Shop Solutions: CAM Puts Design Shop on Route to Success
After graduating from Montana State University (MSU; Bozeman) and applying his degree in mechanical engineering technology to four years’ employment at Jungst Scientific and Blackhawk Industries, Matt McCune started his design, development, and prototype business with a “once in a lifetime” opportunity, a huge project contracted by a large company. Working with SolidWorks CAD software in his living room, he quickly encountered his first hurdle, making parts. Because other prototyping methods were incapable of making intricate working parts with close tolerances, machining would be required.
Machine shops in the area balked, pronouncing the project impossible, or quoting it as a prohibitively expensive experiment without fixed price or guarantee of success. McCune found himself in a time squeeze, which was exacerbated by his inability to persuade the customer to purchase a mill, but alleviated by the customer’s agreement to purchase parts at a fixed price. Leveraging his home equity, McCune purchased a Haas Super Mini Mill for immediate delivery and hired an aspiring machinist, engineering interns, and an accountant.
With a mill in his garage, and four people working in his house, his team began a search for CAM software, contacting vendors, and spending a full week evaluating three packages. While one was considered the most popular, McCune was highly interested in another because, being embedded, it was touted as being fully integrated with his CAD system. The third, GibbsCAM, had been Jungst’s standard for 20 years. The single goal, he emphasized to vendors, was making a very difficult part and nothing else. “Put your best programmer on it. We’ll cut the part. If your software does what we want, we’ll buy it,” he told them.
All three vendors wanted the business, but two tried to coach the group by telephone. According to McCune their guidance amounted to “You can do it, if you know how.” McCune explains: “In contrast, the GibbsCAM Reseller, took our urgency seriously. He lives in another, distant part of the country, but he came, and sat with us in my living room, to program the part while showing us what the software could do. He spent a few days with us.”
Vendor effort aside, the full integration of the first package tempted McCune, who quickly found the number of mouse clicks required to get anything done to be excessive. “The integration was more visual than functional,” he says. “The other packages weren’t embedded, but they didn’t require a lot of steps or remodeling the part. I was attracted by GibbsCAM’s clean interface and the few mouse clicks it required. It was efficient, people seemed to pick up on it quickly, and our GibbsCAM Reseller and Gibbs demonstrated terrific support, helping us to figure out this part.”
McCune describes the part as fitting within a 0.750" (19-mm) cube, extremely complex, with deep channels and irregularly shaped pockets throughout, some intersecting at complex angles, and most requiring surface machining. In production, such a part would be die cast, but as a prototype, it required machining, because other prototyping methods couldn’t produce the part with the critical tolerances and strength required.
For example, he finds that for all the excitement generated by 3-D printing, it isn’t ready to replace machining for functional prototypes. “It has its place, but has limited ability to provide a part that can be tested and manufactured,” he says. “The industry has done a disservice to education, leading students to believe that, with a CAD model and 3-D printing, they can manufacture. When I see students of design engineering proudly display a printed part or assembly, I often think, ‘That’s only 10% done, because manufacturability, material and functioning part haven’t been considered.’”
Working with tiny end mills, machining began. “We were dumb,” says McCune, “but that proved advantageous. Not knowing any better, we did things you just don’t do, things sane people won’t do, because we would spend a whole day replacing broken cutters.”
With small tools, some of 0.025" (0.64-mm) diameter extending to nearly 1" (25 mm), they broke a lot of tools, adjusted parameters, broke more tools, made more adjustments, but learned a lot. “Finally, there we were, holding a part for the customer, with a complete and successful machining process we could manipulate, change, and reprogram,” McCune says. “We had learned how to make a part that experienced machine shops said couldn’t be machined.”
Throughout the trial, McCune became impressed with GibbsCAM. Whenever different parameters, tool motion, machining pattern, entries, or exits were needed, GibbsCAM easily accommodated them. “We put GibbsCAM to the test, the machine to the test, the machinists to the test. Everything was tested, and we delivered the part.”
With a proven process, the parts went through dozens of revisions, and the group prototyped hundreds of parts, without the pain of many broken tools, and with a profit. Success and confidence led to more clients and more equipment. In 2006, the group was incorporated as Autopilot, specializing in product design and development. In 2007, having outgrown all available space in McCune’s shed, garage, and living room, Autopilot purchased a building. Within 18 months, the company had 15 employees, three Super Mini Mills, a TL-1 lathe, and SL-10 turning center from Haas Automation (Oxnard, CA), plus necessary support equipment and three seats of SolidWorks and GibbsCAM.
