发信人: bigapple (红富士), 信区: NTM
标 题: STEP NC—The End Of G-Codes?
发信站: 哈工大紫丁香 (Thu Sep 28 17:34:19 2000) , 转信
STEP NC—The End Of G-Codes?
In the not-too-distant future, the only machine tool input you'll need is a We
b page.
By Mark Albert
------------------------------------------------------------------------------
--
STEP and STEP NC Links
A wealth of information is available from Web sites mainted by STEP Tools Inc.
, NIST, PDES Inc. and other standards forming organizations.
XML: The Latest iTechnology—Internet and Integration
This readable but clear introduction to this next-generation Web language has
links to a surprising number of useful articles and background material. Broug
ht to you by Automotive Manufacturing & Production magazine (www.autofieldguid
e.com).
Visit MMS' Software Zone
Link a Friend to This Page Via E-mail
Imagine this: You call up a Web browser on the PC-based CNC at your machine to
ol. You go to a certain Web site. From a menu on the home page, you select one
of the databases it accesses. A 3D image of a workpiece comes up. You click o
n an icon in the task bar and check a few parameters and default settings on a
pop-up window. Then it's a click on the CYCLE START button. The spindle motor
starts to whir, axes begin to move, coolant spurts out and chips are soon bou
ncing off the Lexan panels in the machine guarding.
According to efforts underway right now, it won't be long—a couple of years a
t most—before this scenario depicts how most shops will be running their mach
ine tools. NC part programs as we've known them for almost 50 years will becom
e passe. All that the machine tool controller will need is the digital product
model represented by the 3D image on the Web page.
The CNC won't use G-codes. Everything it has to know about how to move the cut
ting tool is in the product model's database. There will be no need for creati
ng a new and separate file of tool path data. Tool paths will be figured out i
n the CNC itself, based on the product model. That means there's no need for p
ost processors either. Data will be formatted for execution by the machine wit
hin the CNC. And because the product model won't change, it will be available
for machining "hard copies" whenever and wherever needed.
"Whenever" means as long as the product's life cycle is on-going. Twenty-five
years is a typical life span for aerospace parts, for example. Neither changes
in computer technology nor advances in machine tool technology over the years
would affect the usability of the product model as machine tool input.
"Wherever" means anywhere an adequately equipped shop has authorized access to
the product model database. With the Internet, that access is worldwide. Part
s could be machined anywhere in the world through a global supply chain, with
the digital product model serving as the universal "NC part program."
What will it take to make this dream come true? How much more has to be done t
o get there? How close are we right now?
The First STEP
The biggest step in this direction has already been taken. It's STEP, the STan
dard for the Exchange of Product model data, a comprehensive ISO standard (ISO
10303) that describes how to represent and exchange digital product informati
on. STEP replaces IGES as the means by which graphical information is shared a
mong unlike computer systems around the world. The big difference is that STEP
is designed so that virtually all essential information about a product, not
just CAD files, can be passed back and forth among users.
------------------------------------------------------------------------------
--
3D Model data is not used in manufacturing because even as simple supply chain
must rely on too many interface among unlike computer systems. Suppliers 1 to
4 are subcontractors and job shops. Suppliers A to D are software solution su
ppliers serving various applications.
------------------------------------------------------------------------------
--
The core of the standard is a library of engineering definitions that can be a
ssembled into various "application protocols" customized for the product model
s needed by particular industries and activities. A common library covers geom
etry, topology, tolerances, relationships, attributes, assemblies, configurati
on and other characteristics. New product models can be added as the need aris
es.
An extension to STEP has been created to cover product information related to
CNC machining. This is STEP NC. STEP NC forms the basis for the scenario that
caught your attention at the beginning of this article. The development effort
to make STEP NC product model data usable as direct machine tool input has al
ready progressed substantially.
In May 2000, a prototype for the sets of data required to add machining inform
ation to the product model of a test part was demonstrated. A later phase of t
his project will develop the machine tool controller capable of accepting this
"super model" as input. The current test part is a milled workpiece. Turning
and grinding are on the horizon. The "super model" demonstrated in May used an
emerging Internet language called XML to add information about machining stra
tegy, tool path planning, and tool selection. XML makes the resulting database
"Internet ready'—a key requirement for global e-manufacturing.
