Shop Solutions: Tooling Ups Large-Crankshaft Output
Crankshafts come in sizes matched to the engine or device for which they transmit power. At Ohio Crankshaft (OC; Cleveland), crankshafts are produced that can be as large as 15' (4.6-m) long, weigh a hefty 7500 lb (3402 kg), and are machined from forgings up to 27" (686 mm) in diam.
Machining crankshafts for large-scale machine applications has a long history at Ohio Crankshaft, beginning in 1920, continuing through WWII defense production to the 1990s, when OC became a subsidiary company of Park-Ohio Holdings Corp. (Cleveland).
Sister companies under the aegis of the billion-dollar holding company include the forged and machine products companies, Berwick Forge, Krupp Forge, and Southwest Steel Processing, and capital equipment suppliers, Ajax Tocco Magnethermic, and Ajax/CECO. This combined expertise is capable of providing complete solutions for converting raw steel to precision-machined, finished, heat-treated, and custom critical products.
When company management determined that there was potential for growth in power transmission markets, it decided to expand the rough milling capacity for machining large crankshafts. "Our president, Les Havlik, saw the potential for capturing a larger share of our traditional business, as well as the potential of increasing our business in other market segments," explains Greg Stem, director of engineering and business development.
A team representing engineering, production, and purchasing was put together to identify and justify the capital-equipment investment that would be required to double the company's capacity for rough-milling the largest crankshafts. Machining for processing the large crankshafts is done on an internal milling machine from American GFM (Chesapeake, VA; Steyr, Austria). Other CNC machines and a camshaft milling machine were also acquired under the program to improve the company's competitiveness.
Ohio Crankshaft's objective was to find ways to increase its throughput by a factor of two for machining the large crankshafts. The price tag of the new internal milling machine needed to achieve this objective—about $7 million – sent the OC team on a search for an alternative solution.
"The internal milling machine is unique to the industry, and one that we had a lot of experience with," explains Stem. The working principle of American GFM internal crankshaft milling machines involves a stationary workpiece and a CNC-controlled internal milling cutter. Ohio Crankshaft has three of the American GFM internal milling machines, but only one that can process the largest crankshafts.
Configuration of the cutting tool is equally unusual. The cutting tool is a ring of nitrided tooling steel 52" (1320 mm) in diam and weighing 1350 lb (612 kg), with carbide cutters in the inside diam of the ring for rough-milling the steel forgings. "The cutting tool represents an investment of about $75,000 and takes months to manufacture," remarks Stem. "The design that we had used for years involved a solid cutter. Any crash was costly and time-consuming to get back up and running. For that reason, we had to have backup cutters on hand ready to go." Ohio Crankshaft, in fact, has six of the expensive cutters on hand.
"The largest crankshafts are used for locomotive, stationary power, and marine power applications," says Stem. "A long shaft for a 16-cylinder engine with 1100 in.3 [18,025 cm3] per cylinder can produce 6500 hp [4849 kW] at 1000 rpm. When you hear one of these run at idle, it has a real deep roar that gives you a sense of its power," says Stem. Smaller crankshafts for 8-cylinder engines are about half the size and weight.
When OC began looking for a way to increase its rough milling capacity, production was running around the clock with a mix of parts and production of 5–6 crankshafts/day. "When you talk about crankshaft production at automotive plants, you're talking about popping cranks out of machines in minutes. For our manufacturing, we were talking about removing 2000 lb [907 kg] of material for a large 16-cylinder crankshaft in a four-hour cycle," Stem observes.
"Materials are mostly medium to high-carbon steel. All of the wear surfaces including the main bearings, the pin bearings, and seal surfaces get induction hardened after rough milling and before grinding," says Stem.
Ohio Crankshaft looked to its cutting tool supplier, Greenleaf Corp. (Saegertown, PA) and its representative, Denny Carpenter, for ways to improve the throughput of its largest GFM internal milling machine by improving its tooling performance. The result was a totally new approach to the cutter construction, a switch from a solid body and inner ring to a segmented cutter configuration bolted to the ring's ID.
"When I first started working with Ohio Crankshaft 12 years ago, cycle time for crankshaft rough milling took about 4–4.5 hr and production was 4–5 crankshafts per day," says Carpenter. "Now rough milling is done in about an hour and forty minutes, and crankshaft throughput has increased with two different improvements in the cutter."