According to McCune, SolidWorks, GibbsCAM and CNCs enabled the company to grow like no other prototyping methods could have, allowing Autopilot to take projects from mere concept or napkin sketches, through design, development, and prototype, and even some production work to keep machines busy.
Today, Autopilot has over 200 clients, from very large companies to individual inventors, and is contractually unable to disclose most of their work. A project they can discuss, to a limited degree, is a family of drill bits they are taking through various iterations for a medical client, who tests them between iterations. Machinist and NC programmer Tim Robison turns the round features on 17-4 stainless steel blanks, before parts are sent out for cannulation. Then, on a Super Mini Mill and 5C indexer, he machines three flats on one end, flips the parts, and machines the flutes to completion in four-axis. Deburring, polishing, passivation, and engraving complete secondary operations.
To pursue medical prototyping and production, an interest generated from serving medical clients, Autopilot is very close to achieving ISO 13485 certification. Although not part of McCune’s original plan, the economy drove Autopilot into production as a cushion for survival, with the company once doing a part run of 250,000 pieces.
One project, a combination of prototype and production, allowed the creation of a local company, Bozeman Reel, when, relying on local fishing guides and fly fishing enthusiasts for consulting and testing, Autopilot took on the design and prototype of three lines of fishing reels, which Autopilot also machines and assembles.
McCune says that Autopilot’s goal is to become a world-class design, development and manufacturing company, in Bozeman, and believes the experience and confidence gained by employees provide solid footing to get there, while providing employment for people who share a sense of adventure, design, and manufacturing, with time, of course, to enjoy the great Montana outdoors. To continue growth, and do more production, he believes Autopilot will need CNCs with larger work envelopes, and perhaps a multispindle, multiaxis machine. “It will be interesting to see how GibbsCAM will get those programmed.” ME
For more information from Gibbs and Associates, go to www.gibbscam.com, or phone 805-523-0004.
Grinding Precision Sets Stage for Growth
Delta Gear Inc. (Livonia, MI) acquired TIFCO Gage & Gear, a maker since 1964 of master gears, spline gages, aerospace gears and automotive prototype gears, in 2004. Subsequently in 2005, Delta Gear joined forces with Delta Research Corp., which started out in 1952 as a supplier of prototype transmission and engine components for the Big Three automotive companies. Later, Delta Research went into aerospace work, designing engine test stands and other precision components for aerospace and defense. Today, the two companies employ more than 110.
“The industries each company addresses are entirely different,” says Tony Werschky, Delta partner, “and we are set up this way, between automotive, aerospace and defense, so that our businesses maintain a certain amount of stability regardless of market conditions. At Research, it’s pretty much automotive, truck, and defense work, while Delta Gear is nearly all aerospace, jet engines and rocket components.
“Bob Sakuta, my father-in-law, and son of the initial owner, Alex, came to work here in 1976 and has been here ever since,” says Werschky. “He was running machines out on the floor back then. Now he oversees the companies and manages the growth strategies, while Scott Sakuta and I are responsible for tactics.”
Recently, Delta Gear purchased an old building and turned it into a state-of-the-art, 72,000 ft2 (6689-m2) facility completely devoted to gear production, gear inspection, and metrology. Many of the materials ground are exotic, including super-hardened materials that are very difficult to machine. All the work done on gears and shafts is done in-house, other than heat treatment and specialty coatings. It depends on the complexity of the part, but we may send a part out a number of times for selective heat treating of certain features and plating,” notes Werschky.
“The typical gear, we turn it, rough it, send it out for heat treatment and then finish it. Some parts, however, require more than just heat treating. They may require copper plating or silver plating at different times, and that makes for a longer process or progression from start to finish. We really like the more difficult jobs because there’s more opportunity for failure, which limits the amount of competition that can make that part. There is more risk involved, of course. High risk, high reward. If you’re good, then you do well. If you’re the very best, then you do very well,” Werschky says.
“Our growth for Delta Gear will be in expanding our relationships with OEMs in the industry on precision gears in jet engines and other aircraft components. There has been considerable outsourcing of OEM components recently which has benefited us in our growth. The way Delta Gear is positioned right now, we’re in a good place to handle more growth. So we feel we can handle a significant amount of expansion with the OEMs in the aerospace industry,” says Werschky.
"Most of what we make stays in the US. We have a balancing act in keeping the right amount of engineering staff on hand for parts that are being given to us by our customers. For the size of our company, bringing in a new major OEM every year or two is a very good way to continue our growth organically,” Werschky explains.