STEP Vs. IGES
To understand STEP NC and where it's headed, a look at STEP and its relationsh
ip to IGES is the place to begin. IGES was about exchanging data and only the
data contained in graphics files. STEP is about sharing data, allowing parties
to work together by communicating information interactively.
IGES had its start 20 years ago when designers and engineers were turning to c
omputers to create product designs. Instead of drawing lines and segment of ci
rcles on paper to make graphic representations of what a product should look l
ike, they started making those lines and arcs on a computer screen. The comple
ted design could be saved as a digital file. Although creating the original de
sign file might take longer than preparing the engineering drawing on paper, t
he design file could be quickly copied, modified, printed and otherwise manipu
lated. These time savings more than made up for the extra time it took to prep
are. Moreover, the digital nature of the design file allowed it to contain muc
h more information in a much more flexible format.
STEP allows a single database of 3D product information to be accessed by seve
ral users regardless of the different computer platforms involved. Where sever
al companies may be producing different parts for a complex assembly with crit
ical tolerances, having a single shared database greatly simplifies product da
ta managment and ensures that mating parts and subassembies can be manufacture
d effectively.
One big problem quickly emerged. The computer-aided design (CAD) systems used
to create these digital design files were not compatible with each other. A de
sign created on a Computervision system was meaningless to an Applicon system,
for example. Companies with unlike CAD systems could not exchange CAD data.
The effort to resolve this situation got underway in the spring of 1980. Repre
sentatives of U.S. user groups, vendors and standards organizations began meet
ing regularly to create a neutral, non-proprietary database structure and data
format for CAD files, dubbed the Initial Graphics Exchange Specification (IGE
S). In theory, CAD files translated into IGES could be exchanged with any CAD
system that could translate IGES files into its own proprietary format.
Although IGES eventually became a workable, if imperfect, approach to exchangi
ng CAD files, a major shortcoming with this approach became apparent right awa
y. IGES allowed one system to communicate the lines and symbols of a computeri
zed engineering drawing, but IGES failed to communicate the meaning of the inf
ormation the drawing was intended to convey. It did not provide a reliable mea
ns by which product features could be transmitted with the geometry so that co
mputer-based applications could "understand" the engineering drawing.
While IGES was being developed and gradually made more functional as it moved
through the standards formation process, efforts to develop a true "product da
ta exchange specification" were launched. The goal of this effort was to captu
re and convey "logical" information about product features and provide "physic
al" mechanisms for data exchange. Originally conceived as a U.S. initiative, t
his effort was soon seen as requiring international participation.
By 1984, this international effort to develop a Product Data Exchange Specific
ation had been established under the auspices of ISO, the international standa
rds making body. The goal was to define the methods for creating product data
models that could be interpreted by computers. These models were intended to a
llow the exchange and sharing of product data in a way that the meaning of the
data would not change throughout the product life cycle.
The international standards covering these product data models became known as
STEP. For the last 15 years, various groups and committees (mostly comprising
users rather than vendors) have been meeting regularly to develop standards f
or product data models. They have made considerable progress. Because the STEP
standards are now sufficiently developed to cover all of the original purpose
s of IGES, IGES will receive no further development and refinement. STEP has o
fficially taken its place.
By July 2000, every major and almost all minor CAD system vendors had STEP tra
nslators in the latest releases of their CAD products. Moreover, these transla
tors have been tested for conformance and interoperability. With only a few ex
ceptions did any of the translators fail to operate effectively. (Indeed, one
of the innovative features of the STEP formation process was the early commitm
ent to include testing procedures for assuring that STEP-compliant systems wou
ld truly function as intended. This provision may have slowed development but
it appears to have paid off in the end.) In short, STEP is working. According
to industry analysts, more than one million STEP enabled CAD stations are in p
lace around the world.
Super Model Steps Out Quickly
Although standards-making activities are notoriously protracted, the Super Mod
el project has maintained an intense pace from the outset, surprising and plea
sing members of the Industrial Review Board who attended a review board meetin
g in May.
