by Kip Hanson
Adaptive machining technologies are changing the way we make parts
When I started in the shop, machine tools were dumber than fence rails. Load the paper tape, adjust the rheostats for feedrate and spindle speed, click, click, click the offset thumbwheels, push cycle start and pray. Programs were keyed into a Teletype machine; when changes were needed, the printout was redlined and sent back to the programming office for laborious corrections. CAD/CAM systems and the mainframes to run them cost more than a new house. Probes were something they sent into space, and the only thing high speed was the steel we used to cut parts.
What a difference a few decades make. Today, the combination of complex software, inline probing systems, and increasingly intelligent machine controls give machine tools uncanny decision-making abilities, anything from stopping the machine a split second before one of us error-prone humans crashes it, to automatically adjusting cutting tools and part programs based on the in-process inspection information.
You say tomato
Perhaps the best known of these technologies is in-machine probing, although Dafydd Williams, general manager of Renishaw (Canada) Ltd., Mississauga, ON, hesitates to call it adaptive. “In process measurement using a touch probe or inline CMM is more along the lines of process control,” Williams says. “It’s about checking tool wear and managing temperature fluctuations that may affect part size. Adaptive machining is when the actual part program is automatically modified based on in-process measurement.”
In Renishaw’s definition of adaptive machining, a touch probe is used to validate part location and condition, and the resultant information is sent to some “really deep and very clever software.” This software analyzes part features to determine whether the NC program must be adjusted or “morphed” in to bring part dimensions into spec, or to compensate for certain material conditions. “Let’s say that a customer has an expensive casting, and they need to discover what state the workpiece is in before machining,” Williams explains. “For example, is there too much material, or is the casting warped or oriented incorrectly?. Those arequite complicated problems to deal with on a machine tool.”
That’s where NC-PerfectPart comes in. Metrology Software Products Ltd. (MSP), an associate company of Renishaw located in Northumberland, England, has developed a software program that assesses current part conditions based on data collected via a probing system, and then rotates or shifts the machine coordinate system to suit.
Technical director Peter Hammond explains NC-PerfectPart is used in a variety of manufacturing situations, including mould refurbishment, compensation for part distortion due to clamping forces, and weld machining on remanufactured components.
Aerospace is just one industry with its eye on adaptive machining. MSP and Renishaw have supported BAE Systems and other Tier I aerospace suppliers on a variety of setup problems such as automating the alignment of Airbus A330 ribs prior to machining, scrap reduction on expensive titanium components, turbine blade analysis and repair, and cutting setup time on difficult composite parts. In one instance, MSP helped a firm save more than over half a million dollars on workholding costs, simply because the software is able to “find” the part when loosely fixtured, thus making more expensive workholding unnecessary.
Twist and shout
MSP is currently working with Pratt and Whitney Canada in Montreal on automation of two of its aerospace component production lines. “One example where this technology is helpful is when parts are placed on the machine tool. It’s often difficult to know whether the distortion is real, say a wing spar that is twisted due to internal stress, or if issues exist with part location. Our system is used to probe the workpiece, and then apply a complex fitting algorithm to the data in order to
realign it. We can also generate what’s called a ’condition of supply’ report prior to machining. This helps determine how the part should be programmed, or if it should be machined at all,” Hammond explains.
Someone with a slightly different spin on adaptive machining is Mark Sully, account manager for Delcam North America, Windso, ON, who says there are two distinct approaches to adaptive machining. “It comes down to a basic question: are you aligning the part to the program, or aligning the program to the part?”
In the first case, something that Delcam defines as adaptive fixturing, inspection data, is collected in process and used to determine the best alignment between the part to be machined and the CNC code. The NC code is then automatically aligned to the actual part, thus ensuring maximum accuracy while simplifying the setup process.
