March 20, 2015

MESA starts largest production series in its history

MESAFAB WORKER Penny Moore (1747) loads a cassette into a photolithography machine at the MESAFab, which is now producing base wafers for Application-Specific Integrated Circuits for the nuclear weapons program. Planning and preparation for production took years and involved more than 100 people. (Photo by Lloyd Wilson)

by Sue Major Holmes

Sandia’s Microsystems and Engineering Sciences Applications complex has begun making silicon wafers for three nuclear weapon modernization programs, the largest production series in MESA’s history.

MESA’s silicon fab in October started producing base wafers for Application-Specific Integrated Circuits (ASICs) for the B61-12 Life Extension Program (LEP), W88 ALT (alteration) 370, and W87 Mk21 Fuze Replacement nuclear weapons. Planning and preparation took years and involved more than 100 people.

“We left no one untouched. If you were standing still, you got something to do,” says Jayne Bendure (1747), who was in charge of organizing 1,000 line items that had to be checked off before wafer production began.

Chief of Staff Dave Sandison (110), a senior manager at MESA’s Silicon Fab when the manufacturing readiness process began in July 2011, says detailed requirements for making war reserve-quality ASICs produced a gigantic list “in regular 12-point font about 4 feet long that went down the wall in my office.”

He implemented weekly meetings to review what had been accomplished and what was on tap for the coming week. The meetings focused on hitting deadlines and determining what was needed to stay on track, which convinced him and his successor, Mike Daily, that they needed more people. “You start to see the same people’s names show up every week, and they’re on a gazillion things,” Dave says.

“It was a multiyear program executed with precision and dedication,” says Gil Herrera, director of Microsystems Science &Technology Center 1700. “I’m very, very proud of the team. Now the hard part begins: We’ve got to make the parts.”

Center of Labs’ microsystems work

MESA is the center of Sandia’s investment in microsystems research, development, and prototyping. The 400,000-square foot complex of cleanrooms, labs, and offices houses the design, development, manufacture, integration, and qualification of trusted microsystems for national security applications.

MESA includes the Silicon Fab (SiFab), completed  in 1988, and the Compound Semiconductor MicroFab, completed in 2006.

The SiFab, certified by DoD as a trusted foundry, develops and produces technologies for radiation-hardened complementary metal-oxide semiconductor (CMOS) integrated circuits and MEMS (micro-electro-mechanical systems). The MicroFab is a green-certified plant for III-V compound semiconductor material processing, post-silicon wafer processing, and advanced packaging, and for heterojunction bipolar transistor (HBT) production. Both fabs conduct R&D for future nuclear weapon and broader national security applications.

Kaila Raby (1754), manager for the Product Realization Team for ASICs, calls the production start a huge milestone. The SiFab will make ASICs through 2018 with the plant running its normal schedule of 24 hours a day, five days a week. In addition, the MicroFab is preparing to begin producing HBT integrated circuits in April, the first time MESA will produce HBT products for the stockpile, Kaila says.

Ten different silicon ASIC products go into the work on the B61-12, the W88 ALT, and the Mk 21 Fuze systems, Kaila says. Seven of the 10 ASICs have base wafers that are customized into product-specific designs during later production, she says. It’s these base wafers that the SiFab started manufacturing.

Sandia has invested in new manufacturing tools and processes for MESA, including 2-D marking, similar to bar-coding, for individual devices; an electronic production control system; automated lot acceptance support systems; streamlined quality management; and a greater focus on preventing defects, Gil says.

“A lot of things we’ve been working on for the last few years came together, and we’re doing all this while minimizing impacts to our other work for national security customers and the research mission,” he says.

Identifying production needs  as multiyear project

MESA SiFab finished wafer fabrication for the W76-1 nuclear weapon in 2009, and began identifying manufacturing needs for the B61 and W88 programs, says manager Dale Hetherington (1746). Dale and manager Alan Mitchell (1747) credit success to the hard work of many people within the fab organizations including the equipment maintenance, process, and engineering staff.

