Essentials of a successful CAD-CAM system.
Every Kearney & Trecker product is individual and made to order. "Right now everybody involved in the process makes their own set of drawings," says Warner. "A proposal drawing is a proposal drawing. An engineer gets a copy of it and takes it over to his board to use as a guide. He may even lay a scale over it, but he's better off starting from scratch."
This starting-over process continues throughout the sequence from design to machine-capacity drawings to assembly drawings, according to Warner. Then the assembly drawings have to be broken apart into detail drawings. These go to manufacturing, which writes NC programs from them.
"There's a lot of guesswork, with someone saying, 'Well, this is about an inch and a half and it's 1-48 scale, so he must have meant so many feet from here to there,"' says Warner. "We're hoping to change all that." The company recently installed several CAD/CAM workstations. Warner expects that if all goes smoothly, he'll see vast changes in the complexion of the engineer's job.
For Kearney & Trecker, savings are expected in time, efficiency, and the elimination of guesswork. "With the data base in the system, a person at one stage should know just what his predecessors had in mind," says Warner. "He can go back and ask an earlier contributor if, for example, there is something magic about a particular specification--be it 5 ft or 5"--or whether a change can be made."
Why is a company, whose stock in trade is the edge of technology only now seeking to apply the benefits of an integrated CAD/CAM system to its own work? The reality is that despite wishful thinking, there is no one system that will take care of all of their CAD/CAM needs. And companies such as Kearney & Trecker have investments in expensive, CAD-oriented equiment to consider as well.
This company, like other manufacturers, needs a system that does not seek to displace systems already in place, but rather seeks to integrate the entire workplace both systemically and electronically. This type of system has only recently become available.
A truly integrated CAD/CAM system, one that takes the concept from the design stage to the factory floor, means a highly specialized combination of software and hardware created with the needs of the manufacturing operation in mind. The basic concepts
To understand why this has been slow in coming, a look at the concepts of CAD, CAM, and CAD/CAM is in order. CAD, or computer-aided design, usually refers to a screen-oriented system that allows the user to manipulate images. It is particularly useful for examining two- or three-dimensional objects from different perspectives, rotating them in space, creating complicated designs with precision, and performing other visually oriented operations. In a sense, CAD is a tool for product designers and draftsmen.
The traditional CAD companies entered the factory environment via the electronics industry. Here, the designer's finished artwork actually becomes the tooling that creates a product. In the case of the printed circuit board, the artwork is a photoplot used to create the conductors on the board. In the case of large-scale integrated circuits, the artwork translates into drivers for the development of the masks that create the product. Since the design and the manufacturing processes are already somewhat integrated, CAD/CAM implementation in electronics was very successful.
the story was not so simple in the world of mechanical manufacturing, though. Here the design is only half the story. The function of the drawing in the manufacturing world is not to serve as an end in itself, but as a method of making ideas clearer to others. It is also a vehicle for working out problems.
But true CAM, or computer-aided manufacturing, implies that the computer will actually drive the machine tools that mill or otherwise carry our manufacturing procedures. This requires attention to certain constraints that the product designer doesn't worry about, such as contiguous geometry, nonredundant geometry, and the elimination of construction lines. Need for crisp data
The information driving a machine tool needs to be crisp and contiguous, so that the tool can cut the part. This is where many CAD-oriented systems fall down. Although the information they create is accurate, the continuity necessary for manufacturing just isn't there. The design data base is not the manufacturing data base; therefore NC programs derived from the design data base are insufficient for use in manufacturing applications.
Historically, software has also fallen down in the area of process integration. CAD/CAM software needs to provide a link between the design and the manufacturing process; the output from one application has to serve as input to the other. It is a myth that all one needs is the design data base to serve a manufacturing process. A sense of what is needed is precisely what has been in short supply in most supposed CAD/CAM systems.
There were hardware as well as software factors working against the introduction of CAD/CAM to the manufacturing environment. That's because computers first arrived for business applications as large, central units shared by many users. Such shared environments have traditionally meant slow response times; and the more users, the slower the response.
People used to working with tools--either on the design or the manufacturing side--are accustomed to a quick response from those tools. They expect that same response from computerized equipment, and studies by IBM and others have shown that a response time of more than a fraction of a second can actually detract from the performance of the person using the system.
