C ALIFORNIA

INSTITUTE OF

TECHNOLOGY engineeringE& andS science

Volume LXIII, Number 4,    

IN THIS ISSUE

Living with Faults

Sailing with Light

Cooking with Oil Tensile cracks and fissures W E Pre-1999 surface concrete fence Pre-1999 ground surface 2 m plastic foil Slumping deposits 1 m 1 m

0 ? 0

-1 m -1 m

-2 m 2 m -2 m

0 2.0 4.0 6.0 8.0 10.0

cultivated yellowish fine- to brownish yellowish medium to cobbly gray soil medium-grained sand fine-grained mud coarse-grained sand gravel

Kerry Sieh and his colleagues dug this trench (on cover and at left) exposing the Chelungpu fault in Wufeng, Taiwan, a few months after the fault broke and produced the devastating earthquake of September 1999. At the fault, the earth was shortened a couple of meters and the right side rose more than a meter; the fault’s motions are shown by the red arrows in the drawing above. The red fence in the background was a temporary replacement for a shattered concrete wall. The river gravel at the bottom of the trench shows that the area has also been subject to river floods—one more reason not to build there. See the story beginning on page 8. Cover photo and drawing courtesty of Jian-Cheng Lee of the Academia Sinica in Taiwan.

2 ENGINEERING & SCIENCE NO . X     engineering& and science C a l i f o r n i a Institute

of Technology Volume LXIII, Number 4, 2000 E S

2 Random Walk

8 Acts of God, Acts of Man: How Humans Turn Natural Hazards into Disasters — by Kerry Sieh Earthquakes, tsunamis, and landslides may be unpredictable, but the places where they are likely to occur are not. We now know where not to build, but will we?

18 Solar Sailing: The Next Space Craze? — by Joel Grossman Spacecraft that don’t need rocket fuel riding laser beams to Alpha Centauri at one-tenth the speed of light. Sound nuts? Maybe not.

30 In Defense of Robert Andrews Millikan — by David Goodstein Was he a fraud, a chauvinist, and an anti-Semite?

Departments

39 Obituaries: Brad Sturtevant, J. Harold Wayland

43 Faculty File

Engineering & Science (ISSN 0013-7812) is published Blair A. Folsom quarterly at the California Institute of Technology, 1200 President of the Alumni Association East California Boulevard, Pasadena, CA 91125. Annual J. Ernest Nunnally subscription $10.00 domestic, $20.00 foreign air mail; Vice President for Institute Relations Robert L. O’Rourke single copies $3.00. Send subscriptions to Caltech 1-71, Associate Vice President for Institute Relations Pasadena, CA 91125. Third class postage paid at Pasade- na, CA. All rights reserved. Reproduction of material con- STAFF: Editor — Jane Dietrich tained herein forbidden without authorization. © 2000, Managing Editor — Douglas Smith California Institute of Technology. Published by Caltech Writer/Editor — Barbara Ellis and the Alumni Association. Telephone: 626-395-3630. Contributing Writers — Jill Perry, Robert Tindol PICTURE CREDITS: 2, 5, 7, 39, 44 – Bob Paz; 3, 16, 21, 22, 24, Copy Editors — Emily Adelsohn, 25, 29 – NASA; 10, 20 – Doug Cummings; 10, 11 – C. T. Lee; Michael Farquhar, Elena Rudnev 11, 13, 15, 17 – Kerry Sieh; 12-13 – Aykut Barka; 13 – Timothy Business Manager — Debbie Bradbury Dawson; 14-15 – Earth Consultants International; 16 – Charles Circulation Manager — Susan Lee Rubin; 20 – SOHO, NOAA; 21, 26 – ESLI; 23 – Richard Photographer — Robert Paz Dowling/WSF, Louis Friedman, Planetary Society; 24 – Carnegie Graphic Artist — Douglas Cummings Mellon; 24 – DLR; 27 — RPI; 28 – Energia, Ltd.; 32, 34-35, 37 – Caltech Archives; 36 – American Scientist; 41-42 – Floyd Clark Visit Caltech on the Web at http://www.caltech.edu Random Walk Elachi (left) and Stone at the press conference.

T HE L AB N AMES O NE OF I TS O WN

Charles Elachi (MS ’69, and so knows the school well. of Caltech, and I intend to the Space Shuttle that have PhD ’71) has been named He is an expert in remote continue that tradition. My allowed scientists to see the new director of the Jet sensing, and in recognition commitment is to continue through the clouds that Propulsion Laboratory (JPL), of his work, he was one of the tradition of excellence blanket Earth. (The technol- which Caltech manages for the youngest members ever and boldness in exploring our ogy also penetrates the top NASA, effective May 1. elected to the National solar system, understanding layer of soil in arid regions, President David Baltimore Academy of Engineering. the origin of galaxies, and revealing hints of what lies made the announcement at He has long been a leader of applying that knowledge to below.) He has participated a press conference on January planetary exploration at JPL better understand the changes in a number of archaeological 31, where Elachi and Balti- and is widely respected at the on our own planet.” The new expeditions in the Egyptian more were joined by retiring Laboratory. I look forward to post brings Elachi full circle, desert, the Arabian peninsula, JPL director Edward Stone, having a close working as he recalled being inspired and the Western Chinese the Morrisroe Professor of relationship with him.” as an 11-year-old in Lebanon desert, using satellite data to Physics; and NASA adminis- “Charles Elachi brings by JPL’s launching of Ex- search for old trading routes trator Daniel Goldin. Elachi formidable talents to his new plorer 1—43 years ago to the and buried cities. Some of has served in a variety of job, as both a scientist and a day, he noted. “Maybe that’s these expeditions have been research and management leader,” said Goldin. “In a good omen for me,” he featured in National Geo- positions at JPL since 1971. addition to already being joked. He grew up to receive graphic, on PBS, and in Most recently, he has been responsible for many of JPL’s a BSc in physics from the Caltech News (“The Road head of the Space and Earth missions in solar system University of Grenoble, to Ubar,” April, 1992). He Science Programs; other exploration, Earth sciences, France, and the Dipl.Ing. in has also served as principal positions include manager for and astrophysics, he has led engineering from the Poly- investigator on numerous radar development and leader efforts to create road maps technic Institute, Grenoble, NASA research and develop- of the radar remote-sensing of our exploration strategies both in 1968, and then ment studies and flight team. decades into the future. He earned his Caltech MS and projects. He is currently the Elachi “knows JPL better is both an effective adminis- PhD in electrical engineering. team leader of the Cassini than anyone and will be best trator and a visionary.” He also earned an MBA from Titan radar experiment and a able to lead the Laboratory in Elachi said he was honored USC in 1978, and an MS in coinvestigator on the Rosetta the coming years,” Baltimore to be entrusted with the geology from UCLA in 1983. Comet Nucleus Sounder said. “Charles has an extraor- leadership of JPL. “For the Elachi is perhaps best Experiment. He is the author dinary record of accomplish- last 40 years JPL has enjoyed known for his role in the of more than 200 publica- ment in his 30 years at JPL. a tradition of excellence as a development of a series of tions on space and planetary He is an alumnus of Caltech, NASA center and division imaging radar systems for exploration, Earth observa-

2 ENGINEERING & SCIENCE NO . 4     tion from space, active microwave remote sensing, wave propagation and scattering, electromagnetic theory, lasers, and integrated S NIFF M E A T UNE optics, and he holds several patents in those fields. He has written three textbooks on remote sensing and has taught EE/Ge 157, Introduc- When Hamlet told the iorally relevant to them. tion to the Physics of Remote courtiers they would eventu- But the study likely applies Sensing, since 1982. ally “nose out” the hidden to other animals, including In 1988, the Los Angeles corpse of Polonius, he was humans, because the olfactory Times selected him as one of perhaps a better neurobiolo- systems of most living crea- “Southern California’s rising gist than he realized. Accord- tures appear to follow the stars who will make a differ- ing to research by Caltech same basic principles. After ence in L.A.” In 1989, neuroscientists, the brain placing electrodes in the Asteroid 1982 SU was creates subtle temporal codes brains of individual fishes, renamed 4116 Elachi in to identify odors. Some neu- they were subjected to recognition of his contribu- ral signals change over the sequences of 16 odor compo- tions to planetary exploration. duration of a sniff, giving nents found in foods they Elachi is the second Caltech first a general notion of the normally seek. Analyzing the alumnus to be named director type of odor, then a more signals from the mitral cells of JPL. The first, William subtle discrimination that showed that the information Pickering (BS ’32, PhD ’36), leads to precise recognition the fish could extract about a headed the lab from 1954 to of the smell. In the February stimulus became more precise 1976. ■—JP 2 issue of the journal Science, as time went by. The finding Gilles Laurent, associate was surprising because the professor of biology and signals extracted from the computation and neural receptor neurons located systems, and postdoc Rainer upstream of the mitral cells W. Friedrich, now at the Max showed no such temporal Planck Institute in Heidel- evolution. “It looks as if the berg, Germany, report that brain actively transforms certain neurons respond to an static patterns into dynamic odor through a complicated ones and in so doing, man- process that evolves over a ages to amplify the subtle brief period of time. These differences that are hard to neurons, called mitral cells perceive between static because they resemble patterns,” Laurent says. miters—the pointed hats “Music may provide a use- worn by bishops—are found ful analogy. Imagine that the by the thousands in the olfactory system is a chain of human olfactory bulb. choruses—a receptor chorus, “We’re interested in how feeding onto a mitral-cell ensembles of neurons encode chorus, and so on—and that sensory information,” each odor causes the receptor explains Laurent. “So we’re chorus to produce a chord. On February 12, NEAR Shoemaker became the first spacecraft to land on an less interested in where the Two similar odors evoke two asteroid—all the more impressive when you consider that this legless relevant neurons lie, as very similar chords, making revealed by brain mapping discrimination difficult. orbiter was never designed to land on anything. NASA’s Near Earth Asteroid studies, than in the patterns What the mitral-cell chorus Rendezvous mission, which was renamed in honor of the late Eugene of firing these neurons pro- does is to transform each Shoemaker (BS ’47), the father of planetary geology, had been in close orbit duce and in figuring out from chord it hears into a musical around the 21-mile long Eros for a year. The craft touched down at a these patterns how recogni- phrase, in such a way that the tion, or decoding, works.” difference between these gentle four miles an hour and in an orientation that allowed its solar The researchers used zebra phrases becomes greater over panels to continue to function, so jubilant scientists turned its gamma-ray fish because these animals time. In this way, odors that spectrometer back on to get a close-up analysis of Eros’s surface mineral- have comparatively few initially ‘sounded’ alike pro- ogy. In this image mosaic, taken from an orbital altitude of 200 kilometers, mitral cells and because much gressively become more easily is already known about the identified.” the arrow points to the landing site. Gene’s gotta be happy…. types of odors that are behav- In other words, when we

    ENGINEERING & SCIENCE NO . 4 detect a citrus smell in a garden, for example, the odor is first conveyed by the receptors to the mitral cells. This initial firing allows for little more than the generic detection of the citrus nature of the smell. Within a few tenths of a second, however, new mitral cells are recruited, leading the pattern of activity to change rapidly. This quickly allows us to deter- mine whether the citrus smell is actually a lemon or an orange. T HE I NS AND O UTS OF O VER - AND U NDERVOTES However, the individual tuning of the mitral cells first stimulated by the citrus odor does not become more specif- ic. Instead, the manner in In December, following ballots with hand-marked study is complete, it will which the firing patterns the contentious vote counting votes, lever machines, punch encompass presidential unfold through the lateral in the presidential election, cards, optical scanning elections going back to 1980, circuitry of the olfactory bulb Caltech and MIT decided to devices, and direct-recording and will examine a finer is ultimately responsible for join forces to develop a voting electronic devices (DREs), breakdown of the different the fine discrimination of the system that will be easy to which are similar to auto- technologies, and a break- odor. “Hence, as the system use, reliable, secure, and matic teller machines. down of residual votes into evolves, it loses information modestly priced. The project The study focuses on its two components: over- and about the class of odors, but was the brainchild of the so-called “undervotes” undervotes. A final report becomes able to convey infor- institutions’ two presidents— and “overvotes,” which are will be released in June. mation about precise iden- Caltech’s David Baltimore combined into a group of The analysis is complicated tity,” says Laurent. ■—RT and MIT’s Charles Vest—and, uncounted ballots called by the fact that voting sys- with $250,000 funding from “residual votes.” These tems vary from county to the Carnegie Corporation, include ballots with votes county and across time. faculty from both campuses for more than one candidate, When a system is switched, began collecting data and with no vote, or that are say from lever machines to studying the range of voting marked in a way that is DREs, the number of residual methods across the nation. uncountable. votes can go up due to voter Early in February, the Cal- Careful statistical analysis unfamiliarity with the new tech/MIT Voting Technology shows that there are system- technology. Project submitted a prelimi- atic differences across tech- “We don’t want to give nary report to the task force nologies, and that paper the impression that electronic studying the election in ballots, optical scanning systems are necessarily Florida. Their nationwide devices, and lever machines inaccurate, but there is much study of voting machines have significantly lower room for improvement,” said offers further evidence residual voting rates than Thomas Palfrey, Caltech supporting the task force’s punch-card systems and professor of economics and call to replace punch card DREs. Overall, the residual political science. voting in Florida. The voting rate for the first three “Electronic voting technol- statistical analysis also systems averages about 2 ogy is in its infancy and uncovered a more surprising percent, and for the last two seems the most likely one finding: electronic voting, as systems averages about 3 to benefit significantly from currently implemented, has percent. new innovations and in- performed less well than was This study is the most creased voter familiarity,” widely believed. extensive analysis ever of the states the 13-page report. The report examines the effects of voting technology Other Caltech members effect of voting technologies on under- and overvotes. of the Voting Technology on unmarked and/or spoiled The study covers the entire Project are Michael Alvarez, ballots. Researchers from country for all presidential associate professor of political both universities are elections since 1988, and science, and Jehoshua Bruck, collaboratively studying five examines variations at the professor of computation and voting technologies: paper county level. When the neural systems and electrical

4 ENGINEERING & SCIENCE NO . 4     T HE J UPITER E FFECT

In a study that strengthens Saturn—is no longer present ized” by impacts. “A comet the likelihood that solar sys- in the sun’s stellar vicinity in the size of Hale-Bopp, for tems like our own are still sufficient quantities to form example, would vaporize being formed, an interna- new planets. much of Earth’s oceans if it tional team of scientists has “We looked at only three hit there. The impact from a reported that three young stars, but the results could 500-kilometer object—about stars in the sun’s neighbor- indicate that it’s easier to ten times the size of Hale- hood have the raw materials make Jupiter-sized planets Bopp—could create nearly necessary for the formation than previously thought,” 100 atmospheres of rock of Jupiter-sized planets. Data said Geoffrey Blake (PhD vapor, the heat from which engineering. Participating obtained from the European ’86), professor of cosmochem- can evaporate all of Earth’s from MIT are Stephen Space Agency’s Infrared Space istry and planetary sciences oceans.” Ansolabehere, professor of Observatory (ISO) indicate and professor of chemistry at The researchers did not political science, and Nicho- for the first time that molec- Caltech and the correspond- directly detect any planets las Negroponte, chairman of ular hydrogen is present in ing author of the study, which in the study, but nonetheless the Media Lab. the debris disks around young made the cover of Nature on found that molecular hydro- Check the web at nearby stars. The results are January 3. “There are over gen was abundant in all three www.vote.caltech.edu important because experts 100 candidate debris disks disks. In the disk surround- for further information, or have long thought that within about 200 light-years ing Beta Pictoris, a Southern see www.vote.caltech.edu/ primordial hydrogen—the of the sun, and our work Hemisphere star that formed Reports/report1.pdf for the central building block of gas suggests that many of these about 20 million years ago report itself. ■—JP giants such as Jupiter and systems may still be capable approximately 60 light-years of making planets.” from Earth, the team found The abundance of Jupiter- evidence that hydrogen is sized planets is good news, present in a quantity at least though indirectly, in the one-fifth the mass of Jupiter, search for extraterrestrial life. or about four Neptune’s A gas giant such as Jupiter worth of material. The debris may not be particularly around 49 Ceti, which lies hospitable for the formation near the celestial equator in of life, but experts think the the constellation Cetus, was mere presence of such huge found to contain hydrogen in bodies in the outer reaches of a quantity at least 40 percent a solar system protects small- of the mass of Jupiter. er rocky planets like Earth Saturn’s mass is just under a from catastrophic comet and third that of Jupiter. 49 Ceti, meteor impacts. A Jupiter- which is about 10 million sized planet possesses a years old, is roughly 200 gravitational field sufficient light-years from Earth. Best to kick primordial debris into of all was a 10-million-year- the farthest reaches of the old Southern Hemisphere star solar system, as Jupiter has about 260 light-years away presumably done by sending that goes by the rather perhaps billions of comets unpoetic name HD135344. Alan Alda is Richard Feynman in QED, a new play based on Tuva or Bust! safely away from Earth into That star’s debris disk was the Oort Cloud, which lies found to contain the equiva- and other Feynman tales. Fittingly, the world premiere is at the Mark Taper beyond the orbit of Pluto. If lent of at least six Jupiter Forum, just down the Pasadena Freeway from Feynman’s old haunts. It runs comets and meteors were not masses of molecular through May 13; tickets are available at www.TaperAhmanson.com or (213) ejected by gas giants, Blake hydrogen. 628-2772. On a recent visit to campus to soak up the atmosphere, Alda said, life on Earth (and any The study also confirmed other Earth-like planets) that planetary formation is dined with president Baltimore and assorted campus luminaries. could periodically be “steril- not limited to a narrow