“We consider ourselves to be the leader in precision grinding technologies,” says Werschky. “We aren’t the largest, nor are we the smallest. We are a precision gear facility and not only are our precision gears important, but the journals and faces that ride next to them are in many ways just as important if not more important. When tolerances are so tight, and the expectations of aerospace customers are so demanding, one can’t afford to take chances or cut corners on diameters that have to be 50 millionths in roundness.”
Werschky points out the obvious: “Our facility looks like a showroom for Studer. There is a stable of six Studers from United Grinding Technologies [Miamisburg, OH] in the new facility.”
The S40 was purchased in 2009 and is credited with opening some doors to new clients due to its versatility. “It has a Y axis that allows us to install an ID spindle in a vertical orientation so that we can grind slots and keyways. It’s very important for grinding our most complex parts. It also has C axis for out-of-round grinding. This is essential for grinding cams and lobes, applications in which we really couldn’t compete in the past because we didn’t have the capability. We are able to grind these down to within 0.0002" [0.005 mm] of a degree.
“The C axis also provides thread grinding. We used to outsource that application because there’s so much thread grinding in aerospace. Now we do all of that in-house. It can also do spline grinding as well, but we don’t really use it due to our wide assortment of dedicated gear and spline grinding machines. The turret wheelhead can be swiveled automatically and up to four grinding wheels can be used. The S40 also has a high-resolution B axis,” Werschky points out.
“The master gears that we make are made out of tool steel,” Werschky adds, “so these parts are very hard and they require tight-tolerance cylindrical grinding. When you have diameters that have to be held very tightly, the Studer is exemplary in its ability to meet and to exceed the requirements and our expectations of, for example, 10 millionths in roundness, 50 millionths in flatness, and one or two tenths typical runout.”
Some parts that Delta Gear produces require a super finish to under two microinches. The Studers are capable of grinding down to four microinches. A light polish afterward gets the surface down to under two microinches. So the Studers really help Delta optimize its processes to eliminate time and more rapidly get parts to their destination.
“The materials we grind are nearly all exotics with a hardness of over Rc 60, some over Rc 70. These are very brittle and susceptible to burning and cracking, so we need machines that are not only capable and reliable but also extremely accurate and repeatable. We use rotary diamond dressing tools, allowing us to use different grinding wheels. This helps us with hardened materials, over Rc 60, and it also aids in surface finish control,” Werschky notes.
“Further, we do a lot of shaft grinding in both automotive and aerospace. In automotive, it’s common for these shafts to have multiple diameters. The Studers are very capable of handling multiple diameters in one operation. We do a shaft where the largest diameter might be 6" [152 mm], but most are 3–4" [76–101 mm] in diameter. In aerospace, they’ll go down to 0.500" [12.7 mm] in diameter depending upon which shaft we’re talking about. In automotive shafts, the concentricity might be 0.001–0.002" [0.03–0.05 mm] , and in aerospace it might be one-quarter to one-half of that, depending on the length. If it’s a longer part, you’re looking at 0.002" [0.05 mm] concentricity over 24" [610 mm]. In aerospace, 0.002" over 10" [254 mm]. Typically there are three different diameters per shaft.” ME
For more information on United Grinding Technologies, phone 937-847-1253, or go to www.grinding.com; from Delta Gear Inc. phone 734-525-8000, or go to www.delta-gear.com; from Delta Research Corp. phone 734-261-6400 or go to www.delrecorp.com.
Shop Feels the Joys of Five-Axis Machining
Tri Tech USA (Liberty, SC) acquired a Haas VM3 mill capable of simultaneous five-axis machining via a trunnion table in 2011. The primary objective of this purchase was to further advance the company’s precision and productivity, while allowing the company to acquire new, complex projects. After a year’s experience with precision five-axis machining, this rationale has proven itself to be right on target.
Along the way, however, extracting all the value from its new five-axis capabilities has required a significant amount of effort. The production staff, and Ryan Quinn, an experienced CNC programmer, had almost no prior experience with five-axis programming. Fortunately, the company’s five-axis programming software, Mastercam from CNC Software Inc. (Tolland, CT) provided all the tools Quinn needed for efficient and sophisticated programming. On occasions where Quinn found himself baffled, the team at Barefoot CNC (the local Mastercam reseller) was quick to respond with training and consulting support to ensure he was headed in the right direction.
