Those who wander Berkeley Lab's hallways or surf the Lab's intranet are familiar with the art of Robin Lafever: spectacular images of the proposed SuperNova/Acceleration Probe (SNAP) satellite; the Silicon Vertex Tracker built for the BaBar detector at the Stanford Linear Accelerator Center's B-Factory; GRETINA, the most sensitive gamma-ray detector ever devised, now under construction; and many other intriguing devices.

Robin Lafever with a model, in its early stages, of the focal plane of the SNAP satellite imager (Photo Roy Kaltschmidt, CSO)

Lafever, a member of the Engineering Division, makes most of these images using computer-aided design (CAD). They're not just pretty pictures on a wall or a web page; most can be animated, rotated, pulled apart, and put back together again in ways engineers hadn't even dreamed of in the 1960s, when Lafever was in high school.

Much as he loves animated drawings, however, his real love is a design technique with much deeper roots: physical models in three dimensions, the bigger the better. In fact, Lafever often advocates building a full-scale model before completing engineering drawings. "The closer you can get to the real thing, the better the data," he says.

The focal plane of the SNAP satellite's imager is a case in point. SNAP is designed to discover the nature of the mysterious dark energy that is accelerating the expansion of the universe; to distinguish among theories of dark energy, the satellite will have to be able to detect any tiny variations in the rate of expansion since earliest times. The best way to do this is through precise measurements of the spectra of thousands of distant Type Ia supernovae — bright objects so far away and receding so fast that much of their light has been shifted into the infrared region of the spectrum.

Berkeley Lab scientists and engineers have devised a unique CCD, or charge-coupled device, that is far more sensitive to the entire spectrum than the best astronomical CCDs currently in use, and much more rugged and radiation-resistant as well. These CCDs will account for most of the imaging elements on SNAP's huge focal plane, an imaging surface vastly larger than the Hubble Space Telescope's. The focal plane and its components, including other imaging chips, and the shutters, moving mirrors, and spectrograph that work closely with the camera, are the satellite's heart and reason for being.    

Lafever and his colleague Matt Hoff in SNAP's engineering office have worked with specialists in the Lab's Engineering Division to build a full-scale model of the focal plane and its mounting struts, combining laser-machined stand-ins with real pieces of hardware. Mock-ups of SNAP's unique CCDs and other electronics plug into the support structure, allowing real wiring and cabling to be physically wrestled into place. "That's an area where CAD can't help us," Lafever says.

Around the world

Lafever developed his design philosophy during what he calls "my first two careers." Right out of high school he went to work for the Louis Berger Group, Inc., a globe-spanning company of architects and consulting engineers, and picked up his practical engineering skills on the job. "Our specialty was to go into the boonies, hire the locals, and get to work." 

He helped survey and design a pioneering highway in Nigeria "from Port Harcourt straight into the bush," he says. Other projects in Africa and Asia followed, including the road up Doi Inthanon (also called Ang Ga), at 2,565 meters (8,415 feet) the highest mountain in Thailand — a road now traveled by thousands of tourists from around the world and easily visible from space. "In fact," says Lafever, "it's pretty easy to find on Google Earth." 

The road to the top of Thailand's Doi Inthanon (Ang Ga), which Robin Lafever helped build, is easily visible from space.

During this peripatetic phase of his career Lafever met his wife, Emy, who was working for a design firm in Bangkok. The two decided to move to the San Francisco Bay Area, where Emy started Guillermina Oriental Arts and Antiques, still a going concern, and Lafever earned his degree in industrial design from San Francisco State University. For the next dozen years he was an entrepreneur.

One of his companies, Westedge, spun-off to make high-quality acrylic hinges. Another company, a partnership he'd founded with his design professors, provided services to inventors. One fellow came in with funding for wind power products. Lafever and his partners came up with a bright idea for a windmill generator suitable for sailboats; instead of the usual blades it was made out of rip-stop nylon, shaped like an Archimedean screw, and worked as well in low winds as high. Over the years the gadget earned significant royalties and fees for the inventor and his design partners.

Eventually, says Lafever, "I got tired of leading novice inventors through the learning curve and decided to retire." He started a new business, one that had more to do with destruction than invention, doing forensic work for legal clients. Many of these were so-called subrogation cases, in which insurance companies argued with one another of over who owed what in the aftermath of disastrous losses. Lafever's specialty was fire. "We'd assemble a small team, build a full-scale reproduction, instrument it, set it on fire, and see what happened."