Says IRB member Gary Hargreaves, director of sales at CNC Software, Inc., "It
is amazing to see the progress of the STEP Super Model project . . . going fro
m a vision and discussions in December 1999 to something real, as shown during
the May 2000 Industrial Review Board meeting just five months later. I now ha
ve tangible notes and information on how STEP NC technology can actually be im
plemented and supported for future CAM applications."
Another member, Bob Bean, president and CEO, CADKEY Corporation, had these com
ments after the meeting: "It is clear that with over 30 MCAD products availabl
e in the market today (and more start-ups entering the market every day), CAD
data sharing and interoperability is a major challenge both in the short term
and over the long term. When one looks at the stages of product and machine de
sign such as conceptual, initial design, analysis, outsourcing of manufacturin
g, changes due to manufacturability, mold design, tooling, and so on, it is ap
parent that 3D solid model information needs to be leveraged throughout the de
sign cycle. The STEP Super Model project is designed to facilitate not only th
e early phase of product development, but the full process to production."
Manfred "Whitey" Simon, a 30-year veteran of General Motors Advanced Manufactu
ring Engineering group was likewise encouraged: "The progress shown here is de
finitely in line with the scope of what is to be accomplished in a three year
period. There seems to be lots of attention given to the real user interface—
which is so important."
The IRB meets again in November 2000 at STEP Tools' headquarters in Troy, New
York.
STEP NC
Although STEP NC has not quite completed the formal standards making process,
it is well into the review and approval cycle, which will ultimately settle a
few remaining matters of technical detail. STEP NC defines data representing "
working steps," that is, a library of specific operations that might be perfor
med on a CNC machine tool. In keeping with the STEP concept, these working ste
ps are generic descriptions that can be incorporated into a product model. The
se descriptions are not linked to a specific format or code. However, STEP NC
working steps are roughly equivalent to the machining commands represented by
traditional M and G codes.
STEP NC is the basis, the enabling standard that underlies the potential for u
sing the digital product model as machine tool input. STEP NC allows a complet
e database of machining information to be built around it. The database, then,
dictates what capabilities must exist in the machine tool controller to cut t
he part.
What's called for is a "super model" that includes design information such as
geometry, manufacturing planning information such as form features (holes, slo
ts, contours and so on), plus manufacturing strategy information such as tool
selection, fixture location and so on.
The effort to develop the super model and make it usable as machine tool input
is being spearheaded by STEP Tools, Inc. of Troy, New York. STEP Tools is a d
eveloper of data exchange software for worldwide manufacturing. Although the o
fficial name of the project is the Model Driven Intelligent Control of Manufac
turing, participants are simply calling it the Super Model project. Funding is
coming from the National Institute of Standards and Technology (NIST), an age
ncy of the U.S. Commerce Department's Technology Administration. The Super Mod
el program was formally launched as an Advanced Technology Project with an awa
rd of $2.9 million in October 1999. Participating in the program is an Industr
ial Review Board consisting of manufacturers, software vendors and machine con
trol builders, government and defense agencies, and a range of small- and medi
um-sized job shops from the Hudson Valley in New York State.
The Super Model project has a three year time line. The target for the first y
ear is to build a STEP and STEP NC database containing three kinds of manufact
uring features, and use the database to drive a machine tool controller. The t
arget for the second year is to build a database containing all of the feature
s defined by the STEP NC milling schema and use that database to manufacture t
he STEP NC test part. The target for the third year is to produce a database f
or another machining process such as turning, grinding or electrical discharge
machining.
Dr. Martin Hardwick heads up STEP Tools, Inc., the Troy, New York-based softwa
re development company leading the Super Model project.
The Super Model Database
The challenge for the Super Model project is to create interfaces that bring t
ogether the information defined by STEP and STEP NC. Product geometry can be d
efined by one STEP application protocol. Product features can be defined by an
other STEP protocol. Machining operations can be defined by STEP NC. However,
all three types of data and others must be integrated in a complete product mo
del database. Moreover, this database must be Internet compatible.