    ENGINEERING & SCIENCE NO . 4 “window” early in the life of a star, as previously thought. O F C ELL P HONES , M EMORY C ELLS , AND Because molecular hydrogen F LASHY N ANOCRYSTALS is quite difficult to detect from ground-based observato- ries, experts have relied on measurements of the more easily detectable carbon monoxide (CO) to model the Scientists at Caltech and “flash” memory, which gas dynamics of developing Agere Systems, formerly continues to store informa- solar systems. But because known as the Microelectron- tion even when the device is the CO tends to dissipate ics Group of Lucent Tech- turned off. This information quite rapidly early on, nologies, have developed a could include personal phone researchers assumed that the technique that could result directories in a cellular phone molecular hydrogen likewise in a new generation of reli- or the pictures captured by vanished. This presumed able nanoscale memory chips a digital camera. A typical lack of hydrogen limited the and smaller, less expensive flash-memory chip stores 16 time in which Jupiter-sized cellular phones and digital to 32 million bits of data, planets could form. However, cameras. Announced Decem- with each bit in a separate the new study, coupled with ber 13 at the International “cell.” As chip sizes decrease, recent theoretical models, Electron Devices Meeting, the cells become more diffi- shows that CO is not a the work applies to so-called cult to make leakproof, and particularly good proxy for the total gas mass surround- ing a new star. Blake said the study opens new doors to the understand- ing of planetary growth processes around sun-like stars. He and his colleagues anticipate further progress when the Space Infrared ¿O YE , CHICOS, D ONDE E STÁ EL O BSERVATORIO ? Telescope Facility (SIRTF) and the Stratospheric Obser- vatory for Infrared Astronomy (SOFIA) are launched in 2002. SIRTF, which will Los Angeles-area high Sylmar, Van Nuys, and among the high schools. have its science headquarters school students will team Harvard Westlake High A large array of this type will at Caltech, alone could detect up with Caltech researchers Schools. enable the study of ultrahigh- literally hundreds of stars that to study ultrahigh-energy The research will be energy cosmic rays through still contain enough primor- cosmic rays on their own coordinated by Professor of the detection of “showers,” dial hydrogen in their debris campuses, thanks to a recent Physics Robert McKeown. several kilometers in radius, disks to form Jupiter-sized grant from the Weingart The program will also of secondary particles the rays planets. Foundation, which donated incorporate a high school create in the Earth’s atmo- The other authors are pro- $100,000 to establish the teacher education component sphere. These are the fessor Ewine F. van Dishoeck California HIgh school coordinated by Dr. Ryoichi highest-energy particles ever and Wing-Fai Thi, the Cosmic-ray ObServatory Seki at California State observed in nature, and thus study’s lead author, both of (CHICOS) on four campuses University, Northridge. are of great current interest in the Leiden University in the in the Northridge area initial- Teachers will develop the astrophysics and particle- Netherlands; Jochen Horn ly, expanding to at least 25 curriculum materials to help physics community. Thus, and professor Eric Becklin, and possibly hundreds of sites their students participate in while establishing a state-of- both of the UCLA Depart- eventually. Three of the four this research. Caltech will the-art experimental facility, ment of Physics and As- initial schools have a high host a summer workshop this project will provide an tronomy; Anneila Sargent number of students who where physics teachers and exceptional educational (MS ’67, PhD ’77), professor are underrepresented in students can participate in experience for local high of astronomy at Caltech; the sciences, which means the construction of new school students. When a Mario van den Ancker of the the program may assist in detector stations for deploy- majority of the 25 sites are Harvard-Smithsonian Center increasing the number of ment at additional sites. operating, it is expected that for Astrophysics; and Anto- future scientists in the United The detector hardware, the project will yield signifi- nella Natta of the Astrophysi- States. The schools are the associated electronics, and cant scientific results that cal Observatory of Arcetri Sherman Oaks Continuing computer equipment will will be reported in the in Florence, Italy. ■—RT Education School, and form a networked system scientific literature. ■—JP

6 ENGINEERING & SCIENCE NO . 4     stored data can be lost. professor of applied physics several advantages over the Using an aerosol technique and materials science, and conventional lithographic developed at Caltech, silicon project director. Atwater; techniques used to make nanocrystals were sprayed Richard Flagan, the McCol- today’s flash memory cells. through a bath of high- lum Professor of Chemical Because it requires fewer temperature oxygen to create Engineering; postdoc Mark steps, it is less expensive and memory cells comprised of Brongersma; grad students the chips take less time to silicon on the inside with a Elizabeth Boer (MS ’96), Julie produce. In addition, the silicon dioxide outer shell. Casperson, and Michele aerosol approach will allow The silicon nanocrystals store Ostraat (MS ’98); and Jan de researchers to continue mak- the electrical charge, whereas Blauwe and Martin Green at ing smaller and smaller the insulating silicon dioxide Agere Systems developed a devices. The cells are also shell makes the cells leak- method to break up each cell extremely robust—one cell resistant. “As compared to into 20,000 to 40,000 small- has gone through a million conventional flash memories, er cells. Therefore, even if charge-discharge cycles with- these silicon nanocrystal several of the smaller cells out significant degradation, memories offer higher perfor- spring a leak, the vast major- whereas 10,000 cycles is con- mance, simpler fabrication ity of the charge will not be sidered satisfactory for a tra- processes, and greater promise lost and the bit of data stored ditional chip. The research for carrying memory minia- in the whole memory cell will was supported by the Nation- turization to its ultimate be retained. al Science Foundation and limit,” said Harry Atwater, The aerosol approach has NASA. ■—RT

Caltech scored first, Caltech scored last, but That Other Institute of Technology scored more often in their first ever women’s basketball matchup in front of a packed house in Braun Gym on January 5. I N S EARCH OF THE M IND ’ S E YE Final score: CIT 46, MIT 80.

A study of patients await- surgically removed. human brain are involved The problem has been ing brain surgery has shown During their extended in memory and can respond difficult to address because that humans use the same hospital stay, the patients selectively to a wide variety the techniques that yield very neurons to conjure up mental were asked to look at photos of visual stimuli as well as precise results in animals are images that they use when of famous people, pictures of stimulus features such as generally not suitable for they actually see the real animals, abstract drawings, facial expression and gender. humans, and because the object. In the November 16 and other images. While According to Koch, the study brain imaging techniques issue of Nature, UCLA neuro- they were looking at the helps settle long-standing suitable for humans are not surgeon and neuroscientist images, the researchers noted questions about the nature of very precise, Koch says. Such Itzhak Fried and Caltech the precise neurons that were human imagery. Particularly, techniques can image only neuroscientists Christof active. Then, the subjects the research sheds light on large portions of the brain, Koch, professor of computa- were instructed to close their the process at work when each containing on the order tion and neural systems, and eyes and vividly imagine the humans see things with the of one million very diverse grad student Gabriel images. Again, the research- mind’s eye. “If you try to nerve cells. “Recording the Kreiman report on results ers noted which neurons were recall how many sunflowers activity of single cells allows obtained by questioning nine active. It turns out that a there are in the Van Gogh us to investigate the neuronal patients who had been fitted subset of neurons in the hip- painting, there is something correlates of visual awareness with brain sensors. The pocampus, amygdala, ento- that goes on in your head that at a detailed level of temporal patients, all suffering from rhinal cortex, and parahippo- gives rise to this visual and spatial resolution,” says severe epilepsy uncontrolled campal gyrus would fire when image,” Koch says. “There Kreiman. The work was with drugs, were being the patient looked at the has been an ongoing debate supported by the National observed for a period of one image and also when he about whether the brain areas Institutes of Health, the to two weeks so that the or she imagined the image. involved in perception during National Science Foundation, regions of their brains The results build upon ‘vision with your eyes’ are the and the Center for Conscious- responsible for their seizures work by Fried’s group show- same ones used during visual ness Studies at the University could be identified and later ing that single neurons in the imagery.” of Arizona. ■—RT

    ENGINEERING & SCIENCE NO . 4 Acts of God, Acts of Man: How Humans Turn Natural Hazards into Disasters

by Kerry Sieh Geologists have a particular appreciation of The dramatic increase in human population over Earth’s beauty. That’s not to say that those of you the past several decades has resulted in an enor- who are not geologists don’t appreciate it. We mous increase in the actual dollar losses and in the would probably all agree that the waterfall above loss of human life associated with natural disasters. is a beautiful thing. But some of the same things You see the same basic trend with earthquakes, that make Earth beautiful also make it dangerous, with landslides, with tsunamis, with almost every and to some degree it’s the danger in the beauty hazard—an almost exponential increase over the that attracts the geologist. I’m going to discuss last couple of decades. This trend will most likely the hazardous side, and I’m going to argue for a continue if we don’t change the way we act with different approach to handling the natural hazards respect to our planet’s behavior. that this beautiful Earth puts beneath our feet. Worldwide, about 200 cities with a population In the 20th century, and in fact in the first and of more than 500,000 lie within 100 kilometers of second millennia, we just reacted to natural di- known active faults; Los Angeles is one such city, sasters. We basically cleaned up after they hap- as are San Francisco, Wellington (New Zealand), pened and continued on as before, leaving our and Taipei (Taiwan). Ahmedabad (India) and San great-grandchildren to suffer the same fate, next Salvador only recently suffered large earthquakes. time around. Now, at the turn of the century, the Central America, the Mediterranean countries, the turn of the millennium, I suggest we start looking Middle East, and large parts of Asia are particu- at hazards differently than we have looked at them larly susceptible to earthquake damage. But we in the past. Let’s understand them as the geolo- associate very few of these 200 cities with earth- gist or the earth scientist does, so that they don’t quake hazards, because the Earth’s metabolism is destroy our cities, homes, and lives. Let’s actively so much slower than ours that most of them have reduce our exposure to hazards, rather than being not been devastated by an earthquake in living just reactive to them. memory. We think in terms of years or months,

8 ENGINEERING & SCIENCE NO. 4  Photo by Sheng-Tao Kuo Sheng-Tao Photo by

This is a tale of two countries, Taiwan and Turkey, and the devastating

earthquakes that hit them in 1999. When I look at the waterfall on

Taiwan’s Tachia River, I’m reminded of one of the reasons why I became a

geologist: to understand why this big escarpment rose in the middle of

this river is a joy and a delight.

not decades or centuries. We teach our kids not to cross the street without looking both ways, but we don’t teach them to worry about something that might happen in 50, or 100, or 500 years. Of the earthquakes that we’ve heard about in the last decade, none has a recurrence interval of less than about a thousand years, with the exception of the Turkish earthquake that I’ll be discussing. This low metabolic rate inures us to the fact that the faults are there, that earthquakes happen. Genera- tions come and go, thinking they are perfectly safe, when in fact they are living on a time bomb with a very long fuse. This is a tale of two countries, Taiwan and Turkey, and the devastating earthquakes that hit them in 1999. When I look at the waterfall on Taiwan’s Tachia River above, I’m reminded of one of the reasons why I became a geologist: to under- stand why this big escarpment rose in the middle of this river is a joy and delight. But if I were an The fault scarp, about seven meters high, could be seen clearly from the air, even without a engineer, looking at what happened to my bridge (right), my feelings would be very different. I’d broken bridge for emphasis. The earthquake’s lesson, however, was not learned; see page 13. be very upset. Before the earthquake, the geolo- (Photo provided by Jack, Yung-Wan Lien, Flying Tiger Photographic, Inc.) gist could have told the engineer: “You really

 ENGINEERING & SCIENCE NO. 4 9 (North) Converging boundary American plate Spreading boundary

Turkey Taiwan

Eurasian plate Persian Pacific plate subplate Arabian African plate plate Philippine plate Caroline plate Bismarck plate Somalian (South) subplate American Austral-Indian plate plate

Antarctic plate

300 30 60 90 120 150 180

Taiwan sits on the boundary of the large Eurasian plate and the Philippine Sea plate (above). As they converge, the Philippine Sea plate shouldn’t put your bridge abutment here because The fault that broke on September 21, 1999, rides up over the larger there’s a big fault under it. And when that fault the Chelungpu fault, runs along the western edge plate (right), crunching up moves, it moves with many meters of slip, and of the mountain front in the center of the island. the shallow marine you’re going to lose that $10 million bridge.” A well-known Princeton geologist, John Suppe, sediments into a moun- But the engineer, as is typical, didn’t talk to the determined its geometry in the early 1980s, using geologist until after the earthquake. And we’ll see borings and seismic reflection data collected for oil tainous island (far right). that, even then, the engineer wasn’t encouraged to exploration. The pictures on the opposite page Note also in the map rebuild with any accommodation for future fault give some idea of the damage the fault rupture above the collision of the ruptures. wrought. One side of the fault rode up and over Taiwan spans the boundary between a small the other, forming an escarpment several meters Arabian plate with the plate called the Philippine Sea plate and the great high; fields that once were flat are now three or Eurasian plate, which has Eurasian plate, which runs all the way from the four meters higher on one side than on the other; made Turkey subject to middle of the Atlantic and Iceland eastward to buildings and bridges across the fault were earthquakes. Japan. Taiwan used to be sediment sitting in the wrecked. shallows of the continental shelf on the edge of the Now, fields and rice paddies can be regraded. Eurasian plate. When the continental margin of Those are perfectly good things to have on a fault, the bigger plate started getting jammed down the but cemeteries are a different matter. The Taiwan- subduction zone that separates the two plates, the ese have great reverence for their ancestors, and it Philippine Sea plate began to ride up over the was of great concern that many graveyards were continental shelf, doubling up the sediments ripped asunder by this earthquake. It’s a serious between thrust faults to create the 300-kilometer- matter that they will have to consider in making long island that we see today. Not all of these new zoning laws after this earthquake. faults are active now, but the one that moved to Buildings don’t behave very well if they straddle produce the 1999 earthquake obviously is. The fault ruptures, either. Most didn’t actually col- mountainous mass along the eastern side of the lapse, though playing billiards in some of them island has risen in just a geological twinkling of would be a bit difficult now. Buildings very close an eye—4 million years or so. It’s starting to go to the fault but not directly on it commonly sur- back down again in the north, near Taipei, but it’s vived intact and are still habitable, but buildings still rising up in the central reach of the island and right on the rupture did not fare so well. One is just getting started in the south—the rate of up- estimate is that 35 to 50 percent of the building lift in some places is nearly a centimeter per year. damage was due to ground deformation under

10 ENGINEERING & SCIENCE NO. 4  foot. The fault came up so fast in one building be giving us topography between 60 degrees north that, on the hanging-wall block (the rising side of and 60 degrees south at about 30-meter postings. the fault), the first story simply collapsed, but it Now, if the National Mapping Agency agrees to was left intact on the foot-wall block (the lower release to us civilians the 30-meter data, we’re side). The entire scarp formed in a matter of about going to have a fantastic time mapping many of two seconds. One guy on the first floor of a Earth’s geologic hazards. It will be much easier to building woke up, opened his door, and looked map faults before they break, and then use the maps right into his neighbor’s second-floor window, for land-use planning. I got the shaded-relief map which used to be across the street. from one of my Taiwanese colleagues just before I This earthquake was very important because it went to Taiwan half a year after the earthquake, showed that the engineering problem has been and I thought I’d practice with it, to see how basically solved. The Taiwanese know how to useful the high-resolution SRTM data might be. build buildings that don’t collapse, even during I mapped the fault just from this map, before I some of the heaviest shaking imaginable. Well- ever looked at the map of the ruptures that were engineered buildings didn’t fail at all. Some im- produced after the earthquake. I mapped about properly built structures collapsed, and many 80 percent of it correctly within the 40-meter people died, but the point is that we know basi- resolution of the map, including many of the little cally how to build buildings to survive earthquake secondary faults. shaking. During the Chi-Chi earthquake, as it was But the improper use of land is a problem we named, the Chelungpu fault broke from stem to have barely begun to tackle—land use with stern. It was a beautifully behaved thrust-faulting respect to faulting, with respect to hazards from event. It may well be a good analogy to what will floods, landslides, tsunamis, slumping into the happen when our own Sierra Madre fault ruptures ocean, and so on. Let’s go back to the beautiful and the San Gabriel Mountains fling themselves waterfall on the first page and the bridge across out toward the foothill communities, from San the river. When it failed, people wondered why. Fernando, through Pasadena, to Upland. Paleo- Well, it’s really quite simple: the bridge was built seismic data collected to date suggest that our across the fault. The waterfall is the fault scarp, fault also slips in large events, about five meters and when one side moved up and over the other each, along its entire 60-kilometer length. We side, the bridge fell off its abutments. A large don’t yet know exactly how often this fault breaks, dam upstream was also partially destroyed by the but we’re working on that. But its recurrence Rupture of the Chelungpu fault rupture. The main portion of the dam rose interval is probably measured in thousands of years fault produced the 1999 nine meters over the left abutment. How fortu- rather than hundreds. nate that the dam didn’t fail catastrophically and The mountains of Taiwan’s Central Range, earthquake; the fault is kill thousands of people downstream. which rise steeply east of the Chelungpu rupture, marked by the red line in So, if you lived in central Taiwan and were are very rugged and extremely steep. They have the shaded-relief map building a bridge or a dam, would you prefer been rising for about the same period of time as above. hiring a geological consulting company to inves- the San Gabriel Mountains—three or four million tigate possible problems? Or would you favor years. But they have risen many times faster— Some of the damage from saving the few hundred thousand dollars in con- several millimeters per year. During the 1999 the fault rupture can be sulting fees, and gambling that nothing would earthquake, massive landslides rolled down into seen below: part of what happen? To me the answer is obvious, but, of many of the mountain valleys and caused spectacu- course, as a geologist I’m biased toward the longer lar damage. Of the people caught up in this was formerly a flat field view. myriad of seismically induced landslides, very few rose up four meters (left); At left above is a shaded-relief map made from survived (in all, more than 2,000 people died in buildings tilted, and topographic mapping at 40-meter postings. It’s this earthquake). In the case of the biggest slide, bridges were ripped apart. similar to data that will be produced from the some survivors rode the slide about a kilometer Shuttle Radar Topography Mission (SRTM), which and a half down. The Chi-Chi earthquake shows More than 2,000 died in NASA flew early in 2000. This latest mission will us, once again, that seismically induced landslides the quake.

