model behavior

ASEE Prism, Dec 2004 by Home-Douglas, Pierre

A LEADING EXPERT IN COMPUTER MODELING AND MOLECULAR SIMULATION, VANDERBILT PROFESSOR PETER CUMMINGS IS DEVELOPING ONE OF THE MOST ACCURATE MODELS OF WATER EVER CREATED.

TRYING TO PREDICT the way cancerous tumors will spread, exploring the possibility of how life on Karth may have begun at deep-sea vents, investigating ways to create new materials one molecule at a time: Not your typical idea of what chemical engineers do - especially one chemical engineer. But then Peter Cnmmings is known throughout his field as a quick study, adept at coming up with novel ways to solve diverse problems using mathematical modeling and computer simulation. He is also someone who thrives on allying himself with people in different fields whom he readily admits "know infinitely more about a subject than I do." As Douglas LeVan, chair of the chemical engineering department at Vanclcrbilt University puts it, "Peter collaborates very well." * A lot ofthat versatility has to do with his background. The John R. Hall Professor of Chemical Engineering at Vanderbilt began his career thinking he'd end up in physics. He recalls how at the end of his first year of studying science in his native Australia, the head of the mathematics department approached him and told him his future should be in mathematics. "I was equally successful in chemistry and physics," the 50-year-old father of two recalls, "but the head of the chemistry department never called me in to convince me to switch to chemistry." Cummings says the fact that the mathematics chair was an American was probably not coincidental. "He had the type of aggressive mentality that other department heads didn't have in Australia at that time. Definitely an American thing-to headhunt me out of another department."

Cummings completed his Ph.D. in applied mathematics at the University of Melbourne in 1980 and then went to the University of Guelph in Canada and SUNY Stony Brook as a post-doc in physics and chemistry respectively. When he started looking around for permanent work, colleagues encouraged him to apply for posts in chemical engineering. "Actually I had never published anything in a mathematical journal as a Ph.D. student; it was all in chemistry journals." Cummings says there was a significant shortage of faculty in chemical engineering at the time. "They were looking for new blood." One of the people who helped guide him in his new career-a man he considers a mentor-was Keith Gubbins, now professor of chemical engineering at North Carolina State University. "Hc had originally written to me when he was a graduate student looking for a post-doc. I didn't have anything for him at the time but we stayed in touch," Gubbins says.

Gubbins urged the University of Virginia to invite Cummings for an interview. They did and he got the job. Gubbins says that Cummings is someone who made the transformatior from mathematics to chemical engineering relatively seamlessly, but not everyone can. "It depends on the personality and attitude of the individual. If they are genuinely interested in finding different ways to apply their background to chemical engineering problems then they can make a huge contribution -as Peter has."

Cummings worked for over 10 years at the University of Virginia before taking a joint position as distinguished scientist at Oak Ridge National Laboratory (ORNL) and distinguished professor at the University of Tennessee. In August 2002 Vanderbilt lured him to Nashville, partly on the strength of the university's renowned medical facility and Cummings's interest in biological research and the fact that he could continue working at Oak Ridge. To juggle the two posts, Cuminings keeps an apartment in Nashville as well as a home in Oak Ridge, near Knoxville, where his wife works as a networks manager at the University of Tennessee.

At Vanderbilt he soon linked up with Vito Quaranta, professor of cancer biology, to investigate how cancerous tumors spread. As Quaranta explains, predicting cancer is a little like predicting the weather: You can't be sure how it will develop. Another similarity: "You want some numbers, just like being able to say the chance of rain tomorrow is 20 percent, you want to have some idea of the chance that a cancer is going to spread."

"The reason predictions are not as accurate as they should be," says Quaranta, "is because of the sheer mass of information and the lack of adequate computer power." Enter Peter Cummings with his mathematical modeling to understand the wealth of data. Cummings employs the technique that he uses in other areas of research: looking at a level lower, where things are less complicated - in this case, examining single cancer cells and then using computers to look at their behavior to determine a so-called "emergent collective behavior" that occurs when cells combine to form a tumor.

Plain Old Water

CUMMINGS'S ONE-LEVEL-DOWN" technique has proven particularly helpful in his attempts to understand water. He and his group have worked for the past eight years on designing the most accurate molecular model of water ever developed. Water is ubiquitous and essential to life, but it is far from simple. As Cummings points out, H2O displays lots of anomalies, becoming less dense, for example, as it freezes, unlike virtually all other liquids. Cummings hopes that by creating the world's best model of the water molecule, scientists and engineers around the world will have better predictive power to know how water will behave in different situations. One that he has investigated is high-pressure, high temperature, like the type of water found at the bottom of oceans surrounding hot vents of water gushing up from the ocean floor. It is here that scientists have discovered life that survives not on light, which drives photosynthesis, but a chemical synthesis based on hydrogen sulfide. Theories have surfaced that life on Earth may have begun in similar communities billions of years ago before the ozone layer enveloped the Earth in its protective cover.