Refuse to Lose

NASCAR – The National Association for Stock Car Auto Racing. The key word here is “stock.” The whole point being for the cars to closely resemble the stock, production automobiles the average Joe drives on the street. These are American-made cars, of course, and typically come in three flavors – Chevrolet, Ford, and Pontiac. So, just run down to the local car dealer, pick up a late-model piece of American iron, bolt on a few performance parts, and you’re ready to race, right?

Yeah, right! Any resemblance between a NASCAR Winston Cup race car and the car you’re driving is purely cosmetic. Sure, they both have an engine and four tires and use some of the same parts, but the resemblance ends there. As a matter of fact, the car you’re driving is far more technologically advanced – electronic ignition, computerized fuel injection, dual-overhead- cam multi-valve engines, variable valve timing, and light-weight composite materials. In contrast, NASCAR race cars are more like the muscle cars of the ’70’s – 350 V-8’s, four-barrel Holley carbs, standard ignitions, and steel bodies. So, how do you turn such outdated technology into a vehicle producing 720 horsepower, 525 ft-lb of torque, and capable of reaching nearly 200 miles per hour? That, my friend, is the question pondered by each of the 42 NASCAR teams every weekend.

The object is to build the fastest, most powerful, and best handling vehicle, while staying within the guidelines of the NASCAR rule book. There are rules governing vehicle height, weight, body shape, wheel base, displacement, induction, wheel size and a slew of other things. And, indeed, race teams must use many stock parts as their starting point. Through meticulous massaging modification and assembly, these parts are transformed into the fire breathing monsters that attack asphalt ovals around the nation every Sunday.

v1-i2-Refuse-to-Lose_1

For some insight into the building of successful NASCAR race cars, we visited Hendrick Motorsports (HMS) in Harrisburg, North Carolina.

Hendrick Motorsports currently fields three NASCAR Winston Cup teams, as well as a Chevrolet pick-up truck in the new NASCAR SuperTruck Series. Combined, Hendrick teams have won two Winston Cup championships, more than 75 major races, and three Daytona 500’s. Hendrick drivers Terry Labonte (#5) and Jeff Gordon (#24) finished first and second, respectively, in 1996, and both are currently making serious runs for the 1997 Winston Cup Championship.

One of the keys to Hendrick’s success is their engines. Painstakingly hand-built, these engines are the lifeblood of the race program. Randy Dorton, director of the engine department at Hendrick Motorsports, is an accomplished builder with a long history. He keeps a watchful eye on the engine department. “I’ve been head of the engine program since Rick (Hendrick) began (NASCAR racing) in 1984. There were only about three people then, and now there are 45,” Dorton said.

Dorton has been building race engines since the early ’70’s, when he built a 427 Ford for racing on local dirt tracks. In the early ’80’s, he put together an engine shop of his own, where he built drag-boat engines for Rick Hendrick. Hendrick was interested in getting into NASCAR, so he bought a race car and contracted with Dorton for three engines. Before those three were complete, the order was increased to six. Shortly thereafter, Hendrick bought Dorton’s shop as the core engine group for his NASCAR team, and Hendrick Motorsports was born.

“After Rick (Hendrick) got involved, the company began to grow, and the volume of engines got higher and higher,” said Jim Wall, Engineering Group Manager for Hendrick Motorsports. Wall has worked with Randy Dorton since 1981, and has been with Hendrick Motorsports since their inception in 1984.

According to NASCAR rules, a Winston Cup engine must be a smallblock V-8 displacing a minimum of 350 and a maximum of 358 cubic inches. The engine must be naturally aspirated, and use a Holley single four-barrel carburetor of 750 to 830 cfm. The main engine components – block, cylinder heads, intake manifolds – must start as stock, off-the-shelf, OEM castings. In short, it must be a production engine. Granted, it’s a production engine from the ’50’s, but a production engine nonetheless. Race teams can modify these production engines tremendously – as long as they stay within the rules. Hendrick Motorsports uses Chevrolet engines; and, by the time they hit the race track, very little has been left untouched.

v1-i2-Refuse-to-Lose_2

One of the main ways to gain horsepower in an engine is to improve the air/fuel flow through the cylinder head. The more mixture you can stuff into the combustion chamber, the more horsepower is generated. This is typically done by relieving and shaping the ports and combustion chambers, and polishing the surfaces. It is one of the most time-consuming and expensive parts of building a race engine. “Cylinder heads have always been an expensive component in a racing engine assembly,” Wall said. “They’re usually the things that you wait on.”

