Click on a thumbnail to see the full-size image
Roush Performance Products
By using a new aluminum casting on the air intake manifold of our Roush Performance Products’ Stage 3 Mustang, we reduced machining time for the part from 150 to 90 minutes. The Stage 3 Mustang, a high-performance version of Ford’s Mustang GT, includes a custom air-intake manifold.
When we tried traditional aluminum casting for the manifold, we found that the parts lacked dimensional stability, increasing machining time and adding to the cost. Manifolds produced using the new technology are so accurate that one machining setup works for all parts, reducing machining time by almost half and ultimately reducing the cost of the part. In addition, the smooth finish of the castings produces an attractive surface. The parts produced by Taylor-Pohlman’s new process are so consistent that the machinists just set them up and let them run.
My employer, Roush Performance Products, is a part of Roush Industries, a full-service engineering company, headquartered in the Detroit suburb of Livonia, MI. Although primarily known for providing engineering, management, and prototype services to the transportation industry, Roush has developed a significant role in providing engineering and manufacturing for the electronics, sports equipment, and motorsports industries. Roush Performance Products upgrades commercially available vehicles in both appearance and handling, with track-proven race technologies and backed by advanced automotive engineering.
We make customized versions of a number of vehicles including the Mustang, Cougar, Expedition and Ford F-150 pickup truck.
Roush has been selling Stage 1 and Stage 2 Mustangs, based on the Ford Mustang GT, for a number of years. These vehicles have body, tire and suspension upgrades that change the look of the vehicle but no changes to the powertrain. With the introduction of its Stage 3 Mustang, we now offer a version of the Mustang that combines carefully engineered changes to suspension, braking, and body systems with changes to the powertrain that greatly increase horsepower and torque. We reengineered the 4.6L single overhead cam V8 engine that comes from Ford – boosting horsepower from 260 hp to 360 hp. One automotive writer called the Stage 3 Mustang "one of the strongest street Mustangs ever offered for sale to the public." In recent testing, a manual transmission Stage 3 Mustang accelerated from zero to 60 mph in 4.3 seconds. The automatic version reached 60 in 5.3 seconds. It is an awesome driving car, running with a Corvette Z06 and close to a Viper.
In modifying the Mustang engine, we replaced the standard induction system with an Eaton Roots-type belt-driven supercharger. Instead of the Ford intake manifold, we used a new one designed in-house that incorporates an Allied-Signal dual-core air-to-water heat-exchanger or intercooler with electric water pump.
We also equipped the engine with a modified mass airflow sensor with a greater range, special Bosch fuel injectors, and a BBK throttle body. Our approach is unique in that we are the only vehicle modifier that puts a blower on a Mustang and uses a second drive belt to spin the supercharger and the alternator. Everyone else runs them off the one existing belt.
Our design takes the load off the other front-end accessory drive components and the front main bearing in the engine, dramatically improving durability. Other changes to the Stage 3 Mustang engine include calibration for the spark and fuel systems that eliminate flat spots, which you see in other modified vehicles.
The intake manifold that we designed for this engine is a critical component from both an aesthetic and a geometric standpoint. As it is the first thing the eye sees when the hood is opened, we wanted it completely smooth and free of flecks of sand and other debris. The manifold also needs to be geometrically precise because it mates with the supercharger, the cylinder heads, and the intercooler. It also includes a number of mounting holes for rails and coil packs.
Dimensional stability in the casting is critical because it affects the machining we do to create the mating surfaces and the mounting holes. When a part comes from the foundry as a rough-cast piece, it includes additional material where you will be cutting. This way, when you set the part up and make the initial cut to establish the ‘0 line,’ there is material you can machine away.
On a part that is not tightly dimensionally controlled, you might machine one surface then go to the next feature and the cutter won’t even touch the part. You waste a lot of time moving the part around, or balancing it, to get a consistent zero point for each surface. Dimensionally stable parts, in contrast, eliminate this additional setup time, reducing machining and cycle time and allowing the part to be produced at a lower cost.