Photo provided by Jack, Yung-Wan Lien, Flying Tiger Photographic, Inc.

 ENGINEERING & SCIENCE NO. 4 11 reason, but nonetheless, Istanbul’s population is growing by something like 100,000 a year, as people come in from the countryside in search of a better life. Many of them live in hastily erected buildings at the city’s edges. The North Anatolian fault is similar to the San Andreas only in that it moves horizontally; otherwise it’s quite different, particularly in its segmentation. Four major segments broke in 1999. We don’t have such segments on the San Andreas. The San Andreas is very smooth, which may well be why it produces earthquake ruptures several hundred kilometers long and magnitudes in the upper 7s. Turkey gets mostly 7.5s and less, because the segmentation seems to stop or at least impede the rupture from growing longer than a hundred kilometers or so. In the 1999 earth- quake, four segments with a combined length of about 110 kilometers broke in a single magnitude 7.4 event. Tens of thousands died. Right after the The collision of the Arabian can be a very significant hazard. From the August earthquake, my Turkish colleague, Aykut experience in Taiwan, I would caution against Barka, noted that a short neighboring segment, plate (above) with the dense development within our own precipitous just to the east, near Düzce, was the only remain- Eurasian plate is squeezing San Gabriels and encourage our policy makers to ing segment between the new break and eastern Turkey westward, creating seek geological advice before issuing permits. Turkey that had not yet ruptured. He warned that the 2,000-kilometer-long The Taiwanese fault moved primarily vertically. it might well be the next to go. Sure enough, less Now, let’s go to Turkey and look at a fault that than three months later it too broke, with about North Anatolian fault, moved horizontally, more like our San Andreas four meters of slip, causing a magnitude 7.1 numerous segments of fault. The North Anatolian fault is the indirect earthquake. which have broken in the result of the ongoing collision of the Arabian plate Although the fault zone isn’t paper-thin, it’s past 60 years. The last and the Eurasian plate, which is causing Saudi segment, closest to Arabia and the Persian Istanbul, has yet to break. Gulf to slide under The four segments that Iran. This contraction ruptured on August 17, of the Earth’s surface is squeezing Turkey west- 1999 are shown in green at ward toward Greece right. The red section near and Libya. The fault, Düzce ruptured just three which runs nearly 2,000 kilometers from months later, on the Kurdish part of November 12. Turkey past Istanbul, is the northern margin of this extruding block. Many sections of the North Anatolian fault have broken in the past 60 years. An extraordinary westward progression of earthquakes fairly narrow—only a few meters wide. Every- in 1939, 1942, 1943, 1944, 1957, and 1967 thing right on top of the fault was completely pointed right toward the place where the earth- destroyed, but you didn’t have to be right on the quake happened on August 17, 1999. Several fault to sustain great damage. A resort hotel built published papers had made this long-term on loose, saturated sediments on the shore of Lake forecast. The only sections left to break in this Sapanca is now swamped. The swimming pool remarkable sequence are those that constitute the tips into the lake, and the beautiful little cabanas 300-kilometer portions closest to Istanbul. Unfor- are much wetter than they were ever intended to tunately, this seismic gap lies predominantly be. The hotel itself looks in relatively good shape, under water, so direct access for geological inves- but if you should step up to the bar, you’ll find tigation is impossible. People in Istanbul are that the bar rail is a foot under water. The quite concerned about this forecast, with good sediments are so young and unconsolidated that

12 ENGINEERING & SCIENCE NO. 4  The resort hotel on Lake Sapanca (above left) was not heavily damaged, but when earthquake shaking caused the shallow sedi- ments on which it was built to compact, the land the ground shaking caused them to compact; and landslides, fault rupture, slumping, submergence slumped and the lake when the sediments compacted, the sea rolled in of young sediments. But of course, there are 100 meters or so. The ancient record tells us that others—tsunamis, and even asteroid impacts, waters rose up over the young sediments in such settings are notoriously for example. We can’t really do much about the swimming pool (and into good slumpers. So you’re well advised not to latter, which are fortunately exceedingly rare, the first floor, as well). The build in such places. A parkway or a golf course but we can do something about the others. photo at far right, above, would be just fine there, but it’s best not to build What can we do about the hazard of fault a major metropolitan region right up to the rupture? When I visited Taiwan, six months after of Gölcük shows that it can shoreline. It’s just asking for trouble. the earthquake, I found that in many places the be a costly mistake to In the town of Gölcük, close to the epicenter, houses that had been torn in half along the fault build so close to a shore- the old waterfront is now 150 to 200 meters out had been hauled away, and that new structures in the water. People who lived near the waterfront were being built in the same place. These new line. Many apartment suddenly found themselves sleeping beneath 40 structures will probably be just fine, at least for buildings suddenly slumped meters of water; many hundreds of people and 100, 200, maybe even 300 years. But when the into the sea atop a small many millions of dollars were submerged beneath next earthquake rupture occurs, property will once landslide. The new the waves. again be destroyed, and people may well perish. When the World Bank, along with several other This construction is happening before the coastline in the photo is agencies, made a loan of about $1.7 billion to the government has enacted regulations to guide re- several buildings farther Turkish government for rebuilding, their first building. Remember the photo of the bridge at inland than before the requirement was that a mitigation study be done the beginning of this article? The photo below earthquake. first. They said, “Before you do anything to shows what’s left of it; and it shows an excavator rebuild anything anywhere, do the hazard study digging a deep pit where one of the supporting first to tell you where to put your new buildings and where not to put them.” It’s a step in the right di- Another mistake can be rection, but it remains seen here (right): the to be seen whether or same bridge seen on page not Turkey will use the loan effectively. 9, a few months later being What can we do rebuilt in exactly the same better, or at least dif- place—right on the fault. ferently, in the third millennium to reduce our exposure to these natural hazards? I haven’t given you the entire spectrum of geological disasters, but I’ve given you a taste of a few of them:

 ENGINEERING & SCIENCE NO. 4 13 pillars used to be. According to Clarence Allen, professor of geology and geophysics, emeritus, who was there less than a month after I took the San Bernardino Valley photo, this excavation for the new support pillar had been dug through the fault plane, from the College (right) sits astride hanging wall into the foot-wall block, and the the San Jacinto fault, which fault separating young gravels underneath from runs directly through the the bedrock above had been beautifully exposed. administration building If the bridge lasts more than a few hundred years, it will be there when the fault breaks again, and (the vertical blue rectangle somebody is going to have to spend millions of to the left), the library dollars to rebuild it once more. What is the (next to it in blue and rationale, I wonder, for not rerouting the road now and crossing the fault at a point where it can be gold), the campus center done with a less expensive roadbed rather than (the purple square below with a bridge? it), and the life sciences Let’s turn now to Southern California, where we building (also blue). Other have our own share of earthquake hazards. We can take a holier-than-thou attitude and claim that we buildings, such as the do things right here, but that’s not as true as we’d Greek theater and the like to believe. Nevertheless, let’s take a look at 10,000 years or longer. We located where there auditorium (top, green) are one example of a long-range vision of hazard was deformation going on—tilting and anticlines not right on the fault but mitigation. A few years ago we did a little study and so on. In one of the trenches, we found a fold of the San Jacinto fault, a major fault that runs over a blind thrust, about three meters high. The could be damaged by through Colton and San Bernardino. In fact, it Hollywood fault is doing the same thing near the folding. A schematic runs right through San Bernardino Valley College. Capitol Records building, but with a scarp about drawing of the fault zone I live up in Lake Arrowhead, and a lot of my kids’ 15 meters (50 feet) high. And if you could look friends go to Valley College after high school. It deep under the Library Tower building in down- is shown below: buildings has a beautiful auditorium, one of the nicest town Los Angeles, you’d see the blind-thrust fault on the fold will experience Spanish-colonial-style buildings in Southern there. At the surface it shows up as a big escarp- vertical motion in an California. They built the campus there to avoid ment. At Valley College one little blind-thrust earthquake, while strike- the hazard of flooding, because the location is up fault is about five meters down, deforming older on a little ridge—the fault zone, it turns out. So 10,000-year-old sediments and not deforming the slip motion will affect they were smart about flood hazard, but not about very youngest ones, which are mainly fill contain- those along the fault. earthquake hazard—out of the frying pan and into ing concrete blocks and bricks. When we locate a the fire. fault like this, we can actually see how it’s deform- Back in 1935, after ing the ground. And we can determine what sort the Long Beach earth- of deformation a building will experience, whether quake, Valley College it’s vertical or strike-slip. hired John Buwalda, We can make a map showing precisely, within the first geologist here a foot, where the fault lies. In the main zone, we at Caltech, to come out would expect to have anywhere from a meter to and see if they had any five meters or so of horizontal motion when the problems with regard San Jacinto fault breaks again. We gave Valley to earthquakes. And College a 100-year earthquake scenario and a 400- he said, “Oh, my gosh; year scenario, and told them they had to worry you’ve got a big fault about folding as well as faulting. When the fault going through the breaks, the administration building is going to get campus.” In fact, he ripped in half, as well as the library built in the recommended a early ’70s; the life sciences building is also in thousand-foot-wide zone of no building, which trouble. The campus center, built in 1970, is basically took in almost the entire campus. They right squarely on the fault; the auditorium I like ignored his advice, even though they paid for his so much is off the fault but on the fold. report. A few years ago, the trustees called me This went to the state architect, who spent two (Buwalda’s been gone a long time) and asked what or three years figuring out what the college should they should do. They wanted to have a long-term do in terms of retrofits, building removal and master plan, a 30-, 40-, 50-year master plan for demolition, new building construction, and so on. development. And they knew they had a fault Then Congressman Jerry Lewis (R-San Bernar- problem. dino) managed to get $34 million for the commu- We went out and dug a series of trenches during nity college district, which is going to rebuild. their winter break. We actually pinpointed where The college is going to remove all the buildings the fault traces are, where they’ve been moving for from the fault zone and put replacements else-

14 ENGINEERING & SCIENCE NO. 4  ings near Gölcük, very near the fault and just on the coast. The company is putting half a billion dollars into the construction of this plant, and the earthquake happened about $50 million into it. Fortunately, only a small piece of the fault zone hit the buildings directly, but unfortunately, there was also a lot of warping; the buildings hadn’t been set back far enough from the fault. The body shop, the building closest to the fault, provided us with some information about what occurred during the earthquake. The building’s pillars, spaced about 10 to 15 meters apart, had been surveyed before the earthquake, and knowing their elevations to the nearest millimeter allowed us to reconstruct the folding. (When we first saw what had happened A new layout for San where. And the long axis of the buildings will be to the pillars during the earthquake, we thought oriented parallel to the fault, so that, just in case Bernardino Valley College it was just fantastic, which caused our clients to the geologists didn’t find everything, the build- look at us a bit funny, wondering just what sort of moves all buildings off the ings will present less of a cross section to be hit by consultants they had hired.) Along the fault fault zone and orients the fault. The college is doing a really responsible plane, the vertical dislocation was 1.5 to 2.4 them parallel to the long thing. For those of us who deal with these trage- meters. And where you had the highest vertical dies time and time again, it’s gratifying to see slip, you also had the highest amount of subsid- axis of the fault. someone caring about their children and great- ence. The pits down in the floor of this building grandchildren. were actually under water after the earthquake. In Gölcük, Turkey, a San Bernardino Valley College is a model for Having put $50 million into this already, these partially built automobile how the whole world might behave in the third guys came to us and asked if they should just millennium. What about Taiwan? With the in- abandon the site. The government had given factory sustained the formation we already have, it is eminently possi- them the land; the geologist they had talked to greatest damage—more ble to make a very detailed map of the active had said there were no problems, but one more than two meters of faulting and folding in Taiwan. Once that’s done, earthquake like this and they’re under water. I hope in the next couple of years, someone vertical motion—in its They asked us when this is likely to happen again. wanting to locate, say, a new chip-manufacturing So, should we tell these guys to move or should body shop (right, below), plant could look at what the seismic or other geo- we tell them to stay? Let’s look at the history. In which was visibly evident technical hazards are. If there are problems, they 1509, there was a big earthquake, and we think in the body shop’s sunken still have some choices at that early stage. Prob- that it was produced by the rupture of most of the ably they would choose to put the plant a long segments of the North Anatolian fault near the pillars (right, above). way away from an active fault. Or, if they don’t When rebuilding after the have that choice, a seismologist can calculate quake, the owners, on the “synthetic” seismograms for the potential earth- advice of geologists, moved quake on that fault to estimate likely ground motions, which can be taken into account in the the body shop to higher design. I chose this as an example, because chip ground, farther from prices increased twofold after this last earthquake, the fault. due, I’m told, not to damage to the buildings, but to damage to the actual manufacturing equipment inside the plants. One smart thing to put along a fault is a park, and the Taiwanese are preserving parkland along this fault as a monument. If this were Japan, of course, they would preserve 50 kilometers’ worth of park and create a $50 million museum—this is what they’ve done along the 1995 fault rupture near Kobe! What needs to be done about the slumping and subsidence hazard? Going back to Turkey, an American-Turkish automotive company was building an assembly complex of four large build-

 ENGINEERING & SCIENCE NO. 4 15 The image at left was created by draping Landsat data over a Shuttle Radar Topography Mission map of Pasadena and the San Gabriel Mountains. The Sierra Madre fault, very similar to the Chelungpu fault, runs along their base, right through JPL, at center left. The Rose Bowl in the Arroyo Seco can be seen at lower left.

secondary faulting so that when submergence occurs, they’ll at least have the body shop. They’re doing the right thing for a 100-year vision, but not for the 250-year vision that we had encour- aged. They took a middle road toward mitigation, because they wanted to be making cars for the local market within a year of the earthquake. But they are consciously leaving much of the problem A cross section through the to a future generation. Sierra Madre fault (from What about landslides and liquefaction? Geol- ogists can tell you where these will occur. We can the north end of Lincoln see it in the prehistoric record; we can see it in the Boulevard) shows two geotechnical details of the soil; and we can see it sedimentary deposits (blue close to home. In the SRTM image at left (with and pink) that are Landsat imagery draped over it) of Pasadena and the San Gabriel Mountains, you can see the Arroyo evidence for large fault Seco at lower left. You can see the Jet Propulsion ruptures in the last Laboratory up at the top of the arroyo (it has a big 15,000 years. fault running right through the administration site. It was the most destructive earthquake in building). At left is a trench we dug at Alta Istanbul in the last thousand years. In 1719 Loma Park at the north end of Lincoln Boulevard, another earthquake occurred with, we think, exposing the fault. It shows that there have been almost the same rupture pattern as in 1999. In two five-meter displacements in the last 15,000 1766, there was an earthquake in May that years, where the mountains have shoved up over damaged Istanbul, and later that year another one Altadena. They’re very rare events, only about damaged the Gallipoli Peninsula. So there was every 7,000 or 8,000 years. But the last one was this cluster in the 1700s —bang, bang, bang, all about 8,000 years ago, so the next one could well in a matter of about 50 years. Could this happen happen within the next few centuries. again and mess up this plant if it’s sitting right The California Division of Mines and Geology here? The answer is almost certainly not, because has provided hazard maps for much of the urban it looks as if these earthquake clusters have about part of the state. The maps were mandated after 230-year intervals—almost like clockwork. In our the 1989 Loma Prieta earthquake in Northern report, we said to the automotive company: “Don’t California. These maps show liquefaction to be a worry about the main fault; worry about adjacent potential hazard in the Arroyo Seco, but nowhere earthquakes shaking your facility and about minor, else in Pasadena. Landsliding and rock falls are secondary faulting on your site. If you really can shown to be a problem in parts of the Verdugo afford a longer vision, don’t build here at all, Hills and large parts of the San Gabriels. City because it’s going to submerge in two or three planners are now wondering what to do with this centuries, and somebody will have to deal with the information. Should we be worrying about these problem then.” They decided to deconstruct the things? We have to worry about them now, entire 200-meter by 100-meter body shop and because if something happens and we already had build it higher on the site and away from the the maps from the state and we did nothing, we’re

16 ENGINEERING & SCIENCE NO. 4  We have a clear choice: we can live with our beautiful, dangerous Earth as we

have in past millenia, or we can learn where to put our bridges, campuses,

houses, and factories to minimize the destruction. But what we can plan for, we should. The little Chinese village below is a city that is waiting to die. This city has everything going against it. It sits on an active alluvial fan that could well bury it in a flood. It has an active mountain front; there’s a beautiful normal fault right at the base of the going to get our socks sued off. So we want to act mountain. Talus from seismic shaking could bury responsibly. Municipalities in California are now the city in rocks. There are clear signs of young dealing with this problem. Turkish and Taiwanese landslides. So there’s flooding, faulting, rock falls, municipalities need first to develop the infrastruc- and landslides. This is a place you don’t want to ture just to begin to deal with these sorts of spend a lot of time! As world population grows problems. from 6 billion to 12 billion or whatever it’s going Geologists can map volcanic, flood, and tsunami to be in the next 50 years, we have an opportunity hazards very effectively, but there are some hazards now that will not come around again—an oppor- we probably have to ignore—immense, cata- tunity to choose the places that we’re going to PICTURE CREDITS: strophic volcanic eruptions, for example, hundreds expand into. When the Taiwanese and Turks 10 — Doug Cummings; 10, 11 — C. T. Lee; 11, of times more voluminous and extensive than the expand into the mountains, I hope they avoid 13, 15, 17 — Kerry Sieh; eruption of Mount St. Helens. The valley down- places where landslides can happen. When they 12-13 – Aykut Barka; slope from Mammoth Mountain blew up 700,000 expand further onto the coastal plains, I hope they 13 — Timothy Dawson; 14-15 — Earth Consul- years ago and covered most of the western United are aware of the potential for slumping and tants International; 16 — States with ash. Yellowstone has gone up twice in liquefaction. And when they build near their NASA, Charles Rubin the last million years. But these are very infre- faults, I hope they choose to mitigate against quent events; we’d probably better just duck and ground ruptures. They need to talk to their local cover. Pieces of Hawaii go sliding off into the geologists, who can advise them on how to avoid ocean every once in a while, producing tsunamis all these natural hazards. that are on the order of 350 meters high. I don’t We have a clear choice: we can live with our think we can design around those or plan for beautiful, dangerous Earth as we have in past them. There was a large runout landslide in the millennia, or we can learn where to put our Mojave 17,000 years ago, off the northern flank of bridges, campuses, houses, and factories to mini- the San Bernardino Mountains; this could happen mize the destruction. On my best days, I’m opti- to San Bernardino on the urban side of the moun- mistic that we will choose a new vision for the tains, but I guess we have to cross our fingers and future rather than acquiesce to enduring the hope that doesn’t happen. I don’t see any way to damage and death brought by natural disasters rationally plan for such an immense catastrophe. as we have in the past. ■

Professor of Geology Kerry Sieh has been a member of the Caltech faculty since 1977, the same year he earned his PhD from Stanford; his AB (1972) is from UC Earthquakes, floods, and Riverside. Sieh is a paleoseismologist, that is, he studies landslides: would you want the patterns of earthquakes from the perspective of a to live in this geologist, over hundreds to thousands of years. His Chinese village? initial interest in the San Andreas fault has expanded to include faults all over the world, motivated by a concern for human welfare. This article is adapted from the Watson Lecture he gave in the spring of 2000. Since then, devastating earthquakes in El Salvador and India have provided additional examples of poor land-use planning.

