long run for
one service followed by a separate run for another. McLaughlin assumes that
Ctol, CV and Stovl orders will be mixed up. 
SO HOW DO YOU SWITCH from making CV to Ctol wing boxes without incurring the
high costs of rejigging an entire assembly line? To start, hold design
differences between the boxes to a minimum, which is easier said than done.
Push too hard for commonality and the mission performance of each fighter
degrades. Put too much variance in each model and manufacturing costs
skyrocket. 
"The trick is to optimize the structural loads yet get economies of scale as
if they are the same aircraft," he explained. 
Fighter jets aren't cars cranked out by the thousands. But given
McLaughlin's expectations on how flexible the JSF's production line must be,
the program will need to be able to shift from one model to another with as
little disruption as possible. As an illustration, he pointed to the way
that high-speed numerical controlled (NC) milling machines--which have
spindle speeds of 15,000-30,000 rpm. --produce "as is" surface finishes that
don't require mechanical reworks. A wingbox cut by the new machines is
one-third the price of a conventional product. 
The NC machines get their instructions directly from Dassault Systemes'
Catia computer-aided design and manufacture (Cadam) software. The JSF was
designed completely using Catia's 3D solid modeling, which allows each part
and surface to be represented with absolute fidelity because the model
understands the individual characteristics of each part and how it interacts
with other parts. 
Solid modeling is hardly new. Northrop pioneered its use among military
aircraft more than a decade ago with the B-2 bomber and Boeing did the same
for commercial transports with the 777. But, McLaughlin said, "JSF is the
first aircraft to be designed exclusively with 3D solids." 
Aerospace's advances in information technology this past decade have more to
do with increasing manufacturing efficiencies in line with the goals of lean
manufacturing than they do with the perfection of design or software
expertise. The first-time perfect wing mate of the 777 was a "wow" for the
industry. But infotech applications were not available for the whole
design-to-delivery-to-support enterprise that a program of JSF's complexity
demands. 
For instance, the Catia versions used by Boeing's 777 team were limited to
expensive Unix mainframes. That made data storage a constant worry, recalled
Kevin Fowler, who led the 777 structural program. JSF is using Catia Version
4.2, which supports distributed desktop workstations and overcomes the
storage crunch. As JSF shifts into full production and adds second- and
third-tier suppliers, LMAC expects to transition to the use of Catia's V.5
later this year. That opens the design phase to less costly Windows NT
operating systems. 
Similarly, the blooming of the Internet has made bandwidth available that
the 777 and B-2-era design teams didn't have, so they couldn't employ
real-time file sharing for a global supplier and user base. Global sharing
is the heart of the JSF program. "Our first big enabler is the Web and the
second are neutral protocols," said Larry J. Mestad, senior manager of JSF
airframe's integrated product teams (IPTs). 
There are "islands of software" throughout the JSF program to be managed,
software director Lloyd A. Huff said. As might be expected, the totals are
impressive. Fifteen million lines of code support the aircraft, including 6
million for inflight operations and another 9 million on the ground.
Thirty-five software processes had to be ready to start the program and they
were supported by some 400 software tools. Huff classified 90 of these as
major. 
The JSF team will rely heavily on virtual private networks to carry Catia's
big data sets for real-time performance, but it will use the Web and
wireless applications for document transfer. 
TIME AND TECHNOLOGY advances are cutting costs and manufacturing setup times
on the shop floor. High-speed NC machines, working directly from the Catia
Cadam data, permit production of huge parts such as wing boxes in single
cuttings. The NC machine's extremely fast spindle speeds throw off chips
from the cutting process which carry heat away from the part. Less heat
buildup reduces warping concerns and that eliminates the need for mill
fixtures. Without mill fixtures to change, the process of switching, say,
from a Stovl wingbox to a Ctol design is a matter of punching in a new code
for the NC machine, McLaughlin said. 
Direct programming from the Cadam software has dramatically reduced setup
times as well. Where it once required six weeks of programming and six weeks
of tool trials to verify a cutting procedure, the process now can be
completed in a day. 
Such process improvements have been evolutionary for the industry. The JSF
is benefitting from auto-drilling techniques pioneered by Northrop Grumman
as a partner with Boeing on the F/A-18E/F program. Similarly, Northrop
Grumman's introduction of advanced composite fiber placement techniques that
use Fibersim software to manage the 3D solids data has been shifted to JSF. 
