Do Sparrows Like Bach?: The Strange and Wonderful Things that Are Discovered When Scientists Break Free

by Unknown

  Rocket science has a deserved reputation for being tough, so it should be no surprise that things can get a little bumpy when it comes to designing and testing a launcher. Just ask Elon Musk, the millionaire founder of PayPal and rocket company SpaceX.

  His first rocket, Falcon 1, was scheduled to lift off from the Pacific atoll Kwajalein in November 2005, at which point the problems began. First, unplanned engine tests used up more liquid oxygen (LOX) fuel than expected, then the LOX generator broke down. A shipment of fuel was ordered from Hawaii, but the tanker sprang a leak and arrived only one-fifth full. There was just enough to launch, but a valve was left open during the final preparations, allowing more fuel to waste away, so the launch had to be cancelled.

  A fresh shipment of LOX arrived a month later but disaster struck again. High winds hit the atoll, and to be on the safe side, engineers decided to drain the fuel from the rocket. A faulty pressure valve caused a vacuum to form inside the main fuel tank, sucking in its soft sides like a crushed beer can.

  After three months of repairs, by March 2006 Falcon 1 was ready to go. Seconds after launch, however, the main engine sprang a fuel leak, leaving a trail of flame in the rocket’s wake as it spiralled off course and crashed within sight of the disappointed engineers. An investigation pinpointed the cause: a corroded fuel-line nut was to blame.

  A year later, with the rocket rebuilt from scratch, Falcon 1 finally took off without a hitch. It managed 5 minutes of smooth flight. But a bump as the first and second stages separated confused a control system, causing it to enter an uncontrolled roll, which triggered a premature shutdown of the second-stage boosters. Falcon 1 did reach space, but not with the velocity needed to secure orbit.

  Its next outing was scheduled for January 2008, when it was to carry the cremated remains of 125 people, including actor James Doohan—Star Trek’s Scotty.

  Unfortunately, when Scotty finally blasted off on 2 August 2008, Falcon 1 failed again, bringing his ashes down along with those of 207 other cremated people. Undeterred (or perhaps uninterred), the rocketeers intend to keep on trying. At least they are not alone…

  Schoolkid blunder brought down Mars probe

  In 1999, NASA lost its $125 million Mars Climate Orbiter spacecraft as a result of a mistake that would shame a first-year physics student—failing to convert imperial units to metric. The problem arose from a culture clash between spacecraft engineers and navigation specialists, said Mary Hardin, a spokeswoman for NASA’s Jet Propulsion Laboratory in Pasadena. ‘Propulsion people talk in pound-seconds of thrust and navigators talk in newton-seconds,’ she said.

  The spurious data came from the craft’s attitude-control system, a design which had worked fine on the Mars Global Surveyor. But there was one crucial difference in the system on the orbiter. ‘There was a different propulsion supplier for the Mars Climate Orbiter, and its data package was in imperial units,’ said Noel Hinners, vice-president for flight systems at Lockheed Martin Astronautics in Denver, Colorado. No one adapted this data-processing software for the second probe, so JPL’s navigation software thought the numbers it received were newton-seconds rather than pound-seconds. The attitude thrusters only made small corrections, but the error was enough to leave the probe 100 kilometres too close to Mars when it tried to enter orbit.

  It’s not just space scientists, of course, who screw up big time. The air force, the navy, they’re all at it…

  Back to the future

  In 2007, 12 F-22 Raptors, the US Air Force’s new stealth fighters, left Hickam Air Force Base in Hawaii, bound for Okinawa, Japan, on the high-tech planes’ first overseas outing. Things went smoothly until they reached the 180th meridian—otherwise known as the International Date Line.

  Some of the pilots suddenly found themselves without any navigation aids. With nothing to tell them their compass heading or even whether they were level or not, it was as if the pilots had been instantaneously transported from the cockpit of the world’s most advanced aircraft into that of one dating from the First World War. Fortunately, the skies were clear, so the squadron did an about-face and was able to follow its in-flight refuelling tankers back to Hickam.

