Valve Maker Drops a Bomb on Cost

The engineers who design Marotta Controls' precision pneumatic components sit just a few yards away from the company's manufacturing plant. And that proximity paid off when they designed parts of a new ejection mechanism for the U.S. Air Force's Small Diameter Bomb (SBD) system. This forthcoming system will allow existing fighter jets, including the F-15E, to rack up more kills per sortie by equipping them with a weapons carriage that holds four small bombs rather than one large one. At 250 lbs, each one of the new 70-inch-long small bombs will weigh only half as much as the smallest bomb currently used by the Air Force. Size aside, the new bombs will feature advanced GPS guidance and other targeting systems designed to improve accuracy. And they will feature a new pneumatic bomb ejection system.

Which is where Marotta enters the picture. The company will supply the pneumatic firing valve and accumulator that release these small but potent bombs from the plane. In some ways, the valves represent business as usual for the company. It already supplies precision fluid control components for military, aerospace, and demanding industrial applications. And this lineup of components includes balanced-poppet solenoid valves similar in operation principle to the one used for the SDB firing valve (see http://rbi.ims.ca/4387-556 for more information on balanced poppet designs).

But this particular valve proved especially challenging from both design and manufacturing standpoints. Richard Molesworth, Marotta’s manager for aerospace systems, is intentionally vague about many of the design and performance details, citing confidentiality requirements. But he does note that the SDB approach in general does try to pack a lot of components in a small space. After all, it does squeeze four small bombs and their weapons carriage in roughly the space normally occupied by one big bomb. This packaging constraint ultimately made the valve design more difficult because the valve has to contribute to overall space savings by integrating functions unrelated to its own operation. For example, the valve includes mounting points for and actuates an over-center latch used when the bombs are loaded into the carriage. “The packaging constraints required us to do more mechanical interface work with other parts of the system,” Molesworth says.

From a manufacturing standpoint, the company’s engineers also faced additional challenges with this valve. First, the valve body has tight tolerances and surface finish requirements that are stringent even by Marotta’s standards. Some of the concentricity specifications on the bores are less than 0.001 inch, and the finish requirements for the entire part average 16 rms.

Making just one of these valves bodies would be tough enough, but the second of Marotta’s manufacturing challenges involves much larger production volumes than the company normally encounters. “A lot of our production runs are for two or three units,” Molesworth says. Yet the valves for the SBD will have volumes measured in the “thousands” once the program ramps up fully.

Finally, Marotta had to make this difficult part while meeting a hard-and-fast cost target—an “average unit production price,” as it’s known in the defense business. “The price expectation of our customer is fixed,” Molesworth says diplomatically. Marotta didn’t start this project anywhere near that desired cost. “We wouldn’t have been able to meet our target price with our existing processes and supply model,” says Tony DiGiovanna, a senior program manager.

So Marotta engineers made some changes, taking what boils down to a three-pronged approach to getting this job right.

Focus on Component Design First
Adherence to Design for Manufacturing (DFM) principles played a key role in driving toward production and cost goals. Marotta’s own design team and an independent reviewer scoured the valve design for any threats to manufacturability. And the design aspect that counted the most in this case turned out to be dimensional tolerances.

As Molesworth points out, overly tight tolerances can in general be a big threat to manufacturing yields and ultimately drive up costs. So long before the valve body went into production, Marotta engineers reviewed every single dimensional tolerance and looked for places where those tolerances could be eased up a bit. “You don’t want the tightest tolerances where you don’t need them,” Molesworth says. Of course, this valve still has its share of tight tolerances—largely because it has to operate with response times less than 50 msec and at pressures over 6,000 psi. Most critical were the tolerances that involve sealing surfaces. “But everything else was up for grabs,” Molesworth says.

The company’s engineers didn’t stop there. They also looked for other design features that could be simplified for manufacturing’s sake. For example, they considered a couple of different methods of retaining one of the valve’s O-rings—either a machined undercut in the valve body or a split bushing. They ultimately picked the bushing; though it required an additional component, it avoided a potentially difficult and expensive machining operation.

Get the Right Machine Tools
Of all the costs associated with the valve, the valve body represents the single biggest cost driver. Molesworth notes that it alone is responsible for 65 percent of the hardware cost, and it’s easy to see why given its machining requirements. Made from hardened heat-treated 17-4 PH steel, the valve body has 18 tight tolerance bores, some of them angled, and nine thread holes. Many of the bores have concentricity requirements of 0.001 inch, and a few have even tighter requirements. And the part, which starts off as an investment cast blank, has to be machined on all six sides.

To machine this tricky part without driving up costs, Marotta engineers knew they needed to create an extremely capable manufacturing cell. So they started with an exhaustive search for the right machine tools. DiGiovanna recalls that eight machine tool vendors initially looked good. Yet all but three quickly fell out of contention. “Five of them just couldn’t meet our requirements,” he says. Marotta engineers went on to study four solutions from the three remaining vendors (see table for a detailed look at the evaluation criteria).

One solution, a pair of SuperMiller 400’s from Mori Seiki, “stood above the rest,” DiGiovanna says. These machines feature a five-axis machining center with integrated turning (see http://rbi.ims.ca/4387-557 for the full specs). It’s worth noting that all four of the final machining solutions did pass muster from a technical standpoint; all could produce the valve body parts. But the Mori Seiki shined when it came to operator costs. The SuperMillers can run unattended for up to eight a day, while the other solutions could run for an hour or less without attention.

Smart about Assembly
While most of the valve’s cost can be found in the body, assembly and testing represent the second largest share of the cost. Marotta has contained some of these costs by changing its supply model. For example, some of the soldering and wire harness work that Marotta does in house for lower volume products has been outsourced to a supplier. “Some of the electronics now come in as a module,” Molesworth says.

To further lower assembly and testing costs while contending with the larger production volumes, Marotta also created a stand-alone production cell that makes use of a Kanban supply system. In that cell, Marotta has invested in assembly systems that may have cost more upfront but will ultimately save time and money. The cell includes computerized assembly benches, for instance. These graphically display the correct assembly procedures and collect assembly data—such as bolt torque—for statistical quality control.

Molesworth cites the cell’s laser marking system as another good example. Its initial cost certainly exceeds the cost of a few nameplates, but the laser-marking machine ultimately saves the cost of buying, storing, and affixing those nameplates. And it avoids the cost of mistaken labeling. “Our focus in this cell has been to spend more on upfront tooling that will lower our production costs,” he says.

And it looks like this approach has paid off. Since the beginning of the project, the company has halved the total cost of making these valves, DiGiovanna reports. And he expects that number to fall even further as the production line starts to ramp up next month.