Greenleaf went back to the design basics for the cutter body and the carbide inserts. All the improvements in the cutter involved working with the original rings, which could be salvaged by boring and refitting with the redesigned cutter segments. Greenleaf redesigned the cutter body and insert geometries, and developed different grades of carbide.
"The carbide is a refined grade that has become a standard for Greenleaf, with geometry and edge preps that were specifically designed for the OC application," explains Carpenter. "We looked at speeds and feed rates. Through a combination of tooling grades, geometries, and edge preps, we have been able to run the machine at higher speeds than anyone formerly believed possible," states Carpenter.
The latest cutter design completed the progression from solid cutter to modified design, and has totally eliminated the cartridge and spring clamp and replaced it with a segmented approach. "The new design has totally eliminated the cartridge, and replaced it with eight segments and carbide inserts. There are just three screws to tighten the segment in place. If there is a crash, the segment can be quickly repaired or replaced," Stem explains.
Salvaging the original solid cutter bodies involved boring out the ID using Greenleaf's WG-300 ceramic cutter. "We put a counter bore in it with a small reference lip in the bore. The segments all sit up against the lip to hold the reference base so that programming can be done. There are pocket locations on each segment within tenths for the inserts," says Carpenter.
"With the better geometry, edge prep, and grade of carbide, cycle time dropped, and throughput has increased to 11.12 crankshafts a day compared with 5.6 crankshafts a day," says Stem. "We have, in fact, doubled our throughput with an investment of several hundred thousand dollars versus the millions that it would have cost to buy another machine."
"When I first came over here, we were getting two cranks per index on the carbide. Now we're getting 5–6 cranks per index, reducing carbide by about a half and reducing the tool changes required, so that actual run time has also been increased," says Carpenter.
Stem credits the partnership that Ohio Crankshaft has had with Greenleaf for continuing advances in its tooling. "As a part of our capital program to increase competitiveness, we have invested in other equipment for our general business, including a couple of VMCs and a specialized cam milling machine. Greenleaf was able to tool the cam mill up with cutter bodies and inserts that have succeeded in reducing cycle time by 60% on camshaft production. So our partnership continues to pay big dividends," says Stem.
Prototyping Technology Leaps Ahead
Conventional wisdom says there are two basic technological paths in prototyping die/mold work—EDM and machining.
Yes, 3-D imaging is making inroads, and there are several other prototyping alternatives. But if you are looking for the latest wrinkle in high-speed machining, look no further than a VMC that employs constant velocity (CV) processing rather than relying on brute horsepower and the complexities of controlling variable speeds and feeds.
W.L. Gore's Electronic Products Div. (EPD; Landenberg, PA) is using a Revolution CV4020 VMC from GBI Cincinnati Inc. (Cincinnati) to machine very small parts ranging in size from a finger to a finger tip. Materials include plastics, Delrin, Teflon, nylon, aluminum, and some steel. Ed McCracken, machinist, explains that in the past when he had the time and the machines, he would mill a prototype, then shoot a couple hundred parts (injection molding) using the just-finished prototype mold, checking that the product coming off the mold was what he wanted.
"Now I'm creating prototypes for die/mold work primarily for EPD, but I also handle anything that comes in from R&D, regardless of division," McCracken says. The typical mold shop would likely burn (EDM) these prototypes, a capability Gore has in this facility. "But when I can put the part in the VMC, and the finished prototype meets the specified tolerances [or better], why would I use another technology?" he asks. "I finish these parts right in the Revolution CV4020 VMC. I put the part in one time, machine it, and it comes out finished. And when we set up the fourth axis, these prototypes will come off perfect, every time, ready to ship."
The W.L. Gore facility occupies 106,000 ft2 (9848 m2) and has some 500 employees, representing the bulk of Gore's Electronic Products Div., a recent consolidation of manufacturing plants that make up Gore's eastern US cluster. Other manufacturing facilities are in the western US, Germany, Scotland, Japan, and China.
Founded in 1958, W.L. Gore & Associates Inc. (Newark, DE) is privately held and lists 8000 employees worldwide, and had sales of $2 billion in 2008. Gore technology is found in consumer goods, electronic and electrochemical materials, fibers, geochemical services, cables and cable assemblies, fabrics, filtration, pharmaceutics, medical/healthcare, sealants, and venting, among many others. Its most universally recognized product is Gore-Tex fabric.
