Tooling for the wild blue yonder

In a project nearly 20 years in the making, one man’s vision of injection molded transparencies for Air Force jets spawned a technology that created heretofore unseen tool surface finishes.

When Kevin Roach walked into Bob Pinnell’s office in 1988, he perused the project listings available for newly assigned officers as he entered a different phase of his service career. Roach, a new Air Force second lieutenant, was intrigued by one project in particular. Pinnell was an Air Force program manager who worked at Wright Patterson Air Force Base’s Flight Dynamics Laboratory. In 1985 he started to promote the injection molding of canopies and windshields for fighter jets. Using funding originally allotted to the F-16 program, Pinnell began recruiting skilled personnel to tackle the revolutionary concept, which is why Roach found himself in that room 15 years ago.

Now at the University of Dayton Research Institute (UDRI), Roach has long since completed his transition into civilian life, and that germinal program that piqued his interest in 1988 has now become a reality thanks to earnest efforts from multiple players in a number of industries and the innovative application of technologies.

Molded windshields are scheduled to be in place on the Air Force’s T-38 training jet by the end of 2003, and canopies for the F/A-22 could be next. But before any of this could happen, a revolutionary mold polishing technique, along with technology firsts in many other areas, gave tool surfaces precision finishes never before achieved.

Where No Molding Had Gone Before
Air Force jets currently use a combination of thermoforming, laminating, and vacuum forming of acrylic or PC billets to create canopies. Sheets are stretched to 300 or 400 percent of their original size and then, layer by layer, vacuum-formed over a canopy-shaped mandrel. Problems occur with thickness inconsistencies and layers delaminating, creating 30 to 40 percent scrap rates, in a method that takes up to six to eight weeks.

Applying injection molding to any product can provide repeatability, and Pinnell definitely appreciated this, but several sizable hurdles remained before thermoforming could be replaced. The primary obstacle: Could canopies be molded with the required thickness and still maintain the requisite structural properties and optical clarity?

To find out, parts would have to be created. Pinnell and the Air Force targeted custom molder EnviroTech Molded Products (Salt Lake City, UT) from the start. Using a proprietary bulk injection molding process, EnviroTech had consistently proven the ability to mold substantially thick parts without voids or other inconsistencies (see “Molding Above and Beyond,” October 2002 IMM, for an initial report). Terry Sewell, a senior scientist engineer with Boeing who has worked on the project, says the initial parts from EnviroTech were, at least structurally, a success.

“What [the parts] were able to show,” Sewell says, “much to everyone’s surprise, including the companies that supplied the PC, was that large volumes of plastics could be injection molded, and that the possibility of making a large window by injection molding was [real].”

These early parts were subscale representations of their real-world counterparts and included a 2-by-2-ft, 3/4-inch-thick flat panel and a windshield-shaped conical part with a 1/2-inch-thick optical area. Both parts featured 2-inch-thick tapered edges, created so the transparencies’ attachment hardware could be insert molded, greatly easing installation.

Because of the test run’s promise, more money was allotted and a program to mold a larger canopy, loosely based on the forward half of an F-16 canopy, began in 1989. In 1993, a total of 145 parts (4 ft long, 75 lb each) were molded and subjected to a battery of tests, including bird-strike simulations in which 4-lb chicken carcasses were launched at the canopy at 500 knots. These examinations proved out the parts’ durability and structural soundness, but questions remained about optical clarity since a distortion effect derived from microscopic imperfections on the canopy tool’s surface was visible.

An Una-peel-ing Defect

Delta Tooling Co. (Auburn Hills, MI) created that initial mold, and since thousands of hours of polishing hadn’t created the desired finish, Delta quickly discovered that it faced a project with requirements in uncharted territory.

“There was never a mold that [Delta] had built that required this level of distortion-free optics,” Delta program manager Richard Mozer explains.

Those thousands of hours of handwork on the 1993 tool created a virtual mirror finish, but it simply wasn’t enough. The parts it molded had what the Air Force describes as an orange-peel distortion. Roach says this can be visualized by picturing a straight, crisp, black line on white paper. Under optimal circumstances, the line is well defined and sharp, but through orange-peel optics, its edges become muddled and blurred. So, in spite of months of “high-quality handwork by well-trained craftsmen” at Delta, as Roach describes, the tool had surface imperfections that created an optically unacceptable part.