Even worse than the lag is the consistency factor. If the same procedure that brings a response in 1/2 sec at one point means a 4- or 5-sec wait a few minutes later, the user's work rhythm is broken, and he becomes frustrated. Such experiences do not make for successful CAD/CAM installations.
As if the time-sharing problems weren't enough, computers also arrived at the manufacturing workplace speaking strange languages through intimidating keyboards. To the engineer, who is hardly likely to be a typist, this meant the system had even more built-in roadblocks. In the fast-moving, production-oriented environment, there simply wasn't time to deal with unfamiliar procedures. Benefits of CAM
But if these factors were working against CAD/CAM, they weren't changing the need for its power in the manufacturing world. Good CAM has many benefits from the machinists's point of view. Among these are:
* Increased accuracy.
* Increased repeatability (1000 parts, for example, can all be created alike).
* A decrease in the time it takes to go from a part print to the finished product
* A decrease in scrap due to better utilization of raw materials.
* A shorter production time frame.
* Increased machine flexibility.
An integrated system, with the ability to access the manufacturing data base electronically, would mean that setup time could be minimized, and that an individual machine could run more than one type of part without elaborate resetting. This is the machining-center concept.
At the same time, the data base for any given part could be used by any machine. Such interchangeability would mean eliminating the need to wait to see which machine was available before setup could be worked out. In other words, time could be saved.
A successful CAD/CAM system must deal with the historical drawbacks to bring out these potential benefits. Recent innovations are making this possible.
On the hardware side, the large address space microprocessor and the local-area network, along with a general lowering of costs for electronic memory, have allowed the desk-top engineering workstation to evolve. This provides an effective answer to the response-time problem.
The microprocessor has brought the computing power of former large computers to the desk of an engineer. The local-area network, which connects workstations to each other and to special-purpose mainframes, enables the sharing of data when necessary. At the same time, though, each user can work independently. Thus the actions of one user do not affect those of others.
Hardware innovations also mean that a CAD/CAM system can be--as it must be--easy to use. The idea of CAD/CAM is to make life easier, not more complicated, in the manufacturing environment.
Usually, for an engineer to want to use a system, the benefits of the system must be readily apparent. A computer input device known as a mouse, for example, allows the user to interact with the computer by simply rolling a small plastic device over a table or desk and pushing one, two or three buttons. This input mode makes a keyboard or other complicated input device unnecessary for most operations.
Combined with the proper software, the mouse can be used to interact with familiar engineering symbols and terminology that appear on the screen. The user then merely pushes a button to select the procedure he wants to use.
Today, it is possible for a good CAD/CAM system to duplicate existing factory methods. When the system incorporates elements that the worker will recognize, the worker becomes more efficient, and management spends less time on training and programming. Layering helpful
One element of the software mix should be a layered software approach that allows the user to access--or his company to purchase--only the particular level of software he needs for a job. For example, if he needs to create a 2-D image, he shouldn't have to deal with the 3-D software. Or, if he needs 2-1/2 axis NC programming, he shouldn't have to worry about the sculptured surface segments of the software package.
The direct electronic connection from the drawing or blueprint stage to the NC programmer is another software link that is crucial to an effective mechanical CAD/CAM system. This link has been slow in coming, however.
The need for the link is obvious. In the past, a design engineer would finish a blueprint, and the NC programmer would have to take that blueprint and redefine the geometry before he could start programming. With an integrated system, the programmer can at least start with something in digital form. (As every manufacturing engineer knows, his data base is much larger than the product data base.)
The digital form also gives the programmer more flexibility. For example, if he is creating a family of turned parts on numberically controlled lathes with an integrated system, he would not have to reprogram for each family member. The machining sequences needed to define the shapes would be stored and could be manipulated as needed. This is called family-of-parts programming, and it is where the integration of CAD and CAM really begins to pay off.
A crucial link in this process is the generation of the tool path through the generation of code that actually drives the machine tool. This is called the postprocessing stage; often the postprocessor that will generate codes for a given machine-tool/control combination must be written to customer specifications. Hence, the CAD/CAM vendor must understand machine tools, controls, good machining practice, and the people and conventions of manufacturing.
An effective CAD/CAM system provides a link between mechaniced design, manufacturing engineering functions, and the factory floor. This should be true no matter what variety of mills, lathes, punch presses, and other machine tools make up the factory, and regardless of whether these tools are operated by direct electronic link or a system-generated tape.