Typically, a skilled craftsman – a head porter – would painstakingly carve out a set of cylinder heads from a set of stock castings. Through experience and feel, he would sculpt and polish the ports in an attempt to improve performance through increased airflow. It was a process that took up to a week to complete. Each head was a work of art, and the head porter was the artist.

Like fine art, however, head porting is difficult to duplicate. It’s difficult to get the same results from port-to-port, let alone from head-to-head. As a result, you end up getting a mix of engines – some good, some not. “You’d have ten engines, and you might have two or three that were really, really good,” Wall said. “The others would be all over the place.”

“We would develop a cylinder head, or a pair of heads, that would run real good on our engine,” Dorton said, “and we wanted to reproduce that; make another set like it. But, even though you try to do everything the same, there’s no guarantee the next set will work as well.”

As Hendrick Motorsports grew, this became more and more of a problem To stay competitive, and meet the demands of their growing race program, they needed to produce more good cylinder heads, faster. To meet this need, they farmed the work out to several outside suppliers. But this seriously increased lead times for the heads, and left them too dependent on others for important parts.

The solution was to bring the cylinder- head operations back in-house.

“We realized early on a need to incorporate manufacturing into our program,” Wall said. “The volume was there, and once the volume gets to a certain point, you’ve got to get the manufacturing in-house.” But how would they meet their increasing demand, without bringing in a large stable of head porters? The answer was to automate the process, so they could accurately mass produce the cylinder heads. HMS began investigating the use of computerized machining techniques. Several of their suppliers were already utilizing CNC machines, so they knew the technology was sound.

“We put a proposal together for Mr. Hendrick to purchase a machining centre, a coordinate measuring machine (CMM), and CAD/CAM software as a starting point for producing fully machined assemblies in-house,” Wall said. “We wanted to control our own destiny, so to speak, with cylinder heads. We didn’t want to rely on outside vendors for a key component of our engines.” With hand porting, it’s difficult to maintain consistency and repeat a good design once it’s been developed. There are certain points that are defined – the entrance on the intake manifold, the combustion chambers, the valve seats – but the transition between those points – from the intake flange to the valve seat – is left up to the head porter, who determines how much material to remove. “He usually starts from the original surface of the casting,” Dorton said. “And, from one port, or one head, to the next, there’s a plus or minus 0.030" casting tolerance. So a guy could easily have 0.060" more or less material, depending on what’s been changed in the casting, from head to head and port to port,” he said. “As a result, he could take out too much material, or not enough material. He could start with a port that was plus 0.030" and another one that’s minus 0.030". Now, he takes out 0.060" more material than he intended, based on a template or gauge, and ends up in water.”

To port cylinder heads by machine, you start with a set of heads you already know work well, and digitize their shape through detailed measurement with a coordinate measuring machine. This way, you’re starting with a precise definition of where everything is supposed to be. “It may seem like a point in space three inches down into a port,” Dorton said, “but you have defined exactly where that point is. Once you’ve defined each point, it makes it easy to reproduce the design,” he said. “We needed volume, and this is definitely one method for us to get that.”

“The major components of our engine assembly that require a lot of machining are the blocks, the cylinder heads, the pistons and the intake manifolds,” Wall said. “We began looking for a CNC machine that could do a lot of different things,” he said. “Not necessarily a machine that could do a specific task better, but one that was flexible enough to do all the different jobs we needed to do – blocks, cylinder heads, pistons and manifolds.”