When we were building the prototype engine, we hired a traditional sand-caster to produce several dozen test pieces of the air intake manifold. This was the largest and most complex aluminum casting we needed for this project. And while the prototyping foundry wanted the job, we soon learned that the traditional sand-casting process couldn’t deliver the dimensional stability they needed.
"It took a lot of work to get one of the prototype parts set up and balanced for machining," says Tim Kilgore, director of the machining center at Roush Performance Products. "Then we’d take the next one out of the box and it would balance totally differently. Where we had machine stock on one part there was virtually none on the next part, so we had to go through the entire setup process again, balancing it out with shims." Machining the prototype manifolds took about 150 minutes per part.
More Accurate Manifolds
Then we asked Taylor-Pohlman about producing the air intake manifold, they were the only one who had confidence that they could control the part consistently. They used the new V-Process sand casting technology that was developed specially for producing aluminum castings with thin walls and close tolerances. V-Process is a form of sand casting that uses dry unbonded sand that is finer in grain size than that used in a traditional sand casting. The process is accomplished in the following steps. A pattern, a "positive" of the part to be cast, that has vent holes is placed on a hollow carrier plate. A heater softens a sheet of thermoplastic film 3 to 7 mils thick that has good elasticity and a high deformation ratio.
The softened film drapes over the pattern. A vacuum draws the film skin-tight over the pattern surfaces and assures precise reproduction in the mold. The mold flask is placed on the film-coated pattern. The walls of the flask constitute a vacuum chamber and the flask is filled with fine-grained sand. No binders are used. Vibration is applied to the sand to compact and maximize its bulk density. This avoids the need for jolting or ramming.
The sand is then leveled and a sprue cup formed for the pour. The top of the mold surface is covered with unheated plastic film. Vacuum is applied to the flask to create negative pressure and harden the sand. When the vacuum is released on the pattern, the mold strips easily off the pattern.
The drag half of the mold is similarly formed and the cope and drag halves are assembled to form a plastic-lined cavity with the exact shape of the casting. Still under vacuum, the mold receives the molten aluminum. After cooling, the vacuum is released and the free-flowing sand drops away leaving a clean casting. Sand is then cooled for reuse.
One benefit of having the air intake manifolds produced this way is that they cost less per piece. The vendor that produced the prototype parts charged much less for tooling because they were working with plaster. But the per piece cost was quite high.
With Taylor-Pohlman’s process, the higher tooling cost was balanced out by a lower cost for the 1,000-part run. When the entire fee is converted into a per-piece cost, the V-Process costs less than half as much. The cost of the new manifolds is an even better bargain because machining time for these parts is only 90 minutes, compared with 150 minutes for the prototypes. This is due to the accuracy of the V-Process pieces. "The parts are very consistent, so it takes less time to set them up on the machine," says Kilgore. "Basically, the same setup works for all the parts. Also, because each piece is identical, the machining itself is now completely unattended."
The other advantage of having the manifolds produced by the V-Process is a nicer appearance. Parts produced with traditional sand casting have a porous finish because of the grain size of the sand. The sand used in the V-Process is one-third that size, giving the finished product a smoother appearance. Also, the sand used in the V-Process does not contain other materials such as binding clay that can show up in the finished part. The manifold sits on top of engine so when you pick up the hood it is the focal point. The overall appearance of the part produced by Taylor-Pohlman’s V-Process is very clean and pleasing to the eye.
The Stage 3 Mustang went into production in March 2001. Prices for the car range from about $39,000 to just shy of $50,000.
For Roush Performance Products, having the air intake manifold produced by Taylor-Pohlman using V-Process casting technology has had two key benefits. The part costs less than one produced with traditional aluminum casting methods, yet its appearance is excellent and in keeping with the look of the Mustang.