Lightweight composite structures, manufactured “out of autoclave” without pressurized curing, are a major goal in NASA’s latest technology road map, but the shape of the tanks is bringing a degree of difficulty to this process.

Engineers working on two different NASA-backed composite cryogenic tank demonstrations at Marshall Space Flight Center, Huntsville, Alabama, say strong composite structures can be produced with heat-curing alone, as long as there are open edges that can vent water vapor and other gases that would otherwise create voids when the composite material hardens. In a cylindrical launch-vehicle propellant tank with a dome on the end, those edges do not exist.

“During cure, your volatiles, your entrapped air, all these things that make porosity or voids, travel down those fiber channels to the edges,” says Justin Jackson, who was project engineer on the 5.5-meter (18-ft.) composite liquid-hydrogen tank tested at the NASA field center near Huntsville. “So by trapping off those edges, all of that now has to go through the thickness of the laminate. It is just a more torturous path. You don’t get it out as efficiently as you would traversing down the fiber.”

The Boeing-built tank passed an arduous series of fill-and-drain tests, containing cryogenic liquid hydrogen with acceptable seepage, thanks to a technique that layered two thicknesses of carbon-fiber plies in a process designed to minimize the microcracking that has been a problem with composite cryotanks in the past. Weight savings over aluminum approached the 35% target set by NASA, according to John Fikes, NASA’s deputy project manager, and the out-of-autoclave manufacturing approach was cheaper than curing the tank under pressure.

In an industrial park in nearby Madison, Alabama, Northrop Grumman has used the facilities of subcontractor Griffon Aerospace to build a two-section out-of-autoclave composite tank for liquid oxygen and kerosene to help NASA evaluate the technology for possible use in the advanced strap-on boosters. This modification will raise the capability of the heavy-lift Space Launch System (SLS) to the 130 metric tons mandated by Congress.

As with the Boeing tank, the shape of the Northrop Grumman test article made curing in an unpressurized oven more difficult.

“I would say that that needs additional work,” says Martin McLaughlin, Northrop Grumman’s project manager on the Composite Tank-Set (CTS). “We were successful; we did do it out of autoclave, but if I had to go into serial production, I would probably stay with an autoclave right at the moment.”

In NASA’s latest draft road map outlining the technologies the agency believes it will need in the next 20 years, lightweight composite structures are significant, particularly for the SLS and other large vehicles that must be launched out of Earth’s gravity well for deep-space exploration.

“Manufacturing, especially for composites, is limited by available facility size, and the more complicated the design, the greater the cost and difficulty of manufacturing,” states the new NASA document, which was released May 11. “So concepts that are enabled by non-autoclave processing of composites and with integrated or low-cost tooling are of great importance.”

While it has no plans to begin building new boosters for the SLS until the 2020s, NASA spent $137.3 million in 2012 to begin engineering studies aimed at finding the most affordable way to raise SLS capability to the 130-metric-ton target. Northrop Grumman received $12.2 million of the $137.3 million to demonstrate a relatively cost-efficient out-of-autoclave composite propellant tank.

“The biggest benefit to us from that is that [Northrop Grumman] is identifying what we would call risks to developing an advanced booster,” says Bryan Barley, NASA’s project manager on the CTS. “Even though this one is a composite tank, there is work going on elsewhere, and when you look at the entire body of knowledge we are trying to gather, it is to identify those risks to developing an advanced booster design so we can drive decisions to get us where we want to be in the 2020s.”

Griffon has built an 8-ft.-dia. subscale composite tank structure and is preparing to static-test it with surrogate liquids—liquid nitrogen and diesel fuel—to see if the design approach Northrop Grumman has selected would be scalable to a full-size booster.

The test rig is heavily instrumented, with fiber-optic strain gauges embedded every centimeter along the walls of the tank to collect data at some 15,000 points. A trial run of the test rig was delayed when an early attempt using water as a surrogate delivered anomalous readings and one of the fiber-optic cables broke. The tank is to be removed from its stand, examined with ultrasound and repaired.

McLaughlin says composites are relatively easy to repair, but structures built with them must be carefully designed to “finesse” difficult mechanical loads because, unlike metals, they cannot be easily “beefed up.” In the subscale propellant tank, the company chose a design with a single bulkhead dome between the cryogenic liquid oxygen and the ambient kerosene to save weight and volume, but it required careful design around the bulkhead between the two fluids because of shearing that develops as the propellants are consumed in flight.

Discovering the cause of the test anomaly will add to the data set from the static-test series, which will use hydraulics to simulate flight loads. The testing is scheduled to take about two weeks once everything is ready to go (see photo).

In addition to engineering data, Northrop Grumman’s report will include results of detailed production monitoring by an industrial engineer, and from logs compiled by the technicians who laid up the composite structure by hand following laser projections driven by the computerized design.

That information will help NASA evaluate contractor cost estimates when the advanced boosters are put out for bid, according to Barley. Griffon built the test tank by hand, but the same numeric design can drive an “in-situ” layup tool the company has taken through critical design review. This tool could hold down production costs on a full-scale tank. Instead of moving a tank from workstation to workstation as the processing moves forward, it would stay in the same spot while “minimum-footprint” robots would come to it.

“You would use a single robot or a team of robots in that single cell to do the layup,” says Dawson Vincent, Griffon’s program manager. “You would have an oven that would roll over to that place and a [non-destructive inspection] head that would be on that same system to scan it in place. For any trim or secondary operation, you could put a trim head on those same robots.”

The idea would be to reduce lead time and also make the entire production process relatively portable so the tanks could be built near the SLS launch pad at Kennedy Space Center. In Madison, the tank was built in a portion of a former cotton warehouse leased for the purpose, with a layup room and oven for curing the composite installed inside.

At nearby NASA Marshall, the group that worked with Boeing to develop the 5.5-meter liquid hydrogen tank—a subscale version of the tank that could be added as an upgrade to the SLS upper stage—is considering testing an earlier 2.4-meter-dia. tank to failure.

“We’re looking at possibly doing a burst test of that,” says Fikes. “It would give us even a better understanding of what our margins were; because we didn’t fail the tank, we didn’t get to the capability of the tank.”

Northrop Grumman has been making large composite structures beginning with its work on the B-2 bomber in the 1980s and has autoclave equipment large enough to handle the subscale tank it built for the advanced-booster study. While NASA is using that work to guide future design and procurement for the SLS in the coming decade, the subscale tank could also benefit Northrop Grumman in other, nearer-term product lines.

“It would exactly fit into the XS-1 we are designing for Darpa,” McLaughlin says. “That was not a coincidence.”

Editor's note: This article was edited to change the image and include updates.