Future Tool

Created at the University of Michigan with funding from the National Science Foundation, the Reconfigurable Machine Tool was designed for the needs of the auto industry, where it has yet to catch on.

“In the 1990s,” says Dr. Zbigniew Pasek, operations manager and assistant research scientist at the Engineering Research Center for Reconfigurable Manufacturing Systems, College of Engineering, University of Michigan, “the important issue was making machine tool investments effective in a changing landscape. In short, accommodating fast product changes given a set initial investment in machinery.” Pasek and his colleagues recognized that there was a wild divergence between machine tool lifecycles and product lifecycles: the machines are in use much longer than the products they were developed to make are in production. This led to a National Science Foundation (NSF) grant to help fund the establishment of the Engineering Research Center…and eventually brought about the development of what is called the “Reconfigurable Machine Tool” (RMT).

The RMT would offer functionality that would place it somewhere between general-purpose and dedicated machines. It would have flexibility within applications. “The main premise is that if you can define the part family–generally, an application area that requires a second set of functions–then we offer something that is matched in value to the need,” says Pasek.

Cylinder heads, to name one example, may have major differences based on the included valve angle, number of cylinders, etc., while the basic configuration of each is quite similar. Designing an RMT to perform the machining operations on all members of that part family–even those cylinder heads that have yet to be designed–with equal quality requires modifying the structure in such a way that repeatability is not lost. “The idea is to, within a certain work footprint, change the angular values without having to use a different set of machine tools,” says Pasek. So U of M’s prototype RMT has the spindle on an arcing mount that covers an area from –15°º to +45º°, with hard stops every 15°º of travel. “The hard stops assure accuracy,” he says, “though we can place the spindle anywhere within that angular range. If you know where you are with the spindle–and that is defined very accurately at the hard stops–this value goes into the control system to provide the quality and repeatability required.” (An associated fast calibration project is underway to increase accuracy between the hard stops.)

There are three servo-controlled axes on the RMT: (1) the X-axis along the table, (2) the Y-axis along the column, and (3) the Z-axis along the axis of the spindle, with a manually reconfigurable adjustment of the spindle’s angular position. This combination, it’s claimed, provides an operating envelope with an effective three degrees of freedom that matches the functionality of conventional CNC machines that have orthogonal kinematics with up to five axes. “This machine is built along the lines of adjusting functions, but it isn’t modular,” says Pasek, who adds, “It’s not actually necessary to have modular solutions to have the functional adjustment you need. In fact, it’s probably easier if you don’t.”

The machine, which was shown at the 2002 IMTS show, garnered a lot of interest on the show floor, but created little sustained “buzz.” Pasek wasn’t surprised. The RMT was designed for use in the “ultra-conservative” auto industry where a five- to 10-year use must be guaranteed for any new shop floor technology. However, it also is being studied for possible use in the aerospace industry to tackle its multiple material machining needs. By swapping out spindle units to machine different exotic materials, Pasek and his group believe they can give aerospace manufacturers customized dynamic characteristics in a single machine, not the multiple machines used today. He hasn’t given up on the auto industry, but notes that most of the interest in the RMT has come from Asian machine tool companies and their customers. Only half kidding he talks about the patents U of M holds on the design then says, “Nevertheless, there’s probably a copy or variation out there already.”

X travel
Y travel
Z travel
2,890 x 2,825 mm
2,968 mm
5,000 kg.
1,016 mm
700 mm
500 mm
Maximum Work Piece Volume:
W x L x H600 x 800 x 500 m3
Maximum Speed
Spindle nos
Body diameter
Weiss 175190A
10 kW 
(2,640 to 10,000 rpm)
10,000 rpm
CAT 40
190 mm
Controls and Servo
Resolution 1 µm
Accuracy <10 µm
Controls and Servo Indramat