 ENGINEERING & SCIENCE NO. 4 17 Kepler observed in 1619 that a comet’s tail faces away from the sun, and

concluded that the cause was outward pressure due to sunlight—a force that

might be harnessed with appropriately designed sails.

18 ENGINEERING & SCIENCE NO. 4  Solar Sailing: The Next Space Craze? by Joel Grossman

With a name like solar sailing, the technology launch in the period 2010 to 2015,” said Sarah sounds like it could be Southern California’s next Gavit, associate manager for JPL’s Interstellar beach-sport craze. But the Jet Propulsion Labora- Program and preproject manager of the Interstel- tory (JPL), which is managed for NASA by lar Probe Mission until January 2001. “There are Caltech, is planning to leave the Pacific Ocean many other technologies that need development far behind. Plans on the drawing board run the for interstellar or even near-interstellar travel. For gamut from communications satellites hovering example, communications, autonomy, etc. Sails over Earth’s poles, held in position by solar sails, are getting ready for flight demos, while some of to spacecraft hoisting giant, ultrathin sails for these other technologies are still a ways off.” Adds journeys exploring interstellar space. Chuck Garner, senior engineer in JPL’s gossamer Perhaps the grandest mission of all will be an systems group, propulsion and thermal engineer- A proposed solar-sail interstellar probe. Its destination: Alpha Centauri, ing section, “The sailcraft will be at very great the sun’s nearest neighboring star, approximately distances from Earth, and therefore must operate mission to fly alongside four and a half light-years away. Lasers as power- and navigate itself without the aid of ground con- Halley’s comet in 1986 ful as 10,000 suns, focused on the craft from trollers. And communications hardware must be may have been ahead of its Earth-orbiting satellites, could one day accelerate developed that can transmit data over enormous such a probe to one-tenth the speed of light. At distances, utilizing very little power and requiring time. It was killed by that clip, it would reach Alpha Centauri within very little mass.” To deal with these parallel tech- NASA in 1977, partly on the professional lifetime of a scientist, arriving nology needs, JPL created a Solar Sail Program to the grounds that the there in 40 to 50 years. develop the solar sails, and a separate Interstellar technology was unproven. But that’s going to be a while. JPL scientists Program is developing the other technologies. estimate that a more near-term precursor space- Meanwhile, the National Oceanic and Atmo- But the idea keeps coming craft powered only by sunlight could cover the spheric Administration (NOAA), the National back—sails don’t need to 25 trillion miles, or 273,000 astronomical units Weather Service, the Department of Energy, the carry their own fuel, (AUs; an AU is 93 million miles, or 150 million Air Force, the Department of Defense (DoD), and making them very kilometers, the average distance between Earth the Space Environment Center in Boulder, Colora- and the sun), at a speed of 15 AUs per year. This do, are eyeing solar-sail technology for more attractive to mission is equivalent to flying from Los Angeles to New immediate practical applications. These include planners. And now the York City in 63 seconds, about nine times faster monitoring the sun for magnetic storms that can hardware is starting to than the orbital speed of the Space Shuttle. At knock out communication and Global Positioning 160,000 miles per hour, the solar-sail probe would System (GPS) satellites, and play havoc with such catch up with the hype. be speeding through space five times faster than earthbound things as electrical power grids and Rendering by Ken Hodges. the 3-AU-per-year speed of Voyager 1, a conven- cell phones. This would require an orbit around tionally propelled spacecraft launched in 1977 the sun on a path that keeps the spacecraft directly that is currently our most distant space probe. between the sun and Earth at all times. However, An interstellar precursor mission launched in this disobeys Kepler’s laws of orbital mechanics, 2010, the earliest projected date, would overtake and such a satellite would need to carry a massive the then-41-year-old Voyager 1 in 2018. It would amount of propellant to fight the tendency to take 100 millennia for Voyager 1 to reach Alpha move out of the desired orbit. But propellantless Centauri, but only 20 millennia for the sailcraft. propulsion via solar sails could keep satellites in “We are hoping for a demonstration mission by such orbits—or in stationary orbits over Earth’s 2005, with an interstellar precursor mission poles, which are similarly non-Keplerian. Earlier

 ENGINEERING & SCIENCE NO. 4 19 Right: Kepler’s law says that objects in smaller orbits travel faster, so within, say, a month, a sun- orbiting early-warning spacecraft would be out of line with Earth. Below: A large coronal mass ejection, or solar storm, on April 20, 1998 as seen by the Large Angle and Spectrometric Corona- graph aboard SOHO. The central disk covers the sun, but the white circle in the disk’s middle shows the warnings of inclement space weather would allow Geostorm, which will monitor space weather and sun’s size (about 1,400,000 utilities more time to boost power reserves, pro- the solar wind—a steady sun-produced stream of vide wireless users time to prepare for transmission ions, electrons, and neutral atoms that is perhaps kilometers in diameter) failures, and in extreme cases allow satellites time best known for interacting with the ionosphere to and position. to go into “sleep” mode until the storm passed. create the northern lights. The best space-storm Bottom: Geostorm on And satellites permanently hovering over the warnings currently come from NASA’s Advanced station. North Pole would for the first time allow continu- Composition Explorer (ACE), which along with ous monitoring of areas now only intermittently the joint NASA/European Space Agency (ESA) sampled by orbiting satellites. This round-the- SOlar and Heliospheric Observatory (SOHO) clock surveillance of, for example, polar ice and is “parked” about 0.99 AU from the sun, at the the aurora borealis, or northern lights, could spur point where the gravitational forces of the sun and advances in meteorology, climatology, and ocean- Earth balance each other. (This is called the first ography. Other applications include search and sun-Earth Lagrangian point, or L1.) However, rescue support, communications, and data relay Geostorm will be maintained in a similar position over the poles. 0.98 AU from the sun, doubling the warning time After early input from JPL, both NOAA and for potentially disruptive storms from one to two the DoD are planning a joint mission known as hours. The photon pressure on a solar sail 67 meters on a side will counteract the sun’s “extra” gravity to keep Geostorm sunward of L1. Another mission known as the Solar Polar Imager would further bolster scientific knowledge of the solar wind and geomagnetic storms via an orbit that passes over the sun’s still largely unexplored north and south poles—an orbit that’s hard to achieve conventionally because of the enormous amount of propellant required to leave the equatorial plane in which the planets move. And solar sails could be a boon to planetary ex- ploration. Mercury could be better studied from a sun-synchronous orbit about the planet, for example, which a solar sail would make feasible. Solar sails could also be useful for asteroid rendez- vous, and for reducing the travel time needed to explore the outer planets and their many moons— a particular benefit in the case of sample-return missions. And the cost of transporting large amounts of cargo and equipment between Earth and, for instance, Mars to establish a permanent human presence could be so dramatically reduced that it would become practical financially. Some attribute the idea of hoisting sails in outer space to the aforementioned Johannes Kepler, who

20 ENGINEERING & SCIENCE NO. 4  Estimated flight times to the outer solar system and beyond for a 500-kilogram solar-sail payload launched from a Delta II rocket. (For comparison, the 815-kilogram Voyager 2 took 12 years to reach Neptune after launch on a much larger Titan Centaur. Without the gravity assists from Jupiter, Saturn, and Uranus, it would have taken 30 years.) The sail’s density per unit area is denoted by “g,” and the green bar represents the range traversed by Pluto’s wandering orbit. The Kuiper belt is a zone of icy material that managed to elude the outer planets during their formation.

four centuries ago wrote fellow astronomer Galileo to a sufficient height for the thermal-molecular a letter mentioning space travel using a “ship or effect to work. Then, when the air peters out, the sails adapted to the heavenly breezes.” Kepler sail continues to rise as the photon beam heats its observed in 1619 that a comet’s tail faces away underside enough that atoms in a special coating from the sun, and concluded that the cause was begin to evaporate off it, creating thrust. And outward pressure due to sunlight—a force that upon reaching Earth orbit, it kicks over to purely might be harnessed with appropriately designed photonic propulsion. Harris’s team has demon- sails. At the time, the corpuscular theory of light strated each of these four technologies individu- was in vogue, which turns out to be qualitatively ally, but putting them all together will take an consistent with today’s quantum-mechanical view enormous amount of additional work. that light, in some cases, acts as a particle instead What is now accepted as the first true experi- of a wave. mental verification of the existence of radiation Scientific understanding moved closer to the pressure came from Russian physicist Pyotr mathematical basis for designing a working solar Lebedev’s elegant torsion balance experiments at sail in 1873, when physicist James Clerk Maxwell, the University of Moscow in 1900. Independent who is associated with the wave theory of light, verification came from Ernest Nichols and G. F. predicted the existence of radiation pressure as a Hull at Dartmouth College in 1901. But the consequence of his unified theory of electromag- modern concept of light as photons—massless netic fields. Radiation pressure was independently packets of energy capable of producing momen- shown to exist in 1876 by Adolfo Bartoli, who tum transfer, or radiation pressure—did not evolve derived it mathematically from the second law of until the early 20th century as Max Planck, Albert thermodynamics. Radiometers, invented by Sir Einstein, and others grappled with thermal radia- William Crookes, were initially, albeit mistakenly, tion and the then-puzzling nature of the photo- used to demonstrate the existence of radiation electric effect. pressure in 1873—a classroom demonstration Indeed, as Colin McInnes notes in his book, A carbon disk made by that is still done today. Solar Sailing: Technology, Dynamics and Mission ESLI levitates in the light In fact, the force driving a radiometer is due to Applications, the term “photon” was not even from a 100-watt bulb. thermal-molecular forces, and is several orders of coined until 1925. By this time, Einstein had magnitude greater than radiation pressure. This won the Nobel Prize for explaining the photoelec- type of levitation has recently been demonstrated tric effect, and the equation E = mc2 had become by Energy Science Laboratories, Inc. (ESLI) of San part of the modern lexicon (and Time magazine Diego, under contract to JPL. “These experiments cover material!). The amount of momentum are part of a project to produce a spacecraft that transferred from a photon to a solar sail is derived can be launched from your backyard to the stars, by combining quantum mechanics and Einstein’s using an assortment of different kinds of physics theory of special relativity. Both McInnes and derived from photon interactions,” says Henry former JPL employee Jerome Wright in his earlier Harris, chief scientist for space solar power at JPL book, Space Sailing, detail the mathematics needed and task manager of the project. This type of to calculate radiation pressure, plot trajectories levitation only works where there’s an atmo- and orbits, and begin designing a solar sail, start- sphere—but not too much of one!—so the scheme ing with Planck’s law and Einstein’s mass-energy actually has four components. The fragile sail equivalence. would be gently lofted on a carbon-fiber balloon Perhaps not surprisingly, science-fiction writers

 ENGINEERING & SCIENCE NO. 4 21 velocities attainable via chemical or ion propulsion Earth s Orbit Earth s Orbit for a number of missions. And, for fixed sail Perihelions Perihelions angles relative to the sun, solar-sail orbits were 0.30 AU 0.30 AU 0.47 AU 0.47 AU calculated to be logarithmic spirals. (If the space- 0.61 AU craft is inbound toward the sun, this is the exact Launch equivalent of a moth’s death spiral around a porch Launch light, and for the same reason—the light is kept at a constant angle to the flight path.) But it was famed science-fiction writer Arthur C. Clarke’s melodramatic 1963 short story, “The Wind from 5 AU 5 AU the Sun,” that gave the idea its biggest push, capturing the public imagination and spreading 4.5 Years 2.4 Years the concept among sci-fi–reading engineers—who quickly figured out that the solar wind lacked the propulsive force of the sun’s photon stream. By the 1970s NASA was funding solar-sailing To Pluto in 10.8 years To Pluto in 6.6 years studies, and JPL’s Wright came up with a trajec- Terminal velocity = 19.7 km/sec Terminal velocity = 30 km/sec tory for a solar-sail spacecraft to rendezvous with Halley’s comet and watch it evolve as it neared the sun. Bruce Murray, who was then JPL’s director of the late 19th and early 20th centuries were also and a Caltech professor of planetary science, ele- transfixed by light, and were penning tales of vated the idea to most-favored “purple pigeon” spacecraft propelled by mirrors. That is the prin- status. Two competing designs were produced: A comparison of two sun- ciple of the solar sail: a photon reflecting off a a three-axis-stabilized, square-shaped sail 800 diver trajectories to Pluto. shiny surface gives that surface a push equal to meters on a side; and a “heliogyro” some 15 kilo- “Perihelion” refers to a twice the photon’s momentum. (The surface gets meters in diameter that sported 12 free-spinning, point of closest approach one dose of momentum in slowing the photon helicopter-like blades, designed with the help of down and stopping it, and another one in acceler- helicopter engineers John Hedgepeth and Richard to the sun. The three-loop ating it back in the opposite direction as it is MacNeal. However, NASA dropped the project trajectory at left is for a reflected.) The catch is that because photons and its large payload in 1977, leaving Halley to 270-square-meter sail with have no mass, and thus very little momentum, the conventionally propelled flybys (a much easier mirror has to be featherlight in order to be moved a density per unit area of mission!) by spacecraft from the Soviet Union, significantly. The first really practical writings on Japan, and Europe. One reason the mission was 10 grams per square solar sails are attributed to Konstantin Tsiolkov- shelved was that the deployment of a solar sail meter, and provides an sky, the self-taught father of astronautics in the in outer space was considered a high risk. Indeed, initial acceleration of 0.5 Soviet Union, and his Latvian colleague, Fridrickh a free-flying solar sail has not been deployed in Tsander, a liquid-propulsion rocket pioneer. In the space to this day. millimeters per second per early 1920s, the pair (their work appears to have Nevertheless, solar sailing had by then attracted second. The faster flight at been independent) wrote up their notions of very an almost cultlike following, and several newly right assumes a 380- large, ultrathin mirrors propelled to “cosmic formed international organizations began promot- square-meter sail with half velocities” by the pressure of sunlight, a design ing races to the moon and Mars, eerily reminiscent concept still current. of Clarke’s short story. JPL engineer Robert the areal density and twice After Tsiolkovsky and Tsander, solar sailing Staehle, who is now deputy manager of the Europa the initial acceleration. faded into oblivion for almost three decades, Orbiter Project, helped form the World Space during which liquid-fueled rockets became the Foundation in 1979 and attempted to obtain rage. The first American solar-sailing proposal, private funding for demonstration flights. Euro- in the May 1951 issue of Astounding Science Fiction, peans formed the Union pour la Promotion de la appeared under the byline Russell Sanders—a Propulsion Photonique (U3P) in 1981, and along pseudonym adopted by aeronautical engineer Carl with the Solar Sail Union of Japan promoted a race Wiley to guard his professional reputation. The to the moon. Solar-sail fever was still going strong Sanders/Wiley article proposed using solar sails in 1989, when the Christopher Columbus Quin- instead of rocket propulsion for interplanetary centennial Jubilee Commission attempted to travel, and detailed what is now the usually organize a race to Mars for 1992. Although they accepted plan—to start the journey by spiraling kept alive interest in solar sailing and encouraged in close to the sun to gain maximum momentum designers to advance the art, the prize money and where the photon concentration is highest. space-sailing contests never materialized. During the next decade the matter became a The days of such promotions are over, at least serious scientific subject, and Richard Garwin of for now, though other schemes worthy of the best Columbia University coined the term “solar sail- of P. T. Barnum are taking their place. For ing” in a 1958 article in the journal Jet Propulsion. example, Encounter2001.com, whose Web site Several studies in the next few years showed that also hosts ads for burials in space, is working with solar sailing could theoretically equal or exceed the L’Garde, Inc., of Tustin, California, which special-

22 ENGINEERING & SCIENCE NO. 4  izes in inflatable space structures, on an interstel- lar spacecraft. Encounter2001 hopes for a live, members-only Webcast of its privately launched solar sail, which will double as a giant billboard and carry a payload of photos and greetings from customers who are expected to flock on line to pay by credit card as the launch date nears. For $24.95, you get a photo and a message; for another $25, you can send a “biological signa- ture”—a sample of hair-follicle DNA. Meanwhile, back at the lab, JPL’s New Millen- nium Program (NMP), which solicits proposals to demonstrate new spaceflight technologies and funds the winners, has taken up the torch. NMP’s Space Technology 7 (ST7) announcement in April 2001 is expected to include an opportunity for a full-flight solar-sail mission, says Hoppy Price, manager for solar-sail technology development in the NASA Technology Program Office at JPL. Price hopes to place a solar-sail satellite into a high Above: The World Space equatorial orbit, deploy the sail, and then just let it fly around. The goals are twofold: first, to Foundation displayed a simply prove that it can be deployed; and second, prototype solar sail on the to evaluate attitude-control issues, thrust perfor- floor of the Pasadena mance, structural dynamics, and sail lifetime, and Convention Center at the to prove out the control algorithms in the flight software. This will clear the way for “operational Planetary Society’s first missions” like DoD and NOAA’s Geostorm. Planetfest in 1981. But the title of First Solar Sail in Space may go Right: A panel of the to the nonprofit Planetary Society, whose Pasadena Planetary Society’s current headquarters is just a mile or so from the Caltech campus. As E&S was going to press, the society solar sail, which is being announced that it had contracted with Russia’s assembled near Moscow. Babakin Space Center to build a 600-square- The Mylar sail is a relative- meter, windmill-shaped sail 30 meters in diam- eter. The sail’s deployment mechanism, which ly thick five microns, about uses inflatable booms, will be tested on a subor- one-fifth the thickness of a trash bag; April’s subor- bital flight test will include two such panels. And in a peace dividend from the end of the Cold War, both the suborbital test and the completed spacecraft will be launched in converted intercontinental ballistic missiles fired from a Russian nuclear sub in the Barents Sea. Far right: The completed spacecraft will carry two cameras and several instru- ments and, although just a point of light as seen from Earth, will shine as brightly as the full moon.