  The error was diagnosed as a problem with a ‘partial line of code’ that had pitched the planes’ computers into an infinite loop of trying and failing to calculate their position while dealing with an unexpected date. A fix was issued, and 3 weeks later the planes made their trip to Japan without a hitch.

  ‘Reliance on electronics has changed the flight-test process,’ said Donald Shepperd, once head of the US Air National Guard. ‘It used to be tails falling off, now it’s typos that ground a fighter.’

  The torpedo is not for turning

  In 1993, a torpedo owned by the Institute of Antarctic and Southern Ocean Studies at the University of Tasmania disappeared after heading in the wrong direction under the ice off Antarctica.

  The torpedo was a research tool taken to Antarctic waters on board the Aurora Australis, the research vessel of the Australian Antarctic Division. The ship returned to Hobart without the torpedo and the A$35,000 worth of computer equipment it carried. On its maiden voyage, the torpedo failed to return as it was programmed to.

  The obsolete weapon was bought from the Royal Australian Navy for A$150 and fitted with sonar equipment. The aim was to bounce signals off the bottom of the ice pack to measure its depth. It was programmed to do a U-turn under the ice before being retrieved. In open water, the device performed perfectly, but under the ice it failed the test. ‘There was one bright spot,’ said Garth Paltridge, director of the institute. ‘Part of the exercise was to test the logistics of operating in such a hostile environment. In that sense it worked perfectly.’

  After being lowered into the water and before its motor started, the waves turned the torpedo about 90 degrees from its course. Its guidance system never brought it back on track. Also, it was so cold, about -20 °C, that the ship’s tracking system stopped working.

  Yet, while it’s clear other organisations also make mistakes, NASA can often be seen lurking in the background.

  Memory scramble

  They say ballooning is the least stressful way to fly. Indeed, a balloon seemed the perfect platform from which the $10 million BLAST telescope, funded by NASA, the Canadian Space Agency and the UK’s Particle Physics and Astronomy Research Council, could take far-infrared snaps of star formations.

  For 12 days running up to 2 January 2007, the telescope collected valuable data as it floated 40 kilometres above Antarctica. As it descended, the gondola released the huge balloon and deployed its landing chutes as planned. But the electronics that should have released the parachutes on touchdown failed. Antarctic winds inflated them like giant spinnakers, turning the cargo into a wind-powered sled.

  ‘It was moving as fast as I could run,’ recalled project leader Mark Devlin, who was following the fate of the 1800-kilogram telescope and its support computers from his home in the US. ‘It was absolutely sickening.’ The support plane could only watch as the gondola bounced across the ice, strewing pieces of equipment as it went. It finally came to rest 24 hours later, when it wedged in a crevasse 200 kilometres from its landing site.

  Devlin was emailed a picture of the scene, which he scrutinised for any sign of the hard drives which bore the only copy of the mission’s data. ‘NASA paints everything white,’ he said, so his search was initially in vain. Fortunately, a pilot tracking the furrow gouged by the gondola spotted the package. The damaged drives eventually yielded their irreplaceable data, but the telescope was a write-off. Devlin then planned to fund-raise for a similar mission, with tougher electronics and one other change: ‘I’m thinking fluorescent orange,’ he mused.

  Robot’s return

  In 1993, NASA robotic experts suffered a setback when an experimental robot broke down during a highly publicised demonstration in an Antarctic volcano. Following that, experts planned to try again, with an improved version of the robot and in a different volcano. I
f all went according to plan, Dante II was to explore the crater of Mount Spurr, a volcano in Alaska. The eight-legged robot was studded with video cameras, a laser range finder and sensors that would guide its motion.

  Dante II also had an improved communications tether linking it to its human overlords. The tether on the original Dante broke during its test at Mount Erebus in Antarctica in January 1993. The project was intended to test and demonstrate technologies that could be used in remotely controlled space probes. Dante II was to descend about 600 feet into the volcano’s crater and record temperatures and the make-up of gases there.

  So, what happened on its second expedition?

