Source: Advanced Manufacturing Blog
The road to manufacturing success today runs through the mountain of data that tools are generating in metal-cutting applications, and most importantly communicating and reacting to in real time on the shop floor.
Keys to collecting data include sensor technology that is built into tools and tool networks, software that can seamlessly collect, organize and analyze data, and machine tools that can be networked into a source of factory intelligence. The elements that lead to effective “smart” tool applications can be intimidating unless you engage with companies that can sort out exactly what is needed, what should be avoided, and what else is coming down the pike.
Market Demand Catches Up with Technology
Technology to successfully collect, distribute, and analyze data for adaptive control of machining processes has been available long before the current interest in things Industry 4.0 and IIoT. For some time, attendees at major trade shows have had the opportunity to witness cells put together by Caron Engineering Inc. (Wells, ME) featuring the key factors for successful adaptive machining through data collection and analysis. “Sophisticated shops are really beginning to accept the idea that the more you know about the machining process, the better off your production, quality and efficiency are going to be,” said Rob Caron, president.
The emergence of the MTConnect standard, the emphasis on digital connectivity and better and less expensive sensor technology are making it easier to justify the additional investment required to gather data about every aspect of the machining process. “At current customers, for example, we usually start with RFID tool presetting off-line in the toolcrib,” he said. “All parameters are embedded in the RFID tag and transferred to the machine control, where we monitor all the tools through power, vibration, strain and other parameters. We do some adaptive cutting, modify the feed rate to maintain constant power, and measure the part with electronic devices before sending the data upstream for analysis.”
The reality, Caron added, is that everyone wants the magic bullet that will tell them everything that’s going on in their process. “In that regard, everyone is working on better sensor technology—and better analytics through self-learning and artificial intelligence in software,” he said. Key considerations for sensor development include higher resolution for vibration, power, and coolant flow for high-pressure systems, and better technology for characterizing coolant concentration.
“Most of our products are at the machine level,” Caron pointed out, “where they collect a huge amount of data.” These products include DTect-IT, a Windows application that communicates with Caron Engineering USB sensors to monitor any area of concern on the machine tool or fixture, using custom sensors for vibration, strain and analog applications.
Caron’s Tool Monitoring Adaptive Control (TMAC) system protects CNC machines by reducing the high costs associated with broken tools, lost production and rejected parts by measuring tool wear in real time. The “adaptive” control feature of TMAC reduces cycle time and optimizes cutting time under changing conditions to improve tool life.
Tool Connect automates the transfer of tool presetter data from RFID tags in toolholders to a machine control. AutoComp software uses any electronic gaging device to provide the dimensional measurements and processes gage data to update tool offsets automatically, providing error-free tool offset control.
How Good Can Data Connected Machining Be?
For Wolfram Manufacturing, an Austin, TX-based contract manufacturer, Caron Engineering’s products worked so effectively in the shop that company president Nathan Byman decided to become a distributor for Caron products. Wolfram Manufacturing specializes in medium to high-volume production, manufacturing parts from start to finish, including material sourcing, machining, specialty coating, and specialized packaging.
Wolfram produces parts up to 24″ (609 mm) in diameter, 60″ (1524 mm) in length and weighing 500 lb (226 kg). Because accuracy is a key part of machining, typical tolerances are down to ±0.001″ (0.03 mm) with special features down to ±0.0001″ (0.003 mm).
Byman had experience with Caron Engineering’s TMAC system in large contract machine shops he ran before starting Wolfram, which uses highly flexible machining cells in its production. “At Wolfram, every machine we buy has probing, high-pressure coolant, and is a multifunction mill/turn—machines that are optimized using the TMAC system. TMAC introduces feedback to our processes which lets us automate and made them repeatable.” Wolfram runs Okuma machine tools. “We like the Okuma support structure and the THINC control allows us to get a lot of information in and out of the control, a capability that is especially important for data-generated adaptive machining,” Byman added.
“Today’s sensor technology is better than ever,” said Byman. “The trend in cost of sensors, vibration sensors for example, is coming down and their resolution is getting higher. These systems might add 5–10% to the cost of a mid-size machine tool, or 3–5% for a larger machine, but they can unlock productivity improvement of 20–40%,” he said.
“Our customers for Caron’s products are typically OEMs, who have a deep focus on their machining processes and a heavy commitment to manufacturing engineering. As a result they are more likely to be prepared for what Caron products have to offer. For those that are less sophisticated, education is the first step,” said Byman. “Our approach in dealing with any customer is to recommend working with them to bring their first parts on line and showing them a path to success. We find this works best since most organizations need some help to absorb the thought process and power these tools provide.” To do that, Wolfram has a full engineering capability.
Byman believes that continued sensor development will produce even better resolution, horsepower transducer range, and better response time. “On the software side, TMAC already has a lot going on under the hood in terms of processing algorithms, but there is still a lot to be gained by leveraging the data for trend evaluation and visualization. Right now you program knowledge into the system. It will mature into a place where you are taking knowledge out of the system for use in other places,” he explained.