 ENGINEERING & SCIENCE NO. 4 23 bital flight in April, as will a pioneering inflatable reentry shield the Russians have developed. The spacecraft itself is slated to be launched to an ini- Houfei Fang, a member of the technical staff in JPL’s tial altitude of 850 kilometers between October gossamer systems group, holds two five-meter-long, three- and December 2001 to attempt the first solar-sail inch-diameter STR booms. (One of them is rolled up on a flight. “It’s the Wright Brothers analogy,” says six-and-a-half-inch diameter mandrel, as seen in the inset.) Louis Friedman, the Planetary Society’s cofounder and executive director. “This first sail won’t fly to The boom weighs a kilogram, and can take a maximum a distant location, but we hope it will demonstrate axial force of 75 kilograms. the concept.” (Incidentally, Friedman led the Halley’s comet solar-sail project at JPL.) This commercial venture, named Cosmos 1, is being privately funded by Cosmos Studios, a new-media company started by Ann Druyan, wife of the late Carl Sagan, who cofounded the Planetary Society with Friedman and Caltech’s Bruce Murray; and Joe Firmage, an Internet entrepreneur. Deploying a solar sail in space is fraught with structural-engineering complexities. The packag- ing is particularly tricky, as the stowed sail’s exposed area may expand over a hundredfold as it unfurls. Risks inherent in unpacking solar sails in space include the sudden venting of air trapped in the folds of the sail, with enough force to tear it; tangles; rips in the fabric; electrical arcs or other discharges; and electrostatic forces—static cling, in other words—that may hold the folded sail together. “The folded-up sail is essentially a big capacitor,” says Price. “We do not know how it will behave in space.” Grounding can be viable, including squares, hexagons, and other designed-in to minimize some of the electrical- polygons, as well as disks and hoops. Carnegie discharge risk, but it is “a bit hard to analyze on Mellon University is reviving the heliogyro, albeit the ground.” Thus, actual launches are needed a more modest one with four blades 30 meters to see how it works. Wrinkle management in the long and a meter wide. JPL is starting out with packing and unfurling is vital, because wrinkles a simple design made up of four triangular panels. can cause multiple reflections and intense hot “It looks like a big kite with four booms coming spots that might damage the sail, and wrinkles out from the center,” says Price. The four booms will also decrease the sail’s reflective performance. will deploy themselves straight out from the The sail’s wrinkle potential will probably vary central hub to unfurl the sail. with the manufacturing process; wrinkle-manage- Several types of booms are being considered. ment options being explored include special sail- One option is carbon-fiber booms, developed fabrication machines that fold the sail wrinkle-free by the German Aerospace Center (DLR), that are on a big table as it is built, and methods of pulling lens-shaped in cross section. They lie flat when out the sail with enough tension to remove the rolled up on a spool, but pop open as they unroll, wrinkles. becoming quite stiff. (The prototype sail built by A number of sail shapes are likely to prove the World Space Foundation in 1981 used similar tubes made of much heavier stainless steel.) Another novel design, recently patented by JPL, uses commercial-grade stainless steel carpenters’ measuring tapes (from Sears!) as stiffeners. Called a Spring-Tape Reinforced (STR) aluminum lami- Above: The Carnegie- nate boom, it also rolls up flat, but has a circular Mellon heliogyro. cross section when deployed. Also under consider- Right: The DLR’s lens- ation are carbon or fiberglass rods in the form of a cross-braced, three-sided truss. When twisted, the shaped boom will be used truss coils up to stow compactly into a cylinder in ESA’s solar-sail pro- the size of a small trash can. gram, which is planning a Inflatable booms that blow up like long, skinny balloons have the potential to unfurl a tightly test flight in a year or so. rolled sail from its container, and, if properly stiffened—perhaps by being perfused with an epoxy that cures into a very hard plastic when

24 ENGINEERING & SCIENCE NO. 4  exposed to ultraviolet light—could then function as rigid struts to keep the sail taut. The deploy- ment of an inflatable antenna in orbit has been demonstrated. The Spartan 207/Inflatable Anten- na Experiment, which flew in May 1996 on space shuttle Endeavour, deployed a 14-meter antenna for several hours. “The demonstration was a success, and we learned a lot from it, but the process of inflation was not well controlled,” says Price, which means the technology needs more testing in space. JPL and its industrial and academic partners are currently developing inflatable structures for The overall size of the a variety of applications, including solar arrays, radar and communications antennas, telescopes, inflatable synthetic aper- and all kinds of instruments. So solar sails will ture radar array above is benefit from research directed toward projects like 1.7 by 3.7 meters, yet it the construction of lightweight, inflatable space rolls up into the two telescopes with minimal steel and glass, such as JPL’s proposed Advanced Radio Interferometry scrolls at right for stowage. between Space and Earth (ARISE) mission. The antenna, which is a ARISE is designed to look at the disks of gas flat sheet, acts like a and dust that surround black holes, zooming in traditional parabolic dish on them with a resolution 5,000 times better than that of the Hubble Space Telescope. antenna thanks to a tiny copper dogleg incorpo- rated into each reflective element that steers and Right: The 14-meter focuses the radar beam— Spartan 207/Inflatable the brainchild of John Antenna Experiment in Huang, a principal engineer flight. in JPL’s Spacecraft Telecommunications Equipment Section.

 ENGINEERING & SCIENCE NO. 4 25 The Russians have twice attempted to deploy films and adhesives, and does not degrade until large, sail-like mirrors in space. The wheel-like 670 kelvins (K)—the melting temperature of mirrors, named Znamya (“Banner”), were spun glass, and well above the 520–570 K range con- on motor-driven axles to keep their shape through sidered safe for long-term solar-sail operation. centrifugal force. A 20-meter-diameter version However, even moderate solar-sail performance was successfully tested by Vladimir Syromiatnikov requires films on the order of 2 microns (mil- and colleagues at Energia in 1993, using a Progress lionths of a meter) thick, which tear very easily resupply vehicle that had just undocked from the and are not routinely fabricated in rolls. Kapton Mir space station. However, a 25-meter version film is typically produced in rolls 7.6 microns failed in 1999, when it tangled on an antenna thick. Though small-scale etching tests have jutting out from the Progress spacecraft that was gotten the thickness down to 0.4 microns, lifting deploying it. The antenna had been used in the and handling such ultrathin material without docking maneuver, and was supposed to have been tearing—much less folding, packing, and deploy- retracted before sail deployment. A mission- ing it in space!—is going to be quite a challenge. operations software error was to blame. Suitable Mylar films are easier to come by com- Once a sail is deployed, steering it and control- mercially, and can be had as thin as 0.9 microns. ling its attitude (i.e., pitch, yaw, and roll) become Possible ways to strengthen these films include paramount concerns. Thrusters could be used for special backings, such as crisscrossed Kevlar fibers. steering, but the whole idea behind sails is to Mylar or Kapton needs to be coated with a avoid the weight of propellant and the possibility material like aluminum, which has a reflectivity of running out of it. One approach is to shift the of close to 0.9 (1.0 being perfect), in order to make spacecraft’s mass relative to the sail’s center of it an efficient photon reflector. Even so, some pressure by moving the spacecraft’s electronics photons will be absorbed and the sail will heat package around on the end of a long boom. Alter- up, especially on missions that go close to the sun. natively, the spacecraft could be steered like a Coating the sail’s back side with a substance like sailboat by moving the center of pressure relative chromium, which has an emissivity of order 0.64, to the center of mass by use of adjustable vanes or is one way to shed heat from absorbed photons and flaps on the outer corners of the sail. “We expect extend the sail’s life. (Both metals would be added to see different, competing ideas on how to control to the sail by vapor deposition under high vacu- attitude,” says Price. um, a common industrial process.) And advanced Thin plastics like Mylar and Kapton are the thermal-control technologies such as micro- major near-term candidates for solar-sail fabrica- machined, whisker-like quarter-wave radiators, tion, as they’re lightweight and are commercially which act as antennas at infrared wavelengths, available in wide rolls. Mylar can only be used for would be useful for close approaches to the sun. short-duration missions, as it is rapidly degraded A typical recent design has 0.1 microns of alumi- by ultraviolet light. Kapton is chemically inert, num vapor deposited on 2 microns of Kapton has high radiation resistance, adheres well to metal substrate and 0.0125 microns of chromium

Above: In this microphotograph of ESLI’s carbon-fiber mesh, the scale bar is 20 microns long. The fibers are seven microns in diameter, or about one-tenth that of a human hair. The bulk material comes in sheets one millimeter thick, and typically has a density of about seven grams per square meter. Right: A five-centimeter molybdenum-coated sail sample.

26 ENGINEERING & SCIENCE NO. 4  on the rear side, along with grounding straps to In December 1999, JPL funded Leik Myrabo, guard against electrical discharges between the a mechanical engineering professor at Rensselaer front and rear surfaces that could cause the sail to Polytechnic Institute in Troy, New York, to mount tear, and rip stops to limit tearing should it start. a disk-shaped sample of ESLI’s microtruss on a Since the substrate is mainly needed to allow sensitive pendulum apparatus, stick it in a high- handling, packing, and deployment of the sail, vacuum chamber, and zap it with photons from another strategy involves vaporizing the substrate a high-powered laser at the Wright-Patterson Air after deployment. This would leave a sail com- Force Base in Dayton, Ohio. The mesh was coated posed of a thin reflective metal film, with rip stops with a thin layer of molybdenum on one side to and thick strips of unvaporized substrate left in maximize its reflectance in the infrared, where the strategic places to act as reinforcement. Small- laser operates. The hope was that photonic thrust scale experiments dating back decades show that would deflect the pendulum from the vertical by it is possible to create metal films 0.05 microns a measurable amount. The thrust supplied by the thick, though at some point the film becomes too laser could then be calculated very accurately by thin and starts letting a significant amount of careful measurement of the angle of deflection. light through. One scheme to make metal films However, it was anticipated that the up-to-10- lighter without degrading their optical properties second laser blast would heat the sail enough to involves perforating the films with holes smaller cause atoms to evaporate from its surface, giving the pendulum a “kick” as they left. This kind of thrust—the third step in Henry Harris’s “backyard launch” scheme—could easily be mistaken for pho- tonic thrust. So the samples were weighed before and after each run, and any thrust due to mass loss was than the wavelength(s) of the light being used for calculated. The research, presented last July at The world’s first demon- propulsion. Technology to make these small holes NASA’s 36th annual Joint Propulsion Conference stration of photonic thrust already exists in the semiconductor industry. in Huntsville, Alabama, showed that the pendu- with a real-world laser-sail Since space is a hard vacuum, one could even lum deflected “from 2.4 to 11.4 degrees, measured material. Laser light first fabricate the sail in orbit, using a small vapor- as a function of incident laser powers from 7.9 to deposition unit that would be discarded when 13.9 kW. Laser photon thrust ranged from 3.0 to illuminates, then moves a the job was done. Direct heating would evaporate, 13.8 dynes,” according to the project report five-centimeter specimen say, powdered aluminum, and the metal would published by the American Institute of Aeronau- of ESLI’s molybdenum- condense onto a sail-shaped substrate. After tics and Astronautics. The researchers also found coated microtruss cooling, the metal film could be separated from that heating the sail to about 2600 K caused the substrate. The technology for manufacturing ablation, or mass loss, to occur, and those runs mounted on a magnetically thin metal films already exists, as do methods for produced up to 50 percent more thrust than the suspended pendulum. The controlling their thickness. theoretical maximum available from photon power 150-kilowatt continuous Two years ago there was movement toward alone. “It was amazing what those sails took,” says graphite fibers for sails, and even more recently Myrabo. “They were just incredible.” CO laser is part of the 2 toward stronger and possibly thinner single-crystal In December 2000 another round of tests was Laser-Hardened Materials carbon fibers. Timothy Knowles, president of done, in which the sail was propelled up a vertical Evaluation Laboratory II, ESLI (the company that did the thermal-molecular molybdenum wire. This is considerably more run by the Anteon demonstration mentioned earlier), invented a demanding. The pendulum tests translated into mesh of randomly oriented, crisscrossing graphite a spacecraft acceleration of 0.16 g, with one g Corporation under con- fibers called a microtruss. The material, which being the force Earth’s gravity exerts on you as tract to the Air Force at rather resembles a scrubbing pad, is ultralight, you sit in your armchair reading this. In order to the Wright-Patterson Air yet stiff and strong. When rolled up or folded levitate the sail, it needs to be hit with an upward g Force Base. into accordion pleats, it springs back into a flat force in excess of one . The same size microtruss sheet upon release, greatly simplifying deploy- samples were used again, but they were heavier ment. This carbon mesh also takes the heat much this time because an eyelet to ride the guide wire better than plastics do, which is vital for laser had to be inserted into them and glued into place. propulsion. If you want to get to Alpha Centauri So the sail really had to be zapped hard in order to within your own lifetime, you need to slam so lift it—in fact, some of the specimens wound up many photons into the sail that even a near-perfect fusing to the wire. Nevertheless, says Harris, “we reflector will start to disintegrate from the believe we have demonstrated 1-g photonic accumulated heat it can’t reradiate. acceleration at 2600 K. Motion analysis [of the

 ENGINEERING & SCIENCE NO. 4 27 videotapes and high-speed movie footage taken of the runs] is currently under way to confirm this. We have also demonstrated small-scale deployment and are beginning stability and con- trol analysis. Theoretical analysis suggests that with further development we may be able to achieve 100 g acceleration, which would enable achieving one-tenth the speed of light in approxi- mately eight hours.” (Such acceleration would require a sail with one-tenth the areal density and a laser with ten times more power—not unreason- able goals.) “We have basically demonstrated that it is do- able,” says Neville Marzwell (MS ’71, PhD ’73), technology manager for the Technology and Applications Programs Directorate at JPL, though “there are many technological gaps.” Sail fabric is The Znamya 20-meter mirror used spin to deploy itself and getting stronger, and better able to withstand high pressures and temperatures, and Marzwell expects maintain its shape. to eventually get to the point where the sail is reinforced with carbon fibers in the form of micro- polarization would clearly be superior to the heavy diamonds. Marzwell envisions laser or microwave laser gyros and thrusters used on today’s spacecraft. beams blasting diamond-studded sails to Alpha And the spin could be used to deploy the sail, or Centauri with little need for photons from the sun. to keep it rigid without the use of structural PICTURE CREDITS: However, the same sturdy sailcloth would also be stiffeners, as the Russians did with Znamya. The 20 – Doug Cummings; useful for “sun diver” missions, like those envi- microwave program is using uncoated samples of 21, 22, 24, 25, 29 — sioned by Carl Wiley in his 1951 Astounding Science ESLI’s microtruss, which absorbs about 10 percent NASA; 20 — SOHO, NOAA; 21, 26 – ESLI; Fiction article, where sails tack in close to the sun of the incoming microwave radiation. The sails 23 — Richard Dowling/ to gain maximum momentum. were levitated on a guide wire, and in April 2000 WSF, Louis Friedman, A parallel program studying microwaves instead the first flights were made. It is still unclear, Planetary Society; of lasers is going on in-house at JPL, using a high- however, how much of the thrust was photonic, 24 — Carnegie Mellon; 24 —DLR; 27 — RPI; vacuum chamber in the Advanced Propulsion and pendulum experiments are now under way 28 — Energia, Ltd. Laboratory, which had previously tested the ion to attempt to resolve this. The spin experiments drive for Deep Space 1. The lab also contains a were less ambiguous—the sail spun at exactly the high-power microwave test facility that was adapt- rate predicted for 10 percent absorption, and when ed for the purpose. Microwaves have several the beam’s polarization was reversed, the sail spun potential advantages as a propulsion beam. Micro- in the other direction. wave transmitters have been around a lot longer Either way, Marzwell envisions embedding the than lasers, and very large, high-power arrays are scrubbing pad with nanocomputers 10 microns currently much cheaper to build. (The latter is square (about one-tenth the thickness of a hair), an important consideration, as a microwave array and instruments to match. A micropump could needs to be considerably larger to support a sail of send captured interstellar material to a nano- the same diameter, in order to compensate for the spectrometer for analysis, for example, while longer wavelength of the microwaves.) And if the tiny cameras and dust-mote-sized vibration detectors helped navi- gate the spacecraft and The spacecraft becomes the sail—there’s no need for a separate hull dangling below it to carry the payload. monitor its health. Parallel programs are This is a major departure from today’s paradigm, where 80 to 90 percent of the weight on most missions is looking at the tech- nologies, like nanobat- fuel and the instrument package seems almost an afterthought. teries, needed to make these tiny devices sail is designed to absorb some of the microwaves, work. “You can’t develop a sail without develop- its interaction with a circularly polarized beam ing the instruments,” says Marzwell. In essence, will set it spinning. The torque increases with the spacecraft becomes the sail—there’s no need the beam’s wavelength, so the spinning effect for a separate hull dangling below it to carry the works much better with microwaves than it does payload. This is a major departure from today’s with lasers. The easiest way to keep a spacecraft paradigm, where 80 to 90 percent of the weight on a steady course is to set it spinning perpendicu- on most missions is fuel and the instrument larly to the direction of flight, like a rifle bullet or package seems almost an afterthought. Similarly, a well-thrown football. Controlling the sail’s spin a metallic mesh could be embedded into Mylar, rate and angle by manipulating the beam’s axis of which would be evaporated in orbit to leave the