“A large part of our mission is research and development and work for other customers besides the nuclear weapons program, so while we’ve been getting ready for this NW production and doing a 1,000 line-item production plan, we’ve still been manufacturing prototypes for the NW complex, for other missions around Sandia, and other labs as well,” Dale says. “So it’s not as if we shut down and did nothing but get ready for production.”

Volume for ASIC production will be more than three times that of the W76-1 production. MESA uses a build-ahead process: build many wafers and store them so they’re ready when needed.

“We typically build early because the wafers have to be diced into chips, the chips packaged and delivered to subsystem customers, and those subsystems then integrated into higher-level systems,” Dale says. “We have a production plan that factors in all the various chips for the B61, W88, and W87 programs.”

Production required prototyping in advance so designs were in place and ready to manufacture, and setting up quality systems and making sure they were rigorously documented, Alan says.

Wafer production is a critical part of the ASIC process but is only one step toward the final product and acceptance by NNSA. “While we build the wafers in our fab, when we’re done with them they have to be electrically tested, they need burn-in and reliability evaluation, they need packaging, so there are multiple sets of work activities from initial customer engagement through the design phases and into manufacturing,” Alan says.

MESA does it all: R&D, production, and development

Senior manager Paul Raglin (1210) points out MESA’s fabs are unique: They do R&D and production of ASICs and HBTs, all in-house with trusted components. “We can go directly into the stockpile with all the certifications,” he says.

MESA receives about $50 million a year through NNSA’s Readiness in Technical Base and Facilities (RTBF), Paul says. While RTBF provides base-level funding, the W88 and B61 programs furnish additional money for staff and materials.

The Sandia silicon fab revitalization (SSiFR) program has pledged $150 million over seven years to replace aging tools and transition MESA to 8-inch wafers by about 2020, Paul says.

A manufacturability review in late 2011 evaluated how various areas and tools operated. Curt Dundas (1746) says that snapshot became the basis for deciding what needed improvement and what could affect the SiFab’s ability to increase manufacturing capacity. The team worked down the list of needed improvements, evaluating any significant change to make sure it wouldn’t cause problems, he says.

“In manufacturing, it’s always about continuous improvement,” Curt says. “You’re never done. You just arrive at the point of a new snapshot.”

Jayne began organizing line items in August 2012 so the fab could switch from producing a primarily R&D line to an NW line. She says, however, most processes for controls, repeatability, and documentation already existed. “We reviewed everything, we improved the things we thought we needed to improve for NW, and we worked with our technology development group [1760] to make sure we were solid on what we needed to do.”

The plan required identifying tasks and who could keep those moving and involving enough people to get everything done on time. “It was a lot of work but having it structured like that was the only way to do it. People don’t mind hard work if they see you marching toward a big goal,” Jayne says.

She credits the production start to early planning, the structured plan, and senior management support. “But if not for the ability of the hands-on folks on the floor we would never have been successful in the time we had,” she says.


-- Sue Major Holmes

Back to top of page

Two CRF groups team up to simplify models for improving engines

When UQ met LES — Mohammad Khalil, Joseph Oefelein, Guilhem Lacaze, Layal Hakim, and Habib Najm (all 8351), researchers at Sandia’s Combustion Research Facility, are applying Uncertainty Quantification (UQ) principles to Large Eddy Simulations (LES) to create simpler combustion models. Here the team discusses results for a project on statistical calibration of simplified chemical mechanisms for diesel engine combustion.   (Photo by Randy Wong)

by Holly Larsen

For years, US automakers have been customizing the generic combustion process models from Sandia’s Combustion Research Facility (CRF) to design cleaner and more efficient engines for cars and trucks. To support this process — essential to a sustainable energy future — CRF researchers are constantly seeking to make these basic models not only more accurate and reliable, but easier and faster for industry to use.

This is no small task.

The problem lies in combustion chemistry’s notorious complexity and its interplay with flow turbulence, which encompasses thousands of coupled reactions that must be described over large ranges of pressure and temperature. Further, the chemical concentrations must often be resolved down to parts-per-billion levels for different pollutant species. It’s little wonder, then, that equations that describe the physics of combustion may take a year or more to solve using a high-performance supercomputer — a scale of computing resources that is simply not viable for the private sector.

