Farid F. Abraham

Farid F. Abraham (born May 5, 1937) is an American scientist.

By pioneering new methods of using computer modeling in research, Abraham has made seminal contributions to science in the fields of fracture mechanics, membrane dynamics and phase transformation behavior of matter. He has authored two textbooks and over 200 papers published in international journals, including several cover articles including the Proceedings of the National Academy of Sciences and Nature. He won the Aneesur Rahman Prize in Computational Physics, which is the highest prize given by the American Physical Society.

Abraham is a native of Phoenix Arizona and received both his B.S. (1959) and Ph.D. (1962) degrees in physics from the University of Arizona. He spent two postdoctoral years (1962-63) at the Enrico Fermi Institute at the University of Chicago and two years as a research scientist at the Lawrence Livermore National Laboratory (LLNL) in California. He joined IBM in 1966 as a staff member at its Palo Alto Scientific Center. In 1971, Abraham was named the first Consulting Professor at Stanford University and developed a graduate course in computational applied science in the Materials Science Department. In 1972, he moved to the IBM Research Division's San Jose Research Laboratory, known since 1985 as the Almaden Research Center. During 1994, Abraham held the Sandoval Vallarta Chair at the Universidad Autonoma Metropolitana in Mexico City. For the period of 1995 to 2003, he was awarded several computer grants at the National Science Foundation Computational Centers and Department of Defence Grand Challenge Grants at the Maui High Performance Computing Center (MHPCC). He has been awarded several IBM Outstanding Technical Achievement Awards. Abraham is a Fellow of the American Physical Society and, in 1998/99, was an American Physical Society Centennial Speaker. Abraham was the Chair of the American Physical Society’s Division of Computational Physics in 2000-2001. He was elected the recipient of the Alexander von Humboldt Research Award for Senior Scientists. In March of 2004, he received the Aneesur Rahman Prize for Computational Physics from the American Physical Society. Retiring from IBM in 2004, he joined Lawrence Livermore National Laboratory as a Senior Scientist and was named the Graham-Perdue Visiting Professor at The University of Georgia. In 2010, he retired from LLNL. For over four decades Abraham has pursued a wide range of computational physics applications, mainly in condensed matter physics and chemical physics. Abraham's early interests in nucleation phenomena led him to pioneer the use of Monte Carlo computational methods in the study of microscopic liquid droplets and the liquid-vapor interface. Prior to his work, a molecular understanding of these inhomogeneous fluid states did not exist. His computer simulation studies in the early 1970s resolved an outstanding controversy concerning the pretransition state at the onset of vapor condensation. Also during this period, he wrote an advanced text on nucleation entitled Homogeneous Nucleation Theory.

He then turned his attention to extending the use of Monte Carlo and molecular dynamics computational methods to the simulation of the liquid-solid interface. In 1974, he discovered liquid layering at the boundary of the solid surface neighboring the liquid. This discovery led Abraham and his colleagues to give a conceptually simple but accurate understanding of this behavior using a theory of distribution functions for the liquid state. Computer simulation of clusters, surfaces and interfaces remains a major research activity in the scientific community.

While the nucleation process takes place by a local density fluctuation, it was suggested that under certain conditions sufficiently long wavelength fluctuations might be unstable and initiate the phase transition. This mechanism is called spinodal decomposition and, unlike nucleation, was questioned as a true physical mechanism for phase separation since its introduction in 1961. By extending the state-of-the-art of atomic simulations to thousands of atoms and making extensive use of visualization, Abraham in the late 1970s and early 80s used molecular dynamics computations to demonstrate the validity of the spinodal decomposition process in phase-separating fluids in both two and three dimensions. See: “On the Structure, Thermodynamics and Phase Stability of the Non-uniform Fluid State,” Physics Reports 53, 93 (1979). During the same period, he continued to extend computational physics to many other areas of research, such as solid-surface segregation, the glass transition, hydration in polar liquids, and electron-hole plasmas in semiconductors.

During the early ‘80s, a startling theoretical prediction suggested that melting in two dimensions could be a continuous transition and therefore different from the well-known first-order melting transition in three dimensions. Many early experiments appeared to confirm this prediction. Through novel applications of many different types of simulations, Abraham showed that melting in two and three dimensions are indeed the same and are first-order. He was able to reproduce by computer simulation the experimental results suggesting a continuous transition and explain them to be a consequence of the graphite surface structure and second-layer promotion that made the melting features to appear continuous. This work led to the study of many phases of monolayer films on solid surfaces and to the 1984 simulation of about 200,000 atoms, an impressive `world record' since contemporary simulations at that time used only hundreds to a few thousand atoms. See: “The Phases of Two-Dimensional Matter, Their Transitions & Solid-State Stability,” Physics Reports 80, 339 (1981).

