C. ALLIN CORNELL 1938–2007

BY ROBIN K. MCGUIRE, ROSS B. COROTIS, AND GREGORY B. BAECHER

C. ALLIN CORNELL, who died on December 14, 2007, at the age of 69, was an early proponent of using quantitative probability methods to defi ne structural reliability and safety and, more importantly, using those concepts to make rational engineering decisions.

His work had a fundamental impact on building codes and standards of practice in the design and retrofi tting of structures to withstand earthquakes (commercial buildings, dams, bridges, and power plants), strong winds and waves (offshore oil platforms), and hurricanes (commercial and residential buildings). He was elected to the National Academy of Engineering in 1981, at the age of 42, for the “development of practical methods for application of probability to structural and earthquake engineering.”

Born in Mobridge, South Dakota, in 1938, Allin received an A.B. in architecture and an M.S. and Ph.D. in civil engineering, all from Stanford University. His dissertation, “Stochastic Process Models in Structural Engineering,” and his 1971 book (co-authored with Jack Benjamin), Probability, Statistics, and Decision for Civil Engineers (McGraw-Hill), laid the foundation for his lifelong interest in using stochastic models to represent environmental loads on structures and determine structural responses to those loads. His book is still a standard reference for students and researchers, and its title reflects Allin’s conviction that structural reliability and safety are not abstract concepts but practical applications that must be used to improve engineering decisions. In 1964, Allin became a Ford Foundation Fellow at the Massachusetts Institute of Technology (MIT), where he joined the regular faculty in 1966.

Since then, his hundreds of papers have defi ned the fi eld of structural reliability and safety. His early papers on fi rst-order, second-moment concepts established the fi eld of probability-based codifi ed structural design. As a result of his research on structural reliability and safety techniques, he was the inaugural recipient in 1987 of the CERRA Award from the International Civil Engineering Risk and Reliability Association.

The American Society of Civil Engineers presented him with the Moisseiff Award (1977), Norman Medal (1983), and Fruedenthal Medal (1988) in recognition of his research contributions to solving structural-reliability problems in civil engineering. In 1983, Allin moved back to Stanford as Research Professor, a half-time commitment that gave him time to pursue consulting.

With this arrangement, his consulting advice benefi ted from his research results, and his research directions and interests were guided by the problems faced by practicing engineers and earth scientists. Through his collaboration with industry, Allin developed a basis for the probabilistic design of drilling and exploration platforms and was a strong advocate for ensuring the structural reliability of offshore structures.

He also made signifi cant contributions to “risk-informed” decision making for nuclear power plants (the application of probabilistic descriptions of environmental loads). In addition to reliability and safety problems that could be addressed purely with engineering applications, Allin had a keen interest in the earth sciences and in the applications of reliability and safety concepts to earthquake design.

His seminal paper, “Engineering Seismic Risk Analysis,” published in 1968 in the Bulletin of the Seismological Society, is often cited as a foundational document for the fi eld of probabilistic seismic- hazard analysis. In this paper, Allin argued that optimal engineering decisions on seismic design or retrofi tting had to begin with the rupture on the fault that released crustal-strain energy rather than with the earthquake ground motion at the foundation of a structure.

This paper was the basis for the fi rst seismic-hazard map in the United States that used probability theory; the map was published by the U.S. Geological Survey in 1976. As he pursued these ideas, Allin came into contact with many earth scientists as well as earthquake engineers. From 1986 to 1987, he was president of the Seismological Society of America, which awarded him the Harry Fielding Reid Medal (its highest honor) in 2001.

He was elected a fellow of the American Geophysical Union in 2002, an honor accorded to only a handful of engineers over the years. On the engineering side, he was the Distinguished Lecturer of the Earthquake Engineering Research Institute in 1999 and recipient of the Housner Medal (its highest honor) in 2003. With his brilliant analytical mind, Allin often expressed profound impatience with anyone who used fuzzy terms or took liberties with precise mathematical defi nitions or equations.

For example, in a 2005 paper describing the advantages of mean seismic-hazard calculations, he included an addendum with a correct, precise defi nition of the term “mean frequency,” and he described how the term was often misunderstood or misused in the technical literature. Allin was always more concerned with technical accuracy than with brevity, to the consternation and disapproval of many technical editors.

Allin’s draft manuscripts often included parenthetical comments with qualifi cations or exceptions to statements in the text. A colleague once observed that his parenthetical remarks contained more technical insight than the main theses of many papers. The terms “aleatory uncertainty” and “epistemic uncertainty” are good examples of the importance Allin placed on using precise terms in a precise way.

Many of us initially objected to the use of these terms on the grounds that they were more cumbersome than the commonly used “randomness” and “uncertainty.” Allin’s rejoinder was, “Those common terms have been used so imprecisely and interchangeably in the past that they are useless. If we adopt two completely new terms with precise defi nitions, people will have to use them correctly.” Aleatory uncertainty and epistemic uncertainty are now standard terms in both earthquake engineering and the earth sciences, a tribute to Allin’s vision and persistence. Uninterested in cleverness for its own sake, he was determined that intellectual laziness not be allowed to limit the usefulness of the tools of probability by reducing precise calculations to fuzzy intuition.

Throughout his career, Allin was careful to give his colleagues proper credit for their work. For example, in 2007 he documented his collaboration in the 1960s with Luis Esteva from UNAM in Mexico, who contributed the earthquake-occurrence and ground-motion models that were integral to probabilistic seismic-hazard analysis. Allin’s contribution was to integrate those models in a probabilistic format to obtain unbiased estimates of what is now called “seismic hazard,” but he was insistent that Esteva’s contributions be properly recognized.

When he published technical papers co-authored by graduate students, Allin preferred to list the students’ names fi rst. Throughout his professional career of almost 45 years, Allin was a mentor, colleague, and friend to many engineers and earth scientists. From the highest levels of government to his fi rst-year graduate students, he advised us all with the same even, informative style. Those of us who knew him as a graduate advisor found him to be a tough but fair critic who would accept only our best efforts in developing and documenting our research.

Our careers have been better for this constant reminder that the details of our work matter and for the gift of his friendship and humor. Allin is survived by his wife, Elisabeth Paté-Cornell, Professor and Chair of Management Science and Engineering at Stanford, with similar research interests, and their two children, Phillip and Ariane Cornell; and three children from an earlier marriage, Eric Cornell, Robert Cornell, and Joan Fazzio. He is also survived by two sisters, Joan Scheel of Santa Rosa, California, and Bonnie Bassinger of Edna, Minnesota.