After years of development, researchers have struck gold with a die that reconfigures itself to form a wide variety of parts.

Many of the parts that comprise large manufactured items begin as raw materials that are formed into shapes by dies, punches and form blocks. These forming tools make up a significant portion of the part’s initial production cost. It should come as no surprise, then, that engineers have been trying to create a reconfigurable tool die that could produce a variety of parts since the time that the first patents were issued. But until only recently, the realization of a truly efficient reconfigurable tool die has remained out of reach.

In the 1970s, David Hardt of the Massachusetts Institute of Technology (MIT) built a hydraulic press that configured itself by adjusting a bed of densely packed quarter-inch pins controlled by servomotors. Soon after, C. Robert Crowe of the Advanced Research Projects Agency (now known as DARPA) improved on the concept by developing a reconfigurable tool die that utilizes a surface of 2,688 adjustable pins. The pins, which have rounded ends, are 1.125 inches square by 21 inches long. Researchers at MIT set out to improve the die.

To begin with, even though the pin-ends that made up the die’s surface were rounded, the surface they created would not meet part requirements. In fact, the rounded ends threatened to leave dimples on the part surface. Through experimentation, however, the researchers found that certain polymers could be used to absorb the localized pressure of the pins so that the die created pieces with a smooth overall body shape. Through the process of trial and error, they were able to define parameters in which dimpling was unlikely and thus give themselves a guide by which to choose the best polymeric materials for each job. Further testing confirmed these predictions and today dimpling is no longer a problem with the reconfigurable die.

MIT’s researchers also faced the challenge of making the reconfigurable tool die durable enough to exist in the production shop environment, a place not known for its gentle atmosphere. This was no easy task considering the number of parts that the die contains, including its thousands of moving pins. In addition to bearing the loads of the materials it shapes, the tool also had to be rugged enough to withstand installation and being moved around the shop. Another consideration was that the repositioning of the pins had to be both fast and accurate. The researchers tried a number of different approaches to solve these problems. After foiled attempts to hold the pins in place hydraulically, both from the side and through individual cylinders, the researchers discovered that the pins could bear the weight of the loads if threaded rods supported each one individually. Tiny motors installed on the rods could also position the pins on a vertical basis. Each installed motor controlled eight pins. These eight-pin/one-motor modules were then outfitted with microprocessors and motor controllers. Because they’re installed and removed vertically, the modules can be easily reached for repair. With microprocessors carrying out the operating instructions, as well as relaying any error signals back to the host computer, all of the moveable pins can be controlled with relatively simple instructions and, most importantly, their reconfiguration time had been trimmed down to less than 12 minutes.

To correct tool shape, the reconfigurable die uses both iterative and predictive processes. The iterative process calculates the die shape based on a series of guesses, the predictive process uses finite element analysis to predict and accommodate the elasticity of the interpolation material. Since both processes require a certain degree of familiarity with complicated software, the researchers have developed a ‘user-friendly’ front end to be used by the die’s operators. The software, directed by the operator’s commands at a computer monitor, takes information about the part’s shape from a computer-aided design (CAD) program and prepares a finite element model of the tool, the material and the process. Finally, it submits the job to be prepared for analysis, circumventing unnecessary tool tryouts on the production floor.

The reconfigurable tool die has already been tested in two major aerospace facilities, where cost and benefit analyses has confirmed the dies as being well worth the investment. The long-term benefits of quicker turnarounds and lower assembly costs promise to increase the user’s savings. It’s expected that small-lot production shops will be the most immediate beneficiaries of the reconfigurable tool die.

Source: Tools of Change
John M. Papazian
Mechanical Engineering, Feb. 2002
http://www.memagazine.org/backissues/feb02/features/toolsof/toolsof.html