At six o'clock in the evening on August 19, 1980, a Saudia Airlines Lockheed 1011 TriStar loaded with 287 passengers and a crew of 14 took off from Riyadh's King Khalid International Airport on a routine flight to Jeddah. Seven minutes into the flight, while still climbing, alarms went off and passengers spotted smoke in the rear of the plane. The plane turned around and headed back to the airport. After 10 more minutes there was fire in the cabin, passengers were fighting in the aisles, and the throttle to one of the three engines was stuck on open. The captain landed the plane successfully and brought it to a stop, but by the time ground rescue teams could get into the plane, the interior was fully engulfed in flames and everyone aboard was dead.

"For years there was kind of an urban myth that some passengers had built a campfire in the aisle to make tea," Lafever remarks. "That wasn't the case." His team built a 40-foot duplicate section of the rear of the plane from steel instead of aluminum so they could burn it more than once. They investigated all the ways a fire could start in the aft cargo bay of a climbing L-1011. One possibility involved shifting baggage.

"That's when I learned something railroad baggage handlers had known about for years; they call it 'hot baggage,' which is nothing more than luggage with matchbooks in it, the kind of souvenir matchbooks a traveling salesman might acquire. I got so I could start a fire in a suitcase with matchbooks in it on command, just by stomping in the right place." Lafever had not long finished with the L-1011 investigation when disaster struck again, closer to home.

Lafever used full-sized models to determine the possible causes of major disasters.

On the morning of 21 November, a fire started in the Deli Restaurant of the MGM Grand Hotel in Las Vegas, Nevada. Before the hotel could be evacuated, 84 people had died and 679 were injured. According to the official report of the Clark County Fire Department, "not one, but several factors were present which contributed to the cause of the fire ... the primary cause was electrical."

It's the sort of opinion that leaves lots of room for insurance companies to argue. Lafever and his crew built exact copies of the Deli Restaurant and burned every one of them, documenting every possible combination of factors that could have led to the fatal results. In the process of these and other investigations, Lafever learned that "a full scale model shows you what really happens; extrapolating from small models just creates extra work."

Moreover, "small teams of sharp people can outrun any large company." He and his crews learned from watching movie companies at work; like forensic recreations, movies are projects with a defined scope and the same need for highly skilled people working quickly. Speed was the watchword of Lafever's assignments, 21 days from go‑ahead to results, with everybody working "as long as we could without hurting ourselves."

Meanwhile, adding creativity to destruction, Lafever rented a studio at William Kreysler's custom fabrication shop in San Rafael, where George Lucas's Industrial Light and Magic was also located. While at Kreysler & Associates he created or cast duplicates of large architectural pieces for restoring old buildings and other construction projects and made technically complex, museum-quality replicas of sculptures for museums around the world.

"The guys from Star Wars came over one day. We got to talking and I was impressed; ILM really knew this stuff. I built them a production line for what they called 'crackers,' plastic panels that looked a little like circuit boards, which they used on the insides and outsides of spacecraft at different scales — the textures that made all those sets look complicated and futuristic."

The most fun came "when they asked me to build the power core of the Death Star." He designed and built the large scale model, and at the climax of the action sequence "they blew it to smithereens"; Lafever now possesses the only remaining piece of the Death Star within several galaxies.

A new way to design

"By the late 1980s I was kind of running out of steam and ready to retire" — for the second time — "but I got a call from Craig Peters at LBL inviting me to come over for a while and learn CAD. The Lab had powerful HP computers; I'd been using a Commodore. I jumped at the chance."

	The SNAP satellite has been among Lafever's longest-lasting design projects, and one of the most rewarding.

He's been here ever since, "privileged to be first on the scene on some important design projects, like the BaBar silicon detector. SNAP has been one of the longest running, and one of the most enjoyable," although there is no end to challenging excursions into new terrain. Lafever and a small group of scientists and technicians built a rapid operational prototype of a new kind of detector for finding and characterizing unexploded ordnance, recently field tested at the U.S. Army's Yuma Proving Ground with excellent results.

No doubt there's a personal angle to Lafever's enthusiasm for this kind of device. His son, William Sean, is a senior master sergeant with the 82nd Airborne Division, currently home at Fort Bragg with his wife and daughter after earning a Purple Heart and a Bronze Star during a tour in Afghanistan and two tours in Iraq.

Robin Lafever's enthusiasm for models has never waned; if he has his way, he'll soon be building a full-scale model of the whole SNAP satellite.

"Rapid prototyping gets all the bad actors to come out early," he says, and that's not the only advantage. "Models are an unsurpassed communications and sales tool. The PR value, when you're talking to the nontechnical people who have the money, can be worth the whole project." 

Additional information

• More about SNAP, with links to many design drawings and animations
• More about Berkeley Lab's Engineering Division
• More about Guillermina Oriental Arts and Antiques