Almost everyone who has worked with Chrysler as a supplier during the past few years has known all about the automaker's commitment to the digital domain. If you are one of its suppliers, then it is pretty much a sure thing that you are utilizing CATIA software from Dassault Systemes. For those of you who aren't a supplier to the firm, know that generally speaking, companies supplying DaimlerChrysler (DCX) with physical objects—from suspension components to storage racks—tend to be obligatory users of the CAD/CAE/CAM software tools. (Note: Prior to the merger, Daimler was also a user of CATIA, so DCX is, in this regard, more of the same.)
DCX is taking the implementation of CATIA throughout its supply base to a whole new level, as supplier companies including SHG Inc., Albert Kahn Associates, and BEI Associates—all of Detroit—have all become comparatively new users of the software, in their case, CATIA/CADAM Plant (CCPlant). What's interesting to note about these companies is that they are not suppliers of parts, subassemblies or modules, but are architectural and engineering firms that are involved in the development of facilities including the Jeep plant in Toledo, OH (vehicle production is slated to begin in 2001), and the V-6 engine plant being built adjacent to the new V-8 Mack Engine plant in Detroit.
|Imagine the complexity of trying to discern what is what if this plant structure was rendered in 2D. Which is why DaimlerChrysler is now having the modeling of its factories by its architectural firms done with CATIA in 3D.|
It should also be noted that all of these firms are familiar with computer-aided design. However, the packages that are commonly used by their designers and engineers are usually sourced from companies including Intergraph, AutoDesk, and Bentley Systems. But generally, these architectural and engineering firms have been working in the 2D world, and with CCPlant they've added another dimension.
The importance of going to 3D is that interferences of things—be they pipes, trusses, exhausts, conduits, light fixtures, etc. (plants are, to say the very least, rather complicated buildings)—can be discovered before the construction begins.
Gerald F. Philo of SHG comments, for example, that during work on the Toledo plant design, an interference was discovered related to a process exhaust stack that "would cost as much as $10,000 to $20,000 to change in reality." And this is just one example of several. Digital fixes can translate into hundreds of thousands of dollars in savings—easily.
Catching Problems Early.
Haden Inc. (Auburn Hills, MI), which works with DCX on its paint systems, had designed a 410-ft. E-coat oven system for the Windsor, Ontario, assembly plant with a 2D system then, as a training exercise, modeled the same thing in 3D CATIA. Haden engineers determined that through the use of the 3D system they were able to reduce interferences by 90%. Haden, not surprisingly, is using CCPlant for the development of the paint shop for the Toledo plant.
|One difficulty in designing factories is discovering interferences—Which somtimes aren't discovered until the facility is built. Which is when making changes cost a lot in both time and money. CCPlant software is making interference checks considerably easier.|
According to Daniel J. VandenBossche, senior manager, Manufacturing Technical Support, DaimlerChrysler (DCX), the people who are doing advanced manufacturing engineering at DCX are most encouraged by the implementation of CATIA by the architectural firms that are designing the vehicle manufacturer's facilities. "We have an opportunity here of bringing everything together," he comments, adding, "Everything is going to have to be integrated—the building, the process equipment, and even human beings."
With regard to processing equipment, VandenBossche points out that DCX's major tooling and equipment suppliers—such as Haden, PICO, DCT, Webb—are all using CATIA. The implications of having the digital models of equipment combined with a digital model of a factory are wide ranging. He remarks, for example, "I don't have to worry about vehicles going down the assembly line with their doors open, crashing into columns because I will have simulated that line already." In other words, in the past, it was necessary to actually build a line before there was certainty that doors wouldn't smack into columns. With three dimensional models, there is an understanding of what is likely to happen and the ability to make required modifications before any metal is put on the floor. (In fact, the floor is likely not to exist in a physical form when this work is done.)
What's more, VandenBossche says that beyond doing a digital tryout, having the full digital package—the building, the equipment, the vehicle—allows an array of what-ifs: "We can do studies of putting different vehicles in plants. We are able to determine whether we have clearances for a particular body to go through the body shop, the paint shop, and trim/chassis/final if we want to cross load our facilities."
It's not that they weren't trying to work on these things previously, for they had. But VandenBossche points out that it isn't a simple thing. For example, he explains that one aspect that is often overlooked when doing a 2D analysis of a car going through a plant on a conveyor is that the vehicle's rear end can swing out when it turns a corner. Having a digital model that includes the articulation of the vehicle is vitally important to determining precisely where the conveyor lines will go through a factory.
What VandenBossche wants to have developed is what can be thought of as an entire "library" of components, digital models of the various elements that go into a vehicle assembly plant. This would allow faster modeling of processes. While some people might argue that there are simulation packages available that have simple representations of many pieces of factory equipment, VandenBossche explains, "A model is only as good as the degree of accuracy. If you just model things as boxes and squares, that's not going to be very faithful, and you are not going to get out of it what you want. Our intent is to model things as faithfully as we can and to get as much detail as is practical so we can get a virtual representation of that process.
"That's the whole name of the game here: virtual manufacturing. Prove it out before you go to metal."
And they are also proving it out before they're putting flesh and blood to work along the line. VandenBossche's colleagues are working with a human simulation package developed by Deneb Robotics that is allowing them to improve the ergonomics of assembly operations.
Doing work early in the digital domain prevents problems down the line—literally and figuratively. the contractors start building the structures, the automaker is able to save money precisely the same way that money is saved in vehicle programs, both in terms of reduced late engineering changes and with regard to program timing. Indeed, Bradley J. Harding of Albert Kahn comments, "With CCPlant, it's like cars: we're able to design more quickly."
The architectural firms have all found that they are able to take significant amounts of time out of their projects compared to previous practices.
What's more, they are able to create models that have more complete information. Joel A. Rosenbalm, CADD coordinator with BEI Associates, notes, "When we draw something with CATIA, everything is linked." Not only does this mean that the models are more "intelligent" in that the ramifications or modifications can be easily discerned, but the architectural firms are also in a position of creating entire bills of materials, something they ordinarily don't do.
Although these are still early days of using CATIA for plant design, G.R. Shashaai, manager, Facilities & Construction, Assembly Manufacturing Engineering, DCX, says, "I'm really encouraged with the results." Given that, one can only imagine that just as for the companies supplying DCX with the stampings and the weldments, the plastic moldings and the interior trim, CATIA will become the status quo for the rest of the supply base.