Embracing new opportunities is nothing new to Tri Tech. In 2009, company president Joe Bacigalupo moved his thriving manufacturing business, millions of dollars worth of equipment, and 40 high-skill manufacturing jobs to a newly constructed 40,000 ft2 (3716-m2) facility in Liberty, SC. According to Production Manager John Smith, “An early item on the agenda was becoming AS-9100 certified. Our company has also worked hard to keep our manufacturing balanced among our primary areas of expertise: integration, production machining, and prototyping. These areas tend to feed each other, so keeping the balance is important to us at Tri Tech. Our new five-axis system gives us capabilities that support each of our core competencies.” Tri Tech USA is Tier 1 and Tier 2 in automotive and Tier 2 in aerospace.
Quinn had to immerse himself quickly into the nuances of five-axis machining because his very first project was a challenging one. A bearing manufacturer contacted Tri Tech to machine a difficult vibratory bowl feeder with a track that allows ball bearings to move smoothly up a feeder track during production. The final design included a challenging 60° helical track precisely located on a variable degree ramp.
Quinn says, “I had never programmed a five-axis part like that before. It was an opportunity to develop my five-axis skills on the fly. We had to figure out everything from scratch, including the ramp angles, which we did by using our CMM to capture data from the part the customer gave us. We then imported it into Mastercam to reverse engineer it. We spent the extra time on fixturing because we were going to make a lot of these for six or seven different ball bearing sizes.”
One very important aspect was that the geometry had to be repeatable. That meant the part had to be located very accurately to eliminate any potential for slipping. Bacigalupo came up with the idea of putting a puck on the trunnion. The puck would align with a dead-center hole on the bottom of the casting, and the part would be secured to the puck with a dowel pin. The aluminum feeder bowl casting was very porous. Providing the best possible surface finish was a challenge because it was important that the balls not catch while they were vibrating up the ramp.
The project was completed on schedule and satisfied the customer’s need for precision feeders. Quinn says that this project was a real confidence builder for him. He learned that with the tools available to him in the software he would be able to handle any five-axis project that came his way.
A challenging project involved a gearbox with critical features in 20 separate planes, many with very precise relationships to one or more features in other planes. There would be no way to achieve this degree of precision without five-axis machining. Productivity was also an issue because this part would ultimately be manufactured on a production basis. The prototype gear housing had to be developed using CNC toolpaths optimized for minimized machine cycles.
The customer provided a model of the casting and the finished part, which Quinn then imported into the system. Quinn has never had a problem creating a model in Mastercam, regardless of the source of the data. Workholding is always one of the most important aspects of any five-axis programming job. The part must be held securely so there is no possibility of it slipping. However, the holding devices must not impede access of the tool to all areas of the part, nor can it exert forces that might distort or crush the workpiece.
Quinn uses the CAD capabilities to study the part, find the ideal points of contact, and design the holding fixtures in relation to the parts. On occasion, he will send this model back to the engineering department where they can analyze the clamping forces in SolidWorks to ensure they are neither excessive nor inadequate for the job.
WCS (Work Coordinate System) is used to identify and name various planes from which the necessary toolpaths will be created to complete numerous operations on this one part. Moving from one plane to another is done simply by clicking on the name of the plane from an open menu on the desktop.
The software has a library for both tools and toolholders. If manufacturers don’t offer exactly what is required in terms of reach and rigidity, the programmer can design his own ideal tool or toolholder for the operation. Where exceptional rigidity is required, a single-piece, tool/holder combination can be designed.
Quinn programs his part three times in Mastercam. In the early rough programming stage he quickly lays down the toolpaths that are likely to be used and uses these to estimate machine cycles for quoting. Once Tri Tech gets the project, he programs it a second time for more accuracy. This program is then shop-floor ready, will run about 20% faster, and is suitable for manufacturing one or two pieces.
However, if large production volumes are required, Quinn goes over the program a third time using the Trim function to select and delete unnecessary clearance moves and “air cutting” sequences from the program. This can reduce machine cycles by another 10% and eliminates any need for production people on the floor to optimize machine cycles by altering G-code.
By simulating the machining process using a model of the part as it sits in its fixture, Quinn can see everything that will happen to the part, tool, and toolholder during the entire process. Once he knows the process is safe, he will take screen captures from Mastercam’s simulator. These can be used by operators for duplicating the part setup in the holding fixtures on the machine. ME
For more information from Tri Tech USA, go to www.tri-techusa.com, or phone 864-843-1100; for more information from Mastercam/CNC Software Inc., go to www.mastercam.com or phone 860-875-5006.
This article first appeared in the October issue of Manufacturing Engineering magazine. Click here for a PDF of the article.