  Dante rescued from crater hell

  NASA’S accident-prone robot Dante II was hoisted out of an Alaskan volcano in 1994, more than a week after it lost its footing and tumbled into the volcano’s mouth.

  Dante II’s Alaskan mission made a promising start. The robot trekked down into the crater of Mount Spurr and sampled gas from an open vent. But as Dante was climbing back out of the crater, it lost its footing in the slippery volcanic mud. It fell on its side and, like a giant beetle, was unable to move.

  In the first rescue attempt, a helicopter tried to lift Dante out of the crater by the fibre-optic cable linking the robot to its controllers. The cable snapped, and Dante fell further into the volcano. Further helicopter rescue missions were hampered by fog. Eventually, two workers climbed down into the crater and attached a harness to the robot’s body so that Dante could be winched out by helicopter.

  Things didn’t get much better, did they? Dante was behaving like the embarrassing drunk uncle at the family party. And scientists know full well that alcohol and serious research sometimes don’t mix.

  Particle beams hit the bottle

  Physicists starting up Europe’s most powerful particle accelerator in 1996 hoped to detect fundamental particles not seen since the beginning of the universe. But when the Large Electron Positron (LEP) collider at the CERN particle physics laboratory in Geneva resumed operation after a £210 million upgrade, nothing happened. The fault, it emerged, was due to two empty bottles of lager.

  LEP sits in a 27-kilometre circular tunnel. In the 1980s, physicists used the accelerator to prove the existence of the W and Z particles, which carry the weak nuclear force. With the energy of LEP’s particle beams nearly doubled to 162 giga-electronvolts, the accelerator had a chance of uncovering the Higgs boson, which was thought to endow other particles with mass.

  When LEP2, as the upgraded machine was known, was switched on, beams of electrons and positrons were supposed to travel in opposite directions round the giant circuit, but neither beam made it. CERN’s investigators homed in on the section of the ring causing the trouble. When they dismantled it, they found that the high-vacuum tube through which the beams travel had been blocked by two beer bottles.

  No damage had been done to LEP2. But the incident had a scientific cost. ‘We lost five valuable days of physics,’ said Stephen Myers, who was in charge of LEP2.

  A matter of perspective

  It’s reassuring to know that an engineering screw-up doesn’t always get you into trouble. It can sometimes even dig you out of it, as astronomers found out at the Canada-France-Hawaii Telescope (CFHT) on Mauna Kea, Hawaii.

  In early 2003, researchers began observing the skies using the telescope, equipped with a new digital camera—the biggest in the world at the time. The CFHT was fitted with four precision lenses so the camera could capture crisp images of vast areas of sky.

  The results were disappointing. The images were sharp in the centre, but far more blurred than expected at the sides.

  A laborious investigation followed, with engineers dismantling the optics and reassembling them daily, but finding no answer. Then one day, an engineer mistakenly replaced one of the four lenses back-to-front. The images improved spectacularly.

  ‘The next observations were just “Wow!”,’ said Christian Veillet, the observatory’s director. ‘The image quality was just what it should have been.’

  No one understood why the back-to-front lens worked so well, or why it didn’t work when it was oriented as planned. ‘That has been frustrating, but it would be a waste of resources to investigate, so we decided to just forget about it,’ said Veillet. ‘Now the science that is coming out is exquisite.’

  To be fair, it’s easy to criticise. How many great ideas have you had today? They too may have ended up in a crater like Dante, or an Antarctic crevasse like BLAST. So thank your lucky stars that no one has taken yours seriously. Somebody obviously thought it reasonable to take these researchers from the University of Oklahoma seriously, and look what happened. Whoever thought it was a good idea was just plain, unacceptably stupid.

  Elephants on acid

  What happens if you give an elephant LSD? Researchers solved this mystery in August 1962, when Warren Thomas, director of Lincoln Park Zoo in Oklahoma City, fired a cartridge-syringe containing 297 milligrams of LSD into the rump of Tusko the elephant. With Thomas were two colleagues from the University of Oklahoma School of Medicine, Louis Jolyon West and Chester M. Pierce.