"Each Gore facility operates as its own independent business," McCracken says. "We bought the Revolution strictly for R&D, in particular R&D for EPD. And while EPD is paying for the machine, paying for my time, I pretty much run every R&D project or hot job on the Revolution." McCracken says that he is able to service customers everywhere; the West Coast, Europe, Asia, or China. "You name it. Basically, if I can get the part on the machine, we get the prototype to them with very fast turnaround," he says.
Gore develops and builds its own equipment, including machinery, and McCracken says he uses the Revolution to mill parts in those cases as well. "I use it for everything," he says, "but a major reason we purchased the GBI machine was to have the ability to get our product in our customers' hands, right now. This is what today's customers demand. Everybody wants quick turnaround despite greatly compressed lead times. The days of just-in-time are over. That's not fast enough. Customers now expect on-demand results. If you can't deliver to their expectations, they'll look elsewhere."
The Revolution CV4020 is a 40-taper general purpose/die-mold machine with a 15,000 rpm, 20-hp (15- kW) spindle, 24-tool magazine and double-arm toolchanger (ATC) delivering a 1.9 sec chip-to-chip tool change. X and Y axes feature linear guides, and the Z axis has a box way design.
The control features a high-speed multiprocessor from MTI (Essex, ON, Canada) capable of simultaneous eight-axis, 50,000 blocks/sec processing speed. This allows the Revolution to achieve constant-velocity machining, permitting up to 50% reduction in cycle time when milling complex prismatic parts.
What is so special about the Revolution? "The key for us with this machine is processing speed and constant velocity." McCracken explains that conventional controls have a forward processing capacity of 600–3000 blocks/sec. High-end controls may provide 5000 blocks/sec. But a processing capacity to 50,000 blocks/sec for eight simultaneously controlled axes has a significant effect upon machine tool performance and efficiency.
Because the Revolution's MTI processer can handle the high volume of data that describes, in detail, the cutting tool path, the tool can maintain a nearly constant speed over the workpiece. This constant velocity along the cutting path eliminates the accel/decel experienced by the cutting tool as a conventional control tries to move through a complex prototype contour.
The Revolution's control has 80 smart data buffers, compared with four to five buffers found in conventional controls. Able to run any CAD/CAM brand that can run in a Windows XP environment, the control can handle mid-program restart without difficulty. In fact, such restarts can be handled in four different ways—by line number, block number, percentage of program run, or by having the operator position the cutter over the workpiece and start.
"When we were using a different machine [now retired] to do this work, we'd break a lot of tools," McCracken says. "To change a tool and go back in the program to the point where the tool broke meant wading through code to find just the right point. With the Revolution, we don't do that. When a tool breaks, we go straight in and pick up right where we removed the tool. We save a lot of time, and it's so easy."
Gore uses very small end mills, from 0.010" (0.25 mm) on up. The average end mill is 0.0625" (1.6 mm). "An end mill of 0.125 or 0.250" (3.2 or 6.4 mm) is a big end mill. Plus, these tiny thin tools are generally long, a good 6 to 10 x D. Pre-Revolution, we'd break a lot of these small, thin tools. Now we don't break tools; we change them out when they begin to dull," says McCracken.
A conventional look-ahead system relies on feedback as the machine operates. The Revolution control calculates look-ahead before the machine starts, and adjusts cutter movement as the machine runs. Its data-processing capacity enables the control to monitor operations and update toolpaths in real time.
Gore recently ran a 0.010" (0.25 mm) end mill at 12,000 rpm about 0.100" (2.54-mm) deep in Delrin. "We made the part without a problem. I was shocked," McCracken says. "With an end mill that small, if you don't have through-the-spindle coolant just right, and if the machine isn't rock-stable and the speed variable not constant, you'll break the tool, snap it right off. At that speed and depth, end mills that small can't survive changes in the toolpath. But at constant velocity, they handle complex prismatic work like nothing."
The Revolution is a deceptively quiet machine. McCracken says: "I actually have to check to confirm the Revolution is running—it's so quiet. It's the quietest machine on the shop floor.
NHRA Pit Crew Runs a CNC Shop
It has been a long time since a Ford has ranked high in the NHRA (National Hot Rod Association) Pro Stock Car Racing standings.
In its quest to do something about that, Cunningham Motor Sports turned to CAM software to squeeze more horsepower out of standard Ford engines. Jim Cunningham is a die-hard Ford enthusiast, driver, and co-owner with Gloria Cunningham of the Cunningham Race Team. They have embarked on an aggressive program of engine modification to increase the horsepower and rpm output of standard Ford 500 in.3 (8194 cm3) engines to levels competitive with today's dominant GM and Mopar products.