“The CRF has a lot of experience in developing combustion models that provide accurate and reliable results,” says Joseph Oefelein, who with colleagues Layal Hakim and Guilhem Lacaze (all 8351), applies Large Eddy Simulation (LES) to mathematically model diverse combustion processes. “But now we have to distill new knowledge about combustion into models that can run even faster. Industry consistently makes the case that if a model takes more than a day to run, they can’t use it.”

Streamlining models with uncertainty quantification

The LES construct used by Joseph, Layal, and Guilhem resolves the larger energetic scales of a given flow and models the smaller-scale physics associated with combustion. This significantly reduces the computational burden of the combustion equations. However, LES is very sensitive to a variety of factors. A seemingly minor change to a key parameter can dramatically alter the model’s predictive accuracy. To overcome this challenge, the modelers consulted with fellow CRF scientists Habib Najm and Mohammad Khalil (both 8351), specialists in state-of-the-art application of Uncertainty Quantification (UQ).

Using UQ, this research team hopes to identify and better understand the portions of the LES model that are most sensitive to a design issue in question, such as the formation of soot or nitric oxide emissions within an engine. Instead of seeking a single result, UQ helps to characterize the range of results that occur from running aspects of the LES model multiple times with different values for key parameters. Analyzing the range enables researchers to determine which parameters are most sensitive and must be accounted for more accurately in the models and which are less sensitive and can be abbreviated or omitted. UQ then allows researchers to create a “surrogate” model — a simpler version of the full LES model that captures the essential elements, yielding useful answers while cutting computational time and costs.

So far, the LES-UQ team has demonstrated the feasibility of UQ’s application to LES using the well-studied Sydney bluff-body HM1 flame as a test platform. “By using Bayesian inference methods, we are able to propagate uncertainty through the simulations and understand the effects of various simulation inputs on predicted quantities of interest, such as engine performance and emissions characteristics,” says Guilhem.

This example has fueled the team’s confidence that UQ can dramatically reduce combustion simulation complexity, and therefore runtime, while retaining a useful level of accuracy. The effort has now been extended to include a focus on statistical calibration of simplified chemical mechanisms for diesel engine combustion.

Making multidisciplinary teams work

Multidisciplinary teams have become a reality of the modern world — but working effectively with people in different fields, with different expertise, can be challenging. Joseph credits some of the success of the combined LES-UQ team to experience. “The more we collaborate, the better we understand the optimal interface between various areas of expertise and thus how to share the workflow between team members.”

Also working in the team’s favor is the fact that the UQ techniques being developed are nonintrusive — that is, they do not require changes to the complex simulation code to which they are being applied. Thus, UQ has widespread applicability, from chemistry to materials science to nuclear engineering. Habib finds experience in other fields informs his intuition for applying UQ to LES. “Fortunately, I don’t have to rely solely on experiments in combustion areas. UQ research in entirely different fields is helping me expedite UQ’s application to Large Eddy Simulation.”

Finally Joseph points to the advantages of co-location. “Our workspaces are all close to each other, so it’s easy to get together to talk about ideas and what’s working or not on a near-daily basis. Sometimes all we need is a quick discussion in the hallway to move to the next step.”

For this team, next steps are numerous: acquiring a better understanding of the range of errors that can affect LES itself, regardless of UQ use; decoupling numerical and model errors; quantifying the cold-flow versus combustion attributes of scalar mixing; testing the UQ approach on different flame types; and testing the approach using different LES codes.

The team is optimistic, anticipating that in just a couple of years, this marriage of UQ with LES will be ready for much more routine application. The improved simulation capability couldn’t come at a better time for automakers, who are attempting to implement advanced, low-temperature combustion engine designs. Further, CRF expects its techniques for applying UQ to LES to bear fruit not just throughout the transportation sector — cars, trucks, buses, rail, jets — but in any industry employing computational fluid dynamics.


