Gregory Hyatt is vice president of engineering and chief technical officer for DMG/Mori Seiki USA Inc. (Hoffman Estates, IL), a developer and builder of machine tools and machine controls

Manufacturing Engineering: How critical is software to machine tool users today?

Gregory Hyatt: I would say it's very critical, but it's rather application and market-dependent. Software becomes more critical as the parts, geometries, and toolpaths become more complex. The answer would be somewhat different from a customer in medical or aerospace than it would be from, perhaps, automotive or contract machining. Aerospace would probably be most dependent on it, because not only do they have workpieces of complex geometry, they use alloys that are very difficult to machine, so cutting conditions are particularly sensitive to the analysis that's performed in advance of cutting. And, of course, there are safety-critical parts.

ME: What are some key trends in machine tool software?

Hyatt: Five-axis programming is key, especially simultaneous machining in of five axes or more as opposed to "3 plus 2". There's a lot of complexity involved in programming for multitasking systems, and it's difficult to find good programmers. You would think you could advertise for an applications engineer and get 50 people with experience these days, with the unemployment rate, but it's hard to find someone with good millturn experience. Companies are holding onto them, because there are very few. While conversational programming has been available on machines for years, and that's been a wonderful solution for two-axis lathes, three-axis mills, and four-axis lathes, it is not a solution for five-axis simultaneous machining of complex parts. So we're seeing a step change in the demand for on-machine software, as we move from the simpler machine structures to full five-axis. And the simpler conversational packages that have been popular for some time simply can't support the complex applications. As people move toward more complex machining, they need software that is more user-friendly and more intuitive, to address the shortage of skilled personnel. Once upon a time, people talked about automation and software to eliminate jobs. Now, in many cases, they'd be willing to create the jobs if they could find personnel to fill them. In some cases, the only opportunity for growth is with more sophisticated software that reduces dependence on skill sets that simply are not available in the market.

ME: Describe how your company's Digital Technology Laboratory (DTL) subsidiary enhances development of future machine tools.

Hyatt: DTL really has two main focuses: one is precision engineering and mechanical-based engineering, and the other is based more on software and automation. The precision engineering focus involves looking at technology down the line, at where machine tools are going in terms of accuracy requirements and so on. This group also deals with putting productive solutions on machines, such as thermal-compensation algorithms and other things like the chatter avoidance on the control, which we're working on right now. Our automation group also works on solutions, such as monitoring tool breakage using a camera. Monitoring cameras have been available, but we're looking at automatic detection of a catastrophic tool breakage, or when a tool has a nest of chips wound up around it. If you're running lights-out, you want to have redundant tooling, so if a small-diameter drill breaks or a tap breaks, you can just switch to the redundant tooling without having to stop the line, or ruin the thousand parts that come off during the night while you're away. We're looking at applications that keep machines running, that schedule machines and track material coming into your facility, as well as machine self-maintenance and other things. We're trying to make machines more productive in an automated environment.

ME: Your company's upcoming new MAPPS IV control will have embedded Esprit CAM from DP Technology (Camarillo, CA). How does this help users?

Hyatt: The integration of Esprit addresses the previous shortcoming for complex multiaxis machines, in terms of conversational programming. Now there's a full-blown multiaxis system in the control for programming, editing, and simulating the toolpath of more complex components. But in addition to that, the MAPPS IV control has other features that are also very powerful, independent of the integration of Esprit, such as more sophisticated collision-avoidance software, to self-catch errors in workpiece setup or tooling setup, or errors in toolpath or NC programs. If you're doing on-the-floor programming of these complex parts, it's critical to have a means to check for interferences, to know that the code is good, and that you're not going to spontaneously make an error that's going to crash the machine. The collision-avoidance collision-check was developed at DTL some years ago. We have a team that continually works on it, and works on adding different functions and additional Gcode functionality, and definitely on speeding it up, because as the movement of the axes gets more complicated, the CPU usage gets greedier and greedier. MAPPS IV improves our computing power; the CPU speed itself is a big factor.

ME: How have complex multiaxis machines made advanced simulation capabilities crucial?

Hyatt: As you get more degrees of freedom in simultaneous-axis machining, it becomes more and more difficult for an individual to visualize what's going on. Without simulation, there's an increasing possibility, as the machines become more complex, that the toolpath is not doing what the programmer intended. It becomes especially important to run the simulation, confirm that there's no interference with the workpiece, the workholding, or the machine tool, whereas that could be done almost intuitively with simpler machines. I've talked to people who take great pride that they have scraped paint off parts of their machine, they're running the tolerances so close. I'm sure a lot of people don't like to come within an inch, but when you're doing really complex parts, you're less than a millimeter away from colliding many times. There are so many moving parts—two turning spindles, a milling spindle, a turret—that the chances of a collision are probably an order of magnitude higher than on a regular machine or a three-axis. In the future, for complex multiaxis machines, the simulation capability that's going to be needed for the next step is going to be adding the actual machine dynamics into the simulation itself. Right now, the simulation is based on the commanded path, and the software, the graphics on the screen, indicates that the toolpath goes exactly where you command it. In reality, we know that doesn't happen, right? If you're machining an impeller, for example, when you round over the nose of it, you have to accelerate the rotary axes really fast. The major control manufacturers are pursuing solutions that allow more sophisticated compensation for those five-axis errors. One of these solutions, TCP functionality, is already out.

With the TCP smoothing functionality, we can get beautiful workpieces out of dirty code. Some of the more primitive five-axis codes, which would have resulted in surface-finish blemishes in the past, we can smooth with TCP functionality, and we also get faster cycle time. We've seen applications where we can take 30% out of the cycle time of extremely complex parts with that smoothing functionality. Of course, it could have been done on the CAM side; the most sophisticated CAM solutions, with the most perfect distribution of points, don't need that smoothing to the extent that more primitive systems do. The TCP smoothing was introduced in our MAPPS III control, and it is now standard with all of our five-axis machines, where originally it was an option.

ME: How does the current business climate look for manufacturing?

Hyatt: There's huge variability. Obviously, automotive and other sectors are very publicly struggling; other areas are doing rather well. The medical industry has not been hurt anywhere close to the extent that automotive and some of others have been. Aerospace production is down, so our customers are not typically buying machines to increase the production of widgets, but there are many development programs that are proceeding unabated, on both the civilian and the military side. There's an awful lot of work being done to launch these new platforms, both structure and engine platforms, and we're staying pretty busy on the aerospace side. Things have clearly picked up since the first of the year.there's absolutely no doubt about that. They're looking up; I'm not sure there's going to be a rapid recovery, it may take some time to get back to where we were, but at least the direction is right.

Keywords: Five-axis,