After much research, HMS settled on a Maho 800 universal machining centre – a German-built machine with both horizontal and vertical spindles, and 5-axis capabilities. Costing more than $400,000, this was a serious investment for a race team. In addition, they purchased a Brown and Sharpe CMM, and Point Control CAD/CAM software.

The first project everyone wanted to tackle was porting a cylinder head. They started with a proven cylinder head design; used the coordinate measuring machine to precisely measure, define and digitize the design; then used the CAD/CAM software to write the program for the CNC machining centre. “It’s kind of a tough part to start out with for your first part,” Wall said. “We had to build a lot of fixturing, design some tooling and have it made. It took us right at six months to get a cylinder head design that we had fully machined and ran on an engine,” he said. Once the process was up and running, HMS was able to turn a 5- to 7- day job into an 8- to 10-hour job.

The success HMS found using CNC machines prompted them to bring more and more operations in-house. However, the more jobs they brought in-house, the less machine time they could devote to cylinder-head operations. Although the Maho 800 had served them well, it was no longer fast enough to meet their growing needs. Plus, they were starting to have some service concerns with the machine. “The Maho had done its job,” Dorton said. “We were very satisfied with it. But we had developed more and more things we needed to run on a machining centre.”

This prompted the need for additional CNC equipment. “We realized we needed a turning centre,” Dorton said. “That was coming more and more into play with us. And, we felt we needed to expand ourselves in horizontal and vertical again.” The distributor who sold Hendrick Motorsports the Maho suggested they take a look at machine tools from Haas Automation

“We needed faster machines that were more reliable,” Wall said. We wanted something that was American made; a company that was going to be there for parts and service. We’ve known about Haas for a long time, because of their rotary products,” he said. “The 5-axis work (cylinder heads) relies heavily on rotaries, and Haas has a tremendous amount of experience in that market. One of the contractors we’ve used for years, Frank Wiess Racing Components in Indianapolis, has made a ton of parts for us on Haas equipment,” Wall said. “We knew it was good equipment because they did good work with it.”

At the same time Hendrick Motorsports was looking for new equipment, Haas Automation was looking for a race team to sponsor. “We wanted to find a team that would be a good showcase for our equipment,” said Peter Zierhut, Marketing Manager for Haas Automation. “We were looking for a successful team that would take full advantage of the capabilities of our machines.”

After touring the Hendrick facility, and discussing the needs of Hendrick Motorsports, a sponsorship agreement was reached for the 1997 NASCAR Winston Cup season. Haas agreed to provide three machines – a lathe, a vertical machining centre and a 5-axis horizontal machining centre.

“It sounded like Haas was a good company,” Dorton said, “and it sounded like good equipment. They had an interest in racing, and we had a need for more equipment.”

“I’m really excited about being able to work with a machine-tool manufacturer that’s based in the United States,” Wall said. “They’re American made; I like that,” he said.

The first Haas machine to arrive was an HL-4 lathe with 14.5" x 34" turning capacity and programmable hydraulic tailstock. “There was nothing here in the way of CNC turning equipment,” Dorton said, “we had to tool ourselves up for it. We set ourselves up with jaws and fixtures, and immediately went to work making parts. The first part was a front dampener washer, and we very quickly were turning pulleys and suspension parts,” he said. Just recently, they have begun turning wrist pins for their race pistons – a job they could not have done without the Haas lathe.

Next to arrive was a VF-4 vertical machining centre with 50" x 20" x 25" (xyz) travels. Hendrick wanted this machine mainly for plate work – engine brackets, braces, accessories, small surfacing jobs, drilling, tapping, and profiling.” Wall emphasized, however, “These machines aren’t limited just to engine components. We are using them to manufacture car parts as well. Our long-range plan is to use it for 5-axis work, and make quite a few chassis parts on it,” he said. “The VF-4 lets us make parts now that we couldn’t make before. With 50 inches of X-travel, you can put some big parts on it. By the end of the year, it’s going to be swamped, probably with chassis parts. Those guys do a tremendous number of parts, and they run into the same issues that we do in the engine shop, in terms of turnaround, delivery, and dealing with sub-contractors,” he said. “The VF-4 is a nice machine, and there’s no doubt it has allowed us to improve ourselves.”