Starting with product geometry in the STEP format is the easy part because STE
P translators are built into most CAD systems these days (and they handle 3D g
eometry, doing so more effectively than IGES ever did, apparently). The super
model test part happened to be created in a ProEngineering workstation.
The next step in building the database is adding features to the geometry. For
the sake of demonstration, STEP Tools originally used a Microsoft Excel sprea
dsheet to link STEP-defined feature names to the test part geometry. The Super
Model program is evaluating an automatic feature recognition system being dev
eloped by Honeywell FM & T, one of the subcontractors in the program. Called t
he FBMach Process Planning System (FBMach is short for feature-based machining
), this software reads STEP geometry and automatically determines what feature
s, such as holes, pockets, slots, and so on, are represented by the geometry.
The user interface allows these determinations to be validated before proceedi
ng. The FBMach system is expected to be available commercially by the end of 2
000. Its application to the super model test part will be demonstrated in Nove
mber.
STEP NC establishes a hierarchy of workingstep supertypes/subtypes. In other w
ords, it breaks down every machining operation into the steps required to perf
orm the operation. These steps include actions to be taken as well as data (su
ch coordinates of point-to-point motion) to be applied. These steps are then l
inked to the appropriate part model geometry to fill in the values. STEP Tools
is setting up tables to match workingsteps, workingstep-methods, workingstep
actions, and machined features.
According to Martin Hardwick, president of STEP Tools, the super model databas
e is adapting a modified version of the company's ST Repository product data m
anagement software to structure the database. Each repository uses standard in
terfaces to import and export geometry, features and workingsteps to the tools
used by CAD and CAM engineers.
A key part of STEP Tools approach to the super model database is the use of XM
L in its interfaces. XML, the eXtensible Markup Language, is a vendor-neutral
data exchange language for passing information, not just data, across the Inte
rnet. XML allows data to be "tagged" so that software applications reading the
database can identify what type of information is stored in the database and
extract the data that is needed. HTML, the Hyper Text Markup Language, is a si
milar"metadata" language that the Web uses so that text can be displayed no ma
tter what Internet browser happens to read it. XML offers a comparable level o
f interoperability. An XML standard for STEP is nearing completion. This stand
ard will ensure that all data in a product model is "tagged" in the same way.
For the super model, XML provides a convenient means to link manufacturing str
ategy, tool pathing, and tool selection information to geometry, features and
machining steps in the database. By sorting out data with the appropriate tags
, for example, geometry identified as a hole to be drilled can be linked to op
erations such as rough drilling, boring and counterboring steps. Each of these
steps will require that other data be extracted, such as workpiece material,
surface finish requirements, and so on, to link with speed and feed tables. XM
L provides the tags so that the data is sorted correctly.
Ultimately, XML ensures that a CNC networked to the Internet will be able to f
ind the information it needs from the product model database to machine a part
.
At the May 2000 Industrial Review Board meeting, STEP Tools demonstrated how X
ML transactions had been used to complete a database for its test part, linkin
g the required information for three different types of machined features, inc
luding a hole, a slot and a pocket.
Down To The CNC
The goal for the Super Model project is to show how these features can be cut
on a machining center using the product model as the NC part program, so to sp
eak. One of the subcontractors deeply involved is this phase of the program is
Electro-Mechanical Integrators, Inc. (EMI) of Franconia, Pennsylvania. Engine
ers at EMI are writing new software for a Bridgeport control unit that will en
able it to accept STEP NC data. (The company has considerable experience with
Bridgeport control units and is the factory-authorized support and repair agen
cy for the Bridgeport DX32 control.)
A Bridgeport V3XT milling machine like this one is being outfitted with a CNC
customized with special software. This software will enable it to interpet the
super model database and use the information to machine the part without a co
nventional G-code program.
According to Bart Stater, head programmer of EMI, this effort requires a custo
mized command parser and command interpolator to process STEP NC. "Essentially
we are creating a new CNC protocol to interpret the information in the produc
t model in real time. This software will extract the data it needs to determin
e axis moves, get the specified tool, and issue commands. It will not need or
use G-codes," he says. Otherwise, the I/O structure and servo system of the ma
chine remain the same. He notes that this concept assumes that the CNC will be
networked to a file server that receives and stores data, most likely through
an Internet connection.