-- Holly Larsen

Back to top of page

Iron Rain fell on early Earth, new Z machine data supports

WHEN WORLDS COLLIDE — Sandia’s Z machine provides hard data on the surprisingly low amount of pressure required to metamorphose iron from a massive intrusive force to a gentle vaporized mist as astrophysical bodies collided in space during the later stage of Earth’s formation. That reduced pressure — 40 percent lower than  formerly assumed by theoreticians — helps explain the splashes of iron in Earth’s mantle. This artist’s concept shows a celestial body about the size of our moon slamming at great speed into a body the size of Mercury.   (Image Credit: NASA/JPL-Caltech)

by Neal Singer

Researchers at Sandia’s Z machine have helped untangle a long-standing mystery of astrophysics: why iron is found spattered throughout Earth’s mantle — the roughly 2,000-mile thick region between Earth’s core and its crust. It seemed more reasonable that iron arriving from collisions from  planetesimals ranging from several meters to hundreds of kilometers in diameter, during Earth’s late formative stages,  should have powered bullet-like directly to Earth’s core, where so much iron already exists.

A second, correlative mystery is why the moon proportionately has much less iron in its mantle than does Earth. Since the moon would have undergone the same extraterrestrial bombardment as its larger neighbor, what could explain the relative absence of that element in the moon’s own mantle?

To answer these questions, scientists led by professor Stein Jacobsen at Harvard University and professor Sarah Stewart at the University of California Davis wondered whether the accepted theoretical value of the vaporization point of iron under high pressures was correct. If vaporization occurred at lower pressures than assumed, a solid piece of iron after impact might disperse into an iron vapor that would blanket the forming Earth instead of punching through it. A resultant iron-rich rain would create the pockets of the element currently found.

As for the moon, the same dissolution of iron into vapor could occur, but the satellite’s weaker gravity would be unable to capture the bulk of the free-floating iron atoms, explaining the dearth of iron deposits on Earth’s nearest neighbor.

Looking for experimental rather than theoretical values, researchers turned to the capabilities of Sandia’s Z machine and its Fundamental Science Program, coordinated by Thomas Mattsson (1641). This led to a collaboration between Sandia, Harvard, UC Davis, and Lawrence Livermore National Laboratory (LLNL) to determine an experimental value for the vaporization threshold of iron to replace the theoretical value used over decades.

Rick Kraus at LLNL (formerly at Harvard University), and Sandia researchers Ray Lemke (1641) and Seth Root (1646) made use of Z’s capability to accelerate metals to extreme speeds using high magnetic fields. The researchers created a target that consisted of an iron rectangle 5 mm square and 200 microns thick, against which they launched aluminum flyer plates travelling up to 25 kilometers/second. At this impact pressure, the powerful shock waves created in the iron cause it to compress, heat up, and — in the zero pressure resulting from shock waves reflecting from the iron’s far surface — turn to vapor.

The result, published in Nature Geosciences on March 2 under the title “Impact vaporization of planetesimal cores in the late stages of planet formation,” shows the result: The shock pressure experimentally required to vaporize iron is approximately 507 gigapascals (GPa), undercutting by more than 40 percent the previous theoretical estimate of 887 GPa. Astrophysicists say that this lower pressure is readily achieved during the end stages of planetary accretion.

Emailed principal investigator Kraus, “Because planetary scientists always thought it was difficult to vaporize iron, they never thought of vaporization as an important process during the formation of the Earth and its core. But with our experiments, we showed that it’s very easy to impact-vaporize iron.  This changes the way we think of planet formation, in that instead of core formation occurring by iron sinking down to the growing Earth’s core in large blobs (technically called diapirs), that iron was vaporized, spread out in a plume over the surface of the Earth and rained out as small droplets. The small iron droplets mixed easily with the mantle, which changes our interpretation of the geochemical data we use to date the timing of Earth’s core formation.”

-- Neal Singer

Back to top of page

Download Lab News March 20, 2015 (PDF, 4MB)