28 ENGINEERING & SCIENCE NO. 4  mesh and its instruments. Choosing the sail’s geometry is a complex prob- lem, involving such variables as how the energy couples with the sail material, and the mechanical and elastic properties of the sail. “Shape is crucial because of the need to control the center of mass and center of gravity,” says Marzwell, who believes that a conical, sombrero-like geometry will eventually replace flat sails to obtain maximum photon capture. Such a sail would ride the beam more stably—if the sail’s pointy center started to slip off the center of the beam, the asymmetric pressure on the side of the cone would tend to move the sail back into alignment. Building the laser or microwave facility needed for these missions will be no small feat. Harris’s group estimates that the 40-year trip to Alpha Centauri would require a phased laser array 1,000 kilometers in diameter. (Planetary missions can get by with something more modest—a 15-meter array could send a 50-meter-diameter sail carrying a 10-kilogram payload to Mars in 10 days, he says in an article in Scientific American.) People are working on those issues, too, but that’s another story…. While JPL, the ESA, and the Planetary Society prepare for their first deployment tests with Kapton and Mylar, the advanced concepts and technology people prepare for a more distant, diamond future as they analyze the electrodynamic coupling between high-energy beams and sails varying in shape from thin sheets to balloons. Hopefully, this research will lead beyond missions delivering better space weather reports to explora- tions of interstellar space that will tell us how the stars, the rocky planets, and perhaps life itself evolved in the universe. ■

Joel Grossman is a freelance writer based in Santa Monica, California. He last wrote about solar sailing for the Los Angeles Times Magazine on August 20, 2000. He also wrote a piece about Professor of Electri- These solid-state instruments, which could potentially be cal Engineering Yu-Chong Tai’s bat-sized flying robot, which appeared in the Los Angeles Times Magazine embedded into an interstellar sail, have been developed by on December 10, 2000. JPL’s Center for Space Microelectronics Technology. Top, left: Of course, you’ve got to have cameras, as well as star sen- sors for navigation. This is a complementary metal-oxide semiconductor active pixel sensor. It requires one- For Further Reading hundredth the power of a CCD camera, is lighter, and is less Colin McInnes, Solar Sailing: Technology, susceptible to radiation damage in space. Top, right: A Dynamics and Mission Applications (Springer- microgyro jointly developed by JPL and Hughes for space- Verlag, 1999) craft attitude control. Bottom: A tunable diode laser, Jerome Wright, Space Sailing (Gordon and which can be set like a radio transmitter to emit any given Breach Science Publishers, 1992) Louis Friedman, Starsailing: Solar Sails and frequency within its range, allowing it to look for molecules Interstellar Travel (John Wiley and Sons, 1988) of a specific gas. Scientific American, February 1999

 ENGINEERING & SCIENCE NO. 4 29 30 ENGINEERING & SCIENCE NO. 4  He also has been accused of male chauvinism, anti-Semitism, mistreating his

graduate students, and, worst of all, scientific fraud.

In Defense of Robert Andrews Millikan by David Goodstein

Robert Andrews Millikan was the founder, first Thomson and his colleagues tried to do it by leader, first Nobel Prize winner and all-around observing how an applied electric field changed patron saint of the California Institute of Technol- the rate of gravitational fall of clouds of water ogy, an institution that has given me employment droplets that had nucleated on ions in a cloud for more years than I care to remember. He also chamber. The upper edge of the cloud, which had has been accused of male chauvinism, anti- the smallest droplets, could be assumed to contain Semitism, mistreating his graduate students, and, single charges. In this way, a crude but correct worst of all, scientific fraud. Since we at Caltech estimate of the unit of electric charge could be feel a solemn duty to defend our hero, my purpose obtained. These cloud-chamber experiments were Isaac Newton (framed) and here is to tell his story, look into these various the starting point of Millikan’s efforts. accusations, and, to the extent that I can, mount Working with a graduate student named Louis Robert Millikan: colleagues a defense for Professor Millikan. Begeman, Millikan had the idea of applying a in crime? Speaking of Millikan was born in 1868, son of a Midwestern much stronger electric field than had previously crime, the Newton portrait minister. He attended Oberlin College, got his been used, in the hope of stopping the descent of PhD in physics from Columbia University, did the cloud completely. To Millikan’s surprise, what was stolen from Good- some postdoctoral work in Germany, and, in the happened instead was that nearly all of the drop- stein’s office in 1979. last decade of the 19th century, took a position at lets with their different positive and negative (Photo courtesy of the brand-new University of Chicago in a physics charges dispersed, leaving in view just a few indi- Don Downie, Pasadena department headed by his idol, A. A. Michelson. vidual droplets that had just the right charge to During the next decade, Millikan wrote some permit the electric force to come close to balanc- Star-News.) very successful textbooks, but he made little ing the effect of gravity. Millikan quickly realized progress as a research scientist. This was a period that measuring the charge on individual ionized of crucial change in the history of physics. J. J. droplets was a method far superior to finding the Thomson discovered the electron, Max Planck average charge on droplets in a cloud. kicked off the quantum revolution, Albert Ein- It may have been during this period that stein produced his theories of relativity and the Millikan’s wife, Greta, attending a social event photoelectric effect, and Jean Perrin’s experiments while Millikan spent one of his many long eve- and Einstein’s theory on Brownian motion estab- nings in the lab, was asked where Robert was, lished forever that matter was made of atoms. according to unpublished remarks by Earnest C. Millikan made no contribution to these events. Watson (in the Caltech Archives’ Watson papers). Nearing 40 years of age, he became very anxious “Oh,” she answered, “He’s probably gone to watch indeed to make his mark in the world of physics. an ion.” “Well,” one of the faculty wives was later PICTURE CREDITS: He chose to try to measure the charge of the overheard to say, “I know we don’t pay our assis- 32, 34-35, 37 — Caltech electron. tant professors very much, but I didn’t think they Archives; 36 — American Scientist; 39 Cathode-ray tubes had been around for decades had to wash and iron!” — Bob Paz when, in 1896, Thomson in England succeeded in Unfortunately the single-droplet method had a showing that all cathode rays are electrically serious flaw. The water evaporated too rapidly to charged and have the same ratio of electric charge allow accurate measurements. Millikan, Begeman to mass. This was the discovery of the electron. and a new graduate student named Harvey It was the first demonstration that atoms had in- Fletcher discussed the situation and decided to try ternal parts. The challenge then was to measure to do the experiment with some substance that separately the electric charge of the electron. evaporated more slowly than water. Millikan

 ENGINEERING & SCIENCE NO. 4 31 Millikan’s oil-drop apparatus, shown above in his Chicago laboratory, had many components, including the hefty brass chamber (at right, set up for a much later demon- stration in 1969). A diagram taken from his controversial 1913 paper (bottom right) shows that the chamber contained two metal plates (M and N) to which he applied a high voltage, generated by a bank of batteries (B). Fine droplets of oil produced by a perfume atomizer (A) were fed into the top of the chamber. A tiny hole in the upper plate allowed the occasional droplet (p) to fall through, at which point it was illuminated by an arc lamp (a) and could be seen in magnification through a telescope. A manometer (m) indicated internal pressure. To eliminate differences in temperature (and associated convection currents), Millikan immersed the brass chamber in a container of motor oil (G), and he screened out the infrared components of the illumination using an 80-cm- long glass vessel filled with water (w) and another glass cell filled with a cupric chloride solution (d). An x-ray tube (X) allowed him to ionize the air around the droplet. With this equipment, Millikan could watch an oil drop that carried a small amount of charge rise when the applied electric field forced it upward and fall when only gravity tugged on it. By repeatedly timing the rate of rise and fall, he could determine precisely the electric charge on the drop.

32 ENGINEERING & SCIENCE NO. 4  Millikan proposed that Fletcher be sole author on the Brownian motion work and that he, Millikan, be sole author on the unit-electric-charge work. This is the source of the assertion that Millikan mistreated

his graduate students. No doubt Millikan understood that the measurement of e would establish his

reputation, and he wanted the credit for himself.

assigned to Fletcher the job of devising a way to two men remained good friends throughout their do the experiment using mercury or glycerin or oil. lives, and Fletcher saw to it that this version of the Fletcher immediately got a crude apparatus story was not published until after Millikan’s working, using tiny droplets of watch oil made death and his own. with a perfume atomizer he had bought in a Let us turn now to the question of scientific drugstore. He could view the droplets inside the fraud. In 1984, Sigma Xi published a booklet experimental chamber by illuminating them with called Honor in Science. More than a quarter of a a bright light and focusing a specially designed million copies were distributed before it was telescope on them. Through the eyepiece, he could replaced recently by a newer version. Honor in see the oil droplets dancing around in what is Science includes a brief discussion of the Millikan called Brownian motion, caused by impacts of case that begins, “One of the best-known cases of unseen air molecules. This itself was a phenom- cooking is that of physicist Robert A. Millikan.” enon of considerable current scientific interest. Cooking, meaning “retaining only those results When Fletcher got the busy Millikan to look that fit the theory and discarding others,” is one through his telescope at the dancing suspended of the classic forms of scientific misconduct, first droplets of oil, Millikan immediately dropped all described in an 1830 book by Charles Babbage work on water, and turned his attention to (Reflections on the Decline of Science in England: and its refining the oil-drop method. Causes). According to Honor in Science, it is a well- A couple of years later (around 1910) Fletcher established fact that Millikan cooked his data. and Millikan had produced two results. One was What is going on here? There are really two an accurate determination of the unit electric stories. One concerns the question of what actu- charge (called e) from observing the rate of fall or ally happened back in the period 1910–1917, and rise of oil drops in gravitational and electric fields, the other illustrates how, much more recently, he and the other was a determination of the product came to be accused, tried, and convicted of scien- Ne, where N is a separate constant called Ava- tific fraud. It’s time to tell both of these stories. gadro’s number. The product Ne came out of obser- The accusation against Millikan, very briefly, is vations of Brownian motion. Millikan approached this. After the 1910 paper (with Millikan alone, his student Fletcher with a deal. The academic not Fletcher, as author) presenting his measure- rules of the time allowed Fletcher to use a pub- ment of the unit of electric charge, Millikan found lished paper as his PhD thesis, but only if he was himself embroiled in controversy with the Vien- sole author. Millikan proposed that Fletcher be nese physicist Felix Ehrenhaft. Ehrenhaft, using a sole author on the Brownian-motion work and similar apparatus, found cases of electric charges that he, Millikan, be sole author on the unit- much smaller than Millikan’s value of e (Millikan electric-charge work. This is the source of the refers to these as “subelectrons”). In order to re- assertion that Millikan mistreated his graduate fute Ehrenhaft’s assertion of the existence of sub- students. No doubt Millikan understood that the electrons, Millikan (now working alone; Fletcher measurement of e would establish his reputation, had received his doctorate and left) made a new and he wanted the credit for himself. Fletcher series of measurements, published in 1913, in understood this too, and he was somewhat disap- which the charge on every single droplet studied pointed, but Millikan had been his protector and was, within a very narrow range of error, an in- champion throughout his graduate career, and so teger multiple of a single value of e. The 1913 he had little choice but to accept the deal. The paper succeeded in dispatching Ehrenhaft, and

 ENGINEERING & SCIENCE NO. 4 33 Let us look at some of the pages in Millikan’s private laboratory notebooks. The one at left is dated November 18, 1911. At the top right the temperature is noted as t=18.0ºC (obviously, Millikan’s lab was not well heated for the bitter Chicago weather) and the pressure, 73.45 cm (possibly a stormy day). On the left, we see a column of figures under G, for gravity. These were the times taken for a tiny droplet—a pin- point of light, too small to focus in his telescope— to fall between scratch marks in the telescope’s focal plane. These measurements gave the ter- minal velocity of the drop when the force of gravi- ty was balanced by the viscosity of air. From this measurement alone, he could determine the size of the tiny, spherical drop. contributed significantly to Millikan’s 1923 Nobel Then there is another column under F for field. Millikan’s lab notebooks Prize. An examination, however, of Millikan’s These were the times taken for the drop to rise private laboratory notebooks (housed in the between the scratch marks under the combined provide some insight into Caltech Archives) reveals that he did not in fact influence of gravity, viscosity, and the applied his methods. The page report every droplet on which he recorded data. electric field, which had been turned off during above, dated November 18, He reports the results of measurements on 58 the G measurements. The combined F and G 1911, shows his observa- drops, whereas the notebooks reveal data on measurements made it possible to determine the approximately 175 drops in the period between charge on the drop. We can see that the F mea- tions at left under G, for November 11, 1911, and April 16, 1912. In a surements change from time to time. The first gravity, and F, for field, and classic case of cooking, the accusation goes, he series give an average of 8.83, then 10.06, then his calculations to the reported results that supported his own hypothesis 16.4, and so on. That happens because the charge right. This drop and the of a smallest unit of charge, and discarded those on the drop changes from time to time, when the contrary results that would have supported Ehren- drop captures an ion from the air. Millikan made ones for which comments haft’s position. And, to make matters very much use of the changes to help deduce the number of are excerpted at top right, worse, he lied about it. The 1913 paper present- units of charge on the drop. were not among those ing Millikan’s results contains this explicit asser- To the right of these columns appears a series of tion: “It is to be remarked, too, that this is not a laborious hand calculations (not necessarily done included in Millikan’s selected group of drops, but represents all the drops on the same day that the data were taken), using published paper. The experimented upon during 60 consecutive days, during logarithms to do multiplication and square roots, experiment dated March which time the apparatus was taken down several and then finally, bottom right, the comment, 14, 1912 (opposite page), times and set up anew.” (Emphasis in the origi- “very low something wrong” with arrows to “not nal.) Thus, Millikan is accused of cheating and sure of distance.” Needless to say, this was not one with its exuberant red then compounding his cheating by lying about it of the 58 drops Millikan published. notation, was indeed in one of the most important scientific papers of Another page shows observations on two drops, published. the 20th century. There couldn’t be a clearer case taken November 20 and 22, 1911, with similar of scientific misconduct. columns of figures (excerpt above). To the right at

34 ENGINEERING & SCIENCE NO. 4  the bottom of the first observation we see again Note also the pressure, 16.75 cm—too low for “very low something wrong” and below that, even the stormiest day in Chicago. “found meas[uremen]t of distance to the hole did During these 63 days, Millikan recorded in his not. . . .” Once again, not up to snuff. But on a notebooks data for about 100 separate drops. Of page dated December 20, 1911 (the temperature these, about 25 are obviously aborted during the now a comfortable 22.2ºC—did the university run, and so cannot be counted as complete data turn the heat on in December?), we find the sets. Of the remaining 75 or so, he chose 58 for remark: “This is almost exactly right & the best publication. Millikan’s standards for acceptability one I ever had!!!” (left). were exacting. If a drop was too small, it was Millikan, in his crucial 1913 paper, did not excessively affected by Brownian motion, or at publish any of the drops for which the raw data are least by inaccuracy in Stokes’s law for the viscous shown in these three pages, not even “the best one force of air (more about this later). If it was too I ever had.” This was all part of a warm-up period large, it would fall too rapidly for accurate mea- during which Millikan gradually refined his appa- surement. He also preferred to have a drop change ratus and technique in order to make the best its charge a number of times in the course of an measurements anyone had ever made of the unit of observation, so that he could have changes in electric charge. The first observation that passed charge, as well as a total charge, which had to be muster and made it into print was taken on Feb- integer multiples of a single unit of charge. None ruary 13, 1912, and all of the published data were of this could be determined without actually tak- taken between then and April 16, 1912, actually ing and recording data on a candidate drop. Thus, a period of 63 days (1912 was a leap year). Raw it should not be surprising that Millikan chose to data taken during this period are shown in the use the data on only 58 of the drops he observed notebook page below, dated March 14, 1912. Our during the period when he and his apparatus had eye is immediately drawn to the comment, on the reached near perfection. Furthermore, he had no top center part of the page, “Beauty Publish.” special bias in choosing which drops to discard. A modern reanalysis of Millikan’s raw data by Allan Franklin (see following page) reveals that his result for the unit of charge and for the limits of uncer- tainty in the result would barely have changed at all had he made use of all the data he had, rather than just the 58 drops he used. I don’t think that any scientist, having studied Millikan’s techniques and procedures for conduct- ing this most demanding and difficult experiment, would fault him in any way for picking out what he considered to be his most dependable measure- ments in order to arrive at the most accurate possi- ble result. In the 1913 paper, he cites his result with an uncertainty of 0.2 percent, some 15 times better than the best previous measurement (which reported an error of 3 percent). Furthermore, the value of the charge of the electron today agrees with Millikan’s result within his cited uncertainty of 0.2 percent. The experiment was nothing less than a masterpiece, and the 1913 paper reporting it is a classic of scientific exposition. Nevertheless, it contains the phrase “this is not a selected group of drops, but represents all the drops experimented upon during 60 consecutive days,” which is mani- festly untrue. The question is, why did Millikan mar his masterpiece with a statement that clearly is not true? Many years after the fact, Millikan’s work was studied by historian Gerald Holton, who told the story of the Millikan-Ehrenhaft dispute (“Sub- electrons, Presuppositions, and the Millikan- Ehrenhaft Dispute,” in Historical Studies in the Physical Sciences, 1981) and contrasted Millikan’s published results with what he found in Millikan’s laboratory notebooks. Holton did not accuse Millikan of misconduct of any kind, but instead found in the unpublished laboratory notebooks an