EMI is looking at two approaches to configuration of the CNC. One approach run
s all of the executive software in the CNC's internal processors. Another appr
oach uses a "PC front end" interfaced to the CNC. The PC would process the STE
P NC data and spoon feed it to the CNC, filling a buffer with blocks of data o
n demand. This approach would ensure that the CNC is not starved for data whil
e the product model is processed.
In November 2000, EMI is scheduled to demonstrate actual cutting of the three
selected workpiece features on the test part. Although the concept could be pr
oven with a simulation of the machining operations, cutting chips is a more co
nvincing demonstration. "We want to show the CNC accessing the product databas
e on the Web, finding the features to be machined, then generating commands to
drill the hole, mill the slot and machine the pocket," Mr. Stater insists.
STEP Tools is also working with the Lawrence Livermore National Laboratories,
where a STEP NC interface is being developed for the OMAC (Open Modular Archit
ecture Control) project. Additional shopfloor tests and demonstrations of STEP
NC are set to take place at a production machining facility operated by Gener
al Dynamics Land Systems in Scranton, Pennsylvania. A pilot project at the Jet
Propulsion Laboratory is also in the proposal phase awaiting funds.
Art To Part
Those three words sum up the promise of STEP NC and the Super Model project.
From a shop floor viewpoint, art to part means the intermediate steps of creat
ing an NC program are eliminated. Most of those intermediate steps necessitate
d a transformation of product data, causing data files to proliferate. Part ge
ometry had to be translated, reconstructed or edited. The edited, translated o
r reconstructed geometry had to be processed to generate tool paths. Tool path
files had to be post processed to suit the requirements of the machine tool a
nd control unit combination. Postprocessed files were often edited on the shop
floor. In short, one piece of part geometry begot hundreds by the time the pa
rt was actually cut from metal.
With STEP and STEP NC, the digital product model database replaces all of the
other product data files otherwise created to make the part.
From a design and engineering viewpoint, art to part means that design and man
ufacturing can be managed with a single database. Just as data files need not
proliferate down the supply chain, they need not proliferate across the manufa
cturing organization. Product data can be shared between products, between cor
porate divisions and between applications. The Internet will make this sharing
of data global and virtually instantaneous.
This concept of art to part does make G-code programming obsolete. But this tr
aditional form of programming for machine tools was already on its way out. Ad
vances in CAM software make G-codes less and less visible to programmers and m
achine operators.
This concept also implies that CAD and CAM will have a different relationship
than they did in the past. Dr. Hardwick believes that product models will orig
inate in CAD, with STEP enabling a high degree of collaboration between design
ers and engineers. Feature recognition will be applied at this level as manufa
cturing engineers define the manufacturing process that becomes part of the pr
oduct model. "At this point, the process data will be ready for any machine to
ol but will allow for important local parameters to be defined when it gets to
the machine," he predicts. Selecting cutting tool, setting feeds and speeds a
nd so on will be handled at the machine tool on the shop floor.
------------------------------------------------------------------------------
--
The super model solution seeks to replace hundreds of product data managment f
iles with one database from which data may be shared via the Internet. The tra
nslators, editors, converters and postprocessors require to exchange or reform
ate product files are unnecessary.
------------------------------------------------------------------------------
--
These CAM functions become the domain of intelligent controllers with on-board
CAM software. This software will generate the movements necessary to make the
parts after the appropriate parameters have been set on the CNC. "The on-boar
d CAM software is there to do the last minute custom tool path generation usin
g selections made by the operator," Dr. Hardwick says. Intelligence built into
the software stops the operator from making mistakes or using less than optim
um settings. "It's a huge opportunity for the CAM industry. This software will
be a required component of all future CNCs," Dr. Hardwick contends.
Vision
With the development of STEP NC, what's happening is not simply the re-shaping
of CNC. It is the reshapi
--
※ 来源:.哈工大紫丁香WWW bbs.hit.edu.cn. [FROM: 203.123.8.2]
Powered by KBS BBS 2.0 (http://dev.kcn.cn)
页面执行时间:213.518毫秒