Many shops with CNC lathes have added such accessories as bar feeders to automate material loading. But what about at the other end of the process? This was the question that Mark Flanagan asked when his shop, Central Centerless Grinding (Saugus, Massachusetts) went to bid on a project to build piston rods for a hydraulic cylinder for Teleflex Canada Ltd. (Richmond, British Columbia).
Central specializes in high-end turning work, particularly in making shafts. The company uses a combination of screw machines, grinders and CNC equipment.
Teleflex presented Central with a proposed project calling for lathe work and secondary grinding on a 3/4-inch, 304 stainless steel shaft. Mr. Flanagan expected he would be able to use a bar feeder to accelerate the material input into a twin-spindle turning center. Still, the numbers didn't work out to meet the customer's budget for the project. The bottleneck appeared to lie with the sheer size of the finished part, which at 28 inches long was too large to fall into a parts catcher. "There is no conventional way of keeping the machine running with a shaft that long," said Mr. Flanagan. "All the efficiency gains you get from having a twin-spindle lathe are thrown out the window if you have to stop the machine each time you make a part and unload it."
The only way this would work, thought Mr. Flanagan, would be if he could figure out some way to keep the left spindle of the turning center working as the right spindle was being unloaded. Mr. Flanagan started sketching out ideas for a speedier way to get finished parts out of his CNC lathes. Eventually he hit upon the idea of trying vacuum power to pull the shaft out through the rear of the machine.
Using two-by-fours, PVC pipe and a household vacuum cleaner, the Central staff set up a crude experiment on their shop floor. Would vacuum power alone be enough to reliably pull a 28-inch shaft out of a lathe? It was. So easily, in fact, that Mr. Flanagan was willing to aggressively bid on the Teleflex project, and he was quickly awarded the contract. But now he was faced with the prospect of making his novel theory work on the shop floor.
To handle the added business from the new contract, Central purchased a Nakamura twin-spindle turning center equipped with a magazine-style bar feeder. While awaiting the machine's delivery and installation, Mr. Flanagan and his machinists began creating a vacuum unloading system. They wired it to the lathe's Fanuc CNC and added a control function that would allow the CNC to start and stop the vacuum pressure.
This is how the system works: After the right side spindle of the lathe has finished its cutting operation, the spindle is sent to its home position. A special spindle liner mates with the unloading machine. By using "M" functions, the unloader is activated and a vacuum is created in the spindle, forcing the workpiece out through the rear of the lathe where it is captured, then released onto a receiving table. The unloader is equipped with automatic part eject confirmation that sends the lathe an "all clear" signal so that the next cycle can begin. The entire process takes about 2 seconds.
Mr. Flanagan found that virtually any workpiece that will fit through the spindle could be unloaded. Another version of his machine has been used successfully on certain tubing applications. He also found that the vacuum extraction is ideal for small parts where it is not advisable to drop the part into a parts catcher—for example, a part with fine threading or delicate finishes. For another customer, Central ran a job with chrome-plated bar stock through the unloader without damaging the finish.
It didn't take long before other area production shops got wind of the new unloading system. Chuck Heshion of Brookfield Engineering Laboratories in Stoughton, Massachusetts saw the original "two-by-four" unloader and knew it would work on the lathes he uses to turn metal shafts used in viscosity measuring instruments. At the time, Brookfield needed to have full-time operators at each machine to pull completed parts out every few minutes, requiring the complete shut down of both sides of the machine for every unloading of the right spindle.
Mr. Heshion persuaded Mr. Flanagan to build a commercial version of the vacuum extraction system, and he placed the first order for his shop. With the unloader fitted to its machines, Brookfield's production efficiency gain was staggering. The company went from 85 parts per machine per shift to 225 parts per machine, and with the added bonus of sharply reduced labor costs.
Mr. Flanagan applied for multiple patents on his design and recently formed a new company, K & C Automation, to build vacuum extraction systems for other machine shops. Several installations have been completed in shops from Massachusetts to Ohio. He also was able to design the unloader with a small enough footprint to work on machines equipped with chip conveyors and high-pressure coolant pumps, and that require no more than standard shop air and the lathe's own internal low-voltage power supply.
"We've been through three different designs at this point," Mr. Flanagan said, "refining the system, running and testing in our own shop 20 hours a day, shoring up any weak points and making the unloader more flexible, adjustable." His unloader is now capable of working with most any twin-spindle CNC lathe sold today, and certainly has come a long way from a jury-rigged collection of PVC, two-by-fours, and an old vacuum cleaner.