 | |
| Vol. 17, No. 28 | April 23, 1998 |
| |
| Vol. 17, No. 28 | April 23, 1998 |
More than 50 years ago, researchers stumbled across a curious phenomenon: Under turbulent flow conditions, certain polymers-those long-chain molecules in plastics and rubber-transform even the most sluggish goo into a fast-flowing fluid. This oddity, dubbed the Toms effect, in honor of British chemist B.A. Toms, now sends crude oil zooming through the Alaskan pipeline, where tiny amounts of polymers have reduced pumping costs by up to 50 percent.
Today, UD studies of these so-called "stretchy molecules" or flexible macromolecules, as well as next-generation surfactants that produce rod-like structures capable of rebuilding themselves, may support more powerful firefighting tools, faster submarines, new biomaterials and a host of other uses.
With his graduate students, Antony N. Beris, chemical engineering, uses supercomputers to simulate the complex transformations of flexible macromolecules dropped into turbulent, swiftly flowing fluids.
The work, spanning the past 10 years and supported by the National Science Foundation (NSF) and UD's Graduate Studies Office, is shedding new light on polymer flow behavior. That information, Beris said, should help researchers develop polymers that more efficiently reduce the "drag" or resistance of turbulent fluid flows.
How could solid polymer molecules reduce drag?
Beris and former student Radhakrishna Sureshkumar-now a faculty member at Washington University in St. Louis, Mo.-developed a computer simulation to illustrate polymer deformations in a turbulent fluid flow. The work, building on a theory originally proposed by Arthur B. Metzner, chemical engineering, emeritus, helps explain the behavior of stretchy molecules.
Beris calls it the "dumbbell" model. Certain polymers capable of speeding turbulent fluid flows may look a bit like a dumbbell, he said, with two "spheres" connected by an elastic "spring."
Known in scientific circles as the FENE-P (for finitely extensible nonlinear elastic-Peterlin), the model reveals how a turbulent flow "extends the polymer molecule, which acts like a small spring," Beris said. "The spring has a memory, so it pulls backward, against the flow, thereby changing the momentum of the flow." In this way, he said, "the flow changes the structure of the molecule, and the structure of the molecule changes the flow."
Specifically, Beris said, these dumbbell-shaped molecules seem to slow the formation of swirling eddies adjacent to the interior walls of a pipe. Wall eddies increase the energy level of a turbulent flow.
By reducing the rate of this energy supply, Beris said, polymer additives may reduce drag by up to 70 percent.
Next-generation additives, in the form of "living" polymers, may someday allow thick fluids to flow even faster-reducing drag by as much as 300 percent. Unfortunately, traditional polymer additives for water-based solutions, such as polyacrylamides, eventually lose their effectiveness.
"If you stretch them too far, for a long period of time," Beris said, "they will break, and then you must add more polymers to the flow, which increases costs."
But, researchers world-wide are scrambling to create rod-like arrays of "surfactants," or surface-active agents, designed to reconstruct themselves in the event of breakage. Common surfactant molecules, such as those found in shampoo or dishwashing liquid, are shaped like balloons, with round, water-loving heads and long, water-phobic tails. The whip-like tails of surfactants are ideal for herding dirt and oil into spherical clumps known as micelles. Hard-working surfactants may also be organized into chains just two molecules wide but up to a million molecules long.
"These surfactant chains have the capability of rebirth," Beris said. "They are called living polymers because they are not destroyed by an intense flow environment. They rebuild." The trick, however, is to motivate surfactant molecules to form living chains. Because pressure, temperature and other conditions inside a pipe dramatically affect the self-organizing behavior of surfactants, researchers like Beris's departmental chairperson, Eric Kaler, are investigating optimal environments for "breeding" living polymers.
Ultimately, these polymers may suggest new strategies for controlling different flows, Kaler said.
Already, Beris has documented the ever-changing series of events and structures in flows containing stretchy molecules.
Such massive flow problems can only be expressed mathematically, using many primary differential equations, he said. Solving these equations must be achieved numerically, using powerful supercomputers.
With Sureshkumar, and in collaboration with Robert A. Handler of the Naval Research Laboratory, Beris simulated the Toms effect on a CRAY T3D parallel computer at the NSF's Pittsburgh Supercomputing Center, and later, at the National Center for Supercomputing Applications (NCSA).
Last year, the work was recognized by the UD Graduate Studies Office, which awarded Sureshkumar the Allan P. Colburn Prize for producing the best Ph.D. thesis in the sciences and engineering. Described by Vice Provost John C. Cavanaugh, academic programs and planning, as a "superbly talented student," Sureshkumar completed a post-doctoral assignment at the Massachusetts Institute of Technology before joining the Washington University faculty.
Most recently, his work with Beris made the cover of the NCSA's online magazine. (See http://access. ncsa.uiuc.edu/CoverStories/StretchyMolecules/gasol_1.html.)
Ongoing UD research benefits from the contributions of students such as Constantinos Dimitropoulos and Jaydeep Kulkarni. "The subsequent visibility and validation of this research reflects upon the high quality of UD's graduate program in chemical engineering," Cavanaugh said, "as well as the ability of the University to provide the facilities and support for graduate students."
Beris agreed, noting that high-end computational facilities and support at UD-including, for example, the CRAY J-916/8 and the Silicon Graphics Power Challenge 8 processor systems-allowed him to take full advantage of the supercomputing centers. After using the UD workstations, he said, "I was properly prepared for large production runs at the NSF facilities."
Beginning with special processors acquired with support from former provost Leon Campbell, biological sciences, and a multiprocessor system purchased with Unidel funds obtained by Michael Klein, chemical engineering, early support for UD computing technologies helped the University become "internationally competitive in supercomputer applications," Beris said.
-Ginger Pinholster