With that objective in mind, an engine shop was built in 2005 on the premises of the Cunningham-owned and operated Capitol Raceway (Crownsville, MD). Until recently, however, things were not going that well. As of the fall of 2007, the Cunningham shop had not produced an engine combination that would hold together without some significant component breaking.
Marcus Bowen, who divides his time between being crew chief and working in the engine shop, characterized the situation perfectly: "You have to get something that holds together, because you can't learn if you keep breaking it."
Over-reliance on outside shops with long lead times was one of the biggest problems. "If we wanted to change something, we had to make a phone call, set it up, ship it out, and wait for it to come back," Bowen explains. "When it came back, we had to put it back on the engine and back on the Dyno. That could take several weeks. And when you got it on the Dyno, you may not like what you see and have to take it apart."
Early in 2007, the team purchased a five-axis Millport CNC mill with a Centroid controller from Millsite Engineering (Ravenswood, WV) for head porting and a four-axis RMC V-30 machining center with a Milltronics controller from RMC Engine Rebuilding Equipment Inc. (Saginaw, MI). To develop toolpaths for both CNC systems, Cunningham chose Mastercam X3 software from Mastercam/CNC Software Inc. (Tolland, CT) with Mill, Multiaxis, Solids, Router, and Art modules.
The pit crew was now in the unique position of having to run its own CNC machine shop for engine-component R&D without having any CNC experience. The pit crew spends half of its time for ten months a year on the racing circuit. But the team was undaunted. Team members had visited with Mastercam at the Performance Racing Industry trade show in Orlando. They were convinced that they could master the necessary skills. Developing relationships with other customers they met at the Mastercam booth also proved to be a confidence builder and an invaluable resource.
Bowen's teammate Joe Greenwich took about three months to get acclimated to Mastercam to create toolpaths for head porting and other surfacing applications. Bowen, who had begun working in the shop in November 2007, took on all the fouraxis work. Doug Ewart, head engine builder, who had spent his 30 years in the business doing manual drafting and manual machining, picked up on Mastercam in about two months and "has never looked back."
The aggressive R&D program to improve engine power is producing a rapid turnaround on component availability, with 90% of these components being manufactured in the Cunningham shop to control lead times and costs.
Although the three users are new to CNC programming, the specialized nature of the parts they manufacture is actually a big help. They only have to know a fraction of Mastercam's capabilities, and the userfriendliness of the software allowed them to learn these quickly.
"You don't worry too much with Mastercam, because with its solid modeling and backplotting, when you run the program on your computer screen, what you see is what you get when you then run it on the CNC machine. You don't have to worry about the machine going down because you have misprogrammed something," Bowen explains.
A small number of machining activities have the greatest bearing on improving engine power, so the team has focused on head porting, intake manifolds and runners, the valve train, and pistons. Bowen says, "It's one big puzzle to get the combination put together right."
Piston work provides a telling example of fast turnarounds. Bowen says that the lead time for a set of pistons machined to specification by an outside vendor is 3–6 weeks. Cunningham can machine them on the team's CNCs and have them installed in less than three days.
Head porting is the most sophisticated machining task performed in the engine shop. The CNC is used to rough the port to near-net shape. Then desired port geometry is created with hand tools and tested on the dynamometer until power is believed to be optimal. A Mastercam five-axis program is used to digitize the new head port geometry by automatically collecting 3-D data from the part with a spindle probe. The subsequent CAD model can be used for creating five-axis toolpaths for machining eight identical head ports.
So even if it's not an actual race week, the pace of work at Cunningham Motorsports is always fast. Components are manufactured, tested, and installed in Ford engines, and the engines are then tested on the track a stone's throw away.
The team has gone from having no viable combinations in November 2007 to three reliable engines that have resulted in significant engine performance improvements. Horsepower has been increased from a peak of 1220 to 1360 hp (910.1015 kW). Engine speed has improved from about 9400 to more than 10,000 rpm.
Improvements have been so dramatic that Jim Cunningham has had to alter his driving style a little bit to adjust to the new power at his command.
"We've just got to keep working at it, constantly making changes," Bowen says. "We learn something every day. If the learning ever stops, we are out of a job."
"Maybe we can squeeze out another 50 hp [37 kW] in the next few months. Of course, now that Ford has committed to upgrading its engine block design, we will essentially be starting all over again when the new engines arrive," Bowen concludes.
This article was first published in the April 2009 edition of Manufacturing Engineering magazine.