 ENGINEERING & SCIENCE NO. 4 35 the existence of electrons, and by implication, the existence of atoms, was an issue of burning con- troversy in 1913, with Millikan on one side and Ehrenhaft on the other, and that the whole point of Millikan’s exercise was to prove that “sub- electrons” did not exist. In fact, there were in 1913 a small number of respectable scientists who still insisted that the existence of unseen atoms was an unnecessary and unscientific hypothesis, but they had by then been left far behind by the mainstream of science, and besides, even they would not have chosen Ehrenhaft as their cham- pion. To Millikan, who had seen Brownian motion with his own eyes, the existence of atoms and electrons was beyond question. Every revision Allan Franklin’s reanalysis of his technique, every improvement of his appa- of Millikan’s observations ratus, every word he wrote, public or private, was on published (solid) and directed toward one goal only: the most accurate possible measurement of the charge of the elec- unpublished (open) oil tron. Ehrenhaft and the supposed controversy are drops demonstrates that opportunity to contrast a scientist’s public, pub- never so much as mentioned. And it is worth the scatter in his results lished behavior with what went on in the privacy remembering that history has vindicated Millikan lessened over time. Shown of the laboratory. Holton’s work was seized upon in that his result is still regarded as correct. by two journalists, William Broad and Nicholas Nevertheless, we are still stuck with the blatantly here are estimates of e Wade, who in 1982 published a book about mis- false statement, “[A]ll the drops experimented derived from all the drops conduct in science called Betrayers of the Truth. upon during 60 consecutive days.” for which Millikan obtained Broad and Wade, both of whom were then re- To understand the significance of that state- porters for Science magazine, and both of whom ment, I must make a small digression. Millikan’s adequate data after now write for the New York Times, are the ones who oil drops rose and fell under the influence of three February 13, 1912, the day tried and convicted Robert Millikan of scientific countervailing forces: gravity, electricity, and he gathered observations misconduct. Others, like the writer of Sigma Xi’s viscosity. The first two of these were very well on the first of the 58 Honor in Science, simply bought their argument at understood. For the third, the 19th-century face value. hydrodynamicist George Stokes had produced an drops he ultimately In Betrayers of the Truth, Broad and Wade want exact formula applicable to a sphere moving published. Millikan was to make the point that scientists cheat. Chapter 2, slowly through an infinite, continuous viscous clearly being selective, but “Deceit in History,” starts out with a list of cul- medium. The conditions that would make his choice of drops did not prits: Claudius Ptolemy, Galileo Galilei, Isaac Stokes’s law exact were well-satisfied by Millikan’s Newton, John Dalton, Gregor Mendel, and Robert oil drops in all respects except one: the drops were bias the overall result. Millikan. At the very least, Millikan is in good so small that the air through which they moved company. Of Millikan they say he “extensively could not safely be considered a continuous misrepresented his work in order to make his medium. Instead, the air was made up of mol- experimental results seem more convincing than ecules, and the average distance between molecules was in fact the case.” was not completely negligible compared to the size of an oil drop. For this reason, Stokes’s law could not be depended on as absolutely correct. To deal with this problem, Millikan assumed, Every revision of his technique, every improvement of his apparatus, every entirely without theoretical basis, as he stressed in his paper, that Stokes’s law could be adequately word he wrote, public or private, was directed toward one goal only: the most corrected by an unknown term that was strictly accurate possible measurement of the charge of the electron. proportional to the ratio of the distance between air molecules to the size of the drop, so long as that ratio was reasonably small. To test this idea, he purposely made that damaging ratio larger than it had to be by pumping some of the air out of his I would argue that this statement is profoundly experimental chamber. That is the reason he incorrect. (The accusations against most of the recorded such low pressure in the page from his other scientists on the list are equally spurious— notebook dated March 14, 1912. Then, when he see “Scientific Fraud,” E&S, Winter 1991.) had assembled all of his data, he used a trick that For the statement by Broad and Wade to make would be appreciated by any experimentalist. He sense, Millikan’s principal experimental result plotted a graph of all his data in such a way that, would have to be that there exists a smallest unit if his supposition was correct, all the data points of electric charge. We would have to imagine that would fall on a single straight line, and the posi-

36 ENGINEERING & SCIENCE NO. 4  tion of the line on the graph would give the mag- straight-line test of the correction to Stokes’s law nitude of the unknown correction term. Thus, if described above. On page 138, Millikan discusses it were successful, this procedure would all at once his test of his presumed correction to Stokes’s law. prove that the proposed method of correcting He points out that all of the points do indeed fall Stokes’s law was justified, and give the magnitude on the line, and in fact, “there is but one drop in of the necessary correction. In other words, this the 58 whose departure from the line amounts to procedure, like everything else in this experiment, as much as 0.5 percent.” And then, the very next was designed not to question whether charge came sentence is, “It is to be remarked, too, that this is in units, but rather to measure the unit of charge not a selected group of drops, but represents all with the greatest possible accuracy. the drops experimented upon during 60 consecu- Now let’s turn to Millikan’s actual published tive days . . .” The damning remark is made, not paper. It begins on page 109 of Volume II, No. 2, in regard to whether charge comes in units, but in of the Physical Review. He explains how the experi- regard to getting the correction to Stokes’s law ment is done and (with specific drops as examples) right. What he means to say is, “Every one of how he analyzes his data, using changes in the those 58 drops I told you about confirms my charge on a drop to help determine the total num- presumed formula for correcting Stokes’s law.” ber of units of charge on the drop. Then, on page And, although in Physical Review it comes five 133, he writes: “Table XX contains a complete pages after the remark that qualified the choice summary of the results obtained on all of the 58 of those 58 drops, the intervening pages are tables different drops upon which complete series of and graphs. In the typescript submitted by observations were made during a period of 60 Millikan (which does not survive, to my knowl- consecutive days.” As we have already seen, his edge) it would have followed almost immediately published results came from measurements made after the qualifying statement. Thus a careful over a period of 63, not 60 days, but I think we reading of the context of Millikan’s words greatly can forgive him that lapse. The clear implication diminishes their apparent significance as evidence of the sentence is that there were only 58 drops for of misconduct. which the data were complete enough to be in- In fairness, it should be pointed out that, when cluded in the analysis. Millikan published his book The Electron in 1917, Page 133 is followed by two pages of Table XX, he did take the trouble to confront Ehrenhaft and an additional two pages of the graph of the explicitly and to demolish Ehrenhaft’s arguments

Cosmic rays were the subject of Millikan’s research in the latter half of his career, after coming to Caltech in 1921 at the age of 53. Not only did he give them their name and call attention to their significance, he also lugged detectors such as this one to various altitudes around the world to measure the rays. Millikan died in 1953.

 ENGINEERING & SCIENCE NO. 4 37 rather than if I filled one of my openings with a woman.” In his private correspondence, Millikan also reveals an attitude toward Jews that would not be acceptable today. For example (as noted in Millikan’s School by Judith R. Goodstein), writing from Europe to his wife, Greta, he describes I believe, after reading The Electron, that Millikan’s real rival was never the physicist Paul Ehrenfest (not to be confused with Felix Ehrenhaft) as “a Polish or Hungarian Jew hapless Ehrenhaft, but rather J. J. Thomson—not because they disagreed [Ehrenfest was, in fact, Austrian] with a very short, stocky figure, broad shoulders and abso- scientifically, but because both wanted to be remembered in history as the lutely no neck. His suavity and ingratiating father of the electron. manner are a bit Hebraic (unfortunately) and to be fair, perhaps I ought to say too that his genial open-mindedness, extraordinarily quick percep- tion and air of universal interest are also character- istic of his race.” What are we to make of these lapses? They are certainly not the rantings of a mindless bigot. Undoubtedly Millikan’s biases were typical at the very effectively. He also used verbatim the section time of a man of his upbringing and background. of his 1913 paper on Stokes’s law, thus repeating It should be said that, regardless of whatever the offending assertion of having used every drop, prejudices he harbored, they never interfered with without the earlier qualifying statement. Most his judgment of scientists. His hero A. A. probably by 1917, he had forgotten the very Michelson was Jewish, as were many of the stars existence of the other drops he had observed, how- Millikan personally recruited to Caltech: Paul ever incompletely, between February and April of Epstein, Albert Einstein, Theodore von Kármán, 1912. I believe, after reading The Electron, that and Beno Gutenberg among others. Such actions Millikan’s real rival was never the hapless Ehren- demonstrate that Millikan’s personality was more haft, but rather J. J. Thomson—not because they complex than his detractors acknowledge. Like disagreed scientifically, but because both wanted anyone, he had his strengths and his flaws. He to be remembered in history as the father of the wasn’t generous enough to put his student’s electron. interests ahead of his own at a critical point in his In recent times, Millikan has become a juicy career. In describing the results of his oil-drop target for certain historians because he was very experiment, he let himself get carried away a bit much a part of the establishment, as well as being in demonstrating the correctness of his empirical white and male, and, of course, he is no longer correction to Stokes’s law. And his words about here to defend himself (I’m trying to fill in on that women and Jews grate on modern sensibilities. last point). For example, there is a letter, noted in But Robert Andrews Millikan was not a villain. feminist circles, in which Millikan advised the And he certainly did not commit scientific fraud president of Duke University not to hire a female in his seminal work on the charge of the electron. professor of physics. This occurs much later, in Ladies and gentlemen, the defense rests. ■ 1936, and Millikan is now the famous and power- ful head of the California Institute of Technology. David Goodstein, professor of physics and applied W. P. Few, Duke’s president, had written to Milli- physics, the Frank J. Gilloon Distinguished Teaching kan in confidence, asking his advice on this deli- and Service Professor, and vice provost, has, he claims, cate issue. Millikan’s reply shows his unease: “I the longest title at Caltech and possibly in all of aca- scarcely know how to reply to your letter. . . .” he demia. He has also been a member of the faculty a begins. “Women have done altogether outstand- rather long time, having joined it in 1966 after re- ing work and are now in the front rank of scien- ceiving his PhD from the University of Washington. tists in the fields of biology and somewhat in the Although his own research has been in low-temperature fields of chemistry and even astronomy,” Millikan physics, Goodstein has been a loyal author for E&S over writes later, “but we have developed in this coun- the last couple of decades on topics ranging from scientific try as yet no outstanding women physicists.” He fraud to superconductivity to the excess supply of PhDs, points out that “Fräulein Meitner in Berlin and as well as contributing numerous book reviews. E&S Madame Curie in Paris” are among the world’s hasn’t always been the primary recipient, as is the case best physicists, but that’s Europe, not the U.S. with this article, which is adapted from an address to “I should therefore,” he concludes his confidential the 2000 Sigma Xi Forum, “New Ethical Challenges advice, “expect to go farther in influence and get in Science and Technology,” in November, where he more for my expenditure if in introducing young received Sigma Xi’s John P. McGovern Science and blood into the department of physics I picked one Society Award. The lecture was first published in the or two of the most outstanding younger men, January–February 2001 issue of American Scientist.

38 ENGINEERINGE NGINEERIN & SCIENCE NO. 4  O b i t u a r i e s

B RAD S TURTEVANT 1933 – 2000 Sturtevant with GALCIT’s 17-inch shock tube in the early 1960s.

Bradford (Brad) Sturtevant, “‘Why don’t you all go back lar beam that he built for his the Hans W. Liepmann Pro- to your labs and do an honest PhD thesis under Liepmann, fessor of Aeronautics, died day’s work?’” (Sturtevant was Sturtevant became interested October 20, of pancreatic well known for working on in kinetic theory and then in cancer at the age of 66. Saturdays.) But everyone shock wave structure. He Arriving at Caltech as a stayed, and the memorial went on to apply his shock graduate student in 1955, continued. wave research to motorcycle with a bachelor’s in engineer- Victoria Sturtevant stated noise and sonic booms and to ing from Yale, Sturtevant three things she had learned fields as disparate as geology earned his PhD in 1960 and from her father: “If it hurts, and medicine. With Hor- stayed on at GALCIT (Cal- it’s good for you”; “think nung, he built the T5 Hyper- tech’s Graduate Aeronautical about it and work it in your velocity Shock Tunnel in Laboratory) for the rest of his head to figure out how it 1988. Hornung joked that, career. He was preceded by works before you break it”; throughout construction, his great uncle, Alfred Sturte- and “choose a career where “Brad wanted everything to vant, who, with Thomas you’ll constantly learn.” Brad be done right,” even the Hunt Morgan, was among the Sturtevant took these tenets cleaning up. (A slide showed founders of the Division of extremely seriously himself, Sturtevant directing his boss Biology in 1928. But while as his colleagues, students, at the vacuum cleaner.) the elder Sturtevant studied and friends proceeded to Representing colleagues the genetics of fruit flies (and attest. from the Indiana University irises), Brad’s research was in Hans Liepmann, the Theo- School of Medicine, Andy fluid dynamics, particularly dore von Kármán Professor of Evan spoke of Sturtevant’s shock waves and nonsteady Aeronautics, Emeritus, and work with a group interested “Brad made a magical gas dynamics. He was named former GALCIT director, re- in shock wave lithotripsy. associate professor in 1966, called Sturtevant’s style in The Indiana group had met connection, not only between full professor in 1971, and designing experiments: “very him in 1988, when both were appointed to the Liepmann prepared. He thought he seeking other investigators. aeronautical engineering and chair in 1995. could design an experiment “Brad was very interested in geology and geophysics, but A memorial gathering in that would work the first how shock waves break up remembrance and celebration time,” whereas others, includ- kidney stones,” said Evan. between engineering and of Sturtevant’s life was held ing Liepmann, had a different “We were interested in how Saturday, February 24, fol- approach: “We would first do shock waves caused damage science.” lowing a reception and buffet it lousy and then a little bet- to tissue. It seemed a perfect and Scott Joplin piano music. ter and then a little better.” match for collaboration.” Hans Hornung, the C. L. Liepmann noted that Stur- Despite initial disappoint- “Kelly” Johnson Professor of tevant exemplified GALCIT’s ment in attracting NIH Aeronautics and director of mission: “We do not want to funding, it was Brad’s opti- GALCIT, surveyed the large produce specialists but we mism and contagious enthu- crowd in Dabney Lounge and want to produce people who siasm that kept the project declared that if Sturtevant can specialize wherever they going, said Evan. “I don’t could see this, he would say, want to.” From the molecu- need to remind you how

    ENGINEERING & SCIENCE NO . 4 39 incredibly bright Brad really from Brad is what it is to be a of a house” and competed was, how gifted he was at scientist, and I thank him for in many ocean sailing races analysis, and what a fertile that,” said Brouillette. together. flow of original ideas he Another of those PhD Sturtevant was especially generated.” students, Joe Shepherd (’81), renowned for swimming, His forays into volcanology professor of aeronautics, who open-water swimming races were described by geologist served as master of ceremo- in particular, for which he Sue Kieffer, PhD ’71, who, nies at the memorial and also won numerous prizes and although she had audited one produced the slide show, said, honors. He was a regular lap of Sturtevant’s courses, didn’t “We all learned, I believe, an swimmer at the Caltech pool run into him again until enormous amount from Brad, as well. During the early Seminar Day 1981, when she both on the personal and ‘80s, he worked out three gave a talk on the Mount St. scientific level.” Shepherd times a day, said Michael Helen’s eruption. Sturtevant also led into Sturtevant’s Hoffmann, the James Irvine challenged her conclusions, “other life” as a vigorous ath- Professor of Environmental and the encounter led to a lete. “When I described Brad Science and a fellow member “fruitful collaboration re- Sturtevant to the local news- of the faculty athletic com- sulting in two papers that paper recently,” said Shep- mittee for many years. “He proved a number of things herd, “I found that I had swam in the morning, lifted about the destructive forces almost completely omitted weights and ran at noontime, of the supersonic nature of the fact that he was also a and then swam again in the the blast of 1980.” scientist, and so the headline evening. And each workout “Brad made a magical con- came out that he was a lasted 90 minutes or so. He nection, not only between ‘sportsman.’” His “perfectly also would brag that up until aeronautical engineering and healthy” life was regarded about two years ago, every geology and geophysics, but with awe. “On one thing I year he bested his benchmark between engineering and disagree with him after the time from his days on the science,” said Kieffer. “Nei- fact,” said Liepmann. “He Yale swim team, which is a ther of these is a mean feat died so early; you probably remarkable accomplishment.” given the vastly different should not believe in doing But Hoffmann punctured the content and training of the everything to remain image of Sturtevant’s per- researchers in the different healthy.” fectly healthy life: “I used to fields.” But health wasn’t what it see him in the morning after Interdisciplinary research was all about. “He loved the his workouts, eating dough- may have its downside, how- mountains and the oceans and nuts over at the Greasy.” ever. Added Liepmann, “I, everything outdoors,” said And he drank wine too, for one, firmly believe that it Anatol Roshko (PhD ’52), the said Tim Downes, director of actually reduced the number Theodore von Kármán Pro- athletics, who was pleased to of Brad’s honors and the ex- fessor of Aeronautics, Emeri- see that wine was served at tent of his support. When it tus. Roshko described and the memorial buffet. “Some comes to voting for an award, showed slides of a 1957 hike of my fondest memories of the tendency to keep it with- in the Sierra to the Ionian Brad are of sitting next to in your own narrow group is Basin, which illustrated him at endless athletic con- widespread.” Sturtevant’s penchant for ference meetings after we had Half of the 28 students planning and organization. had a couple of glasses of who received their PhDs “Along the way, he seemed to wine,” said Downes. Sturte- under Sturtevant returned to know every feature, every vant, as well as Hoffmann and attend the memorial, some elevation, every contour, and Downes, represented Caltech traveling long distances. the hike worked out exactly over the years at the Southern Several, including Willie the way he planned it,” said California Intercollegiate Behrens (‘66), Martin Roshko. “Whatever he un- Athletic Conference; Sturte- Brouillette (’85), and Bert dertook, whether it was in his vant also served several terms Hesselink (’77), offered science, or swimming, or hik- as chairman. affectionate reminiscences ing, or whatever, he did it all He was also a key figure in of the man as adviser—his thoroughly.” the planning and construc- “enormously high standards” He had hiked and sailed tion of the Braun Athletic and demanding presence, his since boyhood, and when he Center. And when that was energy, enthusiasm, and his came to California, he also completed, he “re-upped for a insistence on the proper use took up surfing. In the ’60s, second tour of duty,” on the of the English language (even he and his wife, Carol, whom Sherman Fairchild Library of semicolons). But “the most he had met at the Caltech Engineering and Applied important thing I learned pool, “bought a boat instead Science, according to Kim-

