It all started in 1996 when Fanuc came out with a new control specification that was capable of transferring data as a geometry set rather than a bunch of points in the X, Y, Z space. By doing so, the door was opened to transferring NURBS (Non-Uniform Rational B-Splines) as NURBS, basically a free-form curve and not a linear set of data points. Which translated, for part makers everywhere, into parts made with better surface finishes in a shorter amount of time.

Of course, that spec was for three-axis machining, and the advent of high-speed five-axis machining prompted Fanuc to introduce a multi-axis specification a couple of years ago that can accommodate five axes and beyond. Advancements in this area are primarily joint efforts between software developers, machine tool builders and users.

“This is bleeding edge type stuff we are talking about here,” says Makino CAD/CAM Engineer Jeff Wallace. “Only a few are beta testing five-axis NURBS machining, but there are a lot of bugs to be worked out.”

Among the aerospace companies out there currently “working out the bugs” is Boeing, whose Seattle R&D facility has taken on the challenge.

What kind of bugs?

Before you can generate a multi-axis program, you have to be able to define the kinematics of the machine. Unless something strange and dramatic happens in the machine tool industry, machine kinematics differ from builder to builder—and sometimes even from machine to machine.

This means software makers have the dubious task of trying to create CAM programs with mathematical formulas. These will either take such kinematic variables into account or generate CAM programs that are specific to the machine tool builder.

The problem with the first option is that a software program that would attempt to take on all of the kinematic variables of every machine tool would be so large and so cumbersome that neither the post-processor nor the control could handle it. So, the second option is definitely much more feasible.

There are also the kinds of ticks and bugs that occur when you finally have a control, a machine and the software interacting that are anticipated but never defined until it is actually run all together. These occur on nearly every machine, no matter what software or hardware system is incorporated into it, which is why there is a testing cycle before a machine is delivered to a user. As the beta testing cycle on high-speed machining of five-axis NURBS progresses, the bugs and ticks will become more predictable and will be worked out by the hardware, software and machine tool makers—companies like Fanuc or Siemens, currently the only two control builders with five-axis or multi-axis capability; Unigraphics Solutions, typically one of the forerunners of CAM technology; and Makino, who is taking a proactive and aggressive role in developing this capability.

“The core, basic function is in the control,” says Wallace. “So the rest of us are chasing each other in harnessing the function and running with it.”

The way he sees it, at Makino, it is a matter of “how well we can fine tune the machine tool, and how well we understand the interaction between the machine, the control and the software.”

Benefits

Harnessing five-axis NURBS does not give users the ability to program and machine more complex parts than they can now. It probably will give them the ability to machine complex parts faster and with better surface finish.

A reduction in part program size is another benefit, which may simplify post-processing. Three-axis NURBS’ interpolation capability now potentially reduces part programs from two-thirds to one-half the size of linear interpolation part programs—depending upon the CAM system. The move to five-axis will not be that dramatic, but any reduction in program size and the ability to move data quickly is beneficial.

No Secrets?

Unlike other technology races, the pursuit of efficient, reliable use of five-axis NURBS is not a hotly contested, extremely secretive race. “There are no real secrets here,” says Wallace. “We are not really developing new technology so much as we are trying to perfect the use of existing tools.”

Those aerospace companies wanting to try five-axis NURBS for themselves are more than welcome to it. There are no ground rules, and the only requirements are having the right tools to make the attempt.

What you should have, according to Wallace, is the following:

• A CAM system that supports the Fanuc spec-- Keep an eye out for customization options that enable you to set up parameters for kinematics, part complexities and so on. A good example of this is Unigraphics V.17, which has a kinematics library users can refer to, as well as a database system for setting up and storing manufacturing resource data. Also make sure that the look-ahead feature is capable of seeing 100+ blocks into the future. That will ensure that acceleration and deceleration rates are anticipated and accounted for.
• A high-end, 64-bit controller platform--Fanuc and Siemens are the ones with the multi-axis/five-axis specs, so that’s where you need to go. More specifically, the Fanuc 16i- or 160i-series CNCs or the Siemens Sinumeric 810D or 840D lines CNCs have the capability.
• A high-end, five-axis machine with linear guideways rather than box-type guideways-- Linear guideways are faster and more agile than box-type, so you will get more mobility. At Makino, the A-Series horizontal machining centers and the V55-5XA vertical machining center all fit the bill.
• Good communication skills-- Once you embark upon your five-axis journey, you should find you will need to be in fairly constant contact with the makers of all three pieces of the pie (control, software, machine). Make sure everyone knows what you are up to from the get go, so they can all offer advice up front and will be ready to help out when called upon.
• Patience, and a good attitude-- “We’re finding that this is not as easy as we thought it was going to be,” says Wallace. “So you have to be prepared for a lot of trial and error.” He goes on to state that it’s important to keep five-axis NURBS in perspective. “You can’t look at it as the Holy Grail of high-speed machining, because it is simply not,” he says. “Look at it as another tool in your tool box. This is not something you’d need on every part you produce, so you need to keep reminding yourself that this is a capability, not a technological revolution.”

Copyright © 2001 - 2010 by Makino, Inc. All rights reserved.