The link should also allow for a two-way flow of information, as feedback from the manufacturing process to the designer can yield shorter design cycles, more accurate designs, and more cost-effective methods of manufacturing. Costs are down
Cost is, of course, an important factor in management decisions for CAD/CAM application. Even the most useful system can't exceed its own profit potential without being questioned. And undoubtedly the high cost of CAD/CAM systems has traditionally been a barrier to their use.
With costs per console hour running $25 to $50, plus such additional costs as climate control for sensitive components, managers felt they needed to justify the system not only on what it could do but by how many people it could replace. A system in the $8 to $10 per console hour range, on the other hand, can be cost justified for what it is: a tool for existing personnel. Such low-cost systems are now available.
At the same time, software vendors that required the buyer to purchase a whole package--even if the package included more capacity than was needed--also discouraged potential users. This was especially true if the overcapacity also improved overly complicated procedures that had to be dealt with for relatively simple tasks.
Finally, there is the question of longterm benefits and investment protection. A successful system in today's marketplace has to protect a company's investments, both in the skills of its people and equipment itself. The system isn't intended to replace manufacturing skills, but to transfer some of them into a computer system so that life can be easier for everybody.
With a good system, the specialist can share the tricks of the trade he has developed over the years by writing macros that allow less experienced users to take advantage of his (the more experienced user's) skills. This gives freedom from repetitive tasks, yet at the same time those tasks remain closely supervised.
Take the person who must deal with many groove slots. He should be able to create a program to ask a series of questions that will define the slots: how long they are, what the end radii are, and so forth. He can thus remain in control but devote a minimum of time and energy to the project. Compatibility needed
The successful system should also incorporate software links capable of allowing various systems to communicate with each other directly. This is crucial to the integratin of a new system with other systems the company has previously obtained.
If the company is buying the new system to take advantage of state-of-the-art integration, for example, it must be able to utilize the output of, say, the expensive CAD system that is already in place. This capability can also allow direct communication with customers' equipment, thus affording a savings in time, greater efficiency, and a closer relationship with the client.
On the hardware side, it is important that the system be tough enough to function where it is needed. In many shops the programmer's office is on the shop floor. The equipment should therefore be able to withstand the temperature fluctuations and dustry, oily atmospheres that are commonplace in factories.
White lack of attention to any of the above points is a potential barrier to CAD/CAM implementation, general resistance to change cannot be ignored either. That resistance is not without reason, and perhaps the greatest barrier to CAD/CAM has come because the ultimate users of the system--the manufacturing engineers--have rarely been considered in system design.
The organization and work methods of the mechanical engineer have evolved over many years into sensible and efficienct patterns. If they are to become an integral part of the manufacturing operation, new hardware and software must reflect existing work methods.
The most effective CAD/CAM system will be one that can be placed gently over the existing company organization without distrubing it. In a sense, the real world is a local-area network, and the CAD/CAM system can mimic its organization. A successful implementation will encourage widespread communication between existing islands of technology, and allow relationships and essential procedures to remain intact. The implementation will not involve a draw-out learning situation, and it will not disturb the flow of product out the door.
Further, successful implementation will combine technically excellent CAD and CAM capabilities with the ability to communicate with other computer systems in the organization. Specifically, the system will include the various application software and hardware elements a user needs for a given organization.
The local-area network will allow the creation of a more complete data base. This will be a distributed data base built up through the various nodes on the network, not imposed through a centralized computer system. Help, not hassle
In a sense, the successful system should electronically duplicate the way the engineer works. Effective software provides electronic assistance, not hassle. And the level of complexity should rise only with the complexity of the task.
For a company such as Kearney & Trecker, which builds highly customized systems based on modular components, CAD/CAM implementation promises efficiency and increased creativity. Engineers work with a client to create a system configuration. Throughout the process, from design to manufacturing, a data base--founded on the combined experience of the entire organization--can be drawn upon.
Through such a system, both people and machines can be put to their most profitable use.
|Printer friendly Cite/link Email Feedback|
|Publication:||Tooling & Production|
|Date:||May 1, 1984|
|Previous Article:||Mass-finishing fundamentals.|
|Next Article:||New system cuts costs of plasma cutting.|