  The dose was about 3000 times what a human would typically take. Thomas, West and Pierce figured that if they were going to give an elephant LSD they’d better not give it too little. They later explained that the experiment was designed to find out if LSD would induce musth in an elephant—musth being a kind of temporary madness male elephants sometimes experience during which they become highly aggressive and secrete a sticky fluid from their temporal glands. One may also suspect a small element of ghoulish curiosity was involved.

  Whatever the reason for the experiment, it almost immediately went awry. Tusko reacted as if he had been shot by a gun. He trumpeted around his pen for a few minutes and then keeled over. Horrified, the researchers tried to revive him with a variety of antipsychotics, but about an hour later he was dead. In an article published four months after the event in Science, the three scientists sheepishly concluded: ‘It appears that the elephant is highly sensitive to the effects of LSD.’

  The experiment instantly made headlines. Faced with a public relations disaster, the scientists protested their innocence. They had not anticipated the elephant would die, they insisted. In their experience, LSD was a powerful hallucinogen but rarely fatal. West and Pierce helpfully noted that they themselves had previously taken the drug.

  Thomas tried to find a silver lining. They had learned that LSD can be lethal to elephants. So perhaps, he mused, the drug could be used to destroy herds in countries where they are a problem. For some reason, his suggestion has never found any takers.

  We can only hope that the Oklahoma researchers copped all the opprobrium they were due, but for some people—despite their own stupidity—you have to feel a little sympathy. The following story is, we freely admit, not a tale of a researcher being given free rein. But it does illustrate a blunder in the use of technology that we have all surely come across ourselves. How many times have you hit ‘Reply All’ to an email and then thought ‘Aaaaaaaaargh’?

  Naming and shaming

  In 1993, the NatWest Bank admitted that it kept personal information about its customers—such as political affiliation—on computer. But Computer Weekly revealed that a financial institution, sadly unnamed, went one better and moved into the realm of personal abuse.

  The institution decided to mailshot 2000 of its richest customers, inviting them to buy extra services. One of its IT workers wrote a computer program to search through its databases and select the customers automatically. He tested the program with an imaginary customer called Rich Bastard.

  Unfortunately, an error resulted in all 2000 letters being addressed ‘Dear Rich Bastard’. The luckless programmer was subsequently sacked.

  Target practice

  A good story deserves a good airing. Back in the 1990s, two members of the Lothian and Borders traffic police were having a pleasant time out on the Scottish moors, trapping speedi
ng motorists with a radar gun. Suddenly their equipment went crazy, registering a speed of more than 300 miles per hour. It then locked up completely.

  Seconds later, the startled cops understood why, as a low-flying Harrier jet screamed over their heads. Upset that their radar gun had been broken, the policemen put in a complaint to the Royal Air Force—only to discover that the damage could easily have been so much worse.

  The RAF informed them that the Harrier’s target-seeker had locked on to what it had interpreted as enemy radar. This immediately triggered an automatic air-to-surface missile attack. Fortunately for the two policemen, the Harrier was operating unarmed.

  The policemen didn’t blunder, neither did the fighter pilot. But a software developer must now be wiping a brow. They might want to compare notes with the designer of the Swedish navy’s early-warning system.

  Furry submarines embarrass Swedish navy

  When is a sub not a sub? The answer, as an embarrassed Swedish navy had to admit in 1996, is when it’s a mink or an otter. What the navy thought was the sinister sound of Soviet propellers was, in fact, the furious paddling of little legs.

  A scientific commission set up by the government and chaired by the former director of the Swedish Engineering Science Academy, Hans Forsberg, concluded that most of the invading submarines reported by the navy were mythical. Of more than 6000 reports of ‘alien underwater activity’ between 1981 and 1994, the commission found firm evidence for only six incidents.

  In every other case the evidence, often based on sightings by the public, was unreliable. The navy claimed that one underwater noise, similar to the sound of frying eggs, was caused by bubbles from submarines. It was more likely to have been caused by the natural movement of water, said the commission.