New directions in mechanics
Original entry by Bryan Weinstein, Fall 2012
Authors: M. Kassner, S. Nemat-Nasser, Z. Suo, et al.
Keywords: Thin film, Self-Assembly, Modeling, Mechanics, Fluidics, Fracture, Multiscale Simulation
Note to the reader/TA: this was an extremely large article (~30 pages) and I did not want to summarize the entire thing, so I summarized the parts I found interesting. Please let me know if this is an issue. Weinstein
In 2005, the US Department of Energy (DOE) sponsored a workshop to identify areas of research that depended on advances in theoretical and applied mechanics. Three broad areas of research were discussed.
Self Assembly and Fluidics
Ultimately, scientists hope to create self-assembled structures that resemble life itself. The structures will be multi-functional and will have self-preserving attributes such as "healing, self-sensing, and replication." Fluidics will be a large part of these structures. In 2005, the most advanced self-assembly techniques were capable of producing strain-induced quantum dots. Clearly, scientists are far away from creating structures that will mimic life itself. The authors of this paper argue that our understanding of self-assembly is not advanced enough to predict the future of the field.
One challenge that scientists face when studying self-assembly are huge variations in length and time scales. One must extend the effect of microscopic behaviors to the macroscopic regime; this is a difficult task. Several theories have been developed to handle this problem, such as "molecular dynamics (MD), Monte Carlo (MD), and Density Functional Theory (DFT)."
New statistical-mechanics techniques are being developed to tackle this problem as well. Most methods thus far have focused on the "forward" problem of statistical mechanics: finding the macroscopic properties of a large many-body system given the interactions between each particle. The new methods focus on the "inverse problem:" finding the interactions between each particle that lead to a specified structure. One researcher active in this area is Dr. Sorquato at Princeton (I really like his work; I almost worked with him).
Nanofluidics is an emerging area of fluidics. Scientists do not yet understand what happens to a liuid film when its thickness is approximately the same as the molecules themselves. Clearly, at this scale, continuum models of fluids do not apply. This area is still not particularly well understood; much of the results have been phenomenological. Increasingly sophisticated experimental techniques (i.e. powerful x-ray imaging techniques) will hopefully elucidate this issue.
A variety of experimental systems were discussed as well, including Evaporation-Induced Self-Assembly (EISA), Molecular assembly on solid surfaces, and the "molecular car." Many experimental systems are being explored in depth, but many more are still not well understood.
Mechanics in Biological, Bioinspired and Hybrid Material Systems
Scientists are just beginning to utilize and seek inspiration from biology to develop new technologies. In order to mimic biology, we must study the hierarchical structure of biological materials and must consequently develop new multiscale analytic techniques to study them. Scientists must be careful when seeking inspiration from these systems, however, as both robustness and efficiency are important to engineering design while biological systems have only been created for robustness (as demanded by evolution).
Many exciting systems to study fall under this category, such as nanomechanics, cytoskeletal networks, and ultrathin liquid films. Dr. Mahadevan (here at Harvard) is playing a leading role in this emerging field.
Inelasticity and Fracture
Inelastic material behavior, especially involving fracture, must be tackled from a multidisciplinary perspective that takes advantage of powerful computers today. Computers must be used to determine the bulk behavior of solids based on microscopic rules; multiscale modeling will play a large role in the solution of this problem. Continuum models will have to be improved as well; at this point they typically do not include intrinsic length and time scales or probabilistic effects. Furthermore, one must study how to link continuum mechanics to discrete points in space (an atomistic view). Clearly, much work remains to be done in this area.
This paper gave me an excellent overview of many exciting areas of research in mechanics. I am particularly interested in the applications of statistical and classical mechanics to self-assembly and biological systems. This paper helped to teach me what sorts of problems I could study while studying here at Harvard. Though quite long, it was incredibly informative and interesting. I think anyone that is interested in mechanics should read this paper.
 Kassner, M. E. et al. New directions in mechanics. Mechanics of Materials 37, 231–259 (2005).