40 ENGINEERING & SCIENCE NO . 4     berly Douglas, director of that library. “He was not a limelight guy; his name does not appear in the library nor on the gym,” said Douglas, but “these buildings are certainly testimony to his willingness to give up his time and considerable talent to make the Institute a better place for all.” Added Shep- herd, “He had a profound sense of responsibility to the community here at Caltech.” While Sturtevant’s col- leagues, friends, and students had described him as exact- ing, rigorous, creative, imagi- native, energetic, competi- tive, intense, and great fun, the Rev. Douglas Vest, his next-door neighbor in Rubio J . H AROLD W AYLAND Canyon for 25 years, thought 1909 – 2000 one attribute had been left out. He remembered “the tender side of Brad.” When J. Harold Wayland, pro- on the trail to Mount Baldy, he left the house at 5 a.m. for fessor of engineering science, when we were doing what the pool, he would roll his car emeritus, and a pioneer of young graduate students fre- down the hill before turning bioengineering, died October quently did—exploring the on the ignition so as not to 10, 2000, at the age of 91. local mountains,” said Pick- wake the neighbors. Vest, At a memorial service held ering. Their two families who ended the program, also January 29 in Dabney remained close over the fol- told of sitting in silence with Lounge, friends and col- lowing decades, and Picker- his longtime neighbor last leagues remembered his ing recalled fondly a visit to summer after his disease had “Harold was a great man in contributions to science and the Waylands in Strasbourg been diagnosed, “and I to the Caltech community. in 1953. “They started the realized he was comfortable applying classical physics to Wayland’s long member- tradition of globetrotting about his life; he was com- ship in that community dates very early.” fortable about his family; he biology. He had a vision of back to 1931, when he ar- As early as 1936, in fact, was comfortable about the rived at Caltech as a graduate when Wayland left for a year’s unknown.” the new field of bio- student after earning his BS fellowship to Niels Bohr’s from the University of Idaho. Institute for Theoretical An intercollegiate varsity engineering long before the swimming award has been He earned his MS in 1935 Physics in Copenhagen. established in Sturtevant’s word was coined.” and his PhD in physics and Then, after a year as assistant honor, but, like his great mathematics in 1937—with professor of physics at the uncle, he will also have a Robert A. Millikan as his University of Redlands, he physical tribute on campus thesis adviser, according to returned to Caltech as a re- that is uniquely associated George Housner, the Braun search fellow from 1938 to with his life at Caltech. The Professor of Engineering, 1941. He spent the years iris garden north of Gates Emeritus, who spoke of his 1941–1944 degaussing ships Annex, planted in memory 50-year-long friendship with for the Navy in Long Beach, of Alfred Sturtevant, is Wayland. and then joined Caltech’s populated with descendants Bill Pickering had known project for the Navy’s Under- of the irises he bred in his him even longer. Pickering, water Ordnance Division, later genetic studies. In Brad professor of electrical engi- where he was involved with Sturtevant’s memory, a Ja- neering, emeritus, and former other Caltech faculty in de- cuzzi will be constructed at director of the Jet Propulsion signing the torpedo launcher Braun Athletic Center so that Laboratory, met Wayland in at Morris Dam. “he will always have a place the mid ’30s, when both were In 1949, he returned per- on the pool deck.” ■ graduate students. “If I manently to academia as remember rightly, I think the associate professor of applied first time I met Harold was mechanics at Caltech. He

    ENGINEERING & SCIENCE NO . 4 41 studies of microcirculation. Two other professors of bioengineering at UC San Diego also remarked on their early experience working With his microscope (at one time with Wayland. Paul Johnson, referred to as a “megascope”), now emeritus, came to Cal- tech for a year in 1965 to Wayland, shown here in 1971, work on developing quantita- could observe blood flow in a tive techniques for measuring living animal’s tinest blood vessels. blood flow. “Harold had an enormous influence on my life and career. But I think what I’ve described in my own experience you could duplicate for dozens of other people who came and spent time in his laboratory. He was absolutely focused on making the very best research was named full professor in To accomplish this, he built engineering long before the experience for anyone who 1957, professor of engineer- a huge microscope, seven and word was coined, and he was came to work with him.” ing science in 1963, and re- a half feet tall, weighing a evangelical in bringing the And Marcos Intaglietta, tired as emeritus professor in couple of tons (“big enough message of the importance of PhD ’63, who was a student 1979. to hold a goat,” according to physical optics to biology and of Wayland’s, noted that, “in From the early ’50s, Sobin). Wayland also con- medicine.” Taking his cue my cosmology Harold re- Wayland studied streaming ducted research on the impact from George Ellery Hale, said mains the archetype of the birefringence, or double of diabetes mellitus on blood Fung, Wayland preached the professor. He is a rigorous refraction, as a means of flow and on the molecular establishment of international person, unyielding on prin- visualizing complex fluid exchanges between blood and “intravital observatories,” ciples, but vastly human in flows, in particular large tissues that occur at the level which would do for biology education and hospitality.” molecules in fluids. When of the smallest vessels in the what Mount Wilson and Pal- Intaglietta also remarked on Wallace Frasher and Sidney body. omar had done for astronomy. Wayland’s vision: “He en- Sobin started a cardiovascular In the early ’60s, a Caltech This may be Wayland’s great- visioned the contribution that research laboratory, specializ- aeronautics professor, Yuan- est legacy, claimed Fung. “I engineering was poised to ing in small blood vessels, at Cheng Bertram Fung (PhD think the idea is still alive make in the life sciences and USC’s medical school in ’48), became interested in the and will probably wait for acted upon it. . . .His labora- 1957, they looked around Los possibilities of analyzing the future people to come”— tory became the host to tech- Angeles for some assistance in forces and stresses of the most likely in Japan and nicians and physiologists, a mathematics and physics. human body as thoroughly as China, where Wayland had pioneering initiative when They found no one at USC or those of airplanes, and joined many “disciples” and was very ‘interdisciplinary’ was not a UCLA, “so we looked up the the group working on micro- well known, according to household word. . . . He saw hill to Caltech,” recalled circulation. In 1966, Fung Fung. the future and caused it to Sobin. “I knew one person left for UC San Diego, where In later years Wayland’s start.” here, told him we wanted he subsequently laid the work was recognized at home Many of the memorial someone from the physical framework for the new field and abroad. The U.S. Micro- speakers commented on sciences who would work of bioengineering. Many of circulatory Society gave him Wayland’s enthusiasm for with us. He said, ‘Your man’s the original Caltech group its highest honor, the Landis his work and for his many Harold Wayland.’” joined him, including Sobin, Award, in 1981. And in interests outside the lab, So Wayland, working with now professor of bioengineer- 1988, he was awarded the especially music and the Frasher and Sobin, began to ing, emeritus, at UCSD, who Malpighi Award of the Euro- history of playing cards, focus on microcirculation, the credits Wayland with starting pean Society for Microcircula- on which he and his wife, flow of blood through the the spin-off. In December tion (Malpighi, in 17th- Virginia, wrote several books. tiny capillaries. Wayland 2000, Fung was awarded the century Italy, was the first to Ward Whaling, Caltech supplied the knowledge of National Medal of Science. observe blood cells in capil- professor of physics, emeritus, optics, mathematics, engi- (See Caltech News, No. 4, laries through a microscope). described another enthusiasm neering, and fluid dynamics. 2000.) The gold Malpighi medal was of his: “On this campus, His particular contribution “Harold was a great man in presented to him at the Uni- Harold was known first and was in devising quantitative applying classical physics to versity of Maastricht, which foremost for his interest in methods and building instru- biology,” said Fung at the Wayland had helped develop food and wine.” Back in the ments to observe and measure memorial service. “He had a into one of the leading insti- early ’60s, when the Ath- blood flow in living animals. vision of the new field of bio- tutions in the world for enaeum had no bar, recounted

42 ENGINEERING & SCIENCE NO . 4     Whaling, and the dining Faculty File room served plain, inexpen- sive food to postdocs and grad students, Wayland recruited the manager, trained as a chef in France, to collaborate on elegant, “private” dinners, for which Wayland would bring the wines, pour them, and discuss them. He had be- come a connoisseur of wine on his early jaunts to Europe and had struck up acquain- tance with some of the patriarchs of the California wine industry. Other mem- bers of the group, called the H ONORS AND A WARDS Apicians, soon began to take turns planning “private” dinners. “And that,” said Whaling, who along with his Clarence Allen, professor International Prize for Biol- Society’s 2001 Robert L. wife was one of the original of geology and geophysics, ogy. Awarded annually since Coble Award for Young members, “is the way fine emeritus, has been selected to 1985 by the Committee on Scholars. food and wine first made its receive the 2001 George W. the International Prize for Janet Hering, associate way into the Athenaeum, Housner Medal, awarded at Biology, the prize was professor of environmental which is now judged to be the annual Earthquake Engi- presented to Benzer on engineering science, has one of the most elegant din- neering Research Institute November 26 at the Japan received a grant of $100,000 ing rooms in Pasadena. I meeting, February 9, in Academy, in the presence of from the Alice C. Tyler think Harold would count Monterey, California. The the emperor and empress. Perpetual Trust. The grant that as one of his worthy award recognizes his “sus- Michael Brown, assistant will fund Hering’s project, accomplishments, and one tained and significant con- professor of planetary astron- “Environmental Quality Near that his colleagues recognize tributions to earthquake omy, has been selected by the Large Urban Areas,” which as a notable contribution to safety.” American Astronomical So- will examine the effects of a the campus.” Tom Apostol, professor of ciety’s Division for Planetary growing population and the Virginia Wayland died mathematics, emeritus, has Sciences to receive the Harold impact of human interaction January 7, 2001, and on been elected a corresponding C. Urey Prize in Planetary on land and aquatic ecosys- January 26, Whaling and member of the Academy of Science. tems in the San Gabriel Noel Corngold, professor of Athens. The academy is the Richard Ellis, professor of Valley and San Gabriel River applied physics, organized most prestigious scientific astronomy and director of watershed. one last Apicians dinner organization in Greece. Palomar Observatory, has Alice Huang, senior coun- (there had been 146 of them Frances H. Arnold, Dick been appointed the Lans- cilor for external relations and in all) in honor of Harold and and Barbara Dickinson downe Lecturer at the Univer- faculty associate in biology, Virginia Wayland. ■ Professor of Chemical sity of Victoria, Canada. He has been selected to receive Engineering and Biochemis- will deliver three lectures the 2001 Alice C. Evans try, has been elected a fellow there later in the year. Award, which is sponsored by of the American Institute for Sunil Golwala, Millikan the ASM (American Society Medical and Biological Postdoctoral Scholar, has for Microbiology) Committee Engineering. received the American Physi- on the Status of Women in David Baltimore, president cal Society’s Mitsuyoshi Microbiology. of Caltech, has been awarded Tanaka Dissertation Award in Tracy Johnson, postdoctoral the 2000 Warren Alpert Experimental Particle Physics scholar in biology, will be Professor of Political Science Foundation Prize for his work “for his versatile and exten- honored at the Roy Campa- Jeffrey Scot Banks, PhD ’86, “in the development of Abl sive contributions to the nella Humanitarian Award died of complications of a bone kinase inhibitors for use in detectors, hardware, electron- Dinner, on March 29 at the marrow transplant on December the treatment of chronic ics, software, and analysis of Pasadena Hilton Hotel. The 21. He was 42. A memorial myelogenous leukemia.” the results of the Cryogenic award honors “outstanding service will be held April 7 at Baltimore will share the Dark Matter Search (CDMS) leaders who have distin- 3 p.m. in Dabney Lounge. $150,000 prize with four experiment.” guished themselves in their Excerpts from that service will other scientists. Sossina Haile, assistant fields.” appear in the next E&S. Seymour Benzer, Boswell professor of materials science, Jonas Peters, assistant Professor of Neuroscience, has been selected to receive professor of chemistry, is one Emeritus, has received the the American Ceramic of 59 young researchers

    ENGINEERING & SCIENCE NO . 4 43 Observatory and the Interfer- entific publication of out- ometry Science Center, has standing original research in been honored with two invi- mineralogy.” tations, one to be the Univer- Nai-Chang Yeh, professer sity of Edinburgh Science of physics, has been awarded Festival Lecturer for 2001, the Achievement Award by the other to be the Philips the Chinese-American Fac- Visitor at Haverford College ulty Association of Southern for spring 2001. California “for her outstand- Wallace Sargent, Bowen ing contributions to ex- Professor of Astronomy, has perimental condensed matter been awarded the Henry physics, particularly in the Norris Russell Lectureship for areas of high-temperature su- 2001 by the American Astro- perconductivity and state-of- nomical Society. The lecture- the-art frequency standards.” ship is the society’s “most She was also elected a Fellow named by President Clinton prestigious prize and is of the Institute of Physics as a recipient of the Presiden- awarded annually to recog- with the title of Chartered tial Early Career Award for nize a lifetime of preeminence Physicist. ■ Scientists and Engineers. in astronomical research.” Stephen Quake, associate Paul Sternberg, professor of professor of applied physics biology, who is also an S TEVENSON W INS has been named one of the investigator for the Howard F EYNMAN P RIZE “Technology Review Ten” by Hughes Medical Institute, MIT’s Technology Review maga- has been elected a fellow of zine for his innovative work the American Academy of David J. Stevenson, the in the branch of biotechnol- Arts and Sciences. George Van Osdol Professor ogy known as microfluidics, Ersan Üstündag, assistant of Planetary Science, has been which involves manipulating professor of materials science, awarded the 2000–2001 amounts of liquid thousands has also been selected by the Richard P. Feynman Prize for of times smaller than a drop National Science Foundation Excellence in Teaching. The of water, and which may for a Faculty Early Career prize, made possible by an make possible the automation Development (CAREER) endowment established by of genomic and pharmaceuti- award. The award supports Ione and Robert E. Paradise, cal experiments, the perfor- his research into solid-state is awarded annually “to a mance of diagnostic tests, or reactions and phase transfor- professor who demonstrates, the building of drug-delivery mations in materials, particu- in the broadest sense, unusual devices, all on mass-produced larly ceramics, and the me- ability, creativity, and inno- chips. chanical behavior of materi- Steven Quartz, assistant als, especially composites. vation in undergraduate and graduate classroom and laboratory professor of philosophy, has Alison Winter, associate teaching.” It carries a cash award of $3,000 and is matched by been selected by the National professor of history, has been an equivalent raise in the winner’s annual salary. The selection Science Foundation for a Fac- selected to receive the committee detailed the reasons for its choice in the following ulty Early Career Develop- Northeast Victorian Studies citation: ment (CAREER) award, the Association (NVSA) Sonya “Dave Stevenson chaired the faculty committee that imple- NSF’s most prestigious award Rudikoff Award for her book, mented the revised core curriculum, and then seized the oppor- tunity to start a new menu course in Earth and Environment, for outstanding faculty early Mesmerized: Powers of Mind in which embodies the spirit and ideals of the new core. His in their independent profes- Victorian Britain. Given for sional careers. Quartz will be the best Victorian book by a success in achieving this goal can be measured in part by the funded for five years for his first-time author, the award remarkable increase in enrollment, which has risen from 20 research into the mechanisms will be presented in April at students at the start to 165 this year. Dave’s lucid and enthusi- of cognitive development, the NVSA conference at astic teaching style excites student interest, his class notes and enabling him to construct Brown University. supplemental materials provide additional clarity and depth, a computational/robotics Peter Wyllie, professor of and his bringing together of concepts from evolution, chemistry, framework for exploring geology, emeritus, has been and geology make Geology 1 unlike any other course of its kind how the mind emerges from selected by the Mineralogical in the world. The innovative structure of this course also a developing brain’s interac- Society of America as the involves small group projects with individual professors as well tion with environmental Roebling Medalist for 2001. as field trips for first-hand observation. This creates a lasting complexity. impression of how geology research is done, how our earth was The Roebling Medal is the ■ Anneila Sargent, professor society’s highest award “for created, and how our environment evolves.” of astronomy and director of scientific eminence as rep- both the Owens Valley Radio resented primarily by sci-

44 ENGINEERING & SCIENCE NO . 4     No, it’s not a docking maneuver between an Imperial star destroyer and the Death Star, but a hypothetical solar-sail mission that might one day “shadow” an asteroid in its orbit and observe it indefinitely. A sunlight- propelled spacecraft could fly to the asteroid belt and beyond in less time, stay longer, and carry a larger payload than one using chemical fuel. See the story beginning on page 18. Rendering by Tom Hames. Engineering & Science NON-PROFIT ORG. U.S. POSTAGE PAID California Institute of Technology PASADENA, CA PERMIT NO. 583 Pasadena, California 91125