The final machine to arrive was an HS-1R horizontal machining centre with 24" x 20" x 22" travels and a built-in 4th-axis rotary. To better suit the cylinder-head operations, Haas fitted the 4th-axis with an optional 19" x 40" extended table. HMS had already obtained a Haas HRT-210 rotary table and hydraulic tailstock in anticipation of 4-axis work on the VF-4.

Wall fitted the rotary and tailstock to the extended table on the HS-1R as the 5th axis, and designed flanges to hold a cylinder head between them. With only a few adjustments to the existing program, they had the machine running cylinder heads in record time. “What took us three months to set up on the Maho,” Dorton said, “took less than a month on the Haas. Literally within 30 days, the Haas went from the crate, to cutting chips on a head – with 98% Haas components.”

“The focus of the horizontal is cylinder heads,” Wall said “That’s the primary use. We’ll run it around-the-clock unattended if we get in a bind and really need parts.” The higher speed of the Haas (10k-spindle, 710- ipm rapids) has allowed HMS to reduce the cycle time for cylinder heads to about 6 hours, as opposed to 5- to 7-days for hand porting, and 8- to 10-hours on the Maho. “That number is a strong function of how fine of an increment you use,” Wall said. “To make a nice finish, you need an increment of 0.040" max. On the Maho, I had to use about 0.060", because we were very limited by the memory and DNC ability of the control,” he said. “If I got programs that were too big, the Maho had a hard time with them, it would balk. Out of a 10-hour program, you could spend as much as 45 minutes balking, waiting for the control to load code.” This results in dwell marks and reduced surface finish. “We ended up having to code the part around the limitations of the machine tool,” Wall said.

With the Haas, however, “The maximum spacing I used is 0.040", and in some areas I used 0.020",” Wall said, “I was intentionally trying to make the program big. The (cylinder head) program we’re running on the Haas gives nice finishes and the increments are nice. The Haas cuts much more efficiently, because it’s got inverse-time motion, which keeps a consistent chip load on the tool. The program is more than 11 megs,” he said, “and the whole thing fits into the Haas’ memory.” (Hendrick’s HS-1R has 8 MB of optional expanded memory.) “On the Maho, 7 or 8 megs is the maximum you can do, and you have to trade-off balking time. The Haas compresses the program, takes out the carriage returns and line feeds, and strips out the parts of the program you don’t need for the control.” This way, a much larger program can be loaded into the control’s memory.

The Haas control is designed and built in-house features dual, high-speed 32-bit processors. “It’s very user friendly,” Wall said. “I like the way it’s laid out, especially the similarities between the lathe and the mill. The vertical, the horizontal and the lathe all have basically the same control. They function the same and program easy,” he said.

But reduced cycle times are not the most important thing at Hendrick Motorsports. “We’re not in the business to make parts,” Wall said. “We’re in the business to win races. If we can do something to make a better part, but there’s a big penalty in cycle time, we’ll pay the penalty in cycle time to get a better part. We look at it like we’re our own customer, and that’s a double-edged sword. When you’ve got the ability to do it, you better get it done, because if you don’t, you really don’t have any excuses,” he said. “We’re not trying to cut corners. We’re trying to make parts that are as good as they can be.

Hendrick Motorsports continues to grow and continues to win races. Once again, as demand out paces supply, they are looking to acquire more CNC equipment. They continue to bring more and more manufacturing in-house, and are looking to add at least three more machines – another lathe, another horizontal, and possibly Haas’ new HS-2R large horizontal.

“To me, our engine program is a lot like steering a battleship,” Dorton said. “You get in the harbor; it’s pretty dangerous. Sometimes you’ve got to steer quickly. Having the Haas equipment gives us a lot of power to be flexible. That’s one thing that keeps us afloat here, our flexibility. So, if you know design, you have the engineering, you have the raw materials and you have the manufacturing equipment, you’re pretty much ready for whatever comes up,” he said.