In the second example, inspection data is collected from the part prior to or even during machining. “Take a part that’s gone out for heat treating,” Sully explains. “It comes back warped, and the program must be adjusted accordingly. In this case, we’d bolt the part down, probe it, and then do the same best-fit analysis as in adaptive fixturing. But instead of adjusting the coordinate rotation or shifting part location, we send the inspection results to a CAD/CAM system, in our case PowerMILL. Based on the probe results, the NC code is adapted to the new machining conditions. This might mean introduction of a semi-finishing routine, or additional passes with a roughing cutter.”
Like MSP and Renishaw, Delcam sees huge interest in adaptive machining from the aerospace industry, as well as the automotive market. “Many of these customers are working on larger parts, such as moulds and wing spars. In many cases, the challenge is that once they come off the machine, they’re sent to a CMM for inspection. Any problems such as warpage or twist might not be identified until much later, and this can negatively impact production schedules by days, if not weeks. Adaptive machining allows the part program to be modified based on immediate inspection results directly at the machine tool, in some cases eliminating the need for CMM inspection reports.”
Good as new
Sully cites one company performing jet engine repair. Turbine blades deemed “end of life” are scanned and compared to their original CAD model; any areas of wear or damage are identified and digitally marked. This information is fed to a welding robot, which fills in the affected area with weld material. The part is then rescanned, an NC program generated based on those results, and the part is re-machined to original specifications.
Adaptive machining is more than software and touch probes. Machine and control builders continue to improve the intelligence level of their wares: Fanuc offers iAdapt, which, according to the company, automatically maximizes material removal during roughing, thus reducing cycle times by up to 40 per cent. Okuma controls have smart functions such as Tool Posture Compensation and Super-NURBS to more accurately control tool positioning and feedrates during interpolation. DMG MORI has spindle growth sensing, S-Quad smart scanning for in-machine 3D geometry measurement, and application tuning cycles that compensate for varying table loads. The list goes on.
GF Machining Solutions, Lincolnshire, IL, is another builder offering smart machine solutions. “There are a ton of control features available that monitor cutting conditions and verify everything is correct during the machining process,” says product manager Eric Ostini. “Our APS function (Advanced Process System), for example, senses and adjusts for machining vibration, and PFP (Power Fail Protection) performs a controlled stop in the event of a power loss, retracting the tool in the Z-axis in order to prevent damage to the workpiece, cutting tool or machine.”
Down to the wire
One of the more advanced of these adaptive technologies is GF Machining Solutions’ Integrated Vision Unit (IVU Advance), currently available on the company’s wire EDM equipment. According to Ostini, IVU is an optical system that measures the part profile after cutting, and compares the results to the part’s DXF file. “It checks to see if everything is within tolerance, and then rewrites the G-code to compensate for any areas that fall outside spec. It adapts the program so that the next part comes out perfect.”
One final example of adaptive technology comes from Irvine, CA, toolpath simulation software provider CGTech Inc. Marketing communications manager Bryan Jacobs says the company has partnered with Makino and other machine builders to bring setup information from its Vericut software directly to the machine control. “By transferring the virtual fixtures, workpiece and tooling information contained in the toolpath verification file, we can provide the ability to avoid collisions,” Jacobs says. “Makino, for example, has used this approach to develop a proprietary system that actually prevents crashes due to operator mistakes.”
Jacobs explains that on a machine equipped with collision avoidance, there’s no risk of an operator inadvertently jogging a tool into the workpiece, or crashing the spindle into a fixture. Because the machine knows where each clamp, vise and cutting tool sits in 3D space, and what the workpiece looks like both before and after machining, it’s able to look ahead and say, “Uh-oh, we’re going to run into something. Let’s stop the machine.”
Adapt and conquer
It’s clear there are a number of ways to define adaptive machining. To some in the machining world, it’s as simple as a CNC sensing part chatter and autonomously tweaking a programmed feedrate to avoid it, or monitoring spindle and axial growth to improve part accuracy and reduce the time needed for machine warm up. Some consider automatic rotation of a workpiece coordinate system as adaptive, while others say it refers to dynamic G-code adjustments based on in-process inspection results. Whichever camp you belong to, no one would argue that machine tools are getting a little smarter and a lot more capable every year. SMT
Kip Hanson is a contributing editor.