Assembly of the aircraft's flight test articles has demonstrated the promise
of such digitally driven assembly line precision. The mating of the No. 1
JSF flight test article's unitized wing to its center fuselage took four
machinists only 14 min. Its aft-to-center fuselage mating, which joined
components made by BAE Systems and LMAC, took just 16 min. In nondigital
manufacturing techniques, the join time would be measured in days--7-10 in
the case of the fuselage mate. 
Both the Lockheed Martin and Boeing teams proved they could develop a
common, stealthy fighter to meet multiple operational requirements. As the
JSF competition's winner, the LMAC-led team now faces what many consider an
even tougher challenge, holding to a budget that, in 1994 dollars, puts a
lid of $28-38 million on the price of each fighter, depending on operational
configuration. 
For it to succeed, JSF will have to make good on the efficiency promises of
lean manufacturing. Senior managers on the Lockheed Martin team say it will.
When metal cutting begins in the first or second quarter next year, the
company pledges that its advanced fabrication and assembly processes will
have reduced the JSF's tooling time by 90%; manufacturing time by 66%;
manufacturing costs by more than 50% and part counts by 50%. The "build
span" for an F-16, the time from specific order to delivery, is 15 months.
For JSF it will be five months. 
But JSF's performance-based contract puts a premium on life-cycle costs, so
it will take a lot more than state-of-the-art manufacturing processes to
meet the Pentagon's criteria. The Joint Strike Fighter is an evolving
product unlike any other fighter in U.S. history. Once the aircraft is in
the field, the same data used to design and assemble it will support it.
It's the prototype for build-it, support-it, evolve-it programs. 
Although the fighter's customers are auditing these processes, they are not
signing off on milestone reviews in a traditional manner. That's because the
JSF team, not service depots, will be responsible for the operational health
of every tail number for as long as it flies. As a result, bringing senior
noncommissioned maintenance officers into the design loop became as
important as seeking the advice of pilots and assembly-line workers. 
"JSF IS THE FLAGSHIP for a new way to do business" and will push the
company's use of digital technology, commented Mark Peden, LMAC's vice
president for information systems and technology. 
The product uses classified technology that must be distributed over a
global base. It is likely to have a life cycle of three or more decades.
That prompted LMAC to think about supplier and customer needs before it
looked for software to fit them. "Our customers have a great deal to say
about what they want," Peden said. "We didn't sit around saying, 'What's all
the cool software?' We evolved this from what our customer wanted." 
But the JSF's infotech architecture involves 14 basic systems (see p. 54).
Eleven provide the backbone for the digital design and manufacturing
processes. These allow the airframe team to operate as a virtual enterprise
with all-day, any-day real-time access to current file data. The
distribution includes one major partner and--so far--11 major subcontractors
in Europe, with all the attending government concerns about export controls.
Even a program as recent as Lockheed Martin's F-22 can't exchange all of its
data in real time. 
Enterprise resource planning software is the traditional heart of an
extended business operation. As it conducted a Year 2000 review of its
legacy software systems, the Fort Worth JSF team expected to follow a
company-wide effort to implement SAP's R/3 ERP system. But events and
institutional caution intervened. 
As part of a company-wide reorganization, Lockheed Martin was consolidating
its three aeronautics companies in Marietta, Ga., Palmdale, Calif., and Fort
Worth into a single operating unit. The challenge of making this merger work
while competing for the biggest, and perhaps last, manned fighter
competition in U.S. history were daunting. The thought of adding a third
huge task, implementation of SAP with its top-to-bottom remake of a
company's business processes, looked like too much to take on. So LMAC
didn't. 
Instead, the heart of the JSF architecture is SDRC's Metaphase product
life-cycle management (PLM) software. It gets high marks as a flexible, open
architecture that is amenable to handling distributed partnerships, J.S.
Gleeson, an LMAC senior engineering processes manager, said. Furthermore,
LMAC found that customization of Metaphase by Northrop Grumman's Logicon
improved its applications for meeting military specification requirements,
so the modified product was adopted by the JSF program, Gleeson explained. 