“One of the biggest challenges that I’ve overcome in the last 14 years has been to figure out or find out that there aren’t ready specifications to describe what we wanted to do,” Roach says.

Helping him to this realization was Eugene Dahl, a scientist whose company, Precision Engineering Resource Assoc. (PERA), specialized in precision manufacturing and was brought onboard by UDRI to help eliminate the orange peel.

“[Dahl] was the first one to help me realize that there had been no requirement in industry to do what we wanted to do, so there was no hardware or computer system or cutting tool or grinder that could do what we wanted to be done,” Roach says.

Diamonds Are a Bench Hand’s Best Friend
What they wanted to do was achieve slope deviations of no more than 100 microinches/in. Imperfections on the order of tenths of thousandths of an inch would need to be detected and removed to create the necessary ultraprecision surface.

“The key to the program’s success,” Boeing’s Sewell says, “was recognizing that the tool surface had to be virtually defect free. Those surfaces that represent the part you’re making are almost like the Hubbell telescope mirror.”

Drawing on industries like lens polishing, Dahl pooled technologies to create what’s tentatively called a Conformal Bridge Lapping process. Now a proprietary product of Boeing/UDRI/PERA (who are seeking a patent), the lapping system consists of a handheld tool with a flexible diamond media from 3M affixed to a 3- to 4-sq-in ring lap. The ring lap is attached to fixturing and linkages and moved over the tool surface by hand, polishing the mold at rates approaching 2 to 4 ft/sec. The device is flexible enough to conform to curvatures found on a canopy tool, yet stiff enough to smooth out any bumps.

Using a combination of this tool and handwork, Delta, Roach, and Dahl worked on the latest canopy tool in spring 2000, attempting to achieve the elusive surface goals. As the delivery time neared and some technical problems mounted with the conformal lapping process, it was decided to finish polishing by hand until the bugs could be worked out on the new technology.

The tool was delivered on time in June 2000, but the parts molded in August still showed an orange-peel distortion. Working on the side, Dahl, at the behest of UDRI and Boeing, continued to tinker with the lapping device since it had shown such promise.

The Pringles Problem
While the device conformed to the tool’s surface, given the curved geometry of the cavity, it was somewhat prone to twisting or rotating and digging into the sides. To combat this, Dahl created fixturing to prevent rotation, devising what he called a torsional restraint coupling.

To visualize the problem, Roach offers an analogy using curved Pringles brand potato chips. With one chip representing the tool surface and the other the lapping device, either can be moved slightly forward and backward and side to side while maintaining good surface contact, but if you try to rotate one chip relative to the other, it quickly digs into the chip’s side.

Using the updated device, a new demonstration was performed in spring 2001 that polished the mold enough to create parts that exceeded the Air Force’s optical requirements.

While he lived to see the final demonstrations display resoundingly positive results, Dahl is no longer able to watch the project’s progress. “[Dahl] was diagnosed with a terminal illness and was gone a month later [March 2002] at 59,” Roach says, admitting he lost both a friend and a gifted colleague. “But he knew we had [a proven process]. It was just up to the Air Force whether or not it was going to be used.”

Taking Off
As the program moves forward and molded canopies and windshields are adopted in the field by the Air Force, Boeing’s Sewell anticipates monetary benefits, since tooling with the conformal lapping device reduced prices 50 to 70 percent, and the unit cost savings for the jets are as high as 70 to 80 percent, thanks to new efficiencies. But what is truly intriguing is the consistency a molded canopy offers to new technologies, and the doors it opens for heads-up displays and other emerging optical innovations.

“With an injection molded canopy,” Sewell says, “every one is identical, so it opens up a whole new world of optical devices that can be put in and utilized without the additional cost of trying to tune each optical device to each canopy.”

Boeing, Delta, and Roach are currently working to integrate the conformal lapping device with a six-axis, articulating-arm robot from Fanuc, making the process more appealing to more industries.

“[Boeing] feels there are probably other avenues for [the polishing device] in areas of lenses, optics, high-fidelity surfaces, and television screens. The same kind of process could apply,” says Sewell.

Almost 20 years after work began, the injection molded canopy program is still in many ways in its infancy, but the project and the technologies it created are quickly maturing.

“I’ve been working in transparencies for [Boeing] for 27 years and I kind of know what pilots are looking for,” Sewell says. “I’m impressed because I see better optics than on any part I’ve ever seen made before.”