Despite these successes, Abraham realized that the conventional computer architecture represented serious bottlenecks to simulating even larger atomic systems. In the early 1980s, he initiated an IBM project to build a special-purpose parallel supercomputer. At the same time, he asked the computational physics community what they would consider a `super problem' for that supercomputer. Although this approach has now become a popular exercise, termed `grand-challenge' problems, at the time many did not recognize this vision. See: “Computational Statistical Mechanics: Methodology, Applications and Supercomputing,” Advances In Physics, 35, 1 (1986). During this same period, Abraham also contributed to resolving certain outstanding issues in the helium-film phase diagram, biexiton formation in quantum dots and atomic force microscope images using quantum and classical simulations.

By 1990, the behavior of solid, flexible membranes spurred the interests of the theoretical physics community with the suggestion that their natural form would be crumpled up into a ball rather than extended like a sheet. Abraham's molecular dynamics simulations showed that the solid membrane is flat when there are only entropic interactions between atoms, contrary to the theoretical prediction, which was based on the successful Flory theory in polymer physics. This was quite unexpected and initially difficult for many to accept. However, high-resolution x-ray scattering measurements on the solid membrane Spectrin found in red blood cells have verified the simulation result that the natural form is flat. See: Abraham, FF & Nelson, DR, “Diffraction From Polymerized Membranes,” Science 249, 393 (1990); Abraham, FF & Kardar, M, “Folding and Unbinding Transitions in Tethered Membranes,” Science 252, 419 (1991). However, it was proposed that with sufficient attraction between the atoms the flat state would be destroyed and the crumpled state would be achieved. Abraham showed with additional simulations that the solid membrane would fold, but not crumple as anticipated. It has since been suggested that this folding process may provide a method for drug transport. Animations of the dynamics of tethered membranes are available via the World Wide Web:

                   http://www.almaden.ibm.com/st/past_projects/fractures/earlier/ . 

In 1991, Abraham began an 18-month sabbatical at the University of California at Santa Barbara to create and teach an upper-division undergraduate course in computational physics. This course was also taught at UC Davis in the summer of 1993. During that same period, Abraham's conducted research in charged-density waves and the nonlinear dynamics of chaotic oscillators on a lattice. He showed that the dynamics of this intrinsically chaotic system is very rich and interesting. The intricate interplay of coherent and random dynamics in this system suggests a possible analogy with high-Reynolds-number turbulent flow, and that the self-similarity proposed for turbulent flows applies also to this problem. This suggests universality in the dynamics. The significance of this work is that this very simple model may provide a fruitful paradigm for studying dynamical pattern formation in the real world. See: “Turbulent Dynamics of an Intrinsically Chaotic Field” Physical Review 49, 3703.

After returning to IBM in 1993, Abraham initiated an aggressive program in the computational science of materials. He focused on studies in friction, wear and materials failure. Abraham's goal was to develop tools based on computational simulation of microscopic processes important to the materials scientist and technologist. He found that the fracture tip dynamics of a brittle solid under tension undergoes instability at about one third the speed of sound. At the atomic level of the simulation, Abraham could identify the mechanisms associated with the crack instability. This work was extended to the plastic failure of ductile solids. Most recently, he and his LLNL colleagues have simulated work hardening in ductile solids using one billion atoms on the IBM/LLNL ASCI White computer. See: Abraham, Farid F., et. al., “Simulating Materials Failure by Using up to One Billion Atoms and the World’s Fastest Computer, Proceedings of the National Academy of Sciences” 99, 5777 & 5783 (2002). Multimedia versions of atomistic simulation studies of fracture are available via the World Wide Web:

               http://www.almaden.ibm.com/st/past_projects/fractures/ .

For the practical needs of the engineer trying to prevent materials failure, the simulation of "real" structures on much larger space scales must be realized. One way of achieving this is by bringing together continuum, atomistic and electronic structure descriptions of matter into a seamless union. A project called MAAD (Macro, Atomistic, Abinitio, Dynamics) was created by Abraham to accomplish a union of the macroscopic, mesoscopic and microscopic descriptions of matter. The first MAAD application is the rapid brittle fracture of a silicon slab flawed by a crack at the center of the system that is under uniaxial tension. In the "far-field" regions (MACRO region), the continuum is described by the finite-element (FE) method. Around the crack (MESO region) with large strain gradients but with no bond rupture, the molecular dynamics (MD) description for the atomic motion is used. In the region of bond failure (MICRO region), a quantum mechanical description, called tight-binding (TB) is used. MAAD was compose of researchers at IBM, NRL, Harvard and Stanford. See: “Dynamically Spanning the Length Scales from the Quantum to the Continuum,” International Journal of Modern Physics C 11(6), 1135 (2000): Abraham FF, et. al., “Spanning the Length Scales in Dynamic Simulation,” Computers In Physics 12, 538 (1998). A review of his simulation studies in materials failure has recently appeared. See: “How Fast Can Cracks Move? A research adventure in materials failure using millions of atoms and big computers,” Advances In Physics. 52, 727 (2003).

At LLNL, Abraham, with two of his colleagues at IBM, did a large-scale study of protein folding. One of the predictions of the energy landscape theory of protein folding is the possibility of barrier-less, “downhill.” The protein 1BBL was proposed to fold by such a “downhill” mechanism, though this was a matter of some dispute. We carried out extensive replica exchange molecular dynamics simulations on 1BBL in explicit solvent in order to address this controversy and provide a microscopic picture of its folding thermodynamics. Our simulations show two distinct structural transitions in the folding of 1BBL. A low-temperature transition involves a significant rearrangement of the protein’s tertiary structure without loss of secondary structure, corresponding to a shift between native and non-native compact conformations. A distinct, higher temperature transition involved the complete loss of secondary structure and dissolution of the hydrophobic core. Our simulations cannot provide evidence of “downhill” folding in 1BBL, but they clearly show evidence of a complex, non-two-state folding process. See: Pitera, JW, Swope WC, and Abraham FF, “Observation of Non-cooperative Folding Thermodynamics in Simulations of 1BBL,” Biophysical Journal 94, 4837 (2008). In a LLNL second study, Abraham et. al. investigated, by molecular dynamics simulation, the generic features associated with the dynamic compaction of metallic nano-foams at very high strain rates. A universal feature of the dynamic compaction process is revealed as composed of two distinct regions: a growing crushed region and a leading fluid precursor. The crushed region has a density lower than the solid material and gradually grows thicker in time by “snowplowing.” The trapped fluid precursor is created by ablation and/or melting of the foam filaments and the subsequent confinement of the hot atoms in a region comparable to the filament length of the foam. Abraham argues that high-energy foam crushing is not a shock phenomenon even though both share the snowplow feature. This finding has significance to the LLNL National Ignition Facility’s capsule design.

Bibliography

Abraham, Farid F., and Tiller, William A. (1972) An Introduction to Computer Simulation in Applied Science. New York: Plenum Press.

Abraham, Farid F. (1974) Homogeneous Nucleation Theory, New York: Academic Press

Abraham, Farid F. (1979) “On the Structure, Thermodynamics and Phase Stability of the Nonuniform Fluid State,” Physics Reports. 53, 93

Abraham, Farid F. (1981) “The Phases of Two-Dimensional Matter, Their Transitions & Solid-State Stability,” Physics.Reports. 80, 339

Abraham, Farid. F. (1986) “Computational Statistical Mechanics: Methodology, Applications and Supercomputing,” Advances In Physics. 35, 1

Abraham, Farid F. & Nelson, David R. (1990) “Diffraction from Polymerized Membranes,” Science. 249, 393

Abraham, Farid F. and Kardar, M. (1991) "Folding and unbinding transitions in tethered membranes." Science 252, 419

Abraham, Farid. F. et al. (2003) "How fast can cracks move? A research adventure in materials failure using millions of atoms and big computers". Advances In Physics. 52, 727

Abraham, Farid. F. et al. (2003) "How fast can cracks move? A research adventure in materials failure using millions of atoms and big computers". Advances In Physics. 52, 727

Abraham, Farid F. et al. (2002) "Simulating materials failure by using up to one billion atoms and the world’s fastest computer". Proceedings of the National Academy of Science. 99, 5777.

Abraham, Farid F. et al. (2000) "Dynamically spanning the length scales from the quantum to the continuum". International Journal of Modern Physics. C11 (6), 1135 Abraham, Farid F. et al. (1998) "Spanning the length scales in dynamic simulation". Computers In Physics. 12, 538

Abraham, Farid F. (1997) "Portrait of a Crack: Rapid Fracture Mechanics Using Parallel Molecular Dynamics.” IEEE Computational Science & Engineering. 4, 2

Pitera JW, Swope WC, and Abraham FF (2008), “Observation of Non-cooperative Folding Thermodynamics in Simulations of 1BBL,” Biophysical Journal 94, 4837

Abraham, Farid F. and Duchaineau M. A., (2011) “Compaction Dynamics of Metallic Nano-foams,” arXiv:1202.4650