After acquiring SDRC and Unigraphics Solutions, EDS rolled Metaphase into
its Teamcenter suite and created PLM Solutions as a stand-alone company with
$1.2 billion in revenues. The JSF contract is PLM Solutions' third-largest
after Ford and General Motors and has the potential of passing them. "JSF is
really strategic for us," said PLM Solutions President Tony Affusio. The
company is now doing a pilot project at Tinker AFB, Okla., to help develop
wireless technology to support Teamcenter. The result could allow program
managers to send in change orders using secure codes over cellular phones. 
Northrop Grumman in El Segundo, Calif., and BAE Systems in Samlesbury, U.K.,
maintain their own mirror systems of the master PLM. The F119 engine from
Pratt &amp; Whitney (also a Metaphase customer) will be handled as
government-furnished equipment. 
Other key JSF systems include: 
*Telelogic's Dynamic Object Oriented Requirements System provides document
requirements for the aircraft's mission systems, handling work packages,
maintaining traceability and tracking tests and verification documents. 
*The key role of visualization for digital mockups to handle interference
detection, management and graphical-based work instructions is handled by
Visualization Assembly System software from Engineering Animation Inc., now
also part of EDS. 
*Parts management is handled by i2 Technologies/Aspect software. Engineers
search for parts in Aspect and the system keeps the Metaphase database
apprised of what parts are in stock. 
*MSC's MVision plays a similar role as Aspect for materials management. 
*The JSF program is using Catia Data Manager and Catlink, an in-house
product, as an interim interface between Catia and Metaphase. They are to be
replaced by CMI from T-Systems (formerly Debis Systemhaus) because it
supports distribution of data for JSF's customers. CMI also supports
electronic signoffs for International Traffic in Arms (ITAR) regulations.
However, Catia's V.5 Cadam software will require another look at the ITAR
issue. 
*Cost and scheduling requirements are handled by Artimis' Project View/Cost
View. This aspect is a key value for JSF's customers because it gives them a
single-glance view of where they stand on scheduling and budgeting issues. 
*Dassault's Delmia software provides the enterprise with simulation
software. 
*Several in-house systems from Lockheed Martin play key roles. Its
Manufacture Resource Planning II software keeps a master production schedule
and projects requirements for inventory ordering. Atlas analyzes how
products perform in the field. It's been used on the C-130 and F-22 programs
to perform mean time between failure rates. 
Lockheed Martin's Livelink-Open Text provides JSF with a Web-based digital
library. Continuously upgraded, future-user functions are to include
interfaces for accepting data from ruggedized hand-held mobile devices that
will support distribution of technical documents and drawings for line
maintenance operations. 
*Process planning and shopfloor management tasks are handled by HMS Software
Inc.'s SFM and CAPP products. 
*PTC's Windchill/Infoengine provides users an interface to access data from
other systems, giving a window on system costs.
BAE Systems-IFS Ltd. is providing IFS Applications for Web-based
distribution of inflight self-diagnosis by the aircraft. The system is
similar to civil applications in programs like the 777, but will require far
more extensive data management because of the complexities of military
aircraft, BAE Systems-IFS business development manager Wayne Starr said. The
JSF team is still defining what hand-held devices it will use, but they are
to include such features as barcode scanners to automatically record
changeouts of equipment on the flight line. 
Digital technology is at the heart of the affordability aspects of how JSF
will succeed or fail as a lifetime enterprise. Without it, the LMAC team
couldn't possibly meet its affordability targets. Its manufacturing and
support systems could not be as finely tuned, so they would require more
iteration and more tradeoffs, Mestad said. 
It's as if JSF is being held together by an elixir. That elixir is software.

------------------------ Yahoo! Groups Sponsor ---------------------~-->
Sponsored by VeriSign - The Value of Trust
Do you need to encrypt all your online transactions? Find
the perfect solution in this FREE Guide from VeriSign.
http://us.click.yahoo.com/jWSNbC/UdiDAA/yigFAA/kgFolB/TM
---------------------------------------------------------------------~->

------------------
http://all.net/ 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
Next message: Fred Cohen: "[iwar] [fc:More.Online.Security.Woes.For.FBI's.Data.Firm]"
Previous message: Fred Cohen: "[iwar] [fc:Killed.At.Their.Keyboards]"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
This archive was generated by hypermail 2.1.2 : 2002-12-31 02:15:03 PST