BOSTON—Although the seriousness of the consequences may differ, a crack in your coffee cup is produced by the same forces as a crack in the turbine blade of a 747. Before scientists can develop new fracture-resistant materials to prevent the inconvenience of a shattered cup and the tragedy of an airplane crash, they need to understand the fundamental forces between atoms which can bind materials together or split them apart, says University of Michigan researcher David Srolovitz.

” Materials scientists spend a lot of time combining different types of materials, such as metals and ceramics, looking for new composites with specific combinations of strength and ductility or toughness,” said Srolovitz, U-M professor of materials science and engineering. ” The weakest link in these composites is often at the interface between two different materials, especially when impurities contaminate the interface.”

Impurities don’t always have a negative effect, Srolovitz added. As an example, he cited a nickel-aluminum compound that becomes stronger as it gets hotter, but also is brittle and shatters easily. ” Eventually, researchers discovered that boron atoms were the perfect ‘magic fairy dust’ to make interfaces in this material less brittle,” Srolovitz said. ” But they had to work through many possibilities before finding one that worked.”

Srolovitz and his colleague, John R. Smith of the General Motors Research and Development Center, use supercomputers to take some of the guesswork out of materials science. Using complex mathematical computations and hundreds of hours of supercomputer time, they calculate the binding forces between atoms at the interface of different materials and produce a visual image of the bonding which holds the materials together.

Smith and Srolovitz presented research results on the effects of impurities on interfacial adhesion at the Materials Research Society meeting held here Nov. 27-Dec. 1.

” When we join two metals, or a metal and a ceramic, a continuous band of charge accumulates at the interface creating a predominantly metallic bond between materials,” Smith said. ” When impurities like carbon or silicon atoms are present, they push the two materials at the interface farther apart, which changes the bond strength and structure. These impurities can replace some of the metal-like bonds with directional covalent bonds. Covalent bonds are strong, but inflexible, making it easy for the interface to fracture.”

” We can predict which impurities will improve adhesion and which will weaken it,” Srolovitz said. ” However, these predictions are not completely reliable yet. We can use this approach to identify a small number of promising combinations, so materials engineers don’t have to waste time and money testing hundreds of possible combinations before they find one that works.”

With substantial amounts of expensive supercomputer time required to run each calculation, the technique is too costly for widespread use in industry. Srolovitz and Smith, along with other researchers, are currently developing basic sets of algorithms called density matrix methods, which need much less computer time to perform the same calculations and could be adapted for specific manufacturing applications.

Funding was provided by the U.S. Office of Naval Research and the Air Force Office of Scientific Research. The calculations were processed at the San Diego Supercomputer Center and the National Energy Research Supercomputer Center. U-M postdoctoral fellows James Raynolds, Tao Hong and Eric Roddick assisted with the research.

Office of Naval ResearchAir Force Office of Scientific ResearchSan Diego Supercomputer CenterNational Energy Research Supercomputer Center