Although the word “flexibility” trips off of the lips of seemingly every powertrain manufacturing person when the topic comes to the type of equipment they believe is important to have on their factory floors, the transfer line—yes, even the classic machining system that moves product-in-becoming through stations at a measured pace, day-in, day-out—still remains in action. Casting his eye back on what’s occurred during the past several years, Ron Quaile, vice president, Proposal & Estimating, MAG Powertrain (http://www.crosshuellerex-cell-olamb.com/), which consists of the companies Cross Hueller, Ex-Cell-O, and Lamb, all outfits with venerable histories vis-à-vis transfer line technology, acknowledges, “For decades, the only feasible, cost-effective way to make hundreds of thousands of large powertrain components per year was to build sequential-process transfer lines of dedicated, non-programmable machining stations, with part-handling automation that moved parts in ‘lock-step’ from one station to the next.” While he acknowledges that there was the ability to handle minor part variations within the stations, major changes necessitated major rework on the lines. Costly and time-consuming activities. This gave rise, he says, to two different types of systems and a blend of the two.
Specifically, there is the so-called “agile system,” that makes use of CNC machines and (optional) between-machine parts handling automation. There is the dedicated transfer machine. And there is a blend between the two, where CNC machines are integrated into transfer lines. Which type to select, he suggests, is predicated on the analysis of “total lifecycle costs” (TLC). Quaile suggests that while TLC are taken into account, “In general terms, initial capital cost is the first thing people look at. Lifecycle cost becomes a tie breaker. But as we see, it, it helps you make the decision about appropriate technology from day one.”
Quaile says that they’ve created a worksheet to help calculate TLC. Product changes have a significant effect on the costs. They’ve categorized the changes as “minor,” “moderate” and “major,” as in:
So this might mean that the agile system is the way to go, right? Not so fast. Quaile observes, “When the volumes are high, the initial acquisition cost of a transfer line system is less. When you get into the 600,000 parts-per-year range, it can be 40% less.” He also says that operating costs tend to be lower, too. But TLC isn’t just about acquisition costs. He points out that the conversion costs associated with the afore-described changes have a big effect, with major changes to the base product requiring an investment of as much as $20-million or more. Quaile provides an example: “If you’re looking at a cylinder head that is going to see many, many changes over its life cycle, the conversion costs can easily overwhelm the initial capital costs.”
There are other metrics that need to be included in TLC calculation. Quaile enumerates some of them:
So what is the overall picture? According to Quaile, they did an analysis predicated on:
Based on those assumptions, the TLC is lowest for a transfer line—$245-million—if the annual production volume is 600,000 units per year. (The TLC for the agile line is $300-million; it’s $297-million for a hybrid.) The agile line is the most cost-effective for volumes of 300,000 units per year and 400,000 units per year: $161-million and $180-million, respectively. The transfer line costs for those two volumes are $238-million and $242-million. Hybrid systems essentially fall between, at $193-million for a 300,000-unit-per-year system and $211-million for the 400,000 unit system.
|20-Year Lifecycle Costs|
|In $ Millions||Transfer||Agile||Hybrid|
|Acquisition Cost |
(including installation and facility modification)
(retool/banking cost for product changes)
*thousands of parts per year
This is a typical transfer line for a block application. Nowadays, multiple lines are complimented by CNC machines that can pick up features that are either (a) changeable during the life of the product or (b) difficult to machine otherwise. Speaking of the latter, Ron Quaile of MAG Powertrain says that while part designers are doing a good job of minimizing “stragglers”—such as “one hole that’s sitting by itself at some goofy angle that you can’t combine with anything else.” Which would make it expensive to do in a transfer line station. “It makes far more sense to take that operation and put it into a CNC machine with toolchanging and B-axis rotational capability.”
While CNC machines are flexible by design and standard in that they’re manufactured in multiples, one interesting aspect of deployment in production systems is just how standard other aspects of those systems are. Ron Quaile of MAG Powertrain notes that in cylinder head lines, for example, the workholding fixtures are the same for V6 and V8 models. There is an adapter plate bolted to the head, which is then picked up by the workholding system. So the machine is standard. The fixture is standard. The material handling device is standard. And all this in an agile line.
Not by a long shot. As Ron Quaile, vice president, Proposal & Estimating, MAG Powertrain, puts it, “As volumes get higher, transfer lines still have relevance.” While that may be assumed to be the case, there is something else to consider: “When you’re talking about cast iron or compacted graphite iron applications, when you have lower cutting speeds, the advantage or the competitiveness of CNC machines isn’t as significant.” Another aspect is precisely what feature is being machined. Take a deck face. The bolt holes are ideally suited for a transfer line, Quaile observes, because you can have a multi-spindle head performing generation of eight or more holes in a single cycle.
While transfer machines have long been considered to be “special” machines, Quaile suggests that is no longer the case, as there has been a large degree of standardization of everything from bases to axis units to the electrical systems.
Given more frequent changes to products as well as attention to how spending is being performed to achieve them, transfer lines are generally not as monolithic as they once were. While the overall architecture may be the same as it has been for some time, there are places where other types of equipment come into play. “Typically, somewhere in the process we’ll have banks of CNCs under an overhead gantry,” Quaile says, explaining, “The logic is the CNCs take up all the features that are uneconomical to do in a transfer machine”—think, for example, of piston cooling jets in a block, which would require multiple stations for drilling, reaming, tapping—“or subject to variation along the engine’s life”—such as NOx sensor locations, engine mounting attachments, transmission face variations.
Pat McGibbon, vice president-Strategic Information and Research, AMT-The Association for Manufacturing Technology (www.amtonline.org; McLean, VA), remembers a presentation that he’d heard given by financial analyst Eli Lustgarten back in the 1980s. Lustgarten, now the senior vice president of Longbow Research (www.longbowresearch.com; Cleveland, OH), drew a graph. McGibbon describes it as “a prophetic chart about machine tools.” The Y-axis was labeled “Flexibility.” The higher along the axis, the less flexibility. The X-axis was labeled “Output.” The further out, the greater the output. The upper right, McGibbon recalls, was the place where things like automobiles and washing machines were made: high volume, low flexibility. It was the place, he says, where transfer lines were. As you moved in toward the intersection of the two axes, the flexibility increased and the output decreased. The thing was, there were flexible systems available. “Engineering-wise, they were sound,” McGibbon says. But there was a problem. Managing the systems. Back in the early ‘80s, managers of large operations, such as those in automotive, couldn’t think in a way that wasn’t linear. Consequently, they couldn’t take advantage of the flexibility. Flexible manufacturing systems (FMS) pretty much ignominiously disappeared.
But the times, talent and technologies have changed. McGibbon says that they’re seeing more and more CNC-based systems being deployed by larger companies, especially Tier One suppliers, and lower-tier companies, which may find that their contract for a specific job may not last as long as they’d need to pay off a dedicated system. He says that whereas in the ‘80s machining centers were slow compared to transfer lines (he suggests that back then you could take the money necessary for a transfer line and buy machining centers with it and you wouldn’t have the capacity to make the output of the transfer line), they now have the speeds and quality capabilities that make them more competitive.
Another change that is occurring, he notes, are machines that can do so many different applications that “you can’t even classify them.” For example, a single machine can mill, turn, and grind. Or even perform EDM operations. While he acknowledges that the speeds of these machines may be slower, the reduction of both setup time and part handling time combined with higher quality (single setups vs. multiple can reduce part variation) make this equipment competitive, as well.
And then there’s the pace of change. “How many things do